STEMI Critics Are Right. We're Missing Too Many Heart Attacks

Figure: Non-STEMI ECG patterns that nevertheless indicate an occlusion myocardial infarction (OMI). This figure is reproduced from Ricci et al3 under the terms of the Creative Commons CC-BY license.I recently reviewed a case submitted by indignant paramedics. They had diagnosed their patient with an ST-segment elevation myocardial infarction “a STEMI”—and driven like hell to one of our region's cath lab-capable hospitals. The first physician to meet them disagreed with their diagnosis and sent the patient to the waiting room because, well, we're all drowning here. Rhode Island and Massachusetts emergency departments are the third and fourth most crowded in the nation, respectively (behind only Washington, DC and Maryland).1 But the physician also rapidly obtained his own waitingroom ECG. This second electrocardiogram (ECG) convinced everyone, and within 20 minutes, the patient was on a heparin drip headed for the cath lab. He did well. When I saw the ECGs obtained by the paramedics, I agreed that their patient was having an acute occlusive myocardial infarction (OMI). A review by the cardiologist, however, concluded diplomatically that the care was appropriate—because only the patient's second ECG met STEMI criteria. It's disconcerting, but the cardiologist was not wrong. Yes, the patient was having an identifiable OMI. Yes, we have an emergent treatment for this. No, he wasn't eligible for that, because he failed his ECG screening test for it. This could have been much worse. What if that first unimpressed physician hadn't gotten a second ECG? The patient might have been admitted as an NSTEMI, sat upstairs on heparin, lost a lot of myocardium, and possibly even died—and none of it would have been flagged as a quality assurance (QA) issue, because “NSTEMIs don't get emergent revascularizations.” Once upon a time, the Code STEMI concept was revolutionary. Once we realized that ST elevations on ECG are highly predictive of an acutely occluded coronary artery, we could justify mounting a national, and, in fact, worldwide, crusade to identify and treat that. Doing so involved building a huge network of rapidly obtained and interpreted ECGs, regional cardiac catheterization labs, and national databases to track outcomes and performance measures (door-to-ECG times, door-to-balloon times, and so on). The resulting system is a national treasure. It has saved millions of lives by now. But as an intriguing and pleasantly readable paper just published in JACC: Advances argues, it may also finally be a victim of its own success.2 It has generated so many insights and identified so many other ECG findings predictive of acute occlusive MI3 (see Figure 1) that the artificial dichotomy of STEMIs and “not STEMIs” feels ever more incorrect.2 Way too many identifiable acute OMIs are failed today by this approach. This is particularly clear now that, as reviewed in the May 2024 Emergency Medicine News article by Dr Hunter Jackson, artificial intelligence-driven ECG interpretation is coming into its own.4 Computers that play chess better than any grandmaster are also starting to surpass us at ECG reading. (Spoiler alert: They don't need STEMI criteria to see an acute OMI.) The paper's authors include Stephen Smith and Pendell Myers, emergency physicians from Minnesota and North Carolina, respectively, who have advocated for a post-STEMI future in many venues, including in the November pages of Emergency Medicine News.5 Along with lead author Jesse McLaren, a Toronto emergency physician, Smith and Myers call convincingly for a new focus on the disease—an acute coronary occlusion—and not just on one screening test for the disease: the ECG with ST elevations. Specifically, they want a reorientation away from thinking and acting in terms of STEMI/non-STEMI. They prefer the far more precise OMI/non-OMI. While I'm often skeptical of the latest drive to rename something in medicine—what bright spark decided the pseudoseizure would henceforth be the psychogenic non-epileptic seizure? Why did we need that? (How do I pronounce that again?)—the case for the new OMI/non-occlusive myocardial infarction (NOMI) paradigm is powerful. It's far more than just a rebrand. Has any other important pathology been so completely defined by, or even named for, its imperfect screening test? Imagine reclassifying sepsis as either “lactic acid-elevation sepsis” and “non-lactic acid-elevation sepsis”—LASPIS and NLASPIS, dibs on the copyrights!—and then treating only one as an emergency. The stakes are incredibly high. There are about one million MIs each year in the United States.6 It looks like about a third are STEMIs and two-thirds are NSTEMIs.7 So, there are twice as many NSTEMIs as STEMIs. We have been lulled into thinking of those more numerous NSTEMIs as nonevents. (After all, if they were emergencies—well, then they'd be STEMIs.) NSTEMIs also bore us because we've too often seen that yawn of a diagnosis attached as an afterthought to the problem lists of frail, chronically ill patients admitted for other things. If an underlying illness—say, sepsis or end-stage renal disease—causes a troponin elevation, then the NSTEMI diagnosis is often applied. Those patients are obviously not having an acutely occlusive MI. But including the NSTEMI diagnosis buffs the chart for the billing department. So, in it goes. What a disservice this hazily conceived practice does to our collective understanding! It turns out that a startling number of NSTEMIs—25% in one large metanalysis,8 33% in another9—are caused by acute coronary occlusions. That's right. Many of us hazily think of non-STEMIs as nonocclusive, but we are incorrectly conflating those things. Too many NSTEMIs—from one fourth to one third of them—are the same disease process as the STEMIs. But as a rule, they did not get appropriate care, and so they have higher morbidity and mortality than the rapidly revascularized STEMIs. We don't hear much about all of those missed acute occlusive myocardial infarctions (MIs). Why is that? This comes back, again, to the flawed decision to define this important disease process based on the ECG findings. Consider that when a presumed STEMI patient goes to emergent cardiac catheterization and is found not to have an acute occlusion, the discharge diagnosis will reflect that this was a mistake—a “false positive STEMI.” (Say the patient was cathed over ST elevation from a left ventricular aneurysm; the discharge diagnosis would not be “STEMI,” but “LV aneurysm.”) However, the same studies10 that have confirmed this practice also confirm the converse: When an NSTEMI patient eventually has their nonurgent cardiac catheterization and is found—oops!—to have been suffering all along from an acute coronary occlusion, the discharge diagnosis is not changed back to STEMI. The diagnosis remains dictated by the initial ECG findings. There is no acknowledgement of diagnostic error or missed opportunity. It's just an NSTEMI that did poorly. Too bad. Put another way, there are no “false negative STEMIs.” So, tens thousands of OMIs never make it into the STEMI registry databases or into quality improvement initiatives. Tellingly, NSTEMI patients with acute OMIs who were managed lackadaisically are nearly twice as likely to die as their fellow NSTEMIs—even though the acute coronary occlusion cases also tended to be younger and healthier at baseline than the rest of the NSTEMI cohort (think again of all the frail, elderly sepsis patients with “troponin leaks,” aka NOMIs).9 Talking about paradigm shifts can sound grandiose, but those calling for one are on the right track. Consider again those one million MIs a year. I apologize for my ugly, back-of-the-envelope math, but if there are twice as many NSTEMIs as STEMIs, and if at least 25% of the NSTEMIs are acute occlusions, then this suggests we're failing to rapidly treat at least every third acute MI. Just think of the burden on existing cath labs, if for every two MIs they now take emergently as STEMIs, we in the emergency department started to tack on a third. Buy two, get one free! Of course, even the existing system has a loophole to rush an NSTEMI to the cath lab. If, after two hours of nitroglycerin and morphine, the patient is still having refractory chest pain, and also has an MI-defining troponin level, then maybe the cardiologist will be convinced. In my experience, they are more likely to recommend more morphine. As far as waiting for the troponin to help out, even in STEMIs, the first troponin level is normal 26% of the time.11 For that matter, coronary occlusions are dynamic, and whether they identify as STEMIs or NSTEMIs, they often spontaneously reperfuse. Studies show that even among the definitive STEMIs, about one-third have an open artery upon angiography.2 When the STEMI self-resolves in this way, the cardiologist likely is philosophical. But when that NSTEMI rushed to the cath lab is found to have spontaneously reperfused, expect an angry phone call. Hey, if this were easy, it wouldn't be any fun!

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.