A LASSO regression-driven nomogram to predict functional outcomes after surgery for chronic lateral ankle instability

[Development of a nomogram prediction model for re-tear after arthroscopic repair of massive rotator cuff tears].

Paper 07. The Hip Arthroscopy Index Predictive Score: A Preoperative Tool for Predicting Postoperative Outcomes

Objectives: Preoperative clinical factors influencing outcomes after hip arthroscopy are well established but have not yet been consolidated into a single predictive scoring system to estimate the likelihood of meaningful postoperative improvement. The purpose of this study was to develop an index score using readily available preoperative variables to predict the likelihood of achieving clinically significant improvement after hip arthroscopy. Methods: This was approved by our IRB. A cohort of patients undergoing primary hip arthroscopy >= 18 years old was evaluated using readily available preoperative clinical variables including age group (categorized as youngest, ( < 28 ) middle (28 to 37), and oldest ( > 37) tertiles), Tönnis grade (grouped as 0, 1, >=2), sex, BMI category (underweight, normal, overweight), symptom duration (<6 months, 6 months to 2 years, >2 years), enrollment iHOT score groups, and preoperative pain levels. Multivariable logistic regression models assessed these factors as independent predictors of achieving the minimal clinically important difference (MCID) and substantial clinical benefit (SCB) at one year postoperatively. Each variable's contribution was quantified by translating the logistic regression coefficients (log-odds) into a simplified point system. Odds ratios from the model were converted to relative weights, scaled against the smallest meaningful effect, and then rounded to whole numbers. The resulting scoring algorithm stratifies patients by their likelihood of achieving meaningful clinical improvement following hip arthroscopy. Results: 135 patients were included in this study (47.4% female) . In the univariate analysis, younger age, female sex, normal BMI, higher preoperative pain, and lower preoperative iHOT scores were associated with a higher likelihood of achieving the MCID. In the multivariable model for clinical improvement, age group (youngest: OR, 7.13 [95% CI, 1.87-36.08]; middle-aged: OR, 1.79 [95% CI, 0.60-5.58], compared with the oldest group), BMI category (normal BMI 18.5-24.99: OR, 3.57 [95% CI, 1.31-10.31], compared with overweight/obese), and enrollment iHOT score (score 0-39: OR, 5.90 [95% CI, 1.25-31.11]; score 40-59: OR, 1.75 [95% CI, 0.46-6.67], compared with score >=60) were independent significant predictors for clinical improvement (P<0.05). The scoring algorithm demonstrated that a higher total score was associated with an increased likelihood of achieving the MCID. (Table 1). Patients with a total score of 0-2 had a 40.7% MCID achievement rate, those with scores of 3-4 had a 78.6% rate, scores of 5-8 had a 94.0% rate, and a score of 9 had the highest likelihood at 100.0%. Conclusions: Our scoring algorithm, based on preoperative factors such as age, BMI, and enrollment iHOT score, estimates the likelihood of patients achieving the Minimal Clinically Important Difference (MCID)-the smallest change in a patient-reported outcome that reflects a meaningful improvement in symptoms and function. Because the model relies entirely on readily accessible clinical and demographic information rather than imaging or advanced diagnostics, it can be applied broadly and readily across patient populations. Higher total scores were associated with increased odds of reaching this threshold, enabling clinicians to identify, early in the care process, those patients most likely to experience clinical benefit from treatment. This preoperative scoring tool provides a practical approach to predict patient outcomes, facilitating shared decision-making and personalized treatment planning. By identifying patients more likely to benefit from intervention, healthcare providers can improve overall care quality.

Poster 211. Development and Validation of The BRACE (Bankart Risk Assessment & Clinical Estimator) Score – An International Multicenter Study

Objective: Arthroscopic Bankart repair (ABR) is the most widely performed surgical treatment for anterior shoulder instability, and the Instability Severity Index Score (ISIS), while widely used to guide surgical decision-making, has demonstrated inconsistent reliability. Recent evidence suggests that incorporating advanced imaging findings and refining patient-specific risk factors may enhance predictive accuracy and improve clinical utility. The purpose of this study was to develop and validate an ABR recurrence risk calculator in an independent clinical dataset and assess its performance in stratifying patients by recurrence risk. Methods: The BRACE (Bankart Risk Assessment & Clinical Estimator) score was developed, based on age, sex, glenoid bone loss, Hill-Sachs lesion status, number of preoperative dislocations, ligamentous laxity, and sport participation generating 864 distinct clinical scenarios. Odds ratios were extracted from a systematic review of 110 studies and incorporated into a logistic regression model. The model was then evaluated on an international multicenter cohort including 5 institutions from 3 countries. Discrimination (area under the curve (AUC)), Brier score, calibration (slope, Hosmer-Lemeshow), risk stratification (low; 0-14, moderate; 15%-29%, high; 30+%), decision‑curve analysis and operating characteristics from 10%-30% thresholds were reported. Results: The BRACE score was evaluated on 2535 ABR patients. The interaction between glenoid bone loss >=13.5% and an off-track Hill-Sachs lesion was independently associated with higher recurrence risk. The model demonstrated moderate discriminative ability (AUC = 0.70, 95% CI 0.62-0.78) with appropriate calibration across the range of predicted recurrence probabilities (calibration slope = 0.889, 95% CI 0.62-1.16; Hosmer-Lemeshow p = 0.36). The model achieved a Brier score of 0.067 (95% CI 0.052-0.082), indicating good overall accuracy of the predicted probabilities. At a 15% moderate risk threshold, the specificity was 93.4% (95% CI 91.1%-95.4%), and at a 30% high threshold specificity was 98.0% (95% CI 96.7%-99.1%). Model performance remained consistent across subgroups and predicted risk estimates aligned closely with observed recurrence rates. Conclusions: The ABR recurrence risk calculator demonstrated moderate discrimination, good overall accuracy, and appropriate calibration when applied to a large, multicenter clinical cohort. The tool effectively stratified patients into low-, moderate-, and high-risk groups, with decision-curve analysis confirming clinical utility across relevant thresholds.

Poster 197. Predicting Cross-Over to Surgical Management in Patients with Nonoperatively Treated Ulnar Collateral Ligament Injuries: A Machine Learning Analysis

Objective: To develop and validate prognostic machine learning (ML) algorithms to predict progression of ulnar collateral ligament (UCL) repair or reconstruction after a failed trial of nonoperative treatment in patients who sustained UCL injuries. Methods: A total of 253 consecutive patients who presented with a UCL injury between 2017-2024 at a tertiary academic hospital were identified. The primary outcome was surgical intervention consisting of either UCL repair or reconstruction, whereas nonoperative treatment success was defined as successful return to pre-injury activities or sport. Nineteen routinely collected features were explored for predictive value. Six unique ML models were developed on the training set (n=203), internally validated on the hold-out set (n=50), and compared to a standard logistic regression model. Algorithm performance was evaluated through discrimination, calibration, and Brier score. Local interpretable model-agnostic explanations (LIME; analyses exploring dynamic variable interactions that explain individual-level risk) and partial dependency plots (describing shape of association between predictors and outcome) were constructed to provide model decision-making transparency. Results: The incidence of failure of nonoperative treatment among patients with UCL injuries was 32.8%. A combination of nine features optimized prediction accuracy: age, tear quality (sprain, partial or full), symptom onset (insidious vs. traumatic), level of sport competition, specific sport, presence of concomitant injuries, timing of injury relative to play (in-season vs. out-of-season), history of prior elbow injuries, and throwing athlete status. The best performing model used a Random Forest architecture (Figure 1; Area Under the Receiver Operator Cureve, AUROC: 0.79; calibration intercept: -0.30; calibration slope: 0.67; Brier score: 0.018) and was superior to logistic regression (AUROC: 0.33; calibration intercept: 2.29; calibration slope: -0.43; Brier score: 0.53; Table 1). Investigation of model transparency demonstrated that age <16 years, full-thickness tears, traumatic symptom onset, non-recreational competition levels, participation in specific sports (contact sports, tennis, or baseball pitchers), in-season timing of injury, presence of concomitant injuries, history of prior elbow injury, and being a throwing athlete increased the risk of requiring surgery (Figures 2-3). An interactive prognostic application was developed: https://orthoapps.shinyapps.io/UCLINJURY/ Conclusions: Among patients with UCL injuries, several prognostic factors were identified that are associated with failure of nonoperative treatment. A prognostic ML algorithm was developed and validated based on these features to provide real-time personalized assessment of surgical risk. Use of this model and knowledge of these risk factors may allow for more informed assessments of surgical candidacy in patients with UCL injuries and identify patients with a high likelihood of necessitating surgical intervention.

A nomogram for predicting adjacent segment disease after anterior cervical surgery

Adjacent segment disease (ASD) is a recognized complication following anterior cervical surgery and may adversely affect long-term clinical outcomes. This study aimed to develop and internally validate a prognostic nomogram incorporating clinical and radiological variables to predict the risk of ASD after anterior cervical surgery. In this single-center retrospective cohort study, 234 patients who underwent anterior cervical surgery were included. The primary endpoint was the development of ASD during follow-up. Eleven candidate predictors were evaluated. Least absolute shrinkage and selection operator (LASSO) regression was applied for variable selection, followed by multivariable logistic regression to identify independent predictors. A nomogram was constructed, and its performance was assessed using the area under the receiver operating characteristic curve (AUROC), calibration analysis, and decision curve analysis. Internal validation was conducted using bootstrap resampling. ASD developed in 26 patients (11.1%). LASSO regression identified age, treatment modality (cage), mJOA score, disc height, and Cobb angle as the most predictive variables. In multivariable analysis, cage use was independently associated with an increased risk of ASD (odds ratio = 3.56; 95% CI: 1.46–8.68; P = .005). Increased disc height and Cobb angle were also significantly associated with ASD. The nomogram demonstrated good discriminative performance (AUROC = 0.78) with satisfactory calibration and no evidence of overfitting after internal validation. This nomogram provides an individualized risk prediction model for ASD following anterior cervical surgery. By integrating key clinical and radiological parameters, it may assist clinicians in identifying high-risk patients and optimizing postoperative management strategies.

[Early postoperative functional outcomes and influencing factors in patients with rotator cuff injury].