Did you know that New GP data extraction ‘far exceeds’ care.data plans according to privacy campaigners

[stumbledin]

New automatic extractions of data from GP-held patient records, due to come into force in early summer, are ‘far bigger’ and ‘more intrusive’ than care.data, GP privacy campaigners have warned.

NHS Digital announced earlier this month that it will be rolling out a ‘new and improved’ GP data collection system from 1 July called General Practice Data for Planning and Research (GPDPR), with patients wanting to opt out having to do so by 23 June.

GPs don’t have to contact patients but will need to update their privacy notice, for example on their website, NHS Digital has said, adding they may also wish to include information about the change on social media, newsletters and other communications.

This comes in contrast to the scrapped care.data plans which saw every household in England receive a leaflet about the upcoming changes in 2014.

www.pulsetoday.co.uk/news/workload/new-gp-data-extraction-far-exceeds-care-data-plans-warn-privacy-campaigners/

More information at www.foxglove.org.uk/nhs-data

The update from ICO is less than reassuring:

twitter.com/EerkeBoiten/status/1399696743164485634

Will the DWP be able to access these records? making the case that they have a valid and compelling legal reason?

At present they need the claimant’s explicit permission and go to the GP or consultant directly, for example to help verify or explore a claim for sickness or disability benefits (and which I am sure would only give them selective information from a doctor) but I would be horrified to learn that they might be able to get easy access to a database without the patient-claimant even knowing. And even worse to get access to ANY or ALL of a patients records.

Where would we stand on this kind of access?

*Hmm. That doesn't sound great. Is there an easy way to opt out?

Depending on your surgery - most aren't taking an email notification, you have to be able to post/deliver an opt-out form to your surgery.

Form available here: medconfidential.org/how-to-opt-out/*

Personally, if it were me I'd send the notification certified or registered mail, whatever you've got over there. I would want to see that they have received it. I would even try to follow up with a call to see if they followed through.

I have just seen a longer article and there are 2 types of opt out.

I am too tired to try and paraphrase but for anyone wanting to check and how to opt out (1 and/or 2) see digital.nhs.uk/data-and-information/data-collections-and-data-sets/data-collections/general-practice-data-for-planning-and-research/transparency-notice#opting-out-of-nhs-digital-collecting-your-data-type-1-opt-out-

Slight aside but the article says they will be collect "sex" ie not gender!

UmmingandAhhing thanks for your post and for your work. I completely agree that using women’s life course medical records is crucial to improving knowledge, treatments and public services. One of the data fields I understand is often missing from GP data can be about service that are not provided by the NHS but might be relevant for a research to know about a woman, like her use of fertility treatment or her abortions or sexual health clinic records or some mental health issues records - I was told by a nurse that they used to omit these on purpose from the NHS records for the woman’s privacy. She also said that they knew there was underreporting by patients for their own privacy or for other reasons like fear of GP or hospital doctor’s prejudice. Can this new system if it’s individually linked, make it so patients can have some control over what hasn’t been included and add anything that they want a researcher (but maybe not a GP) to know?

Another useful piece on the implications of this:

As NHS Digital can reverse the pseudonymisation process as required, there can be no claim that the personal data being processed are “anonymous”, although in the hands of some Third Parties (e.g. researchers), it may appear to them that they are processing anonymous data.
…
Missing from this non-exhaustive list of purposes are the purposes associated with disclosure of personal data by NHS Digital. Several of these can be found in Section 261 of the Health and Social Care Act 2012 and include: disclosure intended to protect the welfare of an individual; disclosure made to a body exercising public functions for the purposes of those functions (pretty broad that one!); disclosure which relates to the investigation of a crime, or where any other legislation permits disclosure (another broad category of disclosure).

In other words, use of the NHS Digital GP database is not limited to medical related purposes; personal data can be disclosed for non-medical purposes. As with the Electoral Roll, Census and other large public sector personal data collections, the NHS Digital GP database could be the target for other public bodies (e.g. the bulk personal data collecting powers of the national security agencies).

Is this privacy paranoia at its worst? Well no, because in Scotland, the Health Service database of patients – because of its accuracy of the address element– was intended to become the basis for a Scottish population register for general use and to support an ID card arrangement (see references).

amberhawk.typepad.com/amberhawk/2021/06/missing-data-protection-safeguards-with-respect-to-nhs-digitals-national-database-of-medical-records.html

Thank you for all the information EmbarrassingAdmissions

disclosure made to a body exercising public functions for the purposes of those functions

This sounds like the bit that would worry me. In the same vein as the other piece of fairly recent legislation that gave broader access powers re access to communications or whatever that was (sorry am v tired!)

The situation has become rather more confusing as NHS Digital has now removed its original mythbusting page:

twitter.com/EinsteinsAttic/status/1401625760142659590

NHS digital have put an explainer out.
digital.nhs.uk/news-and-events/latest-news/patient-data-saves-lives

Unfortunately it’s meaningless to tell us that it’s everyone’s right to be excluded from this automatic data sharing and not tell us how we can make that actually happen. Why isn’t there a national campaign to inform people about this? I really support health research but you either need positive consent or if that’s too difficult to get, an easy way to withdraw individual consent that has been assumed to be given unless it’s specifically withdrawn.

Unfortunately it’s meaningless to tell us that it’s everyone’s right to be excluded from this automatic data sharing and not tell us how we can make that actually happen.

Exactly. And I should think a large swathe of us would appreciate the sort of choice nuance that would allow us to share our data with clinical trial investigators and similar but to withhold it from speculative searches from the Home Office/a variety of law enforcement agencies/commercial entities etc.

Seems like the lesson they took from the hugely wasteful and damaging Care.data farce (which IIRC, resulted in the rebrand to ‘NHS digital’) lead not that the government and NHS should always be really upfront with the public about the use of our data. And not pass the burden on to our overworked GPs to explain and administrate any withdrawals...instead it seems the only lesson they’ve learnt is not to give concrete details but just to offer meaningless platitudes about security, without offering any contact details or FAQs or how-tos. This crappy and disrespectful handling of public goods like our collective health data really undermines public trust and willingness to help, which is essential for research to succeed. It’s so arrogant.

This has now been delayed for 2 months

www.bbc.co.uk/news/uk-politics-57400902

How can they be allowed to share our personal information like this?

Thread about a NEJM (New England Journal of Medicine) paper that discusses a similar issues in the US: Why is all of your health data being monetized by so many companies w/o your knowledge?

www.nejm.org/doi/full/10.1056/NEJMp2102616?query=WB

I’m glad it’s been delayed. There needs to be more scrutiny of this.

And also I’d like to know where the money goes when this data is sold. If it’s ploughed straight back into the NHS, then fab. But I would hold my breath.

I’m one of the people that will be using this data

Its certainly the case that the NHS cohort data is close to unique in terms of richness and value. Many patients would be happy to share their records for the purpose of research if they were asked and if there were guarantees about no further use without further permissions.

Regarding DV, usually sensitive codes like that are removed from records prior to being shared, but it’s worth contacting NHS digital to check as I can’t say for sure that’s the case here (although I will be very shocked if they aren’t)

Be shocked. NHS Digital is exempt from the standard privacy legislations such as GDPR so are not required to redact. There was good reason for this - tightly controlled data sets with limited users in approved research establishments was the alternative to restricting the data with standard privacy rules.

Pseudonymised data by its very nature is reversible. Only for legal reasons? Well who knows what will be legal in the future. Possibly a commercial contract, DWP, handing the longitudinal record over to Palantir/other commercial interests for track and trace or simply marketing? Its certainly a massively money spinning cohort of data.

Lets say there is a degree of pseudonymisation, some flags are removed. Cross match with commercially available data sets of social media usage, marketing sets, retail sets, census data (commercial orgs can buy it by postcode) and its not difficult to home in on individuals.

This isn't new - it was niche 10 yrs ago when the forerunners of companies like Cambridge Analytica were speaking on the fringes of tech conferences; telling you were you were yesterday at 10:00 as a party trick to show off their data matching. Now its mainstream and its cheap.

Its outrageous that the population of E&W is being opted in to a system allowing commercial sell off of their most sensitive data without even discussion let alone informed consent about the use of their data or where any commercial revenue ends up.

So if you want the data available for research don't shut your eyes to the privacy issues - lobby for proper consent and information.

In the mean time, I've opted out as have all my family and colleagues and their families.

Oh and you must fill in both the forms, including one on paper and they go to two different places. No publicity, difficult to opt out. Anyone would think they had something to hide.

NHS Digital is exempt from GDPR? Evidence please?!

I'm with the minority on here in that I'm happy for my data to be used and I do actually trust that the information is dealt with securely and properly. Benefits are huge and the number of opt outs is going to seriously detriment research particularly of certain groups.

There is a whole bunch of derogations from GDPR which governments can elect to use. In particular under article 23 and articles 85-91 which include:

• A country’s security, defense, and public security
• Enabling and securing judicial independence
• The detection, investigation, and prosecution of crime and the prevention of criminal activity
• To enable enforcement of civil law claims
• The protection of subjects critical to national interests such as budgetary, social, and health matters.
• Freedom of expression and information
• Public access to official documents
• National Identification Numbers
• Personal data of employees
• Data for scientific or historical research
• Archiving in the public interest
• Obligations of secrecy
• Churches and other religious associations

There are more but these offer ample opportunity. However GDPR is only one level of legislation and its poorly understood. There are many data sets with exemptions from GDPR and other legislation and there may be good reason. I've used this type of health and cohort data, there are cases where only complete data sets are useful.

The counterbalance for lack of legal protection of some of the data is strict access controls and an approval process to access the data.

At the moment NHS D will show where individual data sets do have degrees of compliance but they don't have to enforce compliance. So you will note that the permissions you as an individual have are a small subset of what you can do with your regular GP records. You also need to know you are in the data sets and apply to each of them (hundreds) individually. Note that you cannot block them from being used, only request a correction (request, not require).

However the current NHS Digital data set does not have global health records for the UK (or E&W). This is what is happening now. Centralised access to the longitudinal (life time) record of every individual in the country for fuzzily defined purposes including commercial use.

Its not good enough for academics to put their blinkers on and say "I'm sure it will be fine" and a great many are not - they are lobbying for proper protocols.

Its an outrage that this government is planning to scoop up the most sensitive data in the country for commercial profit. Again.

I'll be opting out

I have opted out and will opt back in if my feelings on the subject of sharing personal data changes.

My thanks to the posters who contributed reasons why sharing data will be helpful in research. It is due to your calm reasoning that I am considering future options.

I've opted out.

I expect some nuanced options that will allow me to select research and planning or to participate in pragmatic GP-level trials, clinical trials etc.

I will never consent to 3rd party searches that would undermine the civil rights of either me or others .

On principle, I always agree to anything the NHS asks of me. But I am hesitating at this. It’s not the NHS asking, it’s a government that happily sells off parts of the NHS.

I expect some nuanced options that will allow me to select research and planning or to participate in pragmatic GP-level trials, clinical trials etc.

Exactly this.

This kind of data should always be informed consent and opt-in before being used at detail level.

I've opted out, however I also have a chronic health condition and I've given explicit support for my data to be used by the relevant team, limited to that work. Any extension of use requires additional permissions.

Most people don't realise this is possible but it can be a useful compromise. Most people also underestimate the value not just of the data but of the value of processing that data and building models from it.

Decent thread based on the work in London (a form of pilot, very loosely) and other areas:

twitter.com/axelheitmueller/status/1402363115472310275

Latest update on this continuing saga - archive.is version of a Twitter thread by Phil Booth - there's a Commons Health Committee hearing of oral evidence tomorrow (July 20):

archive.is/zkfVi

twitter.com/EinsteinsAttic/status/1417181046647500806