Relative risk of alternative admission policies for patients with head injuries

During 1975 and 1976, 10 200 patients with head injuries were admitted for observation to sixteen general hospitals serving 2.1 million people. The neurological condition of 60 (0.59%) patients subsequently deteriorated, including 27 (0.26%) who died. The effect of alternative admission policies on outcome and bed occupancy was calculated. If no patients were admitted solely for observation, 35 beds would be released for elective surgery but 30 patients with severe intracranial injuries would be sent home each year. Alternatively, if only patients with a history of loss of consciousness were admitted, 18 beds would be released but 11 patients with severe injuries, including 7 with unrecognised skull fractures, would be sent home. The effect of introducing a standard policy for routine skull radiography has also been determined. If, in England and Wales, post-traumatic skull radiography was restricted to patients with alteration in level of consciousness when examined, with neurological signs, with physical signs (e.g., a palpable fracture), or with a history of loss of consciousness, each year 163 patients with unrecognised skull fractures and potentially lethal intracranial injuries would be sent home and not be admitted for observation. Can the expected annual saving of 3 million pounds on routine skull radiography justify the risk to these patients?