Date: April 10th, 2016

Guest Skeptic: Dr. Salim Rezaie. Salim is an Associate Clinical Professor of Emergency Medicine Internal Medicine University of Texas Health Science Center at San Antonio and the creator/founder of REBEL EM @srrezaie. Salim is also part of The Teaching Course.

Case: A 40 year-old male arrives to the emergency department via EMS with substernal chest pain that has lasted for three days without any relief. He denies shortness of breath, radiation, nausea/vomiting, diaphoresis, or palpitations. He has no past medical history, does not smoke, but states his father died of a myocardial infarction at the age of 70, which is why this chest pain concerned him. He runs four times a week without any difficulty, but has had some increased stress at his work place. Your evaluation in the emergency department shows an ECG with normal sinus rhythm without any other abnormalities and cardiac troponin testing was negative both at arrival and at three hours after arrival. You go to speak to your patient about objective/provocative testing. What will your conversation be?

Background: There are approximately 8 to 10 million patients complaining of chest pain coming to emergency departments in the United States annually. In the US, a very liberal testing strategy is used in order to avoid missing acute coronary syndrome (ACS) in patients with chest pain. This results in over 50% of emergency department patients with acute chest pain receiving serial cardiac biomarkers, stress testing, and/or cardiac angiography at an estimated cost of $10 to $13 billion annually and yet fewer than 10% of these patients are diagnosed with ACS.

To add further angst to emergency department providers the American College of Cardiology/American Heart Association (ACC/AHA) recommends that chest pain patients with concern for ACS should receive serial cardiac markers followed by some sort of provocative/objective cardiac testing.

Using this strategy amongst a low-risk chest pain population unnecessarily uses resources on those least likely to benefit. Low risk chest pain patients have ACS rates of <2%. Provocative/objective cardiac testing is associated with a significant amount of “downstream” testing (i.e. cardiac catheterization) due to false positive tests. These false positives expose patients to potential harms from the testing and treatment.

The value of chest pain characteristics to predict ACS has been studied. None of these individual elements have been shown to have +LR>10 or –LR <0.1 to help us rule-in or out ACS.


Cardiac risk factors like the Framingham Criteria can predict future risk but do not help in the emergency department whether or not the patient is having ACS.

Combining various aspects like: the risk factors, history, physical, labs and ECG findings into a score like the TIMI Score or GRACE Score have been used. A problem with TIMI and GRACE scores is they were not designed to assess whether patients presenting to the emergency department with chest pain are due to ACS. They were designed to risk stratify patients once the diagnosis of ACS had already been made.

A new score has recently been developed in the emergency department to help risk-stratify patients who present with chest pain into who will have a Major Adverse Cardiac Event (MACE) in the next 6 weeks and who will not have a MACE. This new score is called the HEART Score and has five different elements.


To date the HEART Score has examined > 13,000 patients and demonstrated a high negative predictive value for major adverse cardiac events (MACE) at 6 weeks exceeding 98%; but to many providers, a 2% risk is still to high.

Clinical Question: Does the HEART Pathway identify ED patients with acute chest pain who are safe for early ED discharge without objective cardiac testing?

Reference: Mahler SA et al. The HEART Pathway Randomized Trial Identifying Emergency Department Patients with Acute Chest Pain for Early Discharge. Circ Cardiovasc Qual Outcomes 2015

  • Population: Patients ≥21 years of age presenting with symptoms suggestive of ACS that providers ordered an ECG and troponins
    • Key Exclusions: STEMI, hypotension, life expectancy <1 year, and non-cardiac medical, surgical or psychiatric illness needing admission
  • Intervention: The HEART Pathway (Combination of HEART Score with 0- and 3-hr cardiac troponin testing)
  • Comparison: Usual Care (American College of Cardiology/American Heart Association Guidelines) – Serial cardiac biomarkers and objective cardiac testing before discharge
    • Objective cardiac testing was a stress test or angiography
  • Outcome:
    • Primary Outcome: Rate of objective cardiac testing within 30 days of presentation (Any stress testing modality, coronary computed tomographic angiography, or invasive coronary angiography)
    • Secondary Outcomes: Index length of stay, early discharge (discharged from ED without objective cardiac testing), and major adverse cardiac events (all-cause mortality, myocardial infarction, or coronary revascularization) at 30 days.

HEART Score Pathway

Authors Conclusions: The HEART Pathway reduces objective cardiac testing during 30 days, shortens length of stay, and increases early discharges. These important efficiency gains occurred without any patients identified for early discharge suffering MACE at 30 days.

Quality Checklist for Randomized Clinical Trials:checklist

  1. The study population included or focused on those in the ED. Yes. Patients were ≥21 years of age presenting to the ED with symptoms suggestive of ACS
  2. The patients were adequately randomized. Yes. Patients were randomized into 1 of 2 treatment arms with equal probability using a random permuted block randomization
  3. The randomization process was concealed. Yes. Study investigators and staff were blinded to the randomization sequence
  4. The patients were analyzed in the groups to which they were randomized. Yes.
  5. The study patients were recruited consecutively (i.e. no selection bias). No. There were set enrollment hours (6 days a week, excluding Saturdays, 80 hours per week)
  6. The patients in both groups were similar with respect to prognostic factors. Yes
  7. All participants (patients, clinicians, outcome assessors) were unaware of group allocation. No. It would be impossible to blind clinicians to usual care vs. HEART Pathway
  8. All groups were treated equally except for the intervention. Unsure. There was no documentation of who got aspirin, oxygen, nitroglycerin, other medications
  9. Follow-up was complete (i.e. at least 80% for both groups). Yes. Assessment for 30-day events was completed on 96% of study participants.
  10. All patient-important outcomes were considered. Yes and No. The HEART Pathway produced a meaningful reduction in objective cardiac testing, doubled ED rate of early discharge, and reduced the hospital LOS by half a day but was not powered to detect differences in MACE
  11. The treatment effect was large enough and precise enough to be clinically significant. Unsure. Study was powered to detect a 15% reduction in objective cardiac testing within 30 days with 90% power with an expected loss to follow-up rate of 10%. They found a 12% reduction in objective cardiac testing at 30 days.

Key Results: They enrolled 282 patients into the study with 141 in each group. The mean age of the patients was about 53 years with 57% being women. The overall MACE rate was 6% (17/282).

Of the 141 patients randomized into the HEART Pathway there was close to a 50/50 split between patients stratified as low risk (47%) and those stratified as high risk (53%).

There was some significant non-adherence to the HEART Pathway (29% of low-risk patients and 13% of high-risk patients).

  • Primary Outcome: Objective testing with 30 days

HEART Pathway 56.7% vs. Usual Care 68.8% NNT=8

  • Secondary Outcomes:
    • Early Discharge Rate*: HEART Pathway 39.7% vs. Usual Care 18.4% (Absolute Early Discharge Increase of 21.3%) NNT = 5
    • Median LOS: Heart Pathway 9.9 hrs vs. Usual Care 21.9 hrs
    • MACE: ZERO patients identified as low risk by HEART Pathway had an index or non-index MACE.

Screen Shot 2015-04-25 at 3.11.12 PM

Dr. Simon Mahler

Dr. Simon Mahler

Now we have a special treat for the SGEMers. We reached out to the lead author Dr. Simon Mahler from Wake Forest School of Medicine in Winston-Salem North Carolina. He agreed to come on the SGEM and answer some of our questions to get a deep understanding of his research. This 18 minute extra segment takes place between 14-32 minutes of the podcast. You will need to listen to the PODCAST to hear Dr. Mahler’s responses to our questions.

  1. Setting: This was a randomized clinical trial, which makes it a superior design to the previous observational trials and was on a US population rather than European population and so it’s more applicable to North Americans. However, the study was conducted at a single center and this limits the generalizability to other community settings.
    • Listen to Podcast for Response:
  2. Sample Size/Power: The methods say the study was powered to detect a 15% reduction in objective cardiac testing within 30 days with 90% power at the 5% 2-sided level of significance with an expected loss to follow-up rate of 10%. I could not find in the manuscript how many people you needed to recruit but you enrolled 282?
    • Listen to Podcast for Response:
    • You powered your study to find a 15% difference, which you considered clinically significant a priori. Yet only found a 12% effect size. Does that not leave your study a little short of the mark?
    • Listen to Podcast for Response:
    • A 12% absolute reduction in objective testing may be very important to the patient and to the health care system but the p value does not tell us is the precision of the point estimate. For that we need 95% confidence intervals can you provide those.
    • Listen to Podcast for Response:
    • The final thing about sample size and power was about your secondary outcomes. Specifically you make if very clear that the study was not powered to a detect difference in MACE.
    • Listen to Podcast for Response:
  3. Non-Adherence: You left things open in both arms to provide care at the discretion of the provider. In other words you did not mandate the care by the protocol. Why did you make that choice?
    • Listen to Podcast for Response:
    • The non-adherence to the HEART Pathway occurred in 29% (19/66) of low risk patients and 13% of high-risk patients. What would the results have been if there was perfect adherence?
    • Listen to Podcast for Response:
    • So if the physicians strictly followed the HEART Pathway almost half of patients could have been discharged home without objective testing being done prior to emergency department discharge. Do you think this non-adherence represents a more accurate determination of “real world” practice of using clinical decision instruments?
    • Listen to Podcast for Response:
  4. Not all Low Risk HEART Scores are Low Risk: It is possible to have an elevated troponin and all other aspects of the HEART Score be 0 and by definition this would be deemed a low risk patient. Similarly, significant ST-depressions alone would be considered low risk. However, I think some physicians would be uncomfortable stratifying those patients as low risk.
    • Listen to Podcast for Response:
    • Clinical Decision Aids help give us a structure to risk stratify, but the art of medicine is to use clinical judgment. The clinical decision instruments are tools not rules and we need to know how to use the tools to guide our care for individual patient needs.
    • Listen to Podcast for Response:
  5. Shared Decision Making: Evidence based medicine (EBM) is not just about the literature. Evidence based medicine is our clinical judgment, relevant scientific evidence, and patient values/preferences. Dr. Sackett defined EBM as “The conscientious, explicit and judicious use of current best evidence in making decisions about the care of individual patients.” The issue of patient values and preferences was not discussed in this paper but represents an important component of chest pain evaluation and disposition.
    • Listen to Podcast for Response:
HEART Pathway App

HEART Pathway App

HEART Pathway App: Dr. Simon Mahler and his team have developed the HEART Pathway app. It is available to download for FREE at the Apple App Store.

Comment on authors conclusion compared to SGEM Conclusion: We differ a little from the authors. We think using the HEART Pathway in chest pain patients presenting to the ED with suspicion of ACS has the potential to decrease objective cardiac testing, increase early discharge rates, and cut median length of stay. However, based on this study alone it is unclear if it truly results in a zero MACE rate at 30 days.

SGEM Bottom Line: The HEART Pathway appears to have the potential to safely decrease objective cardiac testing, increase early discharge rates and cut median length of stay in low risk chest pain patients presenting to the ED with suspicion of ACS.

Case Resolution: You present the information to the patient letting him know that his ECG does not have any ischemic abnormalities, he is considered a low risk patient based on his HEART Score, and troponin testing at zero and three hours were both negative. At this point you offer him outpatient follow up with his primary care physician vs. objective provocative testing. The patient states the reason he came to the ED is he wanted to make sure he wasn’t having a heart attack and feels comfortable following up with his primary care physician as an outpatient and is discharged home.

Clinically Application: Evaluation of low risk chest pain does not live in the bubble of this one study. Even though this study was not powered to detect the patient oriented outcome of MACE. There have been many studies looking at the use of the HEART Score in the evaluation of chest pain and the sum of these studies shows that we can do amazingly well in a low risk patient with a 30d MACE rate of <2%.


  • Jellema 2013:The results of the entire population of HEART Score 0 – 3 were not included in this manuscript making it hard to know what the exact MACE rate was in this population. There was a subcohort of patients that were admitted to the hospital, for a second troponin test, that were included. Using this population creates a huge selection bias in that although these patients were HEART Scores of 0 – 3, they were admitted because they were thought to be higher risk.
  • Marcoon 2013: The definitions of patient history and ECG used in Marcoon’s paper are different than the original HEART Score definitions.
    • History: Patients with 2 elements typical for ACS (oppressive, burning, left sided, substernal chest pain, radiation, diaphoresis, vomiting, short of breath) were scored as atypical instead of intermediate/suspicious and patients with 3-4 elements typical for ACS were scored as intermediate instead of highly suspicious.
    • ECG: Only ECGs showing ST elevations were scored as a 2, while t-wave inversions/ST depression were scored as a 1, and nonspecific changes were scored as a 0.  Again these definitions would result in underscoring of the ECGs.

What Does Salim Tell His Patient? At this point in your workup with a low risk HEART Score, non-ischemic ECG, and negative zero and three hour cardiac troponins, for every 100 patients who come in to the ED with chest pain, 1 in 100 will have a heart attack or pre-heart attack diagnosis in the next 30 days. Would you like to stay in the hospital for a stress test or schedule an appointment as an outpatient to see your primary care physician?

FOAM logoAdditional FOAMed on HEART Pathway:

Keener Kontest: There was no keener contest winner because the last episode was an SGEM Xtra about the METRIQ Study. If you have not entered the METRIQ yet please help us make FOAMed even better.

Listen the podcast this week for the keener question. If you think you know the answer then email TheSGEM@gmail.com with “keener” in the subject line. The first correct answer will win a cool skeptical prize.

Remember to be skeptical of anything you hear, even if you heard it on the Skeptics’ Guide to Emergency Medicine.

Tom Steele Ad

 Screen Shot 2016-01-10 at 11.16.25 AM

Screen Shot 2016-03-27 at 3.32.09 PM