Predictors of 30-day Readmission in Firearm-related Traumatic Brain Injury- A Nationwide Readmission Database Study
Christopher Chang1, John Lin4, Jia-Shu Chen5, Joshua Feler1, Belinda Shao1, Diana Wang2, Kevin Nguyen6, Damian Sanchez5, Ethan Winkler5, Megan Ranney7, Ali Mahta3
1Department of Neurosurgery, 2Department of Neurology, 3Department of Neurocritical Care, The Warren Alpert Medical School of Brown University, 4Department of Medicine, The Perelman School of Medicine, 5Department of Neurological Surgery, The University of California, San Francisco, 6The Department of Neurosurgery, The Icahn Medical School at Mount Sinai, 7The School of Public Health, Yale University
Objective:
Using the Nationwide Readmission Database (NRD), this study evaluates the predictors of readmission for adult patients with firearm-related traumatic brain injuries (TBIs).
Background:
30-day hospital readmissions (30dRA) are used to measure quality of care, as higher rates are associated with poor outcomes and complications. TBI readmissions disrupt postoperative care and delay recovery. The readmission rates for firearm TBIs - a high lethality injury - are unknown.
Design/Methods:
Firearm injury-related TBI discharges in the NRD were tracked for 30dRA between 2016 and 2021. Patients ≥18 years old with primary TBI diagnoses and co-diagnoses of firearm injury were selected using ICD-10-CM codes. Non-civilian firearm injuries, elective admissions, in-hospital deaths, no available length of stay (LOS), and out-of-state admissions were excluded. NRD discharge weights were applied. Unpaired t-tests and Rao-Scott chi-square tests adjusted for NRD discharge weights were conducted to compare demographics, hospital characteristics, procedures, comorbidities, complications, and discharge patterns in patients with and without 30dRA.
Results:
Our cohort comprised 3654 discharges, including 365 (10%) with corresponding readmissions. Patients with tracheostomy (p<0.01), gastrotomy (p<0.01), decompressive craniectomy (p<0.01), or other CNS drainage procedures (p=0.025) were more likely to have 30dRA. Postprocedural nervous system hemorrhage (p<0.01) and pneumonia (p<0.01) were also associated with 30dRA. Longer LOS (p=0.0131) and transfer to a short-term hospital (p<0.01) had greater probabilities of 30dRA. Routine discharge (p<0.01), as defined by the NRD (e.g. discharged to home or self-care), was associated with a decreased likelihood of 30dRA.
Conclusions:
More invasive procedures–often performed for the treatment of severe TBI–are associated with higher odds of 30dRA outcomes in survivors. Postprocedural neurological and respiratory complications were associated with 30dRA. Longer LOS and transfers to short-term hospitals were associated with 30dRA. These findings may reflect the severity of injury at baseline but also suggest the importance of meticulous procedural selection and discharge management.
Disclaimer: Abstracts were not reviewed by Neurology® and do not reflect the views of Neurology® editors or staff.