Test performance of self-report adherence tools in patients with hypertension: A systematic review and a meta-analysis

What Is Known and ObjectivesAdherence has proved to have a positive influence on achieving plausible treatment outcomes. Self-report questionnaires are widely used in evaluating adherence, creating thus a high-powered research field. This review aims to provide an update of scales used in hypertension, which are compared and analysed against reliability and validity. MethodsPubMed, Web of Science and Cochrane Library were searched in May 2022 to identify studies. We extracted the study characteristics and evaluated their quality. A random-effects model with subgroup analysis was used to calculate estimates and heterogeneity parameters as well as regressions, funnel and forest plots. A bivariate model was selected to conduct validity analyses and draw receiver operating characteristic (ROC) curves. Results and DiscussionFifty-five articles were identified and classified into 22 different reliable and validated tools. Pooled analyses predicted an overall good Cronbach's alpha of 0.76 (95%CI:0.67-0.83), a good ICC of 0.8 (95%CI:0.72-0.86) and an excellent correlation coefficient of 0.91 (95%CI:0.86-0.95), which all showed high heterogeneity and slight detection of asymmetry. Regression analyses showed that only time and the number of items/scale type influenced significantly retest and alpha, respectively. Overall validity showed acceptable sensitivity of 0.65 (95%CI:0.53-0.75) and specificity of 0.57 (95%CI:0.47-0.67) with a good Area Under Curve (AUC) of 0.637. Upon comparison, four tools showed superiority over Morisky's scale. What Is New and ConclusionAdherence is a multi-dimensional phenomenon, which deems scales to be highly variable or complex; thus, complicating the selection process. Adherence to Refills and Medications Scale (ARMS) is the most promising free non-inferior alternative to Morisky, the most used scale.

Automated summary

This review provides an update on the scales used in hypertension to evaluate adherence, comparing and analyzing their reliability and validity. A total of 22 reliable and validated tools were identified, with good overall Cronbach's alpha, ICC, and correlation coefficient values. The validity analysis showed acceptable sensitivity and specificity, with a good Area Under Curve.