The #medtwitter in the US has exploded with cynical and sometimes downright malicious voices after the #RECOVERYtrial announcement. I agree (as does the Trial steering committee based on the press release!) that policy change and implementation should only come after publication.
There are very important data omitted from the press release&the scientific community has every right to demand to see those in a peer-reviewed publication, which will certainly be dissected word-by-word. As a researcher leading this and other RCTs in my Health Board I'm biased.
I haven't seen the data as I'm only a contributor. What I'd like to highlight is the potential differences between UK and US healthcare and ICU in particular, to help understanding the numbers in the study.
There are concerns, that the mortality in the ventilated patients was "too high". The mortality figures in the press release were very close to the @ICNARC reports icnarc.org/DataServices/A… when you look at the overall ICU mortality: 41.6% @ICNARC vs 41% standard care in the trial
However, the patients who were mechanically ventilated had in fact 9% lower mortality in the standard care arm of #RECOVERYtrial at 41% vs 50.1% in the @ICNARC database (Table 12 and 13 in the report).
This own it's own is significant: potential further data to suggest that participating in an RCT is beneficial for the patients? Or is there a significant centre-effect? #RECOVERYtrial recruited from a wide base, but what happened to those and why, who didn't get in?
Back to mortality: it is high, but it is the best universal data available. the UK NHS was not overwhelmed, like some of the Italian and US hospitals were, but there was a big stretch on the system.
Also, lockdown meant patients presented relatively late, many toughed it out at home, as we have learnt it from the @CCPUKstudy the median time to hospital admission from symptom onset was 10 days. By the time they got to critical care their disease could have been well advanced.
ICU bed numbers are also significantly different in the UK vs US, even with the 200-300% increase in bed numbers in the UK just passed the lower IQR maybe scraped the median of ICU bed numbers/100000 population in the US. Also sepsis mortality is about 40% in the UK ICU @msh_manu
Dexamethasone was the first arm which could have been open in many places as it's an ubiquitous drug, so there is a possibility that more patients were randomised into this arm early into the pandemic, when mortality might have been higher as we didn't know the disease.
These could all influence survival in the context of the pandemic and ICU care.
There were other concerns about the fact it was an open label study and there could be spillover to other arms. This will come from the paper, hopefully. Locally, we simply didn't have time to do any "spillover" as we were busy keeping the patients alive.
The UK in general didn't jump on any fancy bandwagon, hence we could deliver this study. Again, locally we adopted a strict "no intervention outside of an RCT" stance and talking to colleagues in research active centres, this seems to be the norm.
I don't want to comment on the rights or wrongs of press release before publication. It's such a thorny issue, which brings up so many problems with current medical journalism, media and public perception, that I can't possibly think about all of them.
