Ottawa Rule for SAH

March 26, 2019

Summary by Alex Chen, MD

Spoon Feed
The application of the Ottawa SAH rule in this relatively small sample of patients had a 100% sensitivity for identifying SAH in the ED but a lower 95% confidence interval of 78.2.  The 100% negative predictive value, with a more promising 95% CI of 98.6-100%, may have been driven by the lower prevalence of SAH in this study population compared to a previous validation study.

Why does this matter?
Headaches suck for patients. Headaches suck for providers when you have to LP them on a busy shift to rule out SAH. Wouldn’t it be nice if there was something that could reliably (in a US less-than-2%-miss-rate kind of way) rule out an otherwise costly and time-consuming evaluation for a subset of patients? A few years ago, the Ottawa group came out with a SAH rule and subsequent validation study (Perry et al., 2017) that seemed pretty promising.

“There’s no chance that this is a subarachnoid.” –Person who signed out the LP
This is a single-center, retrospective chart review of patients presenting to the ED with a primary complaint of headache. Patients were excluded if they had trauma in the past week, abnormal neurological exam, change in level of consciousness, established recurrent headaches, or prior history of aneurysm, SAH, hydrocephalus, or neoplasm. They had a total of 913 patients (92 patients excluded for neurological deficit, 751 for onset greater than 14 days or chronic headache). The prevalence of SAH was 1.6% in this study population. 

While there was 100% sensitivity in this study, the 95% CI (78.2-100%) is wider than I would be comfortable with. In addition, while the NPV of 100% (95% CI: 98.9-100%) is great, the prevalence of the disease was lower than the previous validation study. The validation study by Perry et al. had a prevalence of 5.8%, with a 100% sensitivity and a narrower 95% CI (94.6-100%). They also had a NPV of 100% (95%CI-not available). They pooled this together with their previous data to come up with 100% sensitivity (95% CI: 98.4-100%). I think this study adds another data point regarding the utility of the Ottawa SAH rule, but makes me hesitate a bit.

Source
The Ottawa subarachnoid hemorrhage clinical decision rule for classifying emergency department headache patients.  Am J Emerg Med. 2019 Feb 7. pii: S0735-6757(19)30081-6. doi: 10.1016/j.ajem.2019.02.003. [Epub ahead of print]

Open in Read by QxMD

Reviewed by Thomas Davis