Mild head injury : inhospital observation or computed tomography?

Abstract

Background: Patients with mild head injuries are treated every day in emergency departments around the world. The rationale for any management strategy is to identify with a high degree of safety at reasonable costs, patients at risk of deterioration due to serious intracranial injuries. Traditionally, management has been based on admission for inhospital observation. Recently, computed tomography (CT) imaging for patients with mild head injury has increased. This investigation has often been added to inhospital observation, raising the question whether it would be possible to use CT to triage patients for admission. Thereby, fewer patients would be admitted; those with complications would be detected sooner; and resources would be put to better use in emergency care.

Aim: To compare CT imaging in triage for admission with inhospital observation - in terms of patient outcome, safety, costs, and feasibility - in managing patients with mild head injury.

Methods: In an initial study, the current clinical practise for acute care of patients with mild head injury in Sweden was surveyed. A second study examined the mortality and frequency of complications in patients with mild head injury, in a systematic literature review including a meta-analysis of the findings. The third study assessed the costs of the two strategies of care, by means of a systematic literature review and a decision analysis based on the findings of the meta-analysis. Finally, a pragmatic, multicenter randomised controlled trial (RCT) was conducted to assess patient outcome, and the feasibility and safety of the two strategies.

Results: Patients with mild head injury are routinely admitted for inhospital observation in Sweden, with about 20% also undergoing CT imaging. The literature review identified and critically assessed representative series including 24 000 patients. Based on a metaanalysis of their results, it was estimated that: of 1 000 patients arriving at hospital with mild head injury, 1 will die, 9 will require surgery or other intervention and 80 will have pathological findings on CT. In the decision analysis and literature review it was found that the costs of the CT-strategy was one third lower and might save inhospital bed days for other patients. The RCT randomised 2 602 patients to either CT imaging or inhospital observation. Patient outcome at 3-months was not inferior for the CT strategy compared to inhospital observation, None of the patients with normal findings on the acute CT scan later suffered complications.

Conclusions: The results of the three initial studies were confirmed in the large RCT. Thus, the use of CT imaging to triage for admission is a safe and feasible management strategy for patients with mild head injury. The CT strategy reduces costs by one third, and is not inferior for patient outcome compared to inhospital observation.