Desenvolvimento e validação de escala de percepção de dificuldades com o tratamento antirretroviral: avaliação preditiva da não adesão

Abstract

The aim of this work was to develop and validate a scale of perceived barriers with antiretroviral treatment (ART) in people living with HIV (PLHIV): the PEDIA scale. Five studies were conducted to meet the steps for the development of a new instrument. In study #1, an integrative literature review was conducted to identify perceived barriers instruments used to predict non-adherence to ART. A conceptual matrix of perceived barriers was used to evaluate the instruments’ breadth and depth in representing the construct. No instrument fully captured both the breadth and depth of perceived barriers. Additionally, only two measures demonstrated predictive criterion validity. Next, study #2 was carried out to obtain the pilot version of the scale. A total of 47 items were constructed based on the integrative review and a previous qualitative study and was then submitted to content validation. After semantic analysis with the target population and the evaluation of three judges regarding item relevance, adequacy and dimensionality, the pilot version with 40 items was obtained. In study #3, the pilot PEDIA was administered to 415 HIV-infected adults in the beginning of treatment, recruited from three healthcare facilities in the city of Belo Horizonte. After analyses such as exploratory factor analysis, internal consistency, item response theory analysis, and temporal stability, the measure was reduced to 18 items distributed in three dimensions as follows: cognitive and routine problems (4 items); medication and health concerns (6 items); and patient`s fears and feelings (8 items). Results demonstrate that the PEDIA is internally consistent and yields stable scores over time. In study #4 a systematic review with pairwise meta-analyses was conducted to compare the effects of any two indirect adherence measurements on virologic outcomes. Meta-analyses including 38 studies showed that low-cost and simple adherence measures, such as self-report, predict virologic failure better than or equally well as objective measures. In study #5, patients from study #3 were followed in a prospective cohort to examine the ability of the PEDIA scores to predict non-adherence at follow-up. ART non-adherence was assessed at follow-up (three to four months) by self-report and pharmacy refill. Logistic regression models established that the PEDIA provided good predictive validity. The models were used to propose an index of the probability of non-adherence in the future. The probability of non-adherence calculated by the index was strongly associated with detectable viral load six months later. Findings suggest that the PEDIA is a psychometrically sound tool for evaluating perceived barriers and predicting future non-adherence in adult patients initiating ART. It can be utilized in both research and clinical practice settings for early identification of patients at risk for non-adherence and treatment failure. In addition, the identification of potentially modifiable barriers may assist in the development of behavioral-changing interventions.