The Center for Data Innovation spoke with Richard Corbridge, the chief information officer of the Health Service Executive in the Republic of Ireland. Corbridge discussed the challenges of digitizing healthcare data and the role for wearables, apps and data analysis in public health.
Nick Wallace: You oversaw the introduction of individual health identifies (IHIs) in Ireland, also sometimes called unique patient identifiers elsewhere. What role do these play in making better use of patient data, and improving care?
Richard Corbridge: We put the individual health identifier in place because currently we can’t identify individual patients digitally. A lot of patient records in Ireland are still paper-based, and there is no unique identifier to link them together from hospital to hospital, regardless of whether they are digitized. The IHI is essentially a data set that isn’t clinical; it’s identifying data about the individual. That information can effectively be used as an index that allows us to link information from hospital to hospital, and care-setting to care-setting.
The crucial role of the IFI is twofold. One is patient safety. When we link a referral to an appointment to a prescription to a discharge note, we can use the IFI to make sure that we’ve got the right person. The intention is that the IFI allows those checks to be computer-based rather than eye-based.
Second, in the future, as we start to introduce electronic health records (EHRs), the IHI will be the key to linking information from different hospitals and labs, and will help to remove some of the waste involved in constantly asking the same questions. For example, it means being able to do tests once because the data remains valid afterwards, instead of re-doing tests simply because we don’t have access to the information from the previous tests. Our EHR strategy in Ireland is not to create one single big electronic health record, it’s to link the EHRs that exist, and therefore without the IHI, we wouldn’t be able to do that.
One of the criticisms of the IHI is around privacy. I think it’s fascinating to see that one of the things we actually see the IFI doing is improving privacy, because by having an IHI—as well as an identifier for clinicians—we can now tell who has looked at what data, where, and when. Whereas when the data is on paper, or when it’s digital but without a single identifier, we can’t actually do that.
Some of the objections come from people concerned about linking of government data and sharing of government data. There’s also a misunderstanding. People are worried about their health information. The IHI doesn’t contain any health information. In fact, we use the IHI to establish consent over how health data is used. There are two flags linked to the IHI: one is consent to share data for clinical purposes, and one is consent to share for research purposes. By having those two flags in there it gives people a chance to opt-in and out over time.
But the key privacy concerns always go back to governments having information, linking information, and questions like “who’s to say you’re not going to sell that information to insurance firms?” and “who’s to say you’re not going to use that information in other government activity?” The Irish government’s decision’s decision has been to use the PPSN (Personal Public Service Number)—which effectively is a social security number—as a data item of the IHI. That means we can look up the IHI from the social security number. We have taken the data feed from the Department of Social Protection, because they have the best data set for people in Ireland: we’re not linking the social security information, we’re using it to correct other information.
NW: Some time ago you wrote about the importance of open data for healthcare in Ireland and England, where you worked previously. How would you compare the two countries’ strategies?
RC: Ireland is just starting on that journey, so there’s a difference there from the point of view that there’s a lot that’s been learned from the UK about what data can be shared and why it should be shared. What we’ve tried to do in Ireland is learn from those lessons and apply them. We’ve released around 70 healthcare data sets that are truly identity-free to inspire other organizations that are now using open data to move forward. We are under quite a lot of criticism for not moving quick enough, in as much as Ireland perhaps should have done more in health around open data and released more information. But we have two problems. One is that when we don’t have digital information, making it open is a little more difficult. We also need to be really clear on the strategies for de-identified information.
We aren’t just looking to England for best practice—open data initiatives seem to be springing up all over. But one of the reasons for doing it in England was to try and facilitate the marketplace, particularly SMEs (small and medium-sized enterprises) and startups, to encourage them to use data. We are now seeing that happen in Ireland as well.
NW: What do you think about the prospects for wearable technology in healthcare?
RC: There’s one use we’ve been talking about a fair bit recently. There’s an Irish company called TickerFit who have a wearable device and monitoring tool that can be used to prescribe an amount of activity, particularly targeted at patients who’ve been through heart surgery. It can prescribe an amount of activity, and then with the consent of the patient can monitor that. I think what is fascinating about this is the original assumption was that this would be something where we were pushing for people to do more. But actually, it enables to spot that patients after heart surgery are doing too much, too quickly, and therefore we’re asking them to reduce the amount of exercise in order to keep them out of hospital, and fit and healthy for longer.
There are a couple of other examples less to do with wearables, but still do do with apps and the things people have with them. Something we’ve put in place for the care of bipolar patients is a chatbot. You put the app on the patient’s phone, and with the consent of the patient, you allow that chatbot to ask questions each morning. Because it’s a small piece of AI, it can learn what the patient is saying, and perhaps connect them to a piece of guidance, or connect them to a friend, or indeed connect them to their clinician. So it’s quite a different way of incorporating technology into what we’re doing.
NW: The European Commission recently called for input on priorities for smart wearables, particularly in healthcare. There’s a big question around how best to regulate them, since it is becoming harder to decide where a consumer device ends and a medical device starts. As somebody who is responsible for purchasing and implementing technology in the Irish health service, what would you like to see from policymakers in Brussels and Dublin on this front?
RC: One of the risks is that these are just consumer goods at the moment: patients are coming to see doctors with their Apple watch, with their FitBit, with an app that says X, Y and Z, and we need to be able to give clinicians advice and guidance that they can trust on what’s right and wrong. They need to understand the differences between them: wear a FitBit or an Apple Watch for a day and it’ll give you completely different results on your exercise or your heart rate.
I wonder if it should be more about what the information is used for and less about the device itself. There is a risk that there is so much information coming at clinicians now, we need to give them guidance on what they can use it for, perhaps a little more than regulating devices. There’s a lot of talk about Kitemarks (a UK quality standard) and how we get that right for things like wearables. I wonder if it should be more to do with the information at this point in time.
NW: What would you say are the biggest barriers to more data-driven healthcare in Ireland, and what kind of data do you think matters most in improving things?
Our biggest barrier is that for ten years we have not invested in digital in healthcare, and we are now starting to. But our journey is a slow journey, and appetite and enthusiasm move more quickly than funding and resources to allow Ireland to catch up. Ireland is said to be the last first-world country that doesn’t have an EHR. So you can imagine that sets the tone somewhat for where we are and what we’re going to do. We had our first digital hospital go live in December, which is great, and we now have plans to move another three of our fifty hospitals to that kind of status over 2017. But that’s a slow journey.
The datasets we want to work with are very much defined from the beginning, trying to learn from the United States, and the UK in particular, about why we collect information and what that information is for. The complaints from the United States in particular would be that clinicians are expected to collect information not for care of patients patients, but for…well, let’s call it counting beans, where information is collected for performance management. What we’re saying is that we always must collect information that allows us to deliver care, and a by-product of that could be performance management, or “secondary use” as it’s perhaps commonly known. We try to make sure that the data we collect can, where appropriate, have secondary use, but what matters most is keeping clinicians engaged and giving them systems that build datasets that are about the delivery of healthcare, and less about monitoring performance. That’s our starting point.