Skip to main content

31-10-2010 | Stroke | Article

Decision rules aid SAH diagnosis in headache patients


Free full text

MedWire News: Decision rules based on the clinical characteristics of patients with acute headache can help to identify those who are likely to have subarachnoid hemorrhage (SAH), shows a large prospective study.

The three decision rules developed by Jeffrey Perry (University of Ottawa, Ontario, Canada) and colleagues identified 100% of patients with SAH from a large sample of patients with acute headache - in other words, it was 100% sensitive.

The decision rules were between 28% and 39% specific, meaning that they had a false-positive rate of up to 72%. But the researchers say that use of the decision rules could reduce rates of computed tomography and/or lumbar puncture in headache patients.

"While the rules should not yet be used to explicitly rule out SAH, they certainly can be considered to help to identify high-risk patients," they write in the British Medical Journal.

The decision rules were based on prospective data on 1999 patients presenting to six Canadian hospitals with nontraumatic headache peaking within 1 hour. Of these, 130 proved to have SAH.

Thirteen variables were associated with SAH, as opposed to other causes of headache, and the team used recursive partitioning to develop three decision rules incorporating eight of these variables.

Rule 1 comprised age over 40 years, complaint of neck pain or stiffness, witnessed loss of consciousness, and onset with exertion. Rule 2 comprised arrival by ambulance, age over 45 years, vomiting at least once, and having diastolic blood pressure over 100 mmHg. Rule 3 comprised arrival by ambulance, systolic blood pressure over 160 mmHg, and age 45-55 years.

All three rules identified all 130 SAH patients from the overall cohort, while specificity was 28%, 37%, and 39% with rules 1, 2, and 3, respectively. The researchers say that use of these rules would have provoked further investigations in 74%, 66%, and 64% of patients, respectively.

In an accompanying editorial, Gabriel Rinkel (University Medical Centre Utrecht, The Netherlands) said: "If validated, these decision rules could save one of every three or four patients from having computed tomography scanning and lumbar puncture, respectively."

However, he highlighted several potential problems with the study, including the possible dilution of the sample by the inclusion of patients with headache peaking within 1 hour, rather than earlier, and the reduced applicability to other healthcare settings caused by including arrival by ambulance in two of the decision rules.

Rinkel concluded: "The results of the study are encouraging, but more work is needed before the decision rules are clinically applicable."

MedWire ( is an independent clinical news service provided by Current Medicine Group, a trading division of Springer Healthcare Limited. © Springer Healthcare Ltd; 2010

By Eleanor McDermid