<< Back to the abstract archive
Optimizing Risk Stratification Tools for Patients Undergoing Cranioplasty
Abdulaziz Elemosho MD, Emily Pfahl BS, Taborah Z. Zaramo MD, Jude C. Kluemper BS, Ryan Enslow BS, Caroline Gillespie BS, Kerry-Ann S. Mitchell MD-PhD FACS.
The Ohio State University
2025-01-10
Presenter: Abdulaziz Elemosho
Affidavit:
I certify that the work presented here is the original work of Zed and other members of the Mitchell Lab. And this work has not been presented at any other meeting
Director Name: Gregory Pearson MD
Author Category: Medical Student
Presentation Category: Clinical
Abstract Category: Craniomaxillofacial
Background
Cranioplasty, performed after decompressive craniectomy DC for severe traumatic brain injury (TBI) , stroke etc, has high complication rates (up to 32%). Existing risk stratification tools overlook defect size, though larger cranioplasties often involve frailer patients with higher Risk Analysis Index-A (RAI-A) and mFI-5 scores. This study assesses these scores' predictive utility for outcomes in large defect cranioplasty.
Methods
An IRB approved 9-year retrospective review of adult patients who had large defect cranioplasty was conducted. Patients were stratified into high (RAI >30 and mfi-5 ≥3) and low (RAI ≤30 and mfi-5 ≤2) risk group. Logistic regression and Receiver Operator Curve (ROC) analysis assessed their relationship with postoperative outcome.
Results
The analysis included 302 patients (mean age 50.5 years). Reoperation occurred in 28.8%, Clavien-Dindo III-IV complications in 28.4%, 30-day morbidity in 32.7%, and cranioplasty failure in 11.6%. Patients with RAI >30 or mFI-5 ≥3 had higher reoperation risk (p=0.012, p=0.008, respectively), with both scores showing robust predictive accuracy (AUC: 0.85). Reoperation risk was elevated for patients with prior surgeries (p<0.0001) and frontoparietal cranioplasties (p=0.02). RAI >30 increased 30-day morbidity risk (p=0.033), particularly in patients with frozen autogenous bone grafts (p=0.02), with RAI outperforming mFI-5 (AUC: 0.75). Conversely, mFI-5 demonstrated superior prediction of graft failure (AUC: 0.86), with mFI-5 ≥3 significantly associated with hardware/graft failure (p=0.00012).
Indications for DC was not a significant predictive variable for adverse outcome.
Conclusions
We have shown that RAI and mfi-5 scores can stratify patients risks and help surgeons anticipate complications in patients undergoing cranioplasty.
