Read on Twitter
Short thread on evidence for steroids in #ARDS or lung crit care before #RECOVERYtrial: some discussion on #MedTwitter that "steroids make everything better" or "we already know steroids help" that downplays the impact of this trial. What did we know before? 
1/n
1/n
UK @ics_updates + Faculty of Intensive Care Medicine #FICM ARDS guidelines 2018: Steroids do NOT reduce
hosp mortality RR 0.62 (0.23 - 1.26). 5 studies/n=560 pts Wide CI, possibility of substantial benefit but also c/w harm 2/n
https://www.ics.ac.uk/ICS/ICS/GuidelinesAndStandards/CollaborativeGuidance.aspx
hosp mortality RR 0.62 (0.23 - 1.26). 5 studies/n=560 pts Wide CI, possibility of substantial benefit but also c/w harm 2/n
https://www.ics.ac.uk/ICS/ICS/GuidelinesAndStandards/CollaborativeGuidance.aspx
@bmj_latest 2008 meta-analysis: Established ARDS (5 studies, n=571 pts). Steroids did NOT reduce mortality OR 0.62 (0.23 - 1.26), possible substantial benefit cannot exclude risk moderate harm. 3/n
https://doi.org/10.1136/bmj.39537.939039.BE
https://doi.org/10.1136/bmj.39537.939039.BE
Note apparent protective effect driven by smallest outlier study Meduri 1998 (just n=24, bolus methylpred 2mg/kd/d, weaning dose over 32 days), novel at that time but tiny study not powered for OS 4/n
@Cochranecollab MA 2019: (6 studies, n=574 pts, only 2/6 overlap w BMJ). Established ARDS - steroids did NOT reduce all-cause mortality RR 0.77 (0.57 - 1.05): possible moderate effect, again cannot exclude small risk of harm. 5/n
https://www.cochranelibrary.com/cdsr/doi/10.1002/14651858.CD004477.pub3/full
https://www.cochranelibrary.com/cdsr/doi/10.1002/14651858.CD004477.pub3/full
Again, apparent benefit pulls from smallest studies 25% pts (144/574), including a diff Meduri study (2007) + 2 smaller studies. Meduri - methylpred infusion 1mg/kg/d x28d, powered for surrogates (lung injury score) not OS 6/n
@ANZICS @JAMyburgh @NEJM ADRENAL trial 2018 large modern RCT n=3800 ventilated pts w septic shock (1/3 pulmonary sepsis) hydrocort 200mg/day 7 days did NOT reduce 90 day mortality OR 0.95 (0.82 - 1.10) 7/n https://www.nejm.org/doi/full/10.1056/NEJMoa1705835
Leads us to the #RECOVERYtrial: Dex reduced 28d mortality in vent pts RR 0.65 (0.48-0.88), pts req O2 only RR 0.80 (0.67-0.96). No benefit pts not req resp support. @MartinLandray @Oxford_NDPH @NIHRresearch @OxHealthBRC @wellcometrust @gatesfoundation
https://www.recoverytrial.net/ 8/n
https://www.recoverytrial.net/ 8/n
RECOVERY trial protocol available online now, analysis plan too. https://www.recoverytrial.net/files/recovery-protocol-v6-0-2020-05-14.pdf
Of course need to see data, but this is practice-changing now, great thread by @VPrasadMDMPH on why this press release differs from so many others 9/n https://twitter.com/VPrasadMDMPH/status/1273028969718337536
Of course need to see data, but this is practice-changing now, great thread by @VPrasadMDMPH on why this press release differs from so many others 9/n https://twitter.com/VPrasadMDMPH/status/1273028969718337536
There's been a tantalising promise of benefit from steroids in ARDS / pulm critical care data since the 1990s, but nothing definitive to show a benefit. #RECOVERYtrial changes that - for patients with #COVID19 10/n
Does #COVID19 induce a uniquely inflammatory state that is more responsive to steroids? Is dex different to other steroids (mostly hydrocort or methylpred)? Interesting hypothesis generating Qs - but empiric data should change practice. 11/n
Disclaimer: med onc doc here, but keen student of trials, methodology & evidence base in medicine. Jump in with your thoughts & comments. 12/n
Was this helpful?
Was this helpful?
