Periprosthetic joint infection:A South African perspective

Abstract

Background. South African data on the bacteriology and sensitivity profile of periprosthetic joint infection is lacking.  Current regimens for systemic and local antibiotic therapy are based on international literature. These regimens are different for the United States of America and Europe and might thus not be relevant to South Africa. Objectives.  To determine the characteristics of periprosthetic joint infection in a South African clinical setting by identifying the most common organisms cultured and establishing their antibiotic sensitivities in order to propose the most appropriate empiric antibiotic treatment regimen. In the case of two-stage revision procedures, we aim to compare the organisms cultured during the first stage versus organisms cultured during the second stage in second-stage procedures that had positive cultures.  Furthermore, in these culture-positive second-stage procedures we aim to correlate the bacterial culture with the erythrocyte sedimentation rate/ C-reactive protein result. Methods. We performed a retrospective cross-sectional study looking at all hip and knee periprosthetic joint infections in patients 18 years and older, treated at a government institution and a private revision practice in Johannesburg, South Africa between January 2015 and March 2020. Data were collected from the Charlotte Maxeke Johannesburg Academic Hospital hip and knee and the Johannesburg Orthopaedic hip and knee databanks. Results.  We included 69 patients whom underwent 101procedures relating to periprosthetic joint infection. Positive cultures were found in 63 samples, 81 different organisms were identified. The most common organisms cultured were Staphylococcus aureus (n = 16, 19.8%) and Coagulase negative Staphylococcus (n = 16, 19.8%), followed by Streptococci species (n = 11, 13.6%).  The positive yield in our cohort was 62.4% (n = 63). A polymicrobial growth was found in 19% (n = 12) of the culture positive specimens. Of all the microorganisms cultured, 59.2% (n = 48) were Gram-positive versus 35.8% (n = 29) Gram-negative. The remainder were fungal and anaerobic organisms at 2.5% (n = 2) each. Gram-positive cultures displayed 100% sensitivity to Vancomycin and Linezolid, whereas Gram-negative organisms displayed 82% sensitivity towards Gentamycin and 89% sensitivity towards Meropenem respectively. Conclusion. Our study identifies the bacteriology of periprosthetic joint infections and their sensitivities in a South African setting. We recommend that empiric antibiotic-loaded cement spacers and systemic antibiotic regimens should consist of Meropenem or Gentamycin; Vancomycin and Rifampicin to achieve the broadest spectrum of coverage and most likely success in eradicating infection.