The accuracy of adherence self-report scales in patients on thiopurines for inflammatory bowel disease: a comparison with drug metabolite levels and medication possession ratios

Background: Adherence to maintenance medication for inflammatory bowel disease (IBD) is essential for disease control, albeit often poor. Adherence can be measured by drug metabolites, self-report tools, and prescription data. The aim of this study was to test implementation of self-report tools in IBD clinics by evaluating consistency and to validate them by correlation with drug metabolite levels and medication possession ratios (MPRs). Methods: Ambulatory IBD patients on thiopurine maintenance therapy for >3 months were recruited. Patients self-reported adherence using a visual analog scale (VAS) and Medication Adherence Report Scale (MARS). Thiopurine metabolites levels were assessed using blood, and MPRs were calculated from patient records as the reference standard. Consistency was assessed by McNemar's test (primary outcome), and correlation analysis was performed using Pearson tests. Results: Of 96 patients (58 Crohn's disease, 33 ulcerative colitis, 5 IBD unclassified) 16.6% were classified as nonadherent based on thiopurine metabolites, 14.9% based on VAS, 13.2% based on MARS, and 22.9% based on MPR. VAS and MARS were consistent with thiopurine metabolites (McNemar test P = 0.79, P = 0.45). All 4 methods were consistent with each other when compared directly 1 to 1. Spearman's analysis demonstrated that all 4 methods significantly correlated with each other: (correlation between VAS and thiopurine metabolites: rho = 0.435; P < 0.001; and correlation between MARS and thiopurine metabolites: rho = 0.29; P = 0.005). Conclusions: Self-report tools correlate significantly with thiopurine metabolites and medication possession ratios. The Medication Adherence Report Scale and VAS are validated adherence assessment tools for IBD and can be used as simple screening tools in clinical practice.