Limited Clinical Utility of a Self-Administered Tool for Predicting Medication Adherence Behavior in IBD

Abstract

Background and Aims: Identification of non-adherence is an important component of shortand long-term management of IBD. A validated screening tool for non-adherence in IBD would help determine those patients who may be at highest risk for non-adherent behavior and potentially worse disease outcomes. Our aim was to determine if scores from a selfreported adherence survey of IBD patients correlated with pharmacy refill data, as a reliable measure of medication adherence. Methods: The self-reported tool was an 8-item survey, the Moritsky Medication Adherence Scale, which has been previously validated in patients with hypertension. Each question is worth a point with a maximum score of 8. A score of <4 was defined as a patient at high risk for non-adherence, and a score of 7-8 at no risk for non-adherence. The study was IRB-approved and informed consent was obtained. Surveys were completed at the time of a routine clinic appointment. Patient pharmacies were contacted at 3 time points: time of enrollment for refill information regarding the previous 3 months, then 3 months and 6 months from enrollment. Medications discontinued due to intolerance or non-response before the 6-month time period were excluded. Refill data were recorded for each time interval as the Medication Possession Ratio (MPR) and adherence was defined as greater than 80%. Analysis of variance with Pearson's correlation coefficient was used to determine the relationship between survey scores and MPR by drug class. Results: One hundred fifty consecutive patients were enrolled. Ninety-four patients had Crohn's disease and 56 had ulcerative colitis. Eighty-nine patients (59%) were female. Thirty six percent of patients were on a 5-ASA, 41% on a thiopurine, 11% on infliximab, 6% on an injectable biologic, and 5% on budesonide. The median adherence score was 7 (range, 0-8). Fiftytwo percent stated they rarely missed a dose of medication. However, the adherence by refill data ranged from 25% to 71% by drug class. Only those on a thiopurine had a survey score that positively correlated with adherence (p = 0.02). For all other medication classes, there was no correlation between reported scores and refill data either prior to or 6 months following survey completion. Conclusion: Overall, adherence with IBD therapies was low. Only those on a thiopurine were likely to have an adherence score that predicted high refill behavior. Adherence with therapy for IBD is complex and cannot be predicted in a reliable manner by a self-reported survey tool validated for other chronic conditions.