Risk factors for postoperative fever following cranioplasty in children: a retrospective cohort analysis.

Abstract

<h4>Purpose</h4>To identify and quantify modifiable perioperative risk factors for early postoperative fever (POF) in children undergoing cranioplasty (CP), given that most early POF events are non‑infectious and contribute to longer hospitalization and increased healthcare costs.<h4>Methods</h4>A retrospective single-center cohort of 216 pediatric CP patients (≤ 18 years, 2016-2024) was analyzed. Baseline variables included demographics, implant material, dural repair material, cranial defect size, and so on. Least Absolute Shrinkage and Selection Operator (LASSO) regression with tenfold cross-validation selected candidate predictors; multivariable logistic regression refined the model. Variables with statistically significant associations (p < 0.05) were used to build a nomogram. Model discrimination (C‑index), calibration, ROC curve, and decision curve analysis assessed predictive performance.<h4>Results</h4>POF occurred in 71/216 patients (32.9%). Four variables were independent predictors: CP material (titanium mesh vs. polyetheretherketone [PEEK]: OR = 3.81, 95% CI 1.92-7.92, p < 0.001), dural repair material (synthetic patch vs. intact dura: OR = 7.57, 95% CI 2.45-25.42, p < 0.001; autologous tissue vs. intact dura: OR = 1.77, 95% CI 0.78-4.02, p = 0.171), subcutaneous drainage (with vs. without: OR = 0.29, 95% CI 0.12-0.66, p = 0.004), and cranial defect area (OR = 1.01 per cm² increase, 95% CI 1.00-1.02, p = 0.003). The nomogram had moderate predictive accuracy (C‑index 0.791 ± 0.032; AUC 0.790, 95% CI 0.727-0.853) and demonstrated good calibration and clinical utility.<h4>Conclusion</h4>In pediatric CP, use of titanium mesh, synthetic dural patches, absence of subcutaneous drainage, and larger cranial defect size are modifiable risk factors significantly associated with early POF. Preventive strategies focusing on implant and dural repair material selection, proactive drainage, and consideration of cranial defect size may reduce POF incidence and its clinical impact.