A novel risk calculator predicting surgical site infection after spinal surgery in patients with cerebral palsy

AIM

To develop and validate a risk calculator based on preoperative factors to predict the probability of surgical site infection (SSI) in patients with cerebral palsy (CP) undergoing spinal surgery.

METHOD

This was a multicenter retrospective cohort study of pediatric patients with CP who underwent spinal fusion. In the development stage, preoperative known factors were collected, and a risk calculator was developed by comparing multiple models and choosing the model with the highest discrimination and calibration abilities. This model was then tested with a separate population in the validation stage.

RESULTS

Among the 255 patients in the development stage, risk of SSI was 11%. A final prediction model included non-ambulatory status (odds ratio [OR] 4.0), diaper dependence (OR 2.5), age younger than 12 years (OR 2.5), major coronal curve magnitude greater than 90° (OR 1.3), behavioral disorder/delay (OR 1.3), and revision surgery (OR 1.3) as risk factors. This model had a predictive ability of 73.4% for SSI, along with excellent calibration ability (p = 0.878). Among the 390 patients in the validation stage, risk of SSI was 8.2%. The discrimination of the model in the validation phase was 0.743 and calibration was p = 0.435, indicating 74.3% predictive ability and no difference between predicted and observed values.

INTERPRETATION

This study provides a risk calculator to identify the risk of SSI after spine surgery for patients with CP. This will allow us to enhance decision-making and patient care while providing valid hospital comparisons, public reporting mechanisms, and reimbursement determinations.