Recruiting: A Study of a Positive Emotion Intervention for the Treatment of Long COVID-19 Symptoms, University of California, Lopez

I'll play.
There isn't much detail at the Clinical Trials site.
https://clinicaltrials.gov/ct2/show/NCT05676008


1. waitlist-controlled clinical trial
A wait-list control is not appropriate for reducing the impact of the placebo effect, as people on a wait-list can experience a nocebo effect.

2. vague selection criteria, resulting in heterogeneous sample
People only need to have had a Covid-19 infection more than 3 months ago, be suffering from feeling unwell and have as few as one persistent symptom. The possible symptoms range from coughing to having trouble sleeping. There does not seem to be any requirement for the Covid19 diagnosis or the persistent symptom(s) to be validated by a doctor. Therefore the samples will be poorly characterised. There is potential for later subsetting of the participants, so there might be a finding e.g. that women with persistent coughing benefited from the intervention.

3. biased volunteer selection
As the treatment is delivered online and the trial has been advertised, it is likely that the study will mainly attract people who think that participating in mindfulness training is a useful thing to do. It is unlikely that people who think mindfulness training won't help them will volunteer to be part of the study (unless a whole lot of Science for ME members decide to volunteer to participate ). Therefore, the placebo effect of the intervention and the nocebo effect of the wait-list is increased - there's a big risk of bias.

4. unblinded (although it claims to be 'single masked') combined with subjective primary outcome
The common and major problem of a lack of blinding together with subjective outcome. This maximises the placebo effect i.e. reporting bias resulting from an expectation of a useful effect.

5. large number of primary and secondary outcomes - all subjective,
and 6. no specification of what success is
There are 6 secondary outcomes, assessed at three different times. So, together with the primary outcome, that is 19 possible chances for a positive outcome, before subsetting the sample and subsetting the outcomes, and before different ways of analysing the results.

On the latter, in the protocols that I have seen, there is no specificity about what counts as a meaningful difference in an outcome.
So, if the average absolute endpoint score in the intervention group is 20 points, for example, while the score in the wait-list group is 17 - has the intervention achieved something?
Or, will attention be paid to the percentage of individuals in each group who have a score that is abnormal e.g. 30% of people in the treatment group are deemed to have depression because they have a score over a certain number, while 40% of the waitlist group have depression. If so, what is the diagnostic threshold?
Alternatively, will attention be paid to the change in scores at the individual level, relative to the baseline? So, if the average improvement relative to the baseline was 4 points in the intervention group while the average improvement relative to baseline was 2 points in the intervention group, is that success? It's not even clear that the surveys will be administered prior to treatment.

Will they attempt to correlate wellbeing outcomes with self-reported ongoing practice of the micro-dosed mindfulness?

The lack of specificity about how the data will be analysed creates extremely fertile ground for cherry-picking. In itself, including a substantial number of measures isn't the problem, but there should be some pre-specification of what combination of outcomes will demonstrate success.

7. very short assessment period for the primary outcome
The primary outcome is assessed at the one month mark. This maximises the likelihood that a benefit will be reported that is just the result of the placebo effect, and/or a result that does not last.


8. assumption that the emotions the participants are experiencing require fixing
There seems to be the assumption that the participants are not already applying good coping mechanisms, and that experiencing negative emotions is something that needs to be corrected. People who experience negative emotions might be the ones who apply pressure to make things better, for themselves and others.

9. inadequate description of the intervention
It isn't clear what the people will be taught - is it mindfulness (as in awareness of the reality around one) or is it the cultivation of positive thinking? The nature of the intervention will determine the potential harms that need to be considered.

10. Inadequate consideration of potential harms
This includes the impact of the intervention on the people in the trial (e.g. does the cultivation of positive thinking result in acceptance of difficult symptoms that makes it less likely that the person seeks investigations that might identify a treatable medical condition? does the intervention make people feel they have failed when they still want to stop having disabling symptoms?). It also includes the potential for the poorly designed study to suggest an ineffective practice is useful, and for this intervention to therefore be seen as having delivered a treatment for Long Covid, decreasing pressure for research into useful treatments.

11. Research design doesn't fully address the stated research question

There's no evidence that they will actually monitor whether the training does result in more patients self-microdosing mindfulness, or in individual patients having a higher frequency of microdosed mindfulness each day. None of the outcomes actually measure that. There's no pre-treatment baseline measurement. If the treatment aims to increase the percentage of people taking a moment to appreciate the sunset, or smell a flower, or enjoy the taste of a cup of tea for example, how will they know that people weren't doing that already? If the treatment aims to increase the number of times that people take such moments in a day, how will they know whether the frequency has increased or not?

Because they have no way of knowing whether the (micro)dosing has changed, they won't be able to say that any identified benefits are due to an increase in the practice of mindfulness.

Gosh, I got there. I didn't think I would get eleven mistakes. I didn't even need to use this one:

12. Poor grammar

"that can improve well-being on patients"
"patients suffering of PASC-related symptoms"

 

I don't know, 5-15 seconds in one hit might be over-dosing. This is powerful stuff we are messing with here. Best to start on a few micro seconds per hour.

STAT!

The ultimate proof against mind-over-matter is that we die. Matter (biology) always wins in the long run.

You are more polite than I.

 

Umm, I make that "A WEE", not "AWE".

But I guess "The Power of A Wee" didn't have quite the same ring to it for the title of a self-help book.

 

I was having similar thoughts in relation to S4ME as a safe and sane social media site while listening to a BBC Newscast podcast today in which 3 senior female BBC journalists shared their experiences of receiving awful stuff on social media as part of their job and how they deal with it.

 

Oh, now I feel bad about posting a puerile comment about someone's acronym. Not quite bad enough to delete it.