Cost-effectiveness of angiography performed during surgery for ruptured intracranial aneurysms.

PURPOSE To calculate the incremental cost-utility ratio for routine angiography performed during surgery for ruptured cerebral aneurysms.

METHODS Decision-tree and Markov analyses based on a cohort simulation were used to determine the incremental cost-utility ratio of routine intraoperative angiography versus no angiography. Input data from the literature were estimated for the following variables: frequency of unexpected aneurysmal rests and branch artery occlusions; annual rate of rehemorrhage of partially clipped aneurysms; prevalence of clinically relevant infarction resulting from branch artery occlusion; efficacy of clip repositioning; morbidity associated with intraoperative angiography; morbidity and mortality associated with aneurysmal rehemorrhage; sensitivity of intraoperative angiography for aneurysmal rests; and costs of intraoperative angiography, added duration of surgery, ischemic cerebral infarction, aneurysmal rehemorrhage, and rehabilitation. Sensitivity analyses were performed for all relevant input variables. A societal perspective was used, and cost-utility ratios less than $50000/quality-adjusted life years (QALY) gained were considered acceptable.

RESULTS Baseline input variables resulted in an acceptable cost-utility ratio for routine intraoperative angiography ($19000/QALY). The input variables with greatest influence on the cost-utility ratio were frequency of branch artery occlusions, angiographic morbidity, and cost of angiography. However, the cost-utility ratio remained acceptable even over wide ranges of these input variables. Frequency of unexpected partially clipped aneurysms, efficacy of clip repositioning, and costs of stroke, rehemorrhage, and rehabilitation had relatively little impact on the analysis.

CONCLUSION Routine intraoperative angiography is cost-effective if performed in a manner consistent with low morbidity in a patient cohort harboring at least some unexpected branch artery occlusions that, if uncorrected, would result in clinically relevant cerebral infarctions.