Consequences of Preoperative Cardiac Stress Testing-A Cohort Study

Objective: To understand the consequences of functional cardiac stress testing among patients considering noncardiac nonophthalmologic surgery.

Design: A retrospective cohort study of 118,552 patients who made 159,795 visits to a dedicated preoperative risk assessment and optimization clinic between 2008 and 2018.

Setting: A large integrated health system.

Patients: Patients who visited a dedicated preoperative risk assessment and optimization clinic before noncardiac nonophthalmologic surgery.

Measurements: To assess changes to care delivered, we measured the probability of completing additional cardiac testing, cardiac surgery, or noncardiac surgery. To assess outcomes, we measured time-to-mortality and total one-year mortality.

Main Results: In causal inference models, preoperative stress testing was associated with increased likelihood of coronary angiography (relative risk: 8.6, 95% CI 6.1-12.1), increased likelihood of percutaneous coronary intervention (RR: 4.1, 95% CI: 1.8-9.2), increased likelihood of cardiac surgery (RR: 6.8, 95% CI 4.9-9.4), decreased likelihood of noncardiac surgery (RR: 0.77, 95% CI 0.75-0.79), and delayed noncardiac surgery for patients completing noncardiac surgery (mean 28.3 days, 95% CI: 23.1-33.6). The base rate of downstream cardiac testing was low, and absolute risk increases were small. Stress testing was associated with higher mortality in unadjusted analysis but was not associated with mortality in causal inference analyses.

Conclusions: Preoperative cardiac stress testing likely induces coronary angiography and cardiac interventions while decreasing use of noncardiac surgery and delaying surgery for patients who ultimately proceed to noncardiac surgery. Despite changes to processes of care, our results do not support a causal relationship between stress testing and postoperative mortality. Analyses of care cascades should consider care that is avoided or substituted in addition to care that is induced.