The Modified-Athletic Shoulder Test: Reliability and validity of a new on-field assessment tool

Objectives
To assess the reliability and the validity of a modified version of the Athletic Shoulder Test (M-AST), using a handheld dynamometer instead of a force plate.

Design
Test-retest reliability and concordance between tools.

Setting
University Motion Lab.

Participants
Twenty healthy volunteers who practiced an upper limb sport for at least five hours per week were recruited for the study.

Main outcome measures
The concordance between the Athletic Shoulder Test (AST) and the Modified-Athletic Shoulder Test (M-AST) was assessed with intra-class correlation coefficients (ICC), paired T-Test and Bland-Altman plots. The inter-session reliability was assessed with intra-class correlation coefficients (ICC), SEM, MDC and paired T-Test.

Results
A strong concordance was found between AST and M-AST values (ICC = 0.86–0.97; p > 0.05) in all the positions considered and for both sessions. Bland-Altman plots confirmed these results. However, the inter-session reliability was more variable for both AST and M-AST (ICC = 0.643–0.923; p < 0.05).

Conclusions
The M-AST seems to be a reliable, cheaper and easier to implement alternative to the AST to measure recovery status following matches or training sessions in upper limb athletes. As for the AST, familiarization trials will have to be performed before the assessment to have accurate measurements.