Modeling for policy decisions: potential and problems.

Because decision modeling involves the construction of an explicit, mathematically describable structure of the pertinent elements of a clinical problem, the relative effectiveness of alternative approaches to care can be identified; costly procedures become apparent; new technologies can be assessed in relation to the old in terms of effectiveness and costs; the marginal benefit to be achieved by duplicative or alternative practices can be determined. All of this can be accomplished in terms of patient outcome and without the bias and self-interest of which the profession has been accused. Furthermore, if a resource allocation or reimbursement decision made in the name of cost containment eliminates or limits access to effective diagnostic or therapeutic technologies, the impact of that decision on effective care can be explicitly and quantitatively expressed through decision modeling. When such analyses are based on patient outcome, they are difficult to ignore and provide a pivotal point for discussions and eventual compromise.

Perspectives on clinical decision-making: the application of pattern analysis.

It is the thesis of this presentation that more precise diagnoses and prognostications result when clinical information is analyzed as data-sets or patterns rather than collections of discrete data. Two examples are given in support of the thesis: the classification by computer of QRS complexes in the ambulatory electrocardiogram and the prediction of risk for recurrent ventricular tachycardia. The advantages to be gained from pattern analysis are: (1) significant variables are not preselected and the data are, therefore, unbiased, and (2) nuances in clinical patterns become evident when patients are presented as data-sets. It is proposed that competent physicians undoubtedly use pattern analysis in their decision-making and that expert systems designed to simulate physician behavior might be more accurate if based in pattern analysis.

Cost-effective management of patients following myocardial infarction: the impact of ischemia on alternative approaches.

Following uncomplicated myocardial infarction patients are at varying risk for cardiovascular morbidity and mortality. In order to identify and treat high risk patients, various management approaches can be employed. We performed a decision analysis to examine the cost-effectiveness of seven alternative strategies under the assumption that prognosis is affected by both the location of anatomic obstruction and the degree of myocardial ischemia. Strategies included combinations of angiography and two theoretical diagnostic tests capable of detecting ischemia with different degrees of accuracy. The strategy associated with the lowest overall six month mortality initiated testing with the diagnostic test most sensitive for ischemia, slightly better than proceeding initially to angiography. Initial use of a test sensitive for ischemia was also considerably more cost-effective than proceeding directly to angiography. Future analyses evaluating the role of diagnostic tests in coronary artery disease should incorporate the dimension of ischemia.

An editing method for computer-assisted ambulatory ECG review systems.

Editing of computer-assisted ambulatory ECG reviews is critical for accuracy and quantification of the arrhythmias present. This may be time consuming for high arrhythmia content types or tapes with noise interference. A system that combines probit analysis and principle components transformation coupled with maximum likelihood decision theory, through identification of the complexes most subject to classification error in the initial review process and by correcting residual errors automatically, can decrease the number of complexes which need to be overread by a human editor. Probit analysis assigns a statistically derived value, a z value, to complexes classified as normal sinus or ventricular premature complexes. This permits the grouping of complexes for which the likelihood of being normal sinus or ventricular premature is high and a group which has a greater likelihood of being incorrectly classified. This latter group can be shown to the human editor for verification or correction of classification. The complexes undergo principle components transformation which describes the QRS by a set of derived components. When a classification is approved or changed by the human editor, the computer, utilizing maximum likelihood decision rules, moves ahead in the tape to correct the classification of the remaining unedited complexes on the basis of the similarity of their principle components profile to the edited ones. The system reduced total errors, false positive or false negative, to less than one percent in all of the high arrhythmia and noise content tapes used for this study.

The effect of cost containment on the practice of cardiology: predictions.

When health care cost containment is tied to unit pricing, the system may become price-driven rather than care-driven. Although the incentives engendered by unit pricing may not necessarily result in practices detrimental to the young or the patient with relatively pure disease, the potential for adverse effects on the elderly, the poor and the chronically ill is real. Hospitals will soon emphasize quick turnover, efficiency and intensive care. Diagnostic evaluations and chronic disease care will be moved out of hospitals into physician owned-and-operated facilities and out-of-hospital settings, respectively. The health care system will fractionate, and quality control will require restructuring to achieve the present level of quality assurance. Cardiologists, as well as other physicians, will need to alter their teaching style and teaching locations. Better methods for predicting outcomes will need to be developed; we will no longer have the safety net of following a patient closely and altering management plans according to the patient's response. Cost containment may occur under diagnosis related groups, preferred provider organizations, health maintenance organizations and other prepaid or "capped" systems. There are, however, many issues relative to cost versus quality that need to be resolved if severe detrimental effects on care are to be avoided.

Digitalis cardiotoxicity.

Clinical manifestations of digitalis toxicity were clearly described by Withering in 1785. One hundred years later, certain digitalis-induced arrhythmias were inscribed on the smoked drum, and shortly thereafter with the introduction of the electrocardiograph, manifestations of digitalis toxicity as recognized today were recorded in animals and human beings. With popularization of the direct-writing electrocardiograph in the late 1940s and the introduction of digitoxin in recommended doses (that in retrospect appear inappropriately high), the documented prevalence of digitalis toxicity increased rapidly. With increased understanding of the interaction of electrolytes and digitalis and perhaps, and more importantly, the widespread use of digoxin in doses derived largely from its inotropic action and, thus, inappropriately low for the management of many of the arrhythmias, the prevalence of digitalis toxicity began to decline again. In addition, the advent of serum level determinations and the widespread acceptance of the concept of "therapeutic" levels which, although frequently falling short of the desired clinical end point, served to preclude digitalis toxicity. With the decline in the incidence of digitalis toxicity consequent to these factors, some of the digitalis-related arrhythmias that were common are now rarely observed. This report focuses on arrhythmias that are highly specific for digitalis toxicity and on those that now are less commonly encountered. The discussion and classification of the arrhythmias are based on their most probable electrophysiologic mechanism.

Behavior of the terminal T wave during exercise in normal subjects, patients with symptomatic coronary artery disease and apparently healthy subjects with abnormal ST segment depression.

The Q-T interval and apex of T wave to end of T wave (aT-eT) interval were measured by computer in four age-matched study groups at rest and during exercise to determine whether: the behavior of the aT-eT interval differs in patients with myocardial ischemia when compared with normal subjects, and the behavior of the aT-eT interval differs in subjects with true positive and false positive ST segment responses. Group I consisted of 57 normal subjects. Group II consisted of 41 symptomatic patients with documented coronary artery disease. A group of apparently healthy subjects with asymptomatic ST segment depression during exercise was divided into two additional groups: Group III, those without coronary artery disease; and Group IV, those with coronary artery disease. Subjects were excluded from the study if they had left ventricular hypertrophy or an intraventricular conduction defect or were taking digitalis or type I antiarrhythmic agents. There were no significant differences in the aT-eT interval and aT-eT/Q-T ratio among the four study groups when compared at rest; however, during exercise at similar heart rates, the aT-eT interval was significantly shorter and the aT-eT/Q-T ratio significantly smaller in Groups II and IV, the subjects with coronary artery disease, than in Group I, the normal subjects. The aT-eT interval and aT-eT/Q-T ratio measurements in Group III did not differ from those in Group I at rest or during exercise. In conclusion, the aT-eT interval and aT-eT/Q-T ratio may reflect changes in myocardial repolarization in exercise-induced ischemia and may have potential for future clinical application.