Development and Validation of Futility of Resuscitation Measure in Older Adult Trauma Patients

Sai Krishna Bhogadi,Michael Ditillo,Muhammad Haris Khurshid,Collin Stewart,Omar Hejazi,Audrey L Spencer,Tanya Anand,Adam Nelson,Louis J Magnotti,Bellal Joseph
DOI: https://doi.org/10.1016/j.jss.2024.07.019
Abstract:Introduction: This study aimed to develop and validate Futility of Resuscitation Measure (FoRM) for predicting the futility of resuscitation among older adult trauma patients. Methods: This is a retrospective analysis of the American College of Surgeons-Trauma Quality Improvement Program database (2017-2018) (derivation cohort) and American College of Surgeons level I trauma center database (2017-2022) (validation cohort). We included all severely injured (injury severity score >15) older adult (aged ≥60 y) trauma patients. Patients were stratified into decades of age. Injury characteristics (severe traumatic brain injury [Glasgow Coma Scale ≤ 8], traumatic brain injury midline shift), physiologic parameters (lowest in-hospital systolic blood pressure [≤1 h], prehospital cardiac arrest), and interventions employed (4-h packed red blood cell transfusions, emergency department resuscitative thoracotomy, resuscitative endovascular balloon occlusion of the aorta, emergency laparotomy [≤2 h], early vasopressor requirement [≤6 h], and craniectomy) were identified. Regression coefficient-based weighted scoring system was developed using the Schneeweiss method and subsequently validated using institutional database. Results: A total of 5562 patients in derivation cohort and 873 in validation cohort were identified. Mortality was 31% in the derivation cohort and FoRM had excellent discriminative power to predict mortality (area under the receiver operator characteristic = 0.860; 95% confidence interval [0.847-0.872], P < 0.001). Patients with a FoRM score of >16 had a less than 10% chance of survival, while those with a FoRM score of >20 had a less than 5% chance of survival. In validation cohort, mortality rate was 17% and FoRM had good discriminative power (area under the receiver operator characteristic = 0.76; 95% confidence interval [0.71-0.80], P < 0.001). Conclusions: FoRM can reliably identify the risk of futile resuscitation among older adult patients admitted to our level I trauma center.
What problem does this paper attempt to address?