Sabag Gots Da Rona

Advertisement
Yeah, I saw it this morning.

Two things. The study is not peer reviewed yet. So we have to wait for that. Also, I haven’t seen the study design.

Secondly, and just as importantly, there is no dexamthasone in any of the treatment arms, apparently. Remdesivir, at least in the US, is most frequently administered with dexamethasone.

But anyway, this illustrates the point that all these treatments are experimental.

As an aside, what’s interesting is that Covid mortality seems to be much less today than it was back in March/April. The question is...why?

The paper is linked in the article. I reviewed it.

1) Not sure why the article mentioned issues with study design. They used a randomized adaptive trial whose methodology allows just as it says in the paper to still develop an estimate of interest under the possibility of dropping and adding other drugs on trial. A study design issue would have been looking retrospectively or something else where temporal causality was less likely to be determined.

Unless ‘study design’ refers to population or something else which comes to point 2 below.

Also why is Gilead dismissing this result when this result, despite its lack of clear target may actually be better powered for its given outcome?

2) The protocol has mentioned in the paper did not power on a specific target group. That is concerning. We don’t know if they were looking among those who enrolled in a particular age group (only know age greater than or equal to 18 and we know if a lot of people are younger than 60 then the risk is far less) a particular class of severity of illness. Among the rem cases/comparator group I’d want to what proportion of those cases have a particular severity, what is the mean and variability of age, and/or what are the comorbidities/proportion of subjects with certain comorbidities. Without knowing a lot of those details, the rem-specific results could have a selection bias which, depending on the selection process could bias the results towards the null. Were the comparison groups at least well balanced? Let’s keep going.


3) So their table 1: ( not sure why it reports counts and no percentages for majority of counts) but among the rem ptx we see well balanced distributions among cases and controls by subgroups such as age, history of comorbidities, gender, etc. and very high compliance of treatment which are all good things for quantitative purposes and eventual interpretation. Those are all good things. I will say the large proportion of people are less than 50 whom we know to be much lower risk.

4) the methods for analyzing are okay. A little strange but okay. I would have done the time to event a little differently but I get what they did. I agree with how they stratified. I would definitely look at no ventilation status by age next as a suggestion if I was formally peer-reviewing. There is something there among the lower risk though the numbers may get low.

5)a lot of the meta analysis results are hilarious to me. The variability of most of those trials is so bad I wouldn’t look twice. The paper, however, is correct in saying there may be benefit to using rem among patients who are perceived as lower risk as per figure 4. You can’t outright deny that. Again see above: analyze among the no ventilator status by age. And for God’s sake look at age continuously, too.


6) my concerns:

A)I don’t know their target Population for powering on mortality. Rem looks barely protective on mortality with decent precision of the estimate. The study is very well balanced, though, and I’m not sure enrolling more on a more defined target would really help their result given its so close to the null.

B) Rem May have utility for lower risk patients (no to low ventilation) based on the Meta analysis and actually based on this trial.

Id be more interested in higher risk groups, though. Given what we know now, I would target a population of greater than 50 in age and not stratify by age. I would analyze by ventilator status, ethnicity, ***, and history of comorbidities as well as more detail on comorbidity. I’d be mainly interested in seeing among the most high at risk does rem reduce mortality or lengthen hospital stay? Right now what they did and how they analyzed appears okay. But its such a broad population and includes so many factors that we know are much less at risk for death etc that they should target helping those higher risk subgroups, instead. In fact they could probably look among the 50 and up only and re-analyze.

C) With so many countries represented, I’d want a lot more detail on ethnic makeup, and of course individual biology beyond population health characteristics (genetics, expression, etc) We can see that the highest proportion of patients do come from Asia and Africa. How representative is that distribution to the United States/European population distribution?

7) We are starting to know a lot more about the virus and the concerns are not all about mortality or the overfocus in discussions in this board. There are cohorts being followed of people who don’t die but don’t let go of the virus, either, and it causes neurological issues besides respiratorial issues.
This article by the Washington post points out nicely whats happening without getting into the research details of the long haulers, etc: https://www.google.com/amp/s/www.wa...a-bc45-e5d48ab44b9f_story.html?outputType=amp
 
Apparently all we needed to do to conquer corona was to make sure Saban got it. Diagnosed Wednesday, declared covid free on Saturday.
 
Advertisement
Advertisement
Back
Top