Automated System Using Field ECGs Speeds STEMI Diagnosis, Treatment

Download this article's Factoid (PDF & PPT for Gold Subscribers)

A computerized system that can diagnose ST-elevation myocardial infarction (STEMI) using field electrocardiograms (ECGs) without the need for physician input nearly eliminates treatment delays, according to a study published online April 12, 2013, ahead of print in the American Journal of Cardiology. False-positive STEMI diagnoses are relatively rare and in line with expected rates.

Researchers led by François Gobeil, MD, of Montreal University School of Medicine (Montreal, Quebec), studied an automated system (Zoll Medical Corporation, Mississauga, Canada) that was implemented on a region-wide basis for 2 years starting in January 2010. Ambulance technicians performed field ECGs on any patient presenting with chest pain or dyspnea in the area served by a standalone cath lab in Laval, Canada. Positive tests resulted in cath lab team activation and direct patient transfer. The automated ECG diagnosis was not transmitted or reinterpreted by medical personnel prior to arrival at the PCI center.

Most Diagnoses Appropriate

Of 155 cath lab activations available for analysis, only 17% ultimately were considered to be false positive (5%) or inappropriate (6% resulting from incorrect automated diagnosis, or “machine error,” and 6% from incorrect algorithm application, or “human error”).

Of the inappropriate activations attributed to human error, two-thirds resulted from acting on poor quality tracings. Half of the machine error activations were related to supraventricular tachyarrhythmia of greater than 140 beats per minute.

In all, there were 123 evaluable patients who had a final diagnosis of STEMI. The vast majority had door-to-balloon times that fell within the 90-minute goal. Contact-to-balloon times also were low (table 1).

Table 1. Results in Patients with Final Diagnosis of STEMI


Door-to-Balloon Time

Contact-to-Balloon Time

Median (IQR), mins

46 (38-56)

76 (67-88)

Proportion < 90 Minutes



Proportion < 120 Minutes



Maximum Delay, mins



Abbreviation: IQR, interquartile range.

Procedural success was achieved in 94% of patients who underwent PCI. In-hospital mortality for PCI patients was 2.4%, and the total cohort mortality rate was 3.2%, as 2 other patients died after arrival at the cath lab but before PCI was attempted.

Patients with ECG-inappropriate cath lab activations had higher rates of hypertension (P = 0.0070), known coronary artery disease (P = 0.0008), and elevated heart rate (P < 0.0001) at baseline than those deemed ECG-appropriate. They also tended to be older, though the difference did not reach statistical significance.

Strategy Has the Potential to Spread

“Before starting the protocol, we were assuming 10 to 15 percent of inappropriate activations, and that is exactly what we obtained,” Dr. Gobeil said in an e-mail communication with TCTMD. “We have a lot of experience with ECG transmission and primary PCI at our center, and even with [conventional] ECG transmission, inappropriate activations are [just as] frequent.”

Also in an e-mail communication, Christopher Granger, MD, of Duke University (Durham, North Carolina), told TCTMD that this study confirms that physician-less systems can provide good results for STEMI diagnosis and tamp down time delays. Similar systems are already in use in some regions of the United States, “with good success,” he noted. “Many [also] use a combined approach of the computer interpretation plus paramedic read with excellent success.”

The study authors note that ambulance technicians in Quebec are trained to perform ECGs but not to interpret them. “The most important finding from this study is that even if one only uses the computer interpretation, performance can be very good,” Dr. Granger added.

One limitation of the new research is the lack of data on false negatives, Dr. Gobeil acknowledged. Such cases, where STEMI was present but went undiagnosed on ECG, were not tracked in the study. If the ECG was negative for STEMI, those patients were transported to other hospitals and no data were collected. Dr. Gobeil said, however, that false negatives “definitively exist” and that his group is working on a way to evaluate that aspect of the system.

“We hope that a simple system like ours will be used more in the future,” Dr. Gobeil said. “With adequate staff training, it could be widely spread, even in small remote cities, and greatly improve STEMI management for those patients.”


Potter BJ, Matteau A, Mansour S, et al. Performance of a new “physician-less” automated system of prehospital ST-segment elevation myocardial infarction diagnosis and catheterization laboratory activation. Am J Cardiol. 2013;Epub ahead of print.



Related Stories:

  • Drs. Gobeil and Granger report no relevant conflicts of interest.

We Recommend