Validation of septic arthritis diagnosis and disease characteristics in West Sweden

Abstract

Background: This study aimed to validate the septic arthritis diagnosis using International Classification of Diseases (ICD)-10 codes against the Newman criteria. Additionally, the study presents disease characteristics and compares subgroups. Methods: 475 patients with an ICD-10 code for septic arthritis were identified at a university (n = 265) and a county (n = 167) hospital in West Sweden between 2016 and 2019. Medical records were analysed for adherence to the Newman criteria. Clinical data and culture results were compared between the university hospital and the county hospital, as well as between subgroups categorised by the pathogenetic route of infection, including direct inoculation and haematogenous spread infection. Results: 91% of the patients fulfilled the Newman criteria (52% positive for synovial culture or PCR, 24% positive culture from blood or elsewhere, and 24% with radiological evidence or turbid synovial fluid). The patient population was predominantly male (62%), with a median age of 69 years, and the knee was the most commonly affected joint (39%). Staphylococcus aureus emerged as the predominant pathogen across all patient groups. 80% had a haematogenous spread septic arthritis, while 20% resulted from direct inoculation. In the haematogenous spread groups, patients were older, with fewer males, but higher CRP levels compared to those in the direct inoculation group. Conclusion: ICD-10 codes for septic arthritis have a positive predictive value of 91% in relation to the Newman criteria in Swedish hospital care, making the utilisation of ICD-10 codes a reliable data source for future epidemiological studies.