The Anonymous Widower

Thoughts On The COVID-19 Testing

I must first congratulate all those involved in organising and carrying-out the tests.

As someone, who has analysed many large databases for patterns of perhaps marketing information, product recalls or criminal activity, 100,000 tests per day or million in ten days, is a very large amount of hopefully reliable data, that I believe can be used to answer a lot of relevant questions about the progress of this pandemic and our very boring (For me, at least!) lockdown.

I hope, that the tests collect all the right data to go along with the physical data.

But I suspect that some important scientifically-correct questions won’t be asked. For instance.

  • What is your place of birth?
  • What is your BMI?
  • How much exercise do you do every day?
  • What is your religion?
  • How often do you attend a religious service?
  • How many in your household?
  • How many generations in your household?
  • Do you have a pet that needs exercise?
  • Do you have any drug habits?
  • Do you have any allergies?
  • Do you smoke?
  • How much alcohol do you drink?
  • Are you vegetarian?

Only by collecting a full database alongside the testing process, will we get maximum value out of the testing.

May 2, 2020 Posted by | Computing, Health | , , , , | 4 Comments

Phone Call Cuts Hospital Readmissions

The title of this post is the same as that of an article on page 18 of today’s copy of The Times.

This is the first paragraph.

A single phone call to an older patient who has been discharged from hospital can almost halve the odds of readmission, research suggests.

I have mined health-care data in the past several times and often something simple drops out from a simple analysis.

Some analyses produce the obvious like you gets lots of leg injuries on Saturday afternoons, due to football being played.

I also believe that analysis of health data in an area could pick up more sinister links.

This could be picked up by artificial intelligence scanning the various databases, but until such systems are fully developed, a lot can be picked up by analysts using simple tools. Even Excel can find a lot of problems, if used properly.

 

June 14, 2019 Posted by | Computing, Health | , , , | Leave a comment

A Double Database Cock-Up From The NHS

At three on Sunday morning, I phoned NHS 111 to ask for a bit of help with my terrible cold that was stopping me from even getting to sleep.

I had some advice which helped, but I was also booked in to see a doctor at 09:00 in a surgery a short bus ride away.

So far so good and no complaints.

I duly saw the doctor and he prescribed several drugs, which I took to my local Boots later in the morning.

I should say at this point, that four years ago, I officially changed my name from the one my parents gave me to the one I’ve used continuously since 1968. I was starting to get problems with some airlines, where my passport had a different name to my bank account. My current GP has only ever known me by the latter name and I’m registered with their surgery using it.

When I got to Boots, they initially rejected the prescription, as for some reason it showed by old name, although my address, NHS number and other personal details were correct.

How did the wrong name get on the prescription?

Luckily, Boots were pragmatic and as they recognised me, I got some of thew drugs.

But not all!

The pharmacist recognised that two drugs were incompatible with the Warfarin I take.

So why did the NHS computer system allow the doctor to prescribe the drugs?

As someone who was at the forefront of database technology, I believe, these two problems are inexcusable.

My incorrect name could have led to failure to obtain needed drugs.

The lack of interaction checking, could have led to serious problems for a patient.

January 7, 2019 Posted by | Computing, Health | , , , , | 4 Comments