Evaluation of the impact of the Canadian subarachnoid haemorrhage clinical decision rules on British practice.

INTRODUCTION

Acute headache is among the commonest presenting complaints to emergency departments. While it is estimated that only 1-3% result from subarachnoid haemorrhage (SAH), because the disease carries such significant morbidity and mortality if missed, most clinicians have a low threshold for investigation. A recent prospective cohort study in Canada determined a number of high-risk clinical characteristics for SAH in patients with acute headache. We investigated the potential impact of incorporating the Canadian clinical decision rules on British practice.

METHODS

A retrospective case note review on all adult patients presenting to our emergency department with acute headache between August and October 2011 was conducted. The Canadian decision rules for SAH were applied retrospectively to the cases identified, and the sensitivity, specificity and negative predictive values calculated. The two-tailed McNemar test was used to evaluate differences between proportions of patients undergoing investigations using the clinical decision rules against current practice.

RESULTS

In all, 112 patients met the inclusion criteria in a 3-month period, of which 41 patients (36.6%) underwent unenhanced computed tomography and 4 (3.6%) were found to have SAH. Nine patients subsequently had a lumbar puncture and none demonstrated xanthochromia. None of the patients who were not fully investigated were readmitted to the regional neurosurgical centre within 6 months of discharge with missed SAH. Application of the Canadian clinical decision rules would have led to an investigation rate between 59% and 74%, compared to an actual rate of 37% (p < 0.05).

CONCLUSION

The present study shows that application of the Canadian clinical decision rules for SAH would lead to more patients with acute headache being investigated than current British practice. However, much larger prospective studies are required to determine whether such clinical decision rules may identify patients at risk who would otherwise have been missed.