Fernando Zampieri
@f_g_zampieri
Followers
5K
Following
20K
Statuses
9K
Critical Care Physician. Opinions are my own. Account seldom active. Older tweets do not reflect current opinions.
Edmonton, Alberta
Joined February 2016
Todos os telefones caem em sistema automatizado. Moro fora do Brasil há mais de um ano e não consigo ligar para o 0800. O telefone para quem está fora do Brasil sequer tem uma opção de falar com atendente.
1
0
2
@eemoin Really good when it starts coming together, too cumbersome early on. I find using {dtplyr} to be useful; writing in tidyverse but all data.table code on backend, so you benefit from speed. Also trying to use mode dara.table. Speed is amazing.
1
0
0
RT @CritCareReviews: Last Friday, @f_g_zampieri & @caiodeamtavares presented the results of the #DEFENDERtrial at #CCR24 Paper - @JAMA_cur…
0
7
0
@DrToddLee @PulmCrit Hence running a trial on recovery after ICU admission would be challenging. Feasible, but hard. This could be a nested trial within a larger trial perhaps?
0
0
0
@PulmCrit I feel that a dedicated large trial of SGLT2i for AKI prevention, etc I'm critically ill is definitely needed.
2
2
15
@arthur_alb1 @andrewheiss @mjskay @f2harrell I am definetly not skilled enough to follow many of Harrell's great papers. Will try to have a look, thanks! I have issues reproducing and understanding myself the longitudinal ordinal model myself....
1
0
0
@fluidloading @CritCareReviews Getting placebo is hard and expensive. Ideally should be supplied by manufacturer and they were not interested when we first discussed.... A limitation, for sure.
1
0
0
You may also wonder why not using a proportional odds model for this, and we indeed can! Using the same scale of -2 to 28, the odds for higher (better value) with dapagliflozin is 1.12 (CI 0.82-1.53, p = 0.81). This was not adjusted for site.
1
0
0
The average marginal difference is 0.36, 95% HDI -1.07 to 1.74, hence a pretty neutral result. (+)
1
0
1
Pretty clear that most mass is around 0. Probability of benefit of dapa should be around ~70%. Note that this is cool because if you are doing another trial, a difference in DAF of one may be a good target; it is very much unlikely the effect is above 2 days. (+)
1
0
2