How not to lose 16,000 Covid-19 test results: a data scientist’s view

https://rssdss.design.blog/2020/10/10/real-data-scientists-monitor-how-to-avoid-losing-1600-test-results/

by martingoodson • 6 years ago

8 12 6 years ago

rssdss.design.blog

What has happened is sheer incompetence in a field that has been mature for decades in how to do things reliably, even without grand "data science leader" titles.

A culture of rigour, discipline, and I will add of "getting things done", is sorely needed.

More generally, the track and trace system in the UK is poor. This has a lot to do with the above but also with lack of political courage and leadership.

Unfortunately it is difficult to suggest that the NHS might need to change, it is taboo subject.

martingoodson2• 6 years ago

It's not so much the 'leader' title. It's that somebody sensible needs to have power and be in charge. That doesn't seem to be the case here.

mytailorisrich2• 6 years ago

That doesn't seem to be the case here.

You're being polite.

Silhouette• 6 years ago

Branding aside, AFAICS the track and trace programme has very little to do with the NHS.

One might suggest, however, that when setting up a scheme to deal with a public health matter and to create technologies to assist with that, hiring some experts on public health and technology might have been a good idea. It looks like a lot of the senior figures actually have a background in retail and the recruitment process was more like staffing a commercial call centre than a medical helpline.

(The article is not wrong about the data processing side either, of course. How these systems meet even basic data protection standards, never mind the more stringent requirements for processing sensitive data, is still a mystery to me.)

six2seven• 6 years ago

I'm afraid that what has happened is something unfortunately quite normal in UK and it's a part of the general work culture. That is, everyone is just doing part of their job and following the protocol, but nothing more. Something is outside of the protocol or task description? Not my problem. Shall I double check the number of rows in the excel with Covid data that I am supposed to send to my supervisor? Oh, come on, who cares.

vfclists2• 6 years ago

You are expecting competence from an administration which dole out management positions to their cronies.

Why not call in Crapita or one of the leading consultancy firms to handle it? That will get you even better (worse you mean?) results at wildly inflated prices as well

gnufx• 6 years ago

I don't know that it needs Crapita, given Deloitte, Serco, G4S ... already.

vfclists2• 6 years ago

You are expecting competence from an administrations which dole out management positions to their cronies.

Why not call in Crapita or one of the leading consultancy firms to handle it? That will get you even better results at wildly inflated prices as well.

aaron6952• 6 years ago

Clearly written by someone without any real world experience. Daily checks and plots, and daily phone calls and meetings.....

But how we start by explaining this. Does the author expect us to believe they were using Excel 2003?

"I’m not going to complain about Public Health England (PHE) using excel to merge the test results from each test centre"

What actually happened, then perhaps we move onto solutions.

(The solution most likely being some sort of Hamming code/check sum, the number of cases is also included with the data for instance)

proce55ing• 6 years ago

They did indeed use Excel to process the data. They could have possibly used a more recent version of Excel but saved the data in the legacy file format anyway.

https://www.bbc.co.uk/news/technology-54423988

martingoodson2• 6 years ago

Hamming codes? This comment is clearly written by someone with no real world experience of working in public sector IT. What I've outlined in the article is a basic solution that can be implemented by anyone.

aaron6952• 6 years ago

Sorry.

I didn't see it you who posted your article. That was a cunt comment.

I dont agree with you but I should have worded that differently.

How not to lose 16,000 Covid-19 test results: a data scientist’s view

Related Stories