Pediatric closed head injuries treated in an observation unit.

BACKGROUND: Closed head injury (CHI) is common in childhood and frequently results in hospital admission for observation and treatment. Observation units (OUs) have shown significant benefits for patients and physicians. At our institution, a level 1 pediatric trauma center, patients with CHI are often admitted to an OU for up to 24 hours of observation and treatment. STUDY OBJECTIVES: To describe characteristics of patients with a CHI admitted to a pediatric OU and to identify demographic, historical, clinical, and radiographic factors associated with the need for unplanned inpatient admission (UIA) after OU management. METHODS: Retrospective cohort review of all OU admissions for CHI at Primary Children's Medical Center (PCMC) from August 1999 through July 2001. Data collected included age, gender, mechanism of injury, presenting symptoms, physical examination findings, head computed tomography (CT) results, diagnosis, length of stay, outcome of the injury, and need for UIA. RESULTS: During the study period, 827 patients were seen in the ED for CHI. Two hundred eighty-five patients (34%) were admitted to the OU, 273 (33%) were admitted to an inpatient service, and 269 (33%) were discharged home. OU patients had a median age of 5.2 years, ranging from 2 weeks to 17 years. Sixty-one percent were male. The median admission length of stay was 13 hours. Common mechanisms of injury included: falls (60%), motor vehicle accidents (12%), bicycle accidents (10%), impacts from objects (9%), auto-pedestrian accidents (4.6%), and snow-related accidents (4.6%). Presenting symptoms in the ED included vomiting (39%), loss of consciousness (26%), amnesia to event (19%), persistent amnesia (5%), and seizures (4%). Physical examination findings noted in the ED included altered mental status (45%), facial abnormalities (43%), scalp abnormalities (38%), and neurologic deficits (9%). Two hundred eighty patients (98%) admitted to the OU had a head CT performed. Skull fractures were present in 109 patients (39%) and intracranial pathology (ie, epidural hematoma, subdural hematoma, or intraparenchymal contusion) was present in 38 patients (13%). Only 13 patients (5%) required admission to an inpatient service from the OU for the following reasons: continued need for intravenous (IV) fluids (n = 5), venous thrombosis (n = 2), persistent CSF leakage (n = 3), decreased level of consciousness (n = 1), pain management (n = 1), and clearing of the patient's cervical spine (n = 1). No patient deteriorated or required neurosurgery. Patients with basilar skull fractures, a head laceration (scalp or facial), and patients that needed IV fluids in the ED were more likely to need inpatient admission after a 24-hour observation stay. Logistic regression analysis identified basilar skull fractures (OR 11.61), face/scalp lacerations (OR 7.52), and the need for ED IV fluid administration (OR 4.26) to be associated with UIA. Most children with these findings were successfully discharged within 24 hours, however. Age, sex, loss of consciousness, seizure, vomiting, amnesia, altered mental status, neurologic deficits, intracranial pathology, and skull fractures (aside from basilar skull fractures) were not related to UIA. CONCLUSION: The vast majority (96%) of pediatric OU patients with CHI such as small intracranial hematomas, skull fractures, and concussions were discharged safely within 24 hours without serious complications. The presence of a basilar skull fracture, head laceration, and the need for ED IV fluids were associated with increased risk of UIA. OU admission is an efficient and effective management setting for children with stable intracranial pathology, skull fractures, and concussions.