Takeaway
- Glucocorticoid usage among patients with rheumatoid arthritis (RA) is strongly associated with postarthroplasty infection risk, even at modest doses.
- Biologics usage is associated with generally lower risks, and risks are similar across biologics.
Why this matters
- Hip and knee arthroplasty are common among patients with RA.
- Authors recommend limiting perioperative use of glucocorticoids.
Study design
- Retrospective cohort of patients with RA undergoing knee or hip arthroplasty (primary or revision), using biologics (n=9911) or glucocorticoids (n=about 6500) before surgery, in the Medicare and Truven MarketScan databases.
- Primary outcomes were cumulative incidence of hospitalized infections within 30 days and prosthetic joint infection (PJI) within 1 year of total knee or hip arthroplasty, according to propensity-adjusted analyses using inverse-probability weights.
- Funding: NIH; Bristol-Myers Squibb; others.
Key results
- Biologics:
- Compared with 8.16% predicted risk for hospitalized infection with abatacept, predicted cumulative risk from propensity-weighted models ranged from 6.87% to 8.90% with other biologics (adalimumab, etanercept, infliximab, rituximab, and tocilizumab).
- Compared with 2.14% predicted risk for PJI with abatacept, predicted cumulative incidence ranged from 0.35% to 3.67% with other biologics.
- Glucocorticoids:
- Compared with nonusers, there was a dose-dependent predicted increased cumulative risk for hospitalized infection and PJI, with highest risk at doses >10 mg/day (13.25% and 3.83%, respectively).
Limitations
- Potential residual confounding.
References
References