Actually, sleep problems (isolated from other quality-of-life issues) were not identified in either the qualitative research or the Delphi rounds as a long-term domain or subdomain.

Thank you for flagging.

Loreto

• Does fatigue include severity and impact? Also, will sleep quality, duration, or disturbance be included? (You mention pain interference, but night pain affecting sleep would be important.)
• Can I ask why lifestyle factors were not included? Diet, PA, Smoking/Vaping, Alcohol use, etc
• Also wondering about Mental and Emotional Health, Depression, anxiety, etc
• Why has treatment burden not made the list (Time, complexity, side effect management/ adherence/ cost -  is this in a different outcome?
• Is medical error included in death, especially when we have seen errors come with low or high doses of MXT? Delighted that the patient's voice will be included in safety.

1. None of this is your fault! It's just when you approach RA and the core set, things get difficult. There are reasons that this core set has not been updated yet....

2. I totally agree! But if we want to avoid physicians branching out and replicating your process to get a 'MD core set for LOS', we need to find a perspective to avoid that. I'm looking at the exec and TAG also. The original RA core set has joint damage as a mandatory domain in studies of >=1 year, and I can easily imagine an updated RA core set that includes specifics for LOS along the lines of your work.

If we could add in the final description that your core set is a first step to updating the RA core set by focusing on patient needs for LOS, and that other domains may be added as the update proceeds, I would be fine with endorsing this interim step. But at some point MDs need to be able to weigh in and say, e.g., that in the final RA core set for LOS there should be additional focus on domains of disease activity and damage.

3. The development of foundational domains started at about the same time you started, so it's not strange you weren't (fully) aware of them. Also, I'm not sure we are at the stage that such domains are mandatory in all MSc core sets going forward. I think they need to be seriously considered by the developing groups, and if included, the measurement instruments are already available.

4-6. OK, these are ongoing issues.

Shawna, I'd like the exec/TAG to weigh in here, also.

We do agree that pain is more than intensity and interference, and type can be one. This can be captured by the instruments to measure the domain pain. We’re now in the phase of defining the most important domains to measure and pain (in general) was proposed as very relevant by the patients. I really think that, at the domains stage, it is very difficult to go into this much detail but agree that at the instruments one it is advisable, or at least in specific studies.

Thank you for the question,

Loreto

Thank you for your difficult questions that make us think deep. Let’s see if we can satisfy your needs.

1. On the relationship to the OMERACT-ILAR RA core set.
   We spent near 4 years justifying this core set. Our patient research partners saw a need to focus on the long-term outcomes where punctual disease activity is less of a problem, and complications and life impact become more relevant. We did not aim to update the RA core-set for clinical trials. Whenever we presented our objectives, plans and documentation, like previous attempts to standardize data collection in registries and LOS, at OMERACT, I can’t recall any voice saying that we should do this by updating the RA core-set. It would have been nice, because we were told to focus on a disease, which was not totally the intention when we started, and so we focused on RA. Is this a start? Yes, maybe it’s a knock on our brains, and start seeing each core-set from a long-term perspective, whether they would stand a meaningful long-term study. The generic "This process has culminated in a Draft Core Domain Set. This set outline(s) core domains, a key aspect of health or well-being that is essential to define, measure and report in all clinical trials and other research studies" is part of the endorsement website template. We completely agree that this needs to be corrected for ‘special’ core-sets. Happy to further discuss the implications!!
2. Should we develop another set for physicians/researchers? Please, don’t! First because they also voted here, and were involved in the working group, and understood that the long-term is about consequences and impact, and that patient voice was even more important than in RCTs. Second because this domain set can be translated into instruments reflecting everybody’s perspective, but we were able to discuss independence and financial constraints thinking in the long-term. Most measures have a short-term span. Perhaps this WG was meant to start the needed conversations.
3. This is the first time I hear about the OMERACT initiative for foundational domains. Apologies. Now I understand the stiff process we suffered. Foundational domains (i.e., pain, fatigue, physical function, or quality of life that underpin patient experience regardless of diagnosis) should be included in all condition-related core-sets. On top of those, we could have included others that were left out because of the number of domains you can include in a core-set. Should we already populate the onion template with these?
4. Joint involvement overlaps with the core-set. Yes. This is because, as you said before, it’s a foundational domain. We had ‘joint deformity’ as part of what we called ‘joint involvement’, a subdomain relevant for the patients, but this was considered ‘damage’ and was left out, because we had to include two different aspects of pain and other already set domains and run out of space in the onion.
5. ‘HRQoL’ was included as such after binning and winnowing. We had individual aspects of QoL.
6. ‘Comorbidities’ is in both pathophysiological manifestations and life impact categories. We included it also in life impact because some comorbidities were especially threatening for the patients (as shown in the qualitative studies) and had a huge impact on their lives and their decisions.

I do have some questions, though.

1. I'm not fully clear on what this core set represents, and its relationship to the OMERACT-ILAR RA core set.
I think we all agree that the RA core set is in dire need of an update, and also that it is aimed at clinical trials more than at longitudinal studies (LOS). So such an update should include a specification of any domains that are only intended for trials or for LOS. Also, the update should have the support and endorsement of all stakeholders, signifying that (among other things), patient-important outcome domains have been sufficiently included.
In the absence of this update, I applaud the current initiative to get things started. But I think it should be a step towards an updated RA core set, not a separate core set of patient important domains for use in LOS. Because that will create confusion in guidance as to what domains to include in LOS: the current RA core set plus this patient core set? How to handle the overlap in domains? Etc. 
The generic core set definition shown in step 7 of your process inadvertently illustrates the above difficulties:
"This process has culminated in a Draft Core Domain Set. This set outline(s) core domains, a key aspect of health or well-being that is essential to define, measure and report in all clinical trials and other research studies."
So now all of a sudden you appear to be also redefining a core set for clinical trials...

As an aside, I would add: "...of a disease or health condition." Because I think that is what the term 'core set' is for. We should think of another term for sets of outcomes important to measure processes of care (eg, 'shared decision making').

2. Following on this: the scoping review and qualitative study focuses on patient values. But what about the values of other stakeholders (clinicians, researchers, payors...) in LOS? Would they need to develop their own core sets for LOS?

3. How do the chosen definitions mesh with the OMERACT initiative for foundational domains?

4. Joint involvement: it is hard not to rush to measurement issues. But the group will need to think carefully on how this domain interacts with things currently measured in LOS: joint counts, joint damage; and stiffness never assessed at joint level.

5. I thought OMERACT was moving away from including HRQoL as domain, rather focusing on better specified elements of HRQoL including some that are already present in this core set.

6. Minor: I think 'Comorbidity' should not be considered to be in life impact, but only in (Pathophysiological) Manifestations, as having the disease does not equate with a specific level of impact (eg, consider asymptomatic diabetes).