Cost-Effectiveness of Diagnostic Testing for Coronary Artery Disease
June 1, 1999 6 minutes read
Cost-Effectiveness of Diagnostic Testing for Coronary Artery Disease
abstracts & commentary
Synopsis: For most patients with typical or atypical angina (but not nonspecific chest pain), a noninvasive diagnostic test is a reasonable use of health care resources.
Sources: Kuntz KM, et al. Ann Intern Med 1999; 130:709-718; Garber AM, Solomon NA. Ann Intern Med 1999;130:719-728.
Two recent studies use highly sophisticated cost-effectiveness analysis and Markov decision analysis modeling to assess the cost-effectiveness of a variety of diagnostic techniques for coronary artery disease (CAD). Interpretation of these papers is aided by some familiarity with the concepts of quality-adjusted life-years (QALY) and decision-making analysis. The first publication, by Kuntz and associates, examines the existing data in this field and created decision-analytic models from the perspective of health care policy, which takes into account diagnostic test accuracy and cost, possibility of significant coronary disease, and the cost of subsequent treatment as well as projected clinical events. Long-term risks were estimated based on the need for angiography and revascularization procedures. Health-related quality of life was estimated for a variety of diagnostic strategies. Chest pain patient cohorts were stratified by typical, atypical, or nonspecific chest pain; gender; and age. Kuntz et al, as well as Garber and Solomon, emphasize that the preferred diagnostic strategy will in part depend on how much society is willing to spend for additional QALYs. This study assessed no test, routine exercise testing, exercise echocardiography, and exercise single-photon emission computed tomography (SPECT) in cohorts of individuals with mild to severe chest pain of various ages and sexes with typical and atypical features of angina.
The only group that had a reasonable cost-effectiveness ratio for direct coronary angiography without stress testing were men with a high likelihood of having CAD, such as middle age or older males with classic angina pectoris. In individuals with atypical angina, stress echo was considerably more expensive per QALY then exercise electrocardiography. Cost-effectiveness ratios were high for all test strategies in individuals with a low to moderate pretest probability of CAD, such as women and younger men with nonspecific chest pain. Exercise echocardiography had " a reasonable cost-effectiveness ratio for patients at moderate risk of coronary artery disease." The discussion emphasizes that individual institutions may have differing quality performance among their noninvasive testing modalities. Although this data analysis was based on extensive literature review, some hospitals will have higher quality (increased specificity and sensitivity) with nuclear techniques as opposed to echo and vice versa.
Kuntz et al emphasize that there is a trade-off between improved diagnostic performance of a test and the burden on health care budgets. Thus, only in subgroups with a high probability of CAD was direct angiography cost effective. In subjects with a low probability of coronary disease, all testing strategies were expensive. Exercise echo appeared to be most reasonable for individuals with moderate risk for CAD, with a cost-effective use of resource ratio comparable to generally accepted medical procedures, such as bypass surgery or cholesterol lowering in a man with severe hyperlipidemia. They also conclude that "for most patients with typical or atypical angina (but not nonspecific chest pain), a noninvasive diagnostic test is a reasonable use of health care resources."
The companion article by Garber and Solomon assessed the use of exercise ECG, planar thallium, stress echo, SPECT nuclear testing, and positron emission tomography (PET) in a decision analysis model that sought to assess long-term health care outcomes and costs developing from different strategies. All five strategies were compared to an initial approach of coronary angiography. The analysis was directed at individuals with symptoms and risk factors placing them at intermediate (25-75%) risk or pretest probability of having CAD, such as middle age or older women with typical angina as well as middle age or older men with atypical angina. As in the article by Kuntz et al, sophisticated data analyses were carried out regarding health outcome measures, such as life expectancy and QALY. A management algorithm testing the various strategies was used, and costs were based on projected long-term outcomes. The study concluded that PET was highly sensitive (comparable to SPECT) for significant CAD, but its cost-effectiveness ratio was too high to be recommended. All of the imaging techniques performed better than routine stress electrocardiograms with respect to not missing individuals with severe CAD. Garber and Solomon point out that the testing strategies were all relatively similar with respect to QALY outcomes but that variations in test sensitivity clearly can result in variation in health outcomes.
Total cost of care in this model varied little among the alternative testing strategies. As in the prior report, exercise echo was the least expensive option for men and women of middle age without a high likelihood of CAD. Echocardiography performed better than routine exercise testing and was comparable to nuclear imaging, although the cost of stress echo is greater than routine exercise testing. PET had good outcomes, but at much greater cost than stress echo or SPECT, and was not as cost effective as immediate angiography. Garber and Solomon point out that if high cost-effectiveness ratios were acceptable, with expenses considerably greater than those generally considered to be cost effective, SPECT, angiography, and PET imaging would be reasonable choices. Doing no test at all was felt to be a poor choice, particularly since stress echo has a relatively low QALY cost. As in the previous paper, direct or initial angiography was cost-effective only in men with a high pretest likelihood of CAD. Stress echo appeared to be the most cost-effective strategy for medium- and low-risk individuals. Garber and Solomon conclude that stress "echo, SPECT, and immediate angiography are the most appropriate diagnostic tests for patients at intermediate pretest risk," with exercise testing and planar thallium resulting in poorer outcomes and greater overall costs. In the discussion of exercise echo vs. SPECT, it is emphasized that this should be an institution-dependent decision, and that physicians consider local cost and quality in using these tests.
Prognostic information from noninvasive testing was not considered in either analysis. Garber and Solomon believe that individual physicians should take into account such prognostic factors (such as CAD risk status, severity of angiographic CAD) in their choice of test selection; noninvasive testing may be favored over immediate angiography because of its ability to provide quantitative prognostic information about the future. Garber and Solomon conclude that while all of the noninvasive tests are highly sensitive for detection of severe disease (three vessel and left main), stress echo and SPECT are the most cost-effective options for general use.
Comment by Jonathan Abrams, MD
These two articles require time and patience for the reader; a background in the literature of cost-effectiveness and decision-making analysis is helpful. Nevertheless, the uninitiated reader can readily come away with a similar "take home" message from both studies: all available noninvasive diagnostic techniques operate with an acceptable sensitivity and specificity (except for exercise stress testing, which has a lower sensitivity than the others). The choice of a noninvasive test for a middle-aged or older male with classic angina is unimportant; direct angiography is clearly appropriate and cost effective in many such individuals. However, the large proportion of patients with chest pain evaluated by physicians have a moderate to low pretest probability of CAD, and these are the individuals who use the greatest amount of health care resources in efforts to establish a diagnosis as to whether CAD is present. Routine stress electrocardiography can be appropriately applied in many individuals, but its performance is not as high with respect to specificity as well as QALY adjusted cost-effectiveness as SPECT or stress echo. Thus, one is confronted with a tradeoff of the higher initial expense of these procedures vs. reasonable long-term health outcomes and cost. Stress echocardiography is clearly the winner in these analyses. This modality is not necessarily available in all institutions and cardiology practices; an appropriate alternative, based on the analysis of both studies, is SPECT with thallium and/or sestamibi. For those interested in health care policy, cost-effectiveness issues, as well as a broad overview of the noninvasive literature, careful reading of these publications is recommended. Both analyses conclude that stress echo and SPECT are the preferred noninvasive procedures for individuals at moderate risk for CAD. No specific test is particularly cost-effective in low-probability subjects.
A strategy of immediate coronary angiography was cost-effective for which patient group?
a. Women with intermediate pretest likelihood of CAD
b. Patients with low probability of CAD
c. Men older than age 55 with classic angina
d. Patients with atypical chest pain
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content