Application of conditional probability analysis to the clinical diagnosis of coronary artery disease.

Analysis of multiple noninvasive tests offers the promise of more accurate diagnosis of coronary artery disease, but discordant test responses can occur frequently and, when observed, result in diagnostic uncertainty. Accordingly, 43 patients undergoing diagnostic coronary angiography were evaluated by noninvasive testing and the results subjected to analysis using Bayes' theorem of conditional probability. The procedures used included electrocardiographic stress testing for detection of exercise-induced ST segment depression, cardiokymographic stress testing for detection of exercise-induced precordial dyskinesis, myocardial perfusion scintigraphy for detection of exercise-induced relative regional hypoperfusion, and cardiac fluoroscopy for detection of coronary artery calcification. The probability for coronary artery disease was estimated by Bayes' theorem from each patient's age, sex, and symptom classification, and from the observed test responses. This analysis revealed a significant linear correlation between the predicted probability for coronary artery disease and the observed prevalence of angiographic disease over the entire range of probability from 0 to 100% (P less than 0.001 by linear regression). The 12 patients without angiographic disease had a mean posttest likelihood of only 7.0 +/- 2.6% despite the fact that 13 of the 60 historical and test responses were falsely "positive." In contrast, the mean posttest likelihood was 94.1 +/- 2.8% in the 31 patients with angiographic coronary artery disease, although 45 of the 155 historical and test responses were falsely "negative." In 8 of the 12 normal patients, the final posttest likelihood was under 10% and in 26 of the 31 coronary artery disease patients, it was over 90%. These estimates also correlated well with the pooled clinical judgment of five experienced cardiologists (P less than 0.001 by linear regression). The observed change in probability for disease for each of the 15 different test combinations correlated with their information content predicted according to Shannon's theorem (P less than 0.001 by linear regression). These results support the use of probability analysis in the clinical diagnosis of coronary artery disease and provide a formal basis for comparing the relative diagnostic effectiveness and cost-effectiveness of different test combinations.

Future imperfect: the limitations of clinical prediction models and the limits of clinical prediction.

Stepwise regression procedures are often used to identify a small set of variables that serve as important predictors of clinical outcome and to construct prediction models based on those variables. Several theoretical and practical limitations of this process are discussed and highlighted with a variety of examples from published reports. Wider appreciation of these limitations should encourage the development of more relevant models, and thereby improve the quality of clinical prediction.

Development and validation of a logistic regression-derived algorithm for estimating the incremental probability of coronary artery disease before and after exercise testing.

OBJECTIVES: Our goals were to develop and validate a multivariate algorithm for estimating the incremental probability of the presence of coronary artery disease. BACKGROUND: Multivariate methods, including logistic regression analysis, have been extensively applied to diagnostic exercise testing. However, few previous studies have included both an incremental design and external validation. METHODS: A retrospective collection of clinical, exercise test and catheterization data was performed involving four U.S. referral medical centers. All patients had no prior history of coronary disease and had undergone coronary angiography < or = 3 months after exercise stress testing. An algorithm was developed in one center (590 patients with a 41% prevalence of coronary artery disease) with the use of logistic regression analysis and was validated in the other three centers (1,234 patients, 70% prevalence). The algorithm incorporated pretest variables (age, gender, symptoms, diabetes, cholesterol), exercise electrocardiographic (ECG) variables (mm of ST segment depression, ST slope, peak heart rate, metabolic equivalents [METs], exercise angina) and one thallium variable. Discrimination was measured with receiver operating characteristic curve analysis. Calibration (that is, reliability) was assessed from a comparison of probability estimates and the actual prevalence of disease. RESULTS: The overall incremental receiver operating characteristic curve areas for the validation group were pretest, -0.738 +/- 0.016; postexercise ECG, 0.78 (SE 0.017); and postthallium, 0.82 (SE 0.016); p < 0.01 for both increments. Within the three validation institutions, the institution with a disease prevalence closest to that of the derivation institution had the best incremental receiver operating characteristic curve areas. There was a stepwise incremental improvement in calibration especially from exercise ECG to thallium testing. CONCLUSIONS: An incremental multivariate algorithm derived in one center reliably estimated disease probability in patients from three other centers. The incremental value of testing was best demonstrated when the derivation and validation groups had a similar disease prevalence. This algorithm may be useful in decision making that relates to the diagnosis of coronary disease.

Fee-for-benefit: a strategy to improve the quality of health care and control costs through reimbursement incentives.

OBJECTIVES: The purpose of this study was to determine whether reimbursement in direct proportion to expected therapeutic benefit is capable of improving the utilization and cost of health care. BACKGROUND: The benefit associated with a particular medical or surgical treatment varies widely from patient to patient. Nevertheless, payment to the provider of the treatment is essentially invariant under the current fee-for-service system. Under an alternative fee-for-benefit strategy, empiric data are used to construct a multivariate model to predict the expected benefit to an individual patient from a particular health care service on the basis of conventional clinical descriptors. The payers and the providers of the service then openly negotiate an explicit economic relation between expected benefit and monetary payment such that payment is directly proportional to benefit. METHODS: Computer simulations were performed to determine the potential impact of this fee-for-benefit strategy with respect to medical versus surgical treatment of coronary artery disease. RESULTS: Compared with conventional fee-for-service, fee-for-benefit resulted in a 12% improvement in patient benefit (quality-adjusted survival), a 22% reduction in provider payments and a 55% increase in cost/benefit (the ratio of benefit to payment). CONCLUSIONS: The incentives embodied in a fee-for-benefit strategy can be an effective mechanism for encouraging more appropriate health care utilization while simultaneously controlling health care costs.

Clinician decisions and computers.

The clinician's decisions are subject to numerous distorting influences. Computer decision aids can help avoid these distortions by placing the clinician's limited personal experience into broader perspective through comparison with a larger repository of clinically relevant information; by making explicit the assumptions implied by his or her decisions; and by alerting the clinician whenever the decisions made do not appear consistent with these assumptions, with the available information or with the conventional rules of logic. Practical standards of performance with respect to the development, validation and clinical application of these decision aids are still in evolution, however, and a variety of ethical and legal issues have yet to be addressed. Despite the promise of computer decision aids, it remains to be seen whether their diffusion into medical practice will improve the quality and cost of health care.

Computer-assisted diagnosis in the noninvasive evaluation of patients with suspected coronary artery disease.

A microcomputer program called CADENZA, which employs Bayes' theorem to analyze and report the results of various clinical descriptors and noninvasive tests relative to the diagnosis of coronary artery disease, was evaluated in 1,097 consecutive patients without previous myocardial infarction. With this program, each patient was characterized by a probability for coronary artery disease, based on Framingham risk factor analysis, symptom characterization, electrocardiographic stress testing, cardiokymography, cardiac fluoroscopy, thallium perfusion scintigraphy and technetium equilibrium-gated blood pool scintigraphy. A total of 11,808 probability estimates derived from various combinations of the available observations were analyzed: 2,180 in 170 patients undergoing coronary angiography and 9,628 in 969 patients who completed a 1 year follow-up for coronary events. The predicted probability of disease correlated linearly with observed angiographic prevalence in the 170 patients who subsequently had coronary angiography (prevalence = [0.001 +/- 0.011] + [0.966 +/- 0.019] X probability). The difference between probability and prevalence averaged 3.1%, and the magnitude of this correlation was not affected by the type or amount of data analyzed. The prevalence of multivessel disease in these patients increased as a monotonic function of disease probability. Below a probability of 25%, single vessel disease was slightly more common than multivessel disease. Above a probability of 75%, multivessel disease predominated. In the 969 patients followed up for 1 year from the date of testing, the incidence of cardiac death and nonfatal infarction increased as a cubic function of disease probability (from approximately 0 to 8% per year for each). Above a probability of 90%, however, the standard deviation for predicting these events was wide. These data indicate that Bayes' theorem in general--and CADENZA in particular--is an accurate, clinically applicable means for quantifying the prevalence of angiographic coronary artery disease, the risk of multivessel disease and the incidence of morbid coronary events in the year after testing.

Correspondence of left ventricular ejection fraction determinations from two-dimensional echocardiography, radionuclide angiography and contrast cineangiography.

OBJECTIVES: This study assessed the agreement of left ventricular ejection fraction determinations from two-dimensional echocardiography, radionuclide angiography and contrast cineangiography. BACKGROUND: Previously published reports suggest that two-dimensional echocardiography, radionuclide angiography and contrast cineangiography are equally acceptable methods of assessing left ventricular ejection fraction on the basis of high coefficients of correlation. However, correlation of methods does not necessarily imply agreement. METHODS: In a prospective analysis, 25 consecutive subjects all had two-dimensional echocardiography and radionuclide angiography performed within 10 days of each other in the cardiology department of metropolitan community hospital. A retrospective computer search (Medline) revealed seven studies, using the coefficient of correlation (r), comparing two-dimensional echocardiographic left ventricular ejection fraction (n = 268) with radionuclide angiographic (n = 174) or contrast cineangiographic (n = 119) left ventricular ejection fractions. RESULTS: The eight individual studies (n = 293) comparing two-dimensional echocardiography with either radionuclide angiography or contrast cineangiography exhibited coefficients of correlation ranging from 0.78 to 0.93. Agreement analysis using the method of Bland and Altman was performed by averaging the results obtained from the two techniques and determining how disparate any single ejection fraction was (with 95% confidence limits) from the mean value. Agreement ranged from 23% to 42% around the mean ejection fraction. The average lack of agreement between the two methods for all studies involved was 17%, with an average r value of 0.86. CONCLUSIONS: Left ventricular ejection fraction determinations by means of two-dimensional echocardiography, radionuclide angiography and contrast cineangiography exhibit high correlation and only moderate agreement. High correlation does not always imply high agreement. These results suggest that, when validated by agreement analysis, multiple studies may not be necessary in appropriate clinical situations, potentially reducing costs.

Incremental value of simultaneous assessment of myocardial function and perfusion with technetium-99m sestamibi for prediction of extent of coronary artery disease.

OBJECTIVES: This study assessed the incremental value of technetium-99m myocardial single-photon emission computed tomography (SPECT) and simultaneous first-pass radionuclide angiography, when added to treadmill exercise, for prediction of the extent of coronary artery disease. BACKGROUND: Technetium-99m count statistics permit the simultaneous assessment of myocardial perfusion and function. However, whether this characteristic improves prediction of the extent of coronary artery disease remains unknown. METHODS: We studied 70 consecutive patients who had coronary angiography within 6 months of the scintigraphic study. All patients underwent a symptom-limited treadmill exercise test. Treadmill data were summarized using a previously validated score. Left ventricular ejection fraction and regional wall motion were evaluated from a first-pass radionuclide angiogram acquired at peak treadmill exercise in the anterior view. Perfusion was assessed visually. Extent of angiographic disease was expressed as the presence or absence of multivessel disease (more than two coronary artery territories with > 50% stenosis) and as a score that reflects the location of severe (> 75%) stenosis. RESULTS: Stepwise addition of scintigraphic data (perfusion first, followed by function) to the treadmill score showed significant incremental value for prediction of the angiographic score at each step; exercise ejection fraction alone was the strongest independent predictor. Discriminant accuracy for detection of multivessel disease was also improved by the addition of perfusion information to the treadmill score and addition of regional wall motion analysis to both of them. In this case, ejection fraction failed to show independent value. CONCLUSIONS: The addition of simultaneously performed sestamibi perfusion SPECT and first-pass radionuclide angiography to the treadmill exercise test significantly improved prediction of the extent of coronary artery disease.

Selection of patients for programmed ventricular stimulation: a clinical decision-making model based on multivariate analysis of clinical variables.

OBJECTIVE: This study was conducted to assess the utility of clinical variables in predicting the inducibility of sustained ventricular arrhythmias in a heterogeneous group of patients undergoing programmed ventricular stimulation. METHODS: Variables were considered in a simulated chronologic order to determine the incremental information added by the signal-averaged electrocardiogram (ECG) and left ventricular ejection fraction. All patients undergoing baseline programmed ventricular stimulation for induction of ventricular tachyarrhythmia during a 30-month period were included in the study. Fourteen historical, ECG, signal-averaged ECG and left ventricular wall motion variables were evaluated for their ability in predicting inducibility of a sustained ventricular arrhythmia, a "positive" event, at programmed ventricular stimulation. RESULTS: On univariate analysis of the clinical variables, comparison between patients with positive or negative results showed significant differences in 10 of the 14 clinical variables: major cardiac diagnosis, history of ventricular tachycardia, myocardial infarction by history or ECG, all five signal-averaged ECG variables, left ventricular ejection fraction and presence of left ventricular aneurysm. On multivariate analysis, five independent variables were determined to be important: history of ventricular tachycardia, historical or ECG evidence of myocardial infarction, history of loss of consciousness, filtered QRS duration on the signal-averaged ECG and left ventricular ejection fraction. However, with sequential multivariate analysis, a model based only on historical and conventional ECG data was found to do as well as a model that included signal-averaged ECG and left ventricular ejection fraction data. CONCLUSIONS: Routinely available noninvasive historical, ECG, signal-averaged ECG and left ventricular wall motion variables can be used to accurately predict the outcome of programmed ventricular stimulation. The majority of the predictive power was obtained with the routine model, using only historical and ECG data. The signal-averaged ECG and left ventricular wall motion analysis added no significant incremental information.

A format for integrating the interpretation of exercise ejection fraction and wall motion and its application in identifying equivocal responses.

The conventional interpretation of ejection fraction change with exercise may be limited because it does not consider the rest value, define equivocal responses or integrate wall motion data reproducibly. Thus, a format was developed for combined interpretation of rest and exercise radionuclide ejection fraction and wall motion by reviewing the reported data for the exercise responses of patients without prior myocardial infarction. The ejection fraction data of 202 normal patients and of 259 patients with coronary artery disease were first fitted to beta distributions. The true positive and false positive rates for coronary disease for each combination of rest and exercise ejection fraction were then determined directly from these distributions. A given rest/exercise ejection fraction combination was "normal" if the false positive rate was greater than the true positive rate, or "abnormal" if the true positive rate was greater than the false positive rate, and "equivocal" when the rates were similar (within a 50% confidence interval). This analytic format, which predicted an inverse relation between rest ejection fraction and the change required with exercise, was then validated prospectively in 854 patients without myocardial infarction (557 with and 297 without angiographic coronary artery disease). Using the conventional criterion of an abnormal test result (less than 0.05 absolute rise in ejection fraction with exercise or a wall motion abnormality), sensitivity was 85 +/- 2% and specificity only 42 +/- 3%. The statistical format had a sensitivity of 70 +/- 2% and specificity of 70 +/- 3%, resulting in a twofold increase in information content. This format has at least two advantages over conventional interpretation: 1) it provides an explicit definition of equivocal responses; and 2) it reproducibly integrates discordant ejection fraction and wall motion responses and allows for the combined analysis of other nonscintigraphic observations, such as age and sex.

In vitro validation of three-dimensional intravascular ultrasound for the evaluation of arterial injury after balloon angioplasty.

OBJECTIVES: The hypothesis of this study was that three-dimensional ultrasound imaging would facilitate the evaluation of arterial dissection after balloon angioplasty. BACKGROUND: The presence and extent of arterial dissection occurring at the time of balloon angioplasty may be important predictors of abrupt vessel closure or late restenosis. METHODS: Forty-one human arterial segments obtained after death were imaged in an in vitro system at physiologic pressure (80 to 100 mm Hg) before and after balloon angioplasty. Images were acquired with a 20- to 30-MHz mechanical intravascular ultrasound imaging system (Cardiovascular Imaging Systems) with a constant pullback technique (1 mm/s). Standard 0.5-in. (1.27-cm) video tapes were used for data storage and later playback for analog to digital conversion. Digitized data were reconstructed to three-dimensional images with use of voxel space modeling. The vessels were opened longitudinally and subjected to pathologic examination, photographed and classified histologically as normal, fibrous or calcified. Dissection was defined as a disruption and separation of components of the arterial wall. The length and depth of arterial dissection were evaluated grossly and microscopically. RESULTS: Of the 41 arteries studied, 36 (88%) exhibited dissection on pathologic examination after balloon angioplasty. Three-dimensional reconstruction of intravascular ultrasound images identified dissection in 11 (92%) of 12 normal, 8 (100%) of 8 fibrous and 11 (69%) of 16 calcified arteries. Excellent agreement between ultrasound and pathologic findings was achieved in the evaluation of length and depth of dissection for histologically normal and fibrous arteries (kappa = 0.72 to 1.0). When the vessels were severely calcified, the agreement was not as good (kappa = 0.27 to 0.56), particularly in detection of small, non-raised intimal flaps. CONCLUSIONS: This histopathologic validation study suggests that three-dimensional intravascular ultrasound imaging facilitates the evaluation of both quantitative and morphologic features of arterial dissection induced by balloon angioplasty. The advantage of three-dimensional intravascular ultrasound is its ability to assess the length and morphology of arterial injury over an entire vessel segment.

Incremental prognostic value of exercise thallium-201 myocardial single-photon emission computed tomography late after coronary artery bypass surgery.

OBJECTIVES: This study assessed the incremental prognostic value of exercise thallium-201 myocardial perfusion single-photon emission computed tomography (SPECT) performed > or = 5 years after coronary artery bypass surgery. BACKGROUND: Thallium-201 scintigraphy has shown significant prognostic value in a variety of populations with suspected and known coronary artery disease. However, its value in patients with previous bypass surgery remains unknown. METHODS: We studied 294 patients who were prospectively followed up. Cox proportional hazards models for prediction of "hard" events (cardiac death and nonfatal infarctions) were constructed, with variables considered for inclusion in hierarchic order: clinical and exercise data first, followed by scintigraphic information. RESULTS: Mean (+/- SD) follow-up duration after scintigraphy was 31 +/- 11 months. There were 20 cardiac deaths and 21 nonfatal acute myocardial infarctions. Twenty-nine patients had late (> 60 days after thallium-201 SPECT) revascularization procedures or underwent repeat bypass surgery or percutaneous transluminal angioplasty. Shortness of breath and peak exercise heart rate were the most important clinical predictors of hard events. Two scintigraphic variables added significant prognostic information to the clinical model: the thallium-201 summed reversibility score (summation of segmental differences between stress and redistribution scores) and the presence of increased lung uptake of the radiotracer. The global chi-square statistic for this model was twice as high as that for the clinical/exercise model alone (49.7 vs. 24.2). When a second multivariate Cox model was built adding "soft" events (i.e., late revascularization procedures) as outcomes of interest, the summed reversibility score was selected as an independent scintigraphic predictor of events. The global chi-square statistic for this model was 50.7, three times as high as that for the clinical/exercise model alone. CONCLUSIONS: After evaluation of treadmill and exercise data, thallium-201 myocardial perfusion SPECT provided incremental prognostic information in patients late after bypass.

Gender-related differences in clinical management after exercise nuclear testing.

OBJECTIVES: This study sought to determine the rate of referral to cardiac catheterization in men and women early after nuclear testing as a function of the magnitude of myocardial ischemia by radionuclide perfusion imaging. BACKGROUND: Although many previous studies have suggested that gender-related differences are present in the clinical management of coronary artery disease, the presence of such a difference with respect to referral to catheterization after noninvasive testing is disputed. METHODS: We examined 3,211 consecutive patients (1,074 women, 2,137 men) who underwent exercise dual-isotope single-photon emission computed tomography and had follow-up evaluation performed at least 1 year after nuclear testing (mean [+/- SD] follow-up 19 +/- 5 months) for "hard" events (cardiac death and myocardial infarction) and referral to cardiac catheterization or revascularization within 60 days of nuclear testing. Multiple logistic regression analysis was performed to determine the best predictors of referral to catheterization as well as to examine whether gender itself added further information to this model. RESULTS: Although men were referred to catheterization more frequently than women (10.6% vs 7.1%, p < 0.001) early after exercise nuclear testing, there were no differences in the rate of referral to catheterization or revascularization after stratification by the amount of abnormally perfused myocardium detected by the nuclear scan. Both men and women with normal scan results were infrequently referred to subsequent catheterization. In the setting of severe ischemia, women were referred to catheterization more frequently than men. This higher rate appears to be clinically appropriate because women with severely abnormal scan results had a significantly higher event rate than men (17.5% vs. 6.3%, p < 0.0001). This greater risk in women than in men appeared to be underappreciated because the increased rate of hard events in women with severely abnormal scan results was out of proportion to the smaller increase in their rate of referral to cardiac catheterization. Although gender added information to the multivariate model most predictive of referral to catheterization models when nuclear variables were not included, when nuclear variables were considered, the addition of gender added no further significant information. This finding suggests that adjusting for differences in perfusion scan abnormalities by the use of nuclear testing eliminated the apparent gender-related referral bias. CONCLUSIONS: After controlling for differences in perfusion scan abnormalities, no gender-related referral bias to catheterization was present. In the setting of severe ischemia, women had a greater rate referral to catheterization than men. As a function of risk, both men and women were appropriately referred to catheterization at a low rate when the scan result was normal. However, because women with severe perfusion abnormalities had a greater rate of cardiac death and myocardial infarction then men, women in this high risk subgroup were underreferred to catheterization relative to men. This finding points to the need to better identify women at high cardiac risk.

Incremental value of prognostic testing in patients with known or suspected ischemic heart disease: a basis for optimal utilization of exercise technetium-99m sestamibi myocardial perfusion single-photon emission computed tomography.

OBJECTIVES: This study assessed the incremental prognostic implications of normal and equivocal exercise technetium-99m (Tc-99m) sestamibi single-photon emission computed tomography (SPECT) and sought to determine its incremental prognostic value, impact on patient management and cost implications. BACKGROUND: The prognostic implications of Tc-99m sestamibi SPECT are not well defined, and risk stratification using this test has not been explored. METHODS: We studied 1,702 patients referred for exercise Tc-99m sestamibi SPECT who were followed up for a mean (+/- SD) of 20 +/- 5 months. Patients with previous percutaneous transluminal coronary angioplasty or coronary artery bypass surgery were excluded. The SPECT studies were assessed using semiquantitative visual analysis. Cardiac death and myocardial infarction were considered "hard" events, and coronary angioplasty and bypass surgery > 60 days after testing were considered "soft" events. RESULTS: Of the 1,702 patients studied, 1,131 had normal or equivocal scan results. A total of 10 events occurred in this group (1 cardiac death and 1 myocardial infarction [0.2% hard events]; 4 coronary angioplasty and 4 bypass surgery procedures [0.7% soft events]). The rates of hard events and referral to catheterization after SPECT were similarly low in patients with a low (< 0.15), intermediate (0.15 to 0.85) and high (> 0.85) post-exercise treadmill test (ETT) likelihood of coronary artery disease. With respect to scan type, patients with normal, probably normal or equivocal scan results had similarly low hard event rates. In the 571 patients with abnormal scan results, there were 43 hard events (7.5%) and 42 soft events (7.4%) (p < 0.001 vs. 1,131 patients with normal scan results for both). When the complete spectrum of scan responses was considered, SPECT provided incremental prognostic value in all patient subgroups analyzed. However, the nuclear scan was cost-effective only in patients with interpretable exercise ECG responses and an intermediate to high post-ETT likelihood of coronary artery disease and in those with uninterpretable exercise ECG responses and an intermediate to high pre-ETT likelihood of coronary artery disease. CONCLUSIONS: Normal or equivocal exercise Tc-99m sestamibi study results are associated with a benign prognosis, even in patients with a high likelihood of coronary artery disease. Although incremental prognostic value is added by nuclear testing in all patient subgroups, a testing strategy incorporating nuclear testing proved to be cost-effective only in the groups with an intermediate to high likelihood of coronary artery disease before scanning.

Effective risk stratification using exercise myocardial perfusion SPECT in women: gender-related differences in prognostic nuclear testing.

OBJECTIVES: This study was designed to evaluate the incremental prognostic value over clinical and exercise variables of rest thallium-201/exercise technetium-99m sestamibi single-photon emission computed tomography (SPECT) in women compared with men and to determine whether this test can be used to effectively risk stratify patients of both genders. BACKGROUND: To minimize the previously described gender-related bias in the evaluation of coronary artery disease in women, there is a need to identify a noninvasive testing strategy that is able to accurately and effectively risk stratify women. METHODS: We identified 4,136 consecutive patients (2,742 men, 1,394 women) who underwent dual-isotope SPECT. The incremental value of nuclear testing was determined using both a stepwise Cox proportional hazards model and Kaplan-Meier survival analysis. Receiver operating characteristic curve analysis was performed to determine test discrimination for high risk patients in men and women. RESULTS: The patient population was followed up for 20 +/- 5 months for events (cardiac death or nonfatal myocardial infarction). During this time, 63 myocardial infarctions and 32 cardiac deaths occurred in the men, and 31 myocardial infarctions and 14 cardiac deaths occurred in the women. Nuclear testing significantly stratified both men and women irrespective of their rest electrocardiogram. Cox proportional hazards analysis revealed that nuclear testing added incremental prognostic value in both men and women after inclusion of the most predictive clinical exercise variables (overall chi-square 89 in men vs. 120 in women, p < 0.005). Kaplan-Meier survival analysis demonstrated that nuclear testing further stratified men and women with both intermediate to high and low prescan likelihoods of coronary artery disease (p < 0.005 for all). Receiver operating characteristic curve analysis demonstrated superior discrimination for the nuclear scan results in identifying high risk women than men (area under the curve: 0.84 +/- 0.03 vs 0.71 +/- 0.03 in men, p < 0005). The odds ratio comparing event rates in patients with women than in men, suggesting superior stratification using nuclear testing in women. CONCLUSIONS: Dual-isotope myocardial perfusion imaging yields incremental prognostic value in both men and women. This modality identifies low risk women and men equally well but relatively high risk women more accurately than relatively high risk men and, thus, is able to stratify women more effectively than men.

Extent and severity of myocardial hypoperfusion as predictors of prognosis in patients with suspected coronary artery disease.

The ability of exercise-induced myocardial hypoperfusion on thallium scintigraphy to predict coronary events was assessed in 1,689 patients with symptoms suggestive of coronary artery disease but without prior myocardial infarction or coronary artery bypass surgery. A total of 74 patients had a coronary event in the year after testing (12 cardiac deaths, 20 nonfatal infarctions and 42 referrals for bypass surgery more than 60 days after testing). Stepwise logistic regression identified only three independent predictors: the number of myocardial regions with reversible hypoperfusion (an index of the extent of hypoperfusion), the maximal magnitude of hypoperfusion (an index of the severity of hypoperfusion) and the achieved heart rate (an index of exercise performance). Both extent and severity were exponentially correlated with event rate (r greater than 0.97 and p less than 0.01 for each), whereas achieved heart rate was linearly correlated with event rate (r = 0.79 and p less than 0.05). On the basis of these data, a prognostic model was defined that employs extent and severity as stress-dependent orthogonal variables. Using this model, the predicted coronary event rate ranged over two orders of magnitude--from a low of 0.4% in patients able to exercise adequately without developing severe and extensive hypoperfusion at a low heart rate (less than 85% of their maximal predicted heart rate). Extent and severity of myocardial hypoperfusion, therefore, are important independent variables of prognosis in patients with suspected coronary artery disease.

Should the intent of testing influence its interpretation?

A test is often interpreted as "normal" or "abnormal" by a single criterion, regardless of the intent of testing. The discriminate accuracy of this convention was critically analyzed using information content (I), likelihood ratio and the area under the receiver-operating characteristic curve. Three ejection fraction variables were assessed--ejection fraction at rest, exercise ejection fraction and the change in ejection fraction from rest to exercise--each relative to three intentional goals: diagnosis of coronary artery disease in 929 patients without previous myocardial infarction, prediction of multivessel disease in these same 929 patients and prediction of multivessel disease in 507 patients with previous myocardial infarction. The information content of exercise ejection fraction (IEX) was higher than for ejection fraction at rest (IR) or for the change from rest to exercise (IEX-R), and was relatively constant regardless of the goal of testing. In contrast, neither IR nor IEX-R was constant. IR was lowest for diagnosis of coronary artery disease, whereas IEX-R was highest for this same goal. These empiric observations are consistent with the quantitative relation predicted by information theory: IEX = IR + IEX-R. Thus, ejection fraction at rest has little discriminate value relative to the diagnosis of coronary artery disease, but does have value in evaluating the extent of disease in patients after myocardial infarction. Exercise ejection fraction and change in ejection fraction are nearly equally useful for purposes of diagnosis, whereas the former is most useful for functional evaluation in postinfarction patients.(ABSTRACT TRUNCATED AT 250 WORDS)

Anemia therapy: individual benefit and societal cost.

The US health care system is under increasing pressure to lower costs while maintaining quality of care. Providers will be forced to (1) measure the benefits of a particular therapy, and (2) demonstrate that the benefits justify the costs. The major components of therapeutic benefit are survival and quality of life. Chronic anemia may have little impact on survival, but studies have measured significant decrements in quality of life without therapy and increments in quality of life with therapy. This disease also presents important societal financial concerns due to its many competing therapies. The annual cost of treatment can vary from an average of a few dollars for iron supplementation to an average of $6,000 for a course of recombinant human erythropoietin. Physicians need to integrate information on therapeutic outcomes and cost to maximize individual benefit and justify the costs. The choice of therapy for anemia associated with cancer is complex because the onset of the anemia is multifactorial, and the effects of anemia may be masked by the underlying malignancy. There are insufficient data supporting a specific recommendation for transfusion or recombinant human erythropoietin therapy. The current cost-conscious environment in the United States presents an opportunity for health care providers to formally document the benefits of anemia therapy and justify the societal costs based on those benefits. Anemia is an excellent example of a condition that allows the formal analysis of disease impact and the effectiveness and cost of therapy because (1) it has multiple therapies, (2) the cost of therapy varies widely, and (3) the therapies have variable benefit depending on the individual patient. Using a model based on chronic renal failure, an outcomes structure was developed by which the impact of anemia and the therapies used to manage it can be measured. Its potential application to anemia in patients with cancer is discussed.

Metadiagnosis. An epistemologic model of clinical judgment.

For convenience, clinical findings are often artificially forced into finite pigeonholes such as "positive" or "negative." This convention obscures much of the inherent uncertainty in diagnosis and can result in serious misinterpretation of the significance of certain observations. Unhappy past experience with the poor predictive accuracy of electrocardiographic stress testing in asymptomatic patients is one such example. As an aid to understanding the limitations of clinical diagnosis, an epistemologic model of judgment was developed. According to this model, judgments are viewed as existing on three separate dimensional levels, each of which is rigorously defined and unambiguously quantified. The first dimension expresses diagnostic belief in terms of a numeric probability; the second quantifies the degree of confidence in the probability estimate; and the third defines the information that derives from the probability and confidence. The practical clinical relevance of this conceptual model is illustrated by applying it to a common clinical problem: the interpretation of a "positive" electrocardiographic stress test result in an asymptomatic man. This process--termed herein "metadiagnosis"--provides a new perspective by which the art of diagnosis might be made more accurate and explicit.

Prior restraint: a Bayesian perspective on the optimization of technology utilization for diagnosis of coronary artery disease.

In conclusion, at least 1/3 of patients with suspected coronary artery disease are inappropriately referred for scintigraphic diagnostic testing from a Bayesian such as those described in this report, may be a powerful mechanism for encouraging more appropriate technology utilization while simultaneously controlling costs, and are thereby deserving of a formal prospective demonstration trial. However, since only half the patients currently being tested are referred for diagnostic purposes, analogous strategies must be developed with respect to prognostic and therapeutic evaluation.