Study objective: A rule based on presenting chief complaints can identify patients for a rapid (5-minute) ECG and decrease delays in treatment of patients with acute myocardial infarction (MI).
Methods: The presenting chief complaint was electronically collected on all patients treated in a community teaching hospital emergency department. A rule for ordering ECG on patient presentation to the ED was developed from a model set of patients presenting from July through December 1994 (22,717 patients) and then tested on a validation set of patients from January through May 1995 (18,759 patients). Outcome measures (delay in performance of ECG and delay in administration of thrombolytic agents) were prospectively collected on written data sheets before (April 1993-May 1995, n=67) and after (June 1995-March 1997, n=128) implementation of the rule at the study hospital.
Results: On the model set, 193 patients had the final diagnosis of MI, with 5 chief complaints having the best performance in identifying patients with acute MI and comprising the rapid ECG rule: older than 30 years with chest pain (130 [67.4%] patients); older than 50 years with syncope (5 [1%] patients); weakness (12 [6.2%] patients); rapid heart beat (2 [1%] patients); and difficulty breathing or shortness of breath (20 [10.4%] patients). On the validation set, 142 patients had the final diagnosis of MI, with the rule performing better than chest pain in identifying patients for a "stat" ECG (sensitivity 93.7% versus 67. 4% [95% confidence interval (CI) of the difference, 15.6% to 33.8%]), although a larger percentage of ED patients would receive a stat ECG (7.3% versus 6.3% [95% CI of the difference, 0.7% to 1.7%]). During the model and validation period, 44 (13.1%) of 335 patients with MI received thrombolytic agents. The rule had higher sensitivity on patients with MI treated with thrombolytic agents compared with patients with MI not treated with thrombolytic agents (sensitivity 100% versus 86.4% [95% CI of the difference, 1.7% to 20. 3%] and specificity of 90.4% versus 93.8% [95% CI of the difference, 3.0% to 3.8%]). For the 4-year study period, outcome improved after the implementation of the rule: mean delay in performing ECGs in patients with MI who were administered thrombolytic agents decreased from 10.0 to 6.3 minutes (95% CI of the difference, 1.1 to 6.4), and mean delay in administering thrombolytic agents decreased from 36.9 to 26.1 minutes (95% CI of the difference, 3.5 to 17.7).
Conclusion: Use of a rule based on chief complaints can identify patients with MI for immediate ECG and decrease delays in performing ECGs and administration of thrombolytic agents.