Incidence of Ventricular Shunt Placement for Hydrocephalus with Clipping versus Coiling for Ruptured and Unruptured Cerebral Aneurysms in the Nationwide Inpatient Sample Database: 2002 to 2007

BACKGROUND: Few studies have compared the incidence of ventricular shunt placement for hydrocephalus after clipping versus coiling of cerebral aneurysms. OBJECTIVE: The Nationwide Inpatient Sample (NIS) database was used to compare, on a national level, the incidence of ventricular shunt placement after clipping versus coiling of ruptured and unruptured aneurysms. METHODS: Hospitalizations for clipping and coiling of ruptured and unruptured aneurysms from 2002 to 2007 were collected from the NIS by cross-matching International Classification of Diseases-9 codes for diagnoses of subarachnoid hemorrhage or unruptured cerebral aneurysm with procedure codes for clipping or coiling. The incidence of ventricular shunt placement for hydrocephalus after clipping and coiling was compared using generalized linear models with generalized estimating equations (GEE) to adjust for patient-and hospital-specific factors and correlation between admissions. RESULTS: Of 10,899 ruptured aneurysm patients (6593 clipping, 4306 coiling), clipping had a similar incidence of ventricular shunt placement (9.3%) compared to coiling (10.5%) (odds ratio = 0.984; 95% confidence interval = 0.85, -1.14; P value = 0.833 after adjustment for patient-specific and hospital-specific factors). Likewise, of 9686 unruptured aneurysm patients (4483 clipping, 5203 coiling), clipping had similar incidence of ventricular shunt placement (0.4%) compared to coiling (0.5%) (odds ratio = 0.763; 95% confidence interval = 0.37, -1.58; P value =0.465 after adjustment for patient-specific and hospital-specific factors). Predictors of shunt placement in ruptured aneurysm patients were age, comorbidity score, admission type, payer, and hospital aneurysm volume. Predictors of shunt placement in unruptured aneurysm patients were comorbidity score and admission type. CONCLUSIONS: In an observational study, clipping and coiling of ruptured and unruptured cerebral aneurysms are associated with similar incidences of ventricular shunt placement for hydrocephalus.