Disclosure(s): No relevant financial relationship(s) to disclose.
First Author: Renee Groechel, PhD Co-Author: Emmett McGuire, BA, MD, FACS – Program Director for General Surgery Residency, Swedish Medical Center Co-Author: Gina Berg, PhD, MBA, MSMFT – Director of Research, Wesley Medical Center Co-Author: Christopher Zaw-Mon, MD – Trauma Medical Director, Intermountain Health Lutheran Hospital Co-Author: Carlos Palacio, MD, FACS – Chief of Staff, South Texas Health System Co-Author: David Bar-Or, MD, FACEP – Director of Research, Injury Outcomes Network
Introduction: The Glasgow Coma Scale (GCS) is used to determine the severity of neurologic injury in trauma patients. However, comorbidities, such as dementia, make it difficult to distinguish whether score reductions in the GCS are reflective of impairment due to injury or that of baseline function. We hypothesize that among geriatric patients with similar head injury severity, defined by the Abbreviated Injury Scale (AIS) of the head, those with dementia will have lower GCS scores compared to those without dementia.
Methods: The National Trauma Data Bank (NTDB; 2019 – 2022) was queried for patients aged ≥ 65 years with GCS scores 13 – 15 at hospital admission and head AIS 3, 4, or 5. Demographics, GCS, and other characteristics were compared between patients with and without preexisting dementia. The analytic sample was stratified into six subgroups (3x2) based on head AIS 3, 4, 5 and preexisting dementia (with vs without). Propensity score matching (PSM; 1:1) was performed within each AIS group in models adjusted for demographics, injury severity score, and characteristics related to patient care. Differences in GCS scores were analyzed with Wilcoxon signed-rank tests.
Results: Of 133,384 patients meeting inclusion criteria, 22,481 (17%) had dementia. Patients with dementia had lower GCS scores than those without dementia: GCS 13 (8% vs 4%), GCS 14 (47% vs 20%), GCS 15 (45% vs 76%) (chi-squared p< 0.001). After PSM modeling adjusted for 9 covariates, GCS scores remained lower among patients with dementia at head AIS 3, 4 or 5. In analyses of head AIS 3 (n=12,284 matched pairs), patients with dementia had lower GCS scores (median:14 [IQR:14-15]) than those without dementia (median: 15 [IQR:15–15], p< 0.001); the same significant pattern was observed in analyses of head AIS 4 (n=6,347 matched pairs) and head AIS 5 (n=3,395 matched pairs).
Conclusions: We demonstrate that among patients with similar head injury severity, patients with dementia had lower GCS scores (median 14) than those without dementia (median 15). This discrepancy may reflect that GCS scores in patients with dementia are not reflective of head injury severity. To ensure patients with dementia receive appropriate care for the injury sustained, future work should explore alternatives and/or modifications to the GCS in this population.
