Prevalence of nonadherence with maintenance mesalamine in quiescent ulcerative colitis

OBJECTIVE: There are scant data regarding outpatient adherence in quiescent ulcerative colitis aside from patients enrolled in controlled clinical trials. We conducted a prevalence study to determine the medication adherence rate of maintenance therapy and to identify possible risk factors for nonadherence. METHODS: Outpatients with clinically quiescent ulcerative colitis for >6 months on maintenance mesalamine (Asacol, Procter and Gamble, Cincinnati, OH) were eligible. Patients were interviewed regarding disease history, and demographics were obtained from medical records. Refill information for at least 6 months was obtained from computerized pharmacy records. Adherence was defined as at least 80% consumption of supply dispensed. Using nonadherence as the outcome of interest, stratified analysis and regression modeling were used to identify significant associations. RESULTS: Data were complete for the 94 patients recruited. The overall adherence rate was found to be 40%. The median amount of medication dispensed per patient was 71% (8-130%) of the prescribed regimen. Nonadherent patients were more likely to be male (67% vs 52%, p < 0.05), single (68% vs 53%, p = 0.04), and to have disease limited to the left side of the colon versus pancolitis (83% vs 51%, p < 0.01). Sixty-eight percent of patients who took more than four prescription medications were found to be nonadherent versus only 40% of those patients taking fewer medications (p = 0.05). Age, occupation, a family history of inflammatory bowel disease, length of remission, quality-of-life score, or method of recruitment (telephone interview vs clinical visit) were not associated with nonadherence. Logistic regression identified that a history of more than four prescriptions (odds ratio OR 2.5 1.4-5.7) and male gender (OR 2.06 1.17-4.88) increased the risk of nonadherence. Two statistically significant variables, which were protective against nonadherence, were endoscopy within the past 24 months (OR 0.96 0.93-0.99) and being married (OR 0.46 0.39-0.57). CONCLUSION: Nonadherence is associated with multiple concomitant medications, male gender, and single status. These patient characteristics may be helpful in targeting those patients at higher risk for nonadherence.