BMJ Archives of Diseases in Childhood: ''Editor’s note on correction to Crawley et al. (2018)'', 2019, Nick Brown. (SMILE LP Trial)

BMJ has "corrected" the Lightning Process study but has allowed its conclusions to stand, as far as I can tell. I will have more to say about this. Apparently, the authors have convinced the journal that the outcome swapping had nothing to do with the fact that the revised primary outcome had positive results: "The process has additionally involved seeking assurance from the authors that the change in primary outcome was not influenced by (positive) findings in the feasibility phase."

Everything was an innocent mistake after all!
https://adc.bmj.com/content/early/2019/07/11/archdischild-2017-313375ednote?fbclid=IwAR3-LBW7tdYPwycVC0p0ozmkjdT6bzMx9nz0XVrrfPk97Rp59e1vafba3P0

 

It shouldn't matter whether or not the editor believes it was all an innocent mistake, what matters is that the BMJ should never have published it.

"The process has additionally involved seeking assurance from the authors that the change in primary outcome was not influenced by (positive) findings in the feasibility phase."

Would a pharmaceutical company be allowed to offer such a gentlemanly assurance?

 

wow! what kind of an insane person would devote that kind of energy on such a piece of garbage??

 

How exactly have the BMJ issued a correction in this case?

Which protocol now appears as the original one on the paper and how is history revision permitted in such circumstances?

If they have outlined the correction to show the original protocol doesn't that change the "positive results" when matched to the outcome swapping?

 

The right kind?

 

The scientific method seems now to be the method by which papers come to be published in "scientific" journals.

 

It looks like the original version of the article is still available on PMC, and I archived a copy here in case it's useful: https://web.archive.org/web/20190711172758/https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5865512/

I tried to compare the text here ( bit of a mess - I had trouble getting this to work and I didn't compare the tables - anyone want to try to clean things up? I fear my poor attempt could lead to people missing important changes): https://www.diffchecker.com/AvCo4xsy

Edit: I'm an idiot and wasted my time - they listed the changes here - what follows is just the bits I pulled out on my own so I'd recommend reading their info: https://adc.bmj.com/content/early/2019/07/11/archdischild-2017-313375corr1

It looks like they removed this:

Added stuff:

Finally we get info on their original primary outcome... they didn't have 'capacity' to collect it:

Are they planning to correct the ISRCTN (they got the acronym wrong) page which still falsely claims that the trial was prospectively registered?

 

This is pathetic. How is any of this seriously worth damaging BMJ's reputation like this? I know they expect no consequences for it but inevitably there will be. This would never be tolerated in a non-controversial disease and once the veil of controversy is lifted there will be hard questions.

Basically BMJ's position, like PLOS's position a few years earlier, is that their rules have a secret "or whatever" clause where things that are required can be waived off for inexplicable reasons. Those rules are subject to arbitrary suspension whenever someone feels like it. Just a suggestion, no big deal. Scientific method, shientific shmethod.

The whole study becomes even more meaningless given this. And it remains, its extraordinary conclusions placed on the same level as any large double-blinded placebo-controlled trial, because... reasons. The political demand for this ideology is something to watch. No matter the cost, no matter the consequences, this insane project will be bullied through.

 

So this stands for NICE' s consideration. Wonderful....

 

Just to point out that there's a note listing the changes here (that I'd previously missed): https://adc.bmj.com/content/early/2019/07/11/archdischild-2017-313375corr1

 

Well this is the thing, at what point does a government body decide to consider whether private publications like the BMJ and the Lancet need to be adjudicated over for their own bias?

Or do they go on forever and a day just ignoring that private publications can get things appallingly wrong and then just dig their heels in knowing that they can hide behind, "governments should not interfere with science."

What happens when private publications have made a complete hash of science and realise they have to double down, obfuscate, deny, ignore and worse simply to protect their own reputation.

What happens when they are just incompetent idiots?

Do government agencies that affect public health care just have to ignore the elephant in the room and claim they have to issue advice based on the "scientific evidence" published in the private domain especially when that evidence has been bought and paid for with public funds.

Its such a cluster fuck of the public paying for their own abuse via collection of public funds by the government who are supposedly there to serve the people.

It's pathetic when it becomes even clearer that the, "scientific outcomes" often start out as political desires.

Its an unmerry-go-round.

 

One of the reasons being the changed protocol decided not to use the objective school records as a measure of attendance and just asked the kids if they had attended or not and never matched the two.

 

O. M. G. This is a car crash. A mega-wreck. Just look at it! LOOK AT IT!!!!

They've been allowed to correct it to admit to all sorts of shenanigans. And still keep the same Conclusions. Flippin 'eck!

Cross-linking with my analysis on the other thread.

I'm going to look at this in more detail, and judging by the timeline from the editor's note, they only released the data once re-publication was assured. It should have "expressions of concern" all over it. Like my face at the moment.

lol

 

I always reckon you have to be a little bit mad to stay sane

 

From the editor's note;

Would be interesting to see these reviews as well as the "further unpublished work" that "suggests [school attendance using school records] highly correlated with the self report measure we used" and "could have provided an objective outcome" if only we had the "capacity to check" this. (EC & et al 2018)

(Edited to remove some muddled thoughts).