New Zealand Health Survey - opportunity to include ME/CFS questions

Copied from the News from New Zealand and the Pacific Islands

Just came across the NZ Health Survey.
https://www.health.govt.nz/nz-healt...and-surveys/surveys/new-zealand-health-survey

It looks like it could be a vehicle for getting prevalence figures of ME and CFS in NZ. Would have to advocate for a question to be included in the rotating module (not core) question set I expect.

The current guidelines for topics are defined here...
https://www.health.govt.nz/system/f...bjectives-and-topic-areas-aug09-nov2010_0.pdf

Current Stakeholders (users?) of the Survey have been asked for feedback by 24th July for both the Child and Adult survey. Not sure if the current consultation process is something that ME Awareness NZ can submit on. But we will investigate and try to do so.
https://consult.health.govt.nz/health-survey/adult-nzhs-feedback/
https://consult.health.govt.nz/health-survey/child-nzhs-feedback/

 

@Hutan - The survey is the Behavioral Risk Factor Surveillance System (BRFSS). According to the CDC, it is the US's "premier system of health-related telephone surveys that collect state data about U.S. residents regarding their health-related risk behaviors, chronic health conditions, and use of preventive services."

CDC added two optional questions on "CFS" a few years ago but its up to the states on whether they implement them. CDC did not provide any money to do that.

Paraphrased, the two optional questions are:
Has any medical provider ever diagnosed you with CFS?
Do you still have CFS?

When CDC piloted the questions in 2014 in a few states, they found a lifetime prevalence of 1.6% and a current prevalence of 1.2%. Three times the Jason prevalence estimate and implies a recovery of 25% is much higher than reported in other sources. CDC justified the higher numbers by noting they were similar to another survey from Canada but if I understand, that survey was also inflated.

One perspective is that getting these questions into the survey will raise awareness and build support for a stronger national response to this disease. But in the US medical system - as probably elsewhere - CFS has historically - and is still - used as a wastebin diagnosis and many patients with ME are not diagnosed. IMO, the specific questions added to the BRFSS risk generating erroneous prevalence estimates and fostering false associations with behaviors and risk factors that have nothing to do with ME.

Its possible that different questions would pose less of a problem

One other note... CDC may see BRFSS as its answer to the need for proper epidemiological studies using appropriate methods and properly defined cohorts. But its not and we are pushing for proper epi studies.

Edited to note - I think the BRFSS questions use the term "CFS" but would have to confirm

 

@Simone - you might know?

How about a second question about diagnostic criteria used:
Question 1: Have you ever been diagnosed with ME or CFS?
Question 2: Was the diagnosis made based on [choice of] ICC, CCC, Fukuda, etc. or Don't know.

Not that that would resolve the question of the undiagnosed. The only option there I can see is to send people to the Solve ME/CFS (I think it was) quiz and report their results. Almost certain that wouldn't be acceptable to the creators of the NZ Health Survey but always worth a try.

 

Might be good to get a question about PEM in there?

Just found the info about who is surveyed and how on the site where the data is accessible - as I was unsure who fills the Surveys out, thought it might be the doctors. Turns out that people were interviewed in their homes.
https://minhealthnz.shinyapps.io/nz...11ceee/_w_8a4f8318/_w_7aa52694/#!/methodology

Seems like there is a bit of time for the survey interview, so perhaps the Solve ME/CFS questionaire is a valid thing to suggest.

 

Been skimming through the 2017/2018 adult questionnaire to see what's been done in the past (download here: https://www.health.govt.nz/publicat...ntent-guide-2017-18-new-zealand-health-survey). Unable to read much at the moment but did notice this:

There are several questions about hurdles to accessing healthcare, including dental, such as cost, transport, childcare problems etc. but I didn't see anything relating to being too ill to travel and unable to get home visits. That should surely be included and would also affect people with other conditions I expect.

A more basic question: How many people are interviewed each year? Enough to give a statistically sound result for a relatively rare condition (the others they ask for are very common)? If not it probably wouldn't be a good idea to try to use this particular survey to get data on ME numbers because you wouldn't get any sensible results. The survey could still be useful for more general questions like the one about difficulties accessing health care.

 

So just wondering... what size survey do we believe would be needed given the NZ population of 4,794,436?

 

Yes, Australia committed to funding a health economics study of ME/CFS. Hopefully we shall have an announcement of the successful application soon, which will give you more specifics which you can use.

 

I don't think the population size matters much here. More a question of how many people you need to be able to establish if we're 1/1000 or 4/1000. If the adult survey covers 14000 responses we'd only expect to get between 14 and 56 PwME at those rates and I'm not sure how sound that would be, especially with the diagnosis issues.

Who are the statisticians here? Does anyone know and can tag them? Thanks.

 

Apologies in advance, my mind isn’t all there but submission date is pressing so here goes. I definitely would like the survey to somehow target ME specifically. I think the Solve ME survey does a good job if it could be somehow considered as a basis.
Could we also expand the pool a little by also encouraging the survey to capture all difficult to diagnose conditions? Something along the lines of, “Have you had or do you have a life-altering condition your primary carer has been unable to definitively diagnose?”

Lol and include an appendix listing each for reference. Brilliant! That alone will start some conversation! Perhaps even include as one of the options “unexplained fatigue without experiencing many of the other symptoms required by the CCC or ICC” .

 

Just fyi I plan to make a submission today, probably go with something like...

given the small sample size, but will also raise in a letter that we are drafting to the Minister of Health. Who knows. Might get some support from that angle.
Thanks everyone for your input.
My apologies for the very short notice.

 

I thought you might be interested in this new entry in the Survey...


http://www.washingtongroup-disabili...stion-sets/short-set-of-disability-questions/

This question set would capture people with ME, I would think?

 

I believe @Lucibee is a statistician, I'm not aware of anybody else who is officially trained as such but @Graham and @Adrian usually seem, to me at least, to have a good grasp of numbers.

 

Yes, I think it would. In fact the questions seem awfully familiar. I'm sure I've answered them before somewhere. Just can't think where.

 

I don't know how they select their samples but it seems they do an in-person interview. Whether that means they go to people's homes - which would be the ideal - I do not know (the information may well be there somewhere, but it was too much to read for me).

If we assume perfect conditions - they select a good representative sample and interview people in their homes and ask good questions about PEM - would 14000 be enough for a sound result on ME prevalence? Or 5000 for children?

 

From what I can see over here in the UK, specialists do not use any specific criteria (or if they do, they don't say so), but simply tell a patient that they have ME/CFS.

If the probability was 1 in 1000, and you sampled 14000, there's only a 10% chance of getting more than 18 instead of the 14 you would expect: there's not much chance of confusing it with 4 in 1000. I know that isn't quite what you asked, but it's the easiest to calculate to show that 14000 would be a reasonable size.

There's going to be a far, far greater error due to poor diagnoses, and as such, I'm not sure it is worth doing. Until the diagnosis process is clear, what you are really finding out is how many people have the label rather than the illness.

 

Nope. Not really. I'd ask someone who works for ONS. They'll know all about sampling biases etc.