• Vijay M Ravindra, Adam de Havenon, Timothy C Gooldy, Jonathan Scoville, Jian Guan, William T Couldwell, Philipp Taussky, Joel D MacDonald, Richard H Schmidt, and Min S Park.
• Departments of1Neurosurgery and.
• J. Neurosurg. 2018 Jul 1; 129 (1): 100-106.

AbstractOBJECTIVE The purpose of this study was to compare the unruptured intracranial aneurysm treatment score (UIATS) recommendations with the real-world experience in a quaternary academic medical center with a high volume of patients with unruptured intracranial aneurysms (UIAs). METHODS All patients with UIAs evaluated during a 3-year period were included. All factors included in the UIATS were abstracted, and patients were scored using the UIATS. Patients were categorized in a contingency table assessing UIATS recommendation versus real-world treatment decision. The authors calculated the percentage of misclassification, sensitivity, specificity, and area under the receiver operating characteristic (ROC) curve. RESULTS A total of 221 consecutive patients with UIAs met the inclusion criteria: 69 (31%) patients underwent treatment and 152 (69%) did not. Fifty-nine (27%) patients had a UIATS between -2 and 2, which does not offer a treatment recommendation, leaving 162 (73%) patients with a UIATS treatment recommendation. The UIATS was significantly associated with treatment (p < 0.001); however, the sensitivity, specificity, and percentage of misclassification were 49%, 80%, and 28%, respectively. Notably, 51% of patients for whom treatment would be recommended by the UIATS did not undergo treatment in the real-world cohort and 20% of patients for whom conservative management would be recommended by UIATS had intervention. The area under the ROC curve was 0.646. CONCLUSIONS Compared with the authors' experience, the UIATS recommended overtreatment of UIAs. Although the UIATS could be used as a screening tool, individualized treatment recommendations based on consultation with a cerebrovascular specialist are necessary. Further validation with longitudinal data on rupture rates of UIAs is needed before widespread use.

23 keywords...

hide…