David Tuller: Trial By Error: My Letter about MUS to the British Journal of General Practice

Mistakes are generally quickly corrected. Nobody wants mistakes to remain, especially if they affect policy. This one is very easy and quick to correct and of particular importance.

Deceit, on the other hand, that usually meets some resistance. It requires a lot of scrambling behind the scenes to avoid being caught red-handed.

It's pretty safe to lean on option #2. It's not easy to create a narrative when facts disagree. There is a lot of weight behind that narrative, many careers, promises made that cannot be delivered.

 

So today I received the following e-mail from Professor Roger Jones, editor of British Journal of General Practice. I've been bugging him to correct a false statement in a 2017 editorial. I sent him another nudge on Sunday, given that it had been more than a week since he promised to make the correction. He has let me know that it is in process and should be in place by the end of June.

Here's what he wrote:

"Don’t worry, this is all in hand. It takes a while because it’s not a simple case of updating a webpage; we rely on third parties to update the xml and reprocess the article, and Pubmed etc will only make the correction once the print notice has been published. The print notice will appear in the July BJGP, and the online correction will be completed before the end of this month."

Thanks to Professor Jones for taking responsibility and doing the right thing!

 

Trial By Error: BJGP Correction Coming, BMJ Still Stonewalls

Making corrections when needed is part of the job description of being an editor. Yet as I have come to learn, some editors at major UK journals do not appear to take that part of the job description seriously when it comes to these domains of inquiry. I very much appreciate that Professor Jones is choosing to act in this case. As I try to persuade other editors and authors to correct similar misquotations of the same 2010 article, it will be helpful to be able to point to Professor Jones’ positive example.

In contrast to the resolution of this matter with BJGP, the situation with the Lightning Process paper continues to defy all reason.

 

this is true and definitely good to remember. however, none of those involved had really earned this trust. nor did Professor Jones with his initial responses. I think being concerned about whether the journal really intended to follow through was warranted.

 

As a former web professional, this delay thinh is mostly bunk. Yes, it can take time to push through updates, but that is a choice when the consequences are trivial. It can be expedited easily and they simply choose not to. 99% of the delays are the involvement of human behavior, basically waiting for someone to do the thing. This can all be bypassed with relatively trivial effort.

Better than nothing, but the delays are entirely a matter of choice. Well, 99% of it anyway.

 

I'm a current web professional for a (different) scientific publisher, and I can absolutely believe his story. The terrifying state of the systems that these websites run on would send normal people screaming to hide under the bed. Plus his response suggests that they've outsourced the web production work, which IME makes for some pretty bad times on top. Their website will be a desperate omnishambles. I'm actually quite pleased to get these little glimpses behind the curtain because it reassures me that all the other scientific publishers are just as crap as my employer!

 

Oh with a bad system and no sense of urgency, yes. But you need a terrible system for that and an even worse process. Like, pathetically bad. The kind that is justifiable for a small company. BMJ is not a small company, it is very mature and has all the resources it needs to accomplish its core business, which is publishing. This is a publishing and process issue.

So, basically like an automotive glass business driving around in their van with a cardboard and plastic instead of fixing their broken glass. It's believable, but it shows utter incompetence.

 

When I was collaborating with Healthwise to remove GET and CBT from the Kaiser Permanente website, it took several months for the update to appear live.

Live content updates could only occur during prescheduled rollouts, and these were spaced several months apart.

On rollout, there were some glitches (such as GET and CBT still showing in searches), which had to wait for still another content rollout a couple months later to fix.

All in all, the Healthwise updates took about 6 months (not counting the 2 years it took for them to agree to the changes).

Waiting a month for an update doesn't seem unreasonable.

 

@dave30th. I got the impression at least that Professor Jones had taken the prompt in good humour. I think no harm done.

 

He he. Welcome to scientific publishing! I've got stories that'd make your hair fall out. Journals and their editors are one thing, but they can't do a damn thing to a website without going through the IT people first, and the IT people don't care in the least about the urgency of scientific corrections as its not their industry and has no bearing on their deliverables.

In an episode of Yes, Minister, Sir Ian Whitworth said, "it would be different if the government were a team, but in fact they are a loose confederation of warring tribes." This is also an excellent and accurate summary of what it's like to work for a large academic or scientific publisher!

 

yes, as did I. And I wrote him back thanking him for letting me know. I don't see a reason not to prompt these folks when things seem slow. They don't have a good track record of following up. I'm glad Professor Jones decided, in the end, to take the issue seriously and not continue to downplay it as ten-year-old data.

 

I don't know what BMJ's specific reasons are, but if I had to speculate I'd say it's because it's grunt work, and it needs loads of people to do it. The tools to do the updates are usually very old and inefficient, and it's difficult for managers in charge of that work to build up enough political capital to get improvements to their tools and processes, because it's grunt work. Then the suits start to look askance at the huge number of people working in these inefficient departments and decide it'd be much cheaper and better to outsource it all somewhere else instead. This doesn't solve the problem, and creates a bunch of new problems, but by the time that becomes apparent, the suits responsible for the decision have buggered off to a different company to wreak their havoc there instead.

But like I say, this is speculation! I don't know what it's like at the BMJ or if they ever had an in-house web production team. I wouldn't be surprised if this little piece of speculative fiction is reasonably close to life, though.

 

Work expands to fill the time available.

The main purpose of the editorial team is to introduce typos into your manuscript for you to hunt down. The main purpose of an automated manuscript progressing system is to crash so that you have to have an assistant to deal with all the fallout.

Which is why 'page charges' run at $2000 a manuscript these days.

 

In fairness quite a bit of the delay could be more bureaucratic than technical, given there are third parties being relied upon to implement things. I expect the bureaucracy card might be getting a bit over played, but I think benefit of the doubt might be in order here ... providing the promises are delivered.