Introduction: Comparison of real world clinical outcomes between clipping and coiling for ruptured aneurysm is challenging, mainly due to confounding case severity and variation in physician expertise. We sought to compare outcomes and costs of coiling and clipping by utilizing a large national database.
Methods: Ruptured intracranial aneurysm patients age 18 to 86 years were identified from the National Inpatient Sample (NIS) from 2007 to 2013. The clinical outcomes evaluated were in-hospital mortality, in-hospital stroke, discharge destination and other complications. Also, the total hospitalization cost was compared. Multivariate logistic regression and propensity-score matching were used. Confounding factors such as specific neurovascular comorbidities (e.g. hypertension, diabetes), Charleson comorbidity index (CCI), all patient refined diagnosis related group (APR-DRG) risk of mortality, APR-DRG severity index, and NIS-SAH severity score were also controlled for.
Results: 14379 patients with ruptured intracranial arterial aneurysms were identified (5966 clipping, 8413 coiling). Univariate analysis showed that coiling patients had higher preprocedural risks by CCI (p < 0.001) and APR-DRG mortality (p<0.001). Multivariate analysis showed no difference in in-house mortality (OR 0.96, p = 0.60), but coiling showed lower in-house strokes (OR 0.67, p < 0.001) and more discharge to a rehabilitation facility (OR 0.63, p<0.001). Coiling patients had higher odds of a hematoma (OR 1.44, p = 0.09) and hydrocephalus requiring treatment (OR 1.55, p<0.01), but lower odds of requiring a blood transfusion (OR 0.52, p<0.01). Analysis over time showed that in-house mortality has decreased over time (p=0.01) in coiling patients, but mortality after clipping has not changed. Coiling was slightly more expensive, but the difference was not significant ($2267, p=0.13).
Conclusions: After adjusting for severity, there was no significant difference in mortality between coiling and clipping in ruptured aneurysm treatment. However, coiling patients had a lower chance of an in-hospital stroke and a higher chance of being discharged home. A limitation of the study is the possibility of unmeasured confounders.