4097Validation of seven different exercise treadmill stress testing protocols in a large 25-year stress testing registry

Background While the Bruce protocol has been extensively validated, other modified exercise protocols with less workload burden are commonly used, though their prognostic value is not well established. Purpose We sought to assess whether exercise capacity (or workload achieved in metabolic equivalents of task [METs]) remains predictive of mortality across various exercise stress testing protocols. Methods In a 25-year stress testing registry spanning from 1991 to 2015, we identified 120,705 patients who underwent 7 different standardized symptom-limited exercise stress testing protocols: Bruce, Modified Bruce, Cornell 0%, Cornell 5%, Cornell 10%, Naughton, and modified Naughton. The choice of the protocol was dependent on the supervising exercise physiologist according to purpose of the test and the individual patient. The primary outcome was all-cause mortality. Results Mean age was 53.3±12.5 years and 59% were male. There were 74953 Bruce, 8368 modified Bruce, 2648 Cornell 0%, 9972 Cornell 5%, 20425 Cornell 10% 1226 Naughton, and 3113 modified Naughton individual protocols. A total of 8426 death occurred over 8.7 years of mean follow-up duration. Figure 1 shows that there was an inverse relationship between peak METs achieved and mortality across all 7 protocols. On multivariable analysis, increasing METs remained protective against death [adjusted HR of 0.46; 95% CI (0.44 - 0.48); p<0.001] even after adjusting for the protocol chosen, age, gender, hypertension, diabetes, coronary disease, end-stage renal disease, smoking, and statin use. METS vs. mortality by protocol Conclusion Across 7 different exercise protocols with various workloads, the predicted exercise capacity remained predictive of mortality irrespective of the protocol chosen, patients' demographics and comorbidities. Different testing choices likely represent different estimated functional capacity.