Poster 250. Predictors of Baseline PROMIS Categorization in Patients with Shoulder Injuries

Abstract

Objective: This study aimed to determine the likelihood of patients with shoulder injuries falling into mild, moderate, or severe T-score categories for PROMIS-Pain Interference (PI) and PROMIS-Physical Function (PF). It was hypothesized that the odds of PROMIS categorization would vary based upon shoulder diagnosis and other patient characteristics. Methods: Data were retrospectively analyzed from a prospective clinical registry of patients aged 18 years or older presenting with primary shoulder pain between March 2022 and March 2025. At the initial visit, patients completed PROMIS-PI and PROMIS-PF via computer adaptive testing. Multinomial logistic regression models estimated the odds of falling into mild, moderate, or severe categories for PI and PF versus the within normal limits category, as noted in Table 1. Independent variables included age (continuous), sex (female=1), BMI (>=30=1), operative status (operative=1), and primary diagnosis (rotator cuff pathology=0, arthritis=1, adhesive capsulitis=2, other=reference). Diagnoses categorized as 'other' included AC and SC joint pathology, labral pathology, shoulder instability, muscle and bone injuries, and nerve-related pathology. Less common diagnoses were collapsed into an 'other' category to ensure adequate sample size and preserve inclusion of the clinical population. Statistical significance was set at p<0.05. Results: There were 977 shoulder patients included in the analysis. This sample ranges from ages 18 to 90 with a mean age of 55±16.2, and is 54.5 percent male. Table 2 provides the regression results with associated odds ratios and significance. The reference category of other diagnoses comprised 400 cases. Among shoulder diagnoses, rotator cuff pathology (n=372) was associated with a lower likelihood of severe PI and severe PF impairment compared to other diagnoses. Adhesive capsulitis (n=92) increased the likelihood of mild and moderate PI but decreased the likelihood of severe PF impairment. Shoulder arthritis (n=113) was linked to reduced odds of severe PI and severe PF impairment. Female patients were more likely to report severe PI and both moderate and severe PF impairment compared to male patients. Patients undergoing surgery had a higher likelihood of experiencing mild, moderate, and severe PI, as well as mild, moderate, and severe PF impairment, compared to non-operative patients. Patients with a BMI >=30 were more likely to experience moderate and severe PI and are more likely to fall into mild, moderate, and severe PF impairment. Increasing age was associated with a higher likelihood of severe PI and mild, moderate, and severe PF impairment. Conclusions: This study demonstrates that primary shoulder diagnosis, sex, operative status, BMI, and age are significant predictors of baseline PROMIS-PI and PROMIS-PF category placement in patients with shoulder injuries. Notably, rotator cuff pathology and arthritis are associated with less severe impairments when compared to other diagnoses, while adhesive capsulitis is linked to increased pain interference but less severe functional limitation, possibly due to pain-driven disability. Female patients and those with higher BMI or operative status consistently report greater impairments, suggesting these groups may require targeted interventions. Older age also contributes to worse outcomes, though its effect is less pronounced than other factors. These findings enhance the understanding of baseline patient-reported outcomes, enabling clinicians to tailor preoperative expectations, optimize treatment plans, and prioritize patient-centered care strategies. Future research should explore these predictors in more specific diagnostic subgroups and validate their clinical application in diverse upper extremity populations.