DR MARK WALPORT, PROFESSOR CAROL DEZATEUX AND PROFESSOR SIMON WESSELY

7 JUNE 2007

  Q333 Chairman: Good morning and welcome to the third evidence session of our inquiry into electronic patient records. I wonder if I could ask you, for the record, to give us your name and the position you hold.

  Professor Dezateux: I am Carol Dezateux. I am paediatrician by training. I am a professor of paediatric epidemiology at the Institute of Child Health and an honorary consultant paediatrician at Great Ormond Street NHS Trust.

  Dr Walport: I am Mark Walport. I am the Director of the Wellcome Trust. In a previous life, I was a professor of medicine at Hammersmith Hospital, at Imperial College. I am a member of the board of the UK Clinical Research Collaboration and of the Interim Board of the Office of Strategic Coordination of Health Research (OSCHR).

  Professor Wessely: I am Simon Wessely. I am a professor of psychological medicine at the Institute of Psychiatry at King's College London. I am also here for the Academy of Medical Sciences.

  Q334  Thank you very much and, again, welcome. In broad terms, what benefits can the Secondary Uses Service bring to research in the United Kingdom and what potential do you see in the electronic patient record once it is more widely used?

  Dr Walport: This is all about improving patient care. Good patient care depends on having good medical records. As an example, Choose and Book is something that has been introduced as an early phase of the National Programme for IT but how can you Choose and Book if you do not know what the benefits of seeing different individuals are? That requires good data. Frankly, having good healthcare depends on doing research all the time. It is a question of raising the game. I think research is completely integral to good patient care and electronic patient records are a very powerful way of achieving that.

  Professor Wessely: We would agree completely. It is an ability to use the enormous data resource of the NHS for the public good in a way that has not been possible before. It can answer questions swiftly and accurately on patient safety. Issues such as MMR and autism could have been dealt with very swiftly and properly. The whole range of opportunities that it provides, we welcome. We feel it will amazingly improve the effectiveness of the NHS and patient safety.

  Professor Dezateux: I would like to add to that. Thank you for this opportunity to raise the profile of research using electronic health records. I was a doctor trained in the NHS. Nye Bevan established a health service free at the point of use but the other very important person from my generation was Archie Cochrane who emphasised the need for effective healthcare free at the point of use. England is renowned for initiating the international Cochrane Collaboration which pulls together evidence in an accessible form to inform effective healthcare. For my generation there is no distinction to be made between health services and research for effective health services. Coming as I do from a public health background it is really important that we understand how to protect the health of the public when we are dealing with a much more complex environment where public health concerns require rapid answers. Without an infrastructure built on electronic health records we cannot discharge that public duty effectively. I think it is very important that we get over, using some detailed examples, the important questions that we will not be able to address if we do not have this kind of opportunity in the UK.

  Q335  Chairman: A number of witnesses described access to the national electronic record data as a "unique selling point" for research UK. What is your interpretation of a "unique selling point" and do you agree with that?

  Professor Dezateux: There is indeed a unique opportunity in the UK, based on the fact that we have a National Health Service, we have a diverse, large population, and we have the opportunities through our public services and our public infrastructures to understand and identify reliably effective treatments and also reliably to identify causes of disease. While there are other countries that have developed their electronic record linkage, notably the Nordic countries and Western Australia and some parts of Canada, in general we have a very long and distinguished tradition of using electronic records in the UK and we rather take for granted, in fact, the ability to link to vital events, the ability to link to cancer registers, but we have to remember that some of the most important health discoveries—the link between smoking and lung cancer, et cetera—were all facilitated by these opportunities.

  Dr Walport: The greatest advances in health have come from public health measures, going back to the discovery by John Snow in the 19th century that which water pump you took your water from influenced whether you got cholera or not. The opportunity in England to have potentially 50 million health records with good record linkage offers enormously important opportunities for improving patient health. Let me take another specific example. Drug trials, before a drug is released, offer the opportunities to find out the potential benefits and side effects in populations of a small number of thousands of individuals, but the truth is that drugs that are effective also have side effects and if you want to find the rare side effects then it is important that you can find them when you are looking at large populations. So, again, the opportunity to have good record linkage for a population of 50 million means that you can potentially find rare and uncommon side effects of drugs quickly. For example, the association between some of the new non-steroidal anti-inflammatory drugs such as Vioxx and cardiovascular disease would, I think, have been discovered much more quickly if there had been very good electronic patient records and very good epidemiology surveillance of that. This is all about good health actually.

  Professor Wessely: Researchers from my institution routinely go to other nations with systems, namely the Nordic countries, to do research that we cannot do in the UK. They are just in the middle of looking at a study on maternal health and schizophrenia in children that can only be done in Denmark and Finland because that is the only way you can assess related events during pregnancy to the later outcome of schizophrenia in the child. I think most people would agree that is incredibly important. We cannot do that in the UK. We have someone who has had to be taught Finnish to go and do these studies in Helsinki. Obviously the data here would be much better, because it is much bigger and more comprehensive.

  Professor Dezateux: There has been a huge investment in the human genome project and this country has terrific resource in terms of some large-scale cohort studies linked to biological data. To exploit that and the knowledge and understanding that will lead from that, we really need to capitalise on these opportunities for record linkage to understand the meaning of the genetic changes for future health and to advise the health services wisely about how to use these tests in the future. I think the UK is very uniquely positioned to do that. Already there have been some very important advances building on that and we would like to see that opportunity developed more because we have the opportunity, both in the pharmaceutical industry and in terms of the discovery of causes of disease, to be a world-class leader in the UK in that way. This is really what this will allow us to do.

  Dr Walport: Indeed, it is timely, because the front page of The Independent today and the lead story on the News at Ten last night was a report of an enormous study of genetics of common disease demonstrating new genes for Crohn's disease and diabetes, and in the last few weeks there have been new common genes associated with obesity. These have all benefited from very good cohort studies carried out in the UK.

  Q336  Chairman: The initial aim of setting up the Secondary Uses Services was really a way of performance management and payment by results. Do you think Connecting for Health was very slow to recognise the potential for research?

  Dr Walport: I think it is getting there. The very term "Secondary Uses" is unfortunate because it seems to me that one of the primary uses of Connecting for Health should, for example, be providing expert systems to help doctors and other health practitioners in diagnosis and treatment. I think this should be a primary use. It comes back to the point I made earlier. How is it important, as it were, to be able to choose between five different clinics and five different doctors if you do not know anything other than waiting times?

  Professor Wessely: I quite agree. In the1948 NHS Act research was defined as a primary function of the NHS and I think it is unfortunate they have chosen the term "Secondary Uses Service". However, we do have some concerns on the technical side. To do these kinds of studies takes an enormous amount of expertise, time, and trouble—to do the validation, to do the linkages, to make sure they are accurate and so on. We have some concerns that the Secondary Uses Service, so called, has that ability or expertise or commitment at the moment. That is something that needs to be looked at carefully.

  Professor Dezateux: I think Connecting for Health have been very focused on delivering performance management and business and have been instructed to do so. However, our experience over the simulation exercise that was carried out for the UK Clinical Research Collaboration was a very good co-working in openness with those involved in the Secondary Uses Service, with good receptivity really to the opportunities that this provides, so I hope we can move forward positively, building on this dialogue rather than perhaps perseverate about the past.

  Dr Walport: I think that is exactly right. I think the direction of travel is right; there now is a recognition in Connecting for Health that research does matter. It is very instructive to look at other countries and other systems. If one looks north of the border in Scotland they have extremely good electronic patient records, particularly in areas such as diabetes. We recently held a meeting at the Wellcome Trust on behalf of the UK Clinical Research Collaboration and had presentations from two of the big American systems; from the US Veterans Administration, with many millions of patient care records, and from Partners Healthcare System in Boston, where, again, they have more than two million records. It is quite instructive that in all of those environments they are using them very effectively. I think this will get much better acceptance when the medical profession at large can see it is enabling the profession to do its work better, and that is why, in a sense, I think focusing on the management side meant there was much less ownership of the system by the people who had most to gain from it.

  Q337  Chairman: Carol, you recently took part in a study to simulate the impact of electronic record data on different kinds of research. What were the main findings of the study?

  Professor Dezateux: There were four groups, of which I led one, each of which looked in different ways at different types of study designs, ranging from clinical trials to issues about patient safety to issues about causes of disease. We led our groups in parallel but with a lot of cross-working. The first thing to say is that we were all convinced that there were fantastic strengths in the UK in record linkage and that the opportunities with a largely electronic general practice now, with increasing use of digital information and electronic information in hospitals, needed to be built on, but we looked internationally at leading examples and we felt that political leadership was critical to this. I think we all signed up to that. We felt there was strong consensus across the four areas. The first is identifying patients reliably across these different data sources. We know from the Nordic countries and from Scotland that having a single patient identifier is the most critical factor and mandating its use in all health service encounters reaps enormous benefits. That is the key to the successes in the other countries and where Scotland is ahead of England in this respect. We need to link in all NHS encounters, so the Secondary Uses Services needs to engage proactively the primary care pathology services and, indeed, the private sector. Some of my own interests in assisted reproduction—which is a topical issue at the moment for Parliament—show that we have had, since 1990, a database of more than 1.5 million cycles of treatment. Currently we can link that to nothing. We have learned very little about the long-term effects for mothers and children of assisted reproduction in this country yet there are major public safety issues related to that. We must have this ability to link. We should not have data sets that are not patient-level and that are not linkable, because we cannot answer the important questions that society wants us to address. Following from that, we were all agreed that we needed person-level data. I can give you many detailed examples that show why that is the case and I think this is probably an important point to get across. We want the legacy data—which means we want the data that goes back many years—because one of the important things is the ability to answer questions rapidly and not have to wait to accrue lots of person-years of observation to answer these questions. We need the full demographic record, particularly if we are interested in, for example, environmental exposures relevant to chronic disease, to asthma, to obesity—these really important public health problems that we are going to address. For some opportunities we need real-time access to the data. Clinical trials are an important way to provide reliable evidence about effective treatments. One of the obstacles to delivering those trials in a timely way is that recruitment is low or there are not eligible populations, so we need to be able to look into these data sets and say: "How many people are there out there with these kinds of problems who would be potentially willing to receive an invitation to take part in research?" At the moment it is very, very difficult even to send that invitation to anyone, so we need a "can do" culture shift that allows us to do that but which does not compromise the governance and confidentiality that we all obviously respect. Those points were very common. For the kind of public protection, health protection issues, we cannot invent these infrastructures as the questions arise. This is why we need leadership. We cannot think that somebody is going to come up with an idea about MMR and autism, or somebody is going to think about fire retardants and cot death, or somebody is going to think about pesticides and mal-development of the eye in children. I can produce endless examples of public health concerns where, in order to answer things rapidly, we need an infrastructure such as there is in the Nordic countries where we can look at reliable data and say, "Is there any evidence to support this concern?" If there is: "What kind of more detailed studies should we commission to drill down into this?" We cannot take that first step very easily in this country. I chaired the MRC group looking at the epidemiology of causes in autism. The fastest study came from Denmark. Using data for some two million person-years, they were able to look at children who did and did not have MMR, link them to autism registers, birth registers, and conduct and report a fantastic study in The Lancet. It took us four years to do that in the UK. That is the kind of advantage. Very finally: it is communication, communication, communication. That is the other big conclusion from the simulations. We need to get out there and work with Connecting for Health, to get over these narratives that make it clear to health professionals and the public what the benefits are and what the implications are of opting out or withdrawing from this collective good, if you like, the ability to answer important questions conclusively.

  Q338  Chairman: What did your particular study find in terms of the accuracy of data and also the issue of the standardised form?

  Professor Dezateux: The first thing to say, as a paediatrician, is that one of the outstanding successes has been the introduction of the NHS numbers for babies in 2002 by Connecting for Health. This has enormous potential for us because we have a unique identifier early on. We would like to be able to link mothers to their babies and we would like to be able to link siblings to their siblings because some of the questions that we have are around families. There is not a mandated system for doing that but it is not a technically challenging or difficult to do, given the right leadership and the right go-ahead. There are many examples where there are certain elements of data that are either not routinely collected or where the quality could be improved. I think we need to understand that this is a process. By linking research to clinical practice and embedding this in clinical communities, which is what Scotland has done, there is this constant interchange that research looks and helps us to understand the quality of the information and provides solutions about improving that quality. If the quality serves both clinical care and research, it is a win-win situation. There are interoperable standards, there are technical solutions to this coming out. I do not think those are major obstacles; I think it is the leadership and the political will to make these things happen that is the most crucial thing from our perspective.

  Dr Walport: Could I pick up five points very briefly from Carol's answer. The first is the unique identifier. It is absolutely critical, but it is absolutely critical for clinical care as well. If your name is, let us say, Dr Richard Taylor, there is probably more than one Richard Taylor in the UK.

  Q339  Sandra Gidley: Has there not always been a number? I have had an NHS number for years.

  Dr Walport: But it is not regularly used.