Background/Purpose: Improvement in the medical management of rheumatoid arthritis (RA) over the past two decades may have reduced the need for arthroplasty surgery but the literature to date has reported inconsistent findings. The objective of our study was to compare the annual frequency of hip, knee and other arthroplasty surgery in a prevalent cohort of RA cases and matched controls over 13 years.

Methods: A retrospective cohort study was performed utilizing administrative healthcare data from approximately 1 million people with access to universal healthcare between 1997 and 2010. RA cases were identified using a previously validated RA case definition in the same dataset (1). Each case was matched by age and sex to 4 randomly selected controls. The annual frequency of arthroplasties in cases and controls was compared. In addition the frequency of coronary artery interventions (bypass grafting, angioplasty and stenting) was used as an additional control. Data included physician billings, hospital discharges and patient registry information using ICD-9 and ICD-10. Statistical analysis used least squares regression t-tests and 2-proportion z-tests.

Results: The number (prevalence) of RA cases per year increased from 3,913 (0.42%) to 4,911 (0.52%) over the study. The mean (SD) age changed from 56.7 (15.9) to 60.1 (14.9) years and the proportion of females from 70.8% to 73.9%. In both the first and last years of the study the frequency of all arthroplasty procedures was higher in cases than controls (p< 0.001) (Table). Over time there was a gradual and significant reduction in arthroplasty surgery in RA cases by 51.9% (p<0.001). This was in contrast to controls in whom the frequency of procedures increased by 31.9% (p=0.002) with the exception of hip arthroplasty. For the latter procedure, the frequency decreased by 63% in RA cases (p<0.001) and 35% in controls (p=0.617). In contrast to arthroplasty procedures the frequency of cardiac procedures, which was higher in RA cases in both the first (p=0.013) and final (p=0.003) years of observation, increased in both cases and controls over time although this did not reach statistical significance in either.

Conclusion: There was a striking reduction in arthroplasty surgery in RA cases over 13 years of observation. Lack of similar changes in controls and sustained rates of cardiac procedures over the same time suggests that this was not due to limited surgical access for RA patients. Earlier diagnosis and improved medical treatment of RA are potentially responsible.

Table: Demographics and surgical procedures in RA patients and controls

Reference

1. JG Hanly, K Thompson, C Skedgel. The use of administrative healthcare databases to identify patients with Rheumatoid Arthritis. Access Rheumatology: Research and Reviews 2015:7;69-75.