Clinical Predictors of Subarachnoid Hemorrhage in Patients with Acute Non-traumatic Headache in Emergency Department.
- Author:
Seungmin JEONG
1
;
Yu Jin KIM
;
Kyuseok KIM
;
Sang Do SHIN
;
Joonghee KIM
;
Jin Hee JEONG
;
Chang Woo KANG
;
Soo Hoon LEE
;
Chan Jong PARK
Author Information
1. Department of Emergency Medicine, Seoul National University Bundang Hospital, Gyeonggi-do, Korea. myda02@snubh.org
- Publication Type:Original Article
- Keywords:
Subarachnoid hemorrhage;
Headache;
Clinical prediction rule
- MeSH:
Blood Pressure;
Cerebrospinal Fluid;
Decision Support Techniques;
Emergency Service, Hospital*;
Headache*;
Humans;
Neck;
Neck Pain;
Neuroimaging;
Respiratory Rate;
Retrospective Studies;
ROC Curve;
Subarachnoid Hemorrhage*;
Vomiting
- From:Journal of the Korean Society of Emergency Medicine
2014;25(6):703-714
- CountryRepublic of Korea
- Language:Korean
-
Abstract:
PURPOSE: Missing subarachnoid hemorrhage (SAH) can cause catastrophic results. We aimed to find clinical factors for predicting SAH in neurologically intact patients with acute non-traumatic headache visiting the emergency department (ED). METHODS: This was a retrospective chart review study. Data were collected from September 2006 until October 2011. We included patients aged over 16 with acute non-traumatic headache who had brain imaging work up results during ED visits. Information on candidate clinical predictor variables was obtained from previous reports, and the outcome was confirmed SAH in brain imaging work up or cerebrospinal fluid study. We found the predictors for SAH through multivariable analysis with variables chosen in univariable analysis considering clinical application. Then we simulated possible SAH prediction scoring models using receiver operating characteristic (ROC) analysis and assessed model fit through the Hosmer-Lemeshow test. RESULTS: A total of 3294 patients were enrolled. Seven clinical characteristics were proven for relation of SAH; age, visiting emergency department within six hours from symptom onset time, visiting mode, vomiting, neck pain or neck stiffness, blood pressure, and respiratory rate. We constructed six available SAH prediction scoring models. The area under the ROC curves of each model ranged from 0.810 to 0.834 and all simulated models were good-fit. With these models, we can expect to reduce unnecessary computed tomography use. CONCLUSION: Seven clinical predictors could be helpful in selection of high risk patients of SAH. The proposed SAH prediction models using these characteristics will have to be tested prospectively for external validation.
