Emergency department (ED) physicians can use a simple ECG printout to help predict which patients are at highest risk of cardiac arrest, which is associated with increased mortality. That's the finding of a recent study supported by the Agency for Healthcare Research and Quality (HS06208) and led by Harry P. Selker, M.D., M.S.P.H., of the Tufts-New England Medical Center.
In an earlier Thrombolytic Predictive Instrument (TPI) Study (cited in the August 2002 Research Activities), Dr. Selker and his colleagues used clinical, laboratory, and electrocardiographic (ECG) data available when patients arrive at the ED to develop an ECG-based model to predict the likelihood of cardiac arrest in ED patients with symptoms of heart attack (acute myocardial infarction, AMI).
In the present study, they investigated the clinical implications of the time-dependent predictive features of cardiac arrest in the TPI cardiac arrest model based on a case-control study of ED patients with AMI: 65 cases with sudden cardiac arrest and 258 without cardiac arrest. Within the first hour of AMI symptom onset, increased risk of cardiac arrest was associated with ECG prolonged QTc interval and a greater sum of ST-segment elevation, after adjustment for age, systolic blood pressure, serum potassium, and infarct size.
However, after 1 hour, the effect of ST-segment elevation was much reduced, and prolonged QTc appeared protective. For patients arriving at the ED 30 minutes after chest pain onset, the risk for cardiac arrest for patients with a prolonged QTc (0.50) was more than double (odds ratio, OR 2.20) that of patients with a QTC of 0.44, whereas for those arriving at the ED 1.5 hours after chest pain onset, the risk was 79 percent lower (OR 0.21). Patients presenting to the ED 30 minutes after chest pain onset with a sum of ST elevation of 20 mm had a three-fold higher risk of cardiac arrest than patients with a sum of ST elevation of 5 mm (OR 3.37). However, for those arriving 1.5 hours after chest pain onset, the risk was barely elevated (OR 1.18). These relationships, which were not previously appreciated, provide an example of how multivariable modeling in a health services research project can lead to new pathophysiologic knowledge of a disease, concludes Dr. Selker.
See "Time-dependent predictors of primary cardiac arrest in patients with acute myocardial infarction," by Dr. Selker, Merritt H. Raitt, M.D., Christopher H. Schmid, Ph.D., and others, in the February 1, 2003, American Journal of Cardiology 91, pp. 280-286.
Return to Contents
Proceed to Next Article
540 Gaither Road Rockville, MD 20850