Helping UK health or getting data to sell to Private Health businesses?

FOI? Palantir?

O(f course they are data mining.  thats where the value is.   You should be4 able to track down what they are doing with the data but it will be well obscured

Nearly everybody in Primary to Tertiary health-care thinks that being able to access a patient's entire record would probably be a useful way of making sure your treatment is as good as it can be. Getting any NHS-wide system to be able to do that without the assistance of private businesses is probably - given this country's stellar record of public procurement; more or less impossible. Any deal with any 3rd party provider would stumble over privacy concerns and single provider vulnerability, that and the obvious candidates to build and run such a system are probably based in Silicon Valley, and no-one (not even the Tories TBF) trusts them not to flog the whole thing. After that, there would be a bun-fight between the GPs, the Trusts, and the ICBs about which system that would be, how it would communicate, and what functionality it would provide.

It would be such a shit-show I can't imagine that beyond words, it'll ever get off the ground.

Getting any NHS-wide system to be able to do that without the assistance of private businesses is probably – given this country’s stellar record of public procurement; more or less impossible.

We have pretty much that system in Scotland.  some issues with most GPs only being able to see summaries of hospital treatments.  all NHS owned and the private company whos computer suytems are used are not able to sell the data.  social work is also included

I'm not totally sure of the issues with GPs as some can access TRAK but most only the summaries - my guess its to do with their status as privater contractors.  any secondary care setting anywhere in Scotland has full access to your records I believe

so it can be done and was done in Scotland a decade ago

We have pretty much that system in Scotland.

No, you have the same system we have south of the border, it doesn't have the sorts of complete record sharing that most new systems are suggesting i.e., anything written by any clinician within the NHS is on one record and can be accessed/updated by any clinician, all records be they anything from bloods taken at hospital to your eyesight and dental history, or a NOUS taken at the community centre or the district nurse at your home.

Most local NHS have referral, diagnostics and so on sharing abilities. But things like your GP doesn't see what the (say) registrar in A&E has written and vice versa, the goal is to have that sort of functionality.

No nick - thats simply not so.  the only people that do not have full access to the complete record are some GPs.  community nurses, social work, etc all have full access.  its not the same as in England at all.

GPs can get access to trak but for reasons I do not know most do not.  Its also nationwide.  GPs can install software to get access to Xrays scans etc but most choose not to - again I do not know why

No, everyone has access to TRAK everyone is on the Spine, It's no different down here, I promise. What is being proposed is a step change from that sort of functionality

.

some issues with most GPs only being able to see summaries of hospital treatments.

I thought it was fabulously quaint to come around from aesthetic after surgery earlier this year, be helped into my clothes, be given a slice of heavenly NHS jammy toast (do nurses get special toast training?) and then be given an envelope containing a letter from my surgeon and asked to hand it to my GP on my way home.

🙂

and then be given an envelope containing a letter from my surgeon and asked to hand it to my GP on my way home.

Because it relies on everything being up to date, Manchester for instance has a system that shares pat records with your GP, but only if the two records match...Guess how many folks change their GP? Guess how many times some-one at the trust or at the GP updates the NHS Spine? or tells I dunno; the Radiology dept. becasue for reasons know ones why, they operate their own system that runs off Microsoft 95 and it's millions to update... This is the problem any new system is designed to overcome,  Record Sharing with Everyone isn't the goal. the Goal is One Record shared by Everyone.

That's why it's not going to work.

Check the more recent policy updates, they can sell info now to more people
https://www.theguardian.com/technology/2023/nov/12/private-uk-health-data-donated-medical-research-shared-insurance-companies?ref=biztoc.com

(do nurses get special toast training?)

Hot buttery toast in hospital beats any meal at any Michelin Starred restaurant.

Nick's description of how privacy concerns are hampering modernisation in the NHS rings true... but I also have those concerns, so I guess I'm part of the problem? Who do you trust with your health data? I'm overly cautious about my health privacy, and seeing how other data is used, I'm getting more cautious over time. Avoided Zoe etc for that reason. Coronavirus briefly made me consider the trade off between private data and public good differently, but I've reverted back to being highly suspect about it again already.

Ref the OP i think there are two separate points here. When you sign up for eg Zoe you will probably be signing some terms of use which may allow Zoe to share your data on, albeit in pseudonymised form and likely aggregated with other user's data, and only within what is allowed in GDPR

Then within the NHS there are of course data sharing 'governance' constraints (including need to know etc, not every clinician needs to know all aspects of the patient record), but as others have said the systems are often not interoperable, so free flow of data isn't possible. Reasons for this are many but a significant one is the way health care providers have been set up as separate entities, all buying their own software which isn't necessarily interoperable with other providers software. With the establishment of integrated care boards this *should* improve over the long term, where the objective of defragmenting / reducing the number and complexity of systems across health and care is successful, together with more investment in interop initiatives, joining different software (and providers) together to allow data to flow. Subject of course to those pesky data governance rules, which hopefully achieve the right balance between sharing our data more widely across the NHS in support of our health care (plus for authorised clinical research), but not allowing it to be used for the 'wrong' (purely commercial) reasons.

Just remember that those private companies are (rightly or wrongly) the ones who develop the drugs that save our lives.

I’m overly cautious about my health privacy, and seeing how other data is used, I’m getting more cautious over time

Is this just general paranoia or do you have specific things in mind? Because using your data without your explicit permission is very illegal.

In the way that vaccines are at the same time a matter of private decision and a public health requirements, there's lot in healthcare that's a difficult ethically. Looking at health records at a population level can obviously make a huge difference to the direction of future health care needs, but sometimes it might not be in your best interests to have one record that everyone can look at, and that's way before any dicey company wants to sell your record to some third party. Having said that didn't Garmin get hacked a couple of years ago. They have all sorts of medical history on looooads of folks, loads of women use period tracking apps,...so who knows what's already out there.

For all the reported benefits, I don't think it's doable [for the NHS to have a one record system]

What's depressing is that we have capability to do these analysis ourselves within academia- We have some  of the best research institutes & universities in the world.

But farming it out to the private sector is preferred to funding the research here.

With the establishment of integrated care boards this *should* improve over the long term

You would've thought, eh?  Never going to happen. You may get locally where there's sufficient weight for ICB to ay demand that all practices sign-up to one system, but the best you'll ever achieve is local continuity.

Nearly everybody in Primary to Tertiary health-care thinks that being able to access a patient’s entire record would probably be a useful way of making sure your treatment is as good as it can be. Getting any NHS-wide system to be able to do that without the assistance of private businesses is probably – given this country’s stellar record of public procurement; more or less impossible. Any deal with any 3rd party provider would stumble over privacy concerns and single provider vulnerability, that and the obvious candidates to build and run such a system are probably based in Silicon Valley, and no-one (not even the Tories TBF) trusts them not to flog the whole thing. After that, there would be a bun-fight between the GPs, the Trusts, and the ICBs about which system that would be, how it would communicate, and what functionality it would provide.

Fujitsu would be keen to assist with this!

We have pretty much that system in Scotland.  some issues with most GPs only being able to see summaries of hospital treatments.  all NHS owned and the private company whos computer sytems are used are not able to sell the data.  social work is also included

By definition if most GPs can't access all the information (or in my experience any of it - like maccruiskeen we still seem to be in a world of "here's a letter for your GP".  Then the goal that nickc described is not being met.  I've literally seen a GP take blood and run the same tests the clinic did the week before "because it could be months before they tell me if there's anything strange", and I've been to see a consultant who said (correctly) that the GP had taken bloods from me and expected me to know the result because he didn't have access to it and the GP hadn't told him...   I was never told the result (and unlike england have no way to see my own results).

I’m not totally sure of the issues with GPs as some can access TRAK but most only the summaries – my guess its to do with their status as privater contractors.  any secondary care setting anywhere in Scotland has full access to your records I believe so it can be done and was done in Scotland a decade ago

Yet expectant mothers are still seen walking into Maternity wards carrying their own notes (that may have changed, i've not been near a maternity ward since covid but it was the case more recently that a decade ago...

<5 years ago my daughter was transferred from St Johns to Sick Kids (during the time St Johns peads dept was day only) and I got a letter to give to the staff when we arrived.  They knew she was coming but didn't have any case history and repeated a load of tests on arrival (IIRC including x-rays) because they didn't have access to the data.  This was two wards in the same trust which were routinely transferring multiple patients to each other every day.

So theoretically it might be meant to work how you suggest... but the reality is it doesn't in large parts.

the Goal is One Record shared by Everyone.

Which is what happens in Scotland bar some / most GPs

They knew she was coming but didn’t have any case history and repeated a load of tests on arrival (IIRC including x-rays) because they didn’t have access to the data. This was two wards in the same trust which were routinely transferring multiple patients to each other every day.

Thats just wrong by the trust as both hospitals would be using TRAK then.  I mean I have checked the single record for pateints being transfered to us before they arrive because I could on TRAK.  TRAK creates a single electronic record for every patient and its almost impossible to create a second one

My guess is its poor usage of the system and they discharged the patient on TRAK rather than transferred them - or the nurse / doctor / admin in the receiving ward did not know how to do the transfer / access records assigned to another hospital.

Ever NHS healthworker bar some GPs in Scotland uses TRAK.  its a single patient record for every patient accessible to all healthcare professionals

NICKC - any idea why GPs mainly do not use TRAK?  No technical issue as I have used TRAK in the same office as GPs as a community nurse

Interestingly, The People's Pulse wanted to know about clothes washing habits today.

For me, it's slightly vague paranoia , but mainly 'why should someone other than me make money out of my data'

The prompt was todays "How's your laundry?" survey, and a recent "what do you think about milk then?" survey.

NICKC – any idea why GPs mainly do not use TRAK? 

Because not every trust uses it. I know Stockport uses TRAK where we are we use a system called ICE. This is the problem with integrating all these hospital/GP/Community/social care systems, there's hundreds all over the country. Most GPs use either System One, or EMIS, or Vision, dentists uses something else, and so on and on, hospitals all use something else again.  

Ah - everyone in Scotland bar GPs use TRAK.  all hospitals, community nurses, physio, dietetics everyone.  I do not understand why GPs here do not

One system, single patient record in Scotland bar most GPs

Edit - so contrary to what you said above its obviously not the same here.  any TRAK userr ie every healthcare professional in the NHS bar mo0st GPs can access any and every patients single record via TRAK

Our ~TRAK system also seems to have much more functionality?

So once again bar these GPs a single patient record accessible to all controlled and owned by the NHS is perfectly possible as its what we have here

One system, single patient record in Scotland bar most GPs

And given that the vast majority of NHS patients interact with GPs and not Trusts, you can see the issue. Anyway still not the same. What the project aims is for there to be one patient record on one system that everybody uses; Trusts and GPs alike, not multiple records held at different locations about the same patient that's shared (or not) amongst clinicians that might treat the patient at different times. Or individual silo'd specialities that use some equipment that runs off some obscure system that no-one uses, but is so wildly expensive to update or change, that no one does it, so they print off everything and just  post it to your GP...

Ah – everyone in Scotland bar GPs use TRAK.  all hospitals, community nurses, physio, dietetics everyone.  I do not understand why GPs here do not

One system, single patient record in Scotland bar most GPs

Unfortunately not really true TJ.

All Scottish healthboards do use Trak, but the version used by each area, and its function, and what its used for varies widely - so to say it's a single patient record for the whole of Scotland isnt really accurate sadly.

In my previous trust, it was beginning to function as a proper EPR, but relies on "pulling" data from other systems that you could access via trak (if you had access to that system) for things like blood results, radiology reports etc.

In my current trust, its functionality is much more limited and is really only used as a bed management system. We regularly take patients from other trusts who's EPR we don't have access to so comes in printed version!

Traks not a bad platform (although I believe its original software design was for use in container shipping management!), but as was said before,  trying to get everyone across the board to agree on a platform that meets everyone's needs and demands is I fear somewhat of a pipedream.

For me, it’s slightly vague paranoia , but mainly ‘why should someone other than me make money out of my data’

The prompt was todays “How’s your laundry?” survey, and a recent “what do you think about milk then?” survey.

I've never heard of People's Pulse - but if they are asking question that you don't see value in answering and they aren't explaining to you why its valuable then I think you are just being milked.  Someone somewhere is making money from it, even if its just a professional researcher.  

Zoe is a commercial company.  They played a blinder during covid in getting high adoption as a measure of the real state of the nation - but lets be clear that was about sucking people in to use in the next stage!  

I'd have slightly more faith in BioBank because the rules on human tissues are really quite tight BUT I would feel zero qualms about stopping if you don't think you are getting enough of a positive feedback loop from them on what value/benefit your samples are bringing.  

What the project aims is for there to be one patient record on one system that everybody uses; Trusts and GPs alike, not multiple records held at different locations about the same patient that’s shared (or not) amongst clinicians that might treat the patient at different times

Which is how TRAK functions in Scotland IME - although I note ecampells remarks so clearly some areas are behind others  the only people not having access to the single patient record is some GPs.

I personally have accessed records from other Trusts using TRAK

the only people not having access to the single patient record is some GPs.

But that means that those TRAK patient records aren't the entire patient record, it's missing the GP parts. D'you see? That's the issue, not that some trusts can share the records that they hold. It's not about sharing pt.records, it's about there being just one pt. record.

but mainly ‘why should someone other than me make money out of my data’

Because it might help someone find a cure for some horrible disease?  My data can be infinitely duplicated - once its anonymised, people can fill their boots. I'm not going to jealously guard it. I mean, it's not like I put special effort into being white, mid 40s, anglo-saxon etc so I'm not expecting a reward from it.

A strong disclaimer on this post. The views and comments I'm about to give are completely my own and have nothing to do with my employer.

I've spent the last 20 years working in private health insurance. On every single one of my working days, I've analysed huge amounts of medical claims data, demographic information etc in order to do my job.

All of that information has always been internally collated data relating to people insured by my employers. Not once have we ever imported externally sourced data and matched it up with existing customers.

There are lots of private companies out there trying to sell data to companies like mine. They're not giving us individual people's details. It's all high level stuff, not a single person's medical or health records. I get bombarded by marketing emails for this kind of stuff but it's never selling individual data that could ever be used to identify a person. It's all about making better risk decisions etc. No one is telling me about your STDs or strange rashes.

Believe me, there is zero chance a health insurer could possibly match up a huge dataset of external data with their existing customer base. You're giving far too much credit to us and what service platforms we're using. For context a previous employers customer service platform is running on software from 1988 and is barely serviceable.

NIckc - I think you are misunderstanding me.  There is a single patient record.  for reasons I do not understand most GPs but not all will not use it.  Its not trusts sharing data - its the EPR goes with the patient

Yes I know its not complete if GPs do not use it.  I was looking for an explanation of why not as I have never found one

Anyway - thread digression - apologies

Boardingbob - the problem is that the data can be very granular making it possible to identify individuals even if anonymized.  If its known you are a 45yr old left handed redhead who has had a coronary artery bypass in march in Warrington and is married with 2 children then that sort of detail can lead you to individuals.  O(r at least thats one account of this I have read and why I would opt out of the english system being run by a company well known for poor ethics

NIckc – I think you are misunderstanding me.  There is a single patient record.  for reasons I do not understand most GPs but not all will not use it.  Its not trusts sharing data – its the EPR goes with the patient

No we're taking at cross-purposes about the same thing (as I think our experiences are from each end of the system) There isn't one record, there's at least two: one held by the trust and one held by the GP.

This is a useful website:

https://medconfidential.org/news/

once its anonymised, people can fill their boots.

It's actually quite difficult to anonymise medical data reliably. All the data relating to you has to be grouped together to mean anything - no point know what your cholesterol level is if its not linked to data about age and gender for instance.

Ben Goldacre gave a useful example showing you don't need a lot about someone to find everything about them by using publicly available information about himself - in his example he's male - doesnt narrow things down much. a certain age - doesn't narrow it down much, born in one part of the uk, doesn't narrow it down a lot, moved to another part of the uk in a certain year..... narrows it down a lot, perhaps only looking at dozens of people now, then - he is father to twins. Narrows it down an awful lot. You'd only really need to know one more thing about him to be pretty sure you were looking at all his medical information.

Now he's giving himself as an example as a fairly public figure - and a potentially divisive one if you're in one of the fields of woo that he's critical of. Slighted people can be quite nasty and it's not beyond reason that someone would seek to embarrass him for instant.

But what about you - who cares what you think - why would you have anything to worry about? Us nobodies don't really get a choice as to whether we become news or not.

potentially divisive one if you’re in one of the fields of woo that he’s critical of

And indeed if you're a drug company researcher cherrypicking findings as per Bad Pharma. And if you're in a field of which Goldacre thoroughly approves. He basically just enjoys going round being rude to people. He'd fit in well here.

However, the review he chaired a couple of years ago for government about use of data in research was about how to make data more available to research:
Better, broader, safer: Using Health Data for Research and Analysis

Here's a quote apropos of Biobank:

All of these participants have given written informed consent for their GP records to be accessed by UK Biobank, and linked to the other study data, for access by researchers globally. Despite this consent, UK Biobank’s team has been required to negotiate separately with about 6,500 GP practices – since each is the data controller for some of the participants – in order to obtain separate permission from each GP for each participant. UK Biobank is therefore still not able to make GP data available to support research into the causes and prevention of many different diseases.

Recent special access to these data under pandemic legislation, solely for the purposes of COVID-related research, has now demonstrated just how valuable GP data is for studying the determinants of disease; and, therefore, how much is being lost by the continuing failure to make these data available more widely.

Support biobank and you're supporting life saving research. I've looked at the guardian story - about insurance company researchers accessing data to build predictive models as far as I can see, the sort of thing folks were trying to sell to boardingbob above. I'd personally cock an eyebrow at this - it's not what we'd think of as er lifesaving maybe. But I personally be relaxed about it. It's just risk stratification as per loads of health research. And let's remember that in the round, insurance is a good thing as are effective pharmaceuticals. Industry makes these things and should have some well regulated access to data. to devise and sell equally well regulated products. [should say I'm not particularly an insider, no conflicts to declare, just happen to think biobank is a good thing.]

Manchester for instance has a system that shares pat records with your GP, but only if the two records match…Guess how many folks change their GP? <br /><br />

This is what the Exeter system tracks along with births and deaths. It is like all systems as good as the data entered. The it GPs use is upto them as they are private business. Most use either emis or systmone but they don’t have to

If its known you are a 45yr old left handed redhead who has had a coronary artery bypass in march in Warrington and is married with 2 children then that sort of detail can lead you to individuals.  

You'd be surprised how little information a medical insurer holds about their customers. Names, DOBs, addresses, phone numbers and emails is about it.

No information held on height, weight, lifestyle,smoking status, alcohol consumption etc.

Even detailed medical information is pretty light in terms of detail. Medical records are only accessed if required at the point of a claim if necessary and only with the patient's consent.

Of course, the more data points you have, then any real anomalies could stand out, but as an industry we're not interested in the individual.

I don't want to know or need to know about a single person, I rely on huge data samples for tens of thousands of people.

But again I would stress than no health insurer I've worked for in 20 years has ever used external data that's been collated by a 3rd party. To be honest there's no need. If we're insuring hundreds of thousands of people, we don't need to go elsewhere for data.

The point is that anonymi8sed medical data is still possible to identify individuals from