# Construction of prediction models for prolonged length of postoperative hospital stay in patients undergoing primary total hip arthroplasty via direct anterior approach

## TL;DR

This study identifies risk factors for longer hospital stays after hip surgery and creates a model to help doctors manage patient recovery better.

## Contribution

A new predictive model using LASSO regression to identify risk factors for prolonged hospital stays after hip arthroplasty.

## Key findings

- Higher BMI, longer operation time, and postoperative complications were linked to longer hospital stays.
- The model achieved an AUC of 0.766, showing good predictive accuracy.
- The model's calibration and decision curve analysis confirmed its clinical usefulness.

## Abstract

To investigate the risk factors associated with prolonged postoperative length of stay (PLOS) in patients undergoing primary total hip arthroplasty (THA) via direct anterior approach (DAA) and develop a perioperative dynamic prediction nomogram for optimizing the perioperative management of THA.

This single-center, retrospective cohort study analyzed the perioperative clinical data of patients who underwent primary THA through DAA by a single surgical team at our institution between September 2022 to September 2024. Patients were divided into two groups based on postoperative hospital stay duration: the normal group (PLOS < 6 days) and the prolonged group (PLOS > 6 days). LASSO regression was used to screen variables, multivariate logistic regression was applied to establish the model and then a nomogram was developed. The area under the curve (AUC) of receiver operating characteristic (ROC) curve, calibration curve, and decision curve analysis (DCA) were adopted to evaluate the performance and clinical applicability of the model.

This study included a total of 413 patients. Multivariate logistic regression analysis revealed that higher body mass index (BMI), longer operation time, American Society of Anesthesiologists classification (ASA) > II, postoperative extra opioid use, postoperative nausea and vomiting (PONV), postoperative blood transfusion, lower preoperative albumin (ALB) levels, and no prior contralateral THA history were independent risk factors for prolonged postoperative hospital stay in patients undergoing primary DAA-THA (P < 0.05). The AUC of the established predictive model was 0.766, indicating good predictive performance. The calibration curve demonstrated good consistency between actual delayed discharge rates and predicted probabilities. DCA showed that the model provided maximum net benefit when the threshold probability ranged from 2% to 85%.

BMI, operation time, ASA classification, postoperative extra opioid use, PONV, postoperative transfusion, preoperative ALB, and previous contralateral THA history can be used as predictive factors. The LASSO regression-based model for predicting prolonged hospital stay after primary DAA-THA demonstrates accurate predictive performance and strong clinical utility. It can assist clinicians in stratifying patient risk effectively, thereby supporting enhanced recovery protocols.