BACKGROUND: Osteoarticular infections (OAI) are associated with complications and sequelae in children, whose prediction are of great importance in improving outcomes. We aimed to design risk prediction models to identify early complications and sequelae in children with OAI.
METHODS: This observational study included children (>3 months-17 years old) with acute OAI admitted to a tertiary-care pediatric hospital between 2008 and 2018. Clinical treatment, complications and sequelae were recorded. We developed a multivariable logistic predictive model for an acute complicated course (ACC) and another for sequelae.
RESULTS: A total of 240 children were identified, 17.5% with ACC and 6.0% and 3.6% with sequelae at 6 and 12 months of follow-up, respectively. In the multivariable logistic predictive model for ACC, predictors were fever at admission [adjusted odds ratio (aOR): 2.98; 95% confidence interval (CI): 1.10-8.12], C-reactive protein ≥100 mg/L (aOR: 2.37; 95% CI: 1.05-5.35), osteomyelitis (aOR: 4.39; 95% CI: 2.04-9.46) and Staphylococcus aureus infection (aOR: 3.50; 95% CI: 1.39-8.77), with an area under the ROC curve of 0.831 (95% CI: 0.767-0.895). For sequelae at 6 months, predictors were age ≥4 years (aOR: 4.08; 95% CI: 1.00-16.53), C-reactive protein ≥110 mg/L (aOR: 4.59; 95% CI: 1.25-16.90), disseminated disease (aOR: 9.21; 95% CI: 1.82-46.73) and bone abscess (OR: 5.46; 95% CI: 1.23-24.21), with an area under the ROC curve of 0.887 (95% CI: 0.815-0.959).
CONCLUSIONS: In our model we could identify patients at low risk for complications and sequelae, probably requiring a less aggressive approach.