Associations between hospital characteristics, volume, and reasons for revision: a cohort study of 48,029 unicompartmental knee arthroplasties with 3,397 revisions from the Dutch Arthroplasty Register

BACKGROUND AND PURPOSE: It remains debated whether high annual hospital volumes for unicompartmental knee arthroplasty (UKA) are associated with a low risk of revision, and what explanations may underlie this relationship. We aimed to analyze the association between specific hospital characteristics defined as volume, type, and referral for revision, and frequency and reasons for UKA revision. METHODS: Data from primary UKAs (2007-2022) and their revisions were extracted from the Dutch Arthroplasty Register. Hospitals were categorized by type; academic, top clinical teaching, private, or other general hospitals. Hospitals were grouped by annual UKA volume: ≤24, 25-39, 40-79, and ≥80 procedures. Multiple linear regression assessed the relationship between the number of revisions with hospital volume and type, adjusted for confounders. Chi-squared tests were used to test for differences in revision reasons based on volume and referrals. RESULTS: 48,029 primary UKAs and 3,397 revisions were included. High-volume and top clinical teaching hospitals had a significantly lower risk of revision following primary UKA (P < 0.001). Cementless implants had a lower risk compared with cemented implants. Revision reasons varied by hospital volume and whether revision occurred after referral (P < 0.001). Loosening, progression of osteoarthritis, malalignment, and pain were less common in the highest volume hospitals. If revision occurred after referral, malalignment was more frequently registered as the reason for revision. CONCLUSION: High-volume and top clinical teaching hospitals were associated with lower risk of revision following primary UKA. Differences in revision reasons, with fewer cases of loosening, progression of osteoarthritis, malalignment, and pain, may explain the lower risk of revision at higher volume hospitals.