Análise do risco de viés em estudos de revisão sistemática sobre a relação entre as alterações do valgo ou varo do joelho e a melhora da dor ou função nos distúrbios do joelho

Abstract

Knee pain is the second most prevalent musculoskeletal disorder, with osteoarthritis (OA), patellofemoral pain (PFP) and anterior cruciate ligament injury (ACL) being the most prevalent and frequent. Knee alignment in frontal and transverse planes may predispose and/or accelerate the evolution of OA, PFP and ACL lesion. Despite the potential biomechanical benefit of the various conservative treatments to improve the dynamic alignment of the lower limb in the frontal and transverse planes, it is still unclear whether, in fact, correcting the valgus and/or varus of the knee in patients with knee disorders is associated with improved pain and/or function. An important step in the search for this issue would be to verify the risk of bias in clinical trials that evaluated the valgus/knee varus, pain and/or function. Thus, the main objective of this study was to analyze the risk of bias in studies that evaluated valgus or varus changes in the knee and the improvement of pain or function of people with knee disorders from studies included in a systematic review. For this purpose, a systematic review with risk of bias analysis was performed. The systematic review was conducted in accordance with the recommendations of Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA). Electronic searches were conducted in five databases: Medline and EMBASE via Ovid, CINAHL and SPORTDiscus via EBSCO and Web of Science. Two independent reviewers performed the selection and a third researcher was consulted when there was disagreement. Data extraction and analysis were conducted by two independent researchers. The risk of bias evaluation of the included studies was performed using the Revised Cochrane risk-of-bias tool for randomized trials (RoB 2). Result: Of the 20 studies included, 15% (n=3) were considered low risk of bias, 35% (n=7) presented methodological points (randomization process, deviations from intended interventions, measurement of the result and selection of the reported result) that leave concerns about the risk of bias and 50% (n=10) were considered at high risk of bias. The domains related to the measurement of the result (Domain 4), deviations from the intended interventions (Domain 2) and selection of the reported result (Domain 5) presented the highest percentage of high risk of bias. To reduce the risk of bias, avoiding underestimation or overestimation of the true intervention effect, it is suggested that future studies use evaluators, therapists and blind participants regarding the intervention received by the participants, perform the intention-to-treat analysis, and report the previous data analysis plan.