A Clinical Prediction Model for Complications After Cranioplasty Based on Modified-Brain Collapse Ratio and Comorbidity Burden.

Abstract

Cranioplasty (CP) after decompressive craniectomy is linked to a high complication rate. Although neuroimaging parameters and comorbidity burden are considered potential predictors, no predictive model has been established. This study aimed to develop a clinical prediction model to visualize and ameliorate the occurrence of post-CP complications. Our study retrospectively encompassed 368 adults undergoing unilateral CP after decompressive craniectomy and divided them into 2 groups based on the occurrence of complications. Modified-brain collapse ratio (m-BCR) was calculated by a 3-dimensional way and age-adjusted Charlson Comorbidity Index (aCCI) scores were collected from electronic records. Postoperative complications occurred in 18.48% (68/368) of patients. Multivariable analysis identified 5 independent predictors: m-BCR (odds ratio [OR = 1.670, 95% confidence interval [CI]: 1.150-2.426, P = 0.007), aCCI score (OR= 1.450, 95% CI: 1.233-1.706, P < 0.001), operative duration (OR = 1.005, 95% CI: 1.000-1.010, P = 0.044), intraoperative blood loss (OR = 1.006, 95% CI: 1.001-1.010, P = 0.010), and total serum protein (OR = 0.963, 95% CI: 0.928-0.998, P = 0.040). Receiver operating characteristic analysis showed optimal cutoffs: m-BCR = 1.265 (sensitivity 60.3%, specificity 79.7%) and aCCI = 1.5 (61.8%, 70.3%). The integrated prediction model demonstrated superior discrimination (area under the curve = 0.776, 95% CI: 0.712-0.840, P < 0.001) compared to individual parameters. Based on m-BCR and aCCI as satisfactory risk predictors with significant weights, an effective clinical model was developed to predict complications after CP.