287. The Attributable Mortality of Prosthetic Joint Infection After Primary Hip and Knee Arthroplasty Among Medicare Beneficiaries, 2005–2012

BACKGROUND: Total hip (THA) and total knee (TKA) arthroplasty are the most common elective surgical procedures performed in the USA. Most are performed in older adults and lead to improved quality of life; however, complications such as prosthetic joint infection (PJI) can occur. Little is known regarding the mortality attributable to PJI after THA or TKA. METHODS: Claims data from the 2004 to 2012 Medicare 5% sample Standard Analytic Files were used to find eligible beneficiaries, with ICD-9-CM procedure codes identifying primary THA (81.51) or primary TKA (81.54), and diagnosis code 996.66 indicating PJI during the year following the procedure. Inclusion criteria included traditional Medicare coverage during the year prior and two years following the procedure and original reason for entitlement due to age. Exclusion criteria included missing surgery date, additional primary procedures within 1 year, and PJI diagnosis prior or during index stay. The attributable mortality of PJI during the 2 years following primary hip and knee arthroplasty among Medicare beneficiaries was calculated by fitting Kaplan–Meier survival curves and performing a time-dependent analysis based on PJI timing using an Extended Cox Proportion Hazard model. RESULTS: A total of 248,340 hip and knee arthroplasties were performed on 5% Medicare sample beneficiaries between 2005 and 2012. The final cohort included 117,515 arthroplasties: 37,098 (32%) hip and 80,429 (68%) knee, of which 80,377 (68%) were performed in women and 61,807 (53%) in patients greater than 75 years of age. PJI was diagnosed in one percent of hip (n = 338) and knee (n = 726) arthroplasties, of which 112 (11%) died. The crude mortality rate was 3.2 (95% CI: 2.3,4.2) and 3.7 (95% CI: 2.9,4.8) times greater in patients with PJI than without PJI following THA and TKA, respectively. Controlling for comorbid conditions and the time-dependent nature of PJI, the risk of death with PJI was 2.5 (95% CI: 1.9,3.3) times higher following THA and 2.6 (95% CI: 2.0,4.3) times higher following TKA than for non-PJI. CONCLUSION: Medicare beneficiaries who develop PJI after THA or TKA have an increased risk of death during the first 2 years following the procedure, supporting the importance of better understanding risk factors and preventing PJI following these elective procedures. [Image: see text] [Image: see text] DISCLOSURES: All authors: No reported disclosures.