Related Experiment Video
Updated: Mar 15, 2026
Pre-Chiasmatic, Single Injection of Autologous Blood to Induce Experimental Subarachnoid Hemorrhage in a Rat Model
Published on: June 18, 2021
Subarachnoid hemorrhage admissions retrospectively identified using a prediction model
Shane W English1, Lauralyn McIntyre2, Dean Fergusson2
1From the Department of Medicine (Critical Care) (S.W.E., L.M.), Clinical Epidemiology Program (S.W.E, L.M., D.F., A.F., C.v.W.), Ottawa Hospital Research Institute/The Ottawa Hospital; Department of Anesthesia (Critical Care) (A.T., M.C.), Hôpital de L'Enfant-Jésus, Quebec; and Departments of Medical Imaging (M.P.d.S., C.L.), Surgery (Neuro-Surgery) (J.S.), and Medicine (C.v.W, A.F.), The Ottawa Hospital, Canada. senglish@ohri.ca.
A new model accurately identifies hospitalizations for subarachnoid hemorrhage (SAH) using administrative data. This approach offers a reliable method for creating SAH patient cohorts from health records.
Area of Science:
- Health Informatics
- Epidemiology
- Medical Data Analysis
Background:
- Subarachnoid hemorrhage (SAH) is a critical condition requiring accurate identification.
- Existing methods for identifying SAH hospitalizations may lack efficiency or rely on complex data.
- Health administrative data offers a potentially rich, widely available resource for epidemiological studies.
Purpose of the Study:
- To develop and validate a predictive model for primary subarachnoid hemorrhage (SAH) hospitalizations.
- To utilize readily accessible health administrative data for SAH identification.
- To establish a cost-effective and accurate method for cohort creation in SAH research.
Main Methods:
- Chi-square recursive partitioning algorithm applied to administrative data.
- Combined a complete cohort of primary SAH patients with a random sample of control hospitalizations.
- Internal validation of the predictive model was performed.
Main Results:
- The model achieved high accuracy in identifying primary SAH hospitalizations.
- Sensitivity was 96.5% and specificity was 99.8% in the validation cohort.
- The positive likelihood ratio was 483, indicating strong predictive power.
Conclusions:
- Routinely collected health administrative data can effectively identify patients with a high probability of primary SAH.
- The developed algorithm provides a straightforward and accurate method for creating validated primary SAH cohorts.
- This approach may facilitate research on SAH from ruptured aneurysms or arteriovenous malformations.
