ESC Evidence Levels and Grades

We know clinical guidelines (or recommendations) are pivotal to clinical practice.

Clinical guidelines summarize the current medical knowledge, weigh the benefits and harms of diagnostic procedures and treatments, and give specific recommendations. They can be used for undoubtedly good reasons, like recommending clinicians how to handle patients, but also for less benevolent reasons such as malpractice claims, where guidelines are used and misused as juridical evidence.

The more we learn about how guidelines are made, the more we realize these are not universal truths. When a task force meets to design a guideline, they have to deal with different levels of uncertainty. It is good practice to reflect that level of uncertainty in clinical guidelines.

Here we are going to explore a very straightforward and simple example using the ESC (European Society of Cardiology) framework to assess the evidence.

Let’s start with a clinical question…

Imagine you are a young physician assessing a 72-year-old female patient with congestive heart failure who has anemia (low hemoglobin levels) and you ask yourself…

Shoud I use erythropoietin stimulating agents*?

*a drug that leads the bone marrow, the factory of red blood cells, to make more blood cells and hemoglobin.

In case of doubt, you can the latest recommendations on the subject… In this case, let’s look at the 2021 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure. Here we found this table.

Short answer: Erythropoietin stimulating agents are not recommended in the absence of other indications*

*these “indications” are out of the scope of this article.

But you notice there are two columns with letters “III and B” (with nice colorful backgrounds).

What do they really mean? Do they have any use?

First — Class of recommendation — WHAT does the guideline state?

The class of recommendation reflects the level of agreement between experts in the task force of the guidelines.

We can split them into

• Yes, go ahead, we all agree – Class I;
• Well, it could help if you do it, it is not unanimous but anyway… Class II (IIa and IIb);
• No, just avoid it, we all agree — Class III

Second Level of Evidence — HOW STRONG is the claim?

Now that we know WHAT guidelines try to tell us (class of recommendation), we arrive at the second layer. HOW strong is that claim? Or how well-backed is the recommendation? Here comes the second letter; A, B, or C.

In other Epidence articles, we will take a deeper dive into the design of different studies. For now, randomized clinical trials (RCTs) and metanalysis are the gold standard.

Going back to the previous example, the task force strongly recommends against using erythropoietin-stimulating agents

“So, this means all evidence that isn’t Class I or III and Level A is useless? Right?

WRONG

Just keep in mind the following issues:

• Clinical trials are very expensive — An effective drug may have never been tested for a given indication in a clinical trial just because an RCT is just too expensive to test an unpatented drug (There is no potential return on investment for the sponsor);
• Observational studies are not inferior to clinical trials — Clinical trials are “perfect” because they assure “internal validity” (all participants have to pass a set of inclusion/exclusion criteria) and randomization avoids unknown bias (regarding the primary outcome). However… patients on RCT are usually “too healthy” or “too niche”, meaning that physicians have to take clinical decisions on sicker patients extrapolating data from RCTs carried out in healthier patients. Observational studies represent a more realistic overview of the population;
• Expert opinion is not as useless as it sounds — When we go to a physician we are not looking for someone with a degree to make blind choices following algorithms. We are usually looking for someone with clinical expertise who can make individually tailored medical decisions using the most evidence-based approaches. In other words, we are looking for an expert opinion. Some interventions have a well-known effect for given conditions and that practice is passed through generations even when there is no strong data from SLRs or RCTs;

But I thought there were only 4 levels of evidence. Wasn’t there a level 5?.. . Confused?

Yes, ESC uses a unique method. There are also other methods used for other scientific societies to assess evidence, such as GRADE (Grading of Recommendations, Assessment, Development, and Evaluation) and OCEBM (Oxford Centre for Evidence-Based Medicine).

For the Medium story https://medium.com/evidentebm/how-to-read-and-assess-guidelines-55d8e0d136c2

Disclosure

Sources from McDonagh TA et al. ESC Scientific Document Group. 2021 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure. Eur Heart J. 2021 Sep 21;42(36):3599–3726. doi: 10.1093/eurheartj/ehab368. Erratum in: Eur Heart J. 2021 Oct 14;: PMID: 34447992. are used here for Educational Purposes, according to the statement “The content of these European Society of Cardiology (ESC) Guidelines has been published for personal and educational use only”.

This article does NOT give any kind of individual medical recommendation. If you are seeking medical advice please visit a licensed physician in your country.