The scale of self-efficacy expectations of adherence to antiretroviral treatment: A tool for identifying risk for non-adherence to treatment for HIV

Abstract

Background: Identification of risk for non-adherence to treatment is a challenge for personalized care for people living with HIV. Standardized questionnaires of patients' expectations of their capability to overcome obstacles for treatment adherence may be used as a pre-screening for risk identification. A scale of self-efficacy expectations of adherence to antiretroviral treatment (SEA-ART scale) was previously developed. This study assesses the scale validity in predicting non-adherence to ART in adults living with HIV. Methods and Findings: A prospective cohort study applied a 21-item SEA-ART scale to 275 adults in ART treatment at an outpatient public service for HIV in Southern Brazil. ART medications taken were assessed at one-month follow-up; ART adherence was devised as an intake of 95% and more of the prescribed medication. A SEA-ART score was calculated by adding up the scores of all items. Multivariable logistic regression and the Area Under the Receiver-Operating-Characteristic Curve (AUROC) were applied to examine the ability of the SEA-ART score to predict non-adherence at follow-up. The SEA-ART score varied from 21 to 105; mean 93.9; median 103.0. Non-adherence was 30.3% (n = 81/267). The odds of non-adherence was 8% lower for each unit increase of the SEA-ART score; after adjustment for age, sex, formal education and time in treatment (OR = 0.92; 95%CI 0.90-0.95; LRT for linear trend, p = 0.002). The AUROC was 0.80 (95%CI 0.73-0.87; p<0.001). The SEA-ART optimal cut-off value was 101, providing a sensitivity of 76.5%, a specificity of 73.1 %, a positive predictive value of 55.4% and a negative predictive value of 87.7%. There was no evidence of difference in sensitivity, and specificity among groups organized by age, gender, formal education and time in treatment. Conclusions: The SEA-ART scale appears to have a good capacity to discriminate between adherents and non-adherents at one-month follow-up. Further studies should confirm these results in other populations.