Covert dose reduction is a distinct type of medication non-adherence observed across all care settings in inflammatory bowel disease

Ava i l ab l e on l i ne a t www.sc i enced i r ec t . com

ScienceDirect

Journal of Crohn's and Colitis (2014) xx, xxx–xxx

CROHNS-01063; No of Pages 7

Covert dose reduction is a distincttype of medication non-adherence observedacross all care settings in InflammatoryBowel Disease
Réme Mountifield a,c,⁎, Jane M. Andrewsb,c,Antonina Mikocka-Walus d, Peter Bamptona,c
a Department of Gastroenterology and Hepatology, Flinders Medical Centre, South Australia, Australiab IBD Service, Dept. Gastroenterology and Hepatology & School of Medicine, University of Adelaide at Royal Adelaide Hospital,South Australia, Australiac Flinders University of South Australia, Australiad School of Nursing and Midwifery, University of South Australia, Australia

Received 15 June 2014; received in revised form 29 July 2014; accepted 26 August 2014

⁎ Corresponding author at: DepartmeAustralia. Tel.: +61 8 8204 5511; fax:

E-mail addresses: ramonreme@[email protected]

http://dx.doi.org/10.1016/j.crohns.201873-9946/© 2014 European Crohn's an

Please cite this article as: Mountifieldsettings in Inflammatory Bowel Disea

KEYWORDSMedication adherence;Dose modification;Inflammatory BowelDisease;Complementary andAlternative Medicine;Medication attitudes

Abstract

Background: Non-adherence by dose omission is common and deleterious to outcomes inInflammatory Bowel Disease (IBD), but covert dose reduction (CDR) remains unexplored.Aims: To determine frequency and attitudinal predictors of overall medication non-adherenceand of covert dose reduction as separate entities.Methods: A cross sectional questionnaire was undertaken involving IBD patients in threedifferent geographical regions and care settings. Demographics, medication adherence by doseomission, and rate of patient initiated dose reduction of conventional meds without practitioner

nt of Gastroenterology and He+61 8 8272 3580.m.com.au (R. Mountifield), jan(A. Mikocka-Walus), peter.ba

14.08.013d Colitis Organisation. Publish

R, et al, Covert dose reductionse, J Crohns Colitis (2014), http

knowledge (CDR) were assessed, along with attitudes toward IBD medication.Results: Of 473 respondents (mean age 50.3 years, 60.2% female) frequency of non-adherence was21.9%, and CDR 26.9% (p b 0.001). By logistic regression, significant independent predictors ofnon-adherence were dissatisfaction with the patient–doctor relationship (p b 0.001), depression(p = 0.001), anxiety (p = 0.047), and negative views regarding medication efficacy (p b 0.001) orsafety (p = 0.017). Independent predictors of covert dose reduction included regular complementarymedicine (CAM) use (p b 0.001), experiencing more informative (p b 0.001) and comfortable (p =

patology, Flinders Medical Centre, Bedford Park, South Australia 5042,

[email protected] (J.M. Andrews),[email protected] (P. Bampton).

ed by Elsevier B.V. All rights reserved.

is a distinct type of medication non-adherence observed across all care://dx.doi.org/10.1016/j.crohns.2014.08.013

mailto:[email protected]




http://dx.doi.org/10.1016/j.crohns.2014.08.013

http://dx.doi.org/


2 R. Mountifield et al.

Please cite this article as: Mountifieldsettings in Inflammatory Bowel Disea

0.006) consultations with alternative practitioners, disbelieving doctor delivered information (p =0.021) and safety concerns regarding conventional medication (p b 0.001). Neither the frequency ofnon-adherence (p = 0.569) nor CDR (p = 0.914) differed between cohorts by different treatmentsettings.Conclusions: Covert dose reduction of IBD medication is more common than omission ofmedication doses, predicted by different factors to usual non-adherence, and has not beenpreviously reported in IBD. The strongest predictor of CDR is regular CAM use.© 2014 European Crohn's and Colitis Organisation. Published by Elsevier B.V. All rights reserved.

R, et al, Covert dose reductionse, J Crohns Colitis (2014), http

1. Introduction

Consistent medication adherence yields better outcomes inInflammatory Bowel Disease (IBD). Currently availabletherapies have an important role not only in maintainingremission,1 but also in the prevention of colorectal cancervia inflammation reduction and possibly direct antineoplas-tic pathways.2,3

Non-adherence to IBD medication is common, most studiessuggesting a frequency of 30–45% of patients,4 but a widerrange is reported. Thus many studies have investigated riskfactors for non-adherence, and although results have beeninconsistent,4 some common themes have emerged. Demo-graphic and clinical factors such as younger age,5 employmentstatus,5 being unmarried,6,7 disease duration,8 pill count7 andmedication type9 have been associated with non-adherence,and whilst this is useful to identify at-risk patients, thesefactors are not modifiable.

Studies seeking behavioural reasons for non-adherencedivide causes into categories encompassing forgetfulness(nearly 50%) and deliberate medication avoidance.8 Thelatter relates to patient belief of necessity and concernsregarding medication effects,10,8,9 dissatisfaction with orpoor recall of information regarding medications,11 physi-cian patient discordance,11–14 psychological stress, depres-sion, anxiety,15–17 and poorer QOL.17

Most instruments used to measure medication non-adherence primarily assess dose omission rather than dosereduction. This phenomenon of patient-initiated covertdose reduction (CDR) has not been studied as a separateentity in IBD, although it has been reported amongstpatients prescribed antihypertensive medications.18 Consis-tent under-dosing of IBD medication by CDR is likely to have asdeleterious an effect on disease control as dose omission.19

Therefore CDR is important to identify as distinct fromtraditionally defined non-adherence by dose omission, as itmay reflect different medication attitudes and require adifferent intervention.

Additionally, existing data suggest that there is consider-able variation in non-adherence rates between centres,20

which may affect generalizability of results from single centrestudies. It is unclear whether such variation arises fromcultural, geographic or care structure differences. This studysimultaneously assessed the frequency and attitudinal andpsychological predictors of non-adherence (using existinginstruments) and covert dose reduction of IBD medication,and compared frequencies across three contrasting IBDcohorts in Australia.

is a://

2. Methods

2.1. Subject selection and recruitment

IBD patients from three different care settings in twodistinct geographical locations in Australia were invited toparticipate.

The first cohort came from a large metropolitan teachinghospital IBD Service at Flinders Medical Centre (FMC), whichoffers specialist IBD physician and IBD nurse care. The secondcohort consisted of IBD patients in an overlapping area,receiving their care in a metropolitan Private Practice settingby general Gastroenterologists. The third cohort included IBDpatients cared for via Royal Darwin Hospital (RDH), a publichospital in a very remote location in Northern Australia. Whenthis study was conducted, IBD care in Darwin was undertakenpredominantly by General Practitioners (GPs) and GeneralSurgeons.

Potential subjects were identified from IBD databases/hospital records in each location and mailed a questionnaire.Reminder letters were sent to non-responders after one andthree months.

2.2. Questionnaire content

The questionnaire sought demographic details, views regard-ing conventional IBD medications, Complementary and Alter-native Medicine (CAM), quality of life (QOL), and Psychologicaland Personality traits. Where possible, validated instrumentswere used as described below.

Standard medication adherence was assessed using theMorisky 4 item Self Report Measure of Medication TakingBehaviour (MMAS-4),21,22 a 4 item “yes” or “no” survey thathas been validated in a broad range of diseases.23 Each of the 4items is scored 0 or 1, the sum of the 4 responses yielding atotal of 0 to 4, whereby high adherence is indicated by a scoreof 0, medium adherence by 1–2 and low adherence by 3–4.

Currently no validated tests exist to assess CDR. This wastherefore assessed in two ways; firstly as a dichotomousvariable (yes/no) based on the answer to the question “Itake less than prescribed of my IBD medication withouttelling my doctor”. A continuous variable representing CDRtendency was also generated using factor analysis.

Other medication attitude statements were put tosubjects, seeking the extent of agreement or disagreementusing a Likert scale, and additional free text responses wereencouraged.

distinct type of medication non-adherence observed across all caredx.doi.org/10.1016/j.crohns.2014.08.013


Table 2 Demographic characteristics in contrasting IBDcohorts.

FMC(n = 337)

Private(n = 91)

Darwin(n = 35)

% female respondents 60.2 60.4 60

3Covert dose reduction of medication for Inflammatory Bowel Disease

Anxiety and depression was measured using the HospitalAnxiety and Depression Scale,24 quality of life using theShort Inflammatory Bowel Disease Questionnaire (SIBDQ),25

and the Spielberger State-Trait Personality Inventory26,27

(STPI) was used to assess depressive symptoms, anxiety,anger and curiosity between cohorts.28

% Crohn's disease 55.2 57.1 48.6% indigenous subjects 0.9 1.1 2.9% current smokers 11.1 13.6 17.1% previous smokers 25.8 25.0 42.9% receiving disabilitysupport pension

1.8 1.1 5.7

% employed 58.7 56.7 62.9% currently partnered 92.2 95.3 93.3

2.3. Statistical analysis

Comparisons between cohort means and medians wereperformed using the Kruskal Wallis test for non-normallydistributed values, and two tailed t test or ANOVA fornormally distributed values. Pearson's Chi-square or Fisher'sexact test was applied as appropriate for categorical data.

Significant or trend associations at univariate level(p b 0.10) determined which variables were included inregression analyses. Binary logistic regression was used toassess predictors of non-adherence and CDR as dichotomousdependent variables. For the purposes of binary logisticregression the medium and high categories for MMAS-4 werecombined to create high (score 0–2) and low adherence(score 3–4) dichotomous outcomes.

CDR was also expressed as a continuous variable gener-ated by exploratory factor analysis for ordinal data using MPlus software (V5.2). Multiple linear regression was used todetermine predictors of CDR as a continuous dependentvariable.

A p value of b0.05 was considered statistically signifi-cant. Apart from factor analysis, statistical calculationswere performed using IBM SPSS Statistics for Windows,version 22, 2013 (IBM Corp).

2.4. Ethical considerations

The study was approved by the Flinders Clinical ResearchEthics Committee (FCREC) on behalf of SA subjects andMenzies School of Health Human Research Ethics Committeefor subjects cared for in Darwin.

10

15

20

25

30

35

Percen

tage Public

Private

Remote

p=0.914p=0.569

3. Results

3.1. Demographic data

Response rates to the survey differed between cohorts, with337/612 (55.1%) of FMC and 91/180 (50.5%) of SA Privateinvitees participating, compared with 35/100 (35%) inDarwin (p b 0.0001). Non-respondents did not differ fromrespondents by gender (p = 0.2), but were younger thanrespondents (Table 1). Darwin subjects were more likely becurrent or previous smokers, and to receive a disabilitysupport pension (Table 2).

Table 1 Respondent versus non-respondent characteristics.

Respondents Non-respondents p value

Mean age (years) 50.3 43.7 0.065% female 60.2 55.7 0.2

Please cite this article as: Mountifield R, et al, Covert dose reduction is asettings in Inflammatory Bowel Disease, J Crohns Colitis (2014), http://

3.2. Medication adherence by MMAS-4

“Low adherence” criteria were met by 21.9% of subjectsoverall, with no statistical differences between cohorts (p =0.569) (Fig 1). “Medium” adherence was present in 34.8%and high adherence in 43.3% of subjects.

Of the 241 subjects reporting which medication theywere most likely to omit or forget, 31.7% said oral 5ASA orSulfasalazine, 13.7% azathioprine or 6MP, 10% corticoste-roid, 7.1% enemas, and 1% adalimumab while 36.5% reportednonadherence to other medications, including methotrex-ate, probiotics, loperamide, and antibiotics. These propor-tions correspond approximately with local prescribing ratesof each medication, the use of biologicals being infrequentat this time due to prescribing restrictions in Australia.

Of the 133 subjects offering free text reasons for lowadherence, approximately half (50.6%) cited forgetfulnessand disorganisation as the main reason for missing medica-tion doses. The remaining 49.4% were deliberately lowadherent, citing adverse effects (18.4%), medication cost(13.5%), perceived lack of efficacy (7.7%), finding enema usedisagreeable (7.7%), or lack of convincing benefit based on

0

5

Low adherence Deliberate Dose Reduc�on

Fig. 1 Rates of low adherence and covert dose reduction weresimilar across IBD cohorts.




doctor's explanation (2%), with no difference betweencohorts (p = 0.10).

3.3. Associations of lowadherence — univariate analysis

Demographic factors associated with low adherence includ-ed female gender (27.3% versus 13.6%, p = 0.001), perma-nent employment (26.0% versus 15.9%, p = 0.008), youngerage (mean 44.0 versus 52.0 years, p b 0.001), and high pillburden (p = 0.024). Disease type (UC versus CD) did notaffect adherence (0.388).

Attitudes significantly associated with low adherence onunivariate analysis include perceived medication inefficacy(p b 0.001), adverse effects (p = 0.041), negative relation-ship with the doctor (p = 0.008), missing IBD appointments(p b 0.001), desiring more control over IBD management(p = 0.036), diagnosis of depression (p = 0.007), lack ofsocial support (p b 0.001) and personal relationship dissat-isfaction (p b 0.001). Preference for an alternative practi-tioner consultation style was associated with low adherence(p = 0.047), whilst regular CAM use was not (p = 0.263).Subjects scoring higher for anxiety (p = 0.001), depression(p = 0.002) and anger (p = 0.002) and lower for curiosity(p = 0.043) were more likely to report low adherence, whilstQOL did not affect adherence (p = 0.116).

3.4. Low adherence predictors by logisticregression analysis

After adjusting for known influences on adherence such asage, gender, disease type, employment, relationship status,pill burden, and quality of life, associations that remainedstatistically significant are shown in Table 3. This modelexplained a significant proportion of variance in low adher-ence rates (adjusted pseudo R squared 0.240, goodness of fitby Hosmer Lemeshow p = 0.347). The strongest predictivefactor was a patient perception of a negative relationship withthe IBD doctor. Psychological variables as well as medicationbeliefs such as doubts about efficacy and adverse effects werealso associatedwith low adherence. The strong associationwithmissing IBD appointments likely reflects the same underlyingattitudes contributing to low medication adherence.

Table 3 Independent predictors of low adherence amongst IBD s

Predictor

Negative relationship with IBD doctorMissing IBD appointmentsPerceived medication efficacyLack of social supportSatisfied with level of information re IBD from physicianAntidepressant useSatisfied with personal relationshipsExperienced adverse effects IBD medicationDepression — patient reported diagnosisAnxiety (HADS)Depression (HADS)


3.5. Covert dose reduction

When asking specifically about deliberate dose reduction ofIBD medications without knowledge of their physician, morethan a quarter of the cohort reported this, making it morecommon than low adherence (26.9% vs. 21.9%; p b 0.001),with no difference between cohorts (Fig. 1). Medicationswere proportionally affected by this behaviour in the sameway as reported above for low adherence. Interestingly, thevast majority (68.8%) of subjects reporting regular CDR,reported themselves as highly adherent by MMAS-4, on thebasis that they rarely missed medication doses, despiteconsistently under-dosing, unmasking a clinically relevantgap in this instrument's ability to represent true adherenceto therapy as prescribed.

Significant associations of CDR on univariate analysisincluded regular CAM use (37.1% vs 18.9%, p b 0.001),dissatisfaction with doctor communication (p = 0.037) andinformation provision (p = 0.001), a negative relationship withthe doctor (p = 0.027), personal relationship dissatisfaction(p = 0.035), and increased anxiety (p = 0.037). In contrastwith non-adherence findings, gender did not predict CDR (p =0.306), although younger age was associated with CDR (47.69versus 51.36 years, p = 0.016). Again disease type did notaffect CDR (p = 0.144).

3.6. Covert dose reduction predictors by logistic andlinear regression analyses

Regular Complementary and Alternative Medicine (CAM) usewas the strongest predictor of CDR, the 20.7% of subjectsconsulting an alternative practitioner being no more or lesslikely to dose reduce than those who obtained CAM indepen-dently (p = 0.854). Of all subjects in the study, 45.4% reportedregular CAM use (often or very often).

Adjusting for the same variables as for low adherence,associations of CDR are presented in Table 4 (adjusted pseudoR squared 0.178, goodness of fit Homer Lemeshow p = 0.256).These predictive factors were confirmed in a linear regressionmodel using factor scores for CDR; the overall model fit wasadjusted R squared 0.234 (Table 5).

Types of CAM used are reported in Table 6. A themeemerged from the free text responses of CAM users that“natural therapies” were efficacious enough to allow dose

ubjects using logistic regression.

Odds ratio 95% confidence interval p value

5.425 2.437–12.075 b0.0014.420 3.080–6.342 b0.0010.437 0.339–0.562 b0.0012.175 1.106–4.276 0.0300.500 0.331–0.754 b0.0011.889 1.062–3.363 0.0310.531 0.374–0.756 b0.0011.319 1.050–1.646 0.0171.331 1.071–1.654 0.0101.110 1.001–1.230 0.0471.107 1.057–1.160 0.001



Table 4 Independent predictors of covert dose reduction by logistic regression analysis.

Odds ratio 95% confidence interval p value

Regular CAM use 4.389 2.399–8.028 b0.001Believes what doctor says about IBD 0.321 0.122–0.843 0.021Feels alternative practitioner provides more information about IBD than doctor 3.104 1.650–5.839 b0.001Feels more comfortable with alternative practitioner 2.193 1.254–3.836 0.006Wants IBD doctor to focus more on psychological aspects 1.841 1.111–3.051 0.018Experienced adverse effects convention IBD medication 1.747 1.388–2.199 b0.001HADS anxiety 1.119 1.071–1.170 b0.001


reduction of “stronger” conventional meds, which mightmake them “safer”.

3.7. Psychological variables and QOL in contrastingIBD cohorts

No differences between cohorts were seen for HADSanxiety or depression, Trait anxiety, depression, anger, orQOL. The overall frequency of at least mild depression andmild anxiety (HADS N 7) was 33.7% and 27.7% respectively,consistent with other IBD populations.29 There was atrend toward less curiosity amongst Darwin compared withPrivate subjects (mean Spielberger score 24.0 versus 26.8,p = 0.065).

It was not possible to compare predictors ofnon-adherence and CDR by cohort due to the smaller sizeof Private and Darwin cohorts. On individual questionnaireitems, a trend toward more doctor shopping was observedamongst FMC versus Private subjects (38.3% versus 25.6%,p = 0.080), and increased self-reported depression in Darwinsubjects (45.7% versus 25.8% Private, p = 0.087), althoughno differences were seen between cohorts by HADS.

4. Discussion

This is the first study in Inflammatory Bowel Disease toinvestigate the important issue of covert dose reduction as a

Table 5 Independent predictors of covert dose reduction bylinear regression analysis.

Beta(correlationcoefficient)

pvalue

Regular CAM Use 0.228 b0.001Adverse effects conventionalmedications

0.149 0.007

Believes what doctor says aboutmedication

−0.148 0.006

Wants doctor to focus more onpsychological aspects

0.141 0.010

Feels more comfortable withalternative practitioner

0.145 0.091

HADS anxiety 0.100 0.094Feels alternative practitionerprovides more informationabout IBD than doctor

0.095 0.087


distinct type of medication non-adherence, and identify itsfrequency and attitudinal and psychological predictorsacross three patient cohorts.

Traditional non-adherence by MMAS-4 was reported by21.9% of subjects, associated with anxiety and depression,lack of social support, negative beliefs about efficacy andadverse effects, and negative patient–doctor relationships.Covert dose reduction was reported more frequently(26.9%), and was not detected by MMAS-4 as only 32.2% ofdose reducing subjects in our population met the criteria forlow adherence. Whilst other adherence measuring instru-ments such as MMAS-8 and the Medication AdherenceResponse Scale30 do assess dose reduction, responses tothis contribute to an overall adherence score, which doesnot distinguish between dose omission and reduction. Wealso found that CDR has different behavioural predictors,including regular CAM use, more positive consultationexperiences with alternative practitioners, and scepticismof doctor provided information.

The frequency of non-adherence and of covert dosereduction was similar across the three cohorts from contrast-ing IBD treatment settings, which indicates some generaliz-ability of these results. Supporting this is the similaritybetween predictors of traditional non-adherence betweenour populations and those reported internationally.

Non-adherence rates in this study are at the lower end ofthat reported previously,10,4,31,32,8 which may result fromparticipation bias and the self-reporting methodology. Also,a large proportion of subjects scored as “medium” adherent(34.8%) by self-report, many of whom may be classified aslow adherent using more quantifiable adherence assessmentmethods.

Regarding predictors of non-adherence, the presence of anegative doctor patient relationship was the strongestattitudinal risk factor for non-adherence in this study, andhas been observed in other populations.11,16,33 The finding ofan association between less satisfactory information provi-sion by the doctor and non-adherence contrasts with anotherAustralian study, however.10

The association between medication efficacy doubts andnon-adherence is also consistent with previous studies,9,10,16

as is the association with experience or fear of adverseeffects.10,34,9 In the broader context of other chronic diseases,a recent meta-analysis examining the “Necessity–ConcernsFramework”, confirmed that higher medication adherence isassociated with stronger perceptions of treatment necessityand fewer concerns about treatment.35 As in our study,depression has been identified as a risk factor fornon-adherence in other IBD populations,36–40 but not all.41



Table 6 Most commonly used CAM types reported by IBDsubjects.

Type of CAM Percentageof total CAMreportedoverall

Herbal remedies (e.g. slippery elm, aloevera juice, olive oil extract, greenlipped mussel oil)

30.5%

Probiotics 22.6%Fish oil 12.1%Chinese medicine 10.5%Acupuncture, massage, magnetism 10.5%Other (prayer, meditation, exercise, dietarysupplements, hypnotherapy)

13.7%


Whilst covert dose reduction of IBD medications has beennoted in 18% of patients in a previous non-adherencestudy,32 until now associations of this behaviour as a distinctentity have not been sought. Independent predictors of CDRin this study were notably different from those predictingoverall non-adherence. CAM use did not affect overalladherence in our cohorts, and this lack of association issupported in the literature.42,8,43 It was, however, thestrongest attitudinal predictor of CDR in this study, alongwith a preference for the consultation style of alternativepractitioners in terms of information provision, comfortlevel and believability. This association between CAM useand patient driven conventional medication under-dosagehas been reported recently in hypertension,18 and may applyin other chronic illness settings also. Consistent with thistheme of preferring a holistic approach, many subjects(28.1%) reported wanting more of a psychological focusduring doctor consultations, and this was another predictorof CDR.

Limitations of this study include the self-reporting natureof the questionnaire, which restricted the amount ofverifiable clinical information available, such as the typeand dosage of medication prescribed. The Darwin cohort wassmaller with a lower response rate, likely contributors beingthe lack of both coordinated IBD service and patientdatabase at the time of study which may have resulted inlesser engagement with care and incorrect contact detailsfor potential subjects.

Participation bias is likely, in that a very high proportion(98.7%) of subjects reported a good relationship with theirdoctor. It is likely that invitees with a less sanguine view mayhave chosen not to participate. Non-adherence rates mayhave been even higher in non-participants, with only themost adherent in the smaller Darwin cohort, for example,taking part, thus potentially masking true differences inadherence between IBD populations. Whilst a validated testwas used to assess overall adherence, the CDR testingmethod has not yet been validated.

Covert dose reduction of IBD medication may be a distinctsubtype of non-adherence with different attitudinal predic-tors to dose omission, which is not assessed as a separateentity using current adherence scales. Further work iswarranted to develop validated scales to measure this


phenomenon and to confirm predictors identified in thisstudy. The negative impact of CDR on disease control may beconsiderable and further investigation is justified not only inIBD, but also in the broader chronic disease population.

Conflict of interest

No conflict of interest exists for any authors with regard tothe content of this study.

Acknowledgements

The contribution of Ruth Prosser and Sarah Clark (FMC) fortheir assistance in questionnaire distribution, collection anddata entry is greatly appreciated.

Author contributions: RM was involved in the conception,design, seeking of ethical approval, data acquisition andanalysis, data interpretation, manuscript drafting andmodification and preparation of the final paper. JA and PBwere involved in planning the study, PB in maintaining theFMC database, and both JA and PB in data interpretation andrevising the manuscript. AMW assisted with data entry andanalysis. All authors read and approved the final manuscript.

This study did not receive funding from any source, andwriting assistance was not obtained.

References

1. Higgins PD, Rubin DT, Kaulback K, Schoenfield PS, Kane SV.Systematic review: impact of non-adherence to 5-aminosalicylicacid products on the frequency and cost of ulcerative colitisflares. Aliment Pharmacol Ther Feb 1 2009;29(3):247–57.

2. Chapman CG, Rubin DT. The potential for medical therapy toreduce the risk of colorectal cancer and optimize surveillance ininflammatory bowel disease. Gastrointest Endosc Clin N Am Jul2014;24(3):353–65.

3. Biancone L, Michetti P, Travis S, Escher JC, Moser G, Forbes A,et al. European evidence-based Consensus on the managementof ulcerative colitis: special situations. J Crohns Colitis Mar2008;2(1):63–92.

4. Jackson CA, Clatworthy J, Robinson A, Horne R. Factorsassociated with non-adherence to oral medication for inflam-matory bowel disease: a systematic review. Am J GastroenterolMar 2010;105(3):525–39.

5. Ediger JP, Walker JR, Graff L, Lix L, Clara I, Rawsthorne P, et al.Predictors of medication adherence in inflammatory boweldisease. Am J Gastroenterol Jul 2007;102(7):1417–26.

6. Kane SV, Cohen RD, Aikens JE, Hanauer SB. Prevalence ofnonadherence with maintenance mesalamine in quiescentulcerative colitis. Am J Gastroenterol Oct 2001;96(10):2929–33.

7. Mantzaris GJ, Roussos A, Kalantzis C, Koilakou S, Raptis N,Kalantzis N. How adherent to treatment with azathioprine arepatients with Crohn's disease in long-term remission? InflammBowel Dis Apr 2007;13(4):446–50.

8. Lakatos PL, Czegledi Z, David G, Kispal Z, Kiss LS, Palatka K,et al. Association of adherence to therapy and complementaryand alternative medicine use with demographic factors anddisease phenotype in patients with inflammatory bowel disease.J Crohns Colitis Sep 2010;4(3):283–90.

9. Horne R, Parham R, Driscoll R, Robinson A. Patients' attitudesto medicines and adherence to maintenance treatment ininflammatory bowel disease. Inflamm Bowel Dis Jun 2009;15(6):837–44.


http://refhub.elsevier.com//rf0005





































10. Selinger CP, Eaden J, Jones DB, Katelaris P, Chapman G,McDonald C, et al.Modifiable factors associatedwith nonadherenceto maintenance medication for inflammatory bowel disease.Inflamm Bowel Dis Sep 2013;19(10):2199–206.

11. Lopez San Roman A, Bermejo F, Carrera E, Perez-Abad M,Boixeda D. Adherence to treatment in inflammatory boweldisease. Rev Esp Enferm Dig Apr 2005;97(4):249–57.

12. Van Langenberg DR, AJ. Satisfaction with patient–doctorrelationships in inflammatory bowel diseases: examining patientinitiated change of specialist.World J Gastroenterol 2012;18(18):2212–8.

13. López San Román A, Bermejo F, Carrera E, Pérez-Abad M,Boixeda D. Adherence to treatment in inflammatory boweldisease. Rev Esp Enferm Dig 2005;97:249–57.

14. Linn AJ, van Dijk L, Smit EG, Jansen J, van Weert JC. May younever forget what is worth remembering: the relation betweenrecall of medical information and medication adherence inpatients with inflammatory bowel disease. J Crohns Colitis2013;7(11):e543–50.

15. Nigro G, Angelini G, Grosso SB, Caula G, Sategna-Guidetti C.Psychiatric predictors of noncompliance in inflammatory boweldisease: psychiatry and compliance. J Clin Gastroenterol Jan2001;32(1):66–8.

16. Sewitch MJ, Abrahamowicz M, Barkun A, Bitton A, Wild GE, CohenA, et al. Patient nonadherence to medication in inflammatorybowel disease. Am J Gastroenterol Jul 2003;98(7):1535–44.

17. Shale MJ, Riley SA. Studies of compliance with delayed-releasemesalazine therapy in patients with inflammatory boweldisease. Aliment Pharmacol Ther 2003;18(2):191–8.

18. Lee K, Mokhtar HH, Krauss SE, Ong BK. Hypertensive patients'preferences for complementary and alternativemedicine and theinfluence of these preferences on the adherence to prescribedmedication. Complement Ther Clin Pract May 2014;20(2):99–105.

19. Vrijens B, Urquhart J. Methods for measuring, enhancing, andaccounting for medication adherence in clinical trials. ClinPharmacol Ther Jun 2014;95(6):617–26.

20. Robinson A. Review article: improving adherence tomedication inpatients with inflammatory bowel disease. Aliment PharmacolTher Mar 2008;27(Suppl 1):9–14.

21. Morisky DE, Green LW, Levine DM. Concurrent and predictivevalidity of a self-reported measure of medication adherence.Med Care Jan 1986;24(1):67–74.

22. Morisky DE, Ang A, Krousel-Wood M, Ward HJ. Predictivevalidity of a medication adherence measure in an outpatientsetting. J Clin Hypertens May 2008;10(5):348–54.

23. Kim E, Gupta S, Bolge S, Chen CC, Whitehead R, Bates JA.Adherence and outcomes associated with copayment burden inschizophrenia: a cross-sectional survey. J Med Econ 2010;13(2):185–92.

24. Zigmond AS, Snaith RP. The hospital anxiety and depressionscale. Acta Psychiatr Scand Jun 1983;67(6):361–70.

25. Irvine EJ, Zhou Q, Thompson AK. The Short Inflammatory BowelDisease Questionnaire: a quality of life instrument for commu-nity physicians managing inflammatory bowel disease. CCRPTInvestigators. Canadian Crohn's Relapse Prevention Trial. Am JGastroenterol Aug 1996;91(8):1571–8.

26. Spielberger CD. Cross-cultural assessment of emotional statesand personality traits. Eur Psychol 2006;11(4):297–303.

27. Spielberger C. Manual for the State Trait Personality Inventory.Palo Alto CA: Consulting Psychologist Press; 1970.

28. Forsberg C, Bjorvell H. Swedish population norms for the GHRI,HI and STAI-state. Qual Life Res Int J Qual Life Asp Treat CareRehab Oct 1993;2(5):349–56.


29. Panara AJ, Yarur AJ, Rieders B, Proksell S, Deshpande AR, AbreuMT, et al. The incidence and risk factors for developing depressionafter being diagnosed with inflammatory bowel disease: a cohortstudy. Aliment Pharmacol Ther Apr 2014;39(8):802–10.

30. Jaeger S, Pfiffner C, Weiser P, et al. Adherence styles ofschizophrenia patients identified by a latent class analysis ofthe Medication Adherence Rating Scale (MARS): a six-monthfollow-up study. Psychiatry Res 2012;200(2–3):83–8.

31. Yen L, Wu J, Hodgkins PL, Cohen RD, Nichol MB. Medication usepatterns and predictors of nonpersistence and nonadherencewith oral 5-aminosalicylic acid therapy in patients withulcerative colitis. J Manag Care Pharm Nov–Dec 2012;18(9):701–12.

32. Cerveny P, Bortlik M, Kubena A, Vlcek J, Lakatos PL, Lukas M.Nonadherence in inflammatory bowel disease: results of factoranalysis. Inflamm Bowel Dis Oct 2007;13(10):1244–9.

33. D'Inca R, Bertomoro P, Mazzocco K, Vettorato MG, Rumiati R,Sturniolo GC. Risk factors for non-adherence to medication ininflammatory bowel disease patients. Aliment Pharmacol Ther2008;27(2):166–72.

34. Bokemeyer B, Teml A, Roggel C, Hartmann P, Fischer C,Schaeffeler E, et al. Adherence to thiopurine treatment inout-patients with Crohn's disease. Aliment Pharmacol Ther2007;26(2):217–25.

35. Horne R, Chapman SC, Parham R, Freemantle N, Forbes A,Cooper V. Understanding patients' adherence-related beliefsabout medicines prescribed for long-term conditions: ameta-analytic review of the Necessity–Concerns Framework.PLoS One 2013;8(12):e80633.

36. Goodhand JR, Kamperidis N, Sirwan B, Macken L, Tshuma N,Koodun Y, et al. Factors associated with thiopurine non-adherencein patients with inflammatory bowel disease. Aliment PharmacolTher Nov 2013;38(9):1097–108.

37. Long MD, Kappelman MD, Martin CF, ChenW, Anton K, Sandler RS.Risk factors for depression in the elderly inflammatory boweldisease population. J Crohns Colitis Feb 2014;8(2):113–9.

38. McCombie AM, Mulder RT, Gearry RB. Psychotherapy forinflammatory bowel disease: a review and update. J CrohnsColitis 2013;7(12):935–49.

39. Nahon S, Lahmek P, Durance C, Olympie A, Lesgourgues B,Colombel JF, et al. Risk factors of anxiety and depression ininflammatory bowel disease. InflammBowel DisNov 2012;18(11):2086–91.

40. Gray WN, Denson LA, Baldassano RN, Hommel KA. Treatmentadherence in adolescents with inflammatory bowel disease: thecollective impact of barriers to adherence and anxiety/depressivesymptoms. J Pediatr Psychol Apr 2012;37(3):282–91.

41. Nahon S, Lahmek P, Saas C, Durance C, Olympie A, LesgourguesB, et al. Socioeconomic and psychological factors associated withnonadherence to treatment in inflammatory bowel diseasepatients: results of the ISSEO survey. Inflamm Bowel Dis Jun2011;17(6):1270–6.

42. Weizman AV, Ahn E, Thanabalan R, Leung W, Croitoru K,Silverberg MS, et al. Characterisation of complementary andalternative medicine use and its impact on medicationadherence in inflammatory bowel disease. Aliment PharmacolTher Feb 2012;35(3):342–9.

43. Nousiainen P, Merras-Salmio L, Aalto K, Kolho KL. Complementaryand alternative medicine use in adolescents with inflammatorybowel disease and juvenile idiopathic arthritis. BMC ComplementAltern Med 2014;14(1):124.
































































































































Documents

Covert dose reduction is a distinct type of medication non-adherence observed across all care settings in inflammatory bowel disease