Reproducibility of the Balance Evaluation Systems Test (BESTest) and the Mini-BESTest in school-aged children

This study evaluated the intra-rater, inter-rater and test-retest reproducibility of the Full-BESTest and Mini-BESTest when assessing postural control in children. Thirty-four children aged 7-17 years participated in intra-rater and inter-rater evaluation, and 22 children repeated assessment six weeks later for evaluation of test-retest reliability. Postural control was assessed using the Full Balance Evaluation Systems Test (Full-BESTest) and the short-form Mini-BESTest. Intra-rater, inter-rater and test-retest reproducibility were examined using video assessment. Test-retest reproducibility was also assessed in real-time. Reproducibility was examined by agreement and reliability statistics. Agreement was calculated using percentage of agreement, Limits of Agreement and Smallest Detectable Change. Reliability was calculated using Intra-class Correlation Coefficients. Results showed that the reliability of Total Scores was excellent for the Full-BESTest for all conditions (all ICCs>0.82), whereas the Mini-BESTest ranged from fair to excellent (ICC=0.56-0.86). Percentage of Domain Scores with good-excellent reliability (ICCs>0.60) was slightly higher for the Full-BESTest (66%) compared to the Mini-BESTest (59%). Smallest Detectable Change scores were good to excellent for the Full-BESTest (2%-6%) and for the Mini-BESTest (5%-10%) relative to total test scores. Both the Full-BESTest and Mini-BESTest can discriminate postural control abilities within and between days in school-aged children. The Full-BESTest has slightly better reproducibility and a broader range of items, which could be the most useful version for treatment planning. We propose minor modifications to improve reproducibility for children, and indicate the modified version by the title Kids-BESTest. Future psychometric research is recommended for specific paediatric clinical populations.