Septic Arthritis in Children: A Longitudinal Population-Based Study in Western Australia

To describe the incidence, risk factors and long-term outcomes in children hospitalised with septic arthritis (SA) in Western Australia (WA).

We extracted state-wide longitudinally linked administrative health data for patients aged < 16 years with a first diagnostic code of 711.X (ICD9-CM) and M00.X (ICD10-AM) in WA in the period 1990-2010. Annual incidence rates (AIR) per 100,000 with 95% confidence intervals (CIs), prior conditions during a median lookback period of 63.2 [interquartile range (IQR) 19.8-117.1] months and outcomes, including standardised mortality rates (SMR), during a median follow-up of 10 years are reported.

A total of 891 patients [62% male, median age 6.4 (IQR 1.9-10.6) years with 34% aged < 3 years] were admitted for SA during the observation period. The overall AIR (per 100,000) was 9.85 (95% CI 4.79-14.41), and was higher in Indigenous Australians [34.9 vs. 5.5 (non-Indigenous), p < 0.001] and in males [11.9 vs. 7 (females), p < 0.01]; AIR showed no temporal or seasonal variation. Knees (43.9%), hips (34.6%) and ankles (13.3%) were most frequently affected, with Staphylococci predominant (49%) in patients with positive cultures (41.5%). Prior infection(s) (40.4%) and respiratory disease (7%) were the main pre-existing morbidities. Median hospital stay was 4.0 (IQR 2-8) days, with 1.9% requiring admission to the intensive care unit and 10.4% requiring readmission within 30 days. During follow-up, 26 patients (3.1%) developed osteomyelitis, nine patients were diagnosed with osteoarthrosis (1.1%) and five patients (0.6%) underwent joint replacement. Female patients developed other serious infections more often than male patients (40.5 vs. 27.1%, p < 0.01), as well as other comorbidities (Charlson Comorbidity Index > 0: 34.6 vs. 27.2%, p = 0.02), including diabetes (4.2 vs. 0%; p = 0.001), cardiovascular events (4.2 vs 1.4%, p = 0.002) and chronic arthritis (1 vs. 0%, p = 0.05). The crude mortality rate was low (0.3%), with 99.4% survival at 180 months and no increase in the SMR.

The incidence of SA in children in WA did not change over the 20-year observation period. SA did not lead to excess mortality, but bone and joint complications developed in 5% of patients. The high propensity to comorbid conditions in this young cohort suggests an underlying role of comorbidity in SA development.