HOME SUBSCRIPTIONS CURRENT ISSUE PAST ISSUES CARDIOSOURCE SEARCH HELP FEEDBACK		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Full Text Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Pasternak, R. C. Articles by Houston-Miller, N. Search for Related Content PubMed PubMed Citation Articles by Pasternak, R. C. Articles by Houston-Miller, N.

Task force #1—identification of coronary heart disease risk: is there a detection gap?

View larger version (43K): [in a new window]   Figure 1 Data from the Framingham Heart Study experience. Much of the middle-aged population has a low to intermediate risk for hard CHD (myocardial infarction or CHD death). Even up to age 80 years, more than three-quarters of women experience a 10-year risk of CHD that falls below 10%. The risks are higher for men, and by age 60 the majority of men are at intermediate (10% to 20% per 10 years) or high risk (greater than 20% per 10 years) for CHD. Figure courtesy Peter W. F. Wilson, MD, Framingham Heart Study (unpublished data).

View larger version (19K): [in a new window]   Figure 2 Influence of disease prevalence on predictive accuracy of a typical diagnostic strategy. When used in a population with low disease prevalence, such as an asymptomatic population undergoing cardiovascular screening, even a test with excellent sensitivity (Sens) and specificity (Spec) will yield a poor positive predictive accuracy owing to a larger number of false-positive (FP) than true-positive (TP) results. CAD = coronary artery disease. Reprinted with permission from Gheorghiade and Bonow (43).

View larger version (40K): [in a new window]   Figure 3 (A) The event rate (cardiac death [CD] or nonfatal myocardial infarction [MI] percent per year) is depicted on the y-axis as a function of test result, on the x-axis (DTS = Duke treadmill score as low, intermediate, or high risk score; single-photon emission computed tomography [SPECT] imaging as normal, mildly abnormal, or severely abnormal). For both tests, event rates with a low risk (DTS) or normal (SPECT) result are low: 0.9% per year for DTS, 0.4% per year for SPECT. (B) The data from Hachamovitch et al. (19) are reconfigured to demonstrate the proportion of all patients who had events (y-axis) for each given test result (x-axis). The majority of patients experiencing events (approximately 85%) had a low- or intermediate-risk DTS, whereas a smaller though not inconsiderable proportion of events occurred in patients with a normal or mildly abnormal SPECT (approximately 50%).

View larger version (16K): [in a new window]   Figure 4 Comparison of the magnitude of relative risk of future cardiovascular events associated with abnormal values of different risk factors or combination of risk factors. The data are derived from initially healthy women in the Nurses’ Health Study. In each case, relative risk is shown (black box) for individuals in the top versus the bottom quartile for each factor, 95% confidence intervals are shown by the horizontal lines. (Ridker [30]).

View larger version (43K): [in a new window]   Figure 5 Receiver operating characteristic (ROC) curves for screening guidelines including the first and second National Cholesterol Education Programs (NCEP I and II) and the Canadian Consensus Conference on Cholesterol (CCCC). The 45° line (broken line) represents a nondiscriminating test where the true-positive rate equals the false-positive rate. The NCEP II guidelines performed the best with an area (± SD) beneath the curve of 0.74 ± 0.03, followed by the NCEP I (0.72 ± 0.03), and the CCCC (0.70 ± 0.03). The NCEP II guidelines also performed significantly (p < 0.03) better than the NCEP I in predicting coronary deaths. The computer risk model had an area of 0.85 ± 0.02 and was a significantly (p < 0.03) better discriminator than any of the expert guidelines (Grover et al. [44]).

View larger version (23K): [in a new window]   Figure 6 In this example, pretest probability is estimated by standard coronary heart disease (CHD) risk factor measurements in a multivariable model, such as the Framingham risk score, to predict a future event (dashed line). The solid and dotted lines represent curves generated, depending on whether the subsequent test result is positive (solid line) or negative (dotted line). The arrows represent how a patient with a 15% pre-test probability would have markedly different post-test probabilities depending on whether the additional noninvasive test was positive or negative. In this example, an individual who undergoes a screening evaluation that suggests a 15% chance of having a CHD event in the next 10 years undergoes a second test, and the 15% prediction is modified upward and becomes more than 35% if the test is positive, or is modified downward and becomes less than 5% if the second test is negative. This figure demonstrates how additional test results can either substantially increase or decrease the probability estimate of a future CHD event by increasing the chance that a positive result is a true positive, or that a negative result is a true negative. (Adapted from Greenland et al. [10]).