Evaluation of a Multidisciplinary Integrated Treatment Approach Versus [SMC] for Functional Gastrointestinal Disorders (FGIDS), 2022, Bray et al

This is an Australian study, mostly Brisbane.

What do they mean by functional?

Yeah, it's a bit hard to know. I think they are saying 'we don't really know, but we are happy to assume everything is having an effect and it's certainly not just biological


They note that a 50-60% 'efficiency gain' is reported in patients with IBS from standard pharmacological or diet treatments and that

So, they have no excuse for not realising that surveys of 'how do you feel now' are not going to be reliable.

With currents treatments not really working, why not try a 'multidisciplinary integrated intervention'?

So, it's a retrospective analysis of patient records. I think there's quite a lot of scope for bias in the selection of patients with the integrated care clinic treatment and of the patients who got standard care. Patients attending, and remaining, in the 12 week program of the integrated care clinic were likely to be highly self-selecting. Patients who had already tried dietary changes and exercise and who didn't think psychological issues were causing their condition would be much less to agree to the treatment. And researchers keen on the concept of integrated care may have let bias creep in to their selection of the matched controls.

There was no data on anxiety and depression in the standard medical care cohort - so no control data to anchor the reported levels of anxiety and depression in the integrated care cohort.

Integrated care involved psychological treatment, dietary advice and exercise.The psychological treatment included psychoeducation:

The dietary advice almost certainly was more comprehensive and tried different approaches, compared to the standard medical care.

 

Outcome measures
It won't come as any surprise to hear that the primary outcome measure was a self-reported assessment of symptoms. (Placebo effect (enhanced by a bit of psychoeducation), what placebo effect?)

And the HADS (anxiety and depression) scale, for the integrated care patients only.

 

A mixed linear model for data analysis - always a bit of a black box, a chance for data fiddling.

Just to repeat that, of the eighty patients that they managed to convince to try one session, 35% dropped out before the end of the treatment. No mention of that in the abstract. Do we think that drop out rate and the mixed linear model data fiddling might have influenced results? I certainly do. (The data relates to a time before Covd-19 - the pandemic is not an excuse here.)

Some of the 52 patients who did complete the treatment did not do any psychological sessions at all, some did no exercise sessions at all - we aren't told how many patients were in these groups.

They don't give us the raw data, only the results that popped out of their mixed linear model.

This is indeed psychosomatic research - what you wish to be true becomes the truth. This study looks so good in the abstract, but it takes only a quick read to see that it has so many problems as to not constitute reliable evidence.

 

I honestly approached this study with an open mind - I thought, maybe stress can impact on gastro-intestinal symptoms. But this is a really bad paper, it is unusable as far as evidence goes. I'm going to record the researchers names here.

Nicola A. Bray,
Natasha A. Koloski,
Michael P. Jones,
Anh Do,
Siong Pang,
Jeff S. Coombes,
Sarah McAllister,
Jane Campos,
Leela Arthur,
Paul Stanley,
Katherine DeMaria,
Che-yung Chao,
Rachel Catague,
Amanda Whaley,
Nicholas J. Talley &
Gerald J. Holtmann

Professor Gerald Holtmann is the corresponding author

Funding

 

I think there might be a problem with the presentation of the data in the chart too.



It's hard to tell with no tick marks on the axis, but this is what we are told:

So, the text says that the control's 95% confidence interval for change in symptoms is from 0.6 [worsening] to 4.0 [improvement], with a mean of 1.7. The upper confidence interval marker on the chart does not go anywhere near 4*. And the lower confidence interval marker on the chart is clearly 0, not 0.6. The effect of the differences is to minimise the actual change for the controls in the chart.

* In case you think it's a tick mark issue, look at the upper confidence interval marker that is supposedly representing 4 for the controls, and then look at the lower confidence interval marker for the ICC that is supposedly representing 5.8.

Am I misunderstanding something here? (Edit, I was a bit, see later post, but there are still problems.) If I'm not, then my confidence that their mixed linear model manipulations are reasonable is even lower than it was before.

 

Both subjective self-report.

Stopped there.

 

I've just realised I did misunderstand a bit - the 95% confidence interval for change for the controls was 0.6 worse to 4.0 better. (All the numbers except for the 0.6 are negative, indicating improvements.)

There's still issues with the chart though.

 

This links the SAGIS questionnaire. As usual it's a whole list of symptoms with the patient grading them from 0 (no problem) to 4 (very severe problem).
https://cdn.ueg.eu/ueg-week-2015/posters-and-videos/P1800.pdf

So it looks like it's intended mainly to record which aspects of the gut are causing a patient problems, and how bad they think they are.
Adding them all up to create an overall score seems pretty meaningless to me. And of course it's all subjective, so vulnerable to change as a result of persuasion to interpret symptom severity differently.

As for this so called research, it is so vague as to be useless in terms of what treatment differences there were between the two treatment regimes which were compared on the basis of a retrospective look at patient records before and after the change in provision at the clinic. Maybe they recruited a more persuasive CBT therapist who convinced the patients to fill in their forms differently, or maybe they went in harder on persuading patients with IBS to stick to a low FODMAP diet, or maybe they changed the medication of those with acid reflux. There seems to be nothing in the research to say what they were actually researching.

 

Indeed. It appears that in 2019, Nicola Bray was at a conference in San Diego making presentations with these titles on the basis of this research.

273 – A Multidisciplinary Integrated Treatment Approach is Superior to Standard Care for Functional Gastrointestinal Disorders (FGIDS): A Case-Control Study

Tu1618 – Patients with Functional Gastrointestinal Disorders (FGIDS) Benefit from a Multidisciplinary Integrated Treatment Approach to Reduce Anxiety and Depressive Symptoms.

Yes. The point I made above about us not even knowing if many people had a particular intervention

probably bears elaboration:

Some people had no psychological consultations, for example. Of the psychological consultations, we are told that they might have been for any of these: "cognitive-behavioral therapy (CBT), Acceptance and Commitment Therapy (ACT), relaxation techniques, or Mindfulness".
And so it goes for the diet and exercise interventions too.

It's also mentioned that the following medical people were involved in the treatment:

So, there was also biological treatment, and a GP helping the person access community 'health care services'.

You would hope that something in all of that would help. But the use of subjective outcomes means we can't even say with confidence 'if we offer patients with FGIDS everything we think might be useful, a significant number will get better' on the basis of this study.


It's interesting to look at this study, evaluating a low FODMAP diet for older adults, something just some of the people in the FGID study were taught about.
A Low FODMAP Diet Is Nutritionally Adequate and Therapeutically Efficacious in Community Dwelling Older Adults with Chronic Diarrhoea

The people in the FODMAP study seemed to have gastrointestinal symptoms that were far from normal at baseline, and yet their SAGIS score of 21 at baseline is not much different to the SAGIS score reported for the participants of this study after this extended period of being offered all sorts of therapies (the mean was about 18).

So, the clinic in the FGID study gave patients a wide range of treatments including psychoeducation, surely at considerable expense, and the mean reported SAGIS score after treatment was comparable to the baseline condition of a cohort of older adults with chronic diarrhoea bad enough to have first undergone a colonoscopy. This "Multidisciplinary Integrated Care" is clearly a long way from a cure for "FGIDs".