Objective The purpose of this study was to create SCAT5 reliable change scores in a large sample of professional hockey players.

Design Longitudinal cohort.

Setting National Hockey League (NHL) pre-season medical evaluations.

Participants All NHL players were given the NHL modified SCAT5 prior to both the 2018–19 and 2019–20 seasons. Valid 10-word List Learning (n=986), Delayed List Recall (n=986), modified Balance Error Scoring System (n=828), and Concentration (n=987) data were extracted for players with 2 consecutive valid baselines and no intervening suspected concussion evaluations.

Outcome Measures Paired t-tests were used to examine differences between time 1 and time 2 performance. Pearson correlations were used to assess test-retest reliability and reliable change was calculated using the Iverson method.

Main Results Correlations between repeat administrations had mostly small to moderate effects (r =.29 to .51). Due to significant list-learning form differences, .80, .90, .95, and .99 confidence interval reliable change metrics were developed for 3 list administration orders (1,2; 2,3; 3,1) for English Preference and non-English preference speakers. Reliable change metrics and their associated base rates are presented for mBESS and concentration subscores.

Conclusions This study offers 1-year test-retest reliability and reliable change metrics for a large group of English and non-English preference professional hockey players. Data represent an initial step in evaluating the psychometric properties of the SCAT5 10-word list and Concentration indices – as well as providing reliable change guidelines to aid instrument interpretation.

This abstract has been published in full manuscript format and has the following citation: BMJ Citation https://bjsm.bmj.com/content/early/2022/03/24/bjsports-2021-104851.long