Date: January 16th, 2020

Reference: Troponin Testing and Coronary Syndrome in Geriatric Patients With Nonspecific Complaints: Are We Overtesting? AEM January 2020

Guest Skeptics:

  • Dr. James VandenBerg: James has a master’s degree in clinical investigation from Washington University in St. Louis, and is currently the Chief Resident at Detroit Receiving Hospital.
  • Dr. Andrew Huang: Andy is the Chief Resident at Sinai-Grace Hospital.

Case: As the resident, you have just finished seeing a 78-year-old male who has been brought in by his family over the holidays. The triage nurse has put the reason for the visit as “multiple complaints”. Despite spending 30 minutes in the room, you still are not sure exactly why the patient is here.

Your attending says that if you take a good geriatric history that you can always determine what’s going on. However, 15 minutes later your attending leaves the room defeated. The patient’s complaints are just so nonspecific.

The attending ends up ordering the “geriatrogram” – ticking off every blood test on the form, including the troponin. You turn to the attending and ask, “do you really think this could be acute coronary syndrome (ACS)?”

Background: Patients 65 years and older account for about 15% of emergency department visits in the United States. Their presentations are often complicated as they present with nonspecific symptoms, and there is often obscuring co-morbid conditions, polypharmacy, and cognitive/functional impairment.

Nonspecific symptoms in the elderly usually yield a broad differential and there are no recommended diagnostic algorithms, leading to extensive testing. ACS is usually amongst this differential, as cardiovascular disease is a leading cause of morbidity and mortality in this population.

Additionally, the elderly population with ACS more commonly presents without chest pain compared to younger patients (up to 20% of elderly patients with MI present with “weakness” as part of their chief complaint). While cardiovascular disease is the leading cause of mortality and morbidity in the elderly, the frequency of ACS amongst this population presenting with nonspecific symptoms is unknown.

Clinical Question: What is the frequency of ACS in elderly patients presenting to the ED with nonspecific complaints, and what is the utility of troponin testing in this population?

Reference: Wang et al. Troponin Testing and Coronary Syndrome in Geriatric Patients With Nonspecific Complaints: Are We Overtesting? AEM January 2020

  • PopulationPatients aged 65 years and older presenting to the emergency department with nonspecific chief complaints who underwent troponin testing. “Nonspecific” was designed a priori as including weak or weakness, dizzy or dizziness, fatigue, lethargy, altered mental status, light-headedness, medical problem, examination requested, failure to thrive, or “multiple complaints.”
    • Exclusions: If they had a focal chief complaint (ex. focal pain, injury complaint, shortness of breath, vomiting, diaphoresis, syncope, fever, cough, focal neurologic deficit)or fever of at least 38C at triage.
  • Investigation: Troponin testing
  • Comparison: None
  • Outcomes: There were multiple outcomes of interest:
    1. The proportion of patients with nonspecific complaints who underwent troponin testing.
    2. The proportion of such patients who had elevated troponin.
    3. The proportion of patients with ACS at the index visit or within 30 days.
    4. The utility of troponin testing to diagnose or exclude ACS.
    5. The frequency of other causes of troponin elevation in this population.

Dr. Alfred Wang

This is a LIVE episode of an SGEMHOP which means we have the lead author on the show.  Dr. Alfred Wang is an emergency medicine physician at Indiana University in Indianapolis, IN. With the help from a dedicated team of physician-peers and mentor, Dr. Wang was able to complete this research project.

Authors’ Conclusions: “While consideration for ACS is prudent in selected elderly patients with nonspecific complaints, ACS was rare and no patients received reperfusion therapy. Given the false-positive rate in our study, our results may not support routine troponin testing for ACS in this population.”

Quality Checklist for A Chart Review: There is a quality check list for ED studies that was published by Gilbert et al in Annals of EM 1996. It had eight items. The list was updated and expanded by Dr. Andrew Worster from BEEM to include 12 items.

The authors of this retrospective chart review did a great job and 11 out of 12 answers were yes. The only “no” was that they did not have a management plan described for missing data in the publication.

  1. Abstract Training: Were the abstractors trained before the data collection? Yes
  2. Case Selection Criteria: Were the inclusion and exclusion criteria for case selection defined? Yes
  3. Variable Definition: Were the variables defined? Yes
  4. Abstraction Forms: Did the abstractors use data abstraction forms? Yes
  5. Performance Monitored: Was the abstractors’ performance monitored? Yes
  6. Binding to Hypothesis: Were the abstractors aware of the hypothesis/study objectives? Yes
  7. Inter Rater Reliability (IRR) Mentioned: Was the interobserver reliability discussed? Yes
  8. IRR Tested: Was the interobserver reliability tested or measured? Yes
  9. Medical Record Identified: Was the medical record database identified or described? Yes
  10. Sampling Method:Was the method of sampling described? Yes
  11. Missing Data Management Plan: Was the statistical management of missing data described? No
  12. Institutional Review Board Approved: Was the study approved by the institutional or ethics review board? Yes

A chart review is a type of observational study. We do have an SGEM quality check list for observational studies.

Quality Checklist for Observational Study:

  1. Did the study address a clearly focused issue? Yes
  2. Did the authors use an appropriate method to answer their question? Yes
  3. Was the cohort recruited in an acceptable way? Yes
  4. Was the exposure accurately measured to minimize bias? Unsure
  5. Was the outcome accurately measured to minimize bias? Yes
  6. Have the authors identified all-important confounding factors? Unsure
  7. Was the follow up of subjects complete enough? No
  8. How precise are the results? Precision was poor. The 95% confidence interval for sensitivity was 48-100%. Spec was better at 77-85% but we must remember these measures are CORRELATED, and therefore the poor sensitivity is also a reflection on Specificity. Had they picked a different cut-off for troponin then they could have improved the sensitivity (at a cost to the specificity)
  9. Do you believe the results? Yes
  10. Can the results be applied to the local population? Unsure
  11. Do the results of this study fit with other available evidence? Unsure

Key Results: They initially identified 1,146 potentially eligible patients. After excluding the patients who had a specific complaint listed and those with documented fever, they were left with a total of 594 patients. Of those, 69% had troponins ordered. The average age of the cohort was 78 years old, 58% were female, and 75% were admitted. The most common chief complaints were altered mental status (43%), weakness/fatigue (33%), and dizziness (21%).

  1. The proportion of patients with nonspecific complaints who underwent troponin testing: 412/594 (69%)
  2. The proportion who had an elevated troponin in the ED: 52/412 (12.6%) (Another 30 patients had an elevated troponin at some point during their hospital stay)
  3. The proportion of patients with ACS at the index visit or within 30 days: 5/412 (1.2%) All occurred during the index admission.
  4. The utility of troponin testing to diagnose or exclude ACS. Looking only at the first troponin in the ED, it was 80% sensitive and 88% specific (NPV = 99.7%, PPV = 7.7%) for ACS. The LR+ was 6.67, and LR– was 0.23. Considering all troponins, the sensitivity was 100% (95% CI = 48%–100%), the specificity was 81% (95% CI = 77%–85%), the NPV was 100%, and the PPV was 6.1%.
  5. The frequency of other causes of troponin elevation in this population. There was a long list of non-ACS causes of troponin elevation. The top 3 causes were: dehydration, heart failure, and atrial fibrillation.

We asked Dr. Wang ten questions to get a greater understand of his publication. Listen to the SGEMHOP podcast to hear all of Dr. Wang’s answers.

  1. Dr. James VandenBerg

    Defining “Non-Specific”The definition of “non-specific” symptoms is problematic while at the same time being pragmatic. For instance, “dizzy” could be construed as non-specific, but what if the patient had supporting focalized neurologic complaints? Additionally, some physicians list the chief complaint as the leading sentence a patient provides. This is problematic if a patient initially cites a “non-specific” complaint, but then describes suggestive ACS symptoms in their HPI. Conversely, “focal” chief complaints such as “shortness of breath” can be construed as non-specific in real practice based on the patient’s HPI, but due to the paper’s inclusion criteria, if any triage nurse or physician labeled a chief complaint as “focal” they would be excluded.

  2. Chief Complaints Not Equal: Definitions of nonspecific included a spectrum of complaints, from altered mental status to failure to thrive. I imagine the yield of testing is much higher in altered mental status than it is in failure to thrive. Would there be a benefit of considering these chief complaints separately?
  3. Retrospective Charting: You excluded patients who had nonspecific complaints at triage, but had a focal complaint listed in the ED physician note. The ED physician note might have been written after the troponin result was known. In the presence of a positive troponin, focal complaints might have been emphasized, despite being originally nonspecific. 
  4. Definition of ACS: You did a good job prospectively defining what would count as ACS based on objective measures. However, neither the decision to take a patient for revascularization nor stress testing are perfectly associated with ACS. The result is a possible over call of patients with ACS. On the other hand, based on the information provided, I don’t think we can be 100% certain that the 5 patients diagnosed with ACS truly had ACS.
  5. Use of A Single Troponin: Even before the use of high sensitivity troponins, troponin testing has never been binary. There have always been a large number of patients in a grey area, where clinician judgement or repeat testing is required. Here, you judge the value of troponin testing based on a single test. Do you think that troponin testing would have been more accurate if multiple values or the physicians’ interpretation were considered?
  6. Troponin Assays: The study also utilized two different troponin assays: A troponin I point of care whole blood assay (istat, Abbott) with cutoff of 0.08, based on 99th percentile, was primarily used in the ED. Inpatient troponin testing was performed with a troponin I fourth generation (Access, Beckman Coulter). The cutoff was 0.04, also based on 99th percentile. It is unclear whether these two measurements were of equal accuracy.
  7. Dr. Andrew Huang

    Rise and Fall in Troponin: You describe 30 patients who had a negative troponin in the ED and a positive troponin later during their hospital stay. Part of the definition of MI is a rise and fall in troponin, so these patients seem to fit that definition. What criteria was used to exclude ACS in these patients despite the objective evidence of cardiac ischemia?

  8. Selection Bias: Rather than looking at all patients with non-specific complaints, you only looked at the patients in whom a clinician decided to send a troponin. Presumably, as compared to the patients with the same chief complaints without a troponin drawn, these are higher risk patients.
  9. Positive Predictive Value: Although the sensitivity and specificity numbers look reasonably good, there were more than 10 false positives for every true positive. That results in a positive predictive value of only 6-7%.
  10. Deaths: There were 32 deaths during the 30 day follow up period, as compared to only 5 diagnoses of ACS. Considering the inaccuracy in determining cause of death, might some of these patients actually have been missed ACS, and if so, how would have that altered your results?

Comment on Authors’ Conclusion Compared to SGEM Conclusion: We agree that, although ACS can have atypical presentations in elderly populations, the results don’t support routine troponin testing for all patients with nonspecific complaints.

SGEM Bottom Line: In this retrospective study of elderly patients presenting to the emergency department with nonspecific complaints, the diagnosis of ACS was rare, and troponin testing had limited value.

Case Resolution: Despite the resident’s concern, you decided to order a troponin anyway, and you are relieved you did when it comes back positive. However, 3 weeks later when reviewing the patient’s course, you notice that he had a significant bleed during an angiogram and the cardiologist ultimately determined that the troponin was a false positive.

Clinical Application: The yield of troponin testing was low in this single centre retrospective cohort. However, the troponin testing may have led physicians to change the chief complaint from something non-specific to something focal, eliminating patients from this trial. It is therefore difficult to recommend any practice changes based on these results.

What Do I Tell My Patient? Based on the symptoms you are describing to me, it is very unclear what is causing your symptoms, but I think that a heart attack is very unlikely. We could send a blood test to help check on your heart, but with your symptoms the tests are wrong more often than they are right, so we might end up having to do even more tests. The other option would be to observe you over the next day, and only add the heart tests if we can’t figure out what is going on or you develop new symptoms”

Keener Kontest: Last weeks’ winner was Dr. Cindy Bitter from St. Louis, MO. She knew that ACHIEVE II was the name of the RCT using the same protocol, same authors but different doses of ubrogepant and was published recently in JAMA.

If you know the answer to this weeks’ question then send me an email to with “keener” in the subject line. The first correct answer will receive a cool skeptical prize.

SGEMHOP: Now it is your turn SGEMers. What do you think of this episode on getting troponins on elderly patients with non-specific complaints? Tweet your comments using #SGEMHOP. What questions do you have for Alfredand histeam? Ask them on the SGEM blog. The best social media feedback will be published in AEM.

Also, don’t forget those of you who are subscribers to Academic Emergency Medicine can head over to the AEM home page to get CME credit for this podcast and article. We will put the process on the SGEM blog:

  • Go to the Wiley Health Learning website
  • Register and create a log in
  • Search for Academic Emergency Medicine – “January”
  • Complete the five questions and submit your answers
  • Please email Corey ( with any questions or difficulties.

If you would like a copy of my slides from the LIVE presentation of this #SGEMHOP episode you can click on the LINK or the picture below.

Copy of Slides


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