Prevalence of acute and post-traumatic stress disorder and comorbid mental disorders in breast cancer patients during primary cancer care: a prospective study

Prevalence of acute and post-traumatic stress disorderand comorbid mental disorders in breast cancer patientsduring primary cancer care: A prospective study

Anja Mehnert* and Uwe KochInstitute of Medical Psychology, University Medical Center Hamburg-Eppendorf, Martinistr. 52-S35, 20246 Hamburg, Germany

* Correspondence to: Instituteof Medical Psychology,University Medical CenterHamburg-Eppendorf,Martinistr. 52-S35, 20246Hamburg, Germany. E-mail:[email protected]

Abstract

This study aimed at the identification of acute and post-traumatic stress responses, and

comorbid mental disorders in breast cancer patients. Structured clinical interviews for DSM-IV

(SCID) were conducted post-surgery with 127 patients (t1). Screening measures were used to

assess post-traumatic stress responses, anxiety, and depression at t1 and at 6 months follow-up

(t2). Based on the SCID, prevalence rates were 2.4% for both, cancer-related ASD and PTSD.

Experiences most frequently described as traumatic were the cancer diagnosis itself and

subsequent feelings of uncertainty. Patients with lifetime PTSD (8.7%) were more likely to

meet the criteria for cancer-related ASD or PTSD ðOR ¼ 14:1Þ: Prevalence estimates were

7.1% for Adjustment Disorder, 4.7% for Major Depression, 3.1% for Dysthymic Disorder and

6.3% for Generalized Anxiety Disorder. Using the screening instruments, IES-R, PCL-C and

HADS, we found PTSD in 18.5% at t1 and 11.2–16.3% at t2. The estimates of anxiety and

depression reveal rates of 39.6% (t1) and 32.7% (t2) for anxiety, as well as 16.0% (t1) and

13.3% (t2) for depression (t1) (cut-off58). The diagnosis of a life-threatening illness has been

included as a potential trauma in the DSM-IV. However, it has to be critically evaluated

whether subjective feelings of uncertainty like fears of treatment count among traumatic

stressors, and thus, whether the diagnosis of PTSD is appropriate in this group of cancer

patients. However, a large number of women with emotional distress illustrate the need for

psychosocial counseling and support in this early treatment phase.

Copyright # 2006 John Wiley & Sons, Ltd.

Keywords: PTSD; anxiety; depression; oncology; cancer

Introduction

Since life-threatening illness has been included as apotential traumatic event in the DSM-IV 1994 [1],Post-traumatic stress disorder (PTSD) has beenincreasingly diagnosed as an additional morbidityamong cancer patients. Studies on the prevalenceof PTSD in adult cancer patients predominantlyfocus on breast cancer patients with early andmixed tumor stages, and with a wide time rangebetween initial cancer diagnosis and cancer treat-ment, spanning from a few days to over 11 years.Sample sizes vary between 27 and 209 patients,with the majority of studies being cross-sectional.The prevalence of cancer-related PTSD rangesfrom 0 to 32% [2–5]. Literature also indicates thata high number of patients, up to 80%, are likely toencounter individual PTSD symptoms followingcancer [3,5,6].The consideration of a life-threatening illness as

a potential traumatic event reflects subjectiveexperiences of distress, which many cancer patientsstruggle with. However, it has lead to a controversy

whether a disease, such as cancer, meets the criteriaof a traumatic stressor. Furthermore, this bringsinto discussion the distinction between PTSD andAdjustment Disorder (AD), underlying etiologicmechanisms and the reasonability and suitability ofthe diagnosis in somatically ill patients with regardto evidence-based psychosocial treatments ofPTSD, such as cognitive-behavioral exposureprocedures [7,8].

The conceptual broadening of the stressorcriterion consequently causes a diversity of poten-tial traumatic experiences that occur during thecourse of the illness, ranging from the detection ofsymptoms, and the disclosure of diagnosis topalliative care [2,9]. The life-threatening conditionunceasingly present in the majority of patients doesnot always pose an immediate and direct threat;instead, it differs among patients depending ontumor type and applied therapy. Cancer patientsexperience a certain degree of control over theirillness during the course of treatment and care [2,3].Moreover, the anticipation of impending stressfulevents, such as surgery may alter the impact of

Received: 9 November 2005

Revised: 4 April 2006

Accepted: 6 April 2006

Copyright # 2006 John Wiley & Sons, Ltd.

Psycho-OncologyPsycho-Oncology 16: 181–188 (2007)Published online 19 July 2006 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/pon.1057

those stressors by giving the opportunity to seekand provide support.With regard to the assessment of PTSD, the

evaluation of intrusive symptoms is ambiguoussince cancer patients frequently report future-oriented fears, but rarely flashback episodes orintrusive memories [3,10]. Avoidance behaviorremains difficult to define, as the disease andtreatment procedures imply a continuous confron-tation with potential trauma-related stressors.Furthermore, potential arousal symptoms interferewith cancer- or treatment-related somatic symp-toms, such as sleep disturbances or irritability.Against this background, the aim of this study

was to identify the prevalence of acute and post-traumatic stress responses and comorbid mentaldisorders in breast cancer patients during primarycancer care. In particular, we aimed to identifyevents effectively experienced as traumatic, toclarify the characteristics of cancer- or treatment-related events. In addition, we investigated the roleof premorbid trauma exposure and lifetime PTSDin the occurrence of cancer-related ASD andPTSD.

Method

Study design and participants

The study is based on a prospective design with twoassessment points. A total number of 203 patientswere hospitalized at the Department of Gynecol-ogy, University Medical Center Hamburg-Eppen-dorf with breast cancer or suspicion of breastcancer during the study period. Among thesewomen, 50 (25%) fulfilled the exclusion criteria(no breast cancer, age over 80 years, languageproblems, cognitive problems). Of the remainingeligible 153 patients, 26 (17%) declined to partici-pate. One hundred twenty-seven patients (83%participation rate) were consecutively interviewedpost-surgery, and were asked to fill out self-reportquestionnaires ðn ¼ 108Þ: Six months after theclinical interview (t2), all patients were mailed asecond set of questionnaires to their home address.Ninety-eight patients responded (78%, two pa-tients had died) (Table 1).

The study protocol had been approved by theethics committee of the Hamburg chamber ofphysicians and the regional data protection office.Written, informed consent was obtained fromeligible patients older than 18 years able to speakand read German.Participants and non-participants did not differ

in terms of age and tumor stage (p-values40.14).Participants at t2 did not differ from non-partici-pants in the variables age, tumor stage, PTSDsymptoms, anxiety and depression (p-values40.24).

Measures

Demographic data were obtained by using a semi-structured interview. Medical charts were reviewedto obtain information about disease stage andtreatment characteristics. A semi-structured inter-view was conducted to assess characteristics of thetraumatic experience of the cancer diagnosisfollowed by modules from the Structured ClinicalInterview for DSM-IV (SCID) [11,12]. The follow-ing modules were used: lifetime PTSD, cancer-related PTSD, cancer-related ASD, AD, currentMajor Depressive Disorder (MDD), current Dys-thymic Disorder (DD) and current General Anxi-ety Disorder (GAD). The diagnosis of a cancer-related PTSD was not applied when the time sinceinitial breast cancer diagnosis was less than 4weeks.The SCID was performed by trained psycholo-

gists. Twenty-five interviews were performed withtwo psychologists present, who both independentlyscored the answers. The kappa value was k ¼ 1:0for each of the following diagnosis: MD, DD,GAD, AD, life-time PTSD and cancer-relatedASD. The kappa value for cancer-related PTSDwas k ¼ 0:65: The resulting average kappa value ofk ¼ 0:91 (Cohen’s kappa coefficient) indicates asubstantial concordance of all SCID diagnosesbetween the two raters.The following self-report questionnaires

were used: The Impact of Event Scale}revisedVersion (IES-R) [13,14], a three subscale 22-itemmeasure of intrusive, avoidance and hyperarousalsymptoms experienced over the past 7 days. Itemsare rated on a 4-point scale according tohow frequently each has occurred during the past7 days (0 ¼ ‘not at all’; 1 ¼ ‘rarely’; 3 ¼ ‘sometimes’;5 ¼ ‘often’). The term ‘distressing event’ wasreplaced by the term ‘cancer or cancer treatment’.A cut-off point for a PTSD diagnosiswas calculated on the basis of a regressionformula specific for the German adaptation of theIES-R [14].The Posttraumatic Stress Disorder Checklist}

Civilian Version (PCL-C) [15,16] assesses thePTSD symptom criteria intrusion, avoidance, andhyperarousal according to the DSM-IV. Partici-

Table 1. Study participants (total number, 203)

N %

Exclusion criteria fulfilled 50 24.6

Eligible patients 153 100.0

Participation at timepoint 1

SCID and questionnaire completed 108 70.6

SCID but no questionnaire completed 19 12.4

Participation at timepoint 2

Questionnaire completed 98 78.4a

aTwo patients deceased.

Copyright # 2006 John Wiley & Sons, Ltd. Psycho-Oncology 16: 181–188 (2007)

DOI: 10.1002/pon

182 A. Mehnert and U. Koch

pants estimate the degree to which they have beenconcerned about each symptom during the pastmonth (1 ¼ ‘not at all’ to 5 ¼ ‘extremely’). Theterm ‘distressing event’ was replaced by the term‘cancer or cancer treatment’. Participants areconsidered likely to have a PTSD diagnosis if theyobtain a total score of 50 or above (cut-off scoremethod) or if they meet at least one intrusion, threeavoidance and two arousal symptoms (rated as‘moderately’ or above) (symptom cluster method)[17]. The PCL-C was only used at t2.The Hospital Anxiety and Depression Scale}

German Version (HADS-D) is a 14-item instru-ment for the screening of anxiety and depression insomatically ill patients [18,19]. The item scoresrange from 0 (no distress) to 3 (maximum distress).According to the German Manual of the HADS[19], a score of 8–10 is considered to indicate apossible anxiety or depressive disorder and a scoreof 11 or above is considered to indicate a probableanxiety or depressive disorder.

Statistical methods

Statistical analyses were performed using theStatistical Package for the Social Sciences (SPSS)version 12.0. Content analysis was used to analyzeindividual interviews and written responses. Asso-ciations between variables were calculated using

Pearson’s Product-Moment or Spearman’s correla-tion coefficient. T-tests were conducted for groupcomparisons in normally distributed metric data.Group differences in categorical variables werecalculated using w2-test. Non-parametric tests(Mann–Whitney U test and Wilcoxon test) wereused in ordinal data, and when the assumption ofnormal distribution was in doubt. To provide anestimate of the magnitude of the group differences,Cohen’s standardized effect size (ES) was calcu-lated [20]. Odds ratios (OR) were calculated usingcontingency tables ðCI ¼ 95%Þ: Two-tailed signifi-cance tests were conducted using a significancelevel of p40.05.

Results

Demographic and medical sample characteristics

Table 2 presents demographic and clinical char-acteristics of the sample. For the majority ofwomen, the average time since (recent) diagnosiswas 15 days (SD ¼ 13; range 0–67), however 14patients had received pre-surgical chemotherapyand had known their diagnoses for on average 181days (SD ¼ 95; range 100–380). Those patientswith cancer recurrence had received their initialbreast cancer diagnosis on average 103 months ago(SD ¼ 64; range 17–252). Between t1 and the

Table 2. Demographic and medical characteristics ðN ¼ 127Þ

Variable N %

Mean age in years M¼ 54.87 (SD¼ 10.74, range 29–77)

Marital status

Married 88 69.3

Divorced/separated 27 21.3

Single 12 9.4

Partnership 101 79.5

Educational level

Elementary school 61 48.0

Junior high school 42 33.2

High school certificate/university degree 24 18.8

Employment status

Currently employed 59 46.4

Retired 51 40.2

Housewife/unemployed 17 13.4

Disease phase

Newly diagnosed 98 77.2

Recurrence 29 22.8

pTNM-classification

0 (DCIS) 15 11.8

I 39 30.7

II 58 45.7

III and IV 15 11.8

Type of surgery

Breast-conserving surgery 89 70.1

Mastectomy 21 16.5

Mastectomy with immediate breast reconstruction 17 13.4

Days since recent breast cancer diagnosis* M¼ 14.6 (SD¼ 13.0, range 0–67)

Days since breast surgery M¼ 3.1 (SD¼ 2.5, range 0–13)

pTNM, post-surgery histopathologic classification.

*n ¼ 113 patients without pre-surgical chemotherapy.

183Acute and post-traumatic stress disorder and comorbid mental disorders


DOI: 10.1002/pon

follow-up assessment (t2), 84% of the patientsreceived one or more of the following cancertreatments: radiotherapy (55%), chemotherapy(52%), and anti-estrogen therapy (40%).

Characteristics of cancer-related traumatic events

The majority of patients (87%) experienced thecancer diagnosis as unexpected and overwhelming.Independent of the disease phase ðp ¼ 0:32Þ; 32%said they had already passed the most distressingcancer-related burden, whereas 23% believed thatparticularly distressing events were yet to come,and 43% were undecided.Among the patients, 91% ðn ¼ 116Þ described

that they had experienced at least one traumaticevent related to the breast cancer disease ortreatment; however, 54% additionally respondedwith intense fear, helplessness, or horror and thusfulfilled the DSM-IV A2 criterion. Descriptions oftraumatic events show that the breast cancerdiagnosis (42%) and uncertainty about the future(29%) were most frequently experienced as trau-matic. A comparison of the distribution of reportedevents and differences between patients who didand those who did not fulfill the trauma criterionprovided the following results. Women whofulfilled the criterion A2 more frequently experi-enced the diagnosis as traumatic (p-values50.50),but did not differ in the distribution of theremaining described events with exception towaiting time and the number of other concerns(Table 3).At t2, patients assessed the current and retro-

spective subjective threat of the breast cancerdisease. Patients experienced the cancer disease att2 as significantly less threatening than at t1(M ¼ 3:45; SD ¼ 1:0 vs M ¼ 2:68; SD ¼ 0:98;(t ¼ 7:0; p50.001; ES ¼ 0:8) (item: 1 ¼ ‘no threat’ to 5 ‘severe threat’).

PTSD lifetime prevalence

Based on the SCID, 31% of women ðn ¼ 39Þreported experiencing at least one traumatic event

during their lifetime. These traumatic eventsconsisted of serious accidents ðn ¼ 12Þ; wartimeexperiencesa ðn ¼ 11Þ; physical attacks ðn ¼ 8Þ;natural disasters ðn ¼ 2Þ and sexual assaultðn ¼ 1Þ: Five women reported witnessing a trau-matic event. Estimates of the PTSD lifetimeprevalence in those 39 women reporting a trau-matic event according to the SCID list were 9%(Table 4). The 4-week PTSD prevalence was 5%,i.e. six patients had persistent PTSD symptomsrelated to a non-cancer traumatic event prior totheir cancer diagnosis.

Table 3. Type and frequency of cancer-related traumatic events ðn ¼ 116Þ

Total Trauma criterion A2 fulfilled

Cancer-specific traumatic eventsa n % Yes (n) No (n) p

Breast cancer diagnosis 49 42.2 32 17 0.05

Uncertainty about the future 34 29.3 17 17 0.55

Doctor–patient interaction 17 14.7 11 6 0.36

Fear of breast surgery and mastectomy 15 12.9 7 8 0.52

Fear of chemotherapy 9 7.8 6 3 0.51

Detecting the lump 8 6.9 4 4 0.80

Waiting period prior to breast surgery 7 6.0 1 6 0.05

Other 12 10.3 11 1 0.006

aMultiple responses (151), percentages refer to the sample of n ¼ 116 women.

Table 4. Prevalence of mental disorders at t1 ðN ¼ 127Þ

Psychiatric disorders and symptoms (SCID) N %

Lifetime PTSD total prevalence 11 8.7

Mild 4 3.1

Moderate 5 3.9

Severe 2 1.6

Four-week prevalence 6 4.7

Cancer-related PTSD total prevalence 3 2.4

Mild 2 1.6

Moderate 1 0.8

Cancer-related ASD total prevalence 3 2.4

Mild 3 2.4

Cancer-related PTSD symptoms

A (Traumatic event) 69 54.3

B (Intrusion) 27 21.3

C (Avoidance) 17 13.4

D (Arousal) 27 21.3

E (Duration) 3 2.4

F (Impairment) 12 9.4

Cancer-related ASD symptoms

B (Dissociative symptoms) 20 15.7

Cancer-related adjustment disorder total prevalence 9 7.1

Major depression total prevalence 6 4.7

Mild 3 2.4

Moderate 2 1.6

Severe 1 0.8

Dysthymic disorder total prevalence 4 3.1

Mild 1 0.8

Moderate 3 2.4

Generalized anxiety disorder total prevalence 8 6.3

Mild 1 0.8

Moderate 6 4.7

Severe 1 0.8



DOI: 10.1002/pon

Cancer-related ASD and PTSD prevalence andcomorbid mental disorders

Based on the SCID, 2.4% met the criteria for amild-to-moderate cancer-related PTSD, 2.4% werediagnosed with ASD. Two women met the criteriafor cancer-related PTSD with the exception ofcriterion E (duration of symptoms) and thereforewere diagnosed with AD. Seventy-six women (60%)met at least one PTSD criterion A, B, C or D. Ninewomen (7%) were diagnosed with AD, 5% withmild-to-severe MDD, 3% with mild-to-moderateDD and 6% with mild-to-severe GAD (Table 4).Table 5 shows the frequencies of PTSD, anxiety

and depression assessed with screening question-naires at t1 and t2. There is a significant decrease inthe IES-R score between t1 and t2 (Mt1 ¼ 42:9;SD ¼ 20:5 vs Mt2 ¼ 33:1; SD ¼ 20:9; t ¼ 4:8;p50.001; ES ¼ 0:5). The number of women whomet the cut-off for PTSD did not decreasesignificantly at t2 ðp ¼ 0:61Þ: The relative risk(RR) of having PTSD (IES-R) at t2 in patientsdiagnosed at t1 is 6.0 times higher (95% CI ¼ 2:4–14.9, p50.001) than in patients not having PTSD(IES-R) at t1.The PCL-C mean score was Mt2 ¼ 31:4; SD ¼

11:8: The PCL-C scores based on the symptomcluster and the cut-off score method were moder-ately correlated (rs ¼ 0:63; p50.001); nine patientswere identified as having PTSD by both methods.The concordance between the IES-R and bothPCL-C scores again showed an identical identifica-tion of nine patients with rs ¼ 0:48 (p50.001) forthe symptom cluster score and rs ¼ 0:63 (p50.001)for the cut-off score method. No significantreduction in anxiety (Mt1 ¼ 6:6; SD ¼ 4:1 vsMt2 ¼6:0; SD ¼ 4:4; t ¼ 0:8; p ¼ 0:44) nor in depression(Mt1 ¼ 4:1; SD ¼ 3:3 vs Mt2 ¼ 3:8; SD ¼ 3:5; t ¼�0:7; p ¼ 0:48) was observed between t1 and t2.As presented in Table 6, between 23 and 29% of

the patients met the criteria for at least one(probable) mental disorder at both timepointsmeasured with the SCID or screening question-naires.

Mental comorbidity in patients with PTSD

The risk of having a second mental disorder (MD,DD or GAD) in women diagnosed with a cancer-related ASD or PTSD identified by the SCID wassignificantly higher than in patients not havingcancer-related ASD or PTSD (OR 22.2; 95% CI3.6–136.5; p50.001) (Table 7). The risk for asecond (probable) mental disorder (HADS anxietyor depression) in women who met the cut-off forPTSD was significantly increased as measured bythe following screening questionnaires: For theIES-R at t1 (OR 5.2; 95% CI 1.8–15.1; p50.003)and at t2 (OR 24.8; 95% CI 6.0–101.3; p50.001);as well as for the PCL-C at t2 (cut-off scoremethod: OR 13.5; 95% CI 3.3–55.8; p50.001;symptom cluster method: OR 25.1; 95% CI 6.1–102.6; p50.001).Patients ðn ¼ 11Þ diagnosed with a lifetime

PTSD had a significantly higher likelihood ofmeeting the criteria for cancer-related PTSD orASD (OR 14.1; 95% CI 2.4–81.6; p50.01). Threewomen having a cancer-related ASD or PTSD alsofulfilled the 4-week PTSD prevalence related to anon-cancer traumatic event prior to the cancerdiagnosis (lifetime PTSD). However, the exposureto a non-cancer traumatic event ðn ¼ 37Þ was notsignificantly associated with cancer-related PTSDor ASD (OR 2.6; 95% CI 0.5–13.3; p50.25).

Table 5. Screening of PTSD, anxiety and depression at t1 and t2

Timepoint 1 (n¼108) Timepoint 2 (n¼98)

n % n %

IES-R

PTSD 20 18.5 16 16.3

PCL-C

PTSD (cut-off score method) } } 11 11.2

PTSD (symptom cluster method) } } 16 16.3

HADS

Possible anxiety disorder 25 23.6 19 19.4

Probable anxiety disorder 17 16.0 13 13.3

Possible depressive disorder 9 8.5 5 5.1

Probable depressive disorder 8 7.5 8 8.2

Table 6. Total frequency of mental comorbidity

Clinical interviewa Screeningb

Timepoint 1 Timepoint 1 Timepoint 2

N % N % N %

No diagnosis 98 77.2 77 71.3 73 74.5

One diagnosis 15 11.8 22 20.4 13 13.3

Two diagnoses 10 7.9 5 4.6 5 5.1

Three diagnoses 4 3.1 4 3.7 7 7.1

aPTSD (lifetime, current cancer-related), ASD, MD, DD, GAD, AD (and/or).bPTSD (IES-R or PCL-C) and/or anxiety and/or depression (HADS

cut-off 511).



DOI: 10.1002/pon

Discussion

The findings of this study indicate that a smallproportion of patients, with 2.4% meeting thecriteria for both, PTSD and ASD, usually presentwith mild symptom characteristics. To our knowl-edge, there have not yet been empirical attempts todetermine the prevalence of ASD in cancerpatients. The clinician-administered prevalenceestimate for PTSD is similar to that found byother authors [3,4]. Interestingly, half of thepatients suffering from a cancer-related ASD orPTSD also met the criteria for a PTSD lifetimeprevalence, representing a 14-fold increase of riskin that group. In contrast, lifetime trauma exposurewas not significantly associated with cancer-relatedPTSD or ASD. These findings strengthen assump-tions which indicate that pre-cancer mental dis-orders contribute significantly to cancer-relateddistress [21,22]. This result might also contributeto the explanation of higher PTSD prevalence ratesin cancer patients evident in the literature with nodifferentiation between lifetime, acute, and acutecancer-specific PTSD.One of the aims of this study was to elucidate

characteristics of the cancer-related events experi-enced as traumatic. In this context, it was interest-ing to learn that the breast cancer diagnosis wasindeed experienced as unexpected by the vastmajority of patients. However, there was a broaddivergence in subjective judgment between theseverity of the cancer-related burden alreadyexperienced in the past and the burden expectedin the future.Slightly more than half of the patients experi-

enced the breast cancer disease as a traumatic eventand responded with intense fear, helplessness andhorror (PTSD criterion A2). However, a closerview leads to a few noteworthy findings: First ofall, the diagnosis of breast cancer itself, but alsoovercoming feelings of uncertainty about thefuture, were most frequently perceived as trau-matic. The cancer diagnosis basically refers to theDSM-IV definition of a traumatic event, butfeelings of uncertainty do not, and therefore canbe critically discussed. Second, it needs to bereferred to the comparison of the distribution of

reported events and differences between patientswho did and those who did not fulfill the traumacriterion A2. Although we found that women whofulfilled the trauma criterion (A2), more frequentlyexperienced the breast cancer diagnosis as trau-matic, the two groups did not differ in general inthe distribution of the events mentioned. It can beconcluded, that the responses to the stressor greatlydepend on the subjective experiences of the women.Furthermore, reported fears predominantly re-ferred to the assumed course of the disease anduncertainty rather than to past experiences suchas flashback episodes, and thus should beinterpreted as problems of adjustment rather thanpost-traumatic stress responses.In recent years, a huge body of scientific as well

as non-scientific literature has focused on traumain cancer patients. There is evidence that a numberof patients develop cancer- or treatment-relatedPTSD, however, for the majority of patients, theAD diagnosis seems more appropriate, since thisdiagnosis refers to reactions to an identifiablepsychosocial stressor less specific than a trauma.It would be essential to specify this diagnosis interms of specific symptoms and treatment recom-mendations.The fact that PTSD has been so widely taken

into consideration by professionals and patientsalike may reflect an attempt to ‘validate’ individualgrief and suffering during an unexpected and‘undeserved’ situation. For the majority of patientshaving cancer implies the confrontation withuncertainty, fear of progression and death, a broadrange of cancer- and treatment-related burdens, aswell as the search for causes of the disease andmeaning. Being a ‘victim’ of a cancer traumamay allow for an alleviation of one’s ownemotional responsibilities within the process ofadaptation [23].About one-third of the patients experienced a

traumatic life event according to the DSM-IVdefinition. The estimation of PTSD lifetime pre-valence in those women was 9%, a number thatcorresponds with the PTSD lifetime prevalencerates of women in the general population [24,25].However, it should be considered that traumaexposure and lifetime prevalence were assessedretrospectively and hence, current emotional dis-tress might have interfered with the women’smemory.Slightly one-fourth of the patients (23%) were

diagnosed with one to three mental disorders onthe basis of the SCID during inpatient cancertreatment. Using screening questionnaires, a com-parable amount of patients (between 26 and 29%)was considered likely to have at least one mentaldisorder. The overall prevalence estimates found inthis population are largely consistent with recentfindings from the German National Health Inter-view and Examination Survey [26], which indicate

Table 7. Risk of mental comorbidity in patients with cancer-related ASD or PTSD (SCID)

Second mental disordera Risk

Cancer-related

ASD/PTSD

Yes No Total OR CI 95% p

Yes 4 2 6 22.2 3.6–136.5 0.000

No 10 111 121

Total 14 113 127

aMD, DD0 or GAD.



DOI: 10.1002/pon

that mental disorders are highly prevalent in theadult German population with a 12-month pre-valence total of 31% and a 4-week prevalence of20%.Nevertheless, the prevalence estimates from self-

report questionnaires observed in this study arehigher, particularly since a higher number ofdiagnostic modules was used in the SCID. Adiscrepancy of prevalence estimates between self-and clinical-administered ratings has been widelyfound in the empirical literature [3,27]. Oneexplanation can be seen in the overestimation ofprevalence rates revealed by the IES-R, the PCL-Cand the HADS. However, psychometric valuesavailable for the German adaptations of the IES-Rand the HADS show a sensitivity of 70% and aspecificity of 89% for the IES-R [14], and asensitivity of 78% and a specificity of 71% forthe HADS [28]. Alternatively, the SCID could lacksensitivity in particular in somatically ill patientsand therefore underestimate prevalence.The IES-R-based PTSD prevalence estimates of

almost 19% found during inpatient cancer treat-ment are clearly higher than the SCID prevalencerates, but can be explained by several presump-tions: First, although the IES-R (as well as thePCL-C) refers to a traumatic event, it does notinclude subjective responses of the individual to thestressor, such as intense fear, helplessness, orhorror (criterion A2). In fact, the IES-R prevalencerate decreases to 11% when only those patients,who fulfill the criterion A2 in the SCID, areconsidered. Second, the IES-R refers to thepresence of symptoms during the past 7 days,whereas a PTSD diagnosis according to the DSM-IV requires the presence of symptoms for at least 4weeks. Overall, we would agree with Palmer et al.[4], who suggested the IES measures diffuseemotional distress and adjustment problems ratherthan PTSD symptoms.At t2 prevalence estimates range from 11

to 16%, according to the screening questionnairesand correspond with the 1–20% estimatesreported in the empirical literature [3]. Theprevalence of anxiety and depression reveals highrates of almost 40% for noticeable anxiety and17% for noticeable depression at t1. A small, butinsignificant decrease of this prevalence can beobserved at t2. These results point out a highnumber of distressed women and their need forpsychosocial counseling and support in this earlytreatment phase.The strengths of this study include its long-

itudinal approach, the participation rate of 83%and the use of both, a clinical interview andscreening instruments to measure PTSD, anxietyand depression. One limitation to this study is thatwe have not posed additional questions to clarifycharacteristics of the cancer-related traumaticevents. Furthermore, the use of the SCID was not

possible at both timepoints for organizationalreasons. In conclusion, the appropriateness ofPTSD diagnosis must be discussed more criticallywith regard to this group of patients.

Acknowledgements

This research has been supported by the Cora LobscheidFoundation within the Donors’ Association for the Promo-tion of Sciences and Humanities in Germany, Essen,Nordrhein-Westfalen (grant number T 129-12.171). Wethank Prof. Dr Fritz Janicke and the professional healthcare team at the Department of Gynecology, UniversityMedical Center Hamburg-Eppendorf.

Note

a. Traumatic experiences during World War II or the war informer Yugoslavia.

References

1. American Psychiatric Association (APA). Diagnosticand Statistical Manual of Mental Disorders (4th edn).American Psychiatric Association: Washington, DC,1994.

2. Smith MY, Redd WH, Peyser C, Vogl D. Posttraumaticstress disorder in cancer: a review. Psycho-Oncology1999;8:521–537.

3. Kangas M, Henry JL, Bryant RA. Posttraumatic stressdisorder following cancer. A conceptual and empiricalreview. Clin Psychol Rev 2002;22:499–524.

4. Palmer SC, Kagee A, Coyne JC, DeMichelle A.Experience of trauma, distress, and posttraumatic stressdisorder among breast cancer patients. Psychosom Med2004;66:258–264.

5. Akechi T, Okuyama T, Sugawara Y, Nakano T, ShimaY, Uchitomi Y. Major depression, adjustment disor-ders, and posttraumatic stress disorder in terminally illcancer patients: associated and predictive factors. J ClinOncol 2004;22(10):1957–1965.

6. Matsuoka Y, Nakano T, Masatoshi I et al. Cancer-related intrusive thoughts as an indicator of poorpsychological adjustment at 3 or more years after breastsurgery: a preliminary study. Breast Cancer Res Treat-ment 2002;76:117–124.

7. Foa EB, Keane TM, Friedmann MJ. Effective Treat-ments for PTSD. Guilford Press: New York, 2000.

8. McNally RJ. Progress and controversy in the study ofposttraumatic stress disorder. Annu Rev Psychol2003;54:229–252.

9. Greenberg DB, Goorin A, Gebhardt MC et al. Qualityof life in osteosarcoma survivors. Oncology 1994;8:19–25.

10. Green BL, Rowland JH, Krupnick JL et al. Prevalenceof posttraumatic stress disorder in women with breastcancer. Psychosomatics 1998;39:102–111.

11. Sa� H, Wittchen HU, Zaudig M. Diagnostisches andstatistisches Manual psychischer Storungen (DSM-IV).Gottingen: Hogrefe, 1996.

12. Wittchen HU, Zaudig M, Fydrich T. SKID. Struktur-iertes Klinisches Interview fur DSM-IV. Achse I und II.Handanweisung. Hogrefe: Gottingen, 1997.

13. Weiss DS, Marmar CR. The impact of event scale}revised. In Assessing Psychological Trauma and PTSD,Wilson JP, Keane TM (eds). Guilford Press: New York,1996;399–411.



DOI: 10.1002/pon

14. Maercker A, Schutzwohl M. Erfassung von psychischenBelastungsfolgen: Die Impact of Event Skala}revi-dierte Version (IES-R). Diagnostica 1998;44:130–141.

15. Weathers FW, Huska JA, Keane TM. PCL-C for DSM-IV. National Center for PTSD}Behavioral ScienceDivision: Boston, 1991.

16. Teegen F. Deutsche Ubersetzung der PosttraumaticStress Disorder Checklist (PCL-C) des National Centerfor PTSD. Universitat Psychologisches Institut III:Hamburg, 1997.

17. Andrykowski MA, Cordova MJ, Studts JL, Miller TW.Posttraumatic stress disorder after treatment for breastcancer: prevalence of diagnosis and use of the PTSDChecklist}Civilian Version (PCL-C) as a screeninginstrument. J Consult Clin Psychol 1998;66(3):586–590.

18. Zigmond AS, Snaith RP. The hospital and depressionscale. Acta Psychiatr Scand 1983;67:361–370.

19. Herrmann C, Buss U, Snaith RP. HADS-D}HospitalAnxiety and Depression Scale}Deutsche Version: EinFragebogen zur Erfassung von Angst und Depressivitat inder somatischen Medizin. Huber: Bern, 1995.

20. Cohen J. Statistical Power Analysis for the BehavioralSciences (2nd edn). Erlbaum: New York, 1988.

21. Green BL, Goodman LA, Krupnick JL et al. Outcomeof single vs. multiple trauma exposure in a screeningsample. J Traumat Stress 2000;13:271–286.

22. Mundi EA, Blanchard EB, Cirenza E, Gargiulo J,Maloy B, Blanchard CG. Posttraumatic stress disorderin breast cancer patients following autologous bone

marrow transplantation or conventional cancer treat-ment. Behav Res Therapy 2000;38(10):1015–1027.

23. Stoffels H. Pseudoerinnerungen oder Pseudologien? Vonder Sehnsucht, Traumaopfer zu sein. In Grenzwertigepsychische Storungen. Diagnostik und Therapie inSchwellenbereichen, Vollmoeller W (ed.). Thieme Ver-lag: Stuttgart, 2004; 33–45.

24. Kessler RC, Sonnega A, Bromet E, Highes M, NelsonCB. Posttraumatic stress disorder in the NationalComorbidity Survey. Arch Gen Psychiatry 1955;52(12):1048–1060.

25. Frans O, Rimmo PA, Aberg L, Fredrikson M. Traumaexposure and post-traumatic stress disorder in thegeneral population. Acta Psychiatr Scand 2005;111(4):291–300.

26. Jacobi F, Wittchen HU, Holting C et al. Prevalence, co-morbidity and correlates of mental disorders in thegeneral population: results from the German HealthInterview and Examination Survey (GHS). Psychol Med2004;34:597–611.

27. Sellick SM, Crooks DL. Depression and cancer: anappraisal of the literature for prevalence, detection, andpractice guideline development for psychological inter-ventions. Psycho-Oncology 1999;8(4):315–333.

28. Harter M, Reuter K, Gross-Hardt K, Bengel J. Screen-ing for anxiety, depressive and somatoform disorders inrehabilitation}validity of HADS and GHQ-12 inpatients with musculoskeletal disease. Disabil Rehabil2001;23:737–744.



DOI: 10.1002/pon

Documents

Prevalence of acute and post-traumatic stress disorder and comorbid mental disorders in breast cancer patients during primary cancer care: a prospective study