Methods: This was a retrospective case control study among patients with PJI from 2006-2016. Each fungal PJI case was matched 1:1 with a bacterial PJI control for joint location (hip, knee, shoulder) and year of diagnosis. We compared demographics (age, sex, race), co-morbid conditions (BMI, diabetes, immunosuppression, renal disease, antibiotic use), and clinical characteristics (joint age, wound factors, lab data, previous joint surgeries, and previous PJI) between fungal and bacterial PJI groups using chi square/Fisher’s exact or Wilcoxon rank sum test. Risk factors statistically (p<0.05) or clinically significant were included in a multivariable logistic regression (MVR) model in stepwise fashion (SAS 9.4, Cary, NC).
Results: 41 fungal PJI occurred over the study period and 61% were due to Candida albicans. Median age was 64.7 years, 51% were females, and 87% were White. The hip was involved in 51.2%, followed by the knee (46.3%), and shoulder (2.4%). There were no significant differences in joint age or co-morbid conditions. Compared to bacterial PJI, those with fungal PJI were more likely to have received antibiotics within the past 3 months (70.7% vs. 34%, p=0.001), had wound drainage lasting more than 5 days (48% vs. 9%, p=0.0002), had a lower median CRP (2.95 mg/dl vs. 5.99, p=0.013) and synovial fluid wbc count (13,953 cells/mm3 vs. 33,198, p=0.007), and a higher proportion of prior two-stage exchanges (82.9% vs. 53.6%, p=0.008). After MVR, controlling for the center, presence of wound drainage for more than 5 days (OR, 7.3; 95 % confidence interval [CI], 2.02-26.95) and receipt of antibiotics within the past 3 months (OR, 3.4; 95 % CI, 1.2-9.3) were factors significantly associated with fungal PJI.
Conclusion: In our study, Candida albicans was the most common species in fungal PJIs. Presence of wound drainage for more than 5 days and receipt of antibiotics within the past 3 months were independent risk factors for fungal PJI among a cohort of PJI patients.
L. L. Steed, None
H. Demos, None
C. Salgado, None
D. Osmon, None
C. Marculescu, None