Novice vs expert inter-rater reliability of the balance error scoring system in children between the ages of 5 and 14.

Background: The Balance Error Scoring System (BESS) is a tool to measure balance, however, no studies have shown its reliability between novice and expert raters.

Research Question: What is the inter-rater reliability of BESS measurements when performed by novice raters compared to experts, and does completion of a focused, online training module increase the inter-rater reliability among novice raters?

Methods: In this reliability study, 5 novice volunteers were asked to independently rate BESS tests from 50 random prerecorded BESS videos of normal healthy subjects aged 5-14. Novice raters regraded the same 50 videos after receiving a formal training. The novices' scores before and after the formal training were compared to one another and then the scores were compared to 4 expert scores. Intraclass correlation (ICC) with 95 % confidence intervals or percent agreements were calculated and compared across groups.

Results: For the total BESS score, novice raters showed good reliability (ICC 0.845) which did not change with a formal training (ICC 0.846). Expert raters showed excellent reliability (ICC 0.929). Poor to moderate reliability was noted in the foam stance-single leg in the untrained novice and trained novice group (ICCs 0.452 and 0.64 L respectively).

Significance: BESS testing by novice raters with only written instruction and no formal training yields good inter-rater reliability. In contrast, BESS testing by expert raters yields excellent reliability. A focused training for novice raters conferred a small improvement in the reliability of the scoring of the single leg stance on foam condition but not a significant difference to the overall BESS score. While novices demonstrated promising reliability for overall BESS scores, optimizing clinical research using the BESS with expert raters show the highest reliability.