Abstract

This session is about defining remission. Most examples will be about rheumatoid arthritis (RA) because most experience has been gathered in that disease. But the concept of remission should be viewed from a wider perspective than one disease. The session includes two pro-con debates (on the utility of including imaging and biomarkers in a definition of remission).

As in all things, when embarking on a scientific project, one must ask:

“Why are we doing this?”

To even begin with answering the question, we must first agree on a clear conceptual definition of remission. When we started on the development of the ACR-EULAR definition of remission in RA, we used dictionary sources and discussions to settle on this:

“The state of absence of disease activity in patients with a chronic illness, with the possibility of return of disease activity.” [1]

It is clear that choices are made from the beginning, especially with the concept “disease activity”, and the possibility that disease activity returns (as opposed to “healing”, where this possibility does not exist). Disease activity is a tangible concept adequately defined in RA, but less so in many other rheumatological diseases. Also, disease activity is conceptually separate from (mostly irreversible) consequences of the disease, such as damage. Finally, note that the above concept does not contain the elements “duration” or “treatment”.

If we continue with the above concept, why do we want to proceed to operationalize the remission definition? The two main reasons are research and patient care. For both, it is clear that we are defining a very good, perhaps even the best state a patient can be in, given that we are talking about chronic disease, i.e. the root cause of the disease cannot be taken away to heal the patient. Being in such a good state has immediate benefits (minimal disease impact) and probably also future benefits, if lack of disease activity translates to less consequences (damage etc). In both research and patient care, we want a definition that is both valid (favorable test characteristics; links to prognosis) and feasible (time, costs, interpretability). Validity and feasibility oppose each other to a certain extent (eg, definitions with better sensitivity and specificity are usually more expensive). Research and patient care differ in their use of the definition. In research, validity and feasibility can be lower than in patient care, because research is about groups, and cost and interpretability are less of an issue than in patient care.

Most of the people criticizing the current ACR-EULAR remission definition of RA are confused over its purpose: whereas it was intended for use in trials, they criticize it for lack of validity in the clinic. For instance, it is felt that the patient global criterion is too strict, so that patients with no apparent inflammatory activity but a patient global score of 2 or higher (scale 0–10) are “unjustly” not classified as in remission. Also, the lack of a duration or treatment criterion is felt to be a problem, but this is not an issue in research.

In the following pro-con debates, please consider the following:

Proposals to change existing criteria for remission must also be held to the question: “Why are we doing this?”