Holmstrom, Joshua Aaron

Background Long-term corticosteroid (LTC) use is a known risk factor for adverse outcomes after various surgical procedures. However, the impact of LTC use on surgical complications after TSA remains underexplored. This study aims to assess the influence of LTC use on 90-day and 5-year complications following TSA. Methods A retrospective cohort study utilizing the PearlDiver database was conducted. Patients undergoing primary TSA with a minimum 5-year follow-up were included. LTC users were propensity-matched 1:10 with controls based on demographic variables and comorbidities. Postoperative complications within 90 days of surgery and long-term complications within 5 years were compared between groups. Multivariable logistic regressions were used to compare complications controlling for demographic variables and comorbidities. Results In a matched cohort of 556 LTC users to 4942 controls, corticosteroid users exhibited a 2.91-fold higher risk of pneumonia (95% CI: 1.69–4.80, p < 0.001) and a 2.19-fold higher risk of venous thromboembolism (VTE) (95% CI: 1.30–3.51, p = 0.001) within 90 days following TSA. No significant differences in 5-year surgical outcomes were observed. Discussion LTC users are at a heightened risk for 90-day medical complications, including pneumonia and VTE following TSA. However, their 5-year risk of surgical complications was comparable to patients without LTC use.

Purpose To project the future utilization of open and arthroscopic rotator cuff repair (RCR) in the Medicare population through 2050, incorporating post‐COVID recovery data to provide insight into evolving surgical preferences, economic pressures, and implications for health care delivery. Methods We analyzed 2003‐2024 data from the Centers for Medicare and Medicaid Services Medicare Part‐B National Summary, which captures all Medicare beneficiaries undergoing open or arthroscopic RCR as identified by Current Procedural Terminology codes. Projections were based on 2003‐2019 and 2023‐2024 data, reflecting post‐COVID‐19 recovery and avoiding distortions from 2020 to 2022 pandemic‐related disruptions. Statistical models evaluated included log‐linear, Poisson, negative binomial, and autoregressive integrated moving average. Autoregressive integrated moving average was selected based on the lowest normalized root mean square error, reflecting superior predictive accuracy and its ability to capture time‐dependent trends, and was used to forecast RCR volumes through 2050. Results In 2024, there were 23,154 open and 188,700 arthroscopic RCR procedures, which served as the baseline for modeling. Open RCR is projected to decline on average by 5.91% annually, while arthroscopic RCR is projected to grow by 2.88% annually. By 2050, arthroscopic RCR volume is expected to reach 394,434 procedures (95% forecast interval: 363,424‐425,321), whereas open RCR is expected to decline to 4748 procedures (95% forecast interval: 1243‐43,747). Conclusions Autoregressive integrated moving average modeling projects that by 2050 arthroscopic RCR utilization will increase by 2‐fold and open RCR will decline to one‐fifth compared with 2024 baseline volumes. Level of Evidence Level III, retrospective cohort study.

OBJECTIVES: To evaluate and compare the ability of three popular open-source artificial intelligence platforms to diagnose common trauma-related fractures using radiologic imaging. METHODS: Design: Retrospective diagnostic performance comparison study. Setting: Publicly accessible online radiologic imaging databases. Patient Selection Criteria: Five common orthopedic trauma fractures were assessed: ankle, tibial plateau, intertrochanteric, femoral neck, and humerus. Radiographs and computed tomography (CT) images were collected. Images were randomly selected from confirmed diagnoses on Radiopaedia.org. Outcome Measures and Comparisons: ChatGPT 5, Grok 3, and Claude 4.5 Sonnet were queried with each image. Diagnostic accuracy, sensitivity, specificity, positive and negative predictive values, and performance by modality (X-ray vs. CT) were assessed. The reference standard was the expert-verified diagnosis provided by Radiopaedia.org, limited to cases labeled with a “diagnosis certain” tag. RESULTS: Each model was provided with 30 radiographs and 20 CT images whenever possible. ChatGPT 5, Grok 3, and Claude 4.5 Sonnet accurately diagnosed diseased images in 26.8%, 18.8%, and 22.4% of cases, respectively. By fracture type, ChatGPT 5 demonstrated the highest correct classification rates for ankle (10%), femoral neck (38%), humerus (40%), and tibial plateau (44%) fractures. Grok 3 demonstrated the highest correct classification rate for intertrochanteric fractures (6%). Overall sensitivities were 0.267, 0.187, and 0.223 for ChatGPT 5, Grok 3, and Claude 4.5 Sonnet, respectively. ChatGPT 5 and Grok 3 outperformed Claude 4.5 Sonnet (both P < 0.001). No modality-based performance differences were observed for any model. CONCLUSIONS: Among the publicly available large language models evaluated for radiologic interpretation of orthopedic trauma imaging, ChatGPT 5 demonstrated the highest overall diagnostic accuracy, followed by Claude 4.5 Sonnet and Grok 3. Despite relative variation between the models, overall diagnostic accuracy for fracture detection was low across all platforms (<27%). In their baseline forms, these publicly accessible large language models are not recommended for radiologic imaging interpretation. LEVEL OF EVIDENCE: Diagnostic Level III. See Instructions for Authors for a complete description of levels of evidence.