No clear benefit of patient‐specific instruments over conventional instrumentation in unicompartmental knee arthroplasty: A secondary analysis of a multicentre randomised trial with 5‐year follow‐up

PURPOSE: The objective of the secondary analysis of this randomised controlled trial, conducted at two medical centres, was to compare the revision-free survivorship and patient-reported outcome measures (PROMs) of patient-specific instruments (PSI) versus conventional instrumentation (CI) used in unicompartmental knee arthroplasty (UKA) over a mid-term follow-up period of at least 5-years. The hypothesis was that both groups would show similar outcomes. METHODS: A total of 120 patients from two separate teaching hospitals were recruited and assigned at random to receive either PSI or CI for UKA. The primary outcomes included the revision rate and PROMs, including the Oxford Knee Score, the Western Ontario and McMaster Universities Arthritis Index, the EuroQol-5D (EQ-5D-3L) and Visual Analogue Scale (EQ-5D-VAS), the Numeric Rating Scale for pain (NRS-pain) and the forgotten joint score (FJS-12). RESULTS: The overall incidence of revisions was not statistically significantly different between PSI and CI (14.3% (8/56) vs. 5.1% (3/59), (p 0.094) The main reasons for revision included progression of lateral osteoarthritis (3.6% (2/56) vs. 0%), bearing dislocation (1.8% (1/56) vs. 1.7% (1/59)), persisting knee pain (1.8% (1/56) vs. 1.7% (1/59)) and instability (1.8% (1/56) vs. 1.7% (1/59)). All PROMs improved significantly postoperatively, but no statistically significant difference was observed between the PSI and CI group. CONCLUSIONS: The PSI method showed comparable results regarding patient-reported clinical and functional outcomes and prosthesis survival during the mid-term follow-up. A numerically higher, but not statistically significant, revision rate was observed in the PSI group. Based on the current findings, PSI do not appear to offer clear benefits over CI in UKA; however, this conclusion should be interpreted with caution given the likely limited statistical power of the study to detect differences in revision rates. LEVEL OF EVIDENCE: Level II.