Artikel
Hospital volume-outcome relationship in total knee arthroplasty: a systematic review with dose-response meta-analysis
Suche in Medline nach
Autoren
Veröffentlicht: | 23. Februar 2021 |
---|
Gliederung
Text
Background/research question: The minimum volume threshold for total knee arthroplasty (TKA) is 50 procedures/hospital/year in Germany. It is unclear whether a hospital volume-outcome relationship (higher volumes associated with better outcomes) exists for TKA. Systematic reviews (SR) of the volume-outcome relationship in surgeries often dichotomised volumes into high vs. low resulting in information loss. This SR investigated the hospital volume-outcome relationship for TKA [1], [2] using a dose-response meta-analysis to account for multiple hospital volume categories.
Methods: We included cohort studies with the following PICO-criteria: P: TKA patients; I&C: at least two hospital volumes; O: patient-relevant outcomes. We identified studies from MEDLINE, Embase, CENTRAL, CINAHL and other sources up to 14.02.2020. The risk of bias in individual studies was assessed using the ROBINS-I tool. If adequate data were available, we pooled the best-adjusted effect estimates using a linear, random-effects, dose-response meta-analysis. Heterogeneity was quantified using the I2-statistic. The quality of evidence was assessed using the GRADE approach.
Results: A total of k=68 studies with data from 1985-2018 were eligible. Most study outcomes had a moderate to serious risk of bias. Based on narrative synthesis, higher hospital volumes were associated with less early revisions within 1 year (primary outcome; k=7, n=301,378). Meta-analysis of secondary outcomes showed that higher hospital volumes were associated with lower rates of mortality (odds ratio, OR=0.91, 95% confidence interval [0.87-0.95] for 50 additional TKAs, I2=50%, k=9, n=2,638,996) and readmissions (0.98 [0.97-0.99], I2=44%, k=3, n=830,381). Hospital volumes were not associated with the rates of deep infections (1.03 [0.97-1.09], I2=0%, k=3, n=97,019) nor later revisions at 1-5 years (0.96 [0.86-1.07], I2=98%, k=5, n=163,520). The certainty of evidence was low for early revisions, moderate for mortality and readmissions and very low for deep infections and later revisions.
Conclusion: The dose-response approach with multiple hospital volume categories allowed us to assess the type of a relationship between hospital volumes and outcomes without information loss. We show that a linear association between hospital volumes and some patient-relevant outcomes exists (less mortality and readmissions in high-volume hospitals). The minimum volume threshold for TKA seems justifiable although a cut-off value cannot be established for a linear relationship.
BMBF grant 01KG1805
Competing interests: The authors declare that they have no conflicts of interest.