Comparing costs of non-invasive cardiac diagnostic tests – a population-based study

Introduction — Non-invasive cardiac diagnostic tests for the diagnosis of coronary artery disease (NIT) have been estimated to cost > $3 billion annually in the United States alone and have undergone recent scrutiny over concerns of overuse. Consequently, comparing costs of different NIT testing strategies is of urgent importance to healthcare planning.

Methods — We utilized population-based administrative and clinical data from Ontario, Canada to compare downstream costs between four available NIT testing strategies (exercise stress testing (GXT), stress echocardiography, cardiac computed tomography angiography (CCTA) and myocardial perfusion imaging (MPI) as well as no-testing), amongst patients evaluated for chest pain. To compare costs among the tested (overall and by testing strategy) and non-tested groups, we used a log-gamma generalized linear model to account for the skewed distribution of healthcare cost data, adjusting for relevant clinical covariates.

Results — 2,340,699 patients were included in our cohort, of whom 481,170 (21%) patients received one of the four NITs. Among patients who received a NIT, 254,492 (53%) received a GXT as their initial test, 154,137 (32%) received MPI, 69,160 (14%) received a stress echo and 3,381 (<1%) received a CCTA. After adjustment for differences in baseline patient characteristics, receipt of any NIT was associated with an approximate 12% reduction in downstream 1-year mean costs (cost ratio 0.88, 95%CI 0.87, 0.89) compared to those without any testing. Comparing the different testing strategies with no testing, both GXT (cost ratio 0.80, 95%CI 0.79-0.81) and stress echocardiography (cost ratio 0.82, 95%CI 0.81-0.83) were associated with the lower downstream costs, while both MPI (cost ratio 1.26, 95% CI 1.25, 1.27) and CCTA (cost ratio 1.29, 95% CI 1.23, 1.35) were associated with higher downstream costs.

Conclusions — In a large population-based cohort consisting of >2 million people evaluated for chest pain, we report that receipt of non-invasive testing was associated with a 12% reduction in downstream costs when compared to no testing. GXT and stress echocardiography were associated with the least downstream costs, whereas CCTA and MPI were associated with higher costs when compared to no testing. These findings may help inform testing decisions in chest pain patients.