K. M. Techar1, A. Nguyen1, R. M. Lorenzo1, S. Yang1, B. Thielen1, A. Cain-Nielsen2, M. R. Hemmila3, C. J. Tignanelli4,5,6 4University Of Minnesota,Department Of Surgery,Minneapolis, MN, USA 5North Memorial Medical Center,Department Of Surgery,Robbinsdale, MN, USA 6University Of Minnesota,Institute For Health Informatics,Minneapolis, MN, USA 1University Of Minnesota,Medical School,Minneapolis, MN, USA 2University Of Michigan,Center For Healthcare Outcomes And Policy,Ann Arbor, MI, USA 3University Of Michigan,Department Of Surgery,Ann Arbor, MI, USA
Introduction:
Traumatic Brain Injury (TBI) is responsible for 30% of trauma related deaths each year and is a major cause of permanent disability. Head computed tomography (CT) imaging is the gold standard for diagnosis of intracranial bleeding in TBI, however institutional time to imaging varies significantly, especially in patients without signs of acute distress. Studies have shown earlier clinical intervention is associated with improved outcomes. The objective of this study was to identify the optimal imaging time and its impact on outcomes for elderly patients with head trauma who present to the emergency department (ED) without signs of acute distress.
Methods:
Data from a state-wide quality collaborative was used from 2011-2017 at 29 level 1 and 2 trauma centers. Inclusion criteria were: ICD-9/10 codes for head trauma, age≥50, Glasgow Coma Scale (GCS) ≥14, Injury Severity Score (ISS) ≤20, non-full trauma activation, and head CT imaging time within 1.5 hours of arrival, excluding the initial 5 minutes. Direct admissions and patients who arrived with no signs of life were excluded. Lowess plots were generated to evaluate the association of time to head CT on in-hospital mortality. Data was dichotomized based on these findings into early and late CT cohorts. Logistic regression and negative binomial models were fit to the data to evaluate early vs late CT. Models were risk adjusted for age, gender, race, insurance status, pre-injury anticoagulant use, ED blood pressure, Abbreviated Injury Scale, GCS, and ISS. The primary outcome was in-hospital mortality. Hospital-level factors associated with early CT use were evaluated using logistic regression.
Results:
6,336 patients were included in this study. There was significant variation in time to head CT (μ-45 minutes(m), SD-22m). Mortality nadired at 35 minutes on lowess. Each one minute delay in time to head CT was associated with a 2% increase in mortality (OR 1.02, 95% CI 1.01-1.03, p=0.002). Data was dichotomized into early (≤35m, n=2,535) and late (>35m, n=3,801) cohorts. Early patients had significantly reduced in-hospital mortality (OR 0.58 95% CI 0.35-0.95,p=0.03). Early patients on anticoagulant medications were more likely to receive FFP within 4 hours (OR 1.5,p=0.03). Early patients did not have significantly faster times to neurosurgical intervention (IRR 0.76, 95% CI 0.48-1.2, p=0.2) but did have significantly shorter ED length of stay (IRR 0.89, 95% CI 0.87-0.92, p<0.001). Level 2 (OR 0.46, p<0.001), teaching (OR 0.74, p<0.001), and high-volume trauma centers (OR 0.80, p=0.001), were all less likely to provide early head CTs.
Conclusion:
Each one minute delay in head CT for elderly patients with head trauma is associated with a 2% increase in mortality. This may be due to slower delivery of therapeutic interventions such as anticoagulation reversal. Head CT within 35 minutes for elderly patients with head trauma should be evaluated as a potential quality metric.