To the Editor. —Two recent articles in the Archives suggest ways to cut losses or allocate resources better. Each is perhaps more appealing to hospital administration than to the practicing physician.
In the first, Fesmire and colleagues1 suggest that the application of a simple algorithm would allow easier triage of patients to intensive care units or monitored beds when the diagnosis of myocardial infarction is in question. In this report, 57 of 426 patients were defined as low risk for the development of myocardial infarction and complications of coronary ischemia. Of those 57 patients, 4 required an acute intervention (angioplasty), 5 had complications, and 2 had myocardial infarction. This is a false-negative rate of 10% for low risk. I have little doubt that, extended to include more patients, the death column would also have representation.
While we might assume that these 11 patients misdiagnosed as low risk might not
BURNSIDE JW. Statistical Medicine. Arch Intern Med. 1990;150(4):920-921. doi:10.1001/archinte.1990.00390160160054