UK Genome Wide Association Study (GWAS) project - draft website goes live, feedback sought on recruitment plan, and updates

Andy I don't know if this is any help.wondered if it might be an idea to highlight this is some way on here.either now or later.I saw this on PR, , I rarely go there now .it was the top new post and by somebody I know. there was little interest on pr. I came here [ my usual go to first place when I am online]I first looked under several other threads , research and recruitment threads, was going to give up , before I found it . I went to register interest and had some reluctance when I saw they needed my postcode. I came back to this thread.It is very long now and so I had to try to skim thru and to pick out comments by people whose opinion I have learned to trust. I then went back and registered .Thinking about those severe like me who can be infrequent and limited participants, and have such limited energy and ability . I could easily have missed this .and it does seem research that is severe ME friendly .and maybe any fairly new members. this thread is very long now. Myself I rarely look at research threads because amongst other things I find them difficult to follow.

 

Well, I created this thread, https://www.s4me.info/threads/uk-ge...take-part-and-or-join-the-mailing-list.12968/, to make the call to sign-up more obvious. Thinking on it now, I'll add it as a banner, which once you have seen it can be dismissed permanently, to flag it up to everybody.

 

I just had an idea (apologies @Andy @Chris Ponting if you have already discussed this).

With the questionnaires that people will be asked to fill out, what if you had a set of 'compulsory' questions that are necessary for the basic participation in the GWAS study (e.g. the symptom questions needed for the diagnostic algorithm, and basic patient data that you want to gather), and then also an additional set of 'optional' questionnaires that patients who have the energy to do so are able to fill in.

I am just thinking that, with the sample size aimed for, this could be an amazing opportunity to gather masses of data from a huge number of patients which might help with phenotyping, identifying subsets, understanding disease progression, patterns in patient histories, quality of life, functional impairment, etc. etc.

This could be helpful e.g. to look for correlations with any genetic findings that come from the GWAS, but also even if this study doesn't have the resources to analyse the extra data, if participants consented to it being held in a 'data bank' that other researchers or future studies could use, this could be a cheap way to gather data on thousands of ME patients for use in future studies which often suffer from small sample sizes.

If the online questionnaires were hosted on a 'participant portal' that patients could log in to and fill out section by section over time, they could potentially have, say, a year to fill out all the questionnaires, allowing them to pace themselves, thus maybe making it more accessible for more severe patients to provide additional optional data if they desire.

E.g., to register with the study, patients have to fill out and submit the 'core compulsory' questionnaires. If they are accepted, they send in their saliva sample. Then, while data collection/analysis is going on for the remaining patients, those who want to could log in to the participant portal to fill out additional optional questionnaires, log their disease history, etc. at a pace that suits them.

If this additional data is useful for this study and there are resources to include it in the analysis, then this could happen in the latter stages of the 4 year study duration, once people have had time to fill out as much as they can.

If there is no need for the additional data in this particular study, or no resources to analyse it, then it could be stored in a 'data bank' linked to the DNA samples (and blood samples for CURE ME Biobank participants), using a participant identifier code.

Researchers could then apply to access the data, like they do to the CURE ME Biobank. If participatlnt identifier codes are used, researchers could apply for DNA sequences, questionnaire data, and blood samples (for ME/CFS Biobank participants), or any combination of the three. This could help with the historical problem we have of studies which only have 30 participants, because the researchers couldn't get the funding to collect data and samples from more people.

What do people think about this? Would this be feasible @Andy @Chris Ponting ?

There would obviously be some resource costs involved in secure storage of the data, reviewing applications from researchers who want to use it, and administering release of the data to approved researchers, but perhaps this could be tied in somehow with existing Biobank infrastructure?

Since it would be digital data, the costs associated with gathering, storing and releasing it would be much smaller than if it were additional biological samples.

 

Great idea. I can't confirm details, mainly as nothing has been agreed, but I can say that we are investigating possibilities for a digital platform/app much as you describe.

 

Great! Thanks for your reply!

 

I think this is right. Researchers cannot go searching through records of people who have not volunteered.

 

Yes, the idea is that the 20,000 cohort would be recontactable for further questionnaires and studies - no obligation to participate! All data would be anonymised but bona fide researchers would apply for access to the data. Note that all genetic/questionnaire/biomedical data would be linked to (anonymised) individuals.

 

@Chris Ponting @Andy just wanted to say thank you for your efforts and the engagement. You can probably tell we're all pretty excited by the prospect of you being successful in securing funding!

 

Hi. Regarding "recontactable for further questionnaires and studies". What kind of possible studies do you mean? Drug trials?

 

If there is a compelling argument for such a trial (passed by an ethics committee) then eventually yes, but more likely (and sooner) we expect researchers interested in epidemiology, genetics, molecules ('omics), diagnostic criteria and physiological responses to take advantage of the opportunities given by a recontactable cohort. Not just is this cheaper, but these researchers can easily cross-reference between one type of study (e.g. physiological response) and another (e.g. genetics). Advances are often made at the intersection of different fields -> inter-disciplinarity!

 

That is so very true, and applicable to so many different knowledge domains. Not just medical science but all science, and even outside of science. There can be something well established and understood in one knowledge domain, that might be radical and innovative in the context of a different knowledge domain ... but it may need considerable effort, creative thinking, luck, etc, to make the connection.

 

Thank you very much for your reply, this is good to hear!

Aside from being "recontactable" (which will of course be a great resource for recruitment into future studies), are there plans to have optional questionnaires within the GWAS study, beyond those that are essential, which could be collated in a data bank and used for future analyses either by yourselves or others?

This would save the resources involved in re-contacting participants to gather additional data in future, if patients could have the option to fill out a huge load of questionnaires during the duration of the GWAS study. E.g. on an online portal at their own pace.

These could then be used for epidemiology and other studies without having to obtain the resources to recontact patients to ask them to fill out additional questionnaires.

Patients would just consent to the data being used (anonymised) for future studies.

My thinking is that, even though there aren't funds to substantially expand the CureME Biobank as part of the GWAS study, it could be inexpensive to create a very large 'data bank' based on patients (who are able) completing lots of questionnaires, as part of the GWAS study.

The purpose would be to use the infrastructure of the GWAS study to gather additional data that is not essential for the study, but can be used for future research without having to recontact thousands of patients.

Does that make sense?

 

Thanks. What other optional questionnaires would you like to see? SF-36, for example? Yes, we need to ensure that the questionnaire load for a person with ME/CFS is not too great.

 

Haven't we got a thread on the issues with the SF-36? Or is that a general discussion spread over numerous threads?

 

Yep. I will happily take part in this research, but I have zero tolerance for ambiguous questions.

 

No I don't mean the usual ones used by the BPS crowd. I mean anything that could be helpful for either adding context which may aid the interpretation of any genetic findings (e.g. gathering data on severity, disease course and symptom clusters may be useful to see if there are correlations between e.g. severity or certain symptom clusters or progression patterns and specific genetic mutations, which might aid in sub-typing phenotypes matched to genotypes).

The purpose would be that this data would aid the biomedical research, not be instead of it. I think it is possible that questionnaire data could provide more context for the interpretation of e.g. genetic findings or future studies focusing on the pathways affected by any ME-associated mutations identified.

I think with ME we need to understand the whole picture, and the GWAS will (if funded) be potentially a great step in clarifying the genotypes that contribute to the disease. But phenotype information from clinical assesments, patient histories, etc. is also important for associating genotype with phenotype which could help to further the biomedical research. Clinical assesments en-masse won't be feasible for this study, but questionnaires could provide at least part of the contextual phenotype data to complement the genomic data.

I think things like the SF-36, despite its limitations, could be really useful to gather because it would provide a very large 'snapshot' of the impact of ME on a very large number of patients. This could a). Validate the smaller studies already done on e.g. health-related quality of life of people with ME, and b). provide a useful data set for leveraging more research funds, for more biomedical research. I think it will have more impact to say 'the SF-36 scores of 10,000 ME patients show us that the disease is severely debilitating' than the current studies on this which sampled I think fewer than 200 ME patients?

I'm not suggesting the researchers design and validate new questionnaires as that is extremely labour-intensive and impractical. My suggestion is to first look at incorporating existing questionnaires that have already been validated, which don't have to be the crappy ones like Chalder Fatigue Scale.

I think the SF-36 would be good to include. Off the top of my head I'm not sure what else but would have to look into it, or maybe others can suggest any they know of that might be useful?

For some things there might not be a validated questionnaire in existence (e.g. for disease course ? I think there are some more detailed ones that have been used for symptom clusters but not sure whether they have been validated - would have to check).

I am not sure that self-selecting sample would be a big issue. The sample size would still be very big even if, say, 50% of participants decided to fill out the optional questionnaires. There is a risk that it would bias towards less severe people because it is harder for severe people to complete questionnaires. But there are ways to reduce that e.g. by allowing people the entire duration of the study (e.g. a couple of years) to complete the questionnaires so they can be done at a very slow pace. And it would be possible to see whether those who had filled out the optional questionnaires were representative of the whole 20,000 cohort by looking at e.g. % of severe patients in the whole cohort vs % of severe patients who completed optional questionnaires, so any bias in this way can be identified and reported with any published analyses.

Personally I don't think we should say "SF-36 isn't perfect so we shouldn't use it". I think we should say "this is an opportunity to gather data on a huge number of ME patients that is unprecedented and may not happen again any time soon, so let's gather what useful data we can and account for any potential issues with selection bias or imperfect surveys during analysis".

We could obviously be judicious in the selection of which surveys to use.

EDTA: Because the questionnaires would be linked to each person's genetic data by an identifier code, if the questionnaire data was used to add context to the interpretation of the genetic data, including additional questionnaires wouldn't introduce any additional selection bias that isn't already in the the original sampling for the whole GWAS study. Additional risk of selection bias is only introduced by additional questionnaires if these are analysed without reference to the individual genetic data or to the characteristics of the whole cohort.

 

Yes I think SF-36 would be good. I will have to think about others and get back to you.

I think making the additional questionnaires optional and allowing participants a very long time to complete them (e.g. up to near the end of the study, which if someone was recruited early could be a couple of years) would help to reduce load as people could decline to complete them if they felt it was too much, or complete them at a very slow pace if needed.