O0020: Emergency Department Observation of Children with Minor Head Trauma

Background: The Pediatric Emergency Care Applied Research Network (PECARN) traumatic brain injury (TBI) prediction rules suggest a period of observation prior to decision-making regarding computed tomography (CT) in children with minor blunt head trauma (BHT) who are not immediately identified as high- or very low-risk of clinically important TBI (ciTBI). We sought to evaluate the role of observation in the management of children with minor BHT.

Methods: We performed an a priori subanalysis of a prospective multicenter study of children with minor BHT (Glasgow Coma Scale Score ≥14). Physicians documented whether a child was observed before deciding on CT. We defined ciTBI as a TBI resulting in death, neurosurgical intervention, intubation > 24 hours, or admission for >2 nights due to the TBI in association with a positive CT. Guardians of patients discharged from the emergency department (ED) were contacted one week later to assess for ciTBI. To determine the association of observation on CT use, a multivariable logistic regression model controlling for hospital clustering and patient characteristics was created for patients < 2 and ≥ 2 years old.

Results: 20,316 children (mean age ± standard deviation: 6.5 ± 5.3 years) were enrolled. Clinicians noted if the patient was observed before CT decision-making in 20,037 (99%) patients and 4,564 (23%) patients were observed. CT rate was 619/4,564 (13.6%) in those observed versus 5,779/15,473 (37.3%) in those not observed; difference: 23.8% (95% confidence interval (CI): 22.5 to 25.0%). The rate of ciTBI was 14/4,564 (0.3%) in those observed versus 210/15,473 (1.4%) in those not observed (difference: 1.0%; 95% CI 0.8, 1.2%). The median age in those observed (4.0 years) was lower than those not observed (5.3 years, p< 0.001). After adjustment for hospital and patient characteristics, observation was associated with decreased CT use in patients < 2 years (odds ratio = 0.15 [95% CI 0.12, 0.19]) and ≥ 2 years (odds ratio = 0.15 [95% CI 0.12, 0.17]). The median length of stay (LOS) for patients who had an immediate CT decision (2.6 hours) was shorter than that for observed patients (2.97 hours, p< 0.001). In the 619 patients who underwent CT after observation, the median time from ED arrival to CT was 1.75 hours. 90/619 (15%, 95% CI 12, 18%) had TBI on CT and 14/619 (2.3%, 95% CI 1.2, 3.8%) had ciTBI. 4,182 patients were discharged home after observation without CT and none (0%, 95% CI 0.000, 001%) were subsequently identified with ciTBI.

Conclusion: Observation was associated with a safe decrease in CT use among children with minor BHT with a slight increase in ED LOS. For children who are not immediately identified as having high- or very low-risk of ciTBI, a period of observation is recommended.