CBS: New Research Says The Best Place To Have A Heart Attack Is Not In A Hospital

An overview of a heart and coronary artery showing damage
(dead heart muscle) caused by a heart attack.
Image from NIH via Wikimedia

Today's article comes from CBS; it is about the findings of a new study published in the Journal of the American Medical Association. In this study researchers investigated the death rate from a specific type of major heart attack. They found that heat attack victims who were already in the hospital were less likely to survive than those who arrived at the ER. The news piece and the research article both state that this pattern persisted even when the difference in health of the patients was taken into account (people in the hospital are, on average, sicker than those not in the hospital). But does the data really show that you're better off having your heart attack at home?

Let's start by talking a little bit about heart attacks. The technical name for a heart attack is a myocardial infarction (MI); the myocardium is the muscle of the heart and an infarction is tissue death due to lack of oxygen (suffocation, in a sense). An infarction can occur in any tissue and it's never good. In the heart, these infarctions are almost always caused by something blocking a vessel that supplies blood to the muscle of the heart. While the attacks themselves usually have a sudden onset, the blockage can be sudden or slow. Slow, gradual blockage of the arteries in the heart caused by a build up of cholesterol rich plaque are the type we tend to think of most often. If the heart attack happens in a person with arteries narrowed by such plaque the heart attack in referred to as a NSTEMI; skipping the technical details, this means the only some of the muscle cells in the area are dead, others are just injured. Unstable angina is similar, it feels like a heart attack, but the heart cells are only injured. If the heart attack happens because the plaque ruptures or some other sudden, complete blockage occurs (embolism), it is a STEMI; in this case all the muscle cells in the area are dead or dying and treatment must be fast to save the patient. The difference between a STEMI and NSTEMI is important for the doctors trying to decide how to best treat the patient (a STEMI calls for a "clot buster" and an NSTEMI doesn't) and is differentially diagnosed by an electrocardiogram (EKG).

So what did the authors of the Journal of the American Medical Association find out about heart attack survival?

This study is a larger follow up to a small study of patients who went to the University of NC hospital. For this study they followed patients from several hospitals in California. They looked at all patients who were diagnosed at the hospital with a STEMI (62,021 patients); these patients were then grouped into inpatient STEMI (had a heart attack while in the hospital for something unrelated; 3,068 patients) or outpatient STEMI (had a heart attack and home and arrived at the ER; 58,953 patients). They then compared the outcomes of the two groups.

The inpatient STEMI group was, on average, older, sicker, and more likely to be female (47% vs 32% female) than the outpatient group. The sicker part makes sense; people admitted to the hospital are probably sicker than people not in the hospital.

After doing statistical corrections for the differences in age, sex, and health the researchers found several differences in the heart attack related medical care and outcomes between the two groups. For the inpatient STEMI group, they were less likely than the outpatient STEMI group to get heart stents placed (21.6% vs 65%; OR, 0.23; 95% CI, 0.21-0.26; P-value < .001), but stayed in the hospital longer (mean, 13.4 days [95% CI, 12.8-14.0 days] vs 4.7 days [95% CI, 4.6-4.8 days]). The inpatient group also had a statistically significantly higher higher mortality rate (OR, 3.05; 95% CI, 2.76-3.38; P < .001).

Based on these findings the researchers conclude that patients who developed STEMI while hospitalized, compared with those with onset of STEMI as an outpatient, were less likely to undergo invasive testing or intervention and had a higher in-hospital mortality rate. However, they acknowledge that there are limitations to this study. There could be errors in the records the statistics are based on due to this being a retrospective study of an administrative database (one used for insurance billing); these errors could come from miscoding or from misdiagnosis. Also, the limitations of the database meant that they could not look at how long it took for a patient to be diagnosed and treated after developing STEMI symptoms or the rate of use of "clot busting" (thrombolytic) drugs. They also have no ability to analyze STEMIs in people who did not go to the hospital at all, this may cause an under representation of the mortality rate of people who have a STEMI outside of a hospital. 

What are my conclusions?

Remember how I mentioned that this was a follow up to a study done at the UNC research hospital? To understand my conclusions we need to go back and take a look at that study. That study was done almost the same way; they looked at people who were in the hospital that had a STEMI and people who arrived at the ER because they were having a STEMI. Because it was only one hospital they had a much smaller number of patients to look at, but they had more complete data about each patient. What they found was that the median time from onset of heart attack symptoms to EKG for inpatients was 41 minutes with 25% of patients waiting 11 hours or more. For the outpatients, EKGs were received in 5 min (median), with only 25% patients waiting 10min or more. There were other statistics that showed that even after the STEMI was diagnosed via EKG, the inpatients still received slower care. In the end, after controlling for the age, sex, and health differences between the inpatients and outpatients (much like the bigger study, the inpatients were older, sicker, and more likely to be female), the researchers found that survival was statistically significantly more likely for the outpatient group (96.9% vs 70.3%; P-value <0.001). Sounds like the same story, doesn't it?

There's one last detail in this study, that's missing from the larger one, and I think it's critical to understanding these findings. In the small study they find that 60% of the inpatients first show symptoms of STEMI such as altered mental state (drowsy), low blood pressure, or trouble breathing. These are atypical symptoms for a STEMI and could be caused by many different things. Only 33% of the inpatients actually complained of heart attack like symptoms (chest pain, heart palpitations, or trouble catching their breath). Of these patients, the complainers, the median wait time for an EKG was only 10min. Possibly due to the small sample size, they do not include the survival rate of the inpatient complainers vs non-complainers.

Based on this, I suspect that there are two things going on. Firstly, if the time from EKG to treatment is slower for inpatients that may be because they are not near the ER doctors and cardiologist that are trained to read EKGs. It may also be because they are sicker and the treatment decisions are more complicated. These are things the hospital can and should work to streamline, as time is of the essence when someone has a heart attack.

But more importantly, I bet that most of the patients getting EKGs in the ER came in and said "Help! I think I'm having a heart attack!", where as most inpatients did not. A person, at home, suffering from drowsiness and some trouble breathing may not go to the ER. If I'm right, the data is biased, and what's really going on is "complainers" are more likely to survive than "non-complainers" (who often have atypical symptoms). This is something we already know to be true. In this case it's possible that, controlling for symptom type (typical vs atypical) might show that inpatients with atypical symptoms are more likely to get treatment and survive than people not already at the hospital.

Taking this into consideration, I don't like the news article's headline; it's misleading. I think, probably, the best place to have a heart attack is in the ER (which is in the hospital), but even then the best survival rate is for those that have easily recognized symptoms so that their treatment isn't delayed. I strongly suspect if your only symptom of a STEMI is drowsiness, you are much more likely to survive if you're an inpatient. Also, the story lacked a decent citation. Fortunately, this month's issue of JAMA was having a special on heart disease, so the article was easy enough to find. All in all, I give CBS a C; not terrible, but not good either.

References

Dai, Xuming, et al. "Acute ST‐elevation myocardial infarction in patients hospitalized for noncardiac conditions." Journal of the American Heart Association 2.2 (2013): e000004.

Kaul, P., et al. "Association of Inpatient vs Outpatient Onset of ST-Elevation Myocardial Infarction With Treatment and Clinical Outcomes." Journal of the American Medical Association. (2014) 312.19:1999-2007