Increased Preoperative Stress Test Utilization is Not Associated With Reduced Adverse Cardiac Events in Current U.S. Surgical Practice
David Stone*1, Jesse A. Columbo1, Dan Neal2, Cristina Crippen2, Gilbert R. Upchurch2, Sandra Wong3, Thomas Huber4, David Soybel5, George Sarosi2, Richard J. Powell1, Philip P. Goodney1, Salvatore T. Scali4
1Vascular Surgery, Dartmouth-Hitchcock Medical Center, Lebanon, NH; 2Surgery, University of Florida, Gainesville, FL; 3Surgery, Dartmouth Hitchcock Medical Center, Lebanon, NH; 4Vascular Surgery, University of Florida, Gainesville, FL; 5Surgery, White River Junction VA Medical Center, White River Junction, VT
Objective(s): There is persistent practice variation surrounding the optimal utilization of preoperative stress testing to best identify patients at risk of adverse cardiac events. The purpose of this analysis was to measure the frequency of preoperative stress testing and define its association with perioperative cardiac events across elective U.S. surgical practice.
Methods: We used the Vizient clinical database to study patients who underwent one of eight elective major general, vascular, or oncologic surgical procedures from 2015 to 2019. Procedures were selected using a modified Delphi approach. The primary exposure was preoperative stress testing. The primary outcomes included in-hospital myocardial infarction(MI), mortality, and major adverse cardiac events(MACE), which was a composite of MI, congestive heart failure, dysrhythmia, coronary revascularization, and/or mortality. Centers were grouped into quintiles based on rates of preoperative stress test utilization. We then calculated a modified revised cardiac risk index(mRCRI) to define and compare patient comorbidity profiles among center quintiles. We next compared the incidence of MI, mortality, and MACE across stress testing quintiles and procedures.
Results: We identified 185,612 patients across 133 centers who underwent major surgery. The mean age was 61.7(±14.2) years. 47.5%(n=88,164) were female, and 79.4%(n=144,586) were white. The overall prevalence of stress testing was 9.2% across the entire cohort. Notably, stress testing was most common prior to aortoiliac reconstruction(22.0%), and least common before small bowel resection(4.8%). Preoperative stress test use varied from 3.3% at the lowest quintile centers to 17.4% in the highest quintile centers, despite similar mRCRI comorbidity scores across quintiles(mRCRI >1=15.0% vs. 15.8%)(Figure 1). Overall perioperative MACE rates were similar among low and high quintile centers(8.2% versus 9.4%) despite a 5-fold difference in stress testing rates. These findings were consistent across all procedures(Figure 2). Event rates were also similar for MI(0.5% versus 0.5%) and mortality(1.5% versus 1.0%), despite the notable disparities in stress testing. These findings were again consistent across all procedures.
Conclusions: There is substantial variation in preoperative stress test utilization across the U.S. despite similar patient risk profiles. Moreover, increased testing was not associated with a clinically significant reduction in perioperative MACE either overall, or across eight different surgical procedures. These data suggest that stress testing may be used more selectively to optimize cardiac risk stratification prior to elective major surgery and may indicate a substantial opportunity for systematic cost savings through a reduction of unnecessary testing.
Figure 1: Variation in preoperative stress testing across the 133 centers, and the percent of patients with a modified Revised Cardiac Risk Index (mRCRI) >1 at centers in the lowest quintile versus the highest quintile of stress testing, demonstrating that while preoperative stress test use is highly variable, patient risk profiles are not
Figure 2: Risk of overall perioperative MACE for centers in the lowest quintile (blue) versus the highest quintile (orange) of preoperative stress test use, stratified by procedure type.
Back to 2023 Abstracts