Viés de relato nos títulos de revisões sistemáticas sobre fisioterapia musculoesquelética e neurofuncional: um estudo meta-epidemiológico

Abstract

Introduction: Systematic reviews (SRs) are highly trusted within the scientific community and are often chosen by physiotherapists to support decisions in clinical practice. However, despite their relevance, SRs are not free from reporting biases, especially in the title. Biases in the title can distort information in order to make it more attractive, which can compromise the integrity of the study and influence the reader's perception. Objective: To analyze the prevalence of reporting biases in the titles of SRs on musculoskeletal and neurofunctional physiotherapy. Materials and methods: A meta-epidemiological systematic review was conducted. The top 10 journals in the rehabilitation field were selected and ranked based on the 2022 Journal Impact Factor, as recorded in the Journal Citation Reports (JCR) database. All articles published between March 2020 and March 2023 were extracted from these journals. The SRs in the musculoskeletal and neurofunctional areas were selected using the Rayyan software. The titles of each SR were analyzed, with emphasis on identifying four types of reporting bias in the title: (1) Exclusive citation of secondary outcomes in the title; (2) Specific selection of secondary outcome in the title; (3) Assertion of effect based on secondary outcome; (4) Absence of the expression “Systematic Review”. Each bias was assessed for presence (total or partial) or absence. Descriptive analyses were performed with absolute and percentage values, as well as association analyses using the chisquare and/or Fisher tests. A binary logistic regression was conducted to assess the association between the presence of bias with the journal impact factor and area. All analyses were conducted using Jamovi software. Results: A total of 3,586 studies were identified, of which 3,480 were excluded for not meeting the eligibility criteria. In total, 106 SRs were included. In the musculoskeletal area, of 50 reviews, 20 SRs presented reporting bias in the titles, resulting in a prevalence of 40%. In the neurofunctional area, of 56 reviews, 16 presented bias, corresponding to 29%. There was no association between area and journal with the presence of bias. The main journals with bias in their publications were Disability and Health Journal (100%), Neurorehabilitation and Neural Repair (44%) and Journal of NeuroEngineering and Rehabilitation (40%). Furthermore, there was no statistically significant evidence that the combination of the variables area and impact factor of the journals are associated with the prevalence of bias. Biases related to the secondary outcome were frequent in both areas. Conclusion: The prevalence of reporting biases in the titles of SRs in the areas of musculoskeletal and neurofunctional physiotherapy was identified. However, there was no statistically significant association between the study area, the type of journal, and the impact factor influencing the prevalence of bias. The investigation also revealed that the most common types of biases were related to the lack of specification of primary and secondary outcomes in the text body of the articles. Limitations: This study is a pioneer in this topic; however, it has some limitations, such as the lack of validation of the instrument used to identify reporting biases in the titles. In addition, there is no knowledge about the effects of the presence of these biases on the readers' interpretation or their real impact on clinical practice. Implications: The findings of this research encourage the creation of instruments that provide more efficient guidelines to mitigate bias. Such a measure could ensure the rigorous adoption of good practices in the preparation of studies, especially in the writing of titles, contributing to the improvement of methodological quality. These practices could provide significant benefits to researchers, health professionals and patients in general, ensuring more reliable interpretations and safer clinical practices