GIRD, TRROM, and humeral torsion-based classification of shoulder risk in throwing athletes are not in agreement and should not be used interchangeably.

Objectives: Clinicians commonly interpret the findings of shoulder rotational ranges of motion using various approaches: an apparent reduction in dominant arm internal rotation ("GIRD"), a difference in total rotational range of motion-i.e. differences in the sum of internal and external rotational range (TRROM), and a combination of rotational ROM and torsional difference ("GIRD-torsion"). We have noticed that these approaches are being considered to provide equivalent estimates of shoulder rotational range. This investigation sought to document the extent of agreement of these three different approaches when classifying athletes' "at-risk" status.

Design: Observational cohort study.

Methods: 162 professional male athletes participating in overhead sports (baseball, handball, and volleyball) had their GIRD, TRROM, and GIRD-torsion calculated, and classified as "at risk" using standard cut-points of 20°, 5°, and 10°, respectively.

Results: 25 (15.4%) athletes were classified as "at-risk" using GIRD, 55 (34%) with TRROM, and 30 (18.5%) using GIRD-torsion. Only 3/162 (1.9%) athletes were classified as "at-risk" by all 3 approaches, 4 athletes were concurrently classified as "at-risk" by GIRD and TRROM (Kappa=-0.142, poor agreement), 11 by GIRD and GIRD-torsion (Kappa=0.279, fair agreement), and 11 by TRROM and GIRD-torsion (Kappa=0.025, slight agreement).

Results: 25 (15.4%) athletes were classified as at risk using GIRD, 55 (34%) with TRROM, and 30 (18.5%) using GIRD-torsion. Only 3/162 (1.9%) athletes were classified as at risk by all 3 approaches, 4 athletes were concurrently classified as at risk by GIRD and TRROM (Kappa=-0.142, poor agreement), 11 by GIRD and GIRD-torsion (Kappa=0.279, fair agreement), and 11 by TRROM and GIRD-torsion (Kappa=0.025, slight agreement).

Conclusions: The three described approaches yield demonstrably different findings, and these approaches cannot be used interchangeably. Examples of clinical reasoning are provided to assist with the interpretation of these different measures.