Predictive validity of the Braunwald classification of unstable angina for angiographic findings, short-term prognoses, and treatment selection.

The authors tested the Braunwald classification for its predictive validity for underlying coronary conditions, clinical courses, and responses to treatment. A reliable definition and classification of unstable angina is needed to help physicians make correct diagnoses of patients' conditions and to appraise findings from clinical trials critically. Many clinical trials have been conducted, but it is difficult to compare the results because of different entry criteria. Of 113 consecutive patients admitted with unstable angina, 89 who had primary angina were studied. Braunwald's classification was applied at admission. The outcomes of interest during hospitalization were coronary angiographic findings, short-term prognoses, and the treatment selected. Multivariate analysis showed that the severity class expressed significant positive predictivity for coronary thrombi (adjusted odds ratio [OR], 6.53; 95% confidence interval [CI], 2.82 to 15.1) and progress to impending infarction (OR, 10.43; CI, 3.35 to 32.49). The treatment (OR, 0.02; CI, 0.004 to 0.08) and electrocardiographic (OR, 0.22; CI 0.10 to 0.49) classes showed independent negative predictivity for coronary vasospasm. The treatment (OR, 3.50; CI, 1.94 to 6.33) and electrocardiographic (odds ratio, 3.27; CI, 1.87 to 5.71) classes showed positive predictivity for the necessity for recanalization treatment with coronary angioplasty or bypass grafting. The Braunwald classification used at admission is highly predictive of underlying coronary conditions, progression to impending infarction, and the final selection of treatment. This classification should be considered in determining patient eligibility in clinical trials and studies.