Predictor and complexity indexes as predictors of antiretroviral therapy discontinuation

Objective: To determine the role of complexity and predictor indexes as predictors of antiretroviral therapy discontinuation. Method: Prospective observational study. HIV patients over 18 years attended in a pharmacy department in 2011 were included. The variables collected were: demographic parameters, predictive value, complexity index and the cause of treatment discontinuation according to the Swiss HIV cohort. The predictive value was established using the index published by Morillo et al. This tool classifies patients according to the risk of developing drug related problems. Complexity index was calculated based on the score created by Martin et al. Statistical analysis was performed using SPSS 20.0. Results: 600 patients were included. 79% were male. The average age was forty-five years. 72% had low predictor index and 58% low complexity index. In the patient group who discontinued, 25% had high predictor index and 55% high complexity index. The univariate analysis was conducted in three groups: patients who discontinued for any reason, discontinuation for drug related problems and discontinuation excluding drug related problems. The multivariate analysis was realized in the third group (test of Hosmer and Lemeshow p < 0.05). The complexity index was the only predictor factor of antiretroviral therapy discontinuation increasing the risk 1.4 (95% CI: 1.23-1.58). Conclusions: The high complexity index increased the risk of antiretroviral therapy discontinuation. The predictor index stratifies the patients according to drug related problems but it did not predict the antiretroviral therapy discontinuation