Thoughts since #UKHC15. UK health datasharing.

The world you will release your technology into, is the world you are familiar with, which is already of the past. Based on old data.

How can you design tools and systems fit for the future? And for all?

For my 100th post and the first of 2016, here is a summary of some of my thoughts prompted by . Several grains of thought related to UK heath data that have been growing for some time.

1000 words on “Hard things: identity, data sharing and consent.” The fun run version.

Do we confuse hard with complex? Hard does not have to mean difficult. Some things seem to be harder than necessary, because of politics. I’ve found this hard to write. Where to start?

The search to capture solutions has been elusive.

The starting line: Identity

Then my first thoughts on identity got taken care of by Vinay Gupta in this post, better than I could. (If you want a long read about identity, you might want to get a hot drink like I did and read and re-read. It says it’ll take an hour. It took me several, in absorption and thinking time. And worth it.)

That leaves data sharing and consent. Both of which I have written many of my other 99 posts about in the last year. So what’s new?

Why are we doing this: why aren’t we there yet?

It still feels very much that many parts of the health service and broader government thinking on ‘digital’ is we need to do something. Why is missing, and therefore achieving and measuring success is hard.

Often we start with a good idea and set about finding a solution how to achieve it. But if the ‘why’ behind the idea is shaky to start with, the solution may falter, as soon as it gets difficult. No one seems to know what #paperless actually means in practice.

So why try and change things? Fixing problems, rather than coming up with good ideas is another way to think of it as they suggested at  #ukhc15, it was a meet-up for people who want to make things better, usually for others, and sometimes that involves improving the systems they worked with directly, or supported others in.

I no longer work in systems’ introductions, or enhancement processes, although I have a lay role in research and admin data, but regular readers know, most of the last two years has been all about the data.

More often than not, in #ukhc2015 discussions that focused on “the data” I would try and bring people back to thinking about what the change is trying to solve, what it wants to “make better” and why.

There’s a broad tendency to simply think more data = better. Not true, and I’ll show later a case study why. We must question why.

Why doesn’t everyone volunteer or not want to join in?

Very many people who have spoken with me over the last two years have shared their concrete concerns over the plans to share GP data and they do not get heard. They did not see a need to share their identifiable personal confidential data, or see why truly anonymous data would not be sufficient for health planning, for example.

Homeless men, and women at risk, people from the travelling community, those with disabilities, questions on patients with stigmatising conditions, minorities, children, sexual orientation – not to mention from lawyers or agencies representing them. Or the 11 million of our adult population not online. Few of whom we spoke about. Few of whom we heard from at #ukhc15. Yet put together, these individuals make up not only a significant number of people, but make up a disproportionately high proportion of the highest demands on our health and social care services.

The inverse care law appears magnified in its potential when applied to digital, and should magnify the importance of thinking about access. How will make things better for them, and how will the risks be mitigated? And are those costs being properly assessed if there is no assessment of the current business case and seemingly, since 2012 at least, no serious effort to look at alternatives?

The finish line? We can’t see what it looks like yet.

The #ukhc2015 event was well run, and I liked the spontaneity of people braver than me who were keen to lead sessions and did it well.  As someone who is white, living in a ‘nice’ area, I am privileged. It was a privilege to spend a day with #UKHC15 and packed with people who clearly think about hard things all the time. People who want to make things better.  People who were welcoming to nervous first-timers at an ‘un’conference over a shared lunch.

I hope the voices of those who can’t attend these events, and outside London, are equally accounted for in all government 2016 datasharing plans.

This may be the last chance after years of similar consultations have failed to deliver workable, consensual public data sharing policies.

We have vast streams of population-wide data stored in the UK, about which, the population is largely ignorant. But while the data may be from 25 years ago, whatever is designed today is going to need to think long term, not how do we solve what we know, but how do we design solutions that will work for what we don’t.

Transparency here will be paramount to trust if future decisions are made for us, or those we make for ourselves are ‘influenced’ by machine learning, by algorithms, machine learning and ‘mindspace’ work.

As Thurgood Marshall said,

“Our whole constitutional heritage rebels at the thought of giving government the power to control men’s minds.”

Control over who we are and who the system thinks we are becomes a whole new level of discussion, if we are being told how to make a decision, especially where the decision is toward a direction of public policy based on political choice. If pensions are not being properly funded, to not allocate taxes differently and fund them, is a choice the current government has made, while the DWP seeks to influence our decison, to make us save more in private pensions.

And how about in data discussions make an effort to start talking a little more clearly in the same terms – and stop packaging ‘sharing’ as if it is something voluntary in population-wide compulsory policy.

It’s done to us, not with us, in far too many areas of government we do not see. Perhaps this consultation might change that, but it’s the ‘nth’ number of consulations and I want to be convinvced this one is intentional of real change. It’s only open for a few weeks, and this meet up for discussion appeared to be something only organised in London.

I hope we’ll hear committment to real change in support of people and the uses of our personal data by the state in the new #UkDigiStrategy, not simply more blue skythinking and drinking the ‘datasharing’ kool-aid.  We’ve been talking in the UK for far too long about getting this right.

Let’s see the government serious about making it happen. Not for government, but in the public interest, in a respectful and ethical partnership with people, and not find changes forced upon us.

No other foundation will be fit for a future in which, the phenotype data, is to be the basis for an NHS so totally personalised.

If you want a longer read, read on below for my ten things in detail.

Comment welcome.


Hard things: The marathon version, below.

Ten (longer) things I took away from #ukhc2015 – a 5K

  1. There is no such thing as ‘the data’. (and consent)

There are multiple sources of and collections of individual pieces of information gathered about individual people and how those are created, stored, extracted, labelled, grouped, governed, released, used, communicated, legislated for – “the data” – are in continuous flux as the collections are simultaneously created, changed and deleted.

To try to smooth and simplify this understanding into something easy to grasp can be useful but not when it rubs off some of their distinguishing features which determine how data are recognised. Data as a package, is understood by its characteristics and therefore determines how data are handled* by person or machine. (*apply all the other actions above too).

It is these data features – are data identifiable,  anonymous, sensitive? – that shapes how they are handled in law for one thing.

Saying uses anonymous data is wrong, not because what is shared with a third party – that depends on the onward release – but because at the point of collection and extraction to the HSCIC these data are both identifiable and sensitive.

If you have read the Partridge Report and Sir Nick’s article and opinion afterwards, you know the HSCIC has taken steps to reduce repeating the issues of control and release, it had in the past. (The HSCIC would hardly be investing in building a brand new secure setting if they felt the data had been adequately handled before.)

Talking about permanent ‘fixes to the datasharing landscape’ is therefore a bit like talking about deserts. There is constant change. Some of which is fast and noticeable, some is gradual and almost invisible. Some change is constructive and some is not and much of the time it is wearing.  There is a lot of hot air about and too little is known and too much assumed about how it all works by too few people. It’s easy to get lost, and the whole debate could actually do with some much more data-savvy individuals to correct misunderstandings and propose real data solutions, not simplified “datasharing” as we think of it in lay terms. Collect, store, use as is.

At the National Institute of Health, we heard, large datasets are imputed. Simulation models are used for sexual heath data and sensitive data. Solutions for complex data lie within complex understandings of data.  To date, I’ve heard too many in NHS England leadership making decisions who don’t actually understand data and statistics well. We must understand what we’re talking about.

Simplyfying the boundaries between these data distinctions and definitions in order to find quick ‘n’ dirty solutions to the ‘hard things‘ around consent and identity is therefore counter productive.

And clean up is costly in all sorts of ways.

Things we don’t talk anough about enough about “the data” are data quality and correcting innaccuracy. When billions of data are being combined to make machine learning decisions, I am concerned mistakes will be magnified – in using semantic learning words could have all sorts of different meanings and intentions – and without transparency of “the data” understanding where it comes from, how data are linked,  how may the population see inside these black-box systems to correct innaccuracies, or how might we ever really know of them?

2.  Data ownership doesn’t matter. (and consent)

Debating who “owns” data is rarely helpful I was reminded afterwards, yet I got drawn into that discussion in much further detail than I would have liked, unable as I was to sit and hear people suggest that we have to accept anything done with our data for our own good, and it’s not ours anyway.  It is. Even the recently departed Mr Kelsey and Simon Stevens spoke of this in March 2015.

Tim Kelsey, former Director Patients and Information, NHS England, March 2015 at the King’s Fund:

“We now acknowledge the patient’s ownership of the record […] essentially, it’s always been implied, it’s still not absolutely explicit but it is the operating principle now legally for the NHS.

“So, let’s get back to consent is essentially, their data.”

When I still hear NHS England staff question this ownership, I sigh. It has echoes of “no one who uses a public service should be allowed to opt out of sharing their records. Nor can people rely on their record being anonymised— at the moment sexual health services can be anonymous, and as a result there are almost no measures of performance in that sector.”

Sexual health clinics allow(ed) anonymity for a reason. People would go to them precisely because they would not be known, or followed up, or have their records linked to their regular GP record. That was the personal control that a person could exercise without being concerned that if records were clear, even having been for a test, would be a potential flag to a medical provider.

A woman I met in 2014, let’s call her Clare, asked me about because she was concerned who would access her records and the conclusions they drew after she had been refused top-up medical insurance at a new job. She was flagged as not having completed psychological counseling years earlier, and therefore ineligible.  Counseling she had chosen to leave, together with her GP, because she opted for a different kind of therapy after surviving a near fatal attack in a relationship controlled by ‘domestic violence’.

She wanted to retain control over what decisions were made about her by whom, for good reason. She no longer trusted that records outside her control, and the effects that may have on her,  would be accurate in the eyes of others and their decision making. The trauma of her experience was compounded by its data handling and triggered a mentally difficult year for her. Trusting the NHS manages our data well, matters.

So who owns, controls, and processes data does matter in terms of its governance because without it there is no responsibility or accountability for its management and decision making. You cannot draw lines in the sand of responsibility, if ownership does not matter.

If a patient does not want anyone to have their data other than their GP, it is the extraction from the GP practice that for them is a breach of confidentiality, not who releases it to later. This is why conflicts so badly with the Hippocratic oath. and one reason that doctors voted for confidentiality and for opt-in not out of in 2014.

The autonomy of the decision making must remain with the person if the data belongs to the person and the person, not “the data” comes first. Rights and responsibility stem from here.

3.  The language of datasharing confuses sharing and taking. (and consent)

The whole concept and language of sharing has become abused in the relationship between organisation and citizen, suggested Natasha Siger in the NYT last August.

“Let’s begin with “sharing,” a concept that implies something selfless. Altruistic words may also lend an aura of incontestability.  So it seems inaccurate, even euphemistic, he said, to use “sharing” in the context of commercial enterprise.” 

Good relationships are not about manipulation. Consent to be valid, must be informed and freely given. When is giving not really giving? Is giving truly a freely consensual act of altruism if the behaviour is actively manipulated?

The concepts behind this on sharing cash in the nudge unit work on charitable giving is something about which I feel uncomfortable.

Should the Cabinet Office have any place in the ‘nudge unit’ work on charitable donations? Or the NHS? Organ donation is a good thing. I too am a registered donor, but I want to know that any decision I make about my own body, about choices I make, are mine, and mine alone.  So too for the ‘sharing’ of data.

Where is the transparency over that work? If we are to trust who and what is involved in that work, there should be no secrecy about any ‘behavioural unit’ meant to influence citizens’ free will.

I find it patronising for the state to think it should meddle in the decision making and choice of how I give money to others.

When the report suggests “removing the need for an individual to act at all” on their recommendations for opt out not opt in and that to really increase donations you personalise the request writing ‘dear David’ rather than use an anonymous request, you can see how easy this cumulative effect viewed as a benefit for one party, becomes harassment for the public – it’s not designed that way, but that’s reality in practice – what privacy impact assessment does the nudge unit perform in any of its work, I wonder?

4. Not all datasharing purposes are equal.

Much of #ukhc15 discussion conflated direct and indirect datasharing. Trying to remind people in  discussions to separate the two into the sharing of your information for the purposes of your direct care, and into secondary purposes, was not easy.

Even people who work with data at NHS England I later realised, mixed them back in, saying that planning about your care was also direct care. It’s not that clear cut. Not all ‘planning’ is equal. The consultant planning your treatment with a GP who shares data on the course of action is sharing data for your direct care. A CCG risk stratifying a whole GP practice to find those 6 patients who the system says are most likely to need a consultant is not.

Lumping together data for direct healthcare into ‘planning about your healthcare’ is therefore not simplifying but adds further confusion. The HSCIC FAQs makes a valiant attempt to clarify these.

Direct healthcare and Indirect care are distinct in law and on how they are governed and reading the Caldicott II from April 2013 is clear on this complex area.  There are codes of practice on these things. Not hard, but complex. And while it is currently under review, here is a note from one of the remarks made in the 2013 review:

“The Review Panel found that commissioners do not need dispensation from confidentiality, human rights and data protection law since, with little effort, they can operate perfectly well within it.

Implied consent isn’t a catch all, in the same way that ‘consent’ is not a tick box to then be able to do anything with data.

The reasons why some want to just to take all the data about people or groups of people for wide purposes ( and past history being a lesson why not (Nazis) got brought up in that session, and need serious attention for serious reasons.

These are indeed ‘hard things’.

Purposes for which data are collected must be defined or risk purposes arising for which they were not intended. See above.

The realistic challenge in today’s datasharing plans is that HMRC wants access to everthing. To health, to DVLA data, to DWP data, to education data – and why? Because fraud.

The state seems to have lost its trust in the individual citizen and it’s a darn expensive loss.

On the flip side, the state demands citizens trust the state in everything,  in the NHS that means patients need to be able to trust that the purposes they are signed up to, stay what they sign up to.

You wouldn’t want to say yes to datasharing for direct purposes (like the administration of your child’s schooling), and find your identifiable personal data given to Fleet Street papers. For example.

Bad idea. Utterly stupid in the 21st century, because, identity theft.
Citizens usually suggest in public engagement workshops and surveys, that government departments should collect less and share less between themselves. Why is based on past experience. The example most cited is probably the HM Revenue & Customs data breach involving the missing data disks of half the population.

One example at #ukhc15 was given was that while we may be happy to share personal data for council tax collection, and for refuse collections, and for elections, we do not expect to see all that data joined up by our council in a ‘single customer view’. Yes it may be convenient for the council, but security access is inevitably widened from original specifications and people don’t trust that ‘need to know’ would stay the definition of ‘need’ that it was when they signed up.

The Summary Care Record being opened up to pharmacists without consent is a case in point, of purposes scope creep.

Data sharing reports showing for what an individual’s data has been used by whom, will be almost the only mechanism to achieve this.

For some, there seems to be no choice but it is little discussed what is necessary, versus a government policy choice.  And without transparency or independent audits no one can see whether the benefits of this sharing outweigh the risks. I’d like to know.

Solutions around ‘purposes’ are to be found in transparency of what has been done in the past, transparency of governance of what is planned, and how it is implemented. To do this well, we need education in public and professionals. Where police forces request broad health and social care data to create maps of at risk individuals for predictive policing ‘but not for intervention’ we must question how it will be done. Solutions like transparent data use reporting.

In 2015, after students had found that agreeing to the use of their personal data in their further education application process were used for commercial purposes, UCAS surveyed UK Undergraduate students about their attitudes to data trust and personal data sharing in April 2015.

After intervention by the Information Commissioner’s Office UCAS changed their blanket consent process which enabled and obliged commercial use of personal data. 37,000 students replied and 90% of them said they wanted to be asked before data sharing.

A majority of UCAS applicants agree that sharing personal data can benefit them and support research into university admissions, but they want to keep control over their own records, with nine out of ten saying they should be asked first.

While we continue to lump commercial uses onto all datasharing for public good, we continue to reduce the number of people content to share for the public good.  Consent could accommodate this better.

All or nothing consent boxes fail when signing up to one thing, is abused to enable signing up for anything.  This page 2 of the NHS dental form is a prime pet dislike – sign here to consent to pay us (?), and allow your personal data to be processed by a randomly wide range of government departments and in Sr Lanka or India.

Solutions here are better forms, with proper separation of purposes into processing that is ‘necessary’ and processing which is not.

Which brings us to trust (and consent).

5.  There is no such thing as ‘trust’. (and consent)

Of course there is. But when we talk about trust especially in the context of NIB paperwork it often becomes abstract, as if it’s a simple ‘tick box’, something to be ‘got’ and then move on.

We must also understand who we’re talking about.

In 2016, as we see ever more “big data” I would ask that we remember data sharing must remember to treat individuals as individuals and not simply as ‘the data‘ if those who use it want to continue to have access to it and bring new people onboard.

So many organisations’ representatives say over and over, people withhold information if they can’t be anonymous. It poses risks for public health, risks to their own health, and both come at huge cost.

Trust is all about relationships between people. Organisations that want to be trustworthy, need people speaking for them who are trusted and whose actions back up what is said.  This is not PR, it’s not about ‘communicating the benefits’ it’s understanding human behaviour and responding and iterating for that.

Trust is not an absolute, but layered in time and purposes of trust and  the relationship. I trust my butcher to truss my Sunday joint better than anyone else. I wouldn’t trust him to sew me up after an operation.  And if a shop is beset by a bacteria, we will lose trust in that butcher too. Too much trust, as well as too little, can harm us.

Trust is reactive to what is said and done and built on experience. It’s a basic human survival skill and while knowing what we mean and how to be ‘trustworthy’ is quite simple, doing it, is hard if you suck at population-wide relationships as much as the DH and NHS England public image does, in a way that ‘the NHS’ as a brand does not. Being trustworthy really matters in health.

“Almost a third of people suffered long-term irreversible morbidity and 82 per cent had a worse outcome as a result of delaying seeking help.”

The “Jeremy Hunt effect” is serious. Trust influences decision making.

Trust is layered and in order to meet its needs, consent mechanisms need to be layered too.

Some changes have technical solutions  but the solutions are not of themselves technology; but enabling relationships built on trust.

6.  Solutions are about people.

“There is no one patient copied 60 million times”, was the message from one NHS Open Day and patient online event in summer 2014.

Any system that involves people is hard to model, hard to control and hard to determine a single absolute for, because we are all different.

Understanding and responding to that rather than ignoring it in communications for example, is key.

The aim of “Rebuilding and sustaining public trust” needs refocused to treat the cause, not only the symptoms of the damage done in 2014. Here’s my thoughts why.

Trying to build trust involves dialogue and ongoing work, not a tick box, get consent, and hey ho we’ve sold your daughter’s data to a Scouts group. Yup, really.

Fairness in processing and fairness in seeing how data are used, are important to building a relationship between the government and the public.

Fairness in how and what we want to achieve through the data sharing also needs thought about. The inverse care law appears magnified in its potential when applied to digital, and magnifies the importance of thinking about fairness.

7. And consent. This is among the hard things. We are told there is work-in-progress. Yet of itself it’s easy. If you want to do anything to me, or about me, ask me. How to implement it, is the hard thing.

And that may be based on choices that were made in the past.

Let’s face it, many of today’s problems with using ‘the system’, is that there is no “the system” but a collection of tools never designed to work together and never designed for today’s users. Most GP systems were designed as practice finance systems and then the DH decided they’d quite like the data centrally, and has been all bent out of shape for all sorts of purposes ever since. Personally, I’d love to scrap it all for central use and start again.  Reality might be we end up seeing some sort of bolt-ons for consent to patient online tools.

Let’s see wait and see what the new workstream of work proposes.

If patient facing services are all to be digital by 2020, we’d best get cracking. It’s only 1, 042 work days to go.

8. Big data means big thinking is needed.

Health data sharing does not have to be a trade off between privacy and the public interest and the economic good but big differences between the thinking of government, the drivers of policy, the research community, ethics panels, and the citizens of the country remain.

Increasingly talking about secondary uses of healthcare data we are talking about Big data. Uses of millions of records too big to be anything but machine read. The current state of the data hygiene of some of these big collections is matched only by the shortcomings of the framework of ethical, regulatory and legal frameworks we have created to try to control them, and risks inaccuracy causing harm.

Billions of episode statistics can be shaped into one report whose meaning is summed up in a 300 page report, read in twenty lines of an executive summary, by one decision maker with power to make change, that could affect billions of pounds in spend and impact millions of lives. Lack of education risks causing harm.

People who work with the data, whether in roles like GDS, or commercial entrepreneurs, Government departments and data scientists and all the end users,  have yet to understand what we need and why and what we are looking for.

And at the same time, lack of data risks causing harm. The case in Soham where the caretaker should not have been able to get a job in a school is one many remember, but the case of Baby P as horrific as was, child protection is often misused in data sharing stories.

More data does not of itself protect anyone.

Victoria Climbie’s case worker was overstretched with nineteen cases. The case officer in Baby P’s case had 50% more cases than the council’s limit.  Perhaps the money spent keeping data on every child in the country could be better spent adequately caring for the 1% who really need that attention.

Finding the correct balance between the power that data gives and the responsibility that comes with that power should be added to the list of “hard things”.  Transparency and oversight, again, are key.

9. But it’s all about the benefits.

I do not believe that and the drive for secondary uses of datasharing in health is mostly about the benefit to the patient. It is a top down driven programme to be of benefit to HMT and the taxpayer (who may also be a patient.) Cost reduction is a choice, yet in this time of austerity and the Department of Health unwilling to fund staff as most in the public, never minf the NHS, would like to see, if one asks for the business case of the concrete benefits and costs, they do not exist.  Who plans a big billion pound programme without a budget or cost-benefit analysis? This is our money, with all of our health records, and we should see exactly what is planned, what is problematic and understand why. Minutes should not be secret.

Policy makers should be common sensed and put this kind of planning into the publicdomain. If it’s felt a risk that we wouldn’t understand it, then explain it. If it’s off track, outside perspectives can help. And rescue billions of pounds before it’s too late. Knowledge is power, but all in politics should remember from where that power comes. To control the electorate, is another power game entirely. As Holmes would say, look at the evidence.

We have hard things in the world to solve.

Homelessness is hard, yet some people in London are making things better – supporting 100% attendance at NHS dental appointments for some sleeping on London’s streets.

How to deal with children starving in the Syrian winter is hard,  and of all the people who land in Italy and Greece, only 1.2% get as far as Calais – some people are making things better for them, getting support to those stranded in the Jungle.

Climate change, poverty, conflict, human rights across the world. We have choice and freedom in the UK to support making things better, doing nothing, or worse.

In the UK choice is based on freedoms of liberty and speech and equal rights to education, to access healthcare, to choose to be who we want to be.

Who we are is tied up in what we have been and what we want to be.

To control the personal information about our records, our past and present, and our genomes, our future.  Services are being shaped that are personalised, targeted at us, yet it is a version of ‘us’ that the system has mined, recorded or tracked, not necessarily a complete or truthful picture or one we want to share.

Transparency is not an optional extra. We want to see the costs as well as be told about  ‘the benefits’.

10. Understanding the “hard things” of identity, datasharing and consent will not become easier as we increase the complexity of what is possible to share in future. We need fixes now, fit for future.

Genomics and datasharing need careful clinical understanding and public engagement; two things often absent from their political births. How that information in health, never mind how biometrics are used in education and as identity controls is poorly understood in the population as a whole. There is far too little debate.

Most people are entirely ignorant about how their personal data, and health data, are used by third parties. Take mobile phone data that reveals the people, places and companies your smartphone is talking to and what it’s telling them, but not you.

Yet pages and pages of  complex worded ‘privacy notices’ fail to communicate what should be simple. Who will have my personal data and what will they do with it.

The reasons that some of these “hard things” are hard is in part a self-made problem from providers.  Some may not want it to be easy to understand how our data are used by others because they already know we won’t like it. We might opt out. That might hurt their commercial model or public interest planned public health research. It might. But by taking the choice away from an individual, by making it for them, providers must accept they do not make things ‘easier’ they simply postpone the ‘hard problem’.

‘The hard problem’ is that fundamentally people don’t like the idea of losing their freedoms. And to do that, the providers of commercial services, in government, in health and all sorts of areas where we share data must be prepared to share the power that knowledge creates. As Clare’s example showed, that power can be misapplied.

Knowing how to apply information is much more valuable than facts alone. Once informed, we are empowered to make better decisions, and to have choice. But those decisions can be imperfect. Whether they are made by humans or by machine.

In conclusion:

I agree with Catherine’s idea in her ten (eleven) things post that ‘questions often get you further than answers’.  We must keep asking questions. Change must be made with people, not to them.

Don’t tell me who I am.
Don’t take something from me without consent.
Don’t make a choice on my behalf without telling me.

Basically, don’t use technology to make me less of an individual, self determining, human being.

We can have better data, better privacy and better outcomes for society. These changes are based on choices.

Some things will change, or be changed.  How, why and when we make, and accept, and benefit, from change, is all about people.

We will indeed have a revolution in these hard things of identity, datasharing and consent , but get it wrong, and it may not be the revolution you are looking for.


Thoughts for the NIB / NHS England – could these be fixed?

I also thought about #userneeds How can hospital systems work better for the staff using them?

If the system simply doesn’t meet the doctor’s needs and takes them too long to input or to access information, it can create safety issues, and wastes valuable time. These multiplied across Trusts using the same providers with the same or similar systems, means multiplied loss of time and increased risk and costs.

Design a tracking tool something like the doctor asked at #ukhc2015.  Hard they said, because, whistleblowing. Because, procurement contracts.

Cornwall county council, along with Cornwall Partnership NHS Foundation Trust and Peninsula Community Health CIC recently got out of a ten year contract after two when the provider was not meeting agreed service delivery.

I’d like to think that this would be a warning to providers and a signal of hope to other Trusts, but it is likely that lengthy litigation is something that no one wants to see encouraged.

#NHSEngland #Health2020 #NIBhealth

#paperlessrecords Can political policy changes and practical user needs be reconciled?

The paperless NHS. The hunt for the golden hind of the health system or a mythical unicorn?  Everyone seems to think it exists, should exist, but when you hear some talk about what it should look like no one is quite in agreement.

One attendee from one CCG implementing the political challenge very candidly explained that they really weren’t that clear what reducing paper of itself should achieve.

Identifying what paperless records should achieve and driving towards that, rather than simply ‘being paperless’ would avoid costly mistakes such as this example where a trust had stopped scanning all paper records, suggesting that it would take 11 years to complete at the rate the trust was going, and that for inpatient services, in 80% of cases clinicians weren’t even looking at the notes.

Can NHS Engalnd get a change toolkit on a shared CCG platform of “how to”? Sometimes it hinders local efforts, but for those who don’t know where to start, they need something. Reinventing the wheel is costly in multiple consultancy fees.

#lessonslearned How not to: can case studies in both good datasharing and mistakes or things that went wrong be captured to benefit those working on similar or related programmes in future?

More reference to concrete experiences that have worked in practice are needed.

Projects experiences such as using the e-red book explained by Dr. Alison M. Devlin, Research Associate DALLAS, from some of the poorest areas of Glasgow in which the cost of data was prohibitive for some women to access the app, provides food for thought when “apps” and “data” are so often seen as panacea to solving problems.

Not addressing these issues as real, if hard to solve, wilfully ignores the risk that a digital NHS may further exacerbate the inverse care law through digital exclusion.

#whyarewedoingthis? What is the heart of nhs.alpha and how will that be communicated and embedded with the public to become the success it hopes to be? I’d still like to understand nhs.alpha better. Thoughts from summer.

As the NSPCC Childline has discovered (at 1:40), email interactions where the question and answer need to be written can take “ten times longer” than a conversation. What will the web interface deliver?

Thanks to @harryharrold I read a write up of the session I could not attend while at another.

One public listening event NHS England procured report:


Others’ #ukhc2015 ideas and notes:


And finally. What can I do by the next #ukhc2015 myself? I’m going to get back my own health and fitness back this year.  I ran London 2011, now  only 5 years later I’m not fit for it. How much can change for good or bad, in our own lives and in the #UKDigiStrategy in the course of this year,  or the next 5 is exciting. I hope it’s going in the right direction.