Systematic review and meta-analysis of machine learning-based prediction models for readmission risk after total hip and knee arthroplasty - Takeaways - MDSpire

Systematic review and meta-analysis of machine learning-based prediction models for readmission risk after total hip and knee arthroplasty

  • By

  • Jiabin Feng

  • Min Ma

  • Changliang Ou

  • Kaiwei Zhang

  • July 15, 2026

Share

  • 1

    This study systematically reviews machine learning models for predicting readmission risk after total hip and knee arthroplasty.

  • 2

    Fifteen studies with 57 distinct models were included, yielding a pooled C-statistic of 0.76, indicating moderate predictive performance.

  • 3

    Extreme heterogeneity (I2 = 99.9%) was observed, limiting the clinical utility of the pooled C-statistic and highlighting unpredictability.

  • 4

    Single-center models performed better (C-statistic 0.86) than multicenter models (0.65), and THA-specific models outperformed TKA-specific models.

  • 5

    High risk of bias was identified in most studies due to analytical shortcomings and lack of model recalibration, constraining evidence quality.

Original Source(s)

Systematic review and meta-analysis of machine learning-based prediction models for readmission risk after total hip and knee arthroplasty

  • by Jiabin Feng, Min Ma, Changliang Ou, Kaiwei Zhang

  • July 15, 2026

Related Content

Medical Oddities: Natural Doesn't Always Mean Safe

From the gym to the kitchen to romance, these studies challenge everyday assumptions.

Non-coding RNA in osteoarthritis: mechanistic insights and future perspectives

Unicompartmental knee arthroplasty and total knee arthroplasty in patients with 15 or more degrees of varus deformity: age- and gender-matched comparative study of radiographic and clinical outcomes