Multiple cultures and extended incubation for upper extremity revision arthroplasty affect clinical care: A cohort study

Introduction: Skin flora organisms (SFOs) isolated from 1 to 2 tissue samples during shoulder and elbow revision arthroplasty are difficult to distinguish as contamination or infection. We examined the change in clinical care after implementation of an Arthroplasty Infection Protocol by increasing the number of intraoperative samples held for 10-day incubation to a minimum of 5. Methods: Infection was defined as ≥3 cultures growing the same SFO or any one culture growing any other virulent organism. SFOs growing in 1 to 2 samples were defined as skin flora contaminant. All cases were compared with pre-Arthroplasty Infection Protocol institution standard to determine changes in microbiological diagnosis and resultant antibiotic treatment. Results: Forty cases fulfilled the inclusion criteria: 50% of these were culture negative, and 35% grew Propionibacteria. When compared with the standard of obtaining one sample, this protocol altered the microbiological diagnosis and subsequent antibiotic treatment in 45% of cases (95% confidence interval 29% to 62%). This protocol had a predictive value of joint sterility in 95% of culture-negative cases (95% confidence interval 74% to 99%). Discussion: The addition of 5 or more samples held for 10-day incubation reliably differentiated between joint infection, contamination, and sterility, which changed the course of care in 45% of surgical cases.