A group of Québécois cardiologists, however, decided that they wanted to try a new system, where neither the paramedic, nor the ED physician, would make the decision. Heck, it wouldn't even be a cardiologist. Instead, the turned the decision over to...
|"Beep boop click tweet" (Translate: "I could go for some poutine.")|
The prehospital providers in Laval, Quebec, were trained to obtain a 12-lead ECG on patients with chest pain or dyspnea, but did not receive any instruction on interpretation. If the machine (a Zoll, not actually an R2 unit!) gave an interpretation of *** AMI ***, the EMTs put in a call to the hospital paging system to activate the cath team. The ECG was not immediately analyzed or transmitted, although it was saved.
Over the course of 2 years the EMTs activated the cath lab 157 times, and most of those went to angiography and were found to have coronary occlusion (a "true STEMI"). A few patients (5%) had a very suggestive ECG, but had nothing on cath ("false positive activation"). It happens.
Unfortunately, 12% of the group were found to be "inappropriate activations," meaning that the ECG shouldn't have led to activation. The authors divide these up into human factors (poor quality tracings, lots of artifact), and machine factors (the machine tried to interpret STEMI during a SVT).
What's helpful about this study
It's good that they have distinguished between "false-positive" and "inappropriate" activations, since these concepts often are mixed up in these kinds of studies. They also highlight the importance of obtaining a quality ECG, free of artifact, as well as the perils of diagnosing a STEMI during tachycardia.
What's not so helpful
The authors focus quite a bit on decreasing the rate of non-STEMI cath lab activations, which is of course a worthy goal. But there is little discussion of the STEMIs that EMS did not catch with the automated system. I.e., this study demonstrated the specificity of the system, but said nothing about the sensitivity.
Look at it this way. Laval is a big town, with about 400,000 residents in the city. Based on estimates of the incidence rate of STEMIs (about 100 STEMIs/year per 100,000 people), there should have been about 800 STEMI patients in Laval during the study period.
In other words, where were the other 800 - 128 = 672 STEMIs?
How many of those 672 patients were transported by EMS? How many were "missed," either because of a bad-quality tracing, a misinterpretation by the algorithm, or because they presented with atypical symptoms, and so never even received an ECG by EMS? We have no idea.
So, while decreasing false or inappropriate cath lab activations is a worthy goal, so is making sure that we're not missing anyone who should be sent there!
On the other hand, in Denmark...
A recent article out of Denmark described the experience with a "modern" approach to cath lab activation (e.g., activation based on smaller degrees of ST segment elevation, inclusion of old LBBB patients, use of EMS activation). The Danish cardiologists found a whooping 20% rate of negative coronary caths - and were totally fine with it. In their discussion, they express more concern with missing the opportunity for an early intervention, than with the "false negative"rate. They conclude that
a triage program with acute angiogram for all suspected STEMI patients should anticipate that 1 of 5 patients will not need primary PCI.Now, I'm not endorsing this rate of negative caths as a goal, and some of their comments suggest that they may have swung the pendulum a bit too vigorously (E.g. "liberal access to acute coronary angiography means some of the patients had a “rule-out” angiogram, for example younger patients with chest pain and an ECG compatible with pericarditis.").
Nonetheless, it's clear that the authors are concerned with catching everyone they can in their cardiology "net."
So, should we take medics out of the loop on STEMI?
Well, as usual, "it depends." There are so many local factors that go into the design of every EMS system, and this may be the most appropriate response, say, in an area that will likely never be able to support the training and QA/QI that ALS systems require. And, frankly, if all you want your EMS system to do is identify the obvious STEMI cases, this is probably the way to go.
Except, you probably want your EMS system to do more.
You want to identify and treat the NSTEMI patients who need early intervention, as well as STEMI patients. You want medics to identify the potential, "hyperacute," STEMIs. You want them to find and treat CHF, distinguishing "cardiac asthma" from plain ol' asthma, and even discriminate between primary and compensatory tachycardias.
But the "automated" STEMI activation doesn't provide any benefit beyond getting one small slice of the EMS patient population earlier care. That's fine, great even, but it's still a limited benefit. Why not aim just a bit higher?