Prediction of septic knee arthritis

In this study, we constructed an algorithm to determine the predicted probability of septic knee arthritis based on three independent multivariate predictors. For example, a child's blood test showed ESR > 20 mm/h, CRP > 10 mg/L, and AMONO > 0.74×10⁹/L. When all three factors are present, the probability of a septic knee joint is 94.68%. The area under the ROC curve was 0.87, indicating excellent diagnostic performance of this set of three multivariate predictors in identifying septic knee arthritis. In the calibration curve of the prediction model, the calibration plot shows good prediction accuracy between the actual and predicted probabilities.