Low confidence in the cumulative evidence for the existence of a volume-outcome relationship after revision total knee replacement: A systematic review and meta-analysis

Abstract

Purpose: This systematic review and meta-analysis aimed to establish the relationship between the number of procedures a hospital or surgeon performs with outcomes following revision knee replacement (RevKR).

Methods: MEDLINE and Embase were searched using Ovid silver platter up to December 2024 for randomised controlled trials and cohort studies that reported RevKR volumes, in at least two categories, performed by hospitals and surgeons and their relationship to patient and provider level outcomes. The primary outcome was re-revision rate. Secondary outcomes included mortality, post-operative complications, patient-reported outcomes measures (PROMs), emergency readmissions and hospital length of stay. The effect estimates were pooled and plotted using a random-effects, non-linear dose-response meta-analysis (DRMA). Where limitations in the data prohibited DRMA, a narrative approach was utilised. ROBINS-I and the GRADE approach were used to assess the risk of bias and the confidence in the cumulative evidence, respectively.

Results: A total of 10 cohort studies with data from 1993 to 2021 were included. The confidence in the cumulative evidence exploring the relationship between surgeon/hospital volume and all outcomes after RevKR was very low. An inconsistent relationship was seen between hospital and surgeon volume and re-revision at any point. There was a non-linear dose-response relationship between higher hospital volume and lower odds of adverse post-operative events (p < 0.05, n = 3 studies, n = 35,524 patients). There was no association between increased surgeon volumes and improvements in PROMs (n = 2 studies, n = 2289).

Conclusion: There is a lack of high-quality studies establishing the relationship between the number of procedures a hospital or surgeon performs and outcomes following RevKR. Studies are limited to observational designs and are difficult to effectively power due to the rarity of outcomes. Pooling data from multiple studies provides valuable insights but highlights significant heterogeneity and limitations in the existing literature.

Level of evidence: Level III, systematic review-lowest level of evidence analysed-was from retrospective cohort study of prospectively collected data.

Keywords: complications; hospital volume; revision total knee replacement; re‐revision; surgeon volume.

Figure 1

Prisma flow diagram for new systematic reviews, which included searches of databases and registers only.

Figure 2

Dose–Response relationship for hospital volume and odds ratio (OR) of mortality up to 1 year after RevKR. The black solid line in the middle indicates the OR, and the upper and lower black dotted lines indicate a 95% CI. The blue dotted line at the bottom indicates the OR of the linear model. The risk in the quadratic model is slightly different from the risk in the linear model. The quadratic model has a non‐significant p value of 0.34 indicating there is no strong evidence of a dose–response relationship. The I ² statistic is 62.8% indicating moderate residual heterogeneity. CI, confidence interval; RevKR, revision total knee replacement.

Figure 3

Dose–response relationship for hospital volume and rates of adverse post‐operative events up to 1 year after RevKR. The black solid line in the middle indicates the odds ratio (OR) of adverse post‐operative complications, and the upper and lower black dotted lines indicate a 95% CI. The blue dotted line at the bottom indicates the odds ratio from the linear model. The quadratic model has a significant p < 0.05, indicating when modelled as a quadratic term, hospital volume is related to odds of post‐operative complications. Higher hospital volumes are associated with lower odds of complications, For annual hospital volumes ≥60, the OR and associated 95% CI are below 1.0. The I ² statistic is 33.6% indicating low residual heterogeneity. CI, confidence interval; RevKR, revision total knee replacement.