A method that uses common odds ratios to predict risk of abnormality from symptoms could help doctors to provide more appropriate care for minor head injury.

Pooled odds ratios from published case-control studies of patients with minor head injury (Glasgow coma scale (GCS) 15) with loss of consciousness (LOC) or post traumatic amnesia (PTA) indicated that just three symptoms predicted abnormality on computed tomography. They were vomiting (odds ratio 4.398; 95% confidence interval 2.790 to 6.932), nausea (2.125; 1.467 to 3.057), and severe headache (3.211; 2.212 to 4.584); blurred vision and dizziness were not significant. This finding enabled patients to be subgrouped as: GCS 15a very small risk (no symptoms); GCS 15b low risk (mild headache/blurred vision/dizziness); GCS 15c intermediate risk (LOC/PTA); GCS 15d high risk (severe headache/nausea/vomiting).

The authors carried out a meta-analysis of case-control or nested case-control studies to calculate common odds ratios for five symptoms and abnormality on computed tomography. They searched Medline and PubMed between January 1990 and February 2001 and hand searched several journals. They included only full papers of studies of adults or adults and children with LOC/PTA, all of whom had been examined by computed tomography, whose data allowed common odds ratios to be calculated. Just three papers, covering 3375 patients, met the study criteria.

Minor head injury has most recently been classified as GCS 15, but it still needs to be separated from other injuries within this group. Individual studies that have tried to link symptoms to abnormality on computed tomography are nearly all flawed in some way.