P345 Inconsistency between electronic data of patient's adherence and self reported adherence score

    loading  Checking for direct PDF access through Ovid

Abstract

Background: Long term treatment adherence in chronic diseases is notably poor. In inflammatory bowel disease (IBD) adherence consists of compliance with diagnostic tests, endoscopies, laboratory workup, follow up appointments and adherence to long term medical treatment. There are several methods to evaluate adherence. Morisky score is a subjective questionnaire filled by the patient while electronic data collected by the physician provides an objective tool. This study aimed to compare adherence analyzed by the Morisky score versus the gastroenterologists impression and medication purchase extracted from electronic data.

Methods: IBD patients were asked to fill a questionnaire including: demographic and disease parameters and the Morisky score to evaluate adherence. Physicians filled a questionnaire summarizing the impression of patient adherence as reflected from electronic patient files.

Results: Data from 214 IBD patients was available. Mean age was 37±14 years, 73 (34%) were males, 151 (70.6%) had Crohn's disease 50% recieved biologic treatment. The patients estimation of treatment efficacy was positive in 76%. Main reasons for non-adherence reported by the patients were: busy 19%, forget 36%, does not help 7% and side effects 13%. Electronic data regarding medications was only mildly correlated with the Morisky score (r=−0.0302, p<0.001). The doctor's comprehensive evaluation of patient's adherence was also only mildly correlated with the Morisky score (r=−0.0334, p<0.001). Compared with Morisky score, the adherence in 53% of patients was overestimated by the physician, while 12% were underestimated.

Conclusions: Current modes of adherence estimation are inaccurate and non-consistent. Mismatch between them emphasizes their limitation. An accurate estimation tool to evaluate patient adherence is needed.

Related Topics

    loading  Loading Related Articles