Authors: Camelia E. Marculescu, MD*, Juanmanuel Gomez, MD, MSCR, Kit Simpson, DrPH
Title: Total Joint Replacement Volume and In-Hospital Mortality During Treatment of Prosthetic Joint Infections (PJIs)
Addresses: Medical University of South Carolina, Division of Infectious Disease, 100 Doughty Street, BA/IOP South, suite 210, Po Box 250752, Charleston, SC 29425
Purpose: To determine whether hospital total joint replacement volume, demographic characteristics, severity of comorbid conditions and type of surgical procedure are associated with in-hospital mortality during treatment of PJIs.
Methods: We performed a retrospective analysis of archival billing data from the Healthcare Cost and Utilization Project (HCUP) database between 1999-2000. HCUP database contains a 50% stratified sample of all US hospitals, therefore it is very representative of the US population. A validated case-finding algorithm, using CPT and ICD 9 codes was used to identify all cases of PJI. We assessed the relationship between the hospital total joint replacement (TJR) procedure volume for the 2 years and in-hospital mortality during treatment of a PJI. Analyses were adjusted for age, joint location, Charlson comorbidity score, and type of surgical procedure used to treat PJI.
Statistical analysis was performed with SAS 9.1 (Cary, NC). Comparisons of variables between the group of patients that died during hospitalization and the group of patients that survived were performed using Fisher's exact test (for categorical variables) and Wilcoxon rank sum test (for continuous variables). All tests were two-sided. A p value of 0.05 was considered significant. Variables associated with in-hospital mortality in univariate analysis were included in a stepwise multivariable logistic regression model. Hosmer-Lemeshow test was performed to test the goodness-of-the-fit of the final model.
Results: 119 of 4985 patients (2.3%) died during hospitalization. Pts who died and pts who survived did not differ in gender, race, or the presence of diabetes mellitus, liver disease or S.aureus PJI. Compared to the pts who survived, those who died during hospitalization were significantly older (median 76 vs. 69 yrs, p<0.0001), had more surgical procedures on record (median 6 vs. 3, P<0.001) and were treated in hospitals with lower 2-yr caseloads of TJR (median 295 vs. 348, P = 0.009). Multivariable logistic regression identified pts =60 yrs (OR = 3.6, 95%CI:1.9-6.8), treated with joint aspiration (OR = 2.7, 95% CI: 1.4-4.9), Charlson score =1 (OR = 2.2, 95% CI: 1.4-3.4) and THA PJI (OR = 3.4, 95% CI: 2.0-5.8), as being associated with a higher odds of in-hospital mortality during treatment of PJI, relative to pts younger than 60 years, treated with resection arthroplasty, with a Charlson score <1 and TKA PJI, when adjusting for total number of number of surgical procedures on record. The odds of in-hospital death during treatment of a PJI were lower among patients treated in hospitals with higher 2-year caseloads (=500) of TJR performed (OR = 0.56, 95% CI: 0.36-0.88) relative to those treated in hospitals that performed less than 500 TJR/2 years.
Discussion: This study suggests that the rate of in-hospital mortality during treatment of a PJI is low. In-hospital mortality during treatment of PJIs was associated with a lower 2 -year hospital volume of total joint replacement performed, when controlling for age, Charlson comorbidity score, total number of surgical procedures on record, joint location, and type of surgical procedure to treat PJI. Our findings are similar to the results of previous studies that assessed the association between hospital and surgeon procedure volume and outcomes of total primary and revision hip surgery in the Medicare population.