EXTERNAL VALIDATION OF THE NORWEGIAN ANTERIOR CRUCIATE LIGAMENT RECONSTRUCTION REVISION PREDICTION MODEL USING PATIENTS FROM THE STABILITY 1 TRIAL

Machine learning has emerged as a potential tool for improving outcome prediction accuracy in anterior cruciate ligament (ACL) reconstruction (ACLR). A machine learning-based ACL revision prediction model has been developed using data from the Norwegian Knee Ligament Register (NKLR) but lacks external validation outside of Scandinavia. This study aimed to assess the external validity of the previously published ACL revision prediction model (https://swastvedt.shinyapps.io/calculator_rev/) using data from the Stability 1 randomized clinical trial. The hypothesis was that the model's performance would be similar, indicating validity of the algorithm. This was a level 3 cohort study. The NKLR Cox Lasso model was selected for external validation owing to its superior performance in the original study. Stability 1 patients with all five predictors required by the Cox Lasso model were included. The Stability 1 RCT was a prospective study which randomized patients to receive either a hamstring tendon autograft (HT) alone, or HT plus a lateral extra-articular tenodesis (LET). Since all patients in the Stability 1 trial received HT plus/minus LET, three configurations were tested: 1: all patients coded as HT, 2: HT + LET group coded as bone-patellar tendon-bone autograft (BPTB), 3: HT + LET group coded as unknown/other graft choice. Model performance was assessed via concordance and calibration. In total, 591 patients were included, and 39 patients (6.6%) underwent revision surgery. Model performance was best when patients randomized to HT + LET were coded as BPTB. Validation concordance was similar to the original NKLR prediction model for one- and two-year revision prediction (Stability: 0.71; NKLR: 0.68-0.69). Concordance confidence interval (CI) ranged from 0.63-0.79. The model was well calibrated for one-year prediction while the two-year prediction demonstrated evidence of miscalibration. When patients in Stability 1 who received HT + LET were coded as BPTB in the NKLR prediction model, the concordance was similar to the index study. However, due to a wide CI, the true performance of the prediction model with this Canadian and European cohort is unclear and a larger dataset is required to definitively determine the external validity. Further, model calibration was better for predicting revision surgery within one-year, in keeping with the original study and with prediction modeling in general, as predicting outcome over longer time-periods is typically more challenging due to the increased outcome variability observed over time.