Measure of adherence to direct-acting antivirals as a predictor of the effectiveness of hepatitis C treatment

Background Adherence to direct-acting antivirals could be a predictor response to these treatments in hepatitis C. Objective To assess the ability of three methods of measuring adherence to direct-acting antivirals [pill counts, pharmacy dispensing record and Simplified Medication Adherence Questionnaire (SMAQ)] as predictors of their effectiveness. Setting Study conducted by the pharmacy department of the hospital. Methods: A retrospective study was performed. Patients >= 18 years with hepatitis C that started and completed treatment with direct-acting antivirals between the 1st-April-2015 and 28st-February-2016 were enrolled. To evaluate the predictive ability to obtain a response to treatment, Chi squared test, Mann-Whitney-U test and ROC-curves were used. Main outcome measure Adherence to antivirals was assessed by three methods and response to treatment, which was defined as obtaining a viral load of hepatitis C virus <= 15UI/ml at week 12 after the end of treatment. Results 128 patients were enrolled. The overall average adherence obtained with SMAQ (99.09%) was similar to the pill counts (96.40%, p = 0.043) and pharmacy dispensing record (91.10%, p = 0.02). There was no correlation between the percentage of patients considered as adherent by SMAQ (99.09%) and the achievement of response to treatment (96.40%, p = 0.999). The ROC-curve obtained for the pill count method shows a global area under the curve of 0.53. For pharmacy dispensing record method, patients with an adherence <= 66.66% have a high probability of not achieving response (sensitivity and specificity of 79.00% and 100.00%, respectively). Conclusions Pharmacy dispensing record is shown as the best indicator of adherence to predict therapeutic failure in our study.