Monday, 30 June 2014

GPs spotting Cancer

From the BBC

GPs with a poor record in spotting signs of cancer could be publicly named under new government plans.

Health Secretary Jeremy Hunt wants to expose doctors whose failure to spot cancer may delay sending patients for potentially life-saving scans.

Labour called the idea "desperate" and accused Mr Hunt of attacking doctors.

The Royal College of GPs said it would be a "crude" system and one that could lead to GPs sending people to specialists indiscriminately.

It warned this could result in flooding hospitals with healthy people.

Which is, actually, what we should be doing. Why go to the monkey, instead of the organ grinder?

I caught a few minutes of the Jeremy Vine show where a former cancer sufferer and someone from the BMA were on and talking about how GPs thought her cancer was IBS, or gallstones. When she got to see a pancreatic specialist, he immediately rushed her into surgery. And it turns out that part of the problem is that some cancers in some areas are very rare, which means that GPs, along with handing out antibiotics for chest infections and prescriptions for the pill and signing passport forms are expected to spot that.

Let's make a comparison with the software industry: when you build a large website, you typically have a range of people involved. You have a graphic designer who makes it pretty, a web designer who takes the design and makes the HTML and CSS code, an accessibility specialist who makes it work fine for disabled people, some internationalisation specialists who make it work with foreign languages and deal with things like building sites for Israel where they are left to right. Then you've got Javascript specialists who make the site do clever things when you press a button. Then engineering guys on the back end that build the code that runs on the servers that checks your password and accepts your orders. Security testers supporting them. And behind those are the database specialists who do the bit of actually storing stuff, and supporting them are the database administrators who make sure the databases are performing right. Oh, and server maintenance too.

I mostly work on the back end of things. I can do a bit of the web design stuff too. But what I know is that if I get a problem in that area, it's going to take a lot longer for me to fix say, something not appearing right in Internet Explorer 8 than a couple of friends that are full-time web designers and instantly know what the problem is. It's not because they're smarter than me - it's because they've been specially trained in those bugs or maybe seen it before, or their boss did, or something else.

15 years ago, you could do the whole job. There was 1 browser to worry about, no-one worried about disabled users, SEO or Russian bots that were hacking websites at random. As things got more complicated, we increased specialisation.

The point I'm trying to make is that for all the "GPs must do this better", that isn't going to work. GPs have a limited amount of time each day and limited experience in fields of medical diagnosis as they're expected to cover everything. They simply aren't going to get the training or experience that tells them that something that on the surface might be a gall stone might be a cancer. Because they're only seeing a few cases of what sounds like gall stones each week, and very few of those are cancer.

On the other hand, if all you do each day is look at possible gallstones, you can either be specially trained to look for cancers, or else, you'll gain the experience of symptoms that means you can make a rapid diagnosis.

So, the real answer here isn't to red flag cancers, it's to get rid of most GPs and train up specialists earlier in various areas of the body. People who can be specially trained in one area, people who can gain knowledge and experience in a particular area.


Kj said...

Although I'm all for reducing GPs as much as possible, being that they are generally useless, the question is where you start off. There is some early symptoms that can be anything from cancer to schizophrenia, and you have to have some first point of contact to narrow it down. Maybe some sort of initial level diagnosis specualist in the hospital sector. For early detection of cancers pre-symptoms, I expect that to be so common and cheap that you'll get it as freebies at any point of contact with health care.

Mark Wadsworth said...

"get rid of most GPs and train up specialists earlier in various areas of the body."

Yes, we've said this often enough before and this is another example of why.

mombers said...

The problem with this is exactly what they have in 'Murica. 'Defensive medicine' it's called, doctors do hundreds of unnecessary tests just to cover their asses, and result is double the cost of healthcare per capita for no quantifiable benefit in outcomes