RR11. A Policy of Routine Cardiac Stress Testing Prior to Vascular Surgery Adds Limited Value to Clinical Risk Stratification

Abstract

Objectives: To evaluate utilization of cardiac stress testing (CST) before elective vascular surgery and determine its value in predicting adverse cardiac events. Methods: Within the national Vascular Quality Initiative (VQI), CST use was studied in 26,484 patients undergoing elective carotid endarterectomy (CEA; n = 12,190), endovascular aneurysm repair (EVAR; n = 5255), open abdominal aortic aneurysm (AAA) repair (n = 1,386), suprainguinal (n = 2060), and infrainguinal (n = 5593) bypass between 2010 and 2012. The VQI Cardiac Risk Index was used to stratify patient risk. Hospital-level variation in CST utilization was measured. Propensity matching, receiver-operating characteristic curves, and mixed-effects regression modeling assessed the predictive value of CST for postoperative major adverse cardiac events or 30-day mortality (major adverse cardiac events [MACE-M]). Results: Use of CST varied dramatically (range, 0%-100%) but MACE-M did not significantly differ between high and low utilization centers. These results were confirmed after propensity score and regression modeling. A positive CST statistically improved the predictive value for MACE-M over clinical risk stratification for CEA, and suprainguinal and infrainguinal bypasses, but not for open AAA or EVAR. The number needed to test (NNT) (Figure presented) to prevent MACE-M varied greatly (range, 10-106; Fig). Paradoxically, a normal CST for high-risk EVAR or open AAA patients was associated with higher MACE-M. Conclusions: Among VQI institutions, there is extreme variation in utilization of preoperative CST before vascular surgery. Routine CST does not uniformly improve prediction of MACE-M beyond risk stratification alone and may be falsely reassuring. In an era of cost containment, foregoing an unnecessary CST has the potential to add value to health care delivery by reducing costs and improving resource utilization.