Comparing Common Risk Assessment Tools to Predict Outcomes in Total Knee Arthroplasty
Travis Kotzur,Aaron Singh,Lindsey Peng,Ahmed Makhani,Ali Seifi,Chance Moore
DOI: https://doi.org/10.1016/j.arth.2024.01.052
IF: 4.435
2024-02-10
The Journal of Arthroplasty
Abstract:Introduction A number of tools exist to aid surgeons in risk assessment, including the Charlson Comorbidity Index (CCI), the Elixhauser Comorbidity Index (ECI), and various measures of frailty, such as the Hospital Frailty Risk Score (HFR). While all of these tools have been validated for general use, the best risk assessment tool is still debated. Risk assessment is particularly important in elective surgery, such as total joint arthroplasty. The aim of this study is to compare the predictive power of the CCI, ECI, and HFR in the setting of total knee arthroplasty (TKA). Methods All patients who underwent TKA were identified via ICD-10 code from the National Readmissions Database, years 2016 to 2019. Patient demographics, perioperative complications, and hospital-associated outcomes were recorded. Receiver Operating Characteristic (ROC) curves were created and Area Under the Curves (AUCs) evaluated to gauge the predictive capabilities of each risk assessment tool (CCI, ECI, and HFR) across a range of outcomes. Results A total of 1,930,803 patients undergoing TKA were included in our analysis. For mortality, ECI was most predictive (0.95 AUC), while HFR and CCI were 0.75 and 0.74 AUC, respectively. For periprosthetic fractures, ECI was 0.78 AUC, HFR was 0.68 AUC, and CCI was 0.66 AUC. For joint infections, the ECI was 0.78 AUC, the HFR was 0.63 AUC, and the CCI was 0.62 AUC. For 30-day readmission, ECI was 0.79 AUC, while HFR and CCI were 0.6 AUC. For 30-day reoperation, ECI was 0.69 AUC, while HFR was 0.58 AUC and CCI was 0.56 AUC. Conclusion Our analysis shows that ECI is superior to CCI and HFR for predicting 30-day postoperative outcomes following TKA. Surgeons should consider assessing patients using ECI prior to TKA.
orthopedics