Psychological interventions to improve adherence to antiretroviral treatment. Systematic review and meta-analysis

Abstract

HIV is a chronic disease that affects the CD4 receptors of the immune system, so it is necessary for people with the virus to be under Antiretroviral Treatment (ART) to reduce the viral load in the body. The success of ART is related to adherence to treatment, under-stood as: the intake of antiretroviral drugs, follow-up of medical controls and implementation of healthy lifestyles. There are personal, social and health systems problems that prevent people from adhering to ART in an adequate way. This situation has become a public health problem because inappropriate taking of drugs often leads to drug resistance. In this way, it is necessary to implement effective psychological interventions aimed at increasing adherence to ART, to contribute to a better use of antiretrovirals and to the improvement of the quality of life of people with HIV. Accordingly, the objective was to review the factors associated with the efficacy of psychological interventions, to increase adherence to ART and to aggregate the findings in quantitative estimates of its impact on adherence. For this purpose, meta-analysis was used as a research technique that allows a quantitative analysis of results by calculating specific statistics. The literature search and review were carried out from August 2017 to May 2019 in the PubMed, Academic Search Complete (EBSCO Host) and Springer Link databases, for the identification of experimental studies that met the eligibility criteria (studies published in peer-reviewed indexed journals, experimental or quasi-experimental research, using techniques to increase adherence, published in Spanish and English and reporting the necessary statistics for effect size calculation). Of the 15 studies (total n = 1669) selected, 38 effect sizes were obtained. The calculation of the individual and global effect size was performed using the Comprehensive Meta-Analysis software (Biostat, 2011, v. 2.2.064). All the effect sizes were calculated using the standardized mean difference and a random effects model. To find out if individual effect sizes were homogeneous and therefore represent a similar measure of treatment efficacy, the Cochran Q homogeneity test was applied; based on this, the I2 was calculated. Also, it was performed a publication bias calculation, the objective of which was to estimate the number of non-significant effect sizes that are required for a reduction of the global effect size to a smaller and non-significant one. The overall effect size was moderate (d = 0.56, n = 38, 95 % CI = 0.39 - 0.72, p