Joe’s view of care.data

How embarrassing. I thought there was nobody else in the house, so I turned on the TV and flicked through the channels.

“No one will ever know,” I thought. I put on my tin foil hat and turned up the volume. Suddenly Jake, my 17 year-old son, appeared from nowhere, yelling: “Daaaaad. Not again!”

“Look, it’s important!” I said, caught red-handed. “My mate Tony just tipped me off on Twitter that care.data is being debated live on parliamentary TV. It doesn’t get any more exciting than this.”

“Well, if it’s so important, how come nobody is there? This stuff is like porn to you and all your computery patient safety geek mates. You need to get out more.”

Launched and un-launched

He had a point. I counted only 12 people in the debating chamber. However, I was impressed by the quality of some the points they made, and the member for Easington seemed particularly well informed.

The gist of the debate was that the government is trying to retrofit proper governance arrangements to the care.data project in an attempt to get the governance genie back in the bottle and regain public trust.

It’s having to do this because care.data was launched, came under heavy fire, and got un-launched in short order.

The programme to expand the Hospital Episode Statistics, link them with other datasets, and make the data available to researchers and others received a vicious mauling in all the mainstream press.

It was brilliantly and mercilessly lampooned by members of the privacy zealot community, upset that NHS England was trying to operate an opt-out model, by putting out leaflets that failed to name the programme or include an opt-out form.

Tim Kelsey, NHS England’s director of patients and information, was portrayed as a raving Adolph Hitler figure – berating the privacy lobbyists as “tin foil hat wearers” – in a ‘Downfall’ spoof that received more than 12,000 views on YouTube.

Too important to fail

Ok. So we all had a jolly good laugh at that. But the tragedy is that I suspect most of you, like me, believe in care.data. We got involved in health IT in the desire to harness the power of computing to improve healthcare and outcomes for patients.

We want the Health and Social Care Information Centre to become the highly trusted keeper of all our potentially sensitive data and a hub for world class research.

We want top quality, anonymised data to inform commissioning so that we can get best value for the tax payer in these harsh economic times.

Everybody is talking about the power of big data, but the whole project has had to go on the back burner while the information governance mess gets sorted out.

Not learning the lessons of history  

It wouldn’t be so bad if we hadn’t been here before, and quite recently. We have a Data Protection Act that insists patient identifiable data can only be shared with the subject’s permission.

So, consequently, there was a predictable outcry from the privacy lobby when aspects of the project appeared to challenge that.

In addition, the handling of care.data suggests that we, collectively, learned nothing from the NHS Summary Care Record model of consent debacle.

Wrangles over whether that should be opt-in or opt-out delayed that project for years; and it only wanted to create a limited amount of data for medics who would otherwise have no patient records at all.

It also alienated a number of intelligent and committed campaigners, who wanted nothing more than what they saw as the proper application of the law.

Now, I’m not a lawyer but I know that Parliament passes laws that contradict each other and then decisions have to be made on the basis of case law, at which point judges have to decide which way to jump.

My guess is that the rather higher minded Data Protection Act is likely to trump the practicalities of the Health and Social Care Act if push comes to legal shove.

Consequently, unless there is going to be a major review of legislation – and Dame Fiona Caldicott’s second review of information governance saw no need for this – things are likely to get messy and potentially expensive, no matter how much retrofitting of the governance is done.

Meanwhile, of course, care.data has also found itself at the centre of the perfect storm of paranoia about democratic governments taking liberties with citizen information.

Or Tesco  

That storm has been fuelled by recent revelations from Bradley Manning and Edward Snowden about the extent of US and UK surveillance activities. Yet, surely, someone is doing big data with more success than this.

I looked around for big data projects. Tesco Clubcard was an obvious starting point. I went to the website and looked into getting a Clubcard, I read the company’s privacy policy and assessed the benefits.

Basically, if I signed on the dotted line I agreed to Tesco being able to use all my shopping related data; in return for personalised special offers and rewards.

Tesco had a well sign posted privacy policy which I suspect few bother to read before ticking the consent box. My consent had to be explicit, however.

Similarly, when I checked the consent models for Google – explicit consent required – and for Facebook – explicit consent required – I found the same thing.

Now, if these very large and very profitable companies have all gone for an explicit model of consent, why has the NHS twice tried an opt-out model with predictably negative results?

I am pretty sure that most people would be willing to share their data to further the cause of medical science. But we can’t go on making the same mistake over and again by not getting proper, informed consent.

Hearing clinical views

Maybe an NHS loyalty card scheme would deliver the big medical data we crave. Why not reward members with, say, 20% off prescription charges in return for consenting to share their data?

Those worried about their data being shared could be reassured, and those not so worried would still get something back for their help.

Then we could all move forward and get behind care.data and making the NHS a global leader in healthcare research.

Or perhaps we just need the project to be clinically, rather than technically or managerially led, in order to avoid further pitfalls in the future.

Historically, it has been the case that clinical input to NHS IT has been too little too late; and sometimes ignored if the clinical view isn’t what was ordered.

I do hope that Mark Davies’ slightly mysterious departure from HSCIC and Martin Severs’ arrival doesn’t represent the board shopping around for a different clinical opinion.

Also, that the care.data advisory group that was announced last week, and the new communications manager for the programme, can break the impasse. I want to put that foil hat away. 

Joe McDonald

Joe McDonald is a practising NHS consultant psychiatrist. Over the past five years he has been an NHS trust medical director and national clinical lead for IT at NHS Connecting for Health – a stint that included 18 months as medical director of the Lorenzo delivery team!

His experiences in the National Programme for IT in the NHS have left him with a passion for usability and "end user knowledge networks.” He is the founding chairman of the National Mental Health Informatics Network. Motto: we don't get fooled again. Follow him on twitter @CompareSoftware