Validation of a Model to Identify Women with Chest Pain in Need of Additional Diagnostic Testing after an Initial Negative Emergency Department Evaluation for Cardiac Ischemia

BACKGROUND: Diagnostic testing of women with suspected coronary artery disease (CAD) is challenging due to its poor specificity. OBJECTIVE: Validate a previously derived model identifying women who would benefit from further evaluation of CAD after an initial negative ED evaluation for ischemia. METHODS: A retrospective analysis of women from a prospective registry of patients who presented to a university ED with chest pain from 7/99-3/02. This site was independent from the derivation site. Subjects were excluded if the initial ECG or cardiac injury markers were consistent with infarction or ischemia. CAD was defined as subsequent elevation of cardiac injury markers, a positive diagnostic study, or death during the 30 day follow-up period. Predictors of CAD and their weighted value were hypertension (2), history of CAD (2), hypercholesterolemia (1), age >/=60 (3), high clinical suspicion (6). Low risk was defined as a score 4 and < 10, high risk >/=10. Chi square analyses and logistic regression were used for group comparisons. RESULTS: The validation set comprised 2440 women, mean age 50 yrs, SD12. Compared with the derivation set, the validation set was younger (diff -9 yrs, 95% CI-2.2-12), less likely to have a history of CAD (OR 0.7, 95%CI 0.5-0.8), and less often considered at high clinical suspicion (OR 0.7, 95%CI 0.6-0.8). A final diagnosis of CAD was found in 8/1523 (.5%), 30/618 (5%), and 55/393 (14%) in the low, moderate and high risk groups respectively. The relative risk of CAD was greater in the moderate risk group (OR 10, 95% CI 5-22) and high risk group (OR 35, 95%CI 16-74) compared with the low risk group. CONCLUSION: We successfully validated a model that utilizes cardiac risk factors and clinical suspicion for risk stratification in women after an initial negative ED evaluation. These data suggest this model can identify women who are high risk and would therefore benefit from comprehensive diagnostic testing to identify CAD.