Caring about Sharing – The Do’s and the Wonts

By Mark Conway

Data and choice were both prominent in the news feeds yesterday. There were two specific threads which caught my attention as they highlight perfectly the dilemma and contradictions that surround the sharing of Big Data. If you happened to pick up on either of these threads then it will almost certainly have been the story which was reported on the TV news regarding NHS Digital and the opt out from patient record sharing. The other one was much more niche, but arguably no less important to so many areas of societal wellbeing. However, before we peer more closely at that particular can of worms let us first consider the more widely publicised of the two. The NHS, and how they intend to share your data – or perhaps more accurately, share their data about you.

The NHS is undoubtedly one of the few remaining publicly owned jewels in the crown of our British and UK, socio-economic infrastructure. As a nation we are rightly proud of the colossus of equality and universal caring which has risen from the desolate post war landscape. Very few of us will not know someone who is still alive purely because the NHS existed and was available and able to intervene in a positive way at a time of personal health crisis, with no regards as to that person’s ability to pay for treatment. If you stop to think about it, that is an amazingly wonderful thing, and perhaps the greatest possible endorsement of a society showing that every one of us matters equally. Forget the naysayers and detractors who claim that healthcare is so much better elsewhere, if they are not speaking with distinctly forked and serpentine tongues which seek to profit from the dissolution of existing free at point of delivery healthcare provision, then they are, at best, misunderstanding the crucial role that the NHS plays in coalescing our collective sense of identity and belonging. The National Health Service is what we have, here and now, and what we should continue to cherish and protect as if it were the closest of close family.

Whilst I usually count myself as being amongst those who disdain the insincerity behind much of the jingoistic rhetoric spouted by of some of our leading politicians, I would readily applaud any who stood up and shouted out plaudits for the NHS and suggested that it is a world leading service of which we can all be rightly proud. It is just that. And, as a consequence, it receives many jealous looks from those who are either less fortunate in what provision is available to them as individuals, or who see it as a barrier to further monetising their own health and insurance based businesses. Monetising being a very important word in that last sentence. The NHS has been running for over seventy years now, and during that time has consistently had a ‘customer’ base of upwards of fifty million people. While admittedly the record keeping dating back to the middle of the last century may leave something to be desired, in modern day terms, there is no doubt that the digitised medical and patient records held by the collective parts of the NHS represent an unprecedented source of medical and socio-economic information.

When I mentioned earlier that the NHS was eyed with considerable jealousy by certain other parties I did not just mean those based outside the UK. The NHS has something which is literally worth a fortune to those in the health and pharmaceuticals industries regardless of wherever they are located, and that is Data. Big Data. Indeed, in terms of medical histories for tens of millions of people over decades of treatment then we are talking Really Big Data. If you look at the giants of the current commercial world it is hard to get away from the success stories which are represented by data centric companies. Google, Facebook, Twitter, Amazon, and Alibaba, all have a focus on technology, but what really gives them value as companies is their access to data and their abilities to use that data to create wealth. Data can be used in logistics, making ordering, stock management, and deliveries so much more efficient. Data can be used in marketing, analysing customer’s purchasing and browsing behaviours to identify who they are, what they like, what they may find interesting, and what they are very likely to be willing to pay for. Data analytics gives an unprecedented ability for those controlling the data warehouses to predict, and even control, how society acts. If social media ‘influencers’ are influencing people’s style, living and purchasing choices in a small but obvious manner, then multiply that up by several factors of magnitude and that is what the likes of Google and Amazon are doing to all of us in a subliminal way, every day of the year. If you have ever seen the classic film Glengarry Glen Ross then you might have some idea of how important and valuable a ‘hot lead’ used to be in marketing and sales. Well, they still are important, vitally so, but nowadays the intensely pressurised sales force are less likely to be human and far more likely to be AI informed algorithms which have filtered through the millions of bits of data which your online lives have presented to them in order to determine just what inner buttons you are most likely to react to having pressed.

So, data is big business, and very, very valuable as a monetised commodity. Every web page which we browse will undoubtedly have a privacy and data policy attached. They have to, it is the law in most countries, certainly so in any which are, or have been part of, the European Union. To most of us these policies are just minor irritants which pop up when we first access or use an online service. Most of us simply look to the OK button and click it without thinking, or caring, about just what our mute acceptance of terms and conditions which we will never read might actually mean for us. Those checkboxes which ask us to confirm that we have read the T&Cs are the one-click gateway to yet more of our behavioural data being collected, collated and eventually monetised. We have become so used to having amazing apps and services made ‘freely’ available to us online that we have either forgotten the basic tenet about there being no such thing as a free lunch, or are subliminally aware of it at the back of our minds, but simply choose not to care. Perhaps you think that you are different, that you don’t simply ‘default’ your acceptance in such matters, or that perhaps you don’t avail yourself of this ‘free’ app culture. If so, then I would suggest that instead of the word ‘lunch’ you transpose ‘email’, ‘tweet’, ‘map’, ‘music stream’, ‘news/weather/sports update’, ‘swipe right’, ‘hotel review’, ‘spreadsheet’, ‘chat forum’, ‘job search’, or even, ‘cloud storage’. All of these services are freely available ‘free’. Free in terms of upfront monetary outlay, perhaps, but certainly not free in terms of data outlay.

And for most of us this is an acceptable play-off. We get something which is instantly useful in exchange for something which we either do not particularly value, or which we pay scant interest in the potential uses and longer term consequences of sharing so freely. Although it is all data which is personal to us, we don’t usually think of it as being personal data. So why is the NHS Digital opt out different? And for that matter should it really be different? It is intrinsic human nature to be wary about anything which affects or impacts upon our own health. That includes awareness and information that may exist about our personal health, and even more so when it relates to our ill health. Nobody wants their vulnerabilities and frailties broadcast to all and sundry. If you are sitting on a train, bus, or even attending a theatre show (remember those), then do you really want the person sitting opposite, or two rows down, knowing that you have high blood pressure, or that you had bad piles last year, or that you have a nasty little rash hidden away under a protective layer of clothes. If having that information out there for all to see doesn’t bother you then I applaud your openness, while questioning your sanity. Of course, if I had access to your full medical records then I be able to discover that other people, with greater medical credentials than my own, had also questioned it previously.

All of which leaves us with a slight dilemma. One which is based around the concepts of trust and demarcation. How willing are we to believe the politicians and authorities when they claim that no medical information will be traceable back to us as individuals. How willing are we to trust them when they say that information will not be sold for commercial purposes. And how willing are we to share personally sensitive data which may be of use to society as a whole, through medical research, but which may end up being used to brand us due to anatomical issues which we have no control over. There is a massive public verses commercial interest in anything that involves medical histories and biological facts. You may remember a few years back when genome sequencing first became widely possible. There was an almighty scramble between ‘public’ research projects and commercial interests to try and win the race to sequence as much of the human DNA map as possible. There was a distinct fear that if such information was considered a proprietary right then medical advances of life saving treatments could be held to ransom by corporations claiming copyright to the DNA information needed to develop them. At the time, many bio-scientists were keen to prevent such a thing coming to pass by rushing to publish their findings and so make it publicly available knowledge, and therefore disbarred from licensable copyright protection. Fortunately, sense prevailed, and thankfully we do not find ourselves in the position where the chances of survival of our children and loved ones are based on who has the right to be paid for knowledge about specific genes, as all the data about what makes us who we are is now ‘open source’.

As with capitalism in general, it is too simple to say that public interest is good and commercial interest bad. We live in a world where money is the foundation of everything that we rely on, and almost everything that has changed our societies and individual lives for the better in recent decades will have originally come from someone with an entrepreneurial interest in making money. Commercial interests have long been the innovative spark that has led to life saving medicines and technologies being developed. Medicine is not cheap. Whether it is a new headache cure, a gene-targeting RNA protein vector to splice DNA strands causing harmful diseases such as Sickle Cell Anaemia, or a potential cure for baldness, all cost hundreds of millions to safely develop, trial and license. Big Pharma is Big Business with Big Costs. Pharmaceutical companies are not philanthropic ventures. Everything which they develop has to be paid for somehow. What this does mean is that certain fields of medical science which could have significant benefit to relatively small numbers of people are often neglected in favour of others where there is less benefit but a much wider potential market through which to recoup monies spent on research. That doesn’t mean that very niche drugs and treatments are not worked upon but where they are they tend to be with a view to informing pure science as well as future, more profitable, developments. When they are made available to patients they tend to come with eye-wateringly high costs, sometimes even millions of pounds for treating a single individual with a particularly rare condition.

It is of course almost impossible to view any current topic without factoring in the filter of Covid-19, and in terms of medical progress there is an obvious shining example of how sharing medical data can benefit wider communities. The sharing of genome information on the Covid-19 virus helped significantly in the development of vaccines. The sharing of medical histories on how various forms of vaccines interact within the human physiology helped just as much as it allowed scientists to start from a position of already knowing what was possible, and how the human body would react in known circumstances. Undoubtedly the current success rates of Covid vaccines on a global level point to the immeasurable benefits that data sharing can bring. Just as undoubtedly the geographical limitations in vaccine availability, due to cost and scales of production, point to the importance of economic factors when developing and distributing life saving medicines. The UK government invested heavily in the production of the Oxford Astra-Zenica vaccine which has been proven to be a wise investment, but spare a thought for the countries who did not have the foresight and resources to do so, or who invested in vaccines which failed to work. As indeed could have been the case with the Ox A-Z vaccine. What comfort are such countries and governments to take from the successes of vaccines which are being rolled out in richer areas of the world. To some extent the UK government’s investment in the development and large scale pre-ordering of vaccines was as a direct result of their being forced to gamble big time to make up for failings in protective measures which had led to the UK, for a while, having the highest rates of Covid infection and deaths in the whole world. It now looks like being a gamble which has indeed paid off, the political equivalent of a Hail-Mary throw in American football, and we can all breathe a very big sigh of relief as a result, but that should not gloss over fact that it was a gamble which they felt they had no alternative but to take.

Before we leave the topic of Covid we should perhaps reflect on the slightly darker side to the unprecedented success that so many Covid vaccines are having worldwide. This is the first Coronavirus for which there has been a single successful vaccine produced, and not only one but many. All considerably different from each other in terms of how they act, and how they are vectored for tackling the virus structure. So how has this come to pass when MURS and SARS had no such treatment, and when other virus borne conditions such as Ebola continue to plague lesser developed countries? Unfortunately, it is hard to get away from the suspicion that much of it has to do with self-interest and commercial viability. Covid-19 has unified the world in terms of political will and commercial interest, but perhaps not so for viruses which remain mostly endemic to poorer regions.

Back to NHS Digital, and specifically the current developments around The General Practice Data for Planning and Research, to which the opt out is currently a newsworthy topic of interest, is something which is of importance to everyone. The Health Secretary, Matt Hancock, has received some negative reaction to his seemingly out of hand dismissal of suggestions that all members of the public should be sent letters outlining the proposals for roll out of data sharing and the steps which they can take to opt out. It is entirely probable that, given the negative attention which he has been attracting of late due to revelations and leaks concerning his department’s response to Covid in the early days of the pandemic, that further brickbats will merely bounce off unnoticed, but it is slightly concerning to see such a disregard of the importance and significance of personal data sharing. Most people are generally willing to accept the arguments for greater good if they are presented to them in clear and open language, and this will probably extend to allowing their personal medical records to be shared in anonymised form to enable research to be carried out in future. But the right to choose whether to allow this or not is now such a fundamental precept of life in a modern datacentric world that any precedent which diminishes or marginalises the significance of being able to exercise such a right has a ring of dissonance about it. The Nanny State knows best, it seems.

Perhaps the biggest worry which people have about the proposed data sharing is not about their own personal and private data being made available to all who express a ‘genuine’ interest in accessing it, but in the uses to which it might eventually be put. As we have discovered, Big Data is Big Business. In countries such as the USA there is a strong and vibrant private health economy. Rather than a one-size treats all, nationally funded health service, similar to the NHS, there is instead a massive insurance backed industry providing top quality health provision to those who can afford it, and considerably lesser quality provision to those who can’t. The main reason why successive Democrat governments have struggled to get improved Medicare and Medicaid style programs through congress is that there is a massively powerful lobby movement, backed by the medical insurance industry, who do not want to see their products undercut by free health provision. If you are making a lot of money from selling bottled water in the desert then the last thing you want is for someone to come along and drill a free-to-use artesian well right next to your bottle stand. It may be a laudable social development, but it sure hurts the profit margins.

Another thing that seriously impacts upon the profit margins of insurance companies, whether health care specific or more generally, is having to pay out claims where, with a bit of pre-informed knowledge, they would not have needed to. Or, at least, not for the level of premiums which were being charged. Here again is a major benefit of being able to apply actuarial science to medical data. If you can identify traits and conditions which are more likely to lead to required medical treatment then you can raise the premiums being charged to people exhibiting those traits. Or better still, you can refuse to offer cover to them at all. The classic example of this from recent history is the standard question which appears in all life and medical insurance screens - have you ever smoked? This question is not based on any prejudice against the habit of smoking itself and is entirely founded upon the known statistical links between smoking and the onset of serious health conditions. Put plainly, a heavy smoker is a high risk insurance customer and will be treated as such by any insurance provider. Just as someone with a family history of coronary disease poses a greater risk of needing invasive treatment in future, or indeed of dying from a heart attack, and thus of a claim being made against any life cover held in that person’s name. So medical data, even in anonymised form, can be used to make decisions which can and will have financial impact for those people who might be prone to particular health issues, whether or not they themselves allowed their NHS Digital data to be shared. A discovered link between a particular gene, or a medical condition, and increased liability of medical treatment or lowered life expectancy will become a monetised piece of information. For that matter it doesn’t even have to be a seemingly negative condition which impacts you financially. If it is discovered that your particular genome configuration is linked to longer and healthier life then yes, it may well lead to lower life insurance premiums, but woe betide you if your prospective pension provider finds out. Thus it becomes apparent that your own opt out may not be enough to prevent your personal medical condition becoming a factor which is used against you.

Which is where I totally switch angles and turn to the other ‘data share’ news feed article which caught my eye. For those who know me it will come as no surprise to learn that this involves the gambling industry. As a previous Disordered Gambler I managed to effectively ruin my own and my families finances, and to also abuse the trust of many others including a previous employer, due to my having developed a compulsive addiction to gambling. The upshot (or downshot) is that I gambled away well over a million pounds through online sports betting, the vast majority of which was not my own money. An article in Gambling News yesterday touched upon another unfortunate individual who had followed a similar path to my own and was now being sent to prison as a consequence of crimes which he had committed in order to finance his own Gambling Disorder. The article touched upon the large amounts of money which he acquired and gambled away and also upon something known as the Single Customer View and how, if it were to be introduced, the gambling companies would be in a better position to identify individuals exhibiting signs of extreme disordered and unaffordable gambling.

The idea behind the mooting of SCV is that although individual gambling operators are able to see the level of spend and behaviour carried out by individuals who are customers of their own brands, they are unable to see whether that person is also betting with other companies. This leads to situations where, because an individual is spreading their gambling activities across several betting accounts, it is ‘unreasonable’ to expect the operators to be aware of this and to factor it in to their own mechanisms for identifying potential unaffordable and unhealthy gambling behaviours. More specifically, it also acts as a pretext for excusing them from identifying potential breaches of AML (Anti-Money Laundering) thresholds which they are duty bound to investigate and report via the submission of a Suspicious Activity Report to the National Crime Agency, and also to the UK gambling regulator, the Gambling Commission. This is a known issue in the gambling sector and one which has come in for greater scrutiny of late due to ever increasing numbers of high value, high publicity fraud connected to disordered gambling, and also as a result of the on-going review of the 2005 Gambling Act. The SCV is seen as many as being a vital component in ensuring that gambling harms are minimised and that criminal funds can be kept out of gambling. It would work by all gambling operators submitting details of their own customers, using an anonymised token approach, which would allow customers with accounts across various operators to be matched and thus would give a holistic view of their activities across the whole gambling sector rather than with just individual companies.

In principle this seems a perfectly workable solution to a problem which exists at a non-granular level. It has many elements in common with the parallel issue of the NHS Digital data share roll out, such as the rights of individuals not to be identified, the benefits to scientific research through access to larger data repositories, the commercial interests of private enterprise, and the public interest of creating a more socially just and crime free society. And yet the roles and perspectives of those involved can be viewed as almost being diametrically opposite when comparing the two proposed ventures. In the case of NHS Digital it is the authority representing the public interest who are the gate holders to the data and it is the private enterprise industries who seek to gain access to it. It is also the rights of individuals not to have their data shared and used against them by commercial interests which is being put at risk due to the actions of the government. The SCV scenario flips this about by making it the private enterprise companies who currently hold the data and the public authorities who see benefit in accessing it. The personal rights are the rights of the customer not to have their gambling activities made known to centralised authority against the interests which the government has in ensuring that such information is indeed shared by those with an invested commercial interest in it.

On the face of it the SCV situation should be a win-win scenario for all concerned. By co-operating with each other, and also with the government, the gambling operators can take a giant step towards being able to justifiably claim that they preside over a responsible and crime free leisure industry. Customers harming their own legitimate finances through unaffordable levels of gambling will be able to be identified and to receive intervention designed to prevent their harm continuing. Criminals using gambling sites to launder unlawful proceeds will be detected and stopped from doing so. There will of course be concerns over sensitive commercial data on their customers being potentially made available to all, and to each competitor having a data driven insight into how the other companies are performing, but these are all concerns that can be easily addressed through the creation and appointment of an independent data controller charged with ensuring commercially sensitive data is obfuscated. That way the only people adversely affected are those who actively seek to apply criminal funds to gambling activities. Even there it is not quite as black and white as might appear the case. Speaking as someone who caused a significant level of financial harm to myself and to those around me, I would actually have welcomed someone stepping in and stopping me from perpetuating my disordered and criminal behaviour. Perhaps not immediately, but absolutely so in hindsight. My Gambling Disorder cost almost everyone around me a lot of needless harm. I say ‘almost everyone’ because there is one glaringly obvious and notable exception. The people who were not harmed by my action, and who in fact benefitted greatly from them, were the gambling operators.

And here is the rub of it all. It simply does not fit in with the commercial interests of the gambling industry for a Single Customer View to be implemented successfully. ?Just as it is not in the short term interests of the gambling industry for there to be a total shift to fully responsible and sustainable gambling. Research carried out into the percentage of revenues acquired by operators from ‘problem’ or disordered gamblers is staggeringly high. Official figures for the percentage of individuals who meet criteria for being considered ‘problem’ gamblers tend to sit around the 0.5% mark. Unofficial figures tend to agree that the real figure sits closer to 2%. Both relatively low, one might claim, even though this would place the actual numbers in the UK as being in the order of 1.1 million adults, but looking closer we can see the disproportionate benefit gained by the industry from permitting such gamblers to continue betting. In 2004, Robert J Williams conducted research into the Proportion of Gaming Revenue Derived From Problem Gamblers. Examining it in a Canadian context he utilised research from four countries including Australia, Canada, New Zealand and the USA. That research identified that, respectively, 33%, 30%, 19% and 15% of all gambling revenues in those countries came from ‘problem gamblers’. Bear in mind also that this research predated the explosion in online gambling, a sector which has proven to be more susceptible to disordered gambling behaviour than land based operations, and one which has seem revenues for UK based gambling derived from remote and online products shooting up from approximately ￡600 million to over ￡5 billion in the period since this the Williams report was collated. The UK regulator, the Gambling Commission, have previously acknowledged that disproportionately high levels of revenue, and thus profits, are derived from individuals exhibiting clear signs of gambling disorder. Up to an incredible 83% of deposits were identified as coming from just 2% of customers for one particular operator. Another operator admitted receiving 58% of revenues from just 5% of customers, while a third reported 48% of revenues from 3% of customers.

So why would the operators want to actively help in identifying customers who either may be funding their gambling through criminal activities or who may be ruining their own finances through uncontrolled and harmful levels of disordered gambling? Why indeed, when literally billions of pounds profit are currently being generated by exactly the same people whom a SCV data sharing system would be intentionally trying to prevent gambling.

I therefore make the following prediction. Whereas the UK government will indeed push through the wider sharing of NHS Digital owned data, because in spite of half-hearted official denials it is in the commercial interests of powerful lobbying bodies for them to do so, the implementation of a Single Customer View in the online gambling sector will either fail to occur at all, or will do so in such a watered down manner that it will be reminiscent of the ‘protections’ afforded to members of the public against intrusive media interference through the creation of the Press Complaints Commission. In other words, it will only come to pass if it is the industry policing the industry in a manner which actively stymies the whole rational intent behind the creation of such a body, and all because it is not in the commercial interests of those equally powerful lobbying bodies representing parties who see it as potentially harming their bottom line.

Of course, it could also be argued that the current manifestation of the gambling regulator is just such a partisan creation as it stands, based purely on their total failure to call the operators who knowingly profit from patently vulnerable customers and blatant criminal enterprise to due account.

No spoons were bent in the making of these predictions.

?