Trial Report Stellate Ganglion Block reduces symptoms of SARS-CoV-2-induced ME/CFS: A prospective cohort pilot study, 2024, Duricka & Liu

I think a balance should be struck between skepticism, which is the essence of scientific rigour, and cynicism which can hold back scientific progress.

We should be skeptical of this paper. It is an n=10 uncontrolled, open label trial conducted by a conflicted person so we cannot be certain of the reliability of any of its results. It is partly based on subjective reports of fatigue, which is open to bias - but we can compare the level of bias to other trials.

We should also be skeptical of arguments that these results are due to the placebo effect (and similar effects). I want to point out a few features that suggest these results aren't purely placebo.


Firstly, the effect size is huge. The worst patient went from a subjectively rated physical score of 5 (meaning they answered "limited a lot" for all but one question on physical limitations) to 55 (they answered "limited a little" for all but one question and "not limited at all" to one question) after 2 weeks. After 2 months, that patient further improved to 85 (answered "not at all limited" for 7 questions, "limited a little" for 3 questions and "limited a lot" for 1 question). These questions include bathing yourself, the ability to walk a mile, playing golf and doing vigorous exercise.

So someone who couldn't climb one flight of stairs is now walking a mile and doing mild exercise with no problem, but probably can't do vigorous exercise. This is a dramatic recovery. This is a recovery of 80 points on a 100 point scale.

All the other most severely affected patients also recorded very large jumps in physical score and in vitality score. All but one patient (who recorded the worst physical score) recorded large jumps in social score (and all improved). It's over-egging it to call them cured (as the paper does) but its definitely a miracle level improvement. Indeed, the fact the paper describes them as 'cured' is probably the most misleading thing in the paper. It uses the same trick as PACE, setting a definition of cured that meant that 40% of them had "cured" levels of physical fatigue before the trial started!


Secondly, this effect size dwarfs the effects seen in other biased studies. Consider this retrospective, uncontrolled and open label "trial" of Abilify. Among responders to Abilify, the symptom of fatigue fell from 5.76 to 2.86 (on a 10 point scale) on average. Among both responders and non-responders, the symptoms from from 5.78 to 3.50 on average on a top point scale. By contrast, the SGB improved patients from an average physical score of 39 to 79. If we invert the Abilify scale for all patients, this is a 54% improvement on the baseline compared to a 102% improvement from SGB.

In other words, SGB was twice as effective as Abilify (noting that the latter had n=86 compared to n=10 for SGB) and all patients were responders (unlike Abilify where only 77% responded, again noting the small sample size).


Thirdly, objective improvements matched the subjective improvement in physical and social function, and vitality. There were significant improvements in POTS and cognitive impairment that were objectively measured.

For completeness, I do not that some subjectively reported improvement was not matched by objective measurement. Sleep quality and some cognitive impairment improvements were subjectively reported that did not quite match the objective data.


Fourthly, PEM improved (but only disappeared in one patient). If a placebo can get rid of my PEM, then sign me up.

For these reasons, we cannot be certain this is producing a true result. It would have been nice if OMF had ensured the protocol for this trial was more rigorous prior to funding it though. But it seems promising enough to test in a proper RCT if a procedure can be devised. Let's hope OMF actually enforces some rigour next time.

 

You are right that we need a balanced attitude towards trials like this and notes of caution are always welcome. However, I think the general view expressed in the thread does get the balance right.

This study was looking at post-Covid 19 illness rather than typical ME/CFS. Post-covid symptoms are highly likely to resolve over a two month period (mine did on both occasions) so we should not be surprised by major changes. These are not 'effects' although I realise that 'effect size' has become embedded in stats jargon in a way that does not actually mean effect.

There are all sorts of other reasons for not comparing this trial with the Abilify study. As Hutan points out stellate ganglion block is a sort of 'perfect placebo' in that it is a dramatic interventional procedure. It is maybe comparable to rituximab infusions, which we have seen associated with very major improvements, not confirmed to be specific drug effects.

We have had discussions about placebo effects and its narrower and wider definitions. In terms the wider definition, which includes improvements due to any cause other than a specific biological effect of a treatment agent, this looks very unremarkable to me.

Are you familiar with the people who use stellate ganglion block in other contexts? Are you involved in trials? I ask because I have spent my career involved in trials and in departments with colleagues who use procedures like this. I have never met a physician who uses stellate ganglion block who has any sense of evidence rigour.

I have sat in departmental seminars where colleagues have extolled treatments like this not just with no reliable evidence but with clear dishonesty. I remember one colleague describing to the audience a trial mine in which 60% of patients showed improvement, giving that as evidence for the validity of the treatment. This despite the fact that the placebo control group had shown slightly more than 60% and we had published the trial as clearly negative.

I hate to sound cynical but I think there are times when to maintain 'balance' we need to remember that a man walking down a street trying the door handles on all the cars is unlikely to be an innocent passer by. That may sound harsh in this context but when people are subjecting patients to procedures with very real dangers, including death, being realistic is important.

(I am not sure that the POTS and cognitive measures can really be called 'objective' either. There is likely Mohave been plenty of room for bias to creep in.)

 

It’s the authors responsibility to limit bias as much as possible, and to use adequate control groups to demonstrate what was placebo, and what was not.

 

Anecdotal comment on Reddit —

 

Do we have any evidence that? I wasn't aware of any good evidence that it did anything much. The physiological block effect ought to last about two hours almost.

 

A lot of people I know who claim benefit from it only did the "treatment" once they had "improved enough to make the trip" which sounds like a major confounding factor, if they are doing this during their "natural recovery process".