Justin Richards MBCS, caught up with Dr Nigel Millar, Chief Medical Officer, Canterbury District Health Board, New Zealand, to ask him about the differences in health informatics between the UK and New Zealand, how their new shared care record system stood up in the face of a recent earthquake and what the main challenges will be that face the health sector over the next few years.

Introduce yourself and explain why you are in the UK at the moment?

I’m Nigel Millar, Chief Medical Officer for the Canterbury District Health Board, New Zealand. It’s the health board that covers half a million people I’m the medical adviser to the chief executive of the board and I also work as a general medicine physician part of my time.

I’m here really to learn in the first instant, here in England, but also to bring over some of our ideas and experiences that we’ve been going through recently in Canterbury. You obviously heard about the earthquakes we’ve had and how that’s impacted on the health system, and how we’ve been transforming how we work in Canterbury before the earthquakes and afterwards and also what we’ve been doing with our health information systems.

From your own experiences, what do you see as being the main differences between the UK and New Zealand health industries?

Actually the similarities are quite striking; there’s not that much difference. We have a funding system in New Zealand that I would call ideal, where we have a District Health Board that covers a region and has all the funding for the hospital system, primary care, GPs, community services and residential care for old people. So we have an embracing funding system which is quite useful because the incentive is to get it right, not to transfer the problem somewhere else, that’s probably our major advantage.

There’s a little bit of difference around primary care; there’s a co-payment for patients, apart from that the actual style of practice and the way things hold together is very similar. I meet a lot of UK doctors who work in New Zealand regularly, often as what we call medical resident medical officers or junior doctors and they are quite taken with how similar things are actually and how few differences there are.

Can you tell us how you developed shared care records for New Zealand?

We do it for Canterbury; we haven’t quite managed for New Zealand yet! The share care record has got a bit of history behind it and we’ve been working on it as a concept for a number of years. It came about as a whole series of workshops which were part of the health service. In Canterbury, where we got together large groups of people, sometimes up to 150 on a particular theme involving patients, (GPs, community nurses, people from the hospital) and we must have done about a dozen of those and for every single one, one of the outputs was ‘we’ve got to find a way to share information,’ because they could see how deficient the care was when information wasn’t available at the right point. Information was held in one place and not passed on so that gave us an impetus and we’ve been working on it for a while - how could we do this - as you are probably aware. It’s a big task and has a lot of complexity around it.

Then we had an earthquake. A couple of things happened. I’ve been very enthralled with health information systems for many years and people have often said to me ‘that’s not a good idea, what will happen with an earthquake? All your computers will go down and you’d lose everything’. And, actually, what happened was the opposite, the computers kept going, the hospital information ran right through and scarcely missed a beat, but we couldn’t get into the medical records department, because the shelves had fallen over!

Some of the medical records we couldn’t get for three months. Some of the community providers and community agencies had paper records that were left in a building they weren’t able to re-enter and it had to be demolished with the records still inside and they were lost forever. So it was a bit of a reverse experience of what we expected.

What we learnt in that was where the records weren’t joined up it took a while to rebuild things so general practice was a bit more challenged because the server might have been in the GP’s office. Anyway they were well backed up and we managed to recreate. For example, one GP practice was lost in a catastrophic building collapse and the GP community, with the backups, were able to recreate that practice in another place within 48 hours, which was quite impressive.

What did happen was people went to other parts of the country after the earthquake to go and stay with relatives and friends and they had a problem because they couldn’t ring their GP because the practice wasn’t there, and they didn’t have any records, all they had was what they had in their head, which is a common thing, actually, because if you think about it, people often go to the emergency department and the whole medical record they have between the doctor and the emergency department and the patient is what’s in the patient’s head.

So we had that experience and it was quite disturbing and we realised if we had a better way of sharing information between primary care and the hospital in some secure, centralised platform, we could share it round the country in an emergency. That was the starting of the thought process, but then we realised it was a good thing anyway. After the earthquake of course, we had a huge threat to public health, people living in overcrowded conditions, poor housing with risk of infection; a lot of other stresses that meant we expected a big demand on the health system and we had less resources; we lost wards, lost beds.

So after a whole series of tragedies we realised that if we can really rapidly find a way to share information we can provide a more effective, safe and more efficient health system and therefore ameliorate some of that tension.

After an earthquake you think differently; the small things don’t worry you so much. So we were able to sit around the table and have the difficult conversation about how we can share this information and challenge each other about what the problems were and what the perceptions were, could it be done or not, and what things would be problems. We found that doing it was a lot better than not, and we found a way forward so within a relatively short period of time we had a commitment to act within a week or two and then we set about a rapid agile development process to put together a shared record and one of the key concepts was going to be, or had to be, agnostic.

So it wasn’t owned by any one part of the health system, it was owned by the whole health system, because we wanted to operate this so that no individual provider had an ownership in it except we all owned it together on behalf of the patient and the second part was it had to be patient-centric. That’s a bit of a shift, because most health information systems are provider-centric, so this is a bit of a change and that was the key step and then we set about getting to work on it.

Many people in the UK are sceptical about sharing patient data because of the security implications - how did you get round that in New Zealand?

Those concerns were definitely there; Is this the right thing to do, is it ok? In particular, the GPs find themselves were under pressure because they felt they had this information, but is it ok to share it? But behind that are a whole series of tensions and concerns and worries because the information the GP manages is quite different information, it’s come from a different source than the information that you have in the hospital.

I work as a hospital commissioner and I expect the information I create to be read by various people. It’s normal, it’s what you do. A GP works in a practice where they would be in a room with the door closed with the patient and a computer and basically that becomes an internal process so the information recorded, and the style of it and the content of it is really just for the GP’s eyes only so you can’t just open that up to the world, you have to think: what does that mean? So that’s a real challenge; what information has the GP got that is ok to share in that context?

The second thing is from a GP’s point of view; what are the business implications? Because in New Zealand GPs get a co-payment, about 40 - 50 per cent of their income comes directly from the patients. So if they share information and somebody else has that medical information, the patient could go somewhere else. So, for example, if we share, as we have done, we brought the community pharmacist into the process, if the pharmacist had a good set of clinical information, maybe the patient will go straight there and bypass the GP or we have after hours GP services that work in the evenings to provide cover, maybe the patient can go there because they have the records.

That’s the next thing, that’s quite a challenge, that’s a viral threat. Of course that’s important to us in the District Health Board, we need general practice, we can’t have their business undermined by what we want to do, so we have to acknowledge that and understand it. And the other part is, of course, if we had all the activity information down to the last detail from general practice, maybe as the payer the District Health Board we can take advantage of that, if we acted inappropriately, and therefore we could leverage contract decisions and so on, around having detailed information about the inside workings of the general practice.

There are quite a number of concerns why GPs might think this is a problem, but of course a lot of that might be unspoken or even not fully formed and the privacy thing becomes overrated or a rule of regulation it’s easy to reference to, so the privacy thing is important.

When it comes to the privacy you’ve got to think, what does the patient want first? Well, they want their information to be secure, absolutely, so we met with groups of people from the community and it was very interesting because it was almost universal when we met with a group of consumers; sometime into the conversation they would say ‘oh, we thought you were doing this already’ and we realised there was a bit of almost ‘not quite a deception’, but it felt like a deception, for many of them believed we were already sharing the information and that was quite chastening for us to realise that we had mislead people, not deliberately, but created this illusion and it makes you think. Then we talked about how could we put a control system around it? Is your privacy basis going to be ‘opt in’ or ‘opt out’? We have a shared record system for the community so do we get people to sign up and say I want to be part of it or do we try and make it as widely known as possible and allow people to sign off it? 

After a lot of discussion, including advice from a number of experts, including our privacy commissioner discussions, we really felt the ‘opt out’ was going to be the one that was going to be successful. It does carry the risk of somebody being included in the process who didn’t know about it and feels uncomfortable about it. On the other hand if we build the other protections around that information we allow other ways of protecting it so certain elements from general practise can be controlled, hence the GP can choose things that are private, they can manage the patients’ information to a granular level and if we set the standards around who has access to what, and we set a matrix of every individual data item against the professions, then we can get some controls around the system, we can then feel comfortable. Using a base line permission system, if one of our health professionals accesses information from outside their own domain, so general practitioner access information from the hospital, the hospital will then access the information from the pharmacy, we will, as standard, ask the patient.

But if you don’t ask the patient, you could acknowledge that on the little box that comes up and you can put a reason in and this is subject to audit and the series of protection levels we built in. And the bottom line is these are professionals so the penalty for misusing information would be as much as losing your whole career, so the high stakes involved in it will get people used to the idea to that there is a high trust element in the whole process, but then there always has been in health care.

Do you think UK could follow your model shall we say?

I think so. I wouldn’t profess to be an expert in the details of the quite complex legal and regulatory systems you have around data management control, but I would think the principles are going to be the same. It’s about ethical processes and following some regularity principles. If you sit down and think it through and look out to protect people, the people you want to protect are those people who are the vulnerable and the disempowered.

If you’re sharing information about people and it’s sensitive and might have an impact you’ve got to think about that and work your way through it, but the challenges you will encounter will include some people who feel, as a matter of principle, that their information should not be shared and one wants to give them the easiest possible pathway to choose not to be part of it, so our system has an opt off system where you could talk to your GP or someone else that you encounter and they can help you organise that or simply dial a phone number and start the process to have your information blocked in the system.

Do you think information and information technology is becoming more essential for the underpinning of the health industry?

I think it’s one of the key shifts that are going to happen. In my career, I have seen a number of things happen around information. Probably the first big one was when someone walked into the consulting room with a Google print out of their illness and I realised that as they read it they would know more about their illness than I knew. You can’t read everything so they can have the latest information. Now that’s a bit of a game change really because medicine used to have this whole element of secret knowledge that we would know things about illnesses that patients don’t know. The medical professional has got to move from secret knowledge to being advisers and supporters and interpreters of information rather than just being pure suppliers.

The next level is, of course, we have information of patients’ records they don’t know about, that’s a new challenge for us. How are we going to open that up in a way that’s not going to create problems? But if you look at a complex health environment, it’s no longer about a doctor or a nurse with a patient having an encounter or a continuous care programme going, it is now whole team, the team is distributed.

From the patients’ perspective they move through various domains of care, whether it is their GP, the community agency, the hospital and the information has to be with them. So they’ve either got to carry the folder under their arm, which is one way to do it, but if the folder is under their arm nobody else can look at it, until they give it to them, so you can’t do anything without the patient being present, whereas the information system gives you the option to have the equivalent of the folder under the patient’s arm, but it’s available at a distance from the patient and in multiple places. I think it is a transformational tool; it disrupts what we normally think, even the simplest thing.

One of our analogies we’ve used in presentations is of a GP and I where we would do a sort of simulated consultation for a patient, so rather than sending a referral to me the general practitioner, he would ring me up and we would both look at the same patient record on the same screen at the same time and you can resolve a problem and come up with a management plan in a few minutes, whereas in a traditional model it would involve a week or two or three or four or five weeks of sending things back and forward and doing tests to come to the same conclusion.

Then when you’ve got the information, if it’s real electronic information, I am quite cautious about scanned documents, I don’t think they’re going to give us what we need, except the absence of paper, but you need paper to generate them so you still have paper, and if we have real information then, of course, you’ve got the ability to anonymise it, accumulate it, analyse it and understand much better what’s happening in the health system.

What do you think will be the main issues and challenges for the health informatics sector going forward?

I think one of the big, interesting conundrums is the patient access thing. I start from a process of ‘what would be in the patient’s records I don’t want them to see and how do we manage the process.’ I have always had the good fortune of working in the hospital side where it’s always been read by someone else.

Alistair Brewers in Newcastle once said to me when I was training: ‘always write in the patient’s record for them to read.’ So for me it’s a connation, I don’t see it as challenging, but for you that’s not been the norm, I don’t think. So how are we going to find a way forward for the patients to access their records in a way that’s not going to cause them distress or cause a dissonance in the system? If we can achieve that I think the patients are going to be empowered to become partners in their care and a lot is said for transferring responsibility to patients for their health; maybe think about transferring control. The privacy debate becomes a bit meaningless because then the patient could then, for example, see every single person who accesses the record just from the archive log and then the patient could control and say who has access and who hasn’t.

They need some understanding of the system to realise that a clerical member of staff might get access to the record to book an appointment, might need to know some information, but those things could be explained and understood. We underestimate the ability of patients to take a part in their care. You’ve only got to look at something like renal dialysis patients who do renal dialysis at home, they work the machines. So if they can do that, taking part in and managing your health record is not so challenging.

Watch the video