Our hospital switched to a high sensitivity troponin several months ago. We were given a protocol to use that is a little confusing, so I spent an absurd amount of time reading guidelines to figure out how to safely use this test. This is what I found.

The switch to high sensitivity troponin testing brought with it new protocols. We were given an algorithm to use that consists of a level at time 0, 1, and 3 hours. It seemed excessive as it keeps everyone in the ED for at least 4 hours waiting on the last troponin to result. I set out to find which guideline it was based on and opened a can of worms I was not expecting.

It turns out the ACC/AHA guidelines from 2021 are helpful, but frustratingly vague in some areas. The European guideline was much more clear on details such as what exactly constitutes a significant rise at 1 or 2 hours. The European guideline also gives assay specific recommendations and cutoffs, where the American version just says to use assay specific guidelines without saying what those are. Unfortunately for me, trying to make a reasonable policy for our ED, I fall under the purview of the American guideline.

I did the best I could to use the detail of the European guideline to shore up the gaps in the American guideline and have come up with an algorithm that makes sense to me. I feel every bit of it is guideline compliant. But it’s just my opinion at this point.

I made a lecture outlining all of the details of where every step of it came from. The target audience is residents and attendings in the EM program I am on faculty on, but anyone who watches might find some insights into what the test means and how guidelines dictate protocols. (Mercifully, the speed on the video can be turned up if needed and I think 1.5X is pretty reasonable.)

The algorithm I came up with:

**To be clear, the above protocol is not policy even in the hospital I work in. It’s food for thought at this point.

Let’s summarize the confusing points

What is the goal?

In the ED, the goal is to provide a rapid yet safe way to identify low risk patients who can safely be discharged thus avoiding costly and unnecessary or low value admissions.

On whom should we order the test?

Notice that the first box on the flow sheet says “suspected ACS”. It is not unreasonable to order the test on people with low-ish pretest probabilities. But as we see with D-dimers, ordering them on patients with pretest probabilities of zero or near-zero will lead to false positives and over testing.

When can we say a patient is “low risk” based on the results?

Low risk can be defined in several ways. The ACA/AHA guideline recommends using a clinical decision protocol (CDP), but does not specifically endorse one over others.

HEART is a widely accepted pathway that takes risk factors and other clinical data points into consideration and allows for discharge of low risk patients (HEART score less than or equal to 3). The guideline acknowledges as well that some intermediate risk patients (HEART score 4-6) may be candidates for discharge in the right setting through shared decision making. In terms of safely discharging patients from the ED, the upside of the HEART score is that it is validated with “negative” high sensitivity troponin scores – meaning below the 99th percentile: 17 for women, 35 for men on the test that we use. (That is opposed to the troponin only pathway that requires essentially undetectable levels). The down side is that age and risk factors weigh heavily on the score, which puts some patients in the intermediate risk category even with very unconvincing histories and normal electrocardiograms and troponin levels. (Of note, the HEART pathway uses hs-cTN at time 0 and 2 hours.)

A “troponin only” pathway is the recommendation of the European guideline, and the ACC/AHA recognizes this tool as well. Using a troponin only approach requires the level to be not just “negative” (below the 99th percentile) but “very low” (less than 4) or “low” (less than 5 at one hour or less than 6 at two hours). Numbers like this are very powerful and reassuring, but we don’t see them that often. This sets the bar to an early discharge very high based solely on this pathway. While the “low” and “very low” levels are mentioned in both the ACC/AHA guideline, the “rule in” values in the European guideline (65 ng/L) are not addressed in the American guideline. This leads to some case by case decision making in terms of interpreting trends and deciding who needs a third level. In my interpretation this troponin only pathway is a powerful tool for ruling out MI, but detectable levels of troponin – even below the 99th percentile – require more individualized (i.e. less protocolized) decision making than other clinical decision protocols such at the HEART score.

It seems reasonable to me that in cases where troponin levels are low/very low at the appropriate time intervals, that those levels have been shown to speak for themselves in terms of rapidly ruling out MI in the ED. However, in cases where levels are “negative” but not undetectable, the HEART pathway would land another cohort of patients into a category that can safely be discharged in a guideline supportable way. This is what my suggested protocol tries to convey.

There seems to be a lot of splitting hairs among the clinical decision pathways. Why do some use “low/very low” while others use “negative”?

When people write guidelines, they make their recommendations based off of the research they review. They don’t extrapolate. If a research study uses “very low” in their definition of “low risk,” then that’s what the guideline will use. Admittedly, it feels strange to say that your hs-cTN results have to be essentially “undetectable” (“very low” is less than 4 ng/L – the test’s level of detection), not just “negative” (below the 99th percentile – a level we would consider “normal“) to be ruled out by a troponin only pathway – even if you’ve had symptoms for six hour prior to arrival. But, that’s the nature of guidelines. Perhaps my point of all of this discussion is to educate you on what the guidelines say. If you follow them to the letter and someone had a bad outcome, you could defend your actions by showing that you followed the recommendations set forth by the ACC/AHA.

What about the idea of a significant trend?

For our test, the European guideline says any value at or above 64 is a rule in. In terms of trends, it says that at time 2 hours a delta of 15 is a rule in and a delta of less than 2 is a rule out (as long as the total is less than 64.) Changes between 3 and 14 need a third level, but what do with the third level is not clearly defined. Do I put anyone with a rising third level on the observation unit? If the second level rises by 14 and the third level rises by 6, what do I do with that? Unfortunately, these answers are less clear. Judgment must be exercised. I can’t find where the process is protocolized to quite that granular of a level.

Last thoughts?

Changing to the high sensitivity troponin shifts some of the decision making to the ED provider that we had previously deferred to the observation unit. Clearly, if disposition decisions can be safely be made earlier we need to be on board with it, and we need to do it in a guideline compliant manner. Our first iteration of a protocol called for tests at time 0, 1, and 3 hours. These decisions were probably made with the thought that providers would be more comfortable observing the patients for a longer period given how drastic of a change this would be – a thought that is probably true. However, that protocol does not line up with ACC/AHA guidelines and leads to a very long stay in the ED, which in some ways defeats the purpose of switching tests. As addressed above, picking which protocol to institute is not as straight forward as you might think as resistance to change is a very real phenomenon and multiple options are endorsed. The discharge cutoffs are very strict on protocols that don’t allow for any judgment (troponin only pathways). Leaving room for some individualization (HEART pathway allows for the provider to assign points based on how concerning the history is) moves the discharge bar lower by allowing tests to be “negative,” not just undetectable. None of the protocols call for 3 hour tests on every patient. I feel we can overlap the European and HEART pathways to find patients that can be ruled out with one or two tests, and only selectively get the third. These are my thoughts, and they are a work in progress.

Clearly some of the thoughts in this post were specific to the issues of switching from the old to the new version of the troponin test at our facility. It does, however, bring up universal points about the test characteristics, guidelines, expectations, resource utilization, resistance to change, risk tolerance, risk aversion, and implementing protocols that try to balance all of these issues.