Dynamic track-based modification (Dynamic Track Instability Management Score) of current shoulder instability scores...

Shoulder instability scoring systems (Instability Severity Index [ISI] Score and the Glenoid Track Instability Management Score [GTIMS]) use radiographic or computed tomographic (CT) scan measurements for scoring bone loss. Radiologic methods are known to have variable reliability, and this variability can affect decision making. The purpose of this study was to assess reliability, reproducibility, and diagnostic validity of a new dynamic track-based score, the Dynamic Track Instability Management Score (DTIMS), and to compare these with current gold standard scores for predicting the necessity for a bony reconstruction procedure. The hypothesis was that DTIMS will have greater agreement and inter-observer reliability (IOR) for bone defect assessment, and in decision making for soft tissue vs. bony surgery.

Between February 2023 and October 2024, a total of 108 patients with traumatic anterior shoulder instability were evaluated clinically and using radiographs and CT scans, by 2 independent observers. Radiographs were used to identify glenoid and humeral head defects, and 3D CT scans were used to classify defects as on- and off-track. The data were used to calculate 2 shoulder scores (GTIMS and ISI score). During arthroscopy, a Dynamic Arthroscopic Standardized Test (DAST) was used to classify defects into on- and off-track, and DTIMS was calculated. Cohen kappa was used to evaluate IOR for bone loss assessment for the ISI score (Hill-Sachs on external rotation and loss of inferior glenoid contour on anteroposterior radiograph), and for DTIMS and GTIMS (on- and off-track), and for determining the need for bone or soft tissue surgery for all 3 scores. IOR for total scores was assessed using the intraclass correlation coefficient (ICC). Diagnostic validity of DTIMS was assessed against the ISI score and GTIMS, and sensitivity, specificity, positive and negative predictive values, and diagnostic accuracy were reported.

IOR for total score was excellent (ICC >0.9) for all 3 scores. IOR for identification of significant bone defects was near perfect (κ = 0.94, P < .001) in DTIMS, and was substantial (κ = 0.69, P < .001) in GTIMS. However, IOR was moderate (κ = 0.47 glenoid, 0.58 Hill-Sachs, P < .001) with the radiographic method used in ISI score. Similarly, IOR for surgical decision making was near perfect for DTIMS (κ = 0.91, P < .001) and GTIMS (κ = 0.85, P < .001) and was lower (κ = 0.74, P < .001) for ISI score. DTIMS showed the highest sensitivity (89%, 91%) for predicting a bony procedure, and ISI score showed the highest specificity (78%, 81%) when compared with GTIMS.

Substituting the image-based assessment of bone defects with dynamic track method (ie, DAST) resulted in a highly reliable scoring system (DTIMS) that can be used as an alternative to current gold standard scores (ISI score and GTIMS) for decision-making in anterior shoulder instability.