PDB52 Adherence with Oral Medications for Diabetes Among Brazilian Patients: A Systematic Review of National Literature

Abstract

To identify studies examining adherence with oral diabetes mellitus (DM) medication and the potential association between adherence rates and glycemic control among Brazilian patients. A systematic literature search was performed by two independent reviewers using MEDLINE via Pubmed and LILACS databases (until May 2011) without limits for time or language. Specific filters to identify studies assessing Brazilian population were not used in the search strategy and this assessment was conducted by reviewers. Publications were included only if adequate documentation of adherence and population could be abstracted (adherence outcomes, thresholds used, and characteristics of the populations). The search strategy identified 289 records (Pubmed=174 and LILACS=115), from which only 2 cross-sectional studies met the eligibility criteria and were included in the systematic review. The most recent study (Araujo 2010) was conducted in 2007 and evaluated 79 DM patients using the 4-item Morisky-Green Test. Gimenes 2006 (n=31) was designed to investigate if DM patients have proper knowledge of their prescription and assessed adherence through self-reported patient compliance with medication schedule. Araujo 2010 found that 54.4% of DM patients were considered non-compliant according to the Morisky-Green Test. Taking medication in the wrong schedule and skipping doses were referred by 54.5% and 34% of patients, respectively. Gimenes 2006 observed that 48.4% of patients reported taking medication in the wrong schedule and 71% of them were classified as having unsatisfactory knowledge about their prescription. Studies examining the association between adherence and glycemic control were not found. This review reinforced the lack of adherence data for Brazilian DM population, but the included studies confirmed that a significant group of DM patients were poor compliers with treatment, although their findings should be interpreted with some concerning given the small sample size and explanatory nature.