The balance evaluation systems test (BESTest), mini-BESTest and brief-BESTest as clinical tools to assess balance control across different populations: A reliability generalization meta-analysis

Background: The Balance Evaluation Systems Test (BESTest) and two abbreviated versions, Mini-BESTest and Brief-BESTest are used to assess functioning of balance control systems. Its reliability across different populations remains to be determined. Objective: The present study followed reliability generalization procedures to estimate an average internal consistency and inter and intra-rater reliability for the BESTest, Mini-BESTest and Brief-BEStest. In this study, the heterogeneity of reliability coefficients in each instrument is evaluated. If heterogeneity is significant, a moderator analysis is performed to identify the characteristic which explains such variability. Methods: A search of the PubMed, Embase, PsycINFO, Web of Science, Scopus and CINAHL databases was carried out to February 10th 2024. Two reviewers independently selected empirical studies published in English or Spanish that applied the BESTest, Mini-BESTest and/or Brief-BESTest and reported any reliability coefficient and/or internal consistency with data at hand. Results: Sixty-four studies reported any reliability estimate BESTest, Mini-BESTest and/or Brief-BESTest scores (N. = 5225 participants). Mean Cronbach alpha for the Mini-BESTest and Brief-BESTest (total score = 0.92) indicating no variability in estimated internal consistency. Likewise, no variability was obtained for inter-rater and intra-rater mean agreement of the BESTest (ICC = 0.97; 0.94), Mini-BESTest (ICC = 0.95; 0.94) and Brief-BESTest (ICC = 0.96; 0.95). Mean scores, standard deviation of scores, mean age, gender, population type, mean history of the disorder, disease, raters´ experience, number of raters, rater formation, continent of study and design type presented statistically significant relationships with ICC and/or Cronbach´s alpha for BESTest and the two abbreviated versions. Conclusions: The mean intraclass correlations and Cronbach alpha obtained for BESTest, Mini-BESTest and Brief-BESTest exhibited an excellent inter and intra-rater reliability and internal consistency. The average reliability obtained three scales adequate to be applied for screening balance problems in different populations. Some continuous and categorical moderator variables increase reliability and internal consistency of these scales.