Sonication improves microbiologic diagnosis of periprosthetic elbow infection

Periprosthetic joint infection (PJI) is a relatively frequent and oftentimes devastating complication after total elbow arthroplasty (TEA). Its microbiologic diagnosis is usually based on periprosthetic tissue culture (hereafter referred to as tissue culture), but the sensitivity of tissue culture is variable. Although implant sonication culture has been shown to be superior to tissue culture for the diagnosis of hip and knee PJI, only a single small study (of fewer than 10 infected implants) has assessed sonication for PJI diagnosis after elbow arthroplasty.

We retrospectively analyzed 112 sonicate fluid cultures from patients who underwent revision of a TEA at a single institution between 2007 and 2019, comparing results to those of tissue cultures. We excluded patients who had fewer than 2 tissues submitted for culture. Using the Infectious Diseases Society of America guidelines to define PJI, there were 49 infected and 63 non-infected cases. Median ages in the PJI and non-infected groups were 66 and 61 years, respectively. In the non-infected group, 65% were female vs. 63% in the PJI group. We reviewed clinical characteristics and calculated the sensitivity and specificity of tissue compared with sonicate fluid culture. In addition, we compared the sensitivity of tissue culture to the combination of tissue and sonicate fluid culture.

The most common pathogens were coagulase-negative Staphylococcus sp (49%), followed by Staphylococcus aureus (12%). Sensitivity of tissue culture was 63%, and sensitivity of sonicate fluid culture was 76% (P = .109). Specificity of tissue culture was 94% and specificity of sonicate fluid culture was 100%. Sensitivity of sonicate fluid culture in combination with tissue culture was 84% (P = .002 compared to tissue culture alone).

In this study, we found that the combination of sonicate fluid and tissue culture had a greater sensitivity than tissue culture alone for microbiologic diagnosis of PJI after TEA.