Establishment and Verification of Risk Prediction Model for Adverse Outcomes After Hip Arthroplasty in Elderly Patients.

Background: Adverse outcomes after hip arthroplasty in elderly patients are frequently observed; however, most existing studies concentrate on single complications. Comprehensive predictive models for a wider range of adverse outcomes remain insufficient. This study explores this issue and proposes new approaches for clinical practice.

Purpose: This study aimed to construct and verify risk prediction model for adverse outcomes after hip arthroplasty in elderly patients.

Patients and methods: The TRIPOD checklist was followed to guide the reporting of this study. Data from 620 subjects who underwent hip arthroplasty at a tertiary A-level hospital from January 1, 2021 to December 31, 2023 were used for the modelling group. Additionally, 264 post-hip arthroplasty patients admitted to the orthopaedic department of another tertiary A-level hospital from January 1, 2024 to December 31, 2024 were selected as the validation group. Risk prediction models were constructed by logistic regression, plotted in column line graphs and evaluated for their predictive effectiveness.

Results: The factors included in the prediction model were age, malignancy history, surgical procedure, albumin, prothrombin time, ASA grade, operation duration, and changeover surgery status. Hosmer-Lemeshow test, χ2=5.418, p=0.712, the area under the receiver operating characteristic curve (AUC) was 0.902. The Youden index is 0.668, with a sensitivity of 0.84 and a specificity of 0.828. The correct practical application rate was 83.33%.

Conclusion: The risk prediction model constructed in this study demonstrates favourable predictive performance and can serve as a reference for healthcare professionals in predicting the risk of adverse outcomes after hip arthroplasty in elderly patients.