Variation in Revision Rates Following Primary Knee Arthroplasty Across Danish Hospitals

Overview

This study analyzed baseline data from 1452 patients undergoing primary knee arthroplasty across three Danish university hospitals with differing revision rates. Significant regional variation in 2-year cumulative revision rates (CRRs) was observed, with the Capital Region showing the highest rates. The study aimed to investigate whether differences in patient demographics, knee osteoarthritis severity, implant selection, or patient-reported outcomes could explain these variations.

Background

Knee arthroplasty (KA) quality is often assessed by cumulative revision rates (CRRs), which vary significantly between countries, regions, and hospitals. Danish registry data revealed persistent regional differences in CRRs, notably higher in the Capital Region compared to Central and North Denmark Regions. These variations may reflect differences in patient selection, surgical quality, implant choice, or revision indications. The SPARK study was initiated to prospectively compare patient characteristics, radiographic osteoarthritis severity, implant selection, and patient-reported outcome measures (PROMs) across hospitals with differing revision rates.

Data Highlights

Key Findings

• Significant variation in 2-year cumulative revision rates exists between Danish regions, with the Capital Region having the highest CRR (5.0%) and North Denmark the lowest (1.0%).
• Differences in revision rates are consistent over time and not explained by chance alone.
• Patient demographics, preoperative knee symptoms, and radiographic severity of osteoarthritis are known factors influencing revision risk but had not been compared across hospitals prior to this study.
• The SPARK study enrolled patients from the largest university hospitals in each region to prospectively assess patient characteristics, implant selection, and PROMs before surgery.
• Participation rates and data collection methods were standardized, including exclusion of patients unable to complete preoperative PROMs.
• Baseline PROMs included Oxford Knee Score, UCLA Activity Scale, Copenhagen Knee ROM Scale, EQ-5D-5L, and a global knee function VAS.

Clinical Implications

Clinicians should recognize that revision rates after primary knee arthroplasty vary significantly by region and hospital, potentially reflecting differences in patient selection and surgical practice rather than implant choice alone. Comprehensive preoperative assessment including PROMs and radiographic evaluation may help identify factors contributing to revision risk. Understanding these variations can guide quality improvement initiatives and patient counseling regarding expected outcomes.

Conclusion

The SPARK study highlights substantial regional variation in revision rates following primary knee arthroplasty in Denmark, prompting further investigation into patient and surgical factors influencing these differences. Baseline comparisons of patient characteristics and PROMs provide a foundation for understanding revision risk variation.

References

1. Danish Knee Arthroplasty Register Annual Report 2016 -- Revision Rates Data
2. SPARK Study Protocol -- Variation in Patient Satisfaction and Revision Rates in Knee Arthroplasty