Study designs and statistical approaches for bilateral carpal tunnes syndrome: An overview

Background: Pathologies with bilateral involvement, such as carpal tunnel syndrome (CTS), are relatively common in clinical practice. However, some published data are misleading, as many articles consider only one hand in data analysis. Objectives: This article aims to briefly propose a study design and statistical approach for data analysis of bilateral CTS. Method: Statistical reporting. Results: Although the results of clinical and surgical interventions are usually reported by randomized clinical trials, the main guidelines do not offer recommendations on how to proceed in cases of interventions in patients with bilateral conditions. Additionally, crossover trials may be an alternative, particularly when comparing different interventions in these patients. Considering the statistical approach, traditional tests are not suitable for bilateral conditions, and result in an overestimation of the results. In contrast, regression models, mixed effects analysis, generalized estimating equations, and multilevel modeling analysis are more reliable. Furthermore, in the case of crossover trials, an ANOVA suitable for crossover design should be chosen with normally distributed data of two groups, while a variance-balanced design is the ideal choice for three or more treatments and Cochran’s Statistics. Conclusions: When considering the comparison of different therapeutic intervention/rehabilitation techniques in patients with CTS or other peripheral nerve pathologies, the choice of the most appropriate study design and statistical analysis will provide more reliable evidence.