Background: Missed diagnoses of acute myocardial infarction (AMI) in the ambulatory setting can cause patient suffering and malpractice litigation. Multiple algorithms have been developed to detect the presence of coronary heart disease (CHD) or acute coronary ischemia.
Methods: We performed a case-control study of patients with no prior history of CHD presenting to outpatient practices with potential cardiac ischemia. Malpractice claims files were used to identify 18 cases of patients with missed AMIs. For each case, we identified 3 control patients who had office visits for chest pain during the same month and assessed the association of 4 different prediction tools with missed AMI.
Results: The 18 cases of missed AMI had a 39% 1-month mortality rate. Cases were more likely than controls to be men (67% vs 26%, P = .001), to be smokers (88% vs 39%, P < .001), and to have low HDL cholesterol (39 mg/dL vs 59 mg/dL, P < .001) and elevated total cholesterol (236 mg/dL vs 213 mg/dL, P = .01). A Framingham risk score predicting a 10-year risk of CHD > or =10% and a positive score using the Goldman risk predictor were associated with an increased risk of missed AMI (odds ratio 5.7, 95% CI 1.8-18.4 for Framingham risk score; odds ratio 7.2, 95% CI 1.4-36.8 for Goldman risk predictor).
Conclusions: Among ambulatory patients with possible cardiac ischemia and no prior CHD, multiple algorithms may be useful for improvement of risk stratification.