Clinical findings of bacteremic septic arthritis according to the site of acquisition: The overlap between health care-related and community- and nosocomial-acquired cases

•Almost 20% of bacteremic septic arthritis (SA) was health care-related (HCR).•Characteristics of HCR cases overlapped with those of community-(CA) and nosocomial-acquired (NA).•HCR and NA cases presented older population, and predominance of MRSA and P. aeruginosa.•HCR-SA presented higher mortality than CA but lower than NA.•Identification of HCR-SA is important when planning empirical therapeutical approach.

BackgroundThe site of acquisition of infection may have a major impact on outcome. The health care-related (HCR) environment has recently come under scrutiny. In a group of patients with bacteremic septic arthritis (SA), we compared their characteristics, type of SA, microbiology and prognosis according to the site of acquisition: community-acquired (CA), nosocomial-acquired (NA), and HCR.MethodsWe studied all patients with bacteremic SA seen at our institution between 1985 and 2013. Data were obtained from a protocol of prospectively recorded bacteremia cases.ResultsThere were 273 cases of bacteremic SA (CA: 51%; NA: 31%; and HCR: 18%). NA and HCR sites were more frequent in older and fragile patients. SA of peripheral joints was the most common presentation; infections of the axial skeleton predominated in CA and HCR (24%), and prosthetic joint infection in NA (44%). MRSA and Pseudomonas aeruginosa were mainly found in NA (21% and 6% respectively) and HCR (14% and 8% respectively), whereas Streptococcus spp. was more frequent in CA (30%) and HCR (28%). The 30-day mortality rates were: CA 7%, HCR 18%, and NA 26%.ConclusionThe characteristics of HCR-SA overlapped with those of the CA or NA-SA cases. The HCR and NA cases presented more advanced age, greater fragility, and the predominance of difficult-to-treat microorganisms, while the HCR and CA cases presented an involvement of the axial skeleton, streptococcal etiology, and a lower number of prosthetic joint infections. Our data show that the site of acquisition should be considered when planning diagnostic and therapeutic management for SA.