Who Stays and Who Goes? Symptoms Predicting Admission After Mild Head Injury in Older Adults

Operations pressures in the Emergency Department (ED) have increased the utility of pathways facilitating early identification of patients who will require admission. specifically older adults with non-severe head injury (HI). This study, a secondary analysis of prospectively collected data in a randomized controlled drug trial, aimed to evaluate factors on ED presentation that can predict non- discharged status at 24 hours. The ultimate goal was to establish parameters that can be used to inform ED pathways and admission prediction scores for mild HI in older patients.

Subjects were sampled from the first 1500 cases of “CRASH4”, with eligibility narrowed to Glasgow Coma Score of 15, normal vital signs, and no positive findings on head computed tomography (CT). Research staff assessed five key symptoms dichotomously in every patient (no missing data): amnesia, confusion or drowsiness, loss of consciousness (LOC), nausea without vomiting, and vomiting. Potential confounders assessed include study year, time from injury to ED presentation, age, sex, frailty index (1 best to 7 worst), and exposure to anticoagulant or antiplatelet medications. Admission was defined by presence in the ED or hospital for more than 24 hours.

Of 953 eligible cases (52.5% female, median age 74), 97.8% of whom had pre-HI functional independence (median frailty index 3), 285 (29.9%) were admitted. Variables predictive of admission included age (each year older, OR 1.04, 95% CI 1.02-1.06, p <.001), male sex (OR 1.04, 95% CI 1.06-2.13, p=.021), frailty score (each point higher, OR 1.36, 95% CI 1.19-1.56, p<.001), amnesia (OR 1.88, 95% CI 1.33-2.09, p=.030), LOC (OR 1.47, 95% CI 1.04-2.09, p = .030) and vomiting (OR 2.09, 95% CI 1.25-3.50, p = .005).

Ambulatory, functionally independent patients over 50 with normal GCS and vital signs presenting to the ED with HI have a relatively high risk of admission even if CT is negative. When adjusting for known influencers of admission likelihood (e.g. age, frailty), the strongest symptomatic predictors of admission that can potentially inform development of HI pathways and admission-predictive instruments were vomiting and amnesia, both of which approximately doubled odds of admission. LOC had only intermediate predictive strength, and confusion, drowsiness, and nausea without vomiting do not appear to warrant further evaluation in predictive tools.