Nomogram to predict aneurysmal origin of intracranial subarachnoid hemorrhage: suggesting a user-friendly tool for clinical decision making

BACKGROUND: While aneurysm rupture is the most common etiology of intracranial subarachnoid hemorrhage (aSAH), up to about 20% of cases are reported to be of non-aneurysmal origin or angiogram negative (anSAH). There currently is lack of practical prediction systems to aid in clinical decision making after initial negative angiogram to optimize patient disposition and clinical management. Our objective was to develop a predictive nomogram tool incorporating significant clinical and radiographic predictors of aSAH.METHODS: Demographic, clinical, and radiographic data from sequential non-traumatic SAH patients from institutional database were retrospectively collected and analyzed. Univariate and multivariable statistical analyses were performed yielding significant SAH origin predictors that were incorporated into a predictive nomogram.RESULTS: A total of 167 SAH subjects were identified and randomly split into the training and validation cohorts. Five models were generated to predict aneurysmal origin of SAH. The model including Hunt and Hess grade, modified Fisher grade, and hemorrhage presence/absence in the prepontine cistern, suprasellar cistern, interhemispheric fissure and foramen magnum demonstrated the highest predictive value (area under the curve [AUC] = 0.843, pseudo-R2=0.222). A corresponding graphical nomogram was developed with C-Index of 0.843.CONCLUSIONS: Our nomogram tool incorporates easily identifiable factors assessing the risk of aneurysmal origin of SAH upon patient’s initial presentation. The nomogram can be easily implemented in large, multicenter studies as a point-of-care setting tool aiding in clinical decision making in cases of anSAH.