THE USE OF EMPIRIC CLINICAL-DATA IN THE EVALUATION OF PRACTICE GUIDELINES FOR UNSTABLE ANGINA

Objective.-To determine the applicability to emergency department (ED) clinical practice of a nationally disseminated practice guideline on the disposition of patients with a diagnosis of unstable angina, and t o determine the potential impact of the guideline on hospital admissio ns and demand for intensive care beds. Design.-Application of guidelin e criteria for ED disposition decisions to a validation sample derived from a prospective clinical trial. Setting.-Five hospitals, including 2 urban general teaching hospitals, 2 urban tertiary care university hospitals, and 1 suburban university-affiliated community hospital. Pa tients.-A consecutive sample of 457 patients who presented with sympto ms suggestive of acute cardiac ischemia and who had ''unstable angina' ' or ''rule out unstable angina'' diagnosed by ED physicians. Greater than 90% of eligible patients were enrolled in the clinical trial, fol low-up data sufficient for assignment of a definitive diagnosis were o btained for 99% of subjects. Main outcome Measures.-Acute myocardial i nfarction and unstable angina, based on blind review of initial and fo llow-up clinical data, including cardiac enzyme levels and electrocard iograms. After completion of the trial, without knowledge of final dia gnosis or outcome, the investigators classified patients into risk gro ups specified by the unstable angina guideline. Results.-Of subjects w ith an ED diagnosis of unstable angina, only 6% (n=28) met the guideli ne's criteria corresponding to low risk for adverse events and were th erefore suitable for discharge directly to home. Fifty-four percent (n =247) mel the intermediate-risk criteria; 40% (n=182) met the high-ris k criteria and were identified as requiring admission to an intensive care unit. Actual ED disposition differed from guideline recommendatio ns in 2 major areas: only 4% (1/28) of low-risk patients were discharg ed to home with outpatient follow-up, and only 40% (72/182) of high-ri sk patients were admitted to an intensive care unit. Conclusions.-Alth ough the guideline was intended to reduce hospitalization by identifyi ng a low-risk group, the small size of this group among ED patients su ggests that little reduction in hospitalization can be expected. Indee d, the guideline may increase demand for the limited number of intensi ve care beds to accommodate patients with unstable angina considered h igh-risk but currently placed elsewhere. These results emphasize the n eed to use empiric data from target clinical settings to assess the li kely actual impact of guidelines on clinical care prior to national di ssemination.