Mild head injury - mortality and complication rate: meta-analysis of findings in a systematic literature review

Background. Whether the strategy for care of mild head injury should be in-hospital observation or computed tomography (CT) investigation and home care has been discussed lately. A necessary requirement for guidelines and the design of clinical trials would be knowledge about the risks of the condition. These have not been reliably summarised. The study aims to estimate as accurately as possible the mortality, the complication rates, and the frequency of pathological findings on CT in patients with mild head injury. Methods. Mild head injury was defined as head trauma involving loss of consciousness or amnesia, but where neurological findings on arrival at hospital are normal (GCS 15). Large databases were searched to find relevant scientific literature, and the retrieved studies were critically appraised. Findings were used from all representative patient data sets that met predefined standards for minimum quality. Meta-analysis using the random-effects model was performed on the data collected. Findings. The search yielded 24 studies on 24249 patients fulfilling the requirements. The mean mortality of patients was low, 0.1% (Cl 0.05-0.15). Complications, mostly requiring surgery, occurred in 0.9% (Cl 0.6-1.2) of the cases. In approximately 8% (Cl 6.1-9.5), pathological CT findings, dominated by haemorrhages, were identified in the acute phase. Conclusions. Of 1000 patients arriving at hospital with mild head injury, 1 will die, 9 will require surgery or other intervention, and about 80 will show pathological findings on CT. At least these 8% of patients will probably need in-hospital care.