UvA-DARE (Digital Academic Repository) Monitoring quality ... · Vivian Engelen). Symone Detmar had experience in providing patient reported outcomes (PROs) in adult cancer care,

UvA-DARE is a service provided by the library of the University of Amsterdam (http://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

Monitoring quality of life in paediatric oncology practice

Engelen, V.

Link to publication

Citation for published version (APA):Engelen, V. (2011). Monitoring quality of life in paediatric oncology practice.

General rightsIt is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s),other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulationsIf you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, statingyour reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Askthe Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam,The Netherlands. You will be contacted as soon as possible.

Download date: 24 Aug 2020

https://dare.uva.nl/personal/pure/en/publications/monitoring-quality-of-life-in-paediatric-oncology-practice(613fe2f2-f584-4456-815c-96d8f5704de7).html

Vivian Engelen

Vivian Engelen

Monitoring Q

uality of Life

in Paed

iatric O

ncology Practice

Mon

itorin

g Q

ualit

y of

Life

in P

aed

iatri

c O

ncol

ogy

Pra

ctic

e

Proefkaft3.indd 1 12-08-11 19:47

Vivian Engelen

Mon

itorin

g Q

ualit

y of

Life

in P

aed

iatri

c O

ncol

ogy

Pra

ctic

e

Lenssen aandacht x resultaat GlaxoSmithKline

nancially supported by the Dutch Cancer Society

and the Tom Voûte Fund

nancially supported by:

Dutch Cancer Society (KWF) TNO

Tom Voûte Fund (SKK) Wim Engelen Beheer BV

Academic Medical Centre (AMC) Océ

Carneiro Watson Estro Groep/ Catalpa

Genzyme Hands To Work

Biomedia Interfact

Stichting Artsen voor Kinderen Stoevelaar & Partners

Roche Nestlé Nutrition

Cover: Bregje van Gellecum, BAE Design

Lay-out: Hilde Haverman

Press: GVO drukkers & vormgevers B.V. | Ponsen & Looijen

Copyright © 2011 Vivian Engelen, Amsterdam, The Netherlands

No part of this thesis may be reproduced, stored or transmitted in any form or by any

means without prior permission of the author

Monitoring Quality of Life in Paediatric Oncology Practice

ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad van doctor

aan de Universiteit van Amsterdam

op gezag van de Rector Magnifi cus

prof. dr. D.C. van den Boom

ten overstaan van een door het college voor promoties ingestelde

commissie, in het openbaar te verdedigen in de Agnietenkapel

op vrijdag 7 oktober 2011, te 10.00 uur

door

Vivian Engelen

geboren te Tegelen

Promotiecommissie

Promotor: Prof. dr. H.N. Caron

Co-promotores: Dr. M.A. Grootenhuis

Dr. S.B. Detmar

Dr. H.M. Koopman

Overige leden: Prof. dr. J.B. van Goudoever

Prof. dr. H.S.A. Heymans

Prof. dr. R. Pieters

Prof. dr. N.K. Aaronson

Prof. dr. B.F. Last

Prof. dr. J.M. Koot

Faculteit der Geneeskunde

Contents

Chapter 1

Chapter 2

Chapter 3

Chapter 4

Chapter 5

Chapter 6

Chapter 7

Chapter 8

General Introduction

Health-related quality of life of Dutch children:

psychometric properties of the PedsQL in the Netherlands

Health-related quality of life after completion of successful

treatment for childhood cancer

Psychosocial issues addressed by paediatric oncologists

Development and implementation of a patient reported

outcome intervention (QLIC-ON PROfi le) in clinical

paediatric oncology practice

Reporting health-related quality of life scores to physicians

during routine follow-up visits of paediatric oncology

patients: is it eff ective?

The infl uence of patient reported outcomes on the

discussion of psychosocial issues in children with cancer

General Discussion

Summary

Samenvatting

Dankwoord

Curriculum Vitae

QLIC-ON PROfi le

7

23

37

53

75

97

117

133

151

157

163

169

171

71Chapter 1General Introduction

8

9

1. Origin of this thesisThe Quality of Life In Childhood Oncology (QLIC-ON) study, presented in this thesis, is the end result of a unique collaboration which offi cially started in 2005 with a grant from the Dutch cancer Society. The QLIC-ON project team included members from TNO (dr. Symone Detmar), the Leiden University Medical Centre (dr. Hendrik Koopman) and the Emma Children’s Hospital/Academic Medical Centre (dr. Martha Grootenhuis and drs. Vivian Engelen). Symone Detmar had experience in providing patient reported outcomes (PROs) in adult cancer care, Hendrik Koopman was the fi rst to develop a PRO for children with a chronic disease and Martha Grootenhuis was well known for her psychosocial research in paediatric oncology.

At the time of the start of the current study, it was known that some children with cancer experience psychosocial problems and that these problems are not always identifi ed and discussed by physicians. PROs could possibly bridge this ‘gap’ between child and physician. Therefore, the expertise within the project team was combined and the QLIC-ON study was born: introducing PROs in clinical paediatric oncology practice. A unique project in paediatrics and never done before in paediatric oncology. This general introduction will fi rst elaborate on childhood cancer, health-related quality of life and psychosocial functioning of children with cancer. Secondly, a summary of research fi ndings regarding the eff ectiveness of PROs in clinical practice is provided. Finally, the set-up of the QLIC-ON study and the outline of the current thesis are described.

2. Childhood cancer2.1. DiagnosisIn the Netherlands, approximately 500 children each year are diagnosed with cancer [1]. Cancer is defi ned as the uncontrolled and unrestrained proliferation of cells, which can occur anywhere in the body. The characteristics of childhood cancer diff er from those of cancer in adults in type, histology and anatomic location. Leukaemia, central nervous system tumours (CNS tumours) and lymphoma are the most common in children, representing approximately one-third, one-fourth and one-tenth of the incidence of childhood cancer, respectively. The incidence of several cancer diagnoses varies across age groups and gender. The incidence of some childhood cancer diagnoses (e.g. acute lymphocytic leukaemia (ALL), CNS-tumour, neuroblastoma, Wilms’ tumour) decreases with age, while the incidence of other diagnoses (e.g. Hodgkin’s disease and bone tumour) increases with age. With respect to gender, leukaemia and lymphoma are found more frequently in boys than in girls [2].

chapter 1

10

2.2. TreatmentChildren with cancer are treated according to protocols that have been developed through international cooperation and research. Surgery, chemotherapy and radiotherapy are the major modalities of cancer treatment. Standard practice usually involves a combination of these treatment modalities. Length and type of treatment depend on a number of factors, including the type of cancer, location and stage of the disease.

For children with high risk ALL or relapsed malignancies stem cell transplantation (SCT) is a treatment of last resort. The treatment involves high doses of chemotherapy and/or total body irradiation before the stem cells of a donor are infused. SCT is usually performed with bone marrow from a donor (allogeneic) but in some instances takes place with cells from the patient itself (autologous). SCT is a hazardous treatment, associated with high morbidity and mortality, because children become extremely susceptible to infections. It involves a lengthy hospital admission in an isolated, germ-free environment. In the fi rst four to six months post-SCT, children are still prone to develop infections and are forced to live with restrictions [3].

Cancer treatment and related medical procedures cause pain and distress. Although the development of more eff ective and targeted treatments has reduced the side eff ects to some extent, they are still present. The severity of the side eff ects depends on the type and location of cancer, child age and intensity of treatment. Side eff ects of treatment can occur within days or weeks of initial treatment (early eff ects) or even months to years after the end of treatment (late eff ects) [2]. 2.3. Survival and late eff ectsChances to be cured from childhood cancer have increased over the past several decades; currently, the survival rate is about 75% [4]. Survival rates depend on both diagnostic and prognostic factors. The survival chances for some types of cancer are high. For example the survival chances for Wilms’ tumour and Hodgkin’s disease are between 80% and 90%. On the other hand, the prognosis for some other cancers (e.g. neuroblastoma) is much lower. Although the majority of the childhood cancer patients can be considered cured fi ve years after diagnosis, 10% of these patients die of disease recurrence or treatment related causes within ten years [2].

Studies indicate that between 60% and 75% of all long-term childhood cancer survivors develop one or more adverse events due to the disease or treatment, and approximately one-third of these events are classifi ed as either moderate or severe. The health problems can be categorised into ten main groups: endocrine, organ toxicity, mobility/ orthopaedic, fertility, sensory, cosmetic, fatigue, subsequent neoplasm, psychosocial/ cognitive and neurological [2].

11

3. Health-related quality of life and psychosocial functioning in childhood cancer

3.1. Health-related quality of lifeThe World Health Organisation defi nes quality of life (QOL) as “individuals’ perceptions of their position in life, in the context of the culture and value systems in which they live, and in relation to their goals, expectations, standards and concerns” [5]. The concept of health related quality of life (HRQOL) refers to the impact of health and illness on an individual’s QOL [6]. In paediatrics there are usually four HRQOL domains: physical, emotional, social and school functioning, of which the last three can be summarized as psychosocial functioning [7].

HRQOL can be measured in several manners such as with an interview, a diary or a questionnaire. In general, questionnaires are the most frequently used method to assess HRQOL. There is a diversity of paediatric HRQOL instruments available e.g.: TAPQOL [8], TACQOL [9], TAAQOL [10], DUX-25 [11], Kidscreen [12], PedsQL [7] and CHQ [13]. Internationally, the PedsQL is one of the most frequently used HRQOL questionnaires in research. Dutch reference scores are not on hand however. The PedsQL has a broad age range (2-18 years), a short completion time (approximately 5-10 minutes), good feasibility, validity and reliability and it includes self-report as well as proxy-report [7]. Previous research has demonstrated that children aged 5-18 years can reliably and validly self report their HRQOL when an age-appropriate measurement instrument is utilized [14].

Over the years, HRQOL has become an important outcome measure in paediatric oncology research. Several studies [15-23] have demonstrated that childhood cancer diagnosis can lead to poor HRQOL on all domains compared to healthy peers. The studies included patients from several diagnostic groups, ages and treatment stages. For example, children reported more physical complaints, reduced motor functioning and autonomy and impaired positive emotional functioning 6 weeks after cancer diagnosis. Subsequently, HRQOL improved signifi cantly over the year following diagnosis. However, at 1 year after diagnosis, children still showed reduced motor and emotional functioning. Furthermore, at 6 weeks, children with ALL were most aff ected and at 1 year, children with brain tumours complained about more physical symptoms than the other groups [24]. Another study demonstrated that children treated for high-risk ALL, girls and older children had lower HRQOL. In children treated for standard-risk ALL, those with lower household incomes and unmarried parents experienced lower HRQOL. HRQOL scores were generally constant across phases of ALL therapy, however, children on therapy for ALL have lower HRQOL compared to healthy children [25]. In addition, research has shown that positive expectations of the course of the disease, less frequent parental asking after disease-related emotions of the child and lower levels of family adaptability are psychosocial indicators of a favourable HRQOL [26].

Overall, one can conclude that HRQOL has been the focus of quite a few studies,

chapter 1

12

however, HRQOL shortly after fi nishing treatment has only been investigated to a very limited extent. This is remarkable, since especially this transition phase is known to be very stressful.

3.2. Psychosocial functioningFor almost all children and their parents the diagnostic phase and start of treatment is very stressful. It starts with invasive medical procedures necessary to obtain a diagnosis. The child has to endure pain and parents have to witness their child in pain and fear. The next challenge than is to accomplish a realistic, and for the child an age appropriate, understanding of diagnosis and treatment implications. A second stressor at this stage is learning to deal with the eff ects of medical treatment. Often shortly after the diagnosis chemotherapy starts and the child quickly starts to experience what cancer treatment is about; pain due to medical procedures, sickness because of chemotherapeutic agents and side eff ects such as hair loss. A third stressor, especially at the beginning of treatment, is to inform the social environment. To ensure social support which keeps the family integrated is important because of the long duration of treatment. Families need ongoing support and stay in touch with everyday life as good as possible.

In the fi rst months after diagnosis, many children show adjustment towards treatment and being ill, but they also start to realize the impact of the cancer experience on their daily lives. Most studies in this fi eld indicate that children with cancer as a group are not very diff erent to control groups in terms of adaptation or adjustment and psychopathology in all stages of the disease and treatment [27]. However, there are children who are at risk for diffi culties in adapting to the disease [28], e.g. children who react with withdrawal or social isolation or children with premorbid psychopathology. These children need intensive psychological support during and after treatment.

Children with cancer often feel anger and disappointment because they are restricted to participate in important areas of life, such as going to school or engaging in sports and other social activities. Children who attend school during treatment often feel ‘diff erent’, because of their changed appearance or decline in condition [29]. As a group, however, they hardly seem diff erent from their healthy peers in terms of their classroom behavior and social acceptance [30]. Research shows that children with cancer who receive social support from their teachers, parents and friends report fewer psychosocial problems and a higher self-esteem [31].

The end of treatment in paediatric oncology can be considered a tremendous transition in care and can be experienced as very stressful. Children and parents are supported during the entire treatment by the multidisciplinary team paying attention to all diff erent aspects of functioning. When treatment is fi nished this multidisciplinary guidance quickly diminishes and families must regain their own lives. Studies in this phase show that children of diff erent ages report problems with physical functioning as a consequence of loss of general condition and strength or ongoing tiredness [23]. Next to

13

physical functioning adolescents often experience a sense of distance towards their peers which sometimes makes it diffi cult to reintegrate. Individual counseling can help children and adolescents to express feelings of loneliness or being misunderstood and help them to bridge the gap to ‘normal life’.

Studies in childhood cancer survivors performed during the past 25 years have found little evidence of serious psychological maladjustment. Research on specifi c areas of psychosocial adjustment, however, has found that between 25% and 30% of survivors and their family members have experienced personal, family or social diffi culties that have aff ected their academic achievement, employment, interpersonal relationships or self-esteem [2].

3.3. Psychosocial care and follow-upIn the Netherlands, every academic hospital provides psychosocial care to children with cancer. Psychologists, social workers and/or play therapists are available to support and guide the children and their parents through this diffi cult time of illness and treatment (if needed). However, the organization of this psychosocial care is diff erent for every hospital with respect to manpower and type of care. Some hospitals refer all newly diagnosed children to a psychologist or social worker for a general psychosocial screening; other hospitals do not have enough manpower to do so and are therefore dependent on the medical team involved around the child to detect psychosocial problems and refer to psychosocial care.

As more children with cancer survive into adulthood the importance of monitoring the late eff ects, both physical and psychosocial, has gained recognition. In the Netherlands, all survivors are off ered regular follow-up by long-term follow-up clinics in the academic hospitals, according to guidelines that have recently been designed [32].

From adult research we know that the majority of patients and physicians want to discuss psychosocial issues during cancer treatment. However, physicians do not always address these issues appropriately [33]. How exactly paediatric oncologists perceive their role towards psychosocial functioning is unknown. Neither is known to what extent these psychosocial issues are addressed in paediatric oncology practice. We believe PROs can serve as an helpful aid in identifying and discussing psychosocial problems.

4. Patient reported outcomes in clinical practice

4.1. Defi nition The use of PROs in routine clinical practice is becoming increasingly common. PROs are based on direct reporting by patients without the intervention of an observer. They include the self-assessment of functional status, symptoms, and other concerns such as patient needs and satisfaction with care. HRQOL assessment is a form of PRO that often includes both a patient’s functional status (physical and psychosocial) as well as his or

chapter 1

14

her symptoms [34]. PROs in terms of HRQOL can be included in clinical trials, although in paediatric oncology little experience is available yet.

4.2. Eff ectivenessMonitoring HRQOL by providing PROs in clinical practice has already been the focus of several studies [35-44]. PRO interventions that monitor HRQOL over time have demonstrated to improve patients’ memory and ability to describe their problems. Further, PROs can identify problems that patients might not have raised and that clinicians would therefore assume were not of concern. Completion of a PRO might also make patients feel cared for, thus improving emotional function [44]. Additionally, incorporating PROs in clinical oncology practice appeared to facilitate discussion of HRQOL issues and increase physician’s awareness of their patients HRQOL [35]. PROs also focussed the conversation during the consultation on issues relevant to patient’s experience without extending the visit [40]. Furthermore, there was an impact on physician-patient communication with better HRQOL and emotional functioning for some patients [36]. Moreover, a study in patients with chronic liver disease demonstrated that physicians using a PRO alter their treatment policy signifi cantly more than physicians not using HRQOL feedback [37]. This was also the case for a PRO study involving lung-transplant patients [41]. In addition, it appeared that the introduction of PROs into the daily routine of an out-patient clinic is feasible [42, 43].

Despite these encouraging results, it remains diffi cult to prove eff ectiveness of PROs in clinical practice [44-48]. The feedback of PROs to health professionals has, in some studies, had an impact on the process of care, with less evident impact on health outcomes [45]. Researchers often have the desire and purpose to improve health outcomes (e.g. the patient’s HRQOL) by providing PROs in clinical practice, but this is diffi cult to establish. Several points of consideration to increase the eff ectiveness of PROs in clinical practice were suggested in a previous study [46]: (1) the applied PRO needs to prioritize the views of the individual in order to adequately refl ect the individual’s HRQOL; (2) health care providers other than physicians may also fi nd HRQOL information useful; (3) feedback of HRQOL information should be longitudinally measured and presented over a period of time; (4) a clinically important diff erence or change in HRQOL data does not always resemble physicians’ perceptions of this diff erence or change; and fi nally, (5) commitment and education of the PRO user are needed to address the multiple barriers to the use of PRO in clinical practice. Additionally, another study [44] opted that the integration of PROs in clinical practice can be facilitated by developing stronger theoretical foundations for their use, obtaining buy-in early on from clinicians and patients, and addressing the system-related and methodological barriers that exist. The use of more sophisticated interventions and stronger research designs are needed to move this area of applied research forward.

In sum, profound development and implementation of a PRO intervention is

15

necessary to validly assess the eff ectiveness of a PRO in clinical practice. The current thesis has taken as many of the suggestions above into account in developing and setting up a PRO intervention study (QLIC-ON) in paediatric oncology.

5. The QLIC-ON study

5.1. Background PRO studies are usually conducted in adult patients, with little attention to paediatrics [48]. A recent exception is a diabetes study [49] which demonstrated positive eff ects of a PRO tool in adolescents with diabetes. It is remarkable that no other paediatric populations have profi ted from feeding back HRQOL via PRO tools in clinical practice, since it is known that children with a (chronic) disease are at risk for suff ering from HRQOL problems [50] and that physicians are not always aware of these problems [51].

As stated above, children with cancer can also experience HRQOL problems, which can continue after completion of treatment [23]. It is generally recognized that the fulfi lling of age-specifi c developmental tasks in childhood is of great importance to adjustment in adult life; the burden of cancer interferes with this process [52] and may cause HRQOL problems. PRO tools could possibly bridge the gap between child and paediatric oncologist and serve as an aid in identifying and discussing HRQOL problems, also to prevent developmental delay. Additionally, PRO tools could contribute to eff ective patient–physician communication, which is of crucial importance to psycho-social outcomes in patients with cancer [53, 54].

5.2. AimsThis PhD thesis presents the results of the QLIC-ON study. Aims of the QLIC-ON study included:

• to investigate the eff ectiveness of an intervention that provides a PRO tool about HRQOL (the QLIC-ON PROfi le, p.173) to the paediatric oncologist;

• to examine the eff ect of using the QLIC-ON PROfi le in clinical practice on the type (and amount) of psychosocial topics discussed during a paediatric oncology consultation.

To provide more insight in the issues around PROs in paediatric oncology additional aims of this thesis included:

• to collect Dutch reference data of an international HRQOL measure (PedsQL) and to assess psychometric properties in the Netherlands;

• to determine HRQOL of children with cancer shortly after the end of successful treatment compared with normative values;

• to explore paediatric oncologists’ perception of their role in (1) discussing psychosocial

chapter 1

16

functioning and identifying psychosocial problems and (2) providing emotional support to children with cancer;

• to provide a thorough description of the development and implementation of the QLIC-ON PROfi le in clinical paediatric oncology practice.

5.3. Study designA sequential cohort design was regarded most suitable for the purpose of the QLIC-ON study (Figure 1). Paediatric oncologists participated in the control as well as the intervention period; children and parents participated in either the control period or the intervention period, but not both. First, a control period (Figure 1A) was introduced, in which children or parents (N=99) completed the HRQOL questionnaire while waiting for their consultation in the out-patient clinic. Answers on the HRQOL questionnaire were summarized in the QLIC-ON PROfi le. In the control period the QLIC-ON PROfi le was not provided to the paediatric oncologist, the child and the parent. The paediatric oncologist started the intervention period (Figure 1C) – with a training (Figure 1B) - only as soon as he had seen all patients participating in the control period during three consecutive consultations. The intervention period consisted of a group of 94 children or parents fi lling out the HRQOL questionnaire in the waiting room. This time however, the QLIC-ON PROfi le was provided to the paediatric oncologist, the child and the parent, to be discussed during the three consultations.

5.4. Intervention: the QLIC-ON PROfi leThe QLIC-ON PROfi le is a PRO about HRQOL (p. 173), which concerns a monitoring tool (as opposed to a screening tool). The left top provides information regarding the child: name, date of birth, person that completed the questionnaire and period of time. The main part represents the four HRQOL domains, each marked with a coloured line for easy recognition: physical (blue), emotional (green), social (yellow) and school functioning (pink). The items are depicted for every domain, with the answers summarized in a column behind linked to the completion (=consultation) date. An answer accentuated green (‘never’, ‘almost never’) indicates that the child had no problems regarding the subject, orange (‘sometimes’) points out that there were some problems, and a red answer (‘often’, ‘almost always’) illustrates that a child frequently experienced problems. Most recent answers are printed bold to focus the attention of the paediatric oncologist. At the bottom, scale scores are refl ected in line graphs presented with the corresponding colours of each domain. In every line graph the scale score of the child is indicated with the completion date. Scores can be compared with previous scale scores as well as with the scale score of the healthy norm population (red dotted line).

To optimize eff ectiveness of the QLIC-ON PROfi le, we provided a training which comprised both an individual and a group session for oncologists, and a patient instruction. The training was also part of the intervention.

17

Figure 1. QLIC-ON study: sequential cohort design

Baseline Follow-up

Consultation 1 N=99 Consultation 2 N=86 Consultation 3 N=84

PRO Referral (po) PRO Referral (po) PRO Referral (po)Satisfaction (po, p, c) Satisfaction (po, p, c) Satisfaction (po, p, c)HRQOL topics (p) HRQOL topics (p) HRQOL topics (p)

Communication Consultation duration

Socio-demographics (po, p, c) HRQOL (p or c)HRQOL (p or c)

QLIC-ON PRO le introduction: In-depth use of QLIC-ON PRO le: QLIC-ON PRO le introduction*:

Lay-out Theory: back ground information Lay-outContent Practice: cases on DVD

Group discussion

Educational/supporting material: Educational/supporting material: *After completing PRO beforeDecision tree Syllabus Consultation 1QLIC-ON Pocket Card

Baseline Follow-up

Consultation 1 N=94 Consultation 2 N=93 Consultation 3 N=74

Referral (po) Referral (po)PRO Referral (po) PRO PROSatisfaction (po, p, c) Satisfaction (po, p, c)Satisfaction (po, p, c)HRQOL topics (p) HRQOL topics (p)HRQOL topics (p)

Sound recording:Sound recording:Sound recording:

QLIC-ON PRO le QLIC-ON PRO leQLIC-ON PRO le

HRQOL (p or c)Socio-demographics (po, p, c)HRQOL (p or c)

<± 1-14 days> <± 4-12 weeks> <± 4-12 weeks>

<± 1-14 days> <± 4-12 weeks> <± 4-12 weeks>

<± 1-14 days>

<± 1-14 days>

Control period

March 2006 - January 2008

Intervention period

January 2008 - November 2009

Intervention

Individual training (po) > > Individual pat ient instruction (p , c)

pre post pre post pre post

A

B

C

postpre pre post pre post

CommunicationConsultation duration


Sound recording: Sound recording: Sound recording:Communication

Consultation duration




--

-

Group training (po)

chapter 1

po: paediatric oncologst, p: parent, c: child

18

5.5. Outcome measuresThe majority of the outcome measures applied in the QLIC-ON study are known from previous PRO research: communication, referral, satisfaction and HRQOL [48]. According to an earlier study [46], evidence was found for ‘changes to doctor-patient communication’ when using HRQOL in clinical practice. This thesis assesses communication during the consultation in two (partly innovative) ways: 1) communication about HRQOL domains and 2) identifi cation of HRQOL problems. The fi rst communication outcome was examined with audio recordings of the consultations, which were scored with a self-constructed audio-assessment form. For the second communication outcome, two measures were combined: the QLIC-ON PROfi le and the HRQOL topics checklist (after each consultation the parent indicated if a topic was discussed during the consultation). Unidentifi ed HRQOL problems were retrieved by matching the answers on the QLIC-ON PROfi le with the results of the HRQOL topics checklist. With these two communication outcomes we hope to gain more insight in the eff ect of applying a PRO tool in clinical practice, since these outcomes are possibly more sensitive to changes than the more traditional ones e.g. referral, satisfaction and HRQOL on which changes are more diffi cult to fi nd [46]. Still, we have incorporated these traditional outcomes in the QLIC-ON study as well, because they are well known and important as an outcome. Moreover, as can be concluded from previous research, it is not impossible to fi nd change via some of these measures [36, 49].

6. Outline of the thesisThe general introduction of this thesis is covered in Chapter 1. In chapter 2 Dutch reference data of the PedsQL Generic Core Scales are presented. Results regarding reliability, socio-demographic within-group diff erences and construct validity are also described. Chapter 3 focuses on HRQOL of children with cancer shortly after the end of successful treatment compared with the healthy norm. Chapter 4 elaborates on the paediatric oncologists’ perception regarding their task in discussing psychosocial functioning, identifying psychosocial problems and providing emotional support. A thorough description of the development and implementation of the QLIC-ON PROfi le in clinical paediatric oncology practice is provided in chapter 5. The main fi ndings of the QLIC-ON study are depicted in chapter 6, which investigates the eff ectiveness of the QLIC-ON PROfi le in paediatric oncology practice. Finally, the eff ect of using the QLIC-ON PROfi le on the type (and amount) of psychosocial topics discussed during a paediatric oncology consultation is presented in chapter 7. This thesis ends with a general discussion including main fi ndings, key messages, clinical implications and future perspectives (chapter 8) and a summary of the results.

19

Reference List1. SKION. http://www.skion.nl/. In: 2009.2. Maurice-Stam H. Surviving childhood cancer.

Quality of life, course of life and coping. In: 2007.

3. Vrijmoet-Wiersma CM. Child and parental adaptation to pediatric oncology. In: 2009.

4. KIKA. http://www.kika.nl/. In: 2009.5. WHO. WHOQOL-BREF. Introduction,

administration, scoring and generic version of the assessment. In: 1996.

6. Eiser C, Morse R. Quality-of-life measures in chronic diseases of childhood. Health Technol Assess 2001:5:1-157.

7. Varni JW, Burwinkle TM, Seid M, Skarr D. The PedsQL 4.0 as a pediatric population health measure: feasibility, reliability, and validity. Ambul Pediatr 2003:3:329-341.

8. Fekkes M, Theunissen NC, Brugman E, et al. Development and psychometric evaluation of the TAPQOL: a health-related quality of life instrument for 1-5-year-old children. Qual Life Res 2000:9:961-972.

9. Verrips GH, Vogels TG, Verloove-Vanhorick SP, et al. Health-related quality of life measure for children-the TACQoL. J Applied Therapeutics 1997:1:357-360.

10. Bruil J, Fekkes M, Vogels T, Verrips E. TAAQOL Manual. In. Leiden: Leiden Center for Child Health and Pediatric LUMC-TNO; 2004.

11. Koopman H.M., Theunissen NCM, Vogels TG, et al. The dux-25: A Short-Form Questionnaire for Measuring Health Related Quality of Life of Children with a Chronic Illness. In: 1998. p 619.

12. Ravens-Sieberer U, Gosch A, Rajmil L, et al. The KIDSCREEN-52 Quality of Life Measure for Children and Adolescents: Psychometric Results from a Cross-Cultural Survey in 13 European Countries. Value Health 2007.

13. Landgraf J. Child Health Questionnaire (CHQ). A User’s Manual. 1999.

14. Varni JW, Limbers C, Burwinkle TM. Literature review: health-related quality of life measurement in pediatric oncology: hearing the voices of the children. J Pediatr Psychol 2007:32:1151-1163.

15. Bhat SR, Goodwin TL, Burwinkle TM, et al. Profi le of daily life in children with brain tumors: an assessment of health-related quality of life. J Clin Oncol 2005:23:5493-5500.

16. Eiser C, Vance YH, Horne B, et al. The value of the PedsQLTM in assessing quality of life in survivors of childhood cancer. Child Care Health Dev 2003:29:95-102.

17. Hinds PS, Billups CA, Cao X, et al. Health-related quality of life in adolescents at the time of diagnosis with osteosarcoma or acute myeloid leukemia. Eur J Oncol Nurs 2009:13:156-163.

18. Matziou V, Perdikaris P, Feloni D, et al. Cancer in childhood: children’s and parents’ aspects for quality of life. Eur J Oncol Nurs 2008:12:209-216.

19. Felder-Puig R, Frey E, Proksch K, et al. Validation of the German version of the Pediatric Quality of Life Inventory (PedsQL) in childhood cancer patients off treatment and children with epilepsy. Qual Life Res 2004:13:223-234.

20. Sung L, Klaassen RJ, Dix D, et al. Identifi cation of paediatric cancer patients with poor quality of life. Br J Cancer 2009:100:82-88.

21. Varni JW, Burwinkle TM, Katz ER, et al. The PedsQL in pediatric cancer: reliability and validity of the Pediatric Quality of Life Inventory Generic Core Scales, Multidimensional Fatigue Scale, and Cancer Module. Cancer 2002:94:2090-2106.

22. Eiser C, Eiser JR, Stride CB. Quality of life in children newly diagnosed with cancer and their mothers. Health Qual Life Outcomes 2005:3:29.

23. Stam H, Grootenhuis MA, Brons PP, et al. Health-related quality of life in children and emotional reactions of parents following completion of cancer treatment. Pediatr Blood Cancer 2006:47:312-319.

24. Landolt MA, Vollrath M, Niggli FK, et al. Health-related quality of life in children with newly diagnosed cancer: a one year follow-up study. Health Qual Life Outcomes 2006:4:63.

25. Sung L, Yanofsky R, Klaassen RJ, et al. Quality of life during active treatment for pediatric acute lymphoblastic leukemia. Int J Cancer 2011:128:1213-1220.

chapter 1

20

26. Maurice-Stam H, Grootenhuis MA, Brons PP, et al. Psychosocial indicators of health-related quality of life in children with cancer 2 months after end of successful treatment. J Pediatr Hematol Oncol 2007:29:540-550.

27. Sawyer M, Antoniou G, Toogood I, et al. Childhood cancer: a 4-year prospective study of the psychological adjustment of children and parents. J Pediatr Hematol Oncol 2000:22:214-220.

28. Patenaude AF, Kupst MJ. Psychosocial functioning in pediatric cancer. J Pediatr Psychol 2005:30:9-27.

29. Gartstein MA, Short AD, Vannatta K, Noll RB. Psychosocial adjustment of children with chronic illness: an evaluation of three models. J Dev Behav Pediatr 1999:20:157-163.

30. Brown RT. Comprehensive handbook of childhood cancer and sickle cell disease. A biopsychosocial approach. Oxford: University Press; 2006.

31. Varni JW, Katz ER, Colegrove R, Jr., Dolgin M. Perceived social support and adjustment of children with newly diagnosed cancer. J Dev Behav Pediatr 1994:15:20-26.

32. Stichting Kinderoncologie Nederland. www.skion.nl/bestanden/richtlijn_follow-up_na_kinderkanker. In: 2011.

33. Taylor S, Harley C, Campbell LJ, et al. Discussion of emotional and social impact of cancer during outpatient oncology consultations. Psychooncology 2011:20:242-251.

34. Osoba D. Translating the science of patient-reported outcomes assessment into clinical practice. J Natl Cancer Inst Monogr 2007:5-11.

35. Detmar SB, Muller MJ, Schornagel JH, et al. Health-related quality-of-life assessments and patient-physician communication: a randomized controlled trial. JAMA 2002:288:3027-3034.

36. Velikova G, Booth L, Smith AB, et al. Measuring quality of life in routine oncology practice improves communication and patient well-being: a randomized controlled trial. J Clin Oncol 2004:22:714-724.

37. Gutteling JJ, Darlington AS, Janssen HL, et al. Eff ectiveness of health-related quality-of-life measurement in clinical practice: a prospective, randomized controlled trial in patients with chronic liver disease and their physicians. Qual Life Res 2008:17:195-205.

38. Rubenstein LV, McCoy JM, Cope DW, et al. Improving patient quality of life with feedback to physicians about functional status. J Gen Intern Med 1995:10:607-614.

39. Taenzer P, Bultz BD, Carlson LE, et al. Impact of computerized quality of life screening on physician behaviour and patient satisfaction in lung cancer outpatients. Psychooncology 2000:9:203-213.

40. Berry DL, Blumenstein BA, Halpenny B, et al. Enhancing patient-provider communication with the electronic self-report assessment for cancer: a randomized trial. J Clin Oncol 2011:29:1029-1035.

41. Santana MJ, Feeny D, Johnson JA, et al. Assessing the use of health-related quality of life measures in the routine clinical care of lung-transplant patients. Qual Life Res 2010:19:371-379.

42. Detmar SB, Aaronson NK. Quality of life assessment in daily clinical oncology practice: a feasibility study. Eur J Cancer 1998:34:1181-1186.

43. Wu WW, Johnson R, Schepp KG, Berry DL. Electronic Self-report Symptom and Quality of Life for Adolescent Patients With Cancer: A Feasibility Study. Cancer Nurs 2011.

44. Snyder CF, Aaronson NK. Use of patient-reported outcomes in clinical practice. Lancet 2009:374:369-370.

45. Valderas JM, Alonso J, Guyatt GH. Measuring patient-reported outcomes: moving from clinical trials into clinical practice. Med J Aust 2008:189:93-94.

46. Greenhalgh J, Long AF, Flynn R. The use of patient reported outcome measures in routine clinical practice: lack of impact or lack of theory? Soc Sci Med 2005:60:833-843.

21

47. Greenhalgh J, Meadows K. The eff ectiveness of the use of patient-based measures of health in routine practice in improving the process and outcomes of patient care: a literature review. J Eval Clin Pract 1999:5:401-416.

48. Valderas JM, Kotzeva A, Espallargues M, et al. The impact of measuring patient-reported outcomes in clinical practice: a systematic review of the literature. Qual Life Res 2008:17:179-193.

49. Wit de M, Delemarre-van de Waal HA, Bokma JA, et al. Monitoring and discussing health-related quality of life in adolescents with type 1 diabetes improve psychosocial well-being: a randomized controlled trial. Diabetes Care 2008:31:1521-1526.

50. Grootenhuis MA, Koopman HM, Verrips EG, et al. Health-related quality of life problems of children aged 8-11 years with a chronic disease. Dev Neurorehabil 2007:10:27-33.

51. Astin JA, Soeken K, Sierpina VS, Clarridge BR. Barriers to the integration of psychosocial factors in medicine: results of a national survey of physicians. J Am Board Fam Med 2006:19:557-565.

52. Maurice-Stam H, Grootenhuis MA, Caron HN, Last BF. Course of life of survivors of childhood cancer is related to quality of life in young adulthood. J Psychosoc Oncol 2007:25:43-58.

53. Davidson R, Mills ME. Cancer patients’ satisfaction with communication, information and quality of care in a UK region. Eur J Cancer Care (Engl ) 2005:14:83-90.

54. Street RL, Jr., Makoul G, Arora NK, Epstein RM. How does communication heal? Pathways linking clinician-patient communication to health outcomes. Patient Educ Couns 2009:74:295-301.

chapter 1

232Chapter 2

V. Engelen 1M.M. Haentjens 2S.B. Detmar 3

H.M. Koopman 4

M.A. Grootenhuis 1

BMC PEDIATR 2009,9:68

1 Psychosocial Department, Academic Medical Centre/ Emma Children’s Hospital, Amsterdam,

the Netherlands 2 Department of Education, University of Amsterdam, Amsterdam, the Netherlands3 Prevention and Health Department, TNO, Leiden, the Netherlands4 Medical Psychology Department, Leiden University Medical Center, Leiden, the Netherlands

Health-related quality of life of Dutch children: psychometric properties of the PedsQL in the Netherlands

24

Abstract

Background. Knowledge about psychometric properties of the Pediatric Quality of Life Inventory (PedsQL) in the Netherlands is limited and Dutch reference data are lacking. Aim of the current study is to collect Dutch reference data of the PedsQL and subsequently assess reliability, socio-demographic within-group diff erences and construct validity.

Methods. In this study the PedsQL was administered to Dutch children aged 5 to 18 years. A socio-demographic questionnaire was completed as well. The sample consisted of three age groups: 5-7 years (parent proxy report), 8-12 years and 13-18 years (child self report). Analysis was performed with SPSS 16.0.2. A reliability analysis was done using Cronbach’s alpha coeffi cient. Socio-demographic within-group diff erences were assessed by means of an ANOVA with post hoc Bonferroni correction and t-tests. Subsequently, construct validity was determined by t-tests and eff ect sizes.

Results. For 496 children PedsQL reference data were collected. PedsQL total scores were 84.18 (group 5-7), 82.11 (group 8-12) and 82.24 (group 13-18). Internal consistency coeffi cients ranged from .53 to .85. Socio-demographic within-group diff erences demonstrated that, in group 8-12, children of parents born in the Netherlands had signifi cantly lower scores on several PedsQL subscales, compared to children of parents born in another country. With respect to construct validity, healthy children of group 5-7 and 13-18 scored signifi cantly higher than children with a chronic health condition on all subscales, except for emotional functioning. In group 5-7, the PedsQL total score for healthy children was 85.31, whereas the same age group with a chronic health condition scored 78.80. Eff ect sizes in this group varied from 0.58 to 0.88. With respect to group 13-18, healthy children obtained a PedsQL total score of 83.14 and children suff ering from a chronic health condition 77.09. Eff ect sizes in this group varied from 0.45 to 0.67. No signifi cant diff erences were found in group 8-12 regarding health.

Conclusion. The Dutch version of the PedsQL has adequate psychometric properties and can be used as a health related quality of life instrument in paediatric research in the Netherlands.

25

BackgroundThe World Health Organization defi nes quality of life (QOL) as “individuals’ perceptions of their position in life, in the context of the culture and value systems in which they live, and in relation to their goals, expectations, standards and concerns” [1]. The concept of health related quality of life (HRQOL) refers to the impact of health and illness on an individual’s QOL [2]. It is generally accepted that HRQOL is a multidimensional construct incorporating at least three broad domains: physical, psychological and social functioning [3].

To assess HRQOL in children, diff erent questionnaires are needed for diff erent age groups. Comparison of HRQOL outcomes in children from diff erent age groups or over time is complicated. It is important to consider the use of proxy ratings as substitutes for ratings made by children themselves. A case in point would be when a child is too young or too ill to complete the questionnaires himself [2].

In recent years, several HRQOL questionnaires for children were developed. Examples of questionnaires which can be administered to both healthy and chronically ill children are the TAPQOL [4], TACQOL [5], Kidscreen [6] and CHQ [7]. These are referred to as generic questionnaires. For children with a chronic health condition, disease specifi c questionnaires were designed as well, e.g. the Disabkids [8].

The Pediatric Quality of Life Inventory 4.0 Generic Core Scale (PedsQL) measures HRQOL in both healthy and chronically ill children. Besides a 23-item generic module, the PedsQL has several disease specifi c modules, for instance for asthma [9], cancer [10] and cerebral palsy [11]. The PedsQL encompasses four domains of HRQOL: physical, emotional, social and school functioning and covers a broad age range, including child self-reports as well as parent proxy-reports. Both comprise age versions 5-7, 8-12 and 13-18 years; parent proxy-report also incorporates ages 2-4 years. Feasibility, reliability and validity of the PedsQL are found to be good [12]. Although translated and studied before [13], knowledge about psychometric properties of the PedsQL in the Netherlands is limited and Dutch reference data are lacking.

Aim of the current study is to collect Dutch reference data of the PedsQL and subsequently assess reliability, socio-demographic within-group diff erences and construct validity of this instrument in the Netherlands.

Methods

ProcedureThe PedsQL Generic Core Scale was administered to Dutch children aged 5 to 18 years between October 2006 and January 2007. Stratifi cation criteria for school were geographic location (urban, suburban and rural), percentage of migrant children (low, middle and high) and level of education (low, middle and high).

chapter 2

26

Parents received a letter concerning the purpose and procedure of the study. Confi dential processing of personal information and collected data was explained and guaranteed. Furthermore, a socio-demographic questionnaire and a form for non-participants were enclosed. Parents were requested to complete the socio-demographic questionnaire and send it back to school. If parents did not want to participate they were asked to fi ll out the non-participants form.

Three age versions of the PedsQL were administered in this study: parent proxy-report ages 5-7 years (group 5-7), child self-report ages 8-12 years (group 8-12) and ages 13-18 years (group 13-18). For the purpose of this study, the PedsQL was converted into two computer programs: one stand alone computer version (group 5-7 and 8-12) and an internet version (group 13-18). Layout of both programs was identical and resembled the paper version as much as possible, except that questions were presented one at a time instead of all at once and missing values were not accepted.

Parents and children from group 5-7 and group 8-12 were invited to school to complete the PedsQL on a laptop. Group 13-18 fi lled out the PedsQL via internet in a computer room at school. It concerned a simultaneous procedure, in which several children participated at the same time. Each child logged in on a website with a unique personal number. In all groups, two psychology students clarifi ed the procedure and made sure children did not interfere with classmates. During all administrations PedsQL Manual instructions were observed.

The study was approved by the Medical Ethics Committee of the Academic Medical Centre, the Netherlands.

Measures

PedsQL

The PedsQL accentuates perceptions of children and refl ects their universal concerns on four subscales: physical (8 items), emotional (5 items), social (5 items) and school functioning (5 items). Each of the 23 items states a problem, for example “diffi culty walking”. The parent or child indicates on a 5-point Likert scale to what extent the child had diffi culties with that problem. Answering options are never (0), almost never (1), sometimes (2), often (3) and almost always (4). Each answer is reversed scored and rescaled to 0-100 scale (0=100, 1=75, 2=50, 3=25 and 4=0). A score of 100 represents the best quality of life possible, a score of 0 the worst. Completion time is about 5-10 minutes. Parent proxy-reports are a parallel version of the child self-reports. Diff erences are the use of age appropriate language and fi rst- or third-person tense. The PedsQL refers to two diff erent periods of time: the past month or the past week [12]. The past week version was used in this study.

Socio-demographic questionnaire

The socio-demographic questionnaire consists of fi ve items concerning the parent (date

27

of birth, gender, country of birth, education and employment) and six items with respect to the child (date of birth, gender, presence of a chronic disease, contact with general practitioner and/ or medical specialist over the past year, received psychological care over the past year).

Non-participants form

The non-participants form is a brief list on which parents and children can indicate they are not interested in participating in this study. Parents are additionally requested to complete the fi ve previously mentioned socio-demographic questionnaire items.

Statistical analysis Data were analyzed with SPSS 16.0.2. Diff erences between participants and non-participants with respect to gender, country of birth and employment were analysed by means of Chi2 test. Age diff erences were examined with a t-test; a Mann-Whitney test was performed for education. Diff erence was made between low education (no education, primary school and primary vocational education) middle education (secondary school and secondary vocational education) and high education (higher vocational education and university). PedsQL scores were computed according to the PedsQL Manual and a reliability analysis was done using Cronbach’s alpha coeffi cient based on average inter item correlation (according to SPSS 16.0.2). Socio-demographic within-group diff erences regarding PedsQL scores were assessed by means of an ANOVA with post hoc Bonferroni correction (p<.0167) for age group and education and t-tests for gender, country of birth and employment. Subsequently, construct validity was determined by t-tests and eff ect sizes (es), analysing diff erences in PedsQL scores (per age group) between healthy and chronically ill children and between children that did or did not visit their general practitioner, medical specialist or psychologist over the past year. Eff ect sizes were calculated dividing the diff erence in mean scores between the chronically ill and healthy group by the standard deviation of the healthy group. For the group of children that visited their general practitioner, medical specialist or psychologist eff ect sizes were determined thru the same formula, with the chronically ill group being replaced by respectively the group of children that visited the general practitioner, the medical specialist or psychologist. Eff ect sizes up to 0.2 were considered to be small, eff ect sizes of about 0.5 moderate and eff ect sizes of about 0.8 large [14].

Results

Participants In total, 891 children from Amsterdam and surrounding regions were approached. They attended four elementary schools (three suburban, one urban), four high schools (one rural, one suburban, two urban) and one school for vocational education (urban).

chapter 2

28

544 (61.1%) parents gave informed consent, 53 (5.9%) declined participation and 294 (33.0%) parents did not respond. Eventually, the PedsQL was completed by 496 (55.7%) participants.

Parent gender, ethnicity and education were not evenly distributed among participants and non-participants. There were signifi cantly less fathers in the participating group than in the non-participating group (p<.01). In addition, more participating parents were born in the Netherlands (p<.01) and a larger percentage of the participating parents was highly educated (p<.05), compared to non-participants. Socio-demographic data are summarized in Table 1 and 2.

PedsQL scores and reliabilityPedsQL scale scores ranged from 74.59 (group 13-18, school functioning subscale) to 89.38 (group 13-18, social functioning subscale). Total PedsQL score for group 5-7 was 84.18, for group 8-12 82.11 and group 13-18 scored 82.24 (Table 3).

Internal consistency coeffi cients ranged from .53 to .85 (Table 4). Most subscales had an alpha higher than .60. Overall, reliability was comparable across age groups, with the total score being the most reliable and the school functioning subscale the least.

Socio-demographic within-group diff erences

Age

Results indicate a signifi cant diff erence between age groups for the subscale school functioning (p<.01). Post hoc tests demonstrated that group 5-7 scored higher than group 8-12 (mean diff erence=6.79, p<.001) and group 13-18 (mean diff erence=10.89, p<.001). In addition, group 8-12 had higher scores than group 13-18 (mean diff erence=4.10, p<.01). Gender

There was a signifi cant gender eff ect on the emotional subscale in age group 8-12; here, boys scored higher than girls (mean diff erence=4.24, p<.05).

Ethnicity In group 8-12, children of parents born in the Netherlands had a signifi cantly lower total (mean diff erence=3.94, p<.05), psychosocial (mean diff erence=4.63, p<.01), emotional (mean diff erence=5.39, p<.05) and school functioning score (mean diff erence=4.31, p<.05) compared to children of parents born in another country.

Education

Diff erences were found with respect to education in group 13-18. Post hoc tests revealed that children of low educated parents scored higher than children of parents with a middle education on the total (mean diff erence=6.21, p<.0167), psychosocial (mean diff erence=7.69, p<.01), social (mean diff erence=8.46, p<.01) and school subscale (mean

29

chapter 2

Table 1. Socio-demographic data of participants and non-participants

1 Children and parents that refused participation and completed the non-participants form; or children and parents that consented to participation and fi lled out the socio-demographic questionnaire, but eventually did not complete the PedsQL due to circumstances (e.g. illness, dental visit). 2 Socio-demographic data of parent that completed the questionnaire.

* Sign: p<.05, ** Sign: p<.01.

CHILD N Mean SD N Mean SD p value

PARENT2 N Mean SD N Mean SD p value

Participants Non-participants1

Age 496 11.80 3.20 81 11.44 2.75 .33 N % N %

Gender (boys) 227 45.8 35 43.8 .74

Age 479 44.35 5.73 82 43.16 6.29 .09

N % N % Gender (male) 92 19.4 31 36.5 <.01**Ethnicity (Dutch) 414 85.9 53 62.4 <.01**Education (high) 230 48.4 30 38.0 <.05*Employment 349 72.9 52 62.7 .06

Table 2. Socio-demographic data of participants per age group

1 Most common conditions in total Dutch sample were: asthma (36.4%), congenital defect (13.6%), skin disease (6.1%)

and migraine (6.1%). 2 Socio-demographic data of parent that completed the questionnaire.

Age group Age group Age group 5-7 8-12 13-18 CHILD M SD M SD M SD

Age 6.67 0.73 11.27 1.35 14.98 1.20

N % N % N %

Total 92 18.5 219 44.2 185 37.3Gender (boys) 43 46.7 104 47.5 80 43.2Chronic health condition1 14 15.6 26 11.9 25 14.5Contact general practitioner 54 60.7 91 41.9 88 50.9Contact medical specialist 20 22.2 55 25.2 44 25.6Psychological care 1 1.1 23 10.7 14 8.2

PARENT2 M SD M SD M SD

Age 40.58 4.77 43.67 5.66 47.23 4.78

N % N % N %

Gender (male) 15 16.7 37 17.4 40 23.3Ethnicity (Dutch) 87 95.6 177 81.2 150 86.7Education (high) 57 64.0 97 44.7 76 45.0Employment 61 67.8 158 72.5 130 76.0

30

diff erence=8.99, p<.0167) of the PedsQL. Within the social domain scores of children from parents with a high education were signifi cantly higher compared to children of parents that had a middle education (mean diff erence=5.43, p<.0167).

Employment For none of the age groups and subscales diff erences were found regarding employment.

Table 3. PedsQL scale scores

PedsQL Scale Scores

Age group N PedsQL subscale M SD

5-7 92 Total score 84.18 8.96 Physical health 87.33 10.25 Psychosocial health 82.50 9.89 Emotional functioning 75.82 13.33 Social functioning 86.20 12.21 School functioning 85.49 11.60



Table 4. PedsQL scales: internal consistency reliability

PedsQL subscale Age group 5-7 Age group 8-12 Age group 13-18

α α α

Total score .85 .82 .85Physical health .69 .61 .69Psychosocial health .82 .79 .80Emotional functioning .70 .61 .74Social functioning .73 .67 .76School functioning .59 .53 .62

31

Construct validityTable 5 contains PedsQL scores of children with and without a chronic health condition. Healthy children in groups 5-7 and 13-18 scored signifi cantly higher on all subscales, except for emotional functioning, than peers with a chronic health condition. Eff ect sizes varied from moderate to high. No signifi cant diff erences were found in group 8-12.

Children also scored signifi cantly lower on several PedsQL subscales when they had contact with their general practitioner over the past year. In group 5-7, this was the case for social functioning (p<.05, es=0.49) and in group 8-12 for the total (p<.05, es=0.31), physical (p<.05, es=0.32) and school functioning subscale (p<.05, es=0.37). Children belonging to group 13-18 obtained lower scores on all subscales except for school functioning: total (p<.01, es=0.51), physical (p<.01, es=0.70), psychosocial (p<.05, es=0.37), emotional (p<.05, es=0.34) and social functioning (p<.05, es=0.36).

In some cases, children also had lower PedsQL scores when they had contact with a medical specialist over the past year. For groups 5-7 and 8-12 this diff erence was observed

chapter 2

Table 5. PedsQL scales: construct validity

Healthy sample Chronic health condition

Subscale N Mean SD N Mean SD Effect size p value Age group 5-7 76 14 Total score 85.31 8.59 78.80 9.58 0.76 <.05*Physical health 88.65 9.13 80.58 13.71 0.88 <.01**Psychosocial health 83.53 9.70 77.86 10.22 0.58 <.05*Emotional functioning 76.38 13.80 73.93 11.30 0.18 .53Social functioning 87.76 11.70 80.00 12.56 0.66 <.05*School functioning 86.45 10.76 79.64 15.00 0.63 <.05*

Age group 8-12 192 26 Total score 82.31 8.83 80.64 9.32 0.19 .37Physical health 85.25 8.85 82.21 12.14 0.34 .12Psychosocial health 80.75 10.34 79.81 10.43 0.09 .67Emotional functioning 76.85 13.76 78.85 13.21 -0.15 .49Social functioning 86.51 12.24 83.27 12.80 0.26 .21School functioning 78.88 11.90 77.31 13.13 0.13 .53

Age group 13-18 148 25 Total score 83.14 8.99 77.09 9.40 0.67 <.01**Physical health 86.76 9.21 81.00 12.00 0.63 <.01**Psychosocial health 81.21 10.22 75.00 9.56 0.61 <.01**Emotional functioning 77.53 15.01 71.40 16.62 0.41 .07Social functioning 90.14 11.37 83.40 12.97 0.59 <.01**School functioning 75.95 12.68 70.20 15.17 0.45 <.05*

* Signifi cant p<.05, ** Signifi cant p<.01.

32

on the school functioning subscale (both p<.05, respectively es=0.65 and es=0.31). In group 13-18 the diff erence was apparent on more subscales: total (p<.05, es=0.46), physical (p<.05, es=0.42), psychosocial (p<.05, es=0.42) and school (p<.05, es=0.39).

Finally, in some cases psychological care was an indicator for lower PedsQL scores as well. In group 8-12, a diff erence was measured for the total (p<.05, es=0.51), physical (p<.05, es=0.46) and psychosocial subscale (p<.05, es=0.46). In the oldest group 13-18, these children obtained lower scores with respect to total (p<.05, es=0.67), psychosocial (p<.05, es=0.68) and social functioning (p<.01, es=0.80).

DiscussionIn this study, reference data of the Dutch PedsQL Generic Core Scale were collected and psychometric properties were assessed in a sample of 496 children aged 5 to 18 years and their parents. The psychometric properties of the PedsQL in the Netherlands range from suffi cient to good.

The PedsQL reference scores obtained in this study are an adequate representation of the general Dutch population. According to Statistics Netherlands [15] in 2008 80.4% (of the total population) were from Dutch origin and 77.9% (of population aged 45-55 years) were employed. These fi gures are comparable to the socio-demographic data from our study. Furthermore, it is not uncommon [16] to fi nd a large percentage of highly educated parents (48.4%) in the current study, compared to the Dutch working population (33.5%). It is likely that highly educated parents are better aware of the necessity of this type of research, and thus be more willing to participate. During data collection we experienced that willingness to take part in the study seemed lower on schools with lower educational levels and higher percentages of migrant children. Possibly, parents not born in the Netherlands experienced language problems and for this reason did not fully understand the information letter to participate in the study. These fi ndings are supported by the fact that there were fewer parents with a high education and Dutch ethnicity in the non-participants group, compared to the group that did participate.

The Dutch sample shows the same trend in reliability across subscales as the US [12], however with slightly lower alphas. In both samples, the total score appears to be most reliable and subscale school functioning least. A possible clarifi cation for the lower Dutch reliability fi gures could be the diff erence in sample size. The US sample ranges from 1159 to 2674 children per age group; the Dutch from 92 to 219. More over, the US sample included children from the State Children’s Health Insurance Program (SCHIP) that provides health insurance coverage to uninsured children from low-income families [17]. In the current study a stratifi ed sampling technique was applied in order to create a sample as diverse as possible, which can be generalized to the overall Dutch population. Bastiaansen et al [13] have investigated the psychometric properties of the PedsQL in the Netherlands as well. In their study a sample (n=74) from the general Dutch population (6-

33

18 years) was compared with 310 children referred for psychiatric problems: alpha’s were in concordance with Varni et al [12], however, their sample size was very small.

Regarding the socio-demographic within-group diff erences, this study demonstrates a gender diff erence in group 8-12 on the emotional subscale with girls obtaining lower scores than boys. Similar results in adolescent girls have been reported by Reinfj ell et al [18].

Salient fi nding is that country of birth of the parent had an eff ect on PedsQL scores in group 8-12. Children of parents born in the Netherlands scored signifi cantly lower on all subscales except physical functioning. Previous studies only correspond partly with this result. For instance in the US, White together with Asian children scored higher than Hispanic and Black children [12]. Important to note is that we only collected data regarding country of birth of the parent, and not country of birth of the child. Moreover, this diff erence was not found in groups 5-7 and 13-18.

Results of our study indicate a relationship between educational level and PedsQL scores in group 13-18: children of parents with a low education perceived a signifi cantly better HRQOL. This phenomenon is diffi cult to explain, since previous research mainly pointed out that high quality of life scores were related to high parental education, or that education had no eff ect at all [19,20].

Findings on parental employment are also notable: having a job had no infl uence on the child’s HRQOL in our sample. Previous research has shown that children with low socioeconomic status (SES) functioned worse than children from middle SES backgrounds [21]. Employment and SES are not exchangeable, though. Having a job does not necessarily implicate a middle or high SES.

With respect to construct validity, this study demonstrates that the Dutch PedsQL version diff erentiates between children with and without a chronic condition in group 5-7 and 13-18. Diff erences in group 8-12 were not signifi cant - however, except for the emotional functioning subscale, healthy children obtained higher scores than their chronically ill peers. Although several studies have shown that the PedsQL diff erentiates between children with and without a chronic health condition [9,12,13] it is not exceptional to fi nd adequate functioning for chronically ill children [22]. Another explanation could be that severely ill children did not take part in our study, because they were not present at the time of administration due to illness or that parents did not want to burden them with participation. A further possible reason might be the fact that the presence of a chronic health condition in our sample was determined by the parent and not diagnosed by a physician. Physician-diagnosed chronic health conditions are often stricter than those reported by the parent. Therefore, the 8-12 year old chronic health condition sample in our study might not be completely representative of children with more serious chronic diseases.

Limitations of the study need to be taken into account. First, a considerable number of children were approached, which eventually resulted in nearly 500 participants.

chapter 2

34

However, considering the diff erent age groups and the socio-demographic within-group diff erences, sample sizes per group were relatively small. Furthermore, it is possible that presenting PedsQL items one at a time on a computer - with missing values not being allowed - could have some psychometric implications. Digital administration of the PedsQL has demonstrated equivalent measurement properties to the paper version [23], yet in the study of Varni et al [23] each PedsQL scale was depicted on a separate screen and participants had the option to skip items. The fact that this possibility was lacking in the current study could have forced participants to choose an answer that did not really apply to them. Nonetheless, this probably concerns a minimum of items since Varni et al [23] also demonstrated similar (low) percentages of missing-item responses of digital and paper version PedsQL. Additionally, it would have been interesting to have examined all age versions of the PedsQL involving all regions of the Netherlands, but this was unfeasible for the purpose of our study. Therefore, we recommend that future research with respect to the PedsQL in the Netherlands should include more regions of the country and incorporate the remaining PedsQL versions.

ConclusionWe conclude that the Dutch version of the PedsQL has adequate psychometric properties and can be used as an HRQOL instrument in paediatric research in the Netherlands.

35

Reference List1. WHO. WHOQOL-BREF. Introduction,

administration, scoring and generic version of the assessment. http://www.who.int/mental_health/media/en/76.pdf . 1996.

2. Eiser C, Morse R: Quality-of-life measures in chronic diseases of childhood. Health Technol Assess 2001, 5: 1-157.

3. Sprangers M, Aaronson N: Quality-of-Life Assessment in Oncology. Acta Oncol 2002, 41: 229-237.

4. Fekkes M, Theunissen NC, Brugman E, Veen S, Verrips EG, Koopman HM, Vogels T, Wit JM, Verloove-Vanhorick SP: Development and psychometric evaluation of the TAPQOL: a health-related quality of life instrument for 1-5-year-old children. Qual Life Res 2000, 9: 961-972.

5. Verrips GH, Vogels TG, Verloove-Vanhorick SP, Fekkes M, Koopman HM, Theunissen NCM: Health-related quality of life measure for children-the TACQoL. J Applied Therapeutics 1997, 1: 357-360.

6. Ravens-Sieberer U, Gosch A, Rajmil L, Erhart M, Bruil J, Power M, Duer W, Auquier P, Cloetta B, Czemy L: The KIDSCREEN-52 Quality of Life Measure for Children and Adolescents: Psychometric Results from a Cross-Cultural Survey in 13 European Countries. Value Health 2007, 11: 645-58.

7. Landgraf JM, Abetz L, Ware JE: Child Health Questionnaire (CHQ). A User’s Manual. Boston: HealthAct; 1999.

8. Bullinger M, Schmidt S, Petersen C: Assessing quality of life of children with chronic health conditions and disabilities: a European approach. Int J Rehabil Res 2002, 25: 197-206.

9. Chan KS, Mangione-Smith R, Burwinkle TM, Rosen M, Varni JW: The PedsQL: reliability and validity of the short-form generic core scales and Asthma Module. Med Care 2005, 43: 256-265.

10. Meeske K, Katz ER, Palmer SN, Burwinkle T, Varni JW: Parent proxy-reported health-related quality of life and fatigue in pediatric patients diagnosed with brain tumors and acute lymphoblastic leukemia. Cancer 2004, 101: 2116-2125.

11. Varni JW, Burwinkle TM, Sherman SA, Hanna K, Berrin SJ, Malcarne VL, Chambers HG: Health-related quality of life of children and adolescents with cerebral palsy: hearing the voices of the children. Dev Med Child Neurol 2005, 47: 592-597.

12. Varni JW, Burwinkle TM, Seid M, Skarr D: The PedsQL 4.0 as a pediatric population health measure: feasibility, reliability, and validity. Ambul Pediatr 2003, 3: 329-341.

13. Bastiaansen D, Koot HM, Bongers IL, Varni JW, Verhulst FC: Measuring quality of life in children referred for psychiatric problems: psychometric properties of the PedsQL 4.0 generic core scales. Qual Life Res 2004, 13: 489-495.

14. Cohen J: Statistical power analysis for the behavioral sciences. New York: Academy Press; 1988.

15. Statistics Netherlands. www.cbs.nl. 2009. 16. Hatzmann J, Heymans HS, Carbonell A,

van Praag BM, Grootenhuis MA: Hidden consequences of success in pediatrics: parental health-related quality of life--results from the Care Project. Pediatrics 2008, 122: e1030-e1038.

17. Brach C, Lewit EM, VanLandeghem K, Bronstein J, Dick AW, Kimminau KS, LaClaire B, Shenkman E, Shone LP, Swigonski N: Who’s enrolled in the State Children’s Health Insurance Program (SCHIP)? An overview of fi ndings from the Child Health Insurance Research Initiative (CHIRI). Pediatrics 2003, 112: e499.

18. Reinfj ell T, Diseth TH, Veenstra M, Vikan A: Measuring health-related quality of life in young adolescents: reliability and validity in the Norwegian version of the Pediatric Quality of Life Inventory 4.0 (PedsQL) generic core scales. Health Qual Life Outcomes 2006, 4: 61.

chapter 2

36

19. Kulkarni AV, Cochrane DD, McNeely PD, Shams I: Medical, Social, and Economic Factors Associated with Health-Related Quality of Life in Canadian Children with Hydrocephalus. J Pediatr 2008, 153: 689-95.

20. Sherman EM, Griffi ths SY, Akdag S, Connolly MB, Slick DJ, Wiebe S: Sociodemographic correlates of health-related quality of life in pediatric epilepsy. Epilepsy Behav 2008, 12: 96-101.

21. Varni JW, Burwinkle TM, Seid M: The PedsQL 4.0 as a school population health measure: feasibility, reliability, and validity. Qual Life Res 2006, 15: 203-215.

22. Grootenhuis MA, Koopman HM, Verrips EG, Vogels AG, Last BF: Health-related quality of life problems of children aged 8-11 years with a chronic disease. Dev Neurorehabil 2007, 10: 27-33.

23. Varni JW, Limbers CA, Burwinkle TM, Bryant WP, Wilson DP: The ePedsQL in type 1 and type 2 diabetes: feasibility, reliability, and validity of the Pediatric Quality of Life Inventory Internet administration. Diabetes Care 2008, 31: 672-677.

373Chapter 3

V. Engelen 1H.M. Koopman 2S.B. Detmar 3

H. Raat 4

M.D. van de Wetering 5

P. Brons 6 J.K. Anninga 7 F. Abbink 8M.A. Grootenhuis 1

1 Psychosocial Department, Academic Medical Centre/ Emma Children’s Hospital, Amsterdam 2 Medical Psychology Department, Leiden University Medical Centre, Leiden 3 Prevention and Health Department, TNO, Leiden4 Department of Public Health, Erasmus University, Rotterdam5 Paediatric Oncology Department, Academic Medical Centre/ Emma Children’s Hospital, Amsterdam6 Paediatric Oncology Department, Radboud University Medical Centre, Nijmegen7 Paediatric Oncology Department, Leiden University Medical Centre, Leiden8 Paediatric Oncology Department, VU University Medical Centre, Amsterdam

Health-related quality of life after completion of successful treatment for childhood cancer

PEDIATR BLOOD CANCER. 2011 APR; 56 (4):646-53.

38

Abstract

Background. Previous studies have assessed health-related quality of life (HRQOL) during several treatment stages in children with cancer, but there is limited knowledge about HRQOL shortly after completing therapy. This study determined HRQOL of children with cancer shortly after the end of successful treatment compared with normative values.

Procedure. Several age-specifi c HRQOL questionnaires were administered: the ITQOL (generic, proxy-report, 0 to 4 years), CHQ PF 50 (generic, proxy-report, 5 to 7 years), Kidscreen (generic, self-report, 8 to 18 years) and Disabkids (chronic generic, self-report, 8 to 18 years).

Results. Children with cancer (N=191, mean age 9.25, SD 5.06, 47.1% female) participated. Physical well-being was aff ected for all ages. Compared to normative values zero- to seven-year-olds were rated signifi cantly lower on the majority of the scales. In addition, 12- to 18- year-olds had signifi cantly better HRQOL than the norm on social scales. Compared to chronically ill norms, 8- to 18-years-olds demonstrated no diff erences, except for 12- to 18-year-olds who experienced signifi cantly more physical limitations. Additionally, we found that HRQOL of parents of 0- to 7-year-olds was poorer than the norm.

Conclusion. HRQOL in children with cancer and their parents can be impaired compared with the norm. Therefore, HRQOL should be monitored in clinical practice to make paediatric oncologists aware of these problems. For young children, we recommend checking whether certain HRQOL problems can be explained by parental worries. For older children and adolescents, paediatric oncologists need to consider social desirability and the child’s adaptive style.

39

IntroductionIn the Netherlands, approximately 500 children each year are diagnosed with cancer [1]. Chances to be cured from this disease have increased over the past several decades; currently, the survival rate is about 75% [2].

Health-related quality of life (HRQOL) has become an important outcome measure in paediatric oncology research. The World Health Organisation defi nes quality of life (QOL) as ‘individuals’ perceptions of their position in life, in the context of the culture and value systems in which they live, and in relation to their goals, expectations, standards and concerns” [3]. The concept of HRQOL refers to the impact of health and illness on an individual’s QOL [4].

Several studies [5-13] have already demonstrated that childhood cancer diagnosis can lead to poor HRQOL compared to healthy peers. For example, Varni et al. [11] studied HRQOL in a large paediatric oncology sample. HRQOL scores on all scales (total, physical, psychosocial, emotional, social and school functioning) were signifi cantly lower for children with cancer than healthy controls on both child self-report and parent proxy-report. This study included patients from all diagnostic groups and diff erent treatment stages. Eiser et al. [12] also assessed HRQOL in children with diverse types of cancer about three months after diagnosis. Again, HRQOL of these children was rated signifi cant lower than the norm. Stam et al. [13] investigated HRQOL in children with all cancer diagnoses about two months the after end of successful treatment. Their conclusion was that both patients and parents experience worse well-being than the reference group; this was especially true for motor functioning.

The vast majority of these HRQOL studies in paediatric oncology have been conducted with generic HRQOL measures and with healthy children as a control group. However, little is known about how children with cancer compare to children with a chronic disease. On the surface, there seem to be many similarities between having cancer and having a chronic condition: e.g., frequent hospital visits, receiving treatment, physical discomfort and the mental and social infl uence of the disease on daily life. The main diff erence, however, is the life-threatening nature of cancer. Hence, for researchers as well as paediatricians, it is important to learn how children with cancer compare to chronically ill peers with respect to HRQOL aspects that pertain not to specifi c conditions but to chronic conditions in general (chronic generic HRQOL) [14]. Previous research has focused on comparisons within disease groups and the HRQOL measures used were of a generic nature [15]. Up to now, the paediatric oncology literature has not provided any reports on chronic generic HRQOL.

HRQOL of children suff ering from cancer also depends on treatment status. Thus far, HRQOL has been assessed during diff erent treatment stages (e.g., shortly after diagnosis and at several stages of treatment), but to the best of our knowledge only one study [13] explicitly explored HRQOL of children with cancer shortly after therapy. From a clinical

chapter 3

40

perspective this is remarkable because the period shortly after completion of treatment can be very stressful for patients and parents [13]. This stress can be further exacerbated by fear of relapse when medication has stopped. Finishing treatment also means fewer hospital visits, which can lead to reduced (professional) support for these families in an uncertain time. To provide tailored psychosocial care, more HRQOL research is needed to understand how these children are doing in this phase shortly following the end of treatment.

The aim of the current study was to determine HRQOL of children with cancer shortly after the end of successful treatment, compared to a general and chronically ill norm group.

Methods

ParticipantsChildren with cancer (0-18 years of age) and their parents participated in this study. Data were collected as part of the Quality of Life In Childhood Oncology study (QLIC-ON), a large Dutch multi-centre sequential cohort research project [16]. Participating centres were the Academic Medical Centre/ Emma Children’s Hospital, Leiden University Medical Centre, Radboud University Medical Centre and the VU University Medical Centre, the Netherlands.

QLIC-ON provides patient reported outcomes (PRO) about HRQOL to paediatric oncologists through use of the QLIC-ON PROfi le (p. 173), and investigates whether the QLIC-ON PROfi le is an eff ective tool for identifying and discussing HRQOL problems in children with cancer. The study consists of a control period (March 2006 – January 2008) and an intervention period (January 2008 – November 2009). The child (8-18 years) or parent (for children younger than 8) completed a digital HRQOL questionnaire prior to the fi rst three follow-up consultations with the paediatric oncologist. Results of this questionnaire were summarised in the QLIC-ON PROfi le. The QLIC-ON PROfi le was then presented to the paediatric oncologist for patients in the intervention group during the three consultations. HRQOL was assessed at baseline and at follow-up as well; these measurements were not included in the QLIC-ON PROfi le. The HRQOL data reported in the current study are from the baseline assessment.

ProcedureChildren and parents were invited by mail to participate as soon as the child had fi nished treatment. Children who had undergone stem cell transplantation (SCT) were approached for participation 6 months after SCT. This timeline was determined by a SCT physician and was considered most comparable to the post-treatment stage in children without SCT. Participants completed a baseline package at home including a HRQOL questionnaire and socio-demographic questions. Informed consent was obtained in the out-patient

41

clinic shortly before their next consultation. The study was approved by the Medical Ethics Committee of all participating centres.

Measures

Socio-demographic and medical information

Socio-demographic information collected included: age and gender of the child and parent, and parental country of birth, education and employment status. Information on child age in the non-participants group was retrieved from medical fi les. Medical data for participants and non-participants were gathered by data managers or paediatric oncologists. These data included diagnosis, treatment, treatment duration and time since end of treatment.

HRQOL

Because of the broad age range used in this study and the related varying developmental stages of the child, HRQOL was measured by means of age-specifi c instruments: the Infant and Toddler Quality Of Life Questionnaire (ITQOL, generic, ages 0 to 4 years), the Child Health Questionnaire Parent Form 50 (CHQ PF 50, generic, ages 5 to 7 years) and the Kidscreen (generic, ages 8 to 18 years). Chronic generic HRQOL was assessed with the Disabkids (ages 8 to 18 years).

ITQOL

The ITQOL was completed by parents of children aged 0 to 4 years. The ITQOL assesses generic HRQOL in children and parents and consists of 97 items divided into 9 multi-items and 2 single-item scales: physical functioning, growth and development, bodily pain, temperament and moods, general behaviour, getting along, general health perceptions (parent scale), parental impact-emotional (parent scale), parental impact-time (parent scale), family cohesion, and change in health. Items were rated on 4- to 6-point Likert scales. Raw scores were summed and converted to a 0-100 scale with higher scores indicating better functioning and well-being. The ITQOL scales general behaviour, getting along and change in health are only relevant for children one year and older. Internal consistency, concurrent and discriminative validity of the ITQOL are good; however, there are some concerns about ceiling eff ects and test-retest reliability [17]. Raat et al. [17] provided the Dutch reference data for the ITQOL. The ITQOL recall period of one month was shortened to one week to ensure that all children were off treatment at the time and to equalise the recall period across all measures.

CHQ PF 50

The CHQ PF 50 was completed by parents of children aged 5 to 7 years. The CHQ PF 50 measures generic HRQOL in children (and parents) and consists of 50 items divided into 11 multi-item and 2 single-item scales: physical functioning, role functioning-emotional/

chapter 3

42

behavioural, role functioning-physical, bodily pain, general behaviour, mental health, self-esteem, general health perceptions (parent scale), parental impact-emotional (parent scale), parental impact-time (parent scale), family activities, family cohesion and change in health. Items were rated on 4- to 6-item Likert scales. Raw scores were summed and converted to a 0-100 scale with higher scores indicating better functioning and well-being. Physical and psychosocial summary scores were calculated following the CHQ PF 50 manual [18]. Concurrent and discriminative validity of the CHQ PF 50 are good, but test-retest reliability needs further study [19]. Raat et al. [19] provided the Dutch reference data for the CHQ PF 50. For the reasons noted above, the CHQ PF 50 recall period of one month was changed to one week as well.

Kidscreen

In this study, the Kidscreen was administered as a self-reported questionnaire for children aged 8 to 18 years. The Kidscreen evaluates generic HRQOL in children by means of 52 items divided over 10 dimensions: physical well-being, psychological well-being, moods & emotions, self-perception, autonomy, parent relation & home life, peers & social support, school environment, bullying and fi nancial resources. Items were scored on a 5-point Likert scale [14]. Rasch scores are computed for each dimension and transformed into T-values with a mean of 50 and a standard deviation of 10. Higher scores indicate better HRQOL. The Kidscreen has acceptable levels of reliability and validity [20]. Dutch reference data (means) were provided by the Kidscreen group. The recall period was one week.

Disabkids

The Disabkids (Condition Generic Module) short form was employed as a child self-report in children aged 8 to 18 years. This questionnaire assesses chronic generic HRQOL in children and adolescents with various chronic health conditions. The 12 Disabkids items yield 6 subscales grouped into three major areas: mental (1. independence and 2. emotion), social (3. inclusion and 4. exclusion) and physical (5. limitation and 6. medical treatment). The medical treatment scale was not included in the current study because children participating in our study have recently fi nished therapy. Answers were scored on a 5-point Likert scale, summing to a maximum total raw score of 60 and eventually transforming to a 0-100 scale. High scores indicate better HRQOL. The Disabkids has good internal consistency, split-half reliability and discriminative validity; convergent validity is moderate [14, 21]. Dutch reference data were provided by the Disabkids group. The reference group included children suff ering from bronchial asthma, juvenile idiopathic arthritis, atopic dermatitis, diabetes mellitus, cerebral palsy, cystic fi brosis and epilepsy. The recall period was one week.

Statistical analysisData were analysed with SPSS 16.0.2. Diff erences between participants and non-

43

participants with respect to age group, gender, diagnosis and treatment were analysed by means of Chi2 tests. Diff erences in age and treatment duration were examined with t-tests. ITQOL, CHQ PF 50, Kidscreen and Disabkids scores were computed according to their respective manuals. Subsequently, independent sample t-tests (ITQOL, CHQ PF 50, Disabkids) and one sample t-tests (Kidscreen) were conducted to analyse diff erences in HRQOL scores between the oncology sample and the norm group. To prevent too many fi ndings occurring by chance (Type I error), these tests were carried out at a Bonferroni-adjusted level of signifi cance (p<.01). Eff ect sizes were calculated by dividing the diff erence in mean scores between the oncology sample and the norm group by the standard deviation of the norm group. Eff ect sizes up to 0.2 were considered to be small, eff ect sizes of about 0.5 moderate and eff ect sizes of about 0.8 large [22].

Results

Socio-demographic and medical informationOver the period of March 2006 and April 2009, 272 children with cancer and their parents were approached to participate in the QLIC-ON study. Eventually, 191 (70.2%) participants completed the baseline package (Table 1).

Participants and non-participants diff ered signifi cantly with respect to child age, radiotherapy and SCT. No statistically signifi cant diff erences were observed among the groups for treatment duration, child gender, diagnosis, surgery and chemotherapy.

ITQOL HRQOL of this young oncology sample (aged 0 to 4 years) compared to the Dutch norm population is presented in Table 2. Based on proxy-report, children with cancer scored signifi cantly lower than the normative value on physical functioning, growth and development and temperament and moods. In addition, parents perceived their child’s health as signifi cantly better than a year before. Eff ect sizes varied from 0.58 to 0.91. Parents of these young children also reported signifi cantly poorer HRQOL compared to the norm on all three parent scales. Eff ect sizes were large (0.99 to 2.16).

CHQ PF 50 HRQOL data for the 5- to 7-year-olds are reported in Table 3. Parents rated HRQOL of children with cancer signifi cantly lower on all scales compared to the norm, except for role functioning-emotional/ behavioural, role functioning-physical, general behaviour and family cohesion. Eff ect sizes ranged from 0.55 to 7.84. For parents of children aged 5 to 7 we found the same pattern as for parents that completed the ITQOL: parental HRQOL was signifi cantly lower than the norm with large eff ect sizes ranging from 1.29 to 2.33.

chapter 3

44

Table 1. Socio-demographics and medical information participants and non-participants

*p<.05, **p<.01, ***p<.001

CHILD N M SD N M SD

PARENT N M SD N M SD

Participants Non-participants

Age (years) 191 9.25** 5.06 81 10.93 4.26 Treatment duration (months)1 190 13.63 9.17 78 13.79 8.03 Time off treatment (weeks)2 191 7.60 4.53 N % N % Age Group (years)** 0-4 52 27.2 9 11.1 5-7 36 18.8 13 16.0 8-11 36 18.8 26 32.1 12-18 67 35.1 33 40.7 Gender (female) 90 47.1 34 42.0 Diagnosis3 Leukemia 69 36.1 38 46.9 Lymphoma 23 12.0 9 11.1 Brain tumor 16 8.4 12 14.8 Solid tumor 42 22.0 11 13.6 Bone tumor 28 14.7 8 9.9 Other 13 6.8 3 3.7 Treatment Surgery 101 52.9 39 48.1 Radiotherapy 55* 28.8 34 42.0 Chemotherapy 184 96.3 79 97.5 SCT 18** 9.4 20 24.7

Age (years) 190 40.89 7.12 N % Gender (female) 150 79.8 Country of birth (Netherlands) 166 86.9 Education4 Low 37 19.4 Middle 96 50.3 High 57 29.8 Employed 125 65.4

45

SCT: Stem Cell Transplantation; 1Treatment duration: date of end of treatment minus date of diagnosis; 2Time off treatment: date of completion of the HRQOL questionnaire minus date of end of treatment (only for participants); 3Diagnosis: Leukemia: Lymphoid leukemia and acute myeloid leukemia, Lymphoma: Hodgkin lymphoma and Non-Hodgkin lymphoma, Brain tumor: ependymoma and choroid plexus tumor, astrocytoma and low grade glioma, intracranial and intraspinal embryonal tumor, intracranial and intraspinal germ cell tumor, Solid tumor: neuroblastoma and ganglioneuroblastoma, renal tumor, hepatic tumor, rhabdomyosarcoma, fi broblatoma, peripheral nerve sheath tumor, and fi brous neoplasm, Bone tumor: osteosarcoma, ewing tumor and related sarcoma of bone; 4Education: Low: no education, primary school and primary vocational education, Middle: secondary school and secondary vocational education, High: higher vocational education and university.

Table 2. Results ITQOL (parent report, ages 0-4 years)

*p<.01, **p<.001; Scale: 0-100

Children with cancer Norm Effect size ITQOL N M SD M d

Child scale Physical functioning 49 89.05* 16.22 97.20 0.83Growth and development 51 78.21** 13.35 86.50 0.78Bodily pain 49 77.21 21.64 83.82 0.39Temperament and moods 50 71.05* 14.92 77.18 0.58General behaviour 44 70.68 19.74 72.82 0.17Getting along 45 69.68 11.87 71.42 0.20Family cohesion 48 74.38 21.72 75.31 0.05Change in health 45 72.78* 35.28 56.11 0.91 Parent scale General health perceptions 49 47.77** 19.71 79.02 2.16Parental impact-emotional 50 75.07** 20.41 92.11 1.63Parental impact-time 50 82.10** 15.74 93.00 0.99

chapter 3

Table 3. Results CHQ PF 50 (parent report, ages 5-7 years)

*p<.01, **p<.001; Scale: 0-100

Children with cancer Norm Effect size CHQ PF 50 N M SD M d

Child scale Physical functioning 36 83.33** 21.58 99.48 7.84Role functioning-emotional/behavioural 36 91.05 20.88 98.48 1.34Role functioning-physical 35 88.57 22.06 95.73 0.41Bodily pain 35 78.29* 22.56 88.03 0.65General behaviour 36 71.93 20.65 79.94 0.59Mental health 36 76.46* 13.74 83.03 0.55Self esteem 36 75.21* 13.83 81.52 0.58Family activities 36 79.51* 21.00 89.57 0.98Family cohesion 36 73.06 21.26 74.57 0.08Physical summary 34 46.62** 11.11 56.73 1.88Psychosocial summary 34 50.61* 7.99 54.43 0.60 Parent scale General health perceptions 36 54.80** 18.36 83.28 2.33Parental impact-emotional 36 70.83** 21.32 88.46 1.29Parental impact-time 36 84.26* 18.66 96.11 1.39

46

KidscreenTable 4 shows the results of children with cancer (aged 8 to 11) that completed the Kidscreen compared with the norm. No signifi cant diff erences between the two groups were found, except for physical well-being (eff ect size 1.17).

HRQOL results for adolescents (12 to 18 years) with cancer compared with the norm are depicted in Table 5. Compared to the norm, the oncology sample only reported signifi cant lower scores on the physical well-being scale (eff ect size 0.89). Furthermore, the HRQOL of adolescents with cancer was signifi cantly better than the norm on the parent relation & home life, school environment, bullying and fi nancial resources scales. The eff ect sizes were small to moderate and varied from 0.32 to 0.59.

DisabkidsParticipants between the ages of 8 to 18 also completed the Disabkids (Table 6). Children with cancer (8 to 11 years of age) did not diff er with respect to HRQOL from normative values for children with a chronic disease. Adolescents (12 to 18 years of age), however, perceived signifi cantly more physical limitations than expected from the chronic disease norm group (eff ect size 0.74).

Table 4. Results Kidscreen (child report, ages 8-11 years)

**p<.001; Scale: M=50, SD=10

Children with cancer Norm Effect size Kidscreen 8-11 N M SD M d

Physical well-being 35 46.11** 7.76 57.26 1.17Psychological well-being 36 52.32 10.90 55.74 0.38Moods & emotions 35 51.30 9.76 52.51 0.13Self-perception 34 58.10 8.95 57.40 0.07Autonomy 36 54.86 8.38 56.45 0.18Parent relation & home life 35 56.77 7.55 55.47 0.16Peers & social support 33 53.24 10.69 53.08 0.02School-environment 36 59.94 11.03 58.42 0.15Bullying 36 49.64 9.77 48.08 0.14Financial resourses 28 51.58 11.53 51.36 0.02

47

chapter 3

Table 5. Results Kidscreen (child report, ages 12-18 years)

*p<.01, **p<.001; Scale: M=50, SD=10

Children with cancer Norm Effect size Kidscreen 12-18 N M SD M d

Physical well-being 64 42.10** 8.99 50.42 0.89Psychological well-being 66 52.31 9.14 51.89 0.05Moods & emotions 67 53.98 11.79 50.46 0.37Self-perception 67 49.50 8.99 49.87 0.04Autonomy 65 51.86 7.92 53.48 0.18Parent relation & home life 66 57.77** 7.68 52.29 0.59Peers & social support 67 52.21 10.09 51.74 0.05School-environment 55 54.87* 8.83 50.68 0.48Bullying 61 52.18* 8.57 48.81 0.33Financial resourses 66 55.70* 9.21 52.72 0.32

Table 6. Results Disabkids (child report, ages 8-18 years)

*p<.01; Scale: 0-100

Children with cancer Norm Effect size Disabkids N M SD M d

8-11 years Total score 37 77.23 14.35 80.08 0.18Mental – Independence 37 74.32 20.40 79.17 0.19Mental – Emotion 37 73.31 21.07 71.57 0.07Social – Inclusion 37 77.03 25.77 84.00 0.36Social – Exclusion 37 84.46 20.71 86.03 0.09Physical – Limitation 37 77.03 14.29 79.50 0.12 12-18 years Total score 68 77.61 15.40 80.42 0.24Mental – Independence 68 76.84 22.05 83.07 0.34Mental – Emotion 68 70.96 23.99 66.15 0.26Social – Inclusion 68 80.51 21.53 80.73 0.01Social – Exclusion 68 84.01 17.92 86.46 0.15Physical – Limitation 68 75.74* 18.05 85.68 0.74

DiscussionThis study demonstrates that young children with cancer have lower HRQOL scores shortly after the end of treatment, compared to the norm (proxy-report). Older children and adolescents, however, show minimal diff erences from the general and chronically ill norm groups and even score better on some indices (self-report). In addition, HRQOL of

48

parents of children with cancer 0 to 7 years of age is poor measured up to the norm.A general fi nding across all age groups and all instruments is that children with

cancer have signifi cantly worse physical well-being than the general and chronically ill norm groups. As participants in our study have recently fi nished an intense treatment with a physical burden, this fi nding is not surprising and is in line with previous reports [11-13].

For children with cancer 0 to 7 years of age parents reported signifi cantly worse well-being than the norm on emotional scales (ITQOL and CHQ PF 50). These fi ndings correspond with the results of an earlier study [13]. Furthermore, the positive change in health for the 0- to 4-year-olds with cancer matches our expectations; a year before, most of these children were either only recently diagnosed or still undergoing treatment. Additionally, our fi nding that children with cancer who are 8 years and older were indistinguishable from the norm on the majority of scales (Kidscreen) was consistent with the existing literature [13, 23]. A possible explanation for this phenomenon is benefi t fi nding in children with cancer. Although children have experienced a serious illness, positive outcomes are not uncommon following childhood cancer [24]. Another clarifi cation could be the adaptive style of children with cancer. A higher proportion of repressors have been found in groups of children with cancer relative to healthy controls. Repressors tend to view themselves as well adjusted and lacking in worry. In addition they report better HRQOL [25]. However, considering that older children were less willing to participate in our study, it is also possible that there is a diff erence in HRQOL between adolescents that did and did not take part. Perhaps adolescents that experience fewer problems are more able and prepared to participate, which caused our results to appear more positive than is true of the underlying population. Furthermore, children and adolescents with cancer do not diff er from children with a chronic health condition on measures of chronic generic HRQOL (Disabkids). The subjective perception of these children’s HRQOL is comparable.

A notable outcome was that teenagers with cancer reported signifi cantly and clinically relevant better social HRQOL than the norm (Kidscreen) but displayed no diff erences on equivalent domains compared with the chronic health condition group (Disabkids). This phenomenon might be attributed to response shift [26]. Response shift is the change in the meaning of one’s self-evaluation of a target construct, e.g., HRQOL. Possibly, children confronted with cancer or any other disease change their internal standards of what is important in life, and therefore diff er from the general population norm, but not from each other.

The family cohesion (0- to 7-year-olds) and parent relation & home life scales (8- to 18-year-olds) demonstrate salient results. Children with cancer show either no diff erences from the norm or even obtain higher scores on these scales. It seems that the experience of being confronted with childhood cancer does not necessarily have a negative impact on the family. On the contrary, we believe it is possible that the family system functions as a protective factor for children dealing with this malignant disease, at least for the older

49

chapter 3

children. Goldbeck [27] found similar results shortly following cancer diagnosis.The diff erence in HRQOL between young (ITQOL and CHQ PF 50) and older children

or adolescents (Kidscreen and Disabkids) is striking. This might be clarifi ed by the use of diff erent types of HRQOL instruments but is most likely to be explained by proxy- versus self-report. Previous studies [8, 28] have shown large diff erences between the two forms of report, with parents rating their children lower than children rate themselves. Hence, it is possible to assume that parents and children perceive the phase of fi nishing treatment in divergent ways. Children live in the moment (“I am better and want to go to school and see my friends again”), whereas parents are looking forward (“What are the late eff ects and what if the disease recurs?”). In our study, parental worries are clearly refl ected in ITQOL as well as CHQ PF 50 results: eff ect sizes for the parent as well as child scales are rather large. Kidscreen data, however, tell us that older children and especially adolescents report fewer concerns. This result is confi rmed by van Dijk et al. [29]. From clinical experience we know that children are relieved once treatment is fi nished. They are often resilient, avoid looking back, are fed up with the hospital, and therefore, might overrate their HRQOL. Beside the response shift phenomenon, benefi t fi nding and adaptive style described above, childhood resilience and social desirability might also clarify the signifi cant better social functioning of the adolescents with cancer compared to the norm. Eff ect sizes are small to moderate though, which limits the clinical relevance of these results.

Further results indicate that parents of children aged 0 to 7 years seem to struggle: they obtain low scores with respect to general health perceptions, parental impact-emotional and parental impact-time. Previous research also reports low HRQOL and post traumatic stress in parents of children with cancer [13, 27, 30-32]. Although the child has successfully fi nished treatment, parents have concerns about the future. Moreover, they might be insecure, anxious and sad about what has happened to their child. Furthermore, it is diffi cult for the parent to fi nd some personal time for leisure or relaxation because having a sick child demands a lot of time and energy [33].

One strength of the current study is that we collected HRQOL data in a large sample of children with cancer, representing all childhood ages, at one moment in time, collected at four diff erent academic medical centres in the Netherlands. This supports the generalisability of our fi ndings. Another advantage of our study is that we incorporated a chronic generic HRQOL module in addition to the generic HRQOL questionnaires. This is rather innovative in paediatric oncology HRQOL research and therefore provides interesting information for researchers and paediatricians. A fi nal strength of this paper is that it reports on parental HRQOL as well as paediatric HRQOL.

Our study also has a number of limitations. First, there are some diff erences between participants and non-participants. Older children and adolescents were less willing to take part, likely because they are relieved that treatment is complete. They want to continue their normal lives again and do not want to be confronted with their illness. Younger children, however, did not have a choice; parents decided whether to complete the proxy-

50

report. Additionally, the percentage of children that had received radiotherapy and SCT was larger in the non-participating group. These children are likely to have suff ered more from their treatment compared to chemotherapy or surgery and are still dealing with recovery from intensive therapy. Both parents and children might perceive participation in the current study as too demanding. Additionally, one can assume that children treated with radiotherapy and SCT have poorer HRQOL than children who did not receive those therapies. Because the group of participating children has fewer children that have undergone radiotherapy and SCT, it is possible that our HRQOL results demonstrate an overestimation of the children’s real HRQOL. Further, it would have been interesting to include both a HRQOL self-report as well as a HRQOL proxy-report where possible and to have measured HRQOL of parents of children aged 8 to 18 as well. However, given that this study was part of a larger project (QLIC-ON) with many assessments, we tried limit the amount of questionnaires where possible. Moreover, although we have collected data from a large sample, the diff erent age groups and their associated questionnaires made the sub samples relatively small for comparison with norm data.

General clinical implications of the current study are that paediatric oncology professionals (e.g., physicians, nurse practitioners, nurses, psychologists, social workers) should monitor HRQOL in all children on a regular basis once treatment is complete. More specifi cally, for young children, we recommend professionals check whether certain HRQOL problems can be explained by parental worries refl ected in the child’s psychosocial functioning. Professionals should be aware of the fact that parental mental functioning is known to infl uence their child’s health [34]. Therefore, professionals must be conscious that a HRQOL problem reported by the child can actually be a refl ection of parental diffi culties. Although older children and adolescents seem to be doing well compared with healthy peers, our advice is the opposite: if the child indicates he or she is experiencing no problems this can be a sign of personal growth, still the professional should always verify this well-being with the parents, to ensure the child is not overestimating his or her functioning.

Likewise, we recommend paediatric oncology professionals be aware of HRQOL problems in parents, particularly in the fragile stage shortly following treatment. It is not unusual for parents to worry during this period of time. Therefore we suggest, in line with earlier research [31, 35], not only to monitor parental HRQOL immediately after fi nishing treatment, but in the period afterwards as well, e.g., 3 and 6 months later. Psychosocial intervention is indicated for those parents whose HRQOL still has not improved half a year after treatment completion.

Future research should investigate the factors infl uencing HRQOL outcomes so it can be determined which groups of children with cancer are at risk for HRQOL problems. More knowledge in this area should help paediatric oncology professionals monitor HRQOL of individual patients.

51

Reference List1. SKION. http://www.skion.nl/. In: 2009.2. KIKA. http://www.kika.nl/. In: 2009.3. WHO. WHOQOL-BREF. Introduction,




6. Eiser C, Vance YH, Horne B, et al. The value of the PedsQLTM in assessing quality of life in survivors of childhood cancer. Child Care Health Dev 2003:29:95-102.


8. Matziou V, Perdikaris P, Feloni D, et al. Cancer in childhood: children’s and parents’ aspects for quality of life. Eur J Oncol Nurs 2008:12:209-216.

9. Felder-Puig R, Frey E, Proksch K, et al. Validation of the German version of the Pediatric Quality of Life Inventory (PedsQL) in childhood cancer patients off treatment and children with epilepsy. Qual Life Res 2004:13:223-234.





14. Ravens-Sieberer U, Schmidt S, Gosch A, et al. Measuring subjective health in children and adolescents: results of the European KIDSCREEN/DISABKIDS Project. Psychosoc Med 2007:4:Doc08.

15. Varni JW, Limbers CA, Burwinkle TM. Impaired health-related quality of life in children and adolescents with chronic conditions: a comparative analysis of 10 disease clusters and 33 disease categories/severities utilizing the PedsQL 4.0 Generic Core Scales. Health Qual Life Outcomes 2007:5:43.

16. Engelen V, Haverman L, Koopman H, et al. Development and implementation of a patient reported outcome intervention (QLIC-ON PROfi le) in clinical paediatric oncology practice. Patient Educ Couns 2010.

17. Raat H, Landgraf JM, Oostenbrink R, et al. Reliability and validity of the Infant and Toddler Quality of Life Questionnaire (ITQOL) in a general population and respiratory disease sample. Qual Life Res 2007:16:445-460.


19. Raat H, Bonsel GJ, Essink-Bot ML, et al. Reliability and validity of comprehensive health status measures in children: The Child Health Questionnaire in relation to the Health Utilities Index. J Clin Epidemiol 2002:55:67-76.

20. Ravens-Sieberer U, Gosch A, Rajmil L, et al. The KIDSCREEN-52 quality of life measure for children and adolescents: psychometric results from a cross-cultural survey in 13 European countries. Value Health 2008:11:645-658.

21. The European DISABKIDS group. The DISABKIDS questionnaires: Quality of life questionnaires for children with chronic conditions. Lengerich, Germany: Pabst Science; 2006.

22. Cohen J. Statistical power analysis for the behavioral sciences. New York: Academy Press; 1988.

23. Phipps S, Jurbergs N, Long A. Symptoms of post-traumatic stress in children with cancer: does personality trump health status? Psychooncology 2009:18:992-1002.

chapter 3

52

24. Michel G, Taylor N, Absolom K, Eiser C. Benefi t fi nding in survivors of childhood cancer and their parents: further empirical support for the Benefi t Finding Scale for Children. Child Care Health Dev 2010:36:123-129.

25. Jurbergs N, Russell KM, Long A, Phipps S. Adaptive style and diff erences in parent and child report of health-related quality of life in children with cancer. Psychooncology 2008:17:83-90.

26. Sprangers MA, Schwartz CE. Integrating response shift into health-related quality of life research: a theoretical model. Soc Sci Med 1999:48:1507-1515.

27. Goldbeck L. The impact of newly diagnosed chronic paediatric conditions on parental quality of life. Qual Life Res 2006:15:1121-1131.

28. Cremeens J, Eiser C, Blades M. Factors infl uencing agreement between child self-report and parent proxy-reports on the Pediatric Quality of Life Inventory 4.0 (PedsQL) generic core scales. Health Qual Life Outcomes 2006:4:58.

29. van Dijk J., Huisman J, Moll AC, et al. Health-related quality of life of child and adolescent retinoblastoma survivors in the Netherlands. Health Qual Life Outcomes 2007:5:65.

30. Klassen AF, Klaassen R, Dix D, et al. Impact of caring for a child with cancer on parents’ health-related quality of life. J Clin Oncol 2008:26:5884-5889.

31. Hatzmann J, Heymans HS, Carbonell A, et al. Hidden consequences of success in pediatrics: parental health-related quality of life--results from the Care Project. Pediatrics 2008:122:e1030-e1038.

32. Alderfer MA, Cnaan A, Annunziato RA, Kazak AE. Patterns of posttraumatic stress symptoms in parents of childhood cancer survivors. J Fam Psychol 2005:19:430-440.

33. Hatzmann J, Maurice-Stam H, Heymans HS, Grootenhuis MA. A predictive model of Health Related Quality of life of parents of chronically ill children: the importance of care-dependency of their child and their support system. Health Qual Life Outcomes 2009:7:72.

34. Prince M, Patel V, Saxena S, et al. No health without mental health. Lancet 2007:370:859-877.

35. Vrijmoet-Wiersma CM, Egeler RM, Koopman HM, et al. Parental stress before, during, and after pediatric stem cell transplantation: a review article. Support Care Cancer 2009:17:1435-1443.

534Chapter 4

V. Engelen1

M.C.B. Van Zwieten 2S.B. Detmar 3

H.M. Koopman 4 A.Y.N Schouten – van Meeteren 5 M.A. Grootenhuis 1

1 Paediatric Psychosocial Department, Academic Medical Center/ Emma Children’s Hospital, Amsterdam 2 Department of General Practice/ Medical Ethics, Division Clinical Methods and Public Health, Academic

Medical Centre, University of Amsterdam, Amsterdam 3 TNO Prevention and Health, Leiden4 Medical Psychology, Leiden University Medical Center, Leiden5 Paediatric Oncology Department, Academic Medical Center/ Emma Children’s Hospital, Amsterdam

Psychosocial issues addressed by paediatric oncologists

SUBMITTED FOR PUBLICATION

54

Abstract

Purpose. To examine paediatric oncologists’ perception of their role in (1) discussing psychosocial functioning and identifying psychosocial problems and (2) providing emotional support to children with cancer.

Methods. Twenty-four paediatric oncologists were interviewed. Data were qualitatively analysed using a framework approach.

Results. Paediatric oncologists believed it was their task to discuss psychosocial functioning, to identify psychosocial problems and to provide emotional support to children with cancer. Paediatric oncologists also indicated that this task can be limited due to inherent lack of expertise or time or by the physician’s personality.

Conclusions. Paediatric oncologists felt responsible for providing psychosocial care to the best of their ability.

55

IntroductionFulfi lling age-specifi c developmental tasks in childhood is important for successful adjustment in adult life. The burden of having cancer and cancer treatment can interfere with this process [1]. Although the psychological adjustment of most children with cancer is good [2], some children are at risk for psychosocial diffi culties [3]. Because paediatric oncologists are the primary caregivers of children with cancer, they should be able to identify psychosocial problems so that they can prevent deterioration. Overall, paediatricians believe that they are responsible for the identifi cation of mental health issues [4]; however, many physicians may not be aware of such problems [5, 6].

In this study, we examine the paediatric oncologist’s role with respect to psychosocial functioning, psychosocial problems and emotional support in children with cancer. While no one has previously investigated these topics in paediatric oncology, related studies in adult oncology will be discussed below.

The identifi cation of psychosocial problems by oncologists has received some attention in literature. Bultz & Carlson [7] for instance, stressed the importance of routinely evaluating levels of distress in adult oncology. According to previous research, oncologists’ sensitivity in recognising moderate distress in patients is relatively high but their ability to detect severe distress is low [8]. Furthermore, the ability to detect distress varies between diff erent oncologists, and oncologists often underrate distress in their patients [9].

With respect to emotional support, research has shown that compassion expressed by physicians can reduce anxiety in adult cancer patients [10], and physician attentiveness and empathy have been associated with increased self-effi cacy and reduced emotional distress [11]. Additionally, cancer patients who felt that more attention had been paid to the psychosocial aspects of their illness were more satisfi ed with their consultations [12]. Butow et al [13], however, stated that oncologists are not recognizing or dealing with the emotional needs of their patients. If doctors do not recognise and acknowledge patients’ cues for emotional support, patients will be discouraged from seeking support. This fi nding is in agreement with those reported by Ford et al [14], who concluded that the lack of psychosocial questions from oncologists demonstrates how emotional issues are rarely investigated.

The studies described above were all conducted in adult oncology with no research available for paediatrics in general and paediatric oncology more specifi cally. Crossley and Davies [15], however, described several key components of doctor-patient interactions involving children and their parents. Their research is useful for the current study, because it also includes aspects of the consultation process that are essential for exploring psychosocial functioning, psychosocial problems and providing emotional support in children with cancer. Examples of these aspects are general communication skills, patient-parent-centeredness, interpersonal skills, appropriate family plans, liaisons with other health-care professionals and health-related problems.

chapter 4

56

In sum, the importance of exploring psychosocial functioning, identifying psychosocial problems and providing emotional support in cancer patients is documented, but how these aspects are perceived by paediatric oncologists has not been reported. Therefore, the aim of the current study is to examine paediatric oncologists’ perceptions regarding their role in (1) discussing psychosocial functioning and identifying psychosocial problems and (2) in providing emotional support to children with cancer.

Methods

Participants and contextTwenty four paediatric oncologists were interviewed for the purpose of the current paper, which presents a descriptive and explorative study.

The interviews were performed in the context of a larger project: the multi-centre QLIC-ON study (Quality of Life in Childhood Oncology) [16, 17]. The QLIC-ON study provided a patient reported outcome (PRO) about health related quality of life (HRQOL) (the QLIC-ON PROfi le, p. 173) to paediatric oncologists and investigated whether the QLIC-ON PROfi le was an eff ective tool for systematically identifying and discussing HRQOL problems in children with cancer. A sequential cohort study design was applied in the QLIC-ON study. Paediatric oncologists participated in the control (March 2006 to January 2008) as well as the intervention period (January 2008 to November 2009); children with cancer and their parents participated in either the control period or the intervention period. The QLIC-ON PROfi le was only presented to the paediatric oncologists during the intervention period. Further information and results regarding the QLIC-ON study can be found in previously published papers [16, 17].

The QLIC-ON study is embedded in the HRQOL theme. In paediatrics, there are usually four HRQOL domains: physical, emotional, social and school functioning, of which the last three can be summarised as psychosocial functioning [18]. We were interested in the paediatric oncologists’ perception of psychosocial functioning and therefore, the interview schedule was restricted to the concept of psychosocial functioning (and emotional support) and not HRQOL.

The QLIC-ON study was approved by the Medical Ethics Committees for all participating centres: the Academic Medical Centre/ Emma Children’s Hospital, Leiden University Medical Centre, Radboud University Medical Centre and the VU University Medical Centre in the Netherlands.

Data collectionPaediatric oncologists met the following criteria to be interviewed: 1) none of their patients were still participating in the control period of the QLIC-ON study and 2) their fi rst patient was about to start in the intervention period of the QLIC-ON study.

Eligible paediatric oncologists were invited for the interview by email. Because of the

57

paediatric oncologists’ connection to the QLIC-ON study, the response rate was 100%. The paediatric oncologists were interviewed between January 2008 and January

2009. All interviews were conducted by the fi rst author and took place in a private room in the centre in which the paediatric oncologist was employed. The interviews were audio-taped by means of a voice recorder and verbatim transcribed to a text fi le. Audio and text fi les were saved and confi dentially processed.

At the start of the QLIC-ON study, paediatric oncologists completed a brief socio-demographic questionnaire about themselves.

Interview scheduleThe questions in the interview schedule (Table 1) were used as a guide for each interview. The fi ndings with respect to the questions between brackets are not reported in this paper.

The current paper discusses the themes psychosocial functioning, psychosocial problems (questions 1-5) and emotional support (questions 6-7). To have a notion of what the paediatric oncologists are referring to in the rest of the interview when talking about these themes, we asked the paediatric oncologists to defi ne the themes. To be complete, these defi nitions are reported in this paper as well; however, it was not the primary aim of this study.

Data analysisData were qualitatively analysed using the framework approach [19]. The interview schedule was applied as a framework for data analysis. The software program MaxQDA was used as an aid for analysis.

First, to get familiar with the data set as a whole, 10 interviews were randomly read by the fi rst author. Subsequently, for each question from the interview schedule, open codes were (deductively) created for all 24 interviews. Then, axial codes were generated by (inductively) interpreting, redefi ning and/or merging the open codes, resulting in a code tree (Appendix 1). The fi ndings presented in the current paper are derived from the codes that occurred most frequently.

To ensure that our fi ndings were valid, the analyses were checked by the second author – a senior qualitative researcher - who read 10 interviews and verifi ed the code tree. The remaining authors read 3 interviews and agreed on the content of the code tree. The fi ndings were discussed and approved by all authors. Additionally, the fi ndings were checked for their validity by explicitly asking the fi fth author - a paediatric oncologist who was also interviewed for the purpose of the current study - whether she could recognize the fi ndings and judge these as correct (member’s check) [20].

The paediatric oncologists’ answers concerning questions 2 and 3 from the interview schedule were in close agreement; therefore, we integrated both questions for the analysis.

Eventually, the fi ndings were reported in this paper as quotes, presented in boxes. Each quote was made by one paediatric oncologist.

chapter 4

58

Table 1. Interview schedule

1. How would you define psychosocial functioning?

(Box 1)

2. To what extent do you regard discussing psychosocial functioning with children and parents as your task?

(Box 2a + b)

- (How much attention do you pay to psychosocial functioning during the stage of diagnosis?)

- (How much attention do you pay to psychosocial functioning during the stage of treatment?)

- (How much attention do you pay to psychosocial functioning during the stage of follow-up?)

3. To what extent do you regard identifying psychosocial problems with children and parents as your task?

(Box 2a + b)

4. To what extent do you succeed in identifying psychosocial problems?

(Box 3a + b)

5. What is your strategy of identifying psychosocial problems?

(Box 4)

- (Which criteria do you employ for referral of a child with psychosocial problems to another discipline,

for psychosocial care?)

- (What do you do when you identify psychosocial problems specifically in parents, and thus not in the child?)

6. How would you define emotional support by the paediatric oncologist?

(Box 5)

7. To what extent do you regard providing emotional support (by the paediatric oncologist)

to a child as your task?

(Box 6a + b)

Note: The fi ndings with respect to the questions in parentheses are not reported in this paper.

59

Results

ParticipantsThe study group consisted of 24 paediatric oncologists from four paediatric oncology centres. Thirteen (54.2%) of the participants were male, their mean age was 43.83 years (SD 7.34), and they had been paediatric oncologists for an average of 8.96 years (SD 7.99). The average duration of the interviews was 26 minutes (range 12 – 58 minutes).

Psychosocial functioning and psychosocial problems

Defi nition of psychosocial functioning

Paediatric oncologists defi ned psychosocial functioning as emotional, social and school functioning. Emotional functioning was referred to as psychological or emotional well-being, and the way of coping with (the consequences of ) disease and treatment. Aspects of social functioning included the family situation, contact with friends and peers, and participation in clubs and sports. School functioning was primarily addressed with school presence and global school performance. Some paediatric oncologists also defi ned psychosocial functioning as normal or good psychosocial functioning when the child was able to ‘live a normal life, despite the disease’. Quotes of paediatric oncologists regarding these fi ndings are reported in Box 1.

Perceived task in discussing psychosocial functioning and in identifying psychosocial

problems

Paediatric oncologists agreed that it was their task to discuss psychosocial functioning and to identify psychosocial problems. They provided various arguments for this. Many paediatric oncologists indicated to have a holistic view on medicine; the disease must be treated within the context of the child and its environment, also because psychosocial problems may interfere with treatment. Others thought that this task was an inherent duty to being a paediatrician; exploring psychosocial functioning is a part of the job. Being the child’s primary contact or coordinator during and after treatment was another argument they gave. The responsibility of being the primary contact or coordinator includes exploring functioning and identifying problems so that they could refer their patient to specialised psychosocial care (e.g., psychologist, social worker) (Box 2a).

Perceived task in discussing psychosocial functioning and in identifying psychosocial

problems - limitations

The majority of the paediatric oncologists we interviewed also believed that their task to discuss psychosocial functioning and identify psychosocial problems was limited. The most frequently mentioned limitation was their lack of expertise in dealing with psychosocial functioning and problems. Some of the respondents reported that they do not solve or extensively discuss certain problems and that they employ a low threshold in referring to

chapter 4

60

specialised psychosocial care. Furthermore, a number of paediatric oncologists indicated that the time they have available to pay attention to psychosocial functioning is limited, also due to work pressure. Additionally, some respondents reported that it was not exclusively their responsibility to identify psychosocial problems, but the responsibility of the entire medical team involved with the child (Box 2b).

Practice experience: identifi cation of psychosocial problems

The majority of paediatric oncologists think that they are mostly capable of successfully identifying psychosocial problems because they are familiar with the children and the children’s families, and because the medical team is aware of psychosocial diffi culties(Box 3a).

Practice experience: identifi cation of psychosocial problems - limitations

Some paediatric oncologists pointed out that their ability to identify psychosocial problems successfully is dependent on the amount and type of information provided by the children or parents. Children and parents do not always want to disclose problems or

Emotional, social and school functioning “It is about good and broad functioning in the family, at school, at clubs. That what a normal child would do to lead a normal life within its own social context” (24)

“I think it refers to total well-being, to the total emotional experience of everything actually, of how you are in life” (1)

“The way the child copes with the disease, the way the child functions in the family, how he is doing at school” (8)

“How the child deals with its environment, thus family as well as everything that surrounds it, such as friends” (19)

Living a normal life“That the child is still comparable to peers without the disease, despite the cancer diagnosis” (2)

“That the child feels well, that it resumes normal things in life. And continues being a child and slowly returns to the normal setting. To school, playing with friends and does not feel patient anymore” (7)

“Yes, the way the child is, the way it would be without the disease. How it would be when playing and learning in its environment” (17)

Box 1. Defi nition of psychosocial functioning

Note: For each quote, the respondent number of the paediatric oncologist that made the quote is in parentheses.

61

chapter 4

Having a holistic view“The child is of course not only a cell, a piece of tumour with a child around it” (17)

“It is integrated actually; you treat a child and you are treating the disease. But you are also treating the person and the parents” (12)

“Well, yes, with a beautiful word: holistic medicine” (4)

“Giving chemotherapy is simple. You need to get a bigger bond. It is about day-to-day contact. Building trust” (16)

Interference with treatment“I think it is a necessity for treating the child, to make sure that he or she feels well. To be able to stick to treatment at all” (14)

“Yes, absolutely. It interferes with treatment and guidance” (7)

“If the child dysfunctions psychosocially, it infl uences compliance with therapy and cooperation” (10)

Inherent to being a paediatrician“I fi nd identifying psychosocial problems an important part of our profession, especially in paediatric oncology” (21)

“It is simply inherent to being a paediatrician, because a paediatrician deals with growth and development. This is a primary aim. It is a characteristic of paediatrics. Identifying psychosocial problems belongs to that” (23)

“That is why you become a paediatric oncologist, I think” (3)

Being the primary contact or coordinator“Well, because I am the primary person in charge of treating the child, and I think it is therefore important that I am responsible for the well-being of the child” (13)

“They are not seeing anyone else. Normally, you are their only contact” (18)

“Yes, because the paediatric oncologist is the coordinator of health care” (21)

Referring to specialised care“If I notice there are serious problems, then I will refer quickly” (20)

“You have to know when to refer” (23)

“Because I am the one who has to refer, if I think extra care is needed” (6)

Box 2a. Perceived task in discussing psychosocial functioning and in identifying

psychosocial problems

62

Box 2b. Perceived limitations in task discussing psychosocial functioning

and identifying psychosocial problems

Lack of expertise“But I think you have to be aware of the fact that you only have very limited abilities to do that” (23)

“Yes, but we are less trained and schooled in this. We do it on own experience really” (1)

“Because we do not have a psychosocial background, I think we might overlook certain psychosocial problems” (22)

Not solving psychosocial problems“But solving those problems is often not within my expertise” (18)

“I do not regard it as my task to solve problems; I refer them to a psychologist or social worker” (14)

Not having detailed discussions“It concerns restricted questions; I will not extensively explore the child’s mood” (9)

“I would not question very deeply” (24)

“But not necessarily into detail. If there are any problems then I would like to know what those problems are, how serious they are and the impact it has on child and family. But solving those problems is not my expertise, so to what extent will you discuss?” (18)

Low threshold referral specialised care“There is a weekly signal to the psychologist with the request to telephone certain people and ask them if they need any support” (15)

“When I get the feeling that I don’t have the expertise, I try to pass it on” (18)

Limited time and work pressure“However, sometimes, because of lack of time, discussing psychosocial functioning gets an inferior role, maybe” (12)

“And then of course the work pressure at the out-patient clinic plays a part in this” (7)

Responsibility of whole medical team“…as part of the team, I think. So I am one of the people of the medical team who can identify psychosocial problems” (12)

“Yes, but it is not only my task. I think that other people also have to identify” (9)

63

chapter 4

they may conceal information, which could lead to an underestimation of psychosocial problems. Furthermore, a lack of expertise and time were mentioned as limitations in identifying psychosocial problems successfully. In addition, psychosocial functioning is usually not systematically or structurally discussed during consultation (Box 3b).

Strategies for identifying psychosocial problems

Many paediatric oncologists questioned children and parents to identify psychosocial problems. If they suspected that certain problems existed, they explicitly addressed these issues to learn more about the situation. In addition, some paediatric oncologists explored the child’s daily activities or observed nonverbal signs and the parent-child interaction during the consultation (Box 4).

Emotional support

Defi nition of emotional support

Emotional support provided by the paediatric oncologist was often defi ned as showing empathy, understanding and interest for the child and the family’s situation and feelings (Box 5).

Perceived task in providing emotional support

Nearly all of the paediatric oncologists reported that it was their task to provide emotional support for the children and their families. Many respondents achieved this emotional support by showing interest and talking with the child and family. Some respondents provided emotional support by performing ones job well, giving information, referring to psychosocial care or being straightforward and clear about the prognosis and treatment to diminish uncertainty. Moreover, some paediatric oncologists indicated that it was important to be accessible, to maintain a positive attitude towards the child and family, and to off er physical comfort if appropriate (e.g., a hug when the child is sad) (Box 6a).

Box 3a. Practice experience: successful identifi cation of psychosocial problems

Familiarity with the child and family“Yes, usually, because you know them very well” (23)

“I think so, because the contact we have with the child and family is very intense. I have the impression that I see what is going on in families” (2)

Having an alert medical team“Yes, often. I think that we, the medical team, are quite on top of that” (22)

“I think that we often succeed, mainly because of the team. Someone will point the problem out to you” (21)

64

Box 3b. Practice experience: unsuccessful identifi cation of psychosocial problems

Dependent on information“So I am very dependent on what I hear from parents or nursing. Nurses sometimes know diff erent sides of the story than the doctor” (10)

“What I have noticed quite a few times is that the signals are sometimes relatively weak, or that it does not come through very well. In those cases you identify the problems late or too late” (21)

Undisclosed problems or concealed information“Because not all children and parents are explicit if something is going on” (8)

“I also think, that sometimes parents believe ‘well, this is not something to discuss with the doctor’” (4)

Underestimated problems“‘Oh well, let’s hold on for a while and then it will be okay…’ which is often true. But if it is not, you sometimes think maybe we have been too hopeful that it would resolve by itself” (11)

“It is possible that I will underestimate how big a problem is” (20)

Lack of expertise“My knowledge is too limited” (13)

“Maybe I have not fully learned how to identify problems” (4)

Lack of time“Not always I think, since there is only limited time for each patient” (14)

“Sometimes you are completely focussed on a certain thing, and then you will skip it. There is no time and no energy left for other things” (19)

Lack of structural discussion“What I would want is more structure. Because now I make it up myself…There are no standard questions I ask” (13)

“No, especially if you do not structurally explore it” (19)

Perceived task in providing emotional support – limitations

Although paediatric oncologists generally acknowledged that providing emotional support is one of their tasks, they also stated that there are some limitations in doing this. Several respondents declared that they keep a certain amount of distance from the child and family’s emotions; crying along will not help their patients. In addition, some paediatric oncologists indicated that off ering emotional support depends on the personality and age of the child, and on the personality of the doctor (Box 6b).

65

chapter 4

Box 4. Strategies for identifying psychosocial problems

Discussing, questioning and addressing problems“That is simply asking questions, plainly exploring” (5)

“You need to dare to discuss and you also need to dare to tell it like it is” (1)

“If I hear things I step right up to it and call the parents with me, like a paterfamilias” (21)

“I address what I see” (22)

Exploring daily activities“I ask additional questions, like what do you do all day?” (19)

“To get an impression of the child’s day” (7)

Observing nonverbal signs “I think like an intake or interview, but I also fi nd non-verbal signs very important” (6)

“Or just by looking. If children are gloomy, you can obvious see this very well” (15)

Observing parent-child interactions“Use your eyes, use your ears and simply look what is happening. A child in interaction with the parent” (17)

“What, I think, is very important is the interaction in your consultation room” (2)

Box 5. Defi nition of emotional support

Showing empathy“Showing empathy is most important, I think” (23)

“Being a doctor that is empathic and thinks along” (18)

“Showing some empathy” (4)

Showing understanding“In any case, show that you understand what people are going through” (21)

“That you do not only understand the kind of disease the child suff ers from, but also the impact it has” (11)

Showing interest“Emotional support by the doctor is a bit of empathy, a bit of interest and identifying problems” (13)

“Daily attention. Showing that your attention is sincere” (20)

66

Box 6a. Perceived task in providing emotional support

Showing interest and talking to the children and their families“To keep on communicating, to be aware of each others’ thoughts and feelings” (18)

“By simply exploring what happens per individual and how things are going in the family. So I think that you provide emotional support by talking about it” (21)

Performing your job well“First of all, performing your work well” (9)

“This means providing good substantive work” (18)

Giving information“I think that you can play a part in it by providing patients with information about the disease” (14)

“I think, taking enough time to explain things” (12)

Referring to psychosocial care“There are a number of things you can do yourself, but if you can not, you say ‘I can not do this anymore, I need someone else to do this’” (11)

“And when necessary, if you think it is not suffi cient, you call in the help of other professionals” (21)

Being straightforward“The most important supporting role of a doctor is that he is honest, I think. That he tells it straightforward and performs his work well” (24)

“First of all, you have to be very clear to people in your medical message. I believe when you are clear you can take away a lot of uncertainty. This gives people a lot more rest” (5)

Being available“There is a point of controversy between my colleagues and me, because I always give patients my email address. I fi nd that a very pleasant way of communication” (13)

“Being available in case of emergency; you say ‘if things do not work, then please call me’” (12)

Maintaining a positive attitude“For me the most important thing is to give a child hope… Not always being dramatic or pessimistic about the disease. It must be bearable and there must be bright spots” (7)

“… basically radiate something positive about chances of recovery” (14)

Off ering physical comfort“And then we put our arms around each other and said ‘it is all so shit’” (17)

“It goes as far as putting a hand on someone’s shoulder. Just being physical for a minute” (23)

67

chapter 4

Box 6b. Perceived limitations in task providing emotional support

Keeping ones distance“Let me put it this way: without crying along with the family. I do not think that will help them” (24)

“By showing that you understand their grief, without crying along” (20)

Personality and age of the child“I think providing emotional support depends very much on the patient” (1)

“That is something which is very age-dependent” (8)

Personality of the doctor“I believe that everyone acknowledges that every doctor has a personal style. We respect each others’ style” (20)

“I see great diff erences within our group when it comes to providing emotional support. I would not dare to say which way is best” (18)

DiscussionThis study indicates that paediatric oncologists believe that it is their task to discuss psychosocial functioning, to identify psychosocial problems and to provide emotional support to children with cancer. However, paediatric oncologists also mentioned that fulfi lling this task can be limited due to e.g. lack of expertise or time.

In general, our fi ndings agree with previous research that has demonstrated how paediatricians consider themselves responsible in identifying mental health issues [4]. Paediatric oncologists’ perceptions towards this responsibility are shown in Box 2a (it is ‘inherent to being a paediatrician’ that you are ‘the primary contact’). Our fi ndings are also in line with several key components of doctor-patient interactions involving children and their parents identifi ed by Crossley and Davies [15]. For instance the key component of patient-parent-centeredness regarding sensitivity to unresolved issues and problem identifi cation can be found in Box 2a, 3a and 4 (‘discussing, questioning and addressing problems’ and having an ‘alert medical team’). The key component of interpersonal skills, such as sensitivity to nonverbal cues, was mentioned by paediatric oncologists as well (Box 4, ‘observing non-verbal signs’). Additionally, the key components health-related problems (e.g., relief of distress), psychosocial problems and problems with cognitive diffi culty are applicable to our fi ndings in a more general sense. Some paediatric oncologists defi ned psychosocial functioning as ‘living a normal life, despite having the disease’ (Box 1). Because they also believe that it is their task to identify psychosocial problems (Box 2a), it likely that they will try to achieve an outcome in which distress and psychosocial and cognitive problems are relieved. Liaisons with relevant health-care professionals (e.g., full

68

use of the child’s health team, communication with health-care professionals) is another key component. This was mentioned by paediatric oncologists in terms of the alertness of the child’s medical team (Box 2b and 3a) and referrals to psychosocial care (Box 2a and 2b). Finally, Crossley and Davies [15] stated that making appropriate family plans (e.g., addressing the patient’s needs and problems) is important in paediatric doctor-patient interactions as well; paediatric oncologists also regarded this as their task (Box 4, ‘discussing, questioning and addressing problems’).

With respect to emotional support our study illustrates that paediatric oncologists’ perceptions match the preferences of their patients who indicated to need support and empathy (Box 5 and 6a; ‘showing empathy’ and ‘being available’), and suffi cient time for communication (Box 6a; ‘giving information’, second quote) [21]. Additionally, our fi ndings regarding emotional support agree with some of the key components described by Crossley and Davies [15]. For instance, interpersonal skills ‘empathy and support’ can be found in Box 5 and 6a, and general communications skills ‘listening, clarity of communication and information-giving’ are presented in Box 6a (‘giving information’ and ‘being straightforward’). Finally, the key component making appropriate family plans, e.g. providing psychological support is refl ected in Box 5 and 6a (‘referring to psychosocial care’).

Overall, paediatric oncologists’ views towards their task in psychosocial functioning and emotional support met many of the important components of doctor-patient interactions involving children [15]. However, paediatric oncologists also clearly indicated, that this is a task within boundaries. For instance, some of them do not feel responsible or qualifi ed for solving psychosocial problems (Box 2b), which is in agreement with other research [22]. With respect to providing emotional support, some remain distanced from their patients and their patient’s family (Box 6b). Potentially confounding eff ects on doctor-patient interactions, such as the doctor’s level of training and workload and patient’s age were identifi ed by Crossley and Davies [15]. These limitations were also mentioned by our respondents (Box 2b, 3b and 6b). Other reasons for unsuccessful identifi cation of psychosocial problems were dependency on information, undisclosed problems, underestimated problems and lack of structural discussion of problems (Box 3b). Since paediatric oncologists felt responsible for discussing and identifying psychosocial problems, it is important to fi nd ways to overcome these limitations. In the light of the QLIC-ON study, we therefore suggest considering the use of psychosocial PROs during the consultation as a possible solution. PROs are answers to questionnaires based on direct reporting by patients without the intervention of an observer. PROs can serve as an aid for the physician to systematically identify and discuss psychosocial problems. In a very recent study [17] we demonstrated that the use of PROs in paediatric oncology practice increased the discussion of emotional and psychosocial functioning and enhanced the identifi cation of emotional and cognitive problems, without lengthening the consultation duration.

69

chapter 4

Strength of the current study is that a substantial amount (44%) of the total population of paediatric oncologists working in the Netherlands participated in our study. Therefore, our fi ndings are likely to be representative of all paediatric oncologists working in the Netherlands. Because the organisation and quality of paediatric oncology health care systems can vary, however, our fi ndings may not be representative of other countries.

Our study was limited in that the paediatric oncologists knew that the interviewer was employed at the Psychosocial Department of the Academic Medical Centre/ Emma Children’s Hospital, and they were aware of the purpose of the QLIC-ON study. These circumstances may have biased some of the respondent’s answers (social desirability). Furthermore, it is a study limitation that the paediatric oncologists had no opportunity to refuse participation because the interviews were part of the QLIC-ON study, in which the paediatric oncologists were enrolled. This may have caused bias in our fi ndings as well. However, because the responses were varied and revealed their perceived limitations, we believe that the paediatric oncologists in this study remained sincere and professional and thus provided us with valid information.

We recommend future studies to investigate whether paediatric oncologists are truly successful in discussing psychosocial functioning, identifying psychosocial problems and providing emotional support. This evaluation could be accomplished by comparing paediatric oncologists’ perceptions with the perceptions of the parents and children. Furthermore, it is important to explore the paediatric oncologists’ view regarding their responsibility in parental psychosocial functioning, since parents of children with cancer can also experience psychosocial impairment [23, 24]. Psychosocial problems of parents were included in our interview schedule; however, analysing these data was beyond the scope of this paper.

We conclude that paediatric oncologists are aware of the psychosocial consequences of childhood cancer and that they provide psychosocial care to the best of abilities.

70

Appendix 1

Defi nition of psychosocial functioning

Summary psychosocial functioning

Attitude of parent/ child towards psychosocial care 1

Behaviour 2

Being able to communicate 1

Clubs/ societies/ sports/ hobbies 7

Coping with consequences disease/ treatment 8

Family/ home/ parents/ brothers/ sisters 26

Friends/ peers/ boy-girlfriend 18

Independent functioning 1

Individuality/ being able to be themself 3

Interaction with medical team 1

Normal life 9

Psychological/ emotional well-being 22

Quality of life 1

School/ daycare 23

Self perception 1

Sleeping 1

Support 2

Task to discuss/ identify psychosocial functioning/ problems

Yes 24

Children should live a normal live 3

Fun/ interesting 2

Holistic view on treating children 15

Inherent to being paediatrician 7

Interference with physical functioning/ compliance with therapy 8

Parents/ children disclose problems 3

Parents/ children don’t see problems 2

Preventing delays in development 1

Preventing parental problems (work, fi nances) 1

Primary contact/ coordinator/ coach 9

Responsible for identifying problems 11

Responsible for exploring problems 8

Responsible for providing psychosocial care by paediatrician 6

Responsible for referral to psychosocial care 19

Psychosocial side eff ects of treatment 2

However, limited by

Frequent contact 1

Frequency

71

Limited contact 1

Restricted discussion/ no details 6

Restricted responsibility/ expertise/ training 16

Restricted time/ work pressure 6

Medical team also responsible for identifying problems 5

Not responsible for solving problems 2

Parents/ children don’t disclose problems 2

Type of problem (only disease related) 1

Successfully identify psychosocial problems

Yes/ mostly 13

Medical team is alert 3

Personality paediatrician 1

Well acquainted with parent/ child 2

Not always/ sometimes 10

Blind spot/ not alert enough 2

Depending on info from parents/ child/ medical team 5

Diff erence of opinion in medical team: what is a problem? 1

It takes time to successfully identify 2

Restricted expertise/ training/ knowledge 3

Restricted time 3

No structural discussion 4

Parents/ children conceal problems/ refuse help 7

Underestimation of problems 4

Hopefully 1

Dependent on consultation frequency 1

Strategy of identifying psychosocial problems

Acknowledging problems 1

Being open/ interested 1

Cooperation with medical team 3

Depending on age and type of child/ parents 3

Exploring activities/ daily functioning 6

Exploring body image 1

Exploring crucial moments 2

Intuition 1

Listening 2

Observing 4

Questioning/ communicating/ discussing 23

Standard checklist 2

Frequency

chapter 4

72

Defi nition of emotional support

Summary emotional support

Showing empathy/ interest/ understanding/ acknowledging emotions 24

Task to provide emotional support Yes 23

Asking about things not related to disease 1

Being available 4

Being honest/ clear 5

Being objective 1

Bonding with child/ parents 1

Getting through treatment 1

Identifying problems 2

Listening 3

Making sure the child can be him/ herself 2

Part of being a paediatrician 1

Performing work well 6

Physical comfort 4

Positive/ optimistic attitude 5

Providing info (to prevent uncertainty) 5

Questioning/ communicating/ discussing 10

Referral 4

Regular contact/ continuity of care 3

Relaxing consultation/ joking 3

Showing empathy/ interest/ understanding/ acknowledging emotions 24

Taking action 2

Working together (parent/ child/ doctor) 2

However, limited by

Age or type of child 3

Restricted expertise/ skills/ training 1

Not crying along/ keeping distance/ self protection 7

Personality of doctor 5

Time 2

Restricted discussion 2

No 1

Limited time 1

Frequency

73

Reference List1. Maurice-Stam H, Grootenhuis MA, Caron HN,

Last BF. Course of life of survivors of childhood cancer is related to quality of life in young adulthood. J Psychosoc Oncol 2007:25:43-58.



4. Stein RE, Horwitz SM, Storfer-Isser A, et al. Do pediatricians think they are responsible for identifi cation and management of child mental health problems? Results of the AAP periodic survey. Ambul Pediatr 2008:8:11-17.


6. Fallowfi eld L, Ratcliff e D, Jenkins V, Saul J. Psychiatric morbidity and its recognition by doctors in patients with cancer. Br J Cancer 2001:84:1011-1015.

7. Bultz BD, Carlson LE. Emotional distress: the sixth vital sign in cancer care. J Clin Oncol 2005:23:6440-6441.

8. Sollner W, DeVries A, Steixner E, et al. How successful are oncologists in identifying patient distress, perceived social support, and need for psychosocial counselling? Br J Cancer 2001:84:179-185.

9. Ford S, Fallowfi eld L, Lewis S. Can oncologists detect distress in their out-patients and how satisfi ed are they with their performance during bad news consultations? Br J Cancer 1994:70:767-770.

10. Fogarty LA, Curbow BA, Wingard JR, et al. Can 40 seconds of compassion reduce patient anxiety? J Clin Oncol 1999:17:371-379.

11. Zachariae R, Pedersen CG, Jensen AB, et al. Association of perceived physician communication style with patient satisfaction, distress, cancer-related self-effi cacy, and perceived control over the disease. Br J Cancer 2003:88:658-665.

12. Walker MS, Ristvedt SL, Haughey BH. Patient care in multidisciplinary cancer clinics: does attention to psychosocial needs predict patient satisfaction? Psychooncology 2003:12:291-300.

13. Butow PN, Brown RF, Cogar S, et al. Oncologists’ reactions to cancer patients’ verbal cues. Psychooncology 2002:11:47-58.

14. Ford S, Fallowfi eld L, Lewis S. Doctor-patient interactions in oncology. Soc Sci Med 1996:42:1511-1519.

15. Crossley J, Davies H. Doctors’ consultations with children and their parents: a model of competencies, outcomes and confounding infl uences. Med Educ 2005:39:807-819.

16. Engelen V, Haverman L, Koopman H, et al. Development and implementation of a patient reported outcome intervention (QLIC-ON PROfi le) in clinical paediatric oncology practice. Patient Educ Couns 2010:81:235-244.

17. Engelen V, Detmar S, Koopman H, et al. Reporting health-related quality of life scores to physicians during routine follow-up visits of pediatric oncology patients: Is it eff ective? Pediatr Blood Cancer 2011.


19. Pope C, Ziebland S, Mays N. Qualitative research in health care. Analysing qualitative data. BMJ 2000:320:114-116.

20. Boeije H. Analysis in Qualitative Research. Sage Publications (CA); 2009.

21. Zwaanswijk M, Tates K, van DS, et al. Young patients’, parents’, and survivors’ communication preferences in paediatric oncology: results of online focus groups. BMC Pediatr 2007:7:35.

chapter 4

74

22. Detmar SB, Aaronson NK, Wever LD, et al. How are you feeling? Who wants to know? Patients’ and oncologists’ preferences for discussing health-related quality-of-life issues. J Clin Oncol 2000:18:3295-3301.

23. Kazak AE, Boeving CA, Alderfer MA, et al. Posttraumatic stress symptoms during treatment in parents of children with cancer. J Clin Oncol 2005:23:7405-7410.

24. Engelen V, Koopman HM, Detmar SB, et al. Health-related quality of life after completion of successful treatment for childhood cancer. Pediatr Blood Cancer 2011:56:646-653.

755Chapter 5

V. Engelen 1L. Haverman 1H.M. Koopman 2

A.Y.N. Schouten – van Meeteren 3 E. Meijer – van den Bergh 4 J. Vrijmoet-Wiersma 5

E.M. van Dijk 6B.F. Last 1S.B. Detmar 7 M.A. Grootenhuis 1

1 Academic Medical Centre/ Emma Children’s Hospital, Paediatric Psychosocial Department, Amsterdam 2 Leiden University Medical Centre, Medical Psychology, Leiden3 Academic Medical Centre/ Emma Children’s Hospital, Paediatric Oncology Department, Amsterdam4 Radboud University Medical Centre, Medical Psychology, Paediatric Oncology Department, Nijmegen5 Leiden University Medical Centre, Medical Psychology, Paediatric Oncology Department, Leiden 6 VU University Medical Center, Medical Psychology, Paediatric Oncology Department, Amsterdam 7 TNO Prevention and Health, Leiden, The Netherlands.

Development and implementation of a patient reported outcome intervention (QLIC-ON PROfi le) in clinical paediatric oncology practice

PATIENT EDUC COUNS. 2010 NOV;81(2):235-44.

76

Abstract

Objective. The use of patient reported outcomes (PRO) in routine clinical practice is becoming increasingly common, but there is limited knowledge about the development and implementation of PRO. The objective of the current paper is to provide a thorough description of the development and implementation of a PRO on health related quality of life (HRQOL) – the QLIC-ON PROfi le - in clinical paediatric oncology practice.

Methods. The development of the QLIC-ON PROfi le is explained by elucidating important choices: the HRQOL instrument, the professional that uses the QLIC-ON PROfi le, the optimal form of HRQOL feedback and whether or not a clinically important diff erence is reported. The description of the implementation of the QLIC-ON PROfi le focuses on the education and commitment of the professional that uses the QLIC-ON PROfi le. Study design and outcome measures are also elaborated on.

Results. Important considerations regarding the development and implementation of PRO interventions are reported. These considerations have also resulted in educational material.

Conclusion. Our study adds to current knowledge of PRO research. This paper can be used as a practical guide for researchers and other professionals, who are interested in setting up PRO interventions in any clinical setting.

77

1. Background

The use of patient reported outcomes (PRO) in routine clinical practice is becoming increasingly common, but there is limited knowledge about the development and implementation of PRO. PRO are based on direct reporting by patients without the intervention of an observer. They include the self-assessment of functional status, symptoms, and other concerns such as patient needs and satisfaction with care. Health related quality of life (HRQOL) assessment is a form of PRO that often includes both a patient’s functional status (physical and psychosocial) as well as his or her symptoms [1]. Over the past decade, several studies have focused on the use of PRO in clinical practice [2-7]. Adult psycho-oncology research shows that PRO facilitates discussion of HRQOL issues during medical consultations [2], which can improve the HRQOL of patients [3]. In paediatrics, the impact of PRO has been examined to a limited extent [8]. Recently, however, encouraging results have been found by De Wit et al. [7]: periodic monitoring and discussion of HRQOL in adolescents with diabetes had positive eff ects on their psychosocial well-being and improved satisfaction with care.

Despite these promising fi ndings, it remains diffi cult to prove the eff ectiveness of PRO in clinical practice [9,10]. For this reason Greenhalgh et al. [11] applied a theory-driven approach to the use of PRO to bring together evidence of both whether and how the intervention is eff ective. The authors conclude that (1) the applied instrument needs to prioritize the views of the individual in order to adequately refl ect the individual’s HRQOL; (2) health care providers other than physicians may also fi nd HRQOL information useful; (3) feedback of HRQOL information should be longitudinally measured and presented over a period of time; (4) a clinically important diff erence or change in HRQOL data does not always resemble physicians’ perceptions of this diff erence or change; and fi nally, (5) commitment and education of the PRO user are needed to address the multiple barriers to the use of PRO in clinical practice. Bearing these fi ve conclusions in mind, there is a shortage of literature addressing the development and implementation of a PRO intervention in clinical practice.

In 2005, we initiated the multicenter sequential cohort QLIC-ON study (Quality of Life In Childhood Oncology) to measure the eff ect of the use of PRO in clinical paediatric oncology practice. Participants are children with cancer in the period immediately after end of treatment, which is a phase when HRQOL can be aff ected [12]. Shortly before the fi rst three follow-up consultations with the paediatric oncologist, the child (8 to 18 years of age) or parent (if the child is 0 to 8 years of age) completes a digital HRQOL questionnaire. The outcome (QLIC-ON PROfi le, p. 173) is presented to the paediatric oncologist as a PRO in order to help identify and discuss possible HRQOL problems.

The aim of the current paper is to provide a thorough description of the development and implementation of the QLIC-ON PROfi le based on the fi ve conclusions of Greenhalgh et al. [11] (presented in italics). We also elaborate on the study design and outcome measures. Through this paper we hope to contribute to current knowledge within PRO

chapter 5

78

research to eventually increase the eff ectiveness of PRO use in clinical (paediatric) practice in general.

2. Development of intervention

2.1. HRQOL instrumentThe applied instrument needs to prioritize the views of the individual in order to adequately refl ect the individual’s HRQOL [11]. Possibilities for applying this recommendation in paediatrics are limited due to the cognitive inability of younger children. Such an instrument (e.g., the SEIQoL-DW [13] for adults) is not available for children. Moreover, individualized open instruments are probably more diffi cult to interpret and time consuming than structured questionnaires and therefore not user-friendly. One should carefully consider whether the type of instrument not only matches the patient but also the type of user of the PRO. Selection criteria for an appropriate HRQOL instrument for participants of the QLIC-ON study were therefore the following: generic and structured nature, broad age range, short completion time, good psychometric properties and availability in the Dutch language. Disease specifi c instruments were not applicable, since children are off treatment and cancer modules often refer to treatment related symptoms, e.g. nausea.

The TAPQOL [14], TACQOL [15], TAAQOL [16], DUX-25 [17], Kidscreen [18], PedsQL [19] and CHQ [20] were available in Dutch and compared with each other. The PedsQL (past week - acute - version) appeared to be most appropriate because of its broad age range (5-18 years), its inclusion of self-report as well as proxy-report, its short completion time (approximately 5-10 minutes) and its good feasibility, validity and reliability [19]. The PedsQL self-report form was used for children aged 8 to 12 and 13 to 18, while parents of children aged 6 to 7 completed the PedsQL proxy-report. The 23 items are divided into four subscales: physical, emotional, social and school functioning. Since Dutch PedsQL norm scores were not available, data were collected in a previous study [21].

The TAPQOL was added for parents of children aged 0 to 5 years. This HRQOL instrument focuses on children from 0 to 5 years of age through parent proxy measurement. Completion time is about 5-10 minutes and psychometric properties are satisfying [14,22]. Forty-three items assess the child’s functioning on 12 scales: stomach, skin, lungs, sleeping, appetite, motor functioning (=physical functioning), positive mood, anxiety, liveliness (=emotional functioning), problem behaviour, social functioning (=social functioning) and communication (=communication). Dutch TAPQOL norm scores were available.

2.2. User Greenhalgh et al. [11] stated that health care providers other than physicians may also fi nd HRQOL information useful. The QLIC-ON PROfi le was presented to paediatric oncologists,

79

since they are the child’s primary and, most often, only contact in the phase after treatment. This makes it diffi cult to appoint other PRO users. The paediatric oncologist’s responsibility is therefore not limited to physical functioning; systematic monitoring of psychosocial functioning is essential as well. This is especially important for those families that experience this period as a tense and stressful time, because fear of relapse is present [12].

2.3. HRQOL feedback Feedback of HRQOL information should be longitudinally measured and presented over a period of time. Single assessment is not considered enough; HRQOL should mirror the decision making process of advice and referral itself [11]. In the QLIC-ON study HRQOL outcome of each child was provided to the paediatric oncologist during three consecutive consultations; previous results were also depicted on the QLIC-ON PROfi le to make monitoring possible. The complete development of the QLIC-ON PROfi le is described in the following paragraphs.

2.3.1. Computer program A computer programmer converted the TAPQOL and PedsQL into computerized questionnaires (Figure 1). After completion, individual outcomes (including previous ones) were digitally presented on the QLIC-ON PROfi le. Features to print as well as to back-up and export the HRQOL data to SPSS were also made available. The program was implemented on four laptops, one for each of the four participating centres.

2.3.2. QLIC-ON PROfi le

Inspired by previous studies [2,3] four QLIC-ON PROfi le layout options were designed: 1) line graphs, 2) vertical bars, 3) horizontal bars, and 4) a literal reproduction of the items with reported answers in words. The designs were presented to six paediatric oncologists from the diff erent centres by e-mail and subsequently discussed during a half hour group session with the researchers.

The literal reproduction of the items with reported answers in words was unanimously chosen as the preferred design, since it provided the most detailed individual information. Moreover, this design leads the paediatric oncologist’s attention directly towards the problem by using traffi c light colours to accentuate the direction of the answers. According to the paediatric oncologists, the fi rst three options might prolong consultation time, whereas design 4 was the least likely to do so. The second preferred design, the line graphs, was included in the QLIC-ON PROfi le as well, because it adds longitudinal HRQOL information and the possibility to compare the scores of the child to the healthy norm population.

The fi nal design of the QLIC-ON PROfi le is shown on p. 173 (PedsQL version). The left top provides information regarding the child: name, date of birth, person that completed

chapter 5

80

the questionnaire and period of time. The main part represents the four HRQOL domains, each marked with a coloured line for easy recognition: physical (blue), emotional (green), social (yellow) and school functioning (pink). The items are depicted for every domain, with the answers summarized in a column behind linked to the completion (=consultation) date. An answer accentuated green (‘never’, ‘almost never’) indicates that the child had no problems regarding the subject, orange (‘sometimes’) points out that there were some problems, and a red answer (‘often’, ‘almost always’) illustrates that a child frequently experienced problems. Most recent answers are printed bold to focus the attention of the paediatric oncologist. At the bottom, scale scores are refl ected in line graphs presented with the corresponding colours of each domain. In every line graph the scale score of the child is indicated with the completion date. Scores can be compared with previous scale scores as well as with the scale score of the healthy norm population (red dotted line). The layout of the TAPQOL QLIC-ON PROfi le is essentially the same as the PedsQL.

2.4. Clinically important diff erenceA clinically important diff erence or change in HRQOL data does not always resemble physicians’

Figure 1. Screen lay-out of PedsQL

About my feelings

In the past ONE week, how much of a problem has this been for you...

(problems with...)

I feel afraid or scared

Previous

Next

Never Almost Never

Sometimes Often Almost always

(0) (1) (2) (3) (4)

QUESTIONNAIRE for CHILDREN (age 8-12) © Copyright 1998 JW Vari, PhD. Dutch Translation J.M. Koot & D. Bastiaansen

81

perceptions of this diff erence or change [11]. It is important to carefully consider whether a calculated clinically important diff erence (e.g. cut-off points) adds value to the PRO. This depends on factors such as study aims and PRO user. On the QLIC-ON PROfi le, healthy norm scores for each HRQOL domain are reported (red dotted line) in the graphs. These scores cannot to be used as cut-off points, since PedsQL and TAPQOL were not designed as screening instruments. The healthy norm scores are merely intended as markers to give the paediatric oncologist a sense of the child’s functioning compared to healthy peers and how the child is resuming normal life. For the purpose of systematically monitoring the child’s functioning, paediatric oncologists are instructed to pay attention to results beneath the healthy norm score. The main focus of the paediatric oncologists, however, should be on the items, where individual diff erences over time can easily be detected and discussed with help of the traffi c light system. These changes, though, are not necessarily of a clinically important nature. Thus, if an answer is accentuated red on the fi rst consultation, orange on the second and green when completing the questionnaire for the third time, it does not necessarily imply a clinically important change. The importance of this change can only be determined by the paediatric oncologist after discussing this (diff erence in) score with the child and parent. This is exactly what the QLIC-ON PROfi le is intended to do.

3. Implementing intervention

3.1. Commitment and educationCommitment and education of the PRO user are needed to address the multiple barriers to the use of PRO in clinical practice [11]. The QLIC-ON study observed this recommendation by applying a paediatric oncologist training program (QLIC-ON education) developed by fi ve researchers, four clinical psychologists and one paediatric oncologist. The QLIC-ON education was described in a trainer manual and aimed at enhancing the eff ect of the use of the QLIC-ON PROfi le in clinical practice [23,24]. Goals of the QLIC-ON education included helping the paediatric oncologists to obtain knowledge about the background of the QLIC-ON PROfi le, fostering awareness of the importance of the use of PRO in paediatric oncology and, most importantly, achieving competence in the use of the QLIC-ON PROfi le in clinical practice.

3.1.1. Commitment

Six paediatric oncologists were consulted about their preferences regarding the design of the QLIC-ON PROfi le (as stated earlier) in order to develop a sense of commitment. This also involved consulting one paediatric oncologist regarding the development of the QLIC-ON education, as well as striving for continuity during the consultations by having one paediatric oncologist monitor a child during the three QLIC-ON consultations.

chapter 5

82

3.1.2. Education

The QLIC-ON education contained three parts: an individual training for paediatric oncologists, a group training for paediatric oncologists, and an explanation for children and parents.

Individual training The individual training consisted of an interview and a QLIC-ON PROfi le instruction (duration: 1 hour). The interview explored the paediatric oncologist’s vision of his or her role in identifying and discussing HRQOL problems in children with cancer. The aim of the interview was to put the focus of the paediatric oncologist on the subject of the QLIC-ON study: identifying and discussing HRQOL problems. Subsequently, the QLIC-ON PROfi le was introduced and explained with respect to layout, content, interpretation and use. Layout and content were clarifi ed as described in paragraph 2.3.2 (third indentation). Interpretation and use were illustrated by means of a decision tree (Figure 2), which was developed as an aid for interpretation of the QLIC-ON PROfi le. The decision tree distinguishes three steps:

1) Identify: Is there a HRQOL problem? • Scan the QLIC-ON PROfi le for red and orange coloured items2) Discuss: What is the HRQOL problem exactly? • Clarify and discuss the red and orange coloured items3) Take action: What can be done about the HRQOL problem? • Depending on need for care: patient tailored advice or referral

The QLIC-ON PROfi le can only be experienced as a useful and easy tool if the users are given some liberty to employ it. Therefore, an important message for the paediatric oncologists was to implement the QLIC-ON PROfi le in their consultations in a way that was personally convenient to them and use the decision tree as a tool when needed. Some paediatric oncologists preferred to use the QLIC-ON PROfi le as a conversational guideline during the consultation, while others applied it as a supplement to their usual practice by checking the QLIC-ON PROfi le at the end of the consultation for missed issues.

The decision tree, some important user instructions, and an example of a QLIC-ON PROfi le were summarized in the QLIC-ON Pocket Card (Figure 3), which was distributed to the paediatric oncologist at the end of the individual training as a visual reminder.

83

Figure 2. Decision tree

There are no problems?

Item (or domain) is green

Item (or domain) is red or orange

There are problems?

Can you tell more about it?

When does it bother you?

How often does it bother you?

(What is the reason, do you think?)

How do you feel about it?

What do you do about it? Does it help?

Do you feel supported?

Does it worry you?

Is there a question for help? No End

Worry

Support

Action/ Solution

Feeling/ Opinion

(Reason)

Frequency

Moment/ Location

Explore

QLIC-ON PROfile

Yes, that’s right

Yes, that’s right

No, that’s not right


Yes Dependent on severity and need

Monitor

Advice

Referral

(End)

(End)

(End)

Start

QLIC-ON Support: 020 - 566 82 25, [email protected]

End

1. Id

entit

y2.

Dis

cuss

3. T

ake

actio

n

chapter 5

84

Figure 3. QLIC-ON Pocket Card

QLI

C-O

N

POCK

ET C

ARD

Rem

inde

rs

Ide

nti

fy

Scan

the

QLI

C-O

N P

ROfil

e at

the

curr

ent d

ate.

Whi

ch c

olou

r dom

inat

es?

Gre

en:

nev

er a

pro

ble

m.

Ora

nge:

som

etim

es a

pro

ble

m.

Red:

oft

en a

pro

ble

m.

In w

hich

dom

ain

do y

ou id

entif

y p

rob

lem

s?

In w

hich

item

s do

you

iden

tify

pro

ble

ms?

Is th

ere

imp

rove

men

t or w

orse

ning

com

par

ed to

pre

viou

s as

sess

men

ts?

Gra

phs

: How

is th

e ch

ild d

oing

com

par

ed h

ealth

y p

eers

?

Dis

cu

ss

Dis

cuss

the

QLI

C-O

N P

ROfil

e on

a c

onve

nien

t mom

ent d

urin

g th

e

cons

ulta

tion.

Adv

ice:

at s

tart

.

Use

the

QLI

C-O

N P

ROfil

e as

a g

uide

or s

upp

lem

ent f

or th

e

cons

ulta

tion.

Rem

emb

er: t

imef

ram

e is

the

pas

t wee

k.

Ask

the

par

ents

for c

onfir

mat

ion

freq

uent

ly.

Ope

n th

e di

scus

sion

with

e.g

.: “I s

ee y

ou fi

lled

out,

that

...”

Dis

cuss

the

red

and

oran

ge a

nsw

ers.

Use

the

deci

sion

tree

as

a gu

ide.

If th

e an

swer

s ar

e (p

ract

ical

ly) a

ll gr

een,

ask

for c

onfir

mat

ion.

If th

ere

are

a lo

t of r

ed (a

nd o

rang

e) a

nsw

ers,

disc

uss t

he c

once

rnin

g do

mai

ns o

nly

Che

ck if

chi

ld o

r par

ent w

ant t

o di

scus

s to

pic

s ot

her t

han

thos

e on

the

QLI

C-O

N P

ROfil

e

Sum

mar

ize

the

mos

t im

por

tant

issu

es

Ta

ke

ac

tio

n

Take

act

ion

if th

ere

is a

que

stio

n fo

r hel

p.

Com

e b

ack

to th

is a

ctio

n at

the

next

con

sult

atio

n.

85

Lich

amel

ijk fu

nctio

nere

n

100 80

94

60 40 20 0

85

02-0

320

07

78

16-0

420

07

88

04-0

620

07

s c o r e

Soci

aal f

unct

ione

ren

100 80

85

60 40 20 0

85

87

02-0

320

07

90

16-0

420

0704

-06

2007

s c o r e

Emot

ione

el fu

nctio

nere

n

100 80

50

60 40 20 0

77

02-0

320

07

70

16-0

420

07

65

04-0

620

07

s c o r e

Scho

olfu

nctio

nere

n

100 80

7060 40 20 0

50

79

02-0

320

07

60

16-0

420

0704

-06

2007

s c o r e

It is

har

d fo

r me

to w

alk

mor

e th

an o

ne b

lock

N

ever

N

ever

N

ever

It is

har

d fo

r me

to ru

n

Nev

er

Alm

ost n

ever

N

ever

It is

har

d fo

r me

to d

o sp

orts

act

ivit

y or

exe

rcis

e

Som

etim

es

Ofte

n

Alm

ost n

ever

It is

har

d fo

r me

to li

ft s

omet

hing

hea

vy

Nev

er

Nev

er

Nev

erIt

is h

ard

for m

e to

take

a b

ath

or s

how

er b

y m

ysel

f

Nev

er

Nev

er

Nev

erIt

is h

ard

for m

e to

do

chor

es a

roun

d th

e ho

use

N

ever

N

ever

N

ever

I hur

t or a

che

N

ever

So

met

imes

So

met

imes

I hav

e lo

w e

nerg

y

Nev

er

Alm

ost n

ever

A

lmos

t nev

er

I fee

l afr

aid

or s

care

d

Som

etim

es

Alm

ost n

ever

A

lmos

t nev

erI f

eel s

ad o

r blu

e

Som

etim

es

Alm

ost n

ever

A

lmos

t nev

erI f

eel a

ngry

Som

etim

es

Alm

ost n

ever

A

lmos

t nev

erI h

ave

trou

ble

sle

epin

g

Nev

er

Nev

er

Nev

erI w

orry

ab

out w

hat w

ill h

app

en to

me

A

lmos

t alw

ays

Ofte

n

Alm

ost a

lway

s

I hav

e tr

oub

le g

ettin

g al

ong

with

oth

er k

ids

A

lmos

t nev

er

Nev

er

Nev

erO

ther

kid

s do

not

wan

t to

be

my

frie

nd

Nev

er

Nev

er

Nev

erO

ther

kid

s te

ase

me

So

met

imes

So

met

imes

So

met

imes

I can

not d

o th

ings

that

oth

er k

ids

my

age

can

do

Nev

er

Nev

er

Alm

ost n

ever

It is

har

d to

kee

p u

p w

hen

I pla

y w

ith o

ther

kid

s

Nev

er

Nev

er

Nev

er

It is

har

d to

pay

att

entio

n in

cla

ss

Nev

er

Alm

ost N

ever

A

lmos

t nev

erI f

orge

t thi

ngs

A

lmos

t alw

ays

Ofte

n

Alm

ost a

lway

sI h

ave

trou

ble

kee

pin

g up

with

my

scho

olw

ork

N

ever

N

ever

A

lmos

t nev

erI m

iss

scho

ol b

ecau

se o

f not

feel

ing

wel

l

Alm

ost n

ever

So

met

imes

A

lmos

t nev

erI m

iss

scho

ol to

go

to d

octo

r or h

osp

ital

A

lmos

t nev

er

Som

etim

es

Ofte

n

Phy

sica

l Fun

ctio

ning

(pr

oble

ms

with

...)

02

-03-

2007

16

-04-

2007

04

-06-

2007

Emot

iona

l Fun

ctio

ning

(pro

blem

s w

ith...

)

02-0

3-20

07

16-0

4-20

07

04-0

6-20

07

Soci

al F

unct

ioni

ng (p

robl

ems

with

...)

02

-03-

2007

16

-04-

2007

04

-06-

2007

Scho

ol F

unct

ioni

ng (p

robl

ems

with

...)

02

-03-

2007

16

-04-

2007

04

-06-

2007

A h

igh

scor

e in

the

grap

h im

plie

s hi

gh q

ualit

y of

life

Ther

e ar

e no

pro

ble

ms?

Item

(or d

omai

n) is

gre

en

Item

(or d

omai

n) is

red

or o

rang

e

Ther

e ar

e p

rob

lem

s?

Can

you

tell

mor

e ab

out i

t?

Whe

n do

es it

bot

her y

ou?

How

oft

en d

oes

it b

othe

r you

?

(Wha

t is

the

reas

on, d

o yo

u th

ink?

)

How

do

you

feel

abo

ut it

?

Wha

t do

you

do a

bout

it? D

oes i

t hel

p?

Do

you

feel

supp

orte

d?

Doe

s it

wor

ry y

ou?

Is th

ere

a qu

estio

n fo

r hel

p?

No

End

Wor

ry

Sup

por

t

Act

ion/

Sol

utio

n

Feel

ing/

Op

inio

n

(Rea

son)

Freq

uenc

y

Mom

ent/

Loc

atio

n

Exp

lore

QLI

C-O

N P

ROfil

e

Yes,

that

’s ri

ght

Yes,

that

’s ri

ght

No,

that

’s n

ot ri

ght

No,

that

’s n

ot ri

ght

Yes

Dep

ende

nt o

n se

verit

y an

d ne

ed

Mon

itor

Adv

ice

Refe

rral

(End

)

(End

)

(End

)

Star

t

QLI

C-O

N S

up

po

rt: 0

20 -

566

82 2

5, v

.en

gel

en@

amc.

uva

.nl

End

1. Identity 2. Discuss 3. Take action

chapter 5

Figure 3. QLIC-ON Pocket Card (continued)

86

Group training

A group training was developed in order to establish advanced skills and in-depth use of the QLIC-ON PROfi le. A minimum of three and a maximum of fi ve paediatric oncologists participated per session. Two trainers were present: a researcher and a psychologist.

The theoretical part initially consisted of a presentation elaborating on HRQOL of children with cancer [12,25-29], PRO [1-3,8] and the QLIC-ON study. After the fi rst training sessions, however, paediatric oncologists evaluated that the theoretical part was not as useful as the practical component of the training. Subsequently, the group training was shortened to one hour, and the theoretical part was limited to a core QLIC-ON study description. A short ‘quiz’ regarding the main results of the QLIC-Oncologist Questionnaire was also implemented to enhance awareness of the paediatric oncologists and provide the training with the necessary interaction. The QLIC-Oncologist Questionnaire, derived from two previous studies [30,31], consisted of 12 propositions about the paediatric oncologist’s task regarding HRQOL problems in children with cancer. It was completed by all paediatric oncologists participating in the QLIC-ON study during the control period. The paediatric oncologists were asked to predict what the majority of their colleagues had answered on 5 propositions.

A new dimension in the fi eld of PRO research was introduced in the practical part, which demonstrated three short case studies on fi lm. Each case (duration: ± 5 minutes) showed a real consultation of a child and parent visiting a medical specialist, using an actual QLIC-ON PROfi le. Three patients (cases) with diff erent background and patient characteristics (age, gender, ethnicity and medical condition) were selected. To prevent bias, none of the cases involved a paediatric oncologist consultation. Camera work and production were performed by a nurse specialized in fi lming child and parent interactions in medical settings (Video Interaction Guidance) [32]. The consultations were videotaped as a whole; subsequent editing decisions for educational purposes were made by the project group. The cases were implemented in the group training according to the following three steps:

1) Before demonstration: discussion of QLIC-ON PROfi le. The QLIC-ON PROfi le of the child in the fi lm was briefl y discussed, e.g.: • How would you interpret and discuss this QLIC-ON PROfi le? 2) Before demonstration: watching assignment. A watching assignment was given, to ensure the right focus, e.g.: • Pay attention to the questions of the physician. 3) After demonstration: discussion of use of QLIC-ON PROfi le. To refl ect the content of the cases to the paediatric oncologist’s own clinical practice, questions regarding their personal views and experience were raised, e.g.: • What would you have done diff erently?

87

In addition, each case had its own learning goal: case 1) use of the QLIC-ON PROfi le in general, case 2) use of the line graphs, and case 3) use of the decision tree. These were implemented as an extension of the above mentioned third step.

The paediatric oncologists received a QLIC-ON syllabus at the end of the training. It contained hand-outs of the initial (extensive) theoretical part, QLIC-ON reminders, the decision tree, three examples of a QLIC-ON PROfi le, all outcomes of the QLIC-Oncologist Questionnaire and fi ve relevant papers [2,3,8,12,33]. To create an extra incentive, the group training was accredited.

Parent and child instruction

The third and fi nal part of the QLIC-ON education aimed at instructing children (above 8 years) and parents participating in the intervention period. While waiting for their consultation in the out-patient clinic, the child or parent fi lled out the HRQOL questionnaire on a laptop. After completion, the QLIC-ON PROfi le was printed, provided and explained to the child and parent. The explanation concentrated on clarifying the layout and content of the QLIC-ON PROfi le without interpreting the outcome. The child and parent were informed that the paediatric oncologist would discuss the QLIC-ON PROfi le with them during the consultation. They were also invited to feel free to initiate the discussion about their QLIC-ON PROfi le with the paediatric oncologist themselves. Children and parents participating in the control period only fi lled out the digital questionnaire and received no QLIC-ON PROfi le and instruction.

Children and parents initially received the complete QLIC-ON PROfi le. After a few assessments, however, we noticed that some children became competitive when confronted with the line graphs. They expressed feelings of triumph or disappointment when their score was over or under the red dotted line. This was an undesirable eff ect of our intervention. After consulting several paediatric oncologists, we decided to provide parents and children with a QLIC-ON PROfi le without the line graphs. The paediatric oncologists received the complete QLIC-ON PROfi le containing the literal questions with answers as well as the graphs.

3.1.3. A case study

The use of the QLIC-ON PROfi le and decision tree in clinical practice can best be illustrated by a case study (Box 1).

chapter 5

88

Emma is a 10 year old girl diagnosed with Hodgkin lymphoma. She fi nished treatment 4 months ago

and is now in complete remission.

At the moment she is in the out-patient clinic, together with her father, waiting for her third

follow-up consultation with her paediatric oncologist. She is holding the QLIC-ON PROfi le in her hand,

which she has just completed on a laptop (Figure 1) for the third time. Her father has a copy as well

and explores his daughters’ answers. In the back of his head, however, he is slightly tense about the

consultation to come: what are the blood and urine results? Is the scan result good?

Meanwhile, the paediatric oncologist has rounded off the consultation with his previous patient

and takes a brief look at Emma’s QLIC-ON PROfi le (Figure 2, fi rst step Identify) which is attached to the

patient fi le. He focuses on today’s assessment (June 4th 2007) and immediately identifi es a few (possible)

problems: ache, worries, bullying, memory problems and frequent hospital visits. Especially worrying,

bullying and forgetting seem to have a chronic nature, since they have been going on for a few months

now. On top of that the paediatric oncologist concludes from the graph that school functioning is

deteriorating, particularly compared to healthy peers.

The paediatric oncologist invites Emma and her father in his consultation room. The atmosphere

is familiar, because they know each other well. The paediatric oncologist starts the consultation

with common questions, e.g. “How are you, Emma? How is school?”. When she answers “Alright”, the

paediatric oncologist introduces the QLIC-ON PROfi le: “I see that you fi lled out on the QLIC-ON PROfi le

that you frequently worry about what will happen to you. Tell me about that. What do you worry

about, and when?” (Figure 2, second step Discuss) - a conversation follows. During the course of the

consultation and the physical examination the other orange and red items are addressed. Where

applicable the paediatric oncologist takes action (Figure 2, third step Take action). Apparently, the

worrying and bullying is a persistent problem for Emma. Her father indicates that he and his wife fi nd it

diffi cult to take her worries away, and moreover they weren’t aware of the fact that the bullying was still

a problem. The paediatric oncologist decides that Emma might profi t from the help of a psychologist, so

he refers them. Although Emma’s initial response is a bit hesitant, she agreed to give it a try.

A few months later, when Emma and her father see their paediatric oncologist again, Emma has

had three sessions with the hospital psychologist. She feels comfortable with the psychologist and feels

free to disclose what’s on her mind. The worries haven’t vanished, but they have defi nitely decreased.

Also, the tips and exercises the psychologist gave her, help her with her social skills. She is now more

assertive when classmates tease her. The paediatric oncologist can tell she is much more cheerful now

and is pleased to hear she is doing better.

Box 1. A case study

89

4. Study design and outcome measuresTo ensure optimal evaluation of the eff ect of the QLIC-ON intervention, it was also crucial to carefully consider study design and outcome measures.

4.1. Study designA sequential cohort design was regarded most suitable for the purpose of the QLIC-ON study. Paediatric oncologists participated in the control as well as the intervention period; children and parents participated in either the control period or the intervention period, but not both. First, a control period (March 2006 – January 2008) was introduced, in which children or parents (N=84) completed the HRQOL questionnaire while waiting for their consultation in the out-patient clinic. The QLIC-ON PROfi le was not provided to the paediatric oncologist, the child and the parent. The paediatric oncologist started the intervention period (January 2008 – November 2009) as soon as he had seen all patients participating in the control period during three consecutive consultations. The intervention period consisted of a group of 85 children or parents fi lling out the HRQOL questionnaire in the waiting room. This time however, the QLIC-ON PROfi le was provided to the paediatric oncologist, the child and the parent, to be discussed during the three consultations.

We consider randomization or cross-over designs in PRO studies inappropriate due to the risk of contamination [34]. Randomizing the child is not an option, because in that case the paediatric oncologist has to be trained in advance and be prepared for any child, regardless of control or intervention condition. It is not possible to split or block knowledge: as soon as paediatric oncologists have received the QLIC-ON education and gained some experience in using the QLIC-ON PROfi le, bias could occur when seeing children in the control condition. Randomizing the paediatric oncologists is neither a possibility, because of overlapping responsibilities. Children are frequently discussed in team meetings in which all paediatric oncologists participate and it is not uncommon that paediatric oncologists share the treatment of a child. Finally, randomizing the centres would lead to bias as well. Only four centres participated in our study and not all of them contributed equally to the study (the number of participating patients and paediatric oncologists diff ered greatly). Furthermore, it is not unlikely that there are some diff erences in hospital policy between the centres (e.g. shared care). Therefore, to prevent bias, a sequential cohort design was considered most suitable for the QLIC-ON study. 4.2. Outcome measuresThe majority of the outcome measures applied in the QLIC-ON study are known from previous PRO research: communication, advice, referral, satisfaction and HRQOL [10]. According to Greenhalgh et al [11], evidence was found for ‘Changes to doctor-patient communication’ when using HRQOL in clinical practice. We assessed communication by

chapter 5

90

means of two methods: 1) a self constructed questionnaire (regarding the initiative to talk about the four HRQOL domains), 2) audio taping the consultations (a qualitative method to assess conversational topics by means of a checklist). Especially with the audio tapes we hope to gain more insight in the eff ect of applying the QLIC-ON PROfi le in clinical practice, since it is a less frequently used instrument and it goes into more detail. ‘Changes to clinicians management of the patient’ (e.g. advice and referral), ‘Patient satisfaction’ and ‘HRQOL’ are more diffi cult to fi nd [11]. Still, we have incorporated these assessments in the QLIC-ON study, because they are well known and important as an outcome. Moreover, as can be concluded from previous research, it is not impossible to fi nd change via some of these measures [3,7].

In addition to the more established outcome measures, the QLIC-ON study also assessed health care use and health care needs – comparable to ‘Changes to patient health behaviour’ in the model of Greenhalgh et al [11]. Understanding the use and needs regarding health care might partially explain patient satisfaction and HRQOL outcome.

The last outcome measure focused on feasibility, which was obviously only investigated in the intervention group. This does not actually fi t the model, however, it can be regarded as a direct outcome, immediately linked to ‘Provision of information from health status measures to clinicians’ [11].

5. Discussion and conclusion

5.1. DiscussionThis paper provides a thorough description of the development, implementation, study design and outcome measures of a PRO on HRQOL in clinical practice (QLIC-ON PROfi le). By applying the conclusions of Greenhalgh et al. [11], combined with our own experience, we hope to have maximized the eff ectiveness of the QLIC-ON intervention. Considerations for future PRO studies in clinical practice are summarized in Table 1.

There are some points of interest to consider. First of all, for future research we recommend PRO use by other professionals besides physicians, as Greenhalgh et al. [11] suggested. In our study paediatric oncologists were selected because they are responsible for the after-care trajectory, the phase in which the QLIC-ON PROfi le is studied. In one centre, however, we also included two nurse practitioners as users, since they carried the same responsibility for after-care as the paediatric oncologists. We believe nurses, nurse practitioners, social workers or psychologists could also be eligible candidates for using HRQOL PRO [35]. From a professional point of view they might even be more interested in and capable of discussing HRQOL than physicians. In addition, it is important to keep in mind possible diff erences in the use of the QLIC-ON PROfi le between the four participating centres as a whole. We expect these diff erences to be minor, since the Netherlands is a small country, with only seven paediatric oncology centres that apply the same (international) treatment protocols. However, our aim was for each child to be seen

91

Table 1. Considerations for future PRO studies in clinical practice

Development

Implementation

Study design

Outcome

chapter 5

92

by the same paediatric oncologist for the time of the three QLIC-ON consultations. Due to hospital policy and the reality of clinical practice this was not always possible in one of the four centres, in which shared treatment was common. In the worst case, this could lead to less continuity of care [36] compared to the other centres, which might impact parent and child satisfaction.

Second, the QLIC-ON study proved to be very labour-intensive. Every child completed the PRO during three consecutive consultations. These assessments took place on a stand alone laptop in the outpatient clinic. A researcher was present during every assessment to set up all materials, to guide them through the procedure, to hand a copy of the QLIC-ON PROfi le to the paediatric oncologist and to make sure the consultation was audio taped. We recommend performing future PRO studies and implementations via internet. This gives respondents the opportunity to complete the questionnaire at home and users to retrieve the PRO on the computer in the consultation room.

Third is the timing of assessment. The QLIC-ON sample concerned children aged 0-18 years with diff erent kinds of malignant diagnoses - a very heterogeneous group. For reasons of homogeneity we selected a moment similar for all children: the fi rst three consultations after end of treatment. At this point, however, children and parents are usually familiar with their paediatric oncologist. As a consequence, the QLIC-ON PROfi le might reveal results that are already known to the paediatric oncologist. Furthermore, HRQOL problems might be more pronounced during treatment than after treatment [29]. Both reasons could render the QLIC-ON PROfi le to be less useful and less eff ective than when applied during treatment. Nonetheless, we believe timing of assessment should in essence not matter. Paediatric oncologists care for their patients and must therefore always be aware of possible HRQOL problems.

Another point of consideration is that the QLIC-ON study design consisted of two groups: a(n) (attention) control group and an intervention group. It would probably have gained in strength if the sequential cohort design contained three groups: a control, an attention-control and an intervention group. Velikova et al. [3] found a diff erence in HRQOL subscales between intervention and control, but not between intervention and attention-control. Due to practical limitations (small population size) three groups were not possible in the QLIC-ON study. Furthermore, we realize that our study is at risk for a history eff ect, since intervention and control group data were not collected at the same time. However, the control and intervention period were contiguous and gradually ongoing over time, and treatment protocols remained comparable within the 3-year time span of the study.

Finally, are the outcome measures. Greenhalgh et al [11] already mentioned that there is little evidence for several outcomes in the model. For the more distal outcomes of patient health status and satisfaction with care, numerous intermediate processes need to occur before any impact could be registered. Maybe satisfaction for example does not off er a true refl exion of patient’s experiences [36]. Subsequent research should therefore consider other measures besides the established ones, which until now have not provided satisfying evidence.

93

5.2. ConclusionIn conclusion, each of the fi ve recommendations of Greenhalgh et al. [11] was followed up to a greater or lesser extent, depending on the needs of the QLIC-ON study and our own experience: (1) Although our HRQOL instruments are of a generic and structured nature, we underlined the individual outcomes by presenting the answers of the respondent on all items; (2) The obvious user of the QLIC-ON PROfi le was the paediatric oncologist, yet we acknowledge that other users e.g. nurses, social workers or psychologists might fi t the use of the QLIC-ON PROfi le as well; (3) HRQOL information was collected preceding three consecutive consultations and fed back to the paediatric oncologist longitudinally; (4) The QLIC-ON PROfi le is not intended as a screening device but merely as a monitoring tool, thus no cut-off scores were utilized, and scale scores of the healthy norm population were only presented in the graphs to underline the monitoring aspect; (5) With respect to the implementation of the QLIC-ON PROfi le a lot of attention has been paid to commitment and education. Especially in the group training, we strove for innovation by adding educational fi lm material, which was also emphasized by Guyatt et al. [37].

It is important to realize that only the results of the QLIC-ON study can determine whether our QLIC-ON PROfi le development and implementation have been successful. Nonetheless, this paper provides relevant information for all future PRO studies on how and what to develop and implement, regardless of PRO topic, patient age and disease. In order to increase eff ectiveness of PRO in clinical practice and to improve patient care, we encourage other researchers to give profound insight in development, implementation, study design and outcome measures as well. To achieve broad use of PRO in diverse clinical settings, it is not only necessary to build upon each other’s methods and results, but also upon each other’s unique experiences in the development of a specifi c PRO.

Finally, there is the challenge for the more distant future: closing the gap between research and actual implementation in clinical practice. Most important factors for success are – in our opinion - the user-friendliness of the PRO tool and commitment of the users. Both have been stressed in this paper, but also acknowledged by other researchers [34,38]. The PRO methods applied in research have mainly depended on the work and eff ort of one or more researchers (e.g. logistics, assessment, evaluation). In the reality of clinical practice, there is probably no budget for this available. Internet can provide a step towards a solution, since in the Netherlands 87% of the households has access nowadays [39]. Less staff is needed, if the PRO is directly linked to patient data (e.g. treatments, consultations) and emails with invitations and questionnaires are automatically generated. An internet-based PRO is currently being studied by our research group.

The second important factor is that the users of the PRO have to be convinced of the benefi ts of the tool. If there is no enthusiasm among users, the willingness to successfully implement the PRO in clinical practice is very diffi cult – is our experience. Ways to increase commitment have been described in this paper. In addition, it can be helpful to have at

chapter 5

94

least one ‘user-ambassador’ in the user group to convince the group from within, and not only from external, which researchers often are. Of course, commitment is expected to rise with fi gures: proof of eff ectiveness is likely to encourage medical professionals to use PRO tools in their practice.

95

Reference List1. Osoba D. Translating the science of patient-

reported outcomes assessment into clinical practice. J Natl Cancer Inst Monogr 2007;5-11.

2. Detmar SB, Muller MJ, Schornagel JH, Wever LD, Aaronson NK. Health-related quality-of-life assessments and patient-physician communication: a randomized controlled trial. JAMA 2002;288:3027-3034.

3. Velikova G, Booth L, Smith AB, Brown PM, Lynch P, Brown JM, Selby PJ. Measuring quality of life in routine oncology practice improves communication and patient well-being: a randomized controlled trial. J Clin Oncol 2004;22:714-724.

4. Gutteling JJ, Darlington AS, Janssen HL, Duivenvoorden HJ, Busschbach JJ, de Man RA. Eff ectiveness of health-related quality-of-life measurement in clinical practice: a prospective, randomized controlled trial in patients with chronic liver disease and their physicians. Qual Life Res 2008;17:195-205.

5. Taenzer P, Bultz BD, Carlson LE, Speca M, DeGagne T, Olson K, Doll R, Rosberger Z. Impact of computerized quality of life screening on physician behaviour and patient satisfaction in lung cancer outpatients. Psychooncology 2000;9:203-213.

6. Varni JW, Seid M, Knight TS, Uzark K, Szer IS. The PedsQL 4.0 Generic Core Scales: sensitivity, responsiveness, and impact on clinical decision-making. J Behav Med 2002;25:175-193.

7. Wit de M, Delemarre-van de Waal HA, Bokma JA, Haasnoot K, Houdijk MC, Gemke RJ, Snoek FJ. Monitoring and discussing health-related quality of life in adolescents with type 1 diabetes improve psychosocial well-being: a randomized controlled trial. Diabetes Care 2008;31:1521-1526.

8. Varni JW, Burwinkle TM, Lane MM. Health-related quality of life measurement in pediatric clinical practice: an appraisal and precept for future research and application. Health Qual Life Outcomes 2005;3:34.

9. Greenhalgh J, Meadows K. The eff ectiveness of the use of patient-based measures of health in routine practice in improving the process and outcomes of patient care: a literature review. J Eval Clin Pract 1999;5:401-416.

10. Valderas JM, Kotzeva A, Espallargues M, Guyatt G, Ferrans CE, Halyard MY, Revicki DA, Symonds T, Parada A, Alonso J. The impact of measuring patient-reported outcomes in clinical practice: a systematic review of the literature. Qual Life Res 2008;17:179-193.

11. Greenhalgh J, Long AF, Flynn R. The use of patient reported outcome measures in routine clinical practice: lack of impact or lack of theory? Soc Sci Med 2005;60:833-843.

12. Stam H, Grootenhuis MA, Brons PP, Caron HN, Last BF. Health-related quality of life in children and emotional reactions of parents following completion of cancer treatment. Pediatr Blood Cancer 2006;47:312-319.

13. Hickey AM, Bury G, O’Boyle CA, Bradley F, O’Kelly FD, Shannon W. A new short form individual quality of life measure (SEIQoL-DW): application in a cohort of individuals with HIV/AIDS. BMJ 1996;313:29-33.

14. Fekkes M, Theunissen NC, Brugman E, Veen S, Verrips EG, Koopman HM, Vogels T, Wit JM, Verloove-Vanhorick SP. Development and psychometric evaluation of the TAPQOL: a health-related quality of life instrument for 1-5-year-old children. Qual Life Res 2000;9:961-972.

15. Verrips GH, Vogels TG, Verloove-Vanhorick SP, Fekkes M, Koopman HM, Theunissen NCM. Health-related quality of life measure for children-the TACQoL. J Applied Therapeutics 1997;1:357-360.

16. Bruil, J., Fekkes, M., Vogels, T., and Verrips, E. TAAQOL Manual. 2004.

17. Koopman H.M., Theunissen, N. C. M., Vogels, T. G., Kamphuis, R. P., and Verrips, G. H. The dux-25: A Short-Form Questionnaire for Measuring Health Related Quality of Life of Children with a Chronic Illness. 1998;7:619.

chapter 5

96

18. Ravens-Sieberer U, Gosch A, Rajmil L, Erhart M, Bruil J, Power M, Duer W, Auquier P, Cloetta B, Czemy L, Mazur J, Czimbalmos A, Tountas Y, Hagquist C, Kilroe J. The KIDSCREEN-52 Quality of Life Measure for Children and Adolescents: Psychometric Results from a Cross-Cultural Survey in 13 European Countries. Value Health 2007.

19. Varni JW, Burwinkle TM, Seid M, Skarr D. The PedsQL 4.0 as a pediatric population health measure: feasibility, reliability, and validity. Ambul Pediatr 2003;3:329-341.

20. Landgraf J: Child Health Questionnaire (CHQ). A User’s Manual. 1999.

21. Engelen V, Haentjens MM, Detmar SB, Koopman HM, Grootenhuis MA. Health related quality of life of Dutch children: psychometric properties of the PedsQL in the Netherlands. BMC Pediatr 2009;9:68.

22. Fekkes, M., Bruil, J., and Vogels, T. TAPQOL-manual. 2003.

23. Leaf PJ, Owens PL, Leventhal JM, Forsyth BW, Vaden-Kiernan M, Epstein LD, Riley AW, Horwitz SM. Pediatricians’ training and identifi cation and management of psychosocial problems. Clin Pediatr (Phila) 2004;43:355-365.

24. Baars RM, van der Pal SM, Koopman HM, Wit JM. Clinicians’ perspective on quality of life assessment in paediatric clinical practice. Acta Paediatr 2004;93:1356-1362.

25. Bradlyn AS, Ritchey AK, Harris CV, Moore IM, O’Brien RT, Parsons SK, Patterson K, Pollock BH. Quality of life research in pediatric oncology. Research methods and barriers. Cancer 1996;78:1333-1339.

26. Eiser C, Eiser JR, Greco V. Surviving childhood cancer: quality of life and parental regulatory focus. Pers Soc Psychol Bull 2004;30:123-133.

27. Grootenhuis MA, Koopman HM, Verrips EG, Vogels AG, Last BF. Health-related quality of life problems of children aged 8-11 years with a chronic disease. Dev Neurorehabil 2007;10:27-33.

28. Meeske K, Katz ER, Palmer SN, Burwinkle T, Varni JW. Parent proxy-reported health-related quality of life and fatigue in pediatric patients diagnosed with brain tumors and acute lymphoblastic leukemia. Cancer 2004;101:2116-2125.

29. Varni JW, Burwinkle TM, Katz ER, Meeske K, Dickinson P. The PedsQL in pediatric cancer: reliability and validity of the Pediatric Quality of Life Inventory Generic Core Scales, Multidimensional Fatigue Scale, and Cancer Module. Cancer 2002;94:2090-2106.

30. Ashworth CD, Williamson P, Motano D. A scale to measure physician beliefs about psychosocial aspects of patient care. Soc Sci Med 1984;19:1235-1238.

31. McLennan JD, Jansen-McWilliams L, Comer DM, Gardner WP, Kelleher KJ. The Physician Belief Scale and psychosocial problems in children: a report from the Pediatric Research in Offi ce Settings and the Ambulatory Sentinel Practice Network. J Dev Behav Pediatr 1999;20:24-30.

32. Sibbing H.M.H., Kat C.N., Grootenhuis M.A., Last B.F.. Positief eff ect van video-interactiebegeleiding in het kinderziekenhuis. Tijdschr Kindergeneeskd 2005;73:209-214.

33. Wallander JL, Varni JW. Eff ects of pediatric chronic physical disorders on child and family adjustment. J Child Psychol Psychiatry 1998;39:29-46.

34. Snyder CF, Aaronson NK. Use of patient-reported outcomes in clinical practice. Lancet 2009;374:369-370.

35. Hilarius DL, Kloeg PH, Gundy CM, Aaronson NK. Use of health-related quality-of-life assessments in daily clinical oncology nursing practice: a community hospital-based intervention study. Cancer 2008;113:628-637.

36. Harley C, Adams J, Booth L, Selby P, Brown J, Velikova G. Patient Experiences of Continuity of Cancer Care: Development of a New Medical Care Questionnaire (MCQ) for Oncology Outpatients. Value Health 2009.

37. Guyatt GH, Ferrans CE, Halyard MY, Revicki DA, Symonds TL, Varricchio CG, Kotzeva A, Valderas JM, Alonso J. Exploration of the value of health-related quality-of-life information from clinical research and into clinical practice. Mayo Clin Proc 2007;82:1229-1239.

38. Greenhalgh J. The applications of PROs in clinical practice: what are they, do they work, and why? Qual Life Res 2009;18:115-123.

39. Statistics Netherlands. www.cbs.nl. 2009.

976Chapter 6

V. Engelen 1S.B. Detmar 2H.M. Koopman 3

H. Maurice – Stam1 H.N. Caron 4 P.M. Hoogerbrugge 5 R.M. Egeler 6G.J.L. Kaspers 7M.A. Grootenhuis 1

1 Psychosocial Department, Academic Medical Centre/ Emma Children’s Hospital, Amsterdam2 Prevention and Health Department, TNO, Leiden3 Medical Psychology Department, Leiden University Medical Center, Leiden4 Paediatric Oncology Department, Academic Medical Centre/ Emma Children’s Hospital, Amsterdam5 Paediatric Oncology Department, Radboud University Medical Centre, Nijmegen6 Paediatric Oncology Department, Leiden University Medical Centre, Leiden7 Paediatric Oncology Department, VU University Medical Center, Amsterdam

Reporting health related quality of life scores to physicians during routine follow-up visits of paediatric oncology patients: is it effective?

PEDIATR BLOOD CANCER. 2011 16 MAY

98

AbstractBackground. The aim of the current study is to investigate the eff ectiveness of an intervention that provides health related quality of life (HRQOL) scores of the patient (the QLIC-ON PROfi le) to the paediatric oncologist.

Procedure. Children with cancer participated in a sequential cohort intervention study: intervention N=94, control N=99. Primary outcomes of eff ectiveness were communication about HRQOL domains (t-test, Mann Whitney U test) and identifi cation of HRQOL problems (Chi2 test). Secondary outcomes were satisfaction (multi level analysis), referrals (Chi2 test) and HRQOL (multi level analysis).

Results. The QLIC-ON PROfi le increased discussion of emotional functioning (control M=32.9 vs intervention M=47.4, p<0.05) and psychosocial functioning (M=56.9 vs M=63.8, p<0.05). Additionally more emotional problems remained unidentifi ed in the control compared to the intervention group, e.g. anger (control 26% vs intervention 3%, p<0.01), fear (14% vs 0%, p<0.01) and sadness (26% vs 0%, p<0.001). The intervention had no eff ect on satisfaction and referrals, but did improve HRQOL of patients 5 to 7 years of age with respect to self esteem (p<0.05), family activities (p<0.05) and psychosocial functioning (p<0.01).

Conclusions. We conclude that a PRO is a helpful tool for systematic monitoring HRQOL of children with cancer, without lengthening the duration of the consultation. It is recommended to be implemented in clinical practice.

99

Introduction Monitoring health related quality of life (HRQOL) by providing patient reported outcomes (PROs) as a tool in clinical practice has been the focus of several studies [1-5]. Incorporating PRO tools in clinical oncology practice appeared to facilitate discussion of HRQOL issues and increase physician’s awareness of their patients HRQOL [1]. There was also an impact on physician-patient communication with better HRQOL and emotional functioning for some patients [2]. Moreover, a study in patients with chronic liver disease demonstrated that physicians using a PRO tool alter their treatment policy signifi cantly more than physicians not using HRQOL feedback [3]. In addition, it appears that the introduction of PRO tools into the daily routine of an out-patient clinic is feasible [6].

Previous PRO studies have been conducted primarily in adult patients, with very few studies in paediatrics [7]. A recent exception is the study of De Wit et al. [8] who demonstrated positive eff ects of a PRO tool in adolescents with diabetes. It is remarkable that no other paediatric populations have profi ted from feeding back HRQOL via PRO tools in clinical practice, since it is known that children with a (chronic) disease are at risk for suff ering from HRQOL problems [9] and that physicians are not always aware of these problems [10].

Children with cancer can also experience HRQOL problems, which can continue after completion of treatment [11, 12]. It is generally recognized that the fulfi lling of age-specifi c developmental tasks in childhood is of great importance to adjustment in adult life; the burden of cancer interferes with this process [13] and may cause HRQOL problems. PRO tools could possibly bridge the gap between child and oncologist and serve as an aid in identifying and discussing HRQOL problems, also to prevent developmental delay. Additionally, PRO tools could contribute to eff ective patient–physician communication, which is of crucial importance to psycho-social outcomes in patients with cancer [14, 15].

The aim of the current study was to investigate the eff ectiveness of an intervention that provides HRQOL scores of the patient (a PRO tool) to the oncologist. Primary outcomes of eff ectiveness included communication about HRQOL domains and identifi cation of HRQOL problems. Secondary outcome of eff ectiveness included satisfaction, referrals and HRQOL.

MethodsInterventionThe intervention consisted of providing HRQOL scores of the patient (Quality of Life in Childhood Oncology (QLIC-ON) PROfi le, p. 173) to the oncologist. The QLIC-ON PROfi le is a PRO tool completed by the child or parent [16]. The QLIC-ON PROfi le presents four HRQOL domains (physical, emotional, social and cognitive functioning) by summarising the answers of child or parent on two generic HRQOL questionnaires: the PedsQL Generic

chapter 6

100

Core Scale (Paediatric Quality of Life Inventory) [11, 17] or the TAPQOL (TNO-AZL Preschool children Quality of Life) [18, 19]. The PedsQL self-report form was used for children aged 8 to 18; parents of children aged 6 to 7 completed the PedsQL proxy-report. Dutch PedsQL norm scores were available [11]. The PedsQL has a short completion time (approximately 5-10 minutes) and good utility, validity and reliability [17]. The TAPQOL was used for parents of children aged 0 to 5 years (proxy-report). Dutch TAPQOL norm scores were available as well. Completion time is about 5-10 minutes and psychometric properties are satisfactory [18, 19].

To optimize the eff ectiveness of the QLIC-ON PROfi le, we provided a training which comprised both an individual and a group session for oncologists, and a patient instruction. The training was also part of the intervention [16].

ParticipantsParticipants of the present study were children (0-18 years) with cancer (all diagnoses) shortly after (at 0-3 months; children with stem cell transplantations (SCT) at 6 months) completion of successful treatment. Parents and oncologists took part in the study as well. Children were included in the study after fi nishing treatment because they had diff erent kinds of diagnoses. Because it concerned a very heterogeneous group, we selected a moment similar for all children: the fi rst period after end of treatment.

The QLIC-ON study concerns a large Dutch multi-centre research project in which the Academic Medical Centre/Emma Children’s Hospital, Leiden University Medical Centre, Radboud University Medical Centre and the VU University Medical Center participated. The Medical Ethics Committees of all participating centres approved the study.

ProcedureChildren and parents were invited by mail to participate as soon as the child had fi nished treatment. Written informed consent was obtained in the out-patient clinic shortly before their fi rst consultation. All outcome measures were completed on paper, while the PRO tool was fi lled out digitally in the out-patient clinic prior to each consultation.

DesignA sequential cohort design was used [16]. Oncologists participated fi rst in the control period (from March 2006 to January 2008 the QLIC-ON PROfi le was completed by the child or parent but not provided to the oncologist) and, after that, in the intervention period (from January 2008 to November 2009 the QLIC-ON PROfi le was completed by the child or parent and provided to the oncologist). Children and parents took part in either the control period or the intervention period. The QLIC-ON PROfi le was completed by the child or parent shortly before three consecutive consultations. To prevent contamination of the control period, oncologists only started the intervention period (with a training), as soon as all children from the control period had fi nished their three consultations.

101

Both groups were evaluated before (baseline), during (consultation evaluation) and after (follow-up) the three consultations.

MeasuresSocio-demographic and medical variables

Socio-demographic data of participating children and parents were collected at baseline. Child age and gender in the non-participants group were retrieved from medical fi les. Medical information of participants as well as non-participants was gathered by data managers or oncologists. Socio-demographics of the oncologist were collected at the start of the study: age, gender and years of work experience as a oncologist.

Primary outcomesCommunication about HRQOL domains

All consultations were sound recorded with a digital audio device to assess diff erent aspects of communication about the four HRQOL domains. Reasons for absent sound recordings were no parental or child consent for the recording, drop-out due to relapse of the disease or practical reasons (e.g. technical problems). One third of all audio-tapes were (block) randomly selected for analysis, with an equal number of audio-tapes per consultation (fi rst, second or third) per group (intervention or control). Thus, if a child had a minimum of one or a maximum of three sound recordings available, only one sound recording was used for analysis.

The audio-assessment form, a semi-open instrument, was used to score the consultations. A manual containing defi nitions and rules was applied to ensure reliable and valid scoring of the consultations. Twenty percent of the audio-tapes were double scored by two researchers (the fi rst author and a paediatric psychology student); ICCs were calculated to determine inter rater reliability of the audio assessment form.

The four HRQOL domains were displayed on the audio-assessment form. Each HRQOL domain contained open spaces to register the topics within that particular HRQOL domain. For every registered topic the researcher indicated to what extent the topic was discussed. At the end of analysing each consultation, the researcher completed two summarising questions for every HRQOL domain; it was reported on Likert scales to what extent that HRQOL domain was discussed: 0 not discussed – 5 extensively discussed (Interclass Correlation (ICC) ranged from 0.22 to 0.61) and how much time was spent on it: 0 no time – 3 a lot of time (ICC: 0.29 to 0.75). Scores were transformed to a 0 - 100 scale, with higher scores implying better communication. These scores were reported in the current paper. Total consultation duration was also recorded by means of the digital audio device.

Identifi cation of HRQOL problems

Unidentifi ed HRQOL problems were assessed by matching the answers of each child on the QLIC-ON PROfi le with the results of the HRQOL topics checklist, completed by parents

chapter 6

102

after each consultation. The HRQOL topics checklist contained 27 multiple choice topics, based on the four domains of HRQOL. The parent indicated if the topic was discussed (yes/no) during the consultation.

Green answers (no problems) on the QLIC-ON PROfi le that were or were not discussed during the consultation, according to the parents, received code 1 (=match). Orange (some problems) and red answers (frequent problems) were also assigned code 1 (=match) if the items were discussed. Orange and red answers that were not discussed received code 0 (=mismatch); these were the HRQOL problems that remained unidentifi ed, which are reported in the present paper. The emotional, social and cognitive HRQOL topics of the third consultation (C3) were used for analysis.

Secondary outcomes Satisfaction

Satisfaction with the consultation was assessed with the parent and oncologist at the end of each consultation. The Patient Satisfaction Questionnaire [20, 21] was adapted to the QLIC-ON study. Five questions were answered on a 100 mm visual analogue scale (0 not at all satisfi ed – 100 very satisfi ed). The questions referred to satisfaction with respect to meeting needs, active involvement, provided information, (emotional) support and overall satisfaction. A total satisfaction score was calculated by averaging the responses to the 5 questions. Oncologists completed these questions (about themselves) for both child and parent; parents fi lled out the questions referring to the oncologist.

Referrals

Psychosocial referrals were reported by the oncologist after each consultation. There were two options: (neuro) psychologist and social worker. Scoring range was 0 no referral – 2 two referrals; scores were transformed to a dichotomy scale (0 no referrals - 1 one or more referrals).

HRQOL

To measure the eff ect of the intervention on the child’s HRQOL, HRQOL was investigated at baseline and follow-up with generic standardized HRQOL questionnaires. The Infant and Toddler Quality Of Life Questionnaire [22] (ITQOL) was completed by parents of patients aged 0 to 4 years. The Child Health Questionnaire Parent Form 50 [23, 24] (CHQ PF 50) was fi lled out by parents of children aged 5 to 7 years. Children and adolescents aged 8 to 18 completed the Kidscreen [25, 26].

Statistical analysisSocio-demographics and medical variables

Data were analyzed with SPSS 16.0.2. The diff erence between participants and non-participants, as well as the control and intervention group regarding age group, gender,

103

diagnosis, treatment, country of birth, education and employment was analysed by means of Chi2 tests. Diff erence in age, treatment duration and baseline HRQOL (only for control and intervention group) was examined with t-tests.

Primary outcomes

Several tests were performed to investigate the eff ect of the intervention on communication about HRQOL domains. Independent t-tests were performed to test the diff erence between control and intervention group with respect to the extent of discussing the HRQOL domains and consultation duration. Diff erences between the groups for time spent on HRQOL domains were analysed with Mann-Whitney U tests. Chi2 tests were applied to test diff erences regarding the identifi cation of HRQOL problems.

Secondary outcomes

Multilevel analyses with three levels (consultation, child, oncologist) were performed to test the intervention eff ect on satisfaction with the consultation, employing linear mixed modelling [27]. Satisfaction was predicted by the intervention (1=yes, 0=no), corrected for consultation, child age, age and gender of the oncologist. The models were fi tted with a random intercept per child and per oncologist. Regression coeffi cients of the predictors were considered to be fi xed, because adding random eff ects did not improve the models. Chi2 tests were applied to test diff erences regarding referrals. Multilevel analyses with two levels (child, oncologist) were performed to test the intervention eff ect on HRQOL, employing linear mixed modelling [27]. HRQOL at follow-up was predicted by the intervention, corrected for HRQOL at baseline, child age, age and gender of the oncologist. The models were fi tted with a random intercept per oncologist. For some HRQOL outcomes there appeared to be no intercept variance, these models were fi tted with a fi xed intercept. Regression coeffi cients were considered to be fi xed, because adding random eff ects did not improve the models.

ResultsSocio-demographic and medical variables274 children with cancer and their parents were approached to participate. Eventually, 193 (70.4%) participants completed the baseline assessment (Table 1) [28]. 99 children participated in the control group - 14.3% drop-out; 94 in the intervention group - 21.3% drop-out (Table 2). Main reasons for drop-out were relapse of the disease and interference of end of data collection with unacceptable long intervals between the consultations (especially between consultation 2 and 3 in the intervention group). Regarding baseline HRQOL, the intervention group scored better than the control group in a few domains: change in health (ITQOL) in patients of 0 to 4 years of age (83.7 vs. 61.4; p<0.05), psychosocial summary (CHQ PF 50) in 5 to 7 year old children (53.7 vs. 48.2; p<0.05);

chapter 6

104

physical well-being (Kidscreen) in patients of 12 to 18 years of age (44.9 vs 39.1; p<0.01). A total of 34 oncologists took part; 47.1% were female, average age was 41.9 (SD 10.1) and mean work experience as a oncologist was 8.7 years (SD 7.6).

105

Table 1. Socio-demographic and medical variables: participants vs. non-participants

*p<0.05, **p<0.01,1-4 are explained on p.45.

Participants Non-participants N=193 N=81

Age (years) 193 9.27** 5.04 81 10.93 4.26 Treatment duration (months)1 190 13.63 9.17 78 13.79 8.03 Time off treatment (weeks)2 191 7.60 4.53 N % N % Age Group (years)** 0-4 54 28.0 9 11.1 5-7 34 17.6 13 16.0 8-11 37 19.2 26 32.1 12-18 68 35.2 33 40.7 Gender (female) 92 47.7 34 42.0 Diagnosis3 Leukemia 69 36.1 38 46.9 Lymphoma 24 12.4 9 11.1 Brain tumor 16 8.4 12 14.8 Solid tumor 42 22.0 11 13.6 Bone tumor 29 15.0 8 9.9 Other 13 6.8 3 3.7 Treatment Surgery 101 52.9 39 48.1 Radiotherapy 55* 28.8 34 42.0 Chemotherapy 184 96.3 79 97.5 SCT 18** 9.4 20 24.7 Age (years) 190 40.89 7.12 N %

Gender (female) 150 79.8 Country of birth (Netherlands) 166 86.9 Education4 Low 37 19.4 Middle 96 50.3 High 57 29.8

Employed 125 65.4

CHILD N M SD N M SD


chapter 6

106

Table 2. Socio-demographic and medical variables: control vs. intervention group

Control Intervention N=99 N=94

Age (years) 99 9.13 5.07 94 9.41 5.03 Treatment duration (months) 1 98 13.27 9.37 92 14.02 8.98 Time off treatment (weeks) 2 98 7.17 4.15 93 8.04 4.88 N % N % Age Group (years) 0-4 27 27.3 25 26.6 5-7 20 20.2 16 17.0 8-11 19 19.2 18 19.1 12-18 33 33.3 35 37.2 Gender (female) 46 46.5 46 48.9 Diagnosis 3 Leukemia 35 34.4 34 36.2 Lymphoma 12 12.1 12 12.8 Brain tumor 8 8.1 8 8.5 Solid tumor 21 21.2 21 22.3 Bone tumor 16 16.2 13 13.8 Other 7 7.0 6 6.4 Treatment Surgery 53 54.1 48 51.6 Radiotherapy 31 31.6 24 25.8 Chemotherapy 97 99.0 87 93.5 SCT 13 13.3 5 5.4

Age (years) 97 41.05 6.90 93 40.72 7.38 N % N %

Gender (female) 77 81.05 73 81.1 Country of birth (Netherlands) 84 85.7 82 88.2 Education4 Low 21 21.4 16 17.4 Middle 50 51.0 46 50.0 High 27 27.6 30 32.6

Employed 65 66.3 60 64.5

CHILD N M SD N M SD


1-4 are explained on p.45.

107

Figure 1. Communication: discussion of HRQOL domains measured with the audio

assessment form; control vs. intervention group

Scale: 0 (not discussed) - 100 (extensively discussed); *p<0.05 according to t-test

84

70 6964

83

47

73 71

6469

33*

57*

0

10

20

30

40

50

60

70

80

90

100

Physical Emo onal Social Cogni ve Psychosocialtotal

Total

HRQOL domains

Dis

cuss

ed (m

ean)

Control (N=77) Interven on (N=78)

chapter 6

Primary outcomesCommunication about HRQOL domains

Emotional and psychosocial functioning was signifi cantly (p<0.05) more extensively discussed in the intervention group than in the control group (Figure 1). Additionally, in the intervention group, signifi cantly (p<0.05) more time was spent on emotional and cognitive functioning (Figure 2). The average consultation duration in the intervention group (20 min 34 sec) was shorter compared to the control group (23 min 18 sec), but this diff erence was not signifi cant (p=0.052).

Identifi cation of HRQOL problems

Signifi cantly more emotional and cognitive problems were identifi ed in the intervention group compared to the control group, namely being afraid (p<0.05) (Figure 3, children aged 0-5; TAPQOL), anger (p<0.01), fear (p<0.01), sadness (p<0.001), paying attention (p<0.05) and forgetting things (p<0.05) (Figure 4, children aged 6-18; PedsQL).

108

Figure 2. Communication: time spent on HRQOL domains measured with the audio

assessment form; control vs. intervention group

Scale: 0 (no time) - 100 (a lot of time); *p<0.05 according to Mann-Whitney U test

75 7571

84

7683

68* 68*

0

10

20

30

40

50

60

70

80

90

100

HRQOL domains

Tim

e (m

ean

rank

)

Control (N=74) Interven on (N=77)

Physical Social

Secondary outcomesSatisfaction

Despite a trend towards higher scores for the intervention group (Table 3), multilevel analysis demonstrated no diff erence in satisfaction between the control and intervention group. Regression coeffi cients B for intervention versus control group were as follows: oncologists about the parent B=2.08 (p=0.07) and about the child: B=2.59 (p=0.11); parents about oncologist B=1.15 (p=0.52).

Referrals

There were no signifi cant diff erences in referrals between the two groups. Percentages of referrals for the control versus intervention group were for consultation 1: 2.2% vs. 3.2%, p=0.44; consultation 2: 4.7% vs. 2.3%, p=0.26; consultation 3: 2.6% vs. 2.0%, p=0.85.

HRQOL

Children aged 5 to 7 participating in the intervention group, scored signifi cantly better at follow-up than controls on HRQOL subscales (range 0-100) self esteem (B=10.17; p<0.05), family activities (B=11.11; p<0.05) and psychosocial summary (B=6.45; p<0.01). In 0- to 4- and 8- to 18-year-old children and adolescents no diff erences between the control and intervention group were demonstrated.

109

Figure 3. Unidentifi ed HRQOL problems in children aged 0-5 (TAPQOL). Orange

and red answers on the QLIC-ON PROfi le that weren’t discussed during the third

consultation: control vs. intervention group

*p<0.05 according to Chi2 test

15

39

19

4239

29

13

0

20

10

35

10

20

10

0

39*

0

10

20

30

40

50

60

70

80

90

100

Cheerful Angry Aggressive Rebellious Afraid Tense Self-assured Di cultyunderstanding

HRQOL problems (orange and red answers)

Uni

den

ed (%

)

Control C3 N=26 Interven on C3 N=20

chapter 6

110

Figure 4. Unidentifi ed HRQOL problems in children aged 6-18 (PedsQL). Orange

and red answers on the QLIC-ON PROfi le that weren’t discussed during the third

consultation: control vs intervention group

*p<0.05, **p<0.01, ***p<0.001 according to Chi2 test

50

7 9

0 3 0 0 2 05

13

3

26**

14**

26***21*

32*

0

10

20

30

40

50

60

70

80

90

100

Worries Anger Fear Sadness Bullying Ge ngalong

Payinga en on

Forge ngthings

Schoolprogress

HRQOL problems (orange and red answers)

Uni

den

ed (%

)

Control C3 N=44 Interven on C3 N=42

Control Intervention Satisfaction with consultation N M SD N M SD

Oncologist about parent - Consultation 1 94 71.89 9.95 80 74.85 10.02- Consultation 2 82 71.76 9.25 80 74.13 10.51- Consultation 3 77 72.18 10.27 66 74.69 10.38 Oncologist about child - Consultation 1 56 69.83 13.32 54 72.09 14.08- Consultation 2 53 70.71 10.69 55 73.43 12.23- Consultation 3 47 70.74 13.63 51 70.52 12.81 Parent about oncologist - Consultation 1 82 76.99 15.29 86 80.21 14.73- Consultation 2 75 77.43 14.32 81 77.85 14.85- Consultation 3 69 79.43 11.53 59 80.78 11.24

Table 3. Satisfaction with the consultation of oncologists and parents per

consultation: control vs intervention group

111

DiscussionThis study demonstrates that monitoring HRQOL in paediatric oncology practice by means of providing HRQOL scores of the patient to the oncologist (the QLIC-ON PROfi le) increases discussion of emotional functioning and enhances identifi cation of emotional problems. The QLIC-ON PROfi le does not lengthen consultation duration.

Consistent with previous research [7], the QLIC-ON PROfi le had an impact on communication during the consultation which can infl uence psycho-social outcomes in patients with cancer [14, 15]. Emotional and psychosocial functioning was more extensively discussed with use of the QLIC-ON PROfi le. In addition, more time was spent on emotional and cognitive domains. These are important fi ndings, since physicians may overlook psychosocial problems [10]. This was also the case for several HRQOL problems within the emotional and cognitive domain in the current study. In the control group, e.g., 39% of fear related problems remained unidentifi ed, whereas in the intervention group this amount was reduced to 10%. Hence, the QLIC-ON PROfi le is an eff ective and effi cient tool for systematically discussing and identifying psychosocial problems in paediatric oncology. This is in keeping with the fact that paediatricians often believe they are responsible for identifi cation of mental health issues [29]. Additionally, children with cancer and their parents have indicated in a previous study to feel the need for support and empathy, and suffi cient time for communication. They also preferred the oncologist to be well acquainted with the child and parents [30]. The QLIC-ON PROfi le can be helpful in meeting in these needs.

From clinical experience we know that physicians worry that the use of PROs will lengthen their consultation time. Earlier research has demonstrated no diff erences in consultation duration [1, 6], which is also the case in the current study. Although not signifi cant, we found that the use of the QLIC-ON PROfi le even shortens average consultation duration with almost 3 minutes. We believe that the QLIC-ON PROfi le improves effi ciency, because the oncologist is better equipped to focus quickly on current problems; HRQOL domains in which no problems are reported do not have to be discussed.

Despite a trend towards higher scores for the intervention group, the use of the QLIC-ON PROfi le had no eff ect on satisfaction. Parents were already satisfi ed with the consultation, regardless of QLIC-ON PROfi le use (ceiling eff ect [31]). Previous studies demonstrate mixed results with respect to satisfaction [1, 3, 8]. It is possible that the benefi ts of the QLIC-ON PROfi le will be more evident if it is implemented in clinical practice for a longer period of time.

The QLIC-ON PROfi le had no eff ect on psychosocial referrals. This corresponds with earlier studies [1, 2]. According to Greenhalgh et al. [32], PROs were seen as more useful in overall patient management and physician-patient communication and less so in treatment decision making and referrals. Moreover, children had only recently completed

chapter 6

112

treatment, and oncologists may therefore be more likely to adopt wait-and-see policy with fewer referrals as a consequence. Furthermore, children with problems during treatment had probably already been referred for psychosocial care.

A remarkable fi nding is that only patients of 5 to 7 years of age seem to profi t from the QLIC-ON PROfi le with respect to HRQOL. It is diffi cult to explain why specifi cally this age group improved while 0 to 4 year old and 8 to 18 year old children did not benefi t. The use of diff erent types of HRQOL instruments could clarify this diff erence; it is possible that the CHQ PF 50 is more sensitive in detecting changes in HRQOL than the other measures. However, because of the diff erent developmental stages it was impossible to use the same instrument for all age groups.

An explanation for the overall lack of explicit HRQOL improvement could be the life-threatening nature of cancer, compared to chronic diseases which are not necessarily life-threatening. This may impact on an oncologist’s chance to apply an holistic approach towards their patients which makes them sensitive to HRQOL problems regardless of the use of a PRO tool. Perhaps in paediatric oncology only longitudinal use of a PRO tool, from diagnosis through all treatment stages, will add to the eff ectiveness.

A strength of the current study is that it was a multi-centred study with a relatively large sample size, suffi cient to detect signifi cant diff erences. Additionally, our project was the fi rst PRO study involving parents and children aged 0-18 years in paediatrics in general, and in paediatric oncology specifi cally.

A limitation of our study was that participants were not randomised. However, due to the risk of contamination, randomisation was not desirable [16]. Randomising the child was not an option, because in that case the oncologist had to be trained in advance and be prepared for any child, regardless of control or intervention condition. As soon as oncologists received the QLIC-ON training and gained some experience in using the QLIC-ON PROfi le, bias could occur when seeing children in the control condition. Randomising the oncologists was also not possible, because of overlapping responsibilities. Children are frequently discussed in team meetings in which all oncologists participate and it is not uncommon that oncologists share the treatment of a child. Finally, randomising the centres would lead to bias as well. Only four centres participated in our study and only some contributed equally to the study (the number of participating patients and oncologists diff ered). Furthermore, from practice experience we know that there are some diff erences in hospital policy between the centres (e.g. shared care). Therefore, to prevent bias, a sequential cohort design was considered most suitable for the QLIC-ON study. Nevertheless, our study design might have caused bias in satisfaction outcome, because the control group was actually an attention-control group. A sequential cohort study with three groups (control, attention-control and intervention) would have been ideal, but this was not possible because of practical reasons (limited time and number of patients). Nevertheless, the applied sequential cohort design led to a comparable control and intervention group, since there were no diff erences in socio-demographics and

113

medical variables. Furthermore, it would have been interesting to have applied the PRO during instead of after treatment. It is possible that the eff ect of the intervention would have been more evident during the treatment phase, because children are probably experiencing more HRQOL problems then [33, 34]. In addition, secondary outcome HRQOL was assessed on average two weeks after the third consultation. Perhaps this interval was too short to measure eff ect of the intervention, but time did not allow more consultations or a follow-up assessment more than two weeks after consultation three. Finally, only one third of the sound recorded consultations were analysed, again due to time constraints.

A suggestion for future PRO research is to focus further on analysis of communication (sound recordings), with extension and improvement of the audio assessment form, instead of the more traditional outcomes (referrals and satisfaction). Additionally, other paediatric populations may profi t from PROs in clinical practice as well, using more advanced techniques e.g. PROs via internet [16, 35].

Overall, we conclude that a PRO is a helpful tool for systematic monitoring HRQOL of children with cancer, without lengthening the consultation. It is feasible and recommended to be implemented in clinical practice.

chapter 6

114

Reference List1. Detmar SB, Muller MJ, Schornagel JH, et al.

Health-related quality-of-life assessments and patient-physician communication: a randomized controlled trial. JAMA 2002:288:3027-3034.










11. Engelen V, Haentjens MM, Detmar SB, et al. Health related quality of life of Dutch children: psychometric properties of the PedsQL in the Netherlands. BMC Pediatr 2009:9:68.


13. Maurice-Stam H, Grootenhuis MA, Caron HN, Last BF. Course of life of survivors of childhood cancer is related to quality of life in young adulthood. J Psychosoc Oncol 2007:25:43-58.



16. Engelen V, Haverman L, Koopman H, et al. Development and implementation of a patient reported outcome intervention (QLIC-ON PROfi le) in clinical paediatric oncology practice. Patient Educ Couns 2010.



19. Fekkes M, Bruil J, Vogels T. TAPQOL-manual. In. Leiden: TNO Prevention and Health; 2003.

115

20. Ong LM, Visser MR, Lammes FB, de Haes JC. Doctor-patient communication and cancer patients’ quality of life and satisfaction. Patient Educ Couns 2000:41:145-156.

21. Zandbelt LC, Smets EM, Oort FJ, et al. Satisfaction with the outpatient encounter: a comparison of patients’ and physicians’ views. J Gen Intern Med 2004:19:1088-1095.

22. Raat H, Landgraf JM, Oostenbrink R, et al. Reliability and validity of the Infant and Toddler Quality of Life Questionnaire (ITQOL) in a general population and respiratory disease sample. Qual Life Res 2007:16:445-460.


24. Raat H, Bonsel GJ, Essink-Bot ML, et al. Reliability and validity of comprehensive health status measures in children: The Child Health Questionnaire in relation to the Health Utilities Index. J Clin Epidemiol 2002:55:67-76.

25. Ravens-Sieberer U, Schmidt S, Gosch A, et al. Measuring subjective health in children and adolescents: results of the European KIDSCREEN/DISABKIDS Project. Psychosoc Med 2007:4:Doc08.

26. Ravens-Sieberer U, Gosch A, Rajmil L, et al. The KIDSCREEN-52 quality of life measure for children and adolescents: psychometric results from a cross-cultural survey in 13 European countries. Value Health 2008:11:645-658.

27. Snijders FAB, Boske RJ. Multilevel Analysis. An introduction to basic and advanced multilevel modeling. London: SAGE publications Ltd; 2004.

28. Engelen V, Koopman HM, Detmar SB, et al. Health-related quality of life after completion of successful treatment for childhood cancer. Pediatr Blood Cancer 2010.



31. Darby C. Patient/parent assessment of the quality of care. Ambul Pediatr 2002:2:345-348.


33. Maurice-Stam H, Oort FJ, Last BF, et al. Longitudinal assessment of health-related quality of life in preschool children with non-CNS cancer after the end of successful treatment. Pediatr Blood Cancer 2008:50:1047-1051.

34. Varni JW, Burwinkle TM, Katz ER. The PedsQL in pediatric cancer pain: a prospective longitudinal analysis of pain and emotional distress. J Dev Behav Pediatr 2004:25:239-246.

35. Haverman L, Engelen V, Van Rossum MA, et al. Monitoring health-related quality of life in paediatric practice: development of an innovative web-based application. BMC Pediatr 2011:11:3.

chapter 6

1177Chapter 7

V. Engelen 1M.C.B. van Zwieten 2H.M. Koopman 3S.B. Detmar 4

H.N. Caron 5

P. Brons 6

R.M. Egeler 7G.J.L. Kaspers 8M.A. Grootenhuis 1

1 Psychosocial Department, Academic Medical Centre/ Emma Children’s Hospital, Amsterdam 2 Department of General Practice/ Medical Ethics, Division Clinical Methods and Public Health, Academic

Medical Centre, University of Amsterdam, Amsterdam3 Medical Psychology Department, Leiden University Medical Center, Leiden4 Prevention and Health Department, TNO, Leiden5 Paediatric Oncology Department, Academic Medical Centre/ Emma Children’s Hospital, Amsterdam6 Paediatric Oncology Department, Radboud University Medical Centre, Nijmegen 7 Paediatric Oncology Department, Leiden University Medical Centre, Leiden8 Paediatric Oncology/Hematology, VU University Medical Center, Amsterdam

The infl uence of patient reported outcomes on the discussion of psychosocial issues in children with cancer

SUBMITTED FOR PUBLICATION

118

Abstract

Background. This study investigates the eff ect of using patient reported outcomes (PROs) about health related quality of life (HRQOL) in clinical practice on the type (and amount) of psychosocial topics discussed during a paediatric oncology consultation.

Procedure. Children (N=193) with cancer participated in a sequential cohort intervention study, with a control (no PRO was used) and intervention group (a PRO was used). For each child three consecutive consultations with the paediatric oncologist were audio recorded in order to assess the discussed psychosocial topics. One third of the audio recordings were qualitatively analysed.

Results. The type of the discussed psychosocial topics in the control and intervention group did not diff er from each other. However, the discussion of psychosocial topics increased in the intervention group compared to the control group. In both groups, topics within the social domain occurred most frequently and topics regarding the emotional domain had the lowest incidence.

Conclusions. PROs do not change the psychosocial content of communication. Paediatric oncologists already address psychosocial issues during the consultation, regardless of the use of a PRO. However, with a PRO available they address these issues more systematically and more often.

119

Introduction

Over the years, health-related quality of life (HRQOL) has become an important outcome measure in paediatric oncology research. The World Health Organisation defi nes quality of life (QOL) as “individuals’ perceptions of their position in life, in the context of the culture and value systems in which they live, and in relation to their goals, expectations, standards and concerns” [1]. The concept of HRQOL refers to the impact of health and illness on an individual’s QOL [2]. In paediatrics, there are usually four HRQOL domains distinguished: physical, emotional, social and school (or cognitive) functioning, of which the last three can be summarized as psychosocial functioning [3].

Monitoring HRQOL by providing patient reported outcomes (PROs) in clinical practice has been the focus of several studies [4-10]. PROs are based on direct reporting by patients without the intervention of an observer. HRQOL assessment is a form of PRO that often includes both a patient’s functional status (physical and psychosocial) as well as his or her symptoms [11]. PROs can serve as an aid for the physician to systematically identify and discuss psychosocial problems. Positive eff ects of using PROs were found for e.g. communication, satisfaction with care and HRQOL [4, 5, 10]. Still, it remains diffi cult to prove eff ectiveness of PROs in clinical practice [12].

Children with cancer could also benefi t from the use of PROs in clinical practice. Although psychological adjustment of children with cancer is generally good [13], children at risk for psychosocial diffi culties in adapting to the disease are identifi ed [14] e.g. children with brain tumors. Additionally, several studies [15-21] demonstrate that childhood cancer diagnosis is related to poor HRQOL. Physicians are not always conscious of these problems, though [22, 23]. In a recent paediatric oncology PRO study (the QLIC-ON study) [24], we illustrated that PROs in clinical practice result in more discussion of emotional and psychosocial functioning without lengthening the consultation. Additionally, more emotional and cognitive problems were identifi ed using a PRO, compared to consultations in which no PRO was used. Based on these fi ndings, we were interested in the content of this diff erence, also to understand more about the eff ect of PROs. What types of psychosocial issues are discussed during the consultations and does the use of a PRO change the content of these psychosocial issues? Aim of the current study – which is a continuation the QLIC-ON study [24] - is to investigate the eff ect of using a PRO (about HRQOL) in clinical practice on the type (and amount) of psychosocial topics discussed during a paediatric oncology consultation.

Methods

InterventionThe intervention consisted of providing a PRO about HRQOL (the QLIC-ON PROfi le,

chapter 7

120

p. 173) to the oncologist. The QLIC-ON PROfi le presents four HRQOL domains (physical, emotional, social and cognitive functioning) by summarising the answers of child or parent on two generic HRQOL questionnaires: the PedsQL Generic Core Scale (Pediatric Quality of Life Inventory) [3, 25] or the TAPQOL (TNO-AZL Preschool children Quality of Life) [26, 27]. The PedsQL self-report form was used for children aged 8 to 18; parents of children aged 6 to 7 completed the PedsQL proxy-report. Dutch PedsQL norm scores were available [25]. The PedsQL has a short completion time (approximately 5-10 minutes) and good utility, validity and reliability [3]. The TAPQOL was used for parents of children aged 0 to 5 years (proxy-report). Dutch TAPQOL norm scores were available as well. Completion time is about 5-10 minutes and psychometric properties are satisfying [26, 27].

Paediatric oncologists were instructed to discuss the items in which problems were refl ected (the red and orange answers on the QLIC-ON PROfi le). To optimize the eff ectiveness of the QLIC-ON PROfi le, we developed a training which comprised an individual and a group session for paediatric oncologists, and a patient instruction [28]. The development and implementation of the QLIC-ON intervention (including the training) were extensively described in a previous study [28].

ParticipantsParticipants of the QLIC-ON study were children with cancer (0-18 years) shortly (0-3 months) after completion of treatment, their parents and treating paediatric oncologists. The QLIC-ON study is a large Dutch multi-centre research project in which the Academic Medical Centre/ Emma Children’s Hospital, Leiden University Medical Centre, Radboud University Medical Centre and the VU University Medical Center joined. The Medical Ethics Committees of all participating centres approved of the study.

ProcedureChildren and parents were invited by mail to participate as soon as the child had fi nished treatment. Children with cancer who had undergone stem cell transplantation (SCT) were approached for participation 6 months after SCT. Informed consent was obtained in the out-patient clinic shortly before the fi rst consultation. All measures were completed on paper, except for the QLIC-ON PROfi le which was fi lled out digitally on a laptop in the out-patient clinic prior to three consultations.

All consultations were audio recorded by means of an audio device in order to assess diff erent aspects of communication about HRQOL domains. After the consultation, the audio recording was digitally exported to a computer and saved as the unique child study ID including the date. Only the researchers had access to the key of this code, and they pledged secrecy.

Reasons for absent audio recordings were no parental or child consent for audio recording the consultation, drop-out due to relapse of the disease or practical reasons (e.g. technical problems). One third of the audio-tapes were (block) randomly selected

121

chapter 7

and analysed (Table 1). If available, each child was included once for analysis.

Table 1. Number of (analysed) audio-tapes per consultation

Control Intervention Total

Total audio-taped consultations (N) 216 237 453 Randomised audio-taped consultations (N) 77 78 155Audio-tapes consultation 1 (N) 26 26 52Audio-tapes consultation 2 (N) 24 29 53Audio-tapes consultation 3 (N) 27 23 50

Study designThe QLIC-ON PROfi le was completed by the child or parent and provided to the oncologist during three consecutive consultations by means of a sequential cohort design. As we described in a previous study [28], randomisation was regarded unsuitable due to the risk of contamination. The study consisted of a control period (QLIC-ON PROfi le was completed but not provided) and an intervention period (QLIC-ON PROfi le was completed and provided). Oncologists participated in the control as well as the intervention period; children and parents took part in either the control period or the intervention period. To prevent contamination of the control period, oncologists started the intervention period (with the training), only as soon as all children from the control period had fi nished their three consultations. Both groups were evaluated before (baseline) and during (consultation evaluation) the three consultations.

Measures

Socio-demographic and medical variables

Socio-demographic data of participating children and parents were collected at baseline. Child age and gender in the non-participants group were retrieved from medical fi les. Medical information of participants as well as non-participants was gathered by data managers or oncologists.

Socio-demographics of the oncologist were collected at the start of the study: age, gender and years of work experience as an oncologist. Communication about HRQOL topics

For the current study the (semi-open) audio-assessment form was developed to score the audio recordings of the consultations. The form entailed four HRQOL domains (physical, emotional, social and cognitive functioning), each containing open spaces on which the researcher depicted all topics that came up during the consultation. For example, topics such as ‘skin’, ‘sleeping’ and ‘walking’ were categorised within the physical domain; topics such as ‘fear’, ‘anger’ and ‘coping’ were included within the emotional domain; issues concerning ‘friends’, ‘family’ and ‘hobbies’ were classifi ed within the social domain and

122

fi nally, topics such as ‘school progress’, ‘teacher’ and ‘exams’ were written down within the cognitive domain. The few topics that could not be covered by the HRQOL domains were classifi ed within the ‘Other’ category. These were mainly issues unrelated to the child and its disease, e.g. the weather, traffi c or work of parent. Additionally, the audio-assessment form included some closed questions to rate the consultation, however these data are reported elsewhere [24].

The researcher that scored the consultations on the audio-assessment form remained as close as possible to the phrasing that the persons in the consultations used. Furthermore, the researcher was unaware if the consultation originated from the control or intervention group. To ensure reliable and valid scoring of the consultations a manual containing defi nitions and rules (regarding which types of topics belong to which of the domains) was applied.

Analysis

Socio-demographic and medical variables

Socio-demographic and medical data were analyzed with SPSS 16.0.2. The diff erence between participants and non-participants, as well as the control and intervention group regarding age group, gender, diagnosis, treatment, country of birth, education and employment was analysed by means of Chi2 tests. Diff erence in age and treatment duration was examined with t-tests. Communication about HRQOL topics

The software program MaxQDA was used as an aid for analysis. Where possible and appropriate, the principles of framework analysis [29] were applied by the fi rst author to the depicted topics gathered with the audio-assessment forms. Framework analysis was only partly applicable, though, because the collected topics already summarized parts of the conversation; the audio-assessment forms did not provide literal transcripts but a list of topics. Per HRQOL domain these topics were given codes, and if necessary sub codes, for the control and intervention group separately. This resulted in a code tree per HRQOL domain per group.

The number of times that a code occurred was counted per group (control vs intervention) and not per consultation. It is therefore possible that some codes came up more than once during a consultation, e.g. ‘sports’. If a child mentions to play soccer and judo and also goes to swimming class, than all three topics are depicted on the audio-assessment form and thus the frequency of the code ‘sports’ is 3. Subsequent sub codes are ‘soccer’, ‘judo’ and ‘swimming’.

Subsequently, codes with a frequency of less than four were excluded from analysis, since their value is limited. The remaining codes (and belonging sub codes) of both groups were then compared per HRQOL domain by the fi rst two authors to decide whether there was a substantive diff erence. The physical functioning domain was excluded from analysis,

123

chapter 7

because we were especially interested in psychosocial functioning (emotional, social and cognitive functioning).

Eventually the codes are presented (as topics) in tables divided over three categories: (1) topics that were discussed in both groups, (2) topics that were discussed in both groups, but the frequency of the topics is less than four in one of the groups and (3) topics discussed in only one group.

Results

Socio-demographic and medical variables274 children with cancer and their parents were approached to participate. Participants and non-participants diff ered signifi cantly with respect to child age (p<.01), radiotherapy (p<.05) and SCT (p<.01). For treatment duration, child gender, diagnosis, surgery and chemotherapy no statistically signifi cant diff erences were observed among the groups [24].

Eventually, 193 (70.4%) participants completed the baseline assessment. 99 children participated in the control group - 14.3% drop-out; 94 in the intervention group - 21.3% drop-out. Main reasons for drop-out were relapse of the disease and interference of end of data collection with unacceptable long intervals between the consultations (especially between consultation 2 and 3 in the intervention group. There were no statistically signifi cant diff erences with respect to socio-demographic and medical variables between participants in the control and the intervention group (Table 2).

A total of 34 paediatric oncologists took part; 47.1% were female, average age was 41.9 (SD 10.1) and mean work experience as an oncologist was 8.7 years (SD 7.6).

Emotional topicsEmotional topics that were discussed in the control and intervention group are presented in Table 3. In the control group the topics ‘tension’, ‘coping’ and ‘anger’ occurred most frequently during the consultations; in the intervention group ‘fear’, ‘sadness’ and ‘anger’ came up most often. ‘Emotional functioning’ and ‘blue’ were only discussed in the intervention group.

Social topicsTable 4 demonstrates the social topics that were discussed in the control and intervention group. In the control group as well as the intervention group the topics ‘sports’ and ‘holiday/travel’ had the highest frequency. Additionally, in the control group the topic ‘playing’ was also discussed quite often; in the intervention group the incidence of the topic ‘behaviour’ was fairly high. The topic ‘going out’ only came up in the intervention group.

124

Table 2. Socio-demographic and medical variables: control vs intervention group

Control Intervention N=99 N=94

Age (years) 99 9.13 5.07 94 9.41 5.03 Treatment duration (months) 1 98 13.27 9.37 92 14.02 8.98 Time off treatment (weeks) 2 98 7.17 4.15 93 8.04 4.88 N % N % Age Group (years) 0-4 27 27.3 25 26.6 5-7 20 20.2 16 17.0 8-11 19 19.2 18 19.1 12-18 33 33.3 35 37.2 Gender (female) 46 46.5 46 48.9 Diagnosis 3 Leukemia 35 34.4 34 36.2 Lymphoma 12 12.1 12 12.8 Brain tumor 8 8.1 8 8.5 Solid tumor 21 21.2 21 22.3 Bone tumor 16 16.2 13 13.8 Other 7 7.0 6 6.4 Treatment Surgery 53 54.1 48 51.6 Radiotherapy 31 31.6 24 25.8 Chemotherapy 97 99.0 87 93.5 SCT 13 13.3 5 5.4

Age (years) 97 41.05 6.90 93 40.72 7.38 N % N %

Gender (female) 77 81.05 73 81.1 Country of birth (Netherlands) 84 85.7 82 88.2 Education4 Low 21 21.4 16 17.4 Middle 50 51.0 46 50.0 High 27 27.6 30 32.6

Employed 65 66.3 60 64.5

CHILD N M SD N M SD


1-4 are explained on p.45.

125

chapter 7

Table 3. Type and frequency of discussed topics within the emotional HRQOL

domain: control vs intervention group

Control group Intervention group

Domain freq Domain freq Emotional 63 Emotional 114 Topic freq Topic freq anger 6 anger 12 tension 6 tension 11 coping 6 coping 6 cheerful 5 cheerful 7 fear 4 fear 18 crying 4 [crying] [1] [sadness] [3] sadness 13 [worries] [3] worries 6 emotional functioning 6 gloomy 8

Dis

cuss

ed in

bot

h gr

oups

Dis

cuss

ed in

bo

th g

roup

s bu

t fre

q <4

in o

ne g

roup

Dis

cuss

ed in

one

gro

up

Cognitive topicsThere was no diff erence in cognitive topics that occurred most frequently when comparing the control and intervention group (Table 5). In both groups the topics ‘school (unspecifi ed)’, ‘school attendance’ and ‘school progress’ had the highest incidence. The topic ‘fatigue’, however, only came up in the intervention group.

Other topicsWith respect to the ‘other’ category, there was only one topic that had a frequency higher than three, namely the ‘parents’, which was discussed in both groups: control (frequency 25) and intervention (frequency 13).

126

Table 4. Type and frequency of discussed topics within the social HRQOL domain:

control vs intervention group


Domain freq Domain freq Social 288 Social 308 Topic freq Topic freq sports 71 sports 98 holiday/ holiday/ travel 53 travel 46 playing 21 playing 12 friends 17 friends 23 siblings 16 siblings 14 family 16 family 7 trips 14 trips 5 behaviour 13 behaviour 25 parents 12 parents 14 birthday/ birthday/ party 10 party 11 social social interaction 7 interaction 16 (creative) (creative) hobby 7 hobby 8 “make a “make a wish” wish” activity 5 activity 7

job 4 [job] [3]

going out

Dis

cuss

ed in

bot

h gr

oups

Dis

cuss

ed in

bo

th g

roup

s bu

t fre

q <4

in o

ne g

roup

Dis

cuss

ed in

one

gro

up

127

chapter 7

Table 5. Type and frequency of discussed topics within the cognitive HRQOL

domain: control vs intervention group


Domain freq Domain freq

Cognitive 165 Cognitive 229 Topic freq Topic freq school (unspecified) school attendance school progress day care/ after school care exam/ test teacher intelligence/ school performance type of education/ continued education attention/ concentration school/ class transition additional guidance/ remedial teaching writing/ drawing reading

[speech (therapy)] [course]

[social- emotional functioning]

[homework]

[forgetting]

Dis

cuss

ed in

bot

h gr

oups

D

iscu

ssed

in b

oth

grou

ps

b

ut fr

eq <

4 in

one

gro

up

43

13

11

10

9

8

8

7

5

5

5

5

5

[3]

[3]

2

[1]

[1]

school (unspecified) school attendance school progress day care/ after school care exam/ test teacher intelligence/ school performance type of education/ continued education attention/ concentration school/ class transition additional guidance/ remedial teaching writing/ drawing [reading]

speech (therapy) course

social- emotional functioning

homework

forgetting

31

20

21

8

12

15

10

14

15

5

15

7

[2]

4

6

5

9

6

128

DiscussionThe current study demonstrates that using a PRO (about HRQOL) in clinical paediatric oncology practice increases the amount of psychosocial topics, but it does not infl uence the type of psychosocial topics discussed during a consultation.

The discussion of emotional, social and cognitive functioning increased in the intervention group compared to the control group. In both groups, topics within the social domain occurred most frequently and topics regarding the emotional domain had the lowest incidence. The amount of discussed topics within the ‘other’ category decreased in the intervention group in comparison with the control group. Overall, these fi ndings are in agreement with previous research [4, 5]. PROs have a positive impact on patient-physician communication; they increase systematic attention to psychosocial topics. From previous studies we know that eff ective patient–physician communication can be of crucial importance to psychosocial outcomes in patients with cancer [30, 31].

In general, the same type of topics regarding emotional, social and cognitive functioning came up during a paediatric oncology consultation, regardless of the use of a PRO. This implies that paediatric oncologists already address all the relevant psychosocial topics. In a previous interview study (submitted), we found that paediatric oncologists defi ned ‘psychosocial functioning’ as emotional (e.g. psychological well-being and coping), social (e.g. family, friends, peers, clubs and sports) and school functioning (e.g. school presence and school performance). This defi nition validates the results of the present study, since comparable topics were discussed during the consultations and it concerns the same sample of paediatric oncologists. Still, the interview study also taught us that paediatric oncologists sometimes perceive hesitations and uncertainties towards their task in psychosocial functioning, e.g. lack of time or expertise. The current study demonstrates that paediatric oncologists are well aware of - and address - psychosocial issues in children with cancer. However, these issues are not always consequently discussed during the consultation. Our results illustrate that PROs are a way to make sure that psychosocial functioning is systematically and not occasionally addressed during the consultation. Systematic attention for psychosocial functioning is important in order to identify children at risk for psychosocial diffi culties and the paediatric oncologist has a responsibility of identifying psychosocial diffi culties in children with cancer [32, 33]. Thus, PROs are helpful, but from practice experience and literature we know that physicians are sometimes reluctant or hesitant towards the use of PROs [34, 35]. Therefore, we underline that PROs are intended to help the physician identify and not solve psychosocial problems. Psychosocial problem solving is mostly beyond the scope of the expertise of a physician. In case of a psychosocial problem, the physician’s task is mainly limited to referring to psychosocial care [36] or giving tailored advice - if he feels competent.

Strength of this study is that it concerns the fi rst paper that qualitatively reports on communication about psychosocial topics within a paediatric oncology consultation. For

129

chapter 7

paediatric oncologists it is interesting to learn what type of psychosocial issues are usually discussed during a consultation, that these topics concern relevant issues and how PROs can help them discussing these issues more systematically. For PRO researchers this paper provides new qualitative information with respect to the eff ect of PROs on the content of communication. This was never investigated before. They can use our results to further prove the eff ectiveness of PROs, which is still diffi cult to establish [12]. Another vigour of our study is that it is a multi-centre study with a relatively large sample size.

Still, our study has some limitations. The researcher who analysed the consultations with the audio-assessment form was blind to the consultation (control or intervention group). However, it was fairly easy to recognise a consultation from the intervention group, since the QLIC-ON PROfi le is sometimes literally mentioned during the consultation. Nonetheless, we believe we could minimize the bias because analyses were performed according to the instructions in the manual. Furthermore, only one third of the audio recorded consultations (N=155) were analysed, due to time constraints. Although analysing all consultations (N=453) would have provided us with more data, we do not expect this would have changed or added to our current results, because we subtracted the consultations as random as possible. Additionally our results may not be representative of other countries, because the organisation and quality of paediatric oncology health care systems can vary.

A suggestion for future research is to examine the eff ect of e.g. physician’s gender and experience on patient-physician communication with the use of PROs. Physician’s level of training is known to have a confounding eff ect on doctor-patient interactions [36]. There are probably individual diff erences between physicians with respect to communication; however, these weren’t investigated in the current study. More knowledge about physician’s characteristics will lead to more knowledge of how to train physicians with diff erent characteristics to use PROs. Furthermore, it is important to note that we only used generic questionnaires (PedsQL and TAPQOL) as PROs. We strongly recommend future studies to include disease specifi c questionnaires as a PRO as well, to give the physician more specifi c information about the (eff ects of the disease on the) child. In addition, we advise to apply more advanced techniques such as PROs via internet [28, 37], which will benefi t the user friendliness of the PRO assessment.

We conclude that PROs do not change the psychosocial content of communication. Paediatric oncologists already address psychosocial issues during the consultation, regardless of the use of a PRO. However, with a PRO available they address these issues more systematically and more often.

130

Reference List1. WHO. WHOQOL-BREF. Introduction,











11. Osoba D. Translating the science of patient-reported outcomes assessment into clinical practice. J Natl Cancer Inst Monogr 2007:5-11.










21. Engelen V, Koopman HM, Detmar SB, et al. Health-related quality of life after completion of successful treatment for childhood cancer. Pediatr Blood Cancer 2011:56:646-653.

131


23. Fallowfi eld L, Ratcliff e D, Jenkins V, Saul J. Psychiatric morbidity and its recognition by doctors in patients with cancer. Br J Cancer 2001:84:1011-1015.

24. Engelen V, Detmar S, Koopman H, et al. Reporting health-related quality of life scores to physicians during routine follow-up visits of pediatric oncology patients: Is it eff ective? Pediatr Blood Cancer 2011.

25. Engelen V, Haentjens MM, Detmar SB, et al. Health related quality of life of Dutch children: psychometric properties of the PedsQL in the Netherlands. BMC Pediatr 2009:9:68.


27. Fekkes M, Bruil J, Vogels T. TAPQOL-manual. In. Leiden: TNO Prevention and Health; 2003.

28. Engelen V, Haverman L, Koopman H, et al. Development and implementation of a patient reported outcome intervention (QLIC-ON PROfi le) in clinical paediatric oncology practice. Patient Educ Couns 2010:81:235-244.

29. Pope C, Ziebland S, Mays N. Qualitative research in health care. Analysing qualitative data. BMJ 2000:320:114-116.



32. Wiener LS, Pao M, Kazak AE, et al. Quick reference for pediatric oncology clinicians: the psychiatric and psychological dimensions of pediatric cancer symptom management. In. Charlottesville, VA: IPOS Press; 2009.


34. Baars RM, van der Pal SM, Koopman HM, Wit JM. Clinicians’ perspective on quality of life assessment in paediatric clinical practice. Acta Paediatr 2004:93:1356-1362.

35. Velikova G, Awad N, Coles-Gale R, et al. The clinical value of quality of life assessment in oncology practice-a qualitative study of patient and physician views. Psychooncology 2008:17:690-698.



chapter 7

1338Chapter 8General Discussion

134

135

1. AimsThis PhD thesis presented the results of the Quality of Life In Childhood Oncology (QLIC-ON) study. Aims of the QLIC-ON study included:

• to investigate the eff ectiveness of an intervention that provides a patient reported outcome (PRO) tool about health related quality of life (HRQOL) (the QLIC-ON PROfi le, p. 173) to the paediatric oncologist (chapter 6);

• to examine the eff ect of using the QLIC-ON PROfi le in clinical practice on the type (and amount) of psychosocial topics discussed during a paediatric oncology consultation (chapter 7).

To provide more insight in the issues around PROs in paediatric oncology additional aims of this thesis included:

• to collect Dutch reference data of an international HRQOL measure (PedsQL) and to assess reliability, socio-demographic within-group diff erences and construct validity in the Netherlands (chapter 2);

• to determine HRQOL of children with cancer shortly after the end of successful treatment compared with normative values (chapter 3);

• to explore paediatric oncologists’ perception of their role in (1) discussing psychosocial functioning and identifying psychosocial problems and (2) providing emotional support to children with cancer (chapter 4);

• to provide a thorough description of the development and implementation of the QLIC-ON PROfi le in clinical paediatric oncology practice (chapter 5).

2. Main fi ndingsAn overview of the studies presented in this thesis is provided in Table 1. This paragraph describes the main fi ndings of each study.

In chapter 2 for 496 children from the general Dutch population HRQOL reference data were collected by means of the PedsQL. PedsQL total scores were 84.18 (group 5-7), 82.11 (group 8-12) and 82.24 (group 13-18). Internal consistency coeffi cients ranged from .53 to .85. With respect to construct validity, healthy children of group 5-7 and 13-18 scored signifi cantly higher than children with a chronic health condition on all subscales, except for emotional functioning. No signifi cant diff erences were found in group 8-12 regarding health. We conclude that the Dutch version of the PedsQL has adequate psychometric properties and can be used as a HRQOL instrument in paediatric research in the Netherlands.

Chapter 3 demonstrates that HRQOL in children with cancer (N=191, mean age 9.25,

chapter 8

136

SD 5.06, 47.1% female) shortly after completion of treatment and their parents can be impaired compared with the norm. Physical well-being was aff ected for all ages. Compared to normative values 0- to 7-year-olds with cancer were rated signifi cantly lower on the majority of the scales. In addition, 12 to 18 year old children with cancer had signifi cantly better HRQOL than the norm on social scales. Compared to chronically ill norms, children with cancer aged 8 to 18 years demonstrated no diff erences, except for 12- to 18-year-olds who experienced signifi cantly more physical limitations. Additionally, we found that HRQOL of parents of 0 to 7 year old children with cancer was poorer than the norm. We conclude that HRQOL in children with cancer can be impaired and that paediatric oncologists should be aware of these problems.

Twenty-four paediatric oncologists were interviewed for the purpose of chapter 4. In general, paediatric oncologists believed that it was their task to discuss psychosocial functioning, to identify psychosocial problems and to provide emotional support to children with cancer. Provided perceptions with respect to this task were e.g. that it is ‘inherent to being a paediatrician’ and that they apply ‘a holistic view on medicine’ and feel ‘responsible for psychosocial referrals’. Paediatric oncologists also indicated that this task can be limited due to e.g. their inherent lack of expertise, lack of time or by the physician’s personality. We conclude that paediatric oncologists felt responsible for providing psychosocial care to the best of their ability.

The QLIC-ON PROfi le is a PRO tool applied in paediatric oncology. Important considerations regarding the development and implementation of the QLIC-ON PROfi le intervention are reported in chapter 5. These considerations have also resulted in educational material. Our fi ndings add to current knowledge in PRO research. This chapter can be used as an initial guide for researchers and other professionals, who are interested in setting up PRO interventions in clinical practice.

Chapter 6 describes the eff ectiveness of the QLIC-ON PROfi le. The QLIC-ON PROfi le increased discussion of emotional functioning (control M=32.9 vs intervention M=47.4, p<0.05) and psychosocial functioning (M=56.9 vs M=63.8, p<0.05). Additionally more emotional problems remained unidentifi ed in the control compared to the intervention group, e.g. anger (control 26% vs intervention 3%, p<0.01), fear (14% vs 0%, p<0.01) and sadness (26% vs 0%, p<0.001). The intervention had no eff ect on satisfaction and referrals, but did improve HRQOL of patients 5 to 7 years of age with respect to self esteem (p<0.05), family activities (p<0.05) and psychosocial functioning (p<0.01). We conclude that a PRO is a helpful tool for systematic monitoring HRQOL of children with cancer, without lengthening the duration of the consultation.

Finally, chapter 7 demonstrates that the type of the discussed psychosocial topics in the QLIC-ON control and intervention group did not diff er from each other. However, the discussion of psychosocial topics increased in the intervention group compared to the control group. In both groups, topics within the social domain occurred most frequently and topics regarding the emotional domain had the lowest incidence. We conclude that

137

paediatric oncologists address psychosocial issues in clinical practice, but with a PRO available they address these issues more often.

3. Key messagesAltogether, the main fi ndings of this thesis lead to three key messages:

• PROs in clinical practice increase identifi cation and discussion of psychosocial problems

• Successful use of the QLIC-ON PROfi le in clinical practice: making the right choices• Paediatric oncologists feel responsible for the identifi cation of psychosocial problems

3.1. Refl ections on key messages

3.1.1. PROs in clinical practice increase identifi cation and discussion of


This thesis demonstrates that the use of PROs in paediatric oncology practice increased the discussion of emotional and psychosocial functioning, increased the time spent on emotional and cognitive domains and enhanced the identifi cation of emotional and cognitive problems, without lengthening the consultation duration (chapter 6).

It is known that eff ective patient–physician communication is of crucial importance to psycho-social outcomes in patients with cancer [1, 2]. Therefore, we regard the increased identifi cation and discussion of psychosocial problems as the most important eff ect of our PRO intervention. This is supported by a previous study which demonstrated that if adult cancer patients felt that more attention had been paid to psychosocial aspects of their illness they were more satisfi ed with their consultations [3]. Additionally, other research concluded that showing compassion by physicians can reduce anxiety in cancer patients [4]. However, oncologists in adult cancer care are not always recognising or dealing with emotional needs. If doctors do not recognise and acknowledge patients’ cues for emotional support, patients will be discouraged from seeking that support [5]. This is in line with a previous study [6] illustrating that the lack of psychosocial questions from oncologists demonstrates how emotional issues remained mostly uninvestigated. With a PRO tool available, psychosocial issues will gain more systematic attention which will eventually benefi t secondary outcomes such as HRQOL and overall psychosocial functioning, even though we haven’t been able to prove eff ectiveness of PROs on these secondary outcomes. However, if psychosocial problems are detected in an early stage, appropriate intervention can prevent the child from having more serious psychological diffi culties at a later time. This is supported by a previous study [7] which demonstrated that the benefi cial eff ects of a HRQOL PRO tool intervention in adolescents with diabetes largely disappeared one year after withdrawing the HRQOL monitoring procedure. The authors underscore the importance of integrating a HRQOL PRO tool in routine care for

chapter 8

138

Table 1 Overview of studies presented in this thesis

Aim

Ch

ap

ter

2C

ha

pte

r 3

Ch

ap

ter

4C

ha

pte

r 5

Ch

ap

ter

6C

ha

pte

r 7

Colle

ctin

g D

utch

re

fere

nce

data

of t

hePe

dsQ

L

Ass

essi

ng H

RQO

L of

ch

ildre

n w

ith c

ance

rsh

ortly

aft

er

com

plet

ion

of

trea

tmen

t

Sample

n=49

6

child

ren

from

ge

nera

l pop

ulat

ion

(5-1

8 ye

ars

of a

ge)

n=19

1

child

ren

with

canc

er

(0-1

8 ye

ars

of a

ge)

n=24

paed

iatr

ic

onco

logi

sts

n.a.

Con

trol

gro

up:

n=99

chi

ldre

n w

ith c

ance

r

Inte

rven

tion

grou

p:

n=94

chi

ldre

n w

ith c

ance

r

(0-1

8 ye

ars

of a

ge)

Con

trol

gro

up:

child

ren

with

can

cer

cons

ulta

tions

of n

=78

cons

ulta

tions

of n

=77

Inte

rven

tion

grou

p:

child

ren

with

can

cer

(0-1

8 ye

ars

of a

ge)

Methods Main findings

HRQ

OL

(Ped

sQL)

HRQ

OL

(ITQ

OL,

CH

Q P

F 50

an

d Ki

dscr

een)

Inte

rvie

ws

Qua

litat

ive

anal

ysis

n.a.

Com

mun

icat

ion

abou

t HRQ

OL

Iden

tifica

tion

of H

RQO

L pr

oble

ms

Sat

isfa

ctio

n R

efer

ral

HRQ

OL

(ITQ

OL,

CH

Q P

F 50

an

d Ki

dscr

een)

Sou

nd re

cord

ings

of

con

sulta

tions

Q

ualit

ativ

e an

alys

is

The

Dut

ch v

ersi

on o

fth

e Pe

dsQ

L ha

s ad

equa

te

psyc

hom

etric

pr

oper

ties

for

paed

iatr

ic re

sear

ch

Expl

orin

g pa

edia

tric

onco

logi

sts’

perc

eptio

nre

gard

ing

psyc

hoso

cial

func

tioni

ng o

f chi

ldre

nw

ith ca

ncer

Des

crib

ing

the

deve

lopm

ent a

ndim

plem

enta

tion

of th

e Q

LIC-

ON

PR

Ofil

e

Exam

inin

g th

e eff

ect

of th

e Q

LIC-

ON

PRO

file

on th

e ty

pe (a

ndam

ount

) of d

iscu

ssed

psyc

hoso

cial

topi

cs

Inve

stig

atin

g th

e eff

ectiv

enes

sof

the

QLI

C-O

N P

ROfil

e

HRQ

OL

in c

hild

ren

with

can

cer a

nd th

eir

pare

nts

can

be

impa

ired

com

pare

dw

ith n

orm

sho

rtly

afte

r end

of t

reat

men

t. H

RQO

L sh

ould

be

mon

itore

d in

clin

ical

pr

actic

e

Alth

ough

with

in li

mits

,pa

edia

tric

onc

olog

ists

be

lieve

it is

thei

rta

sk to

dis

cuss

ps

ycho

soci

al

func

tioni

ng, t

o id

entif

y ps

ycho

soci

al p

robl

ems

and

to p

rovi

de

emot

iona

l sup

port

in

child

ren

with

can

cer

This

cap

ter i

s pr

actic

al g

uide

for

sett

ing

up P

RO

inte

rven

tions

in

clin

ical

pra

ctic

e

The

QLI

C-O

N P

ROfil

e: in

crea

sed

disc

ussi

on o

f em

otio

nal a

nd p

sych

osoc

ial

func

tioni

ng im

prov

ed id

entifi

catio

n of

em

otio

nal p

robl

ems

had

no

effec

t on

cons

ulta

tion

dura

tion,

sa

tisfa

ctio

n an

d re

ferr

al im

prov

ed H

RQO

L of

5- t

o 7-

year

-old

s

The

type

psy

chos

ocia

l to

pics

doe

s no

t ch

ange

. Pae

diat

ric

onco

logi

sts

addr

ess

psyc

hoso

cial

issu

es in

cl

inic

al p

ract

ice,

but

w

ith a

PRO

ava

ilabl

eth

ey a

ddre

ss th

ese

issu

es m

ore

ofte

n

139

adolescents with diabetes. Benefi ts of PROs in clinical practice have been demonstrated before [8-17], but

it remains diffi cult to prove eff ectiveness [17-21]. The feedback of PROs to health professionals has, in some studies, had an impact on the process of care, with less evident impact on health outcomes [18]. This was also the case in the QLIC-ON study. Even though we incorporated a PRO and a training that were carefully developed, we did not fi nd a convincing eff ect of PROs on secondary outcomes HRQOL, satisfaction and referrals. How come? HRQOL was assessed on average two weeks after the third consultation. Perhaps this interval was too short to measure eff ect of the intervention, but time did not allow more consultations or a follow-up assessment more than two weeks after consultation three. Or maybe the HRQOL instruments we used were not sensitive enough to detect a diff erence. Another explanation could be that most children fi nished treatment months ago and that their HRQOL is close to normal again [22], which also makes it diffi cult to demonstrate a diff erence. The use of the QLIC-ON PROfi le had no eff ect on satisfaction either, despite a trend towards higher scores for the intervention group. Parents were already satisfi ed with the consultation, regardless of QLIC-ON PROfi le use (ceiling eff ect [23]). Previous studies demonstrate mixed results with respect to satisfaction [8, 10, 24]. It is possible that the benefi ts of the QLIC-ON PROfi le will be more evident if it is implemented in clinical practice for a longer period of time. Furthermore, our intervention had no eff ect on psychosocial referrals. This corresponds with earlier studies [8, 9]. According to Greenhalgh et al. [19], PROs were seen as more useful in overall patient management and physician-patient communication and less so in treatment decision making and referrals. Moreover, children had only recently completed treatment, and oncologists may therefore be more likely to adopt wait-and-see policy with fewer referrals as a consequence. Additionally, children with problems during treatment had probably already been referred for psychosocial care.

Nonetheless, the fact remains that PRO tools increase identifi cation and discussion of psychosocial problems in clinical practice, which benefi ts patient-physician communication. Therefore we recommend implementing PROs in clinical practice.

3.1.2. Successful use of the QLIC-ON PROfi le in clinical practice: making the

right choices

We believe that various factors contributed to the success of the QLIC-ON PROfi le (chapter 5). To begin with, the QLIC-ON study was the fi rst study to report HRQOL outcome on item level with traffi c light colours on the PRO tool. This gave the physician the opportunity to immediately focus on the actual problem. With the more traditional PROs, mostly only HRQOL scale scores were presented in graphs. This left the physician sometimes guess what the actual problem was within that HRQOL domain, since graphs are less specifi c compared to items.

Furthermore, as described in this thesis, the intervention consisted of more than simply

chapter 8

140

providing the QLIC-ON PROfi le to the paediatric oncologist. We developed an individual training and a group training, in which a decision tree (Figure 1) and case studies on video were applied. The decision tree was intended as an aid for interpretation of the QLIC-ON PROfi le; it distinguished three steps: (1) identifying a problem, (2) discussing the problem and (3) taking appropriate action towards the problem. Additionally, the case studies on video represented a real consultation of a child and parent visiting a medical specialist, using an actual QLIC-ON PROfi le. The cases were implemented in the group training according to the following three steps: before demonstration (1) the QLIC-ON PROfi le that occurred in the video was discussed and (2) a watching assignment was given, after demonstration (3) the use of the QLIC-ON PROfi le in the video was evaluated. In addition, each case had its own learning goal: case 1) use of the QLIC-ON PROfi le in general, case 2) use of the line graphs, and case 3) use of the decision tree.

The visual lay-out of the QLIC-ON PROfi le and the developed training can be considered innovative in PRO research. The training was necessary in order to introduce physicians to the PRO tool, to demonstrate them the advantages and to give them examples and instructions of how to use it. An additional purpose of the training was to increase commitment towards the use of PROs. However, to further enhance the chance of successful use of PROs in clinical practice, several other important decisions were made, e.g. with respect to the study design, logistics, the type of PRO questionnaires and the person that uses the PRO.

First of all, it was important to consider which study design was most appropriate for the QLIC-ON study. We applied a sequential cohort design with a control and intervention group. Randomisation was not desirable due to the risk of contamination. Randomising the child was not an option, because in that case the paediatric oncologist had to be trained in advance and be prepared for any child, regardless of control or intervention condition. As soon as paediatric oncologists received the training and gained some experience in using the QLIC-ON PROfi le, bias could occur when seeing children in the control condition. Randomising the paediatric oncologists was neither a possibility, because of overlapping responsibilities. Children are frequently discussed in team meetings in which all paediatric oncologists participate and it is not uncommon that paediatric oncologists share the treatment of a child. Finally, randomising the centres would lead to bias as well. Only four centres participated in our study and not all of them contributed equally to the study (the number of participating patients and oncologists diff ered). Furthermore, from practice experience we knew that there are some diff erences in hospital policy between the centres (e.g. shared care). Therefore, to prevent bias, a sequential cohort design was considered most suitable for the QLIC-ON study.

Secondly, the logistical organisation of a PRO intervention needs good thought. The QLIC-ON intervention was very labour-intensive. Every child completed the PRO during three consecutive consultations. These assessments took place on a standalone laptop in the outpatient clinic. A researcher was present during every assessment to set up all materials, to guide them through the procedure, to hand a copy of the QLIC-ON PROfi le

141

chapter 8

Figure 1. Decision tree

There are no problems?

Item (or domain) is green

Item (or domain) is red or orange

There are problems?

Can you tell more about it?

When does it bother you?

How often does it bother you?

(What is the reason, do you think?)

How do you feel about it?

What do you do about it? Does it help?

Do you feel supported?

Does it worry you?

Is there a question for help? No End

Worry

Support

Action/ Solution

Feeling/ Opinion

(Reason)

Frequency

Moment/ Location

Explore

QLIC-ON PROfile

Yes, that’s right

Yes, that’s right



Yes Dependent on severity and need

Monitor

Advice

Referral

(End)

(End)

(End)

Start

QLIC-ON Support: 020 - 566 82 25, [email protected]

End

1. Id

entit

y2.

Dis

cuss

3. T

ake

actio

n

142

to the paediatric oncologist and to make sure the consultation was audio taped. To increase the chance of successful implementation of PROs in clinical practice in the future, this logistical barrier can to be overcome by making use of the internet [25]. This gives respondents the opportunity to complete the questionnaire at home and physicians to retrieve the PRO on the computer in the consultation room.

Thirdly, one has to determine the type of questionnaires that are used in the PRO tool. In the QLIC-ON study the PedsQL and TAPQOL were employed. These instruments have a generic nature and therefore they also included items not directly relevant to the physician. Some items could have been left out e.g. “I miss school to go to the doctor or hospital” (PedsQL, cognitive domain). The answer to this question is for almost every child yes and there is nothing that the physician can do about it. Moreover, a generic PRO might lack essential information that for example would be provided by a disease specifi c HRQOL questionnaire, e.g. the PedsQL Cancer Module [26]. However, because participants of the QLIC-ON study had recently fi nished treatment this disease specifi c instrument was not suitable for our study sample. The PedsQL Cancer Module would be applicable if HRQOL was monitored during treatment. Overall, we advice that a PRO should be fl exible. Besides age appropriate (which was the case in the QLIC-ON study), it should also be tailored to the type of disease and treatment phase. Additionally, the time frame of the questionnaire applied as a PRO should be considered. Most HRQOL questionnaires have a time frame of one month. However, this was problematic for the QLIC-ON study since participants’ HRQOL was monitored during three consecutive consultations and for some children (especially children that had undergone stem cell transplantation) the time interval between the consultations was less than one month. For that reason we chose to use the PedsQL acute version which has a time frame of one week (chapter 2).

Finally, the person that is using the PRO must be taken into account. Before applying a PRO tool in clinical practice one should consider who the most appropriate user of the PRO is. This can be a physician, but also e.g. a nurse or a social worker. It depends on the content and purpose of the PRO. The user must be committed and equipped to work with the PRO. In this thesis, paediatric oncologists were the most suitable PRO users, since it concerned children in the follow-up trajectory after treatment and at that stage paediatric oncologists were the only medical contacts for them. However, if it concerned children during treatment, than e.g. a nurse practitioner would have been an adequate user of the PRO as well. On the whole, it is important to make sure that the PRO user is convinced of the benefi ts of the PRO tool. If the user lacks enthusiasm regarding the PRO, the willingness to successfully implement the PRO in clinical practice is very diffi cult. Often, a change in behaviour is required to make the use of PRO tools succeed. Additionally, a ‘user-ambassador’ in the PRO user group could be necessary to convince the group from within. Other ways to increase commitment have been described in chapter 5.

In sum, the use of PROs in clinical practice is all about making tailored choices. Development should take into account e.g. the type of disease, the type of treatment, the

143

type of PRO user and age of the patient. Every researcher or health professional that aims at successfully implementing a PRO in clinical practice should always bear these multiple choices in mind.

3.1.3. Paediatric oncologists feel responsible for the identifi cation of


It might seem obvious that most paediatric oncologists feel responsible for the identifi cation of psychosocial problems, after all - they are the child’s primary care taker. However, this is a statement based on assumptions; it was never actually investigated. From practice experience we knew that in general paediatric oncologists acknowledge that psychosocial problems can exist in children confronted with cancer, but - while developing and implementing the QLIC-ON intervention – we also noticed some hesitations regarding this issue: “I am a physician, not a psychologist” and “I want to discuss psychosocial functioning, but what if a cesspool of emotions opens up?”. Therefore, it was time to explore the paediatric oncologists’ perception regarding this topic.

This thesis demonstrates that paediatric oncologists believe that it is their task to discuss psychosocial functioning and to identify psychosocial problems in children with cancer. Paediatric oncologists also indicated, however, that fulfi lling this task can be limited due to e.g. lack of expertise or time (chapter 4).

In general, our fi ndings agree with previous research that has demonstrated how paediatricians consider themselves responsible in identifying mental health issues [27]. Furthermore, our study showed that paediatric oncologists’ perceptions match preferences of children with cancer and their parents who feel the need for support and empathy, and suffi cient time for communication [28]. From previous research we know that eff ective patient–physician communication is of crucial importance to psycho-social outcomes in patients with cancer [1, 2]. Additionally, paediatric oncologists’ views towards their task in psychosocial functioning met many of the important components of paediatric doctor-patient interactions [29], e.g. general communication skills, patient-parent-centeredness, interpersonal skills and liaisons with other health-care professionals. Still, paediatric oncologists also clearly indicated, that their task in providing psychosocial care is limited. For instance, some of them do not feel qualifi ed for solving psychosocial problems, which is in agreement with other research [30]. Most of them apply a low threshold for referral to psychosocial care. With respect to providing emotional support, some remain distanced from their patients and their patient’s family. Potentially confounding eff ects on doctor-patient interactions, such as the doctor’s level of training and workload and patient’s age [29] were also mentioned by our respondents. Other reasons for unsuccessful identifi cation of psychosocial problems were dependency on information, undisclosed problems, underestimated problems and lack of structural discussion of problems.

Since paediatric oncologists felt responsible for providing psychosocial care to the best of their abilities, it is important to fi nd ways to overcome these limitations.

chapter 8

144

Therefore, we suggest considering the use of PROs about psychosocial functioning during the consultation as a possible solution. PROs can serve as an aid for the physician to systematically identify and discuss psychosocial problems, without lengthening the consultation. The benefi ts of PROs in paediatric oncology practice were demonstrated in the present thesis (chapter 6).

The fact that paediatric oncologists feel responsible does not imply they are the only ones responsible for detecting psychosocial diffi culties. We underline that this is de responsibility of the entire medical team involved with the child: nurses, nurse practitioners, social workers, psychologists and - maybe even beyond the hospital - teachers. Still, as the paediatric oncologists indicated themselves, they have the role of being the coordinators of care. Psychosocial care is part of that.

During the implementation of the QLIC-ON intervention we sometimes noticed that there was wavering towards the use of the PRO tool. Paediatric oncologists said e.g. “I don’t need a PRO. I already discuss psychosocial issues in my practice” or “I have a busy practice, the use of a PRO takes too much time”. We understand and acknowledge their arguments; however, their hesitation is based on incorrect assumptions towards the use of PROs. Our suggestion to use a PRO is not a statement against the way in which paediatric oncologists currently provide psychosocial care. Paediatric oncologists already address psychosocial issues during the consultation, which is underlined by the present thesis (chapter 7). PROs can only add to that because it will enable them to monitor psychosocial functioning even more systematically, without increasing the duration of the consultation. To get used to the use of PRO tools we believe it is necessary to introduce PROs during the physicians’ education at university. To further increase commitment towards the use of PRO tools and to enhance the possibility of successful implementation physicians should get more actively involved with the development and implementation of PROs. In that way, physicians (or other users) can apply their preferences regarding the content and presentation of the PRO which will give the PRO a better chance of becoming part of routine practice.

4. LimitationsThe fi ndings of the QLIC-ON study must be considered in the light of some limitations. Most of them were already mentioned in paragraph 3 of this Discussion section.

First of all, there is a limitation to the study design we applied. A disadvantage of a sequential cohort design is that the control and intervention group were not assessed simultaneously. Therefore, our study is at risk for a history eff ect. However, the control and intervention group were contiguous and gradually ongoing over time, and treatment protocols remained comparable within the 3-year time span of the study. It would have been ideal to have a sequential cohort study with three groups (control, attention-control and intervention), but this was not possible because of limited time and limited number of

145

patients. Nevertheless, the applied sequential cohort design led to a comparable control and intervention group; there were no diff erences in socio-demographics and medical variables.

Additionally, as stated earlier, the logistical procedure of the QLIC-ON study was very labour-intensive. A PRO internet application would have been more user-friendly and more professional. However, the QLIC-ON study was one of the fi rst PRO studies in paediatrics and at the time of the start of the study (2005), the use of internet for PRO purposes was not common yet. Moreover, current PRO internet applications are based on experiences and advancing insights derived from studies such as QLIC-ON.

Finally, as mentioned before, it is remarkable that the eff ect of our PRO intervention on HRQOL was limited. Only 5 to 7 year old patients seemed to profi t from the QLIC-ON PROfi le with respect to some HRQOL domains. Perhaps the eff ect of the intervention would have been more evident during treatment, because children are probably experiencing more HRQOL problems [22, 31] than after treatment when the QLIC-ON PROfi le was applied. A general explanation for this overall lack of HRQOL improvement could be the life-threatening nature of cancer, compared to chronic diseases which are not necessarily life-threatening. This may impact on a paediatric oncologist’s chance to apply a holistic approach towards their patients which makes them sensitive to HRQOL problems regardless of the use of a PRO tool. Maybe in paediatric oncology only longitudinal use of a PRO tool, from diagnosis through all treatment stages, will add to the eff ectiveness.

5. Clinical implications

5.1. Implementation of PROs in clinical practiceBased on the fi ndings of this thesis the main clinical implication is to implement PROs in routine clinical paediatric oncology practice. This thesis can be used as an initial guideline in achieving this goal.

5.1.1. Implementation during all treatment stages

Although the present research focused on children with cancer after end of treatment, we advise to implement a PRO tool during all cancer treatment stages. In that way, psychosocial functioning can be monitored throughout the course of the disease. In the QLIC-ON study only out-patients were involved. However, in-patients will probably profi t even more of the benefi ts of PRO tools, since they have a greater risk of suff ering from psychosocial and HRQOL problems [22, 31] because they are still undergoing intense treatment.

We suggest administering the PRO tool once every three months in children with cancer, starting one month after diagnosis, continuing during treatment and follow-up. The user of the PRO should be a professional who has regular contact with the child and its family, who has a complete overview of the care program in which the child is involved

chapter 8

146

and who has a coordinating task within the multidisciplinary team.

5.1.2. Implementation in all paediatric populations

The use of PROs in clinical practice should not only be reserved to children with cancer. Children with chronic diseases (e.g. rheumatic arthritis, sickle cell disease, diabetes [24]) may also benefi t. The chronic nature of their disease makes them well suited candidates for PRO tools. The condition they suff er from is not immediately life-threatening; still the impact of the disease on their lives can be immense [32]. It is therefore of great importance that their psychosocial functioning is monitored. Additionally, compliance with therapy may benefi t from the use of PRO tools. Questions referring to medicine use can be incorporated into the PRO. This information will help the physician to provide the patient with more tailored advice regarding disease management.

5.1.3. Incorporating internet use

The PRO tools applied in research have mainly depended on the work and eff ort of one or more researchers (e.g. logistics, assessment, evaluation). In the reality of clinical practice, however, there is probably no budget available for this. Internet can provide a step towards a solution, since in the Netherlands 87% of the households has access nowadays [33]. Less staff is needed, if the PRO is directly linked to patient data (e.g. treatments, consultations) and emails with invitations and questionnaires are automatically generated. Recently, a PRO website (www.hetklikt.nu) was developed by our research group and tested in children with rheumatic arthritis [25].

6. Future perspectives

6.1. PRO implementation researchUp to now, PRO studies have mainly aimed at investigating the eff ectiveness of PROs in clinical practice. Even though it has sometimes been diffi cult to demonstrate eff ectiveness on health outcomes, there are quite a few studies that illustrated the eff ectiveness of PROs on the process of care. The results of the present thesis for example confi rm that the use of PROs in clinical practice increases identifi cation of HRQOL problems and improves patient-physician communication. Both factors are important for the short as well as the long term well-being of the patient. Our fi ndings – together with several other benefi ts of PROs - suggest taking PRO research one step further: future studies should shift from assessing eff ectiveness to assessing implementation of PROs in clinical practice. But how? We advice to use the Deming circle (plan-do-check-act cycle) as presented by the Netherlands Youth Institute [34]. The cycle is a guide through the process of change, namely the implementation of PROs in clinical practice. It will help the professional to determine the implementation question, to analyse the persons and setting involved, to determine the plan of how to implement and how to create support among users, to

147

actually facilitate PRO use, to monitor and evaluate, to anchor the use of PROs in clinical practice and subsequently to continuously adjust and improve.

6.2. Parent PROsLogically, PRO studies have focussed on patients. However, in paediatrics the role of the parent and the impact of childhood disease on the parent may not be overlooked. It is known that parenting children with chronic diseases is highly demanding and has practical and emotional consequences. In a previous study [35] HRQOL of caregivers of children with a chronic disease was determined. Among others, parents of children with end stage renal failure, metabolic diseases, asthma, muscular dystrophy, sickle cell disease and also parents of children treated for paediatric brain tumours were included. Results demonstrated that these parents had a signifi cantly lower HRQOL. Subgroup analysis showed lower HRQOL on sleep, social functioning, daily activities, vitality, positive emotions and depressive emotions in disease-specifi c groups. On average, 45% of the parents were at risk for HRQOL impairment. This implicates screening of parents at risk in clinical practice, and more attention for family functioning in both research and clinical practice. Additionally, research has indicated that parents of children treated for childhood cancer report signifi cantly more psychological distress than the norm about two months after end of treatment [36]. This is confi rmed by the results presented in the current thesis (chapter 3) which demonstrate that parents of 0- to 7-year-old children with cancer have signifi cantly more emotional problems and perceive the general health of their child as less positive shortly after fi nishing treatment.

Altogether, the psychosocial impact of childhood cancer on parents should not be underestimated and deserves attention. For future research we suggest monitoring psychosocial functioning of parents. This can be done with the use of PRO tools by e.g. a social worker at set moments throughout the course of cancer treatment and follow-up.

6.3. ScreeningIt is important to note that the QLIC-ON PROfi le was not intended as a psychosocial screening device but merely as a psychosocial monitoring tool. Thus no cut-off scores were utilized, and scale scores of the healthy norm population were only presented in the graphs to give the physician a sense of how the child scored compared to healthy peers. The purpose of a PRO monitoring tool in clinical practice is to help physicians and patients discuss and subsequently identify psychosocial problems; it is a communication aid. The purpose of a PRO screening tool in clinical practice is to actually screen for psychopathology with the use of cut-off scores; it is a ‘diagnostic’ aid. Both types of PRO tools diff er from each other with respect to the type of questionnaires and also with respect to the type of users. A PRO monitoring tool can be regarded as a low-profi le instrument (e.g. HRQOL questionnaires), which can be used by diff erent types of professionals (e.g. physicians, nurses, social workers) with relatively little training. It provides an indication of a problem,

chapter 8

148

since individual scores cannot actually be compared with norm scores. The scores can facilitate discussion and subsequently the professional can give tailored advice (if he feels competent) or refer to specialised psychosocial care. A PRO screening tool is much stricter (e.g. Strength and Diffi culties Questionnaire, SDQ [37]), since individual scores can be compared with the norm population. Therefore, they can only be used by professionals (e.g. psychologists) that are familiar with the screening instrument and that are able to interpret and (if necessary) to provide treatment related to the outcome.

Tools for screening on psychosocial problems in children with cancer are not available in the Netherlands. In order to adequately screen HRQOL problems as well as adequately intervene in children and families who are at greater risk for psychosocial maladjustment, consistent assessment with well validated instruments and evidence based care are required. Structured attention for these problems and identifi cation of children and parents at risk for mental health problems is important. Once these children and parents are identifi ed, tailored psychosocial support can be off ered at an early stage, aimed at minimizing long-term mental health problems.

In the United States, a psychosocial screening tool (the Psychosocial Assessment Tool, PAT) that closely fi ts with these requirements was developed. The PAT systematically screens for psychosocial problems in families after the diagnosis of cancer in their child by assessing distress, problems in beliefs regarding the disease, availability of social support, symptoms of traumatic stress, other psychosocial strains and other kind of severe concerns within the family of children with diff erent kinds of cancer [38, 39]. The PAT is the most frequently used screening instrument in paediatric oncology and it seems practically applicable but its reliability and its validity remains to be proven in the Netherlands. Therefore, we suggest future cancer PRO studies in the Netherlands to incorporate the PAT.

In addition to the three preceding paragraphs, our research group has recently obtained a KWF grant to implement monitoring of HRQOL by paediatric oncologists (as a direct follow-up of the QLIC-ON study) and to screen for psychosocial risks (with the PAT) in children with cancer and their families during treatment by their psychosocial team. The study will start in September 2011.

In conclusionThe QLIC-ON study was the fi rst study to investigate the eff ectiveness of PROs about HRQOL in clinical paediatric oncology practice. It concerned a multicentre project with a large sample size. The PRO intervention was carefully developed, using innovative methods with respect to the design and the use of the PRO. Positive eff ects were found for patient-physician communication and identifi cation of psychosocial problems. PROs are recommended to be implemented in clinical paediatric oncology practice.

149

Reference List1. Davidson R, Mills ME. Cancer patients’

satisfaction with communication, information and quality of care in a UK region. Eur J Cancer Care (Engl ) 2005:14:83-90.


3. Walker MS, Ristvedt SL, Haughey BH. Patient care in multidisciplinary cancer clinics: does attention to psychosocial needs predict patient satisfaction? Psychooncology 2003:12:291-300.

4. Fogarty LA, Curbow BA, Wingard JR, et al. Can 40 seconds of compassion reduce patient anxiety? J Clin Oncol 1999:17:371-379.

5. Butow PN, Brown RF, Cogar S, et al. Oncologists’ reactions to cancer patients’ verbal cues. Psychooncology 2002:11:47-58.

6. Ford S, Fallowfi eld L, Lewis S. Doctor-patient interactions in oncology. Soc Sci Med 1996:42:1511-1519.

7. de WM, Delemarre-van de Waal HA, Bokma JA, et al. Follow-up results on monitoring and discussing health-related quality of life in adolescent diabetes care: benefi ts do not sustain in routine practice. Pediatr Diabetes 2009.






13. Berry DL, Blumenstein BA, Halpenny B, et al. Enhancing patient-provider communication with the electronic self-report assessment for cancer: a randomized trial. J Clin Oncol 2011:29:1029-1035.

14. Santana MJ, Feeny D, Johnson JA, et al. Assessing the use of health-related quality of life measures in the routine clinical care of lung-transplant patients. Qual Life Res 2010:19:371-379.


16. Wu WW, Johnson R, Schepp KG, Berry DL. Electronic Self-report Symptom and Quality of Life for Adolescent Patients With Cancer: A Feasibility Study. Cancer Nurs 2011.

17. Snyder CF, Aaronson NK. Use of patient-reported outcomes in clinical practice. Lancet 2009:374:369-370.

18. Valderas JM, Alonso J, Guyatt GH. Measuring patient-reported outcomes: moving from clinical trials into clinical practice. Med J Aust 2008:189:93-94.


20. Greenhalgh J, Meadows K. The eff ectiveness of the use of patient-based measures of health in routine practice in improving the process and outcomes of patient care: a literature review. J Eval Clin Pract 1999:5:401-416.

chapter 8

150


22. Maurice-Stam H, Oort FJ, Last BF, et al. Longitudinal assessment of health-related quality of life in preschool children with non-CNS cancer after the end of successful treatment. Pediatr Blood Cancer 2008:50:1047-1051.

23. Darby C. Patient/parent assessment of the quality of care. Ambul Pediatr 2002:2:345-348.







30. Detmar SB, Aaronson NK, Wever LD, et al. How are you feeling? Who wants to know? Patients’ and oncologists’ preferences for discussing health-related quality-of-life issues. J Clin Oncol 2000:18:3295-3301.

31. Varni JW, Burwinkle TM, Katz ER. The PedsQL in pediatric cancer pain: a prospective longitudinal analysis of pain and emotional distress. J Dev Behav Pediatr 2004:25:239-246.


33. Statistics Netherlands. www.cbs.nl. In: 2009.34. Netherlands Youth Institute.

http://www.nji.nl/eCache/DEF/1/19/241.html. In: 2011.

35. Hatzmann J, Heymans HS, Carbonell A, et al. Hidden consequences of success in pediatrics: parental health-related quality of life--results from the Care Project. Pediatrics 2008:122:e1030-e1038.


37. Bourdon KH, Goodman R, Rae DS, et al. The Strengths and Diffi culties Questionnaire: U.S. normative data and psychometric properties. J Am Acad Child Adolesc Psychiatry 2005:44:557-564.

38. Pai AL, Patino-Fernandez AM, McSherry M, et al. The Psychosocial Assessment Tool (PAT2.0): psychometric properties of a screener for psychosocial distress in families of children newly diagnosed with cancer. J Pediatr Psychol 2008:33:50-62.

39. Alderfer MA, Mougianis I, Barakat LP, et al. Family psychosocial risk, distress, and service utilization in pediatric cancer: predictive validity of the Psychosocial Assessment Tool. Cancer 2009:115:4339-4349.

151SSummary

152

153

BackgroundThe use of patient reported outcomes (PROs) in routine clinical practice is becoming increasingly common. PROs are answers to questionnaires based on direct reporting by patients without the intervention of an observer. Health-related quality of life (HRQOL) assessment is a form of PRO which often includes physical, emotional, social and cognitive functioning of the patient. By using PROs in clinical practice, the HRQOL of patients can be systematically monitored. PRO studies are usually conducted in adult patients and the results are promising. Thus far however, there has been little attention for PROs in paediatric practice. This is remarkable since it is known that children with a (chronic) disease are at risk for suff ering from HRQOL problems and that physicians are not always aware of these problems.

Children with cancer can also experience HRQOL problems. PRO tools could bridge the gap between child and paediatric oncologist and serve as an aid in identifying and discussing HRQOL problems. Additionally, PRO tools may contribute to eff ective patient–physician communication, which is of crucial importance to psycho-social outcomes in children with cancer.

This thesis covers the Quality of Life In Childhood ONcology (QLIC-ON) study. The main aim of the QLIC-ON study was to investigate the eff ectiveness of an intervention that provides a PRO about HRQOL (the QLIC-ON PROfi le, p. 173) to the paediatric oncologist. More information about the background of the QLIC-ON study is given in the general introduction (chapter 1) of this thesis. It elaborates on childhood cancer, HRQOL and psychosocial functioning of children with cancer. A summary of research fi ndings regarding the eff ectiveness of PROs in clinical practice is also provided. The study aims, study design and intervention are described in this chapter as well.

Methods

A sequential cohort study design was applied in the QLIC-ON study. Paediatric oncologists participated in the control (March 2006 to January 2008) as well as the intervention period (January 2008 to November 2009); children with cancer (0-18 years, shortly after completion of treatment) and their parents participated in either the control period or the intervention period. First, a control period was introduced, in which 99 children or parents (if the child was younger than 8 years of age) completed a HRQOL questionnaire on a standalone computer while waiting in the out-patient clinic prior to three consecutive consultations with the paediatric oncologist. Answers on the HRQOL questionnaire were summarized in the QLIC-ON PROfi le. In the control period the QLIC-ON PROfi le was not provided to the paediatric oncologist, the child or the parent. The paediatric oncologist started the intervention period – with a training - only as soon as he had seen all patients

summary

154

participating in the control period (to prevent contamination bias). The intervention period consisted of a group of 94 children or parents fi lling out the HRQOL questionnaire. This time however, the QLIC-ON PROfi le was provided to the paediatric oncologist, the child and the parent, to be discussed during the three consultations. All participants completed (evaluating) questionnaires after each consultation; there was also a baseline and follow-up assessment. In addition, the consultations were audio taped. The eff ectiveness of the QLIC-ON PROfi le was assessed with the outcome measures communication about HRQOL domains, identifi cation of HRQOL problems, the child’s HRQOL, the amount of psychosocial referrals and satisfaction with communication.

ResultsBefore presenting the main results of the QLIC-ON study, this thesis fi rst reports on four related studies.

To begin with, the reference data of the Dutch version of the PedsQL are described in chapter 2. The PedsQL is a widely used generic paediatric HRQOL measure (which was applied as a PRO in the QLIC-ON study). It was therefore important to investigate the PedsQL within the general Dutch population. For 496 children (5 to 18 years) PedsQL reference data were collected. The psychometric properties of the PedsQL proved to be adequate and it can therefore be used as a HRQOL instrument in paediatric research in the Netherlands.

Secondly, chapter 3 demonstrates that the HRQOL of children with cancer (0 to 18 years, N=191) shortly after completion of treatment can be impaired compared with the norm. This is also the case for parents of children with cancer aged 0 to 7 years. HRQOL was assessed by means of the ITQOL (generic, proxy-report, 0 to 4 years), the CHQ PF 50 (generic, proxy-report, 5 to 7 years), the Kidscreen (generic, self-report, 8 to 18 years) and the Disabkids (chronic generic, self-report, 8 to 18 years). The fi ndings in this chapter underline the importance of systematically monitoring HRQOL in paediatric oncology, not only during but also after end of treatment and not only in children but also in parents.

Thirdly, 24 paediatric oncologists were interviewed for the purpose of the QLIC-ON study in chapter 4. Data were qualitatively analyzed by means of a framework analysis. In general, paediatric oncologists believed that it was their task to discuss psychosocial functioning, to identify psychosocial problems and to provide emotional support to children with cancer. However, they also indicated that this task is limited due to e.g. their inherent lack of expertise, lack of time or by the physician’s personality. It is concluded that paediatric oncologists feel responsible for providing psychosocial care to the best of their ability.

Fourthly, an extensive description of the development and implementation of the QLIC-ON intervention as well as the rationale behind the study design and the content of the training is provided in chapter 5. This chapter can be used as an initial guide for

155

researchers and other professionals, who are interested in setting up PRO interventions in clinical practice.

The main results of the QLIC-ON study are reported in chapter 6. They demonstrate that the QLIC-ON PROfi le increased discussion of emotional and psychosocial functioning without lengthening the consultation duration. Additionally more emotional problems were identifi ed in the intervention group compared to the control group e.g. problems related to anger, fear and sadness. Furthermore, the QLIC-ON PROfi le improved HRQOL of patients 5 to 7 years of age with respect to self esteem, family activities and psychosocial functioning. However, the QLIC-ON PROfi le had no eff ect on the amount of psychosocial referrals. Neither did the QLIC-ON PROfi le improve parent and physician satisfaction regarding communication, possibly because this was already high. Overall, it is concluded that a PRO is a useful tool for systematic monitoring HRQOL of children with cancer.

Finally, the fi ndings of a qualitative analysis of one third (N=155) of the audio recorded consultations are presented in chapter 7. It appeared that the QLIC-ON PROfi le did not change the psychosocial content of communication. The nature of the psychosocial topics in the intervention group was equal to the nature of the psychosocial topics in the control group. However, the discussion of psychosocial topics increased in the intervention group compared to the control group. Thus, paediatric oncologists already address psychosocial issues, regardless of the use of a PRO. However, with a PRO available they address these issues more systematically and more often.

DiscussionThe fi ndings presented in this thesis are discussed in the general discussion (chapter 8). The fi ndings lead to three key messages:

• PROs in clinical practice increase identifi cation and discussion of psychosocial problems

• Successful use of the QLIC-ON PROfi le in clinical practice: making the right choices• Paediatric oncologists feel responsible for the identifi cation of psychosocial problems

Chapter 8 refl ects on these three key messages. Limitations in study design, logistics and study population are described as well. Additionally, the general discussion elaborates on the clinical implication to implement PROs in routine clinical practice, incorporating the use of internet, during all treatment stages of childhood cancer. It is also advised to implement PROs within other paediatric populations. The chapter ends with future perspectives that focus on PRO implementation research (instead of investigating eff ectiveness), a PRO for parents of children with a (chronic) disease and developing a PRO screening tool (in addition to the current monitoring tool).

summary

156

ConclusionThe QLIC-ON study is the fi rst study to investigate the eff ectiveness of PROs about HRQOL in paediatric oncology practice. It concerned a multicentre project with a large sample size. The PRO intervention was carefully developed, using innovative methods with respect to the design and the use of the PRO. Positive eff ects were found for communication and identifi cation of psychosocial problems. PROs are recommended to be implemented in paediatric oncology practice.

157SSamenvatting

158

159

AchtergrondHet gebruik van patient reported outcomes (PROs) in de klinische praktijk komt steeds vaker voor. PROs zijn antwoorden op vragenlijsten direct gerapporteerd door de patiënt zonder tussenkomst van een derde persoon. Een kwaliteit van leven (KvL) meting is een vorm van PRO die meestal het fysiek, emotioneel, sociaal en cognitief functioneren van de patiënt weergeeft. Door PROs in de klinische praktijk te gebruiken, kan de KvL van patiënten systematisch gemonitord worden. PRO studies worden vooral gedaan bij volwassen patiënten en de resultaten zijn veelbelovend. Tot nu toe was er weinig aandacht voor PROs in de kindergeneeskunde. Dit is opmerkelijk, want het is bekend dat kinderen met een (chronische) ziekte risico lopen op KvL problemen en dat artsen zich hier niet altijd van bewust zijn.

Kinderen met kanker kunnen ook KvL problemen ondervinden. Een PRO instrument zou als hulpmiddel kunnen dienen bij het identifi ceren en bespreken van deze problemen. Daarnaast zou een PRO instrument een bijdrage kunnen leveren aan eff ectieve communicatie tussen de patiënt en de arts, wat van cruciaal belang is voor psychosociale uitkomsten bij kinderen met kanker.

Dit proefschrift presenteert de Quality of Life In Childhood ONcology (QLIC-ON) studie. Het hoofddoel van de QLIC-ON studie is het onderzoeken van de eff ectiviteit van een interventie waarbij een KvL PRO (het QLIC-ON PROfi el, p. 173) tijdens een kinderoncologisch consult gebruikt wordt.

Meer informatie over de achtergrond van de QLIC-ON studie is te vinden in de algemene introductie (hoofdstuk 1) van dit proefschrift. Dit hoofdstuk gaat over kinderkanker, KvL en psychosociaal functioneren van kinderen met kanker. Een samenvatting van onderzoeksbevindingen betreff ende de eff ectiviteit van PROs in de klinische praktijk wordt ook gegeven. Tevens worden de doelen van het onderzoek, het studiedesign en de interventie in dit hoofdstuk beschreven.

Methoden

In het QLIC-ONderzoek is een sequentieel cohort studie design toegepast. Daarbij namen kinderoncologen aan zowel de controle- (maart 2006 tot januari 2008) als interventieperiode (januari 2008 tot november 2009) deel; kinderen met kanker (0-18 jaar, kort na einde behandeling) en hun ouders namen alleen aan de controle- of de interventieperiode deel. Er werd gestart met de controleperiode waarbij 99 kinderen of ouders (indien het kind jonger dan 8 jaar was) een KvL vragenlijst invulden op een laptop in de wachtkamer vlak voorafgaand aan drie achtereenvolgende consulten met de kinderoncoloog. De antwoorden op de KvL vragenlijst werden vervolgens samengevat op het QLIC-ON PROfi el. In de controleperiode werd het QLIC-ON PROfi el niet aan de

samenvatting

160

kinderoncoloog, het kind of de ouder getoond. Om de interventie zuiver te houden begonnen de kinderoncologen pas met de interventieperiode - met een training - als ze alle kinderen uit de controleperiode gezien hadden. De interventieperiode bestond uit een groep van 94 kinderen of ouders die de KvL vragenlijst in de wachtkamer invulden. Deze keer werd het QLIC-ON PROfi el wel aan de kinderoncoloog, het kind en de ouder meegegeven, zodat het besproken kon worden tijdens het consult. Alle deelnemers vulden na afl oop van de consulten (evaluerende) vragenlijsten in; er was tevens een begin- en een eindmeting. Daarnaast werden er geluidsopnamen van de consulten gemaakt. De eff ectiviteit van het QLIC-ON PROfi el werd gemeten door middel van de uitkomstmaten communicatie over KvL domeinen, signalering van KvL problemen, KvL van het kind, het aantal psychosociale verwijzingen, het type besproken psychosociale onderwerpen en tevredenheid met communicatie.

ResultatenVoordat de belangrijkste resultaten van het QLIC-ONderzoek worden gepresenteerd, rapporteert dit proefschrift eerst over vier gerelateerde studies.

Om te beginnen, worden de normdata van de Nederlandse versie van de PedsQL beschreven in hoofdstuk 2. De PedsQL is een veelgebruikt generiek pediatrisch KvL instrument (welke in de QLIC-ON studie toegepast is als PRO). Het was daarom belangrijk om de PedsQL binnen de algemene Nederlandse populatie te onderzoeken. Bij 496 kinderen (5 tot 18 jaar) zijn PedsQL normdata verzameld. De psychometrische eigenschappen van de PedsQL blijken adequaat en het kan daarom gebruikt worden als KvL instrument voor pediatrisch onderzoek in Nederland.

Ten tweede laat hoofdstuk 3 zien dat de KvL van kinderen met kanker (0 tot 18 jaar, N=191) kort na afronding van de behandeling aangedaan kan zijn vergeleken met de norm. Dit is ook het geval bij de ouders van kinderen met kanker in de leeftijd van 0 tot 7 jaar. KvL werd onderzocht door middel van de ITQOL (generiek, ouderrapportage, 0 tot 4 jaar), de CHQ PF 50 (generiek, ouderrapportage, 5 tot 7 jaar), de Kidscreen (generiek, zelfrapportage, 8 tot 18 jaar) en de Disabkids (chronisch generiek, zelfrapportage, 8 tot 18 jaar). De bevindingen in dit hoofdstuk onderstrepen het belang van systematisch monitoren van KvL in de kinderoncologie, niet alleen tijdens de behandeling maar ook daarna en niet alleen bij kinderen maar ook bij hun ouders.

Ten derde werden 24 kinderoncologen geïnterviewd in het kader van het QLIC-ONderzoek in hoofdstuk 4. Data werden kwalitatief geanalyseerd door middel van een frameworkanalyse. Over het algemeen bleek dat kinderoncologen van mening waren dat het hun taak is om psychosociaal functioneren te bespreken, psychosociale problemen te signaleren en emotionele steun te bieden aan kinderen met kanker. Zij gaven echter ook aan dat deze taak beperkt was door bijvoorbeeld gebrek aan expertise en tijd of door de persoonlijkheid van de arts. Geconcludeerd wordt dat kinderoncologen zich

161

verantwoordelijk voelen voor het verschaff en van psychosociale zorg voor zo ver dit binnen hun vermogen ligt.

Ten vierde wordt in hoofdstuk 5 een uitgebreide beschrijving van de ontwikkeling en implementatie van de QLIC-ON interventie gegeven alsmede de gedachte achter het studiedesign en de inhoud van de training. Dit hoofdstuk kan gebruikt worden als een eerste handleiding voor onderzoekers en andere professionals die geïnteresseerd zijn in het opzetten van PRO interventies in de klinische praktijk.

De belangrijkste resultaten van het QLIC-ONderzoek worden gerapporteerd in hoofdstuk 6. De resultaten laten zien dat er meer aandacht was voor het emotioneel en psychosociaal functioneren als het QLIC-ON PROfi el gebruikt werd, zonder dat het consult hierdoor langer duurde. Verder werden er meer emotionele problemen gesignaleerd in de interventiegroep in vergelijking met de controlegroep, bijvoorbeeld problemen gerelateerd aan boosheid, angst en verdriet. Daarnaast verbeterde de KvL van 5 tot 7 jarige patiënten door het gebruik van het QLIC-ON PROfi el op het gebied van zelfvertrouwen, familieactiviteiten en psychosociaal functioneren. Het gebruik van het QLIC-ON PROfi el had echter geen eff ect op het aantal psychosociale verwijzingen. Ook trad er geen verbetering op in tevredenheid van de ouder en de kinderoncoloog wat betreft communicatie, waarschijnlijk omdat deze al hoog was. Over het algemeen wordt geconcludeerd dat een PRO een nuttig hulpmiddel is voor het systematisch monitoren van KvL bij kinderen met kanker.

Tot slot worden de bevindingen van de kwalitatieve analyse van een derde (N=155) van de geluidsopnamen van de consulten gepresenteerd in hoofdstuk 7. Het blijkt dat het QLIC-ON PROfi el de psychosociale inhoud van de communicatie niet verandert. De aard van de psychosociale onderwerpen in de interventiegroep was gelijk aan de aard van de psychosociale onderwerpen in de controlegroep. Echter, er werd in de interventiegroep wel vaker over psychosociale onderwerpen gesproken vergeleken met de controlegroep. Kinderoncologen laten psychosociale onderwerpen dus aan de orde komen, ongeacht het gebruik van een PRO. Echter, deze onderwerpen komen wel vaker en meer systematisch aan de orde als ze een PRO beschikbaar hebben.

DiscussieDe bevindingen die dit proefschrift presenteert, worden besproken in de algemene discussie (hoofdstuk 8). De bevindingen leiden tot drie kernboodschappen:

• PROs in de klinische praktijk zorgen voor een toename van het signaleren en bespreken van psychosociale problemen

• Succesvol gebruik van het QLIC-ON PROfi el in de klinische praktijk: een kwestie van de juiste keuzes maken

samenvatting

162

• Kinderoncologen voelen zich verantwoordelijk voor het signaleren van psychosociale problemen

Hoofstuk 8 refl ecteert op deze drie kernboodschappen. Beperkingen in studiedesign, logistiek en studiepopulatie worden ook beschreven. Verder wijdt de algemene discussie uit over de klinische implicatie PROs te implementeren in de klinische praktijk, waarbij het gebruik van internet geïncorporeerd wordt, gedurende elke fase van behandeling bij kinderkanker. Tevens wordt geadviseerd om PROs binnen andere pediatrische populaties te implementeren. Het hoofdstuk eindigt met toekomstperspectieven die zich focussen op PRO implementatieonderzoek (in plaats van eff ectiviteitonderzoek), een PRO voor ouders van kinderen met een (chronische) ziekte en het ontwikkelen van een PRO screeninginstrument (in aanvulling op het huidige monitoringinstrument).

ConclusieHet QLIC-ONderzoek is de eerste studie die eff ectiviteit van PROs over KvL in de klinische pediatrische oncologiepraktijk onderzocht heeft. Het betrof een multicenter project met een aanzienlijke steekproefgrootte. De PRO interventie is zorgvuldig ontwikkeld, waarbij gebruik is gemaakt van innovatieve methodieken wat betreft het ontwerp en gebruik van de PRO. Positieve eff ecten werden gevonden voor communicatie en signalering van psychosociale problemen. Aanbevolen wordt PROs te implementeren in de pediatrische oncologische praktijk.

163DDankwoord

164

165

DankwoordDit proefschrift beschrijft het QLIC-Onderzoek, een multi-centerstudie. Multi mensen om te bedanken dus!

Allereerst, en bovenal, dank ik alle kinderen met kanker en hun ouders die aan het QLIC-ONderzoek hebben deelgenomen. Het is niet niks om (weer) aan een onderzoek mee te doen, terwijl je net klaar bent met je behandeling. Ik realiseer me dat het afronden van de behandeling een bijzondere en spannende periode kan zijn. Dat jullie desondanks de tijd en moeite hebben genomen om aan onze studie deel te nemen, waardeer ik zeer.

Ook wil ik alle kinderen, ouders en scholen bedanken voor hun bijdrage aan de PedsQL studie (hoofdstuk 2). Met jullie medewerking hebben we onze interventie nog beter kunnen uitvoeren.

Dit proefschrift was er niet gekomen zonder de medewerking van alle kinderoncologen uit het EKZ/AMC, LUMC, St. Radboud en VUmc. Het volgen van trainingen, het toepassen van het QLIC-ON PROfi el, het invullen van vragenlijsten, het beantwoorden van de vele e-mails en het geven van een interview... We hebben veel van jullie gevraagd. Hartelijk dank voor alle inspanningen!

In het bijzonder wil ik Prof.dr. H.N. Caron bedanken voor het faciliteren van mijn onderzoek in het AMC en de ruimte die we daarin kregen om het naar wens en behoefte vorm te geven. Prof.dr. R.M. Egeler, Prof.dr. P.M. Hoogerbrugge en Prof.dr. G.J.L. Kaspers dank ik voor de gelegenheid om het QLIC-Onderzoek uit te breiden naar respectievelijk het LUMC, St. Radboud en VUmc. Die uitbreiding was noodzakelijk en waardevol.

Graag dank ik Prof.dr. H.S.A. Heymans, Prof.dr. B.F. Last, Prof.dr. R. Pieters, Prof.dr. J.M. Koot, Prof.dr. J.B. van Goudoever en Prof.dr. N.K. Aaronson voor de bereidheid om zitting te nemen in de promotiecommissie.

Dan een woord van dank aan mijn begeleiders:

Lieve Martha, zó fi jn dat je me uiteindelijk toch hebt aangenomen! Onder jouw vleugels heb ik mogen leren en groeien. Ik heb ontzettend prettig met je samengewerkt, met bewondering voor je enthousiasme en bevlogenheid. Dank voor alles - in het bijzonder voor je vertrouwen, de vrijheid, je kritische noten, je betrokkenheid, de mentale lucht èn niet te vergeten de reisjes en je gezelligheid! Oh ja, en groejtes…;)

Lieve Symone, kort door de bocht gezien was mijn onderzoek een kopie van jouw onderzoek. Je advies en ervaring waren voor mij dan ook zeer waardevol. Je stijl was to-

dankwoord

166

the-point, vriendelijk, opbouwend en concreet. Prettig. Veel dank!

Lieve Hendrik, qua samenwerking moesten we soms aan elkaar wennen. Een andere manier van werken, een andere kijk op dingen. Uiteindelijk hebben we er een mooie weg in gevonden. Bedankt voor je visie, maar zeker ook veel dank voor je belangstelling en altijd positieve bekrachtiging.

Zonder velen was dit proefschrift er niet gekomen, maar zonder Anne en Lotte was het er zeker nooit gekomen. Als onderzoeksassistenten lag het grootste deel van de dataverzameling in jullie handen. Laptopje, printertje, voice recorder, stapels vragenlijsten… op naar een poli in den lande voor een meting. En dat meer dan 500 keer. Een leuk, maar tijdrovend en niet te onderschatten logistiek gebeuren. Thx! Nu allebei jullie eigen promotietraject, zeker niet minder tijdrovend. Lieve Anne, heel jammer dat je er niet bij kunt zijn vandaag – een gemis. Lieve Lot, met jou heb ik het langst samengewerkt en veel gedeeld – vooral denkend aan New Orleans, Sid en the Uggs-hunt ;). Bedankt dat je altijd zo attent bent en zeker ook dank voor je tomeloze energie. Fijn dat je zo dicht naast me staat vandaag!

Veel hulp heb ik ook gehad van studenten: Machteld, Marleen, Christine, Marjolein, Jaimy, Ruben, Stijn, Mees, Iris, Renée en Marcelle. Allemaal hebben jullie een bijdrage geleverd aan QLIC-ON: uitvoeren van metingen, uitwerken van interviews, scoren van interviews, schrijven van scripties en werkstukken. Ik dank jullie!

Zeker een groot woord van dank voor alle balie- en polimedewerkers, verpleegkundigen en nurse practitioners van de afdeling kinderoncologie van het AMC/EKZ, LUMC, St. Radboud en VUmc. Jullie assistentie was zeer welkom bij het werven van patiënten en het uitvoeren van de metingen. Lang niet eenvoudig in de hectiek van een poli. In het bijzonder bedank ik Greet, Vera, Sabine, Suzan, Tony, Armanda en natuurlijk Marietje.

Daarnaast wil ik dr. Paul Brons uit Nijmegen danken voor het praktisch toegankelijk maken van de kinderoncologie in het St. Radboud voor mijn studie.

De uitvoering van het QLIC-Onderzoek werd tevens vergemakkelijkt door Pia de Koning, Ilse van Gils, Richard Heinen en Arjan Nooteboom. Als datamanagers verschaften jullie ons op gezette tijden lijsten met namen van patiënten die we konden benaderen. Veel dank daarvoor.

Onderdeel van de QLIC-ON interventie was een training. Met de deskundigheid, inspiratie en hulp van Bob Last, Jantien Vrijmoet - Wiersma, Esther Meijer - van den Bergh, Wynie van de Berg, Netteke Schouten en Alice van Dijk hebben we er een mooi product van kunnen maken. Veel dank!

167

In de training werd onder andere gebruik gemaakt van fi lmmateriaal: Alina Harris, bedankt voor het draaien en monteren. De kinderen, ouders en artsen (dr. Jaap Groothoff , dr. Marion van Rossum en dr. Eric Haarman) die in de fi lms fi gureren wil ik bij deze ook hartelijk danken.

Frits Schiphorts van Zonder Boter, bedankt voor het muisje Floris Hekma van Biomedia, dank voor de online versie van de PedsQL.

En niet te vergeten, Robert Segaar, bedankt voor het programmeren van de QLIC-ON software. Het was een lang traject van voortdurend fi ne-tunen en aanpassen aan onze wensen, maar het resultaat was er naar.

Bregje van Gellecum wil ik heel hartelijk danken voor het vormgeven van de buitenkant van dit boekje. Ik vind het prachtig geworden!

Beste Hilde, de binnenkant van dit proefschrift is door jou opgemaakt. Bedankt voor de strakke lay-out!

Een interventiestudie kan niet zonder een goede ambassadeur. Lieve Netteke, als kinderoncoloog was jij de mijne. Niets was te veel. Geen vraag was te gek. Veel dank voor je toegankelijkheid, je hulpvaardigheid, je betrokkenheid, je interesse en je vriendelijkheid.

Lieve Myra, waar een cursus kwalitatief onderzoek al niet naar kan leiden. Nou, twee artikelen dus (hoofdstuk 4 en 7). Bedankt voor het plezierige contact en de goede samenwerking!

Statistiek: onlosmakelijk verbonden met onderzoek, maar iets minder met mezelf. Daarom veel dank aan Nan van Geloven en natuurlijk Heleen voor de hulp bij de analyses.

En dan die fi jne (oude en nieuwe) Basement-Babes & Boy: Heleen, Marieke, Madelon, Channa, Lotte, Eefj e, Liesbeth, Willemijn, Janneke, Anne, Lobke, JP, Hedy en Perrine. De afgelopen 6 jaar was het een ware eer om één van jullie te mogen zijn. Jullie waren de kers op de PSA-taart, de jus over de AMC-kelder. Zoveel lol, zoveel feestjes, zoveel life-events, zoveel gedeeld. Het was mooi!

Basement-Buren: bedankt voor jullie aandeel in bovenstaande feestvreugde!

Lieve Eef, ook jij staat vandaag dicht naast me, zoals je dat eigenlijk in het dagelijks leven ook doet. Dank voor je warme vriendschap!

dankwoord

168

Lieve collega’s van boven: ik vond het een gezellige en leerzame tijd. Jullie gaven mij een blik in de praktische psychologenkeuken. Bedankt!

Tot slot, zijn er altijd een aantal mensen die eigenlijk niet zoveel met de totstandkoming van een proefschrift te maken hebben, maar die desondanks toch in het dankwoord genoemd worden. Gewoon, omdat dat leuk en fi jn is. En in mijn geval ook omdat ze mijn leven meer kleur geven.

Om te beginnen mijn lieve vriendinnetjes en vriendjes die ik uit Venlo, Hoensbroek, Eindhoven, Utrecht en Amsterdam ken. Vakanties, uitstapjes, feestjes, etentjes, terrasjes, eindeloos kletsen over niets, maar zeker ook over alles. Heerlijk! Laat het voortgaan.

Lieve Liek, Lon, Daan, Anna, Nies en “Patsy”: intens, echt, inspirerend, hartverwarmend – er altijd zijn. Zo fi jn. Bedankt voor de mooie vriendschap!

Dan, lieve Piet en Mia. Al enige jaren kom ik bij jullie over de vloer en ik voel me altijd welkom. Hoogtepunt is nog steeds het vijfsterrenontbijt op zondagochtend. Bedankt voor de warmte!

Lief zusje, lieve San. Afgezaagd, maar er zijn geen andere woorden: een beter zusje kan ik me niet wensen. Bedankt dat je er altijd bent. Lieve Robert, je bent van toegevoegde waarde voor de Engeltjes! Bedankt voor je prettige aanwezigheid en je vaak culinaire en alcoholische initiatieven!

Lieve pap, lieve mam. Nooit kunnen vermoeden dat het zo bijzonder is dat jullie er vandaag bij zijn. Aan jullie dank ik heel veel, maar vooral het eindeloze vertrouwen dat jullie in mij hadden en hebben. Het fundament van mijn zijn. Bedankt en liefs!

Als laatste, mijn kleine en mijn grote man,

Lieve kleine Boaz, zo waanzinnig prachtig mooi dat je er bent. Ik Geniet!!

Mijn allerliefste John. Zo belangrijk, zo liefdevol. Ik vind mijn leven meer dan heerlijk met jou. Samen de toekomst in – ik heb er ongeloofl ijk veel zin in! Lieve zoen.

CVCurriculum Vitae

Vivian Engelen werd geboren op 2 oktober 1978 in Tegelen. Ze groeide op in Belfeld

en behaalde in 1997 haar VWO diploma aan het Stedelijk College Den Hulster te Venlo.

Haar studententijd begon aan de Hogeschool Limburg in Hoensbroek met de studie

Ergotherapie. Na het behalen van haar propedeuse was de tijd rijp om de Limburgse

grond achter zich te laten en startte ze in 1998 met de studie Fysiotherapie aan de

Fontys Hogescholen te Eindhoven. Het avontuur begon pas echt toen ze in 2000 vier

maanden stage ging lopen in Mt. Wilga Private Hospital in Sydney Australië. Een jaar later

rondde ze de opleiding af en vertrok vervolgens naar Utrecht om daar Pedagogische

Wetenschappen te gaan studeren aan de Universiteit Utrecht. Haar onderzoekstage

deed ze in het Wilhelmina Kinderziekenhuis te Utrecht, waarna ze in 2004 afstudeerde.

Tijdens haar studie werkte ze als fysiotherapeut in de revalidatie en in de particuliere

sector. Na haar afstuderen wilde ze graag aan de slag als psychosociaal onderzoeker. Via

een omweg in het geneesmiddelenonderzoek is dit uiteindelijk gelukt. In 2005 werd haar

reis voortgezet richting Amsterdam, waar ze op de Psychosociale Afdeling van het Emma

Kinderziekenhuis AMC als promovendus startte. Ze richtte zich op het QLIC-ONderzoek,

wat geresulteerd heeft in dit proefschrift.

171

QQLIC-ON PROfi le

172

173

Respondent: 692Name: xxxDate of birth: 31-12-1996Completed by: Child Child ChildTime period: Last week

Remarks:

It is hard for me to walk more than one block It is hard for me to run

It is hard for me to take a bath or shower by myself It is hard for me to do chores around the house I hurt or ache

02-03-2007 16-04-2007 04-06-2007

02-03-2007

Almost always

16-04-2007 04-06-2007

Almost always

I feel afraid or scaredI feel sad or blue

I worry about what will happen to me

Other kids do not want to be my friend Other kids tease me

It is hard to keep up when I play with other kids

02-03-2007 16-04-2007 16-04-2007

02-03-2007

Almost always

16-04-2007 04-06-2007

Almost always

Emotional Functioning

100

80

50

60

40

20

0

77

02-032007

70

16-042007

65

04-062007

Social Functioning

100

8085

60

40

20

0

85

02-032007

90

16-042007

04-062007

School Functioning

100

80

7060

40

20

0

50

02-032007

60

16-042007

04-062007

7987

78

9488

85

Physical Functioning

100

80

60

40

20

002-032007

16-042007

04-062007

score

score

QLIC-ON PROfi le

Documents

UvA-DARE (Digital Academic Repository) Monitoring quality ... · Vivian Engelen). Symone Detmar had experience in providing patient reported outcomes (PROs) in adult cancer care,