[Can subarachnoid haemorrhage be recognised based on clinical characteristics?].

Use of prediction rules may reduce the number of patients with acute headache undergoing a CT scan and lumbar puncture to rule in or rule out subarachnoid haemorrhage from a ruptured aneurysm. A Canadian study group that studied patients admitted to emergency departments from tertiary centres shows preliminary evidence that is theoretically promising. However, before such rules can be implemented they require prospective validation in the same and other health care systems and a cost-effectiveness analysis that balances the risks of missed instances of subarachnoid haemorrhage versus the risk of radiation-induced cancer. External validation particularly holds true for the Dutch health care system, where most of these patients are initially seen by a general practitioner.