Computer glitch has led to incorrect advice on statins | Hacker News Reader

Computer glitch has led to incorrect advice on statins

newscientist.com

32 pointsgulda18 comments

"About one in three UK family doctors’ practices use a computer program made by the IT company TPP" instead of online calculators such as www.qrisk.org

"Last month the firm told the UK government that there is a problem with the software, and it has been overstating the risk for some people, while understating it for others. The government informed GPs of this on Wednesday."

Statins are controversially prescribed to people deemed at risk of illnesses such as stroke and heart disease e.g. if they exhibit risk factors such as smoking, or family history of these diseases.

"The software has been used since 2009."

There's some political "flavour" in the last half of the article about time-pressures faced by GPs, targets, guidelines, and the fact that risk-factors should not be used as the sole justification for prescribing statins.

Can't find anything on "TPP".

Non-paywalled versions of the story:

http://www.telegraph.co.uk/news/2016/05/11/statins-glitch-me...

http://www.bbc.co.uk/news/health-36274791

http://www.tpp-uk.com/

They apparently make the god-awful "SystmOne" system that my local GP uses for appointments.

There are some pretty crazy stories online about what TPP is like internally.

https://www.glassdoor.co.uk/Reviews/Employee-Review-The-Phoe...

I guess it shouldn't come as much of a surprise that a large healthcare software company is horrifyingly dysfunctional.

Should I get off statins?

You can get a coronary calcium scan to see if you actually have arterial plaques.

Big point.

It seems the doctors (As in none????) didn't notice it was making mistakes. All of them, not noticing since 2009.

The glitchy program still was better than the doctors knowledge base.

I'd pick the program anyday. Imagine the failure rate of the 2/3 who don't use the program.

How about not relying on a Dr. or a program and taking your health care into your own hands? Research and review and ask questions and don't just take w/o questioning the advice either give?

Statins have serious side effects. I would never take them without doing my own research.

EDIT: I don't mean "research" in the scientific sense, but in the "look into it" sense.

This method needs to be compared in effectiveness against the above 2 approaches.

Assume I'm a run of the mill dude without a biomedical background. I decide to read up on the current literature relating to statins. After going through these highly technical studies I find a conflict between what the doctor/program is telling me and what I thought I read.

Now what are the odds I'm right and the doctor/program are wrong?

You have the great advantage of all of the information. That clear advantage will greatly benefit you.

EDIT: by "all the information" I mean the sum total of all your health information. Even if you are at an HMO (e.g., Kaiser), not everything is written in your record and not every Dr. will read your entire record.

Also, only you can order a 2nd opinion, based on the facts you know. A Dr. will never order a 2nd opinion. I have literally never heard of this, but if it does happen, it is exceedingly rare.

High

"glitch" is such an awful, handwavey word. Please, describe the cause in a more intelligible way than by using this stupid word.

Seriously... It seems like a convenient way for journalists to avoid having to explain how technology works. At the same time I keep technology looking like magic and not science. "Glitch" makes it sound like bugs are caused by cosmic rays or something.

From the Reagan administration: "Mistakes were made."

The program by TPP connects to a medical risk calculator called QRISK2. Apparently they(TPP) have some error in this interface.

No Agenda?

Hopefully someone will run the numbers from this. Effectively it is a large scale test of outcomes for patients with equal risk factors, some treated with statins and some untreated.

I had the exact same thought--it's unfortunate that this happened, but some good might come of it as a randomized control trial...

This bug is worrying. EU Spreadsheet Risk Interest Group (EUSPRIG) released a report in 2006 about spreadsheet errors in medicine, which some people might be interested in.

http://www.eusprig.org/2006/spreadsheets-in-clinical-medicin...

Reply on news.ycombinator.com