roubaixtuesday wrote: mjr wrote:
Only 60 patients, many details subscription only, but a study design which excludes long-term accumulation effects and won't readily capture exercise-induced effects. The authors and journal should be ashamed of the damage they will do to patients' mental health with this.
Very odd comment.
How is it bad for mental health to understand more about what's going on?
I never said it was. Maybe that misinterpretation is why you think it an odd comment.
The damage to mental health will come from news reports about this study being used to browbeat some patients who experienced real side-effects into attempting statin therapy again, not only by medics (not the experts in hospitals and not even the better GPs, but there will be some GPs who are influenced by this), but by ordinary people, maybe family, who see this reported in the news and tell someone who came off statins when they suffered memory problems or could no longer stand up that it was all in their mind and they were imagining it all.
Many of those patients already have mental and/or physical scars from their experience with statins. I certainly do. I know I'm a bit excitable about this — but I'm not sure you'll find someone who's missing years of memory who doesn't find it a bit unjust and want others to avoid the same fate. It's not often people recover from such things.
The study looks weak/niche and the reporting seems worse. For example, the BBC headline is "Most statin problems caused by mysterious 'nocebo effect', study suggests" but I'm pretty sure that the authors weren't considering "most statin problems" (I see no analysis in the public materials of what problems have what prevalence, for example) rather than simply looking for any reported short-term side-effects, so most of that headline is the reporter's invention.
It's a small study with pretty tight selection criteria (described below - the 60 in the study were all that were used from a pool of about 300 potential participants). Could anyone really defend the BBC headline?
Sure, the study could be larger, wider in scope and longer, but nevertheless, it shows very clear results, obviously beyond the level of statistical significance.
The small sample mainly means that the study won't tell us much about "most statin problems" because many of the long list of known problems occur in 1% or less of patients, at least according to statin patient information leaflets.
But more than that, the design looks seriously flawed to me, both based on my education as a statistician and my experience as a patient. It can be scarily easy to get statistical significance for what is arguably a rewording of a null hypothesis (no effect) with a bad design. It's worth remembering that most of the classical statistical methods are based around proving that there probably wasn't no effect (rejecting the null hypothesis), so it's hotly debated how well they work for actually finding evidence for no effect.
On top of that, the participant selection criteria (Appendix S1) only includes patients who withdrew from statins because of perceived side effects developed within 2 weeks of starting taking them. The exclusion criteria also excludes anyone who suffered side effects taking more than 2 weeks to present, as well as a list of the most serious side effects. So people like me already wouldn't be in this study. There was also a catch-all option to exclude anyone the study doctor judged should not be enrolled.
It reads like patients had to go to Hammersmith Hospital twice. That could also bias the study. Of the 222 of 308 possible participants who declined to attend, 24 gave travel as an explicit reason, but I suspect it may have been a factor in the 31 "too busy", the 22 who gave no reason and some of the others like "anxiety about travelling to site". It also seems a bit suspicious to me that no-one declined due to it being a smartphone-app-based study, which makes me wonder what was in the recruitment adverts.
The drug and dose was Atorvastatin 20mg, which means that it's probably limited to medium-or-higher risk patients, not the far larger population taking a 5 or 10mg dose daily. It's stated elsewhere that their average LDL-C is just 4.16mmol/L.
And it's downright curious to me that none of the list of adverse events look like typical oft-reported statin problems that I've heard about at patient forums. Did the participant selection criteria effectively exclude those who suffer real statin side-effects?
There are more questions than answers... but the people who will promote this study hardest don't seem to be interested in the questions, only their preferred answer.