A New Year's Resolution

You choose!

Image of fireworks is from Wikimedia

My most recent post^⌘ concluded a series of 8, where I explored the obvious defects in modern scientific journals, with a focus on Medicine. We started rather provocatively with Sex, Lies and Robots,^⌘ followed by predatory journals,^⌘ papermills,^⌘ Hirsch hacking,^⌘ bad players,^⌘ the deceptions of drug companies,^⌘ and how AI has helped^⌘ to muck things up. And we ended with my speculation about fixes.^⌘

As I said then, there are several options for where we now take Mostly Wrong. I can see three that appeal:

1. I simply carry on writing about those things I enjoy: a mix of Medicine, Science and programming, with a smattering of philosophy and maths.

2. I do a deep dive into one possible solution for the “Journal Problem” I’ve identified in the past 8 posts.

3. I systematically work through my take on why computerisation of Medicine has been such a bust.

All of these appeal to me in one way or another. I am also acutely aware of the potential benefits and down side of each choice. So …

I’d like your opinion

I could simply ask you to vote for one of those three options now. And if you’re in a hurry, yep, you can skip to the bottom and vote. But in all fairness, I’ve decided to give you a bit more detail about what each choice implies. So you can decide based on the best information I can give you (and keep me honest).

1. More of the same?

The first option—just bash on—should be pretty obvious. If you’ve stuck with me for the past year or so, then you might simply want more of the same. That, I can deliver. I sit down every weekend, and write a piece inspired by stuff that I’ve encountered during the week. I throw in the odd mid-week post too, from time to time.

The other two options need a bit of fleshing out.

2. Journalocracy—a deep dive?

I’ve been aware of the ‘Journal problem’ for years. A bit more than a year ago, I even took some time off to explore a possible solution. From the past 8 posts, you’re now familiar with several of the issues; you also know that others have tried to fix things. Some of these fixes have even had moderate success.

My idea was to tie the successful bits together, and amp things up. Here’s what we need:

1. Wide coverage of both good and bad studies. Ideally, we should “mark” every research article.

2. Spread the word. Disseminate the mark-up of the bad and the good studies.

3. Get trustworthy expert review, with clarity about provenance, performance and conflicts of interest.

4. Join up all of the data: reference existing criticism.

5. Engage competent, enthusiastic people.

My idea was to ‘gamify’ the solution. I’m well aware of a multitude of ‘journal clubs’ around the world, where medical professionals sit down and criticise important journal articles. I’m sure that every day, thousands of insightful criticisms are made—and then lost.

So how might we do better? I thought “Doctors in particular tend to have competitive, Type A personalities, so why not reward good insights into study defects?” And who better to do this than other doctors, through participatory voting?

I made the database. I wrote the front end. It all worked. There are some catches. When I presented this ‘Journalocracy’, the universal response from my colleagues was “What a great idea! For someone else.” The problem is that this looks too much like hard work for most. Some of my design choices were likely poor, as well.

I’m happy to revitalise this application—and explain how it works, and indeed, explain my failings, if you’re interested. The flipside is that the details are pretty technical. It will be a deep dive. There is a lot of scutwork involved. And it failed.

3. Computerising Medicine

Another area that I’ve put a lot of effort into is understanding why attempts to ‘computerise’ Medicine have been such a bust. Humongous projects like NPfIT in the UK failed, wasting at least £14 billion. In the US, $29 billion was spent, resulting in widespread computerisation—but it decreased face-to-face time with patients, increased the workload of clinicians, and made them profoundly depressed (‘burnout’). There are many other examples of partial successes, abject failures, and projects that initially worked brilliantly, and then degraded.

There’s a multitude of things you need to get right for computerisation of Medicine to succeed. And guess what? We’ve got most of these wrong.

If you wish, I can work through my perception of what’s wrong, and how to fix things. I find this topic fascinating, but it’s also challenging to explain. I’ll give it my best shot. Not everyone will agree. I have some strong opinions here—but each one is logically supported, and open to scientific criticism.

It’s now your call

So where shall we go?

It’s very much up to you.

My 2c, Dr Jo.

^{⌘ This symbol identifies another post of mine where I explore in detail. Click on the link.}