Select Committee on Science and Technology Minutes of Evidence


Examination of Witnesses (Questions 620-639)

THURSDAY 13 MARCH 2003

SIR JOHN PATTISON, DR MIKE CATCHPOLE, DR JAMES PATON AND MR GRAHAM SPITTLE

Lord Turnberg

  620. I should declare my interest as an ex-Chairman of the Public Health Laboratory Service, having retired from that post last July. The funding for research in the PHLS, and now the new HPA, was a very important area, even though the proportion of total funding for PHLS, and I presume for HPA, was relatively small from central funds, although larger from external funds. In the change that is described, there are two sorts of fears that were expressed in PHLS, and I presume HPA may well have expressed them to you as well. One is that the funding that was used for research within the PHLS, using peer-reviewed schemes, might be relatively neglected in the general pool of funding for the NHS R&D. That was a fear in the background. I am sure, with you in charge, it is unlikely to happen in that way, but one can never tell for the future. The second fear was that much of the research was done on many—very many—small technological advances in techniques, in how to do tests better—try this, try that, which is better? Those sorts of research areas are unlikely to get high recognition amongst the major areas that you describe, which are clearly very important. The question is, how does an organisation like the HPA do that very important work?
  (Sir John Pattison) I think one can be reassuring about both of those, at least to start with, because the reassurance is based on the assumption that when we look at the portfolio of work and the forward strategy of the new organisation, which will be derived from the previous organisations, we will see there an appropriate portfolio of high-quality work. That is our a priori assumption. If and when that is revealed not to be the case, of course, then one would wish to make some changes to it, and certainly there is no intention to start off by putting all the albeit limited resource into a general pool and asking the HPA to fight for it. My personal view of this—and I stress that simply because this has not yet been formalised into any agreement between the Department and a body which does not come into existence until 1 April—is that there will be an alignment between the research that is going on in the bodies that have come together to form the HPA, and that the likelihood is that they will be able in any competitive system to justify that funding, and indeed, they may well be able to justify more funding than that, in which case what is threatened, if you like, are the other areas of research that are currently being funded from the Policy Research Programme and NHS R&D. I suppose I should also declare the most fundamental principle of the Department of Health's R&D strategy, which is that it will go out and secure the best quality research that it can, wherever that may be, in order to address the questions which it feels to be important. So in that sense, it is exposing the R&D of the former organisations to more external scrutiny than has been the case in the past, but once we have done that in the first instance, it is my view that we will do it on the quinquennial basis. We will also have to write it into the contract, which we have considerable experience of in terms of running the 13 units that we have in the Policy Research Programme, and that is to have some long-term research going on at the same time as having call-off contracts whenever urgent things come up and we require a rapid piece of work, and I think that model would suit the HPA very well indeed.

Chairman

  621. Perhaps we might go to question number one now, and that is, to what extent do you believe that the medical community is aware of the unrealised potential of information technology for the management of infectious disease?
  (Dr Paton) Before getting straight into the answer, this is actually a very exciting time for IT in the NHS potentially. It is certainly dramatic, and there is, I think, a tremendous opportunity to lay some markers down as to what we are looking for in IT in the NHS, including surveillance, and it will be key to make sure that certain things are not forgotten as the planning gets under way. There is a tremendous central drive, much of it due to Sir John in setting up this national programme for IT for the NHS. If it delivers its potential, it will, I think, transform the infrastructure and the equipment that is available on the ground in hospitals and in general practices. That is the plan. What is absolutely essential is that some of the clever things in the thinking and the surveillance and the national aspects have to be brought along in parallel to that. The medical community, I think, is very ready for much greater electronic involvement than there has been in the past. I think many have really been demanding for a long time the resources to move things on. In a lot of hospitals and primary care trusts there are champions, if you like, of clinical IT, who have been very keen to make things happen, but have been frustrated by what has seemed very difficult to make this happen over the past years, for various reasons. The IT experience varies enormously across the country, and even between parts of the profession, so in managing infectious diseases, the schemes that involve microbiology laboratories, NHS laboratories anyway, reporting on various pathogens to regions and then on to the Communicable Disease Surveillance Centre, are maturing very well. The software available is now very much better than it was, and those in labs and microbiologists are very IT-literate and aware, and that seems to be moving on very well. When you move away from the very discrete laboratory side of things to, for example, notifications in the community, I think that is a very different story, and although, again, there are electronic routes and electronic modules available to do that reporting, the quality of the data and the amount of reporting that is done is not great. There is a lot that is missed. There are some districts that do not report at all and so on. So it is a very different story there when you get to notification. When you move on to looking at clinical cases of disease and symptoms, patients presenting to A&E with pneumonia, for example, as soon as you get into that realm, we have almost nothing; we have almost nothing in terms of the data being collected within hospitals for example, electronically by and large, certainly not in a form that is consistent, that uses a terminology that means the same thing from one hospital to another, or from one hospital to general practice. There are great issues there, and there is certainly no electronic system of surveillance that works automatically, and perhaps I will have the opportunity to talk through some ways that we might address that, but across the board, it is a very patchy story, and the more you get into real clinical diseases and symptoms, the electronic route of surveillance is very weak indeed.

Lord Oxburgh

  622. Could I ask a question on that? I must declare an interest as a former member of the Hammersmith Hospitals Trust and an occasional adviser to Fujitsu Laboratories (Europe). You said there was great enthusiasm and a significant degree of literacy, and that is great, and that is a start. The procedures by which we manage information in hospitals today were largely laid down at the time of paper and pencil, and of the typewriter and the telephone. There can be an enormous danger of simply saying we will take the same system and implement it using software and hardware, when really what one should do is say, "What are we trying to do? What are the end points that we are trying to achieve?" and see whether one might actually want to do things totally differently. Would you care to comment on that and the extent to which you see that being realised?
  (Dr Paton) That is certainly generally true, that when an electronic system goes into place, you need to take a look at the process and say, "Actually, do we need to continue with what we have done?" What needs to happen, I think, is that up front, the clinician who is recording information about a patient who is in front of them has to feel it makes sense at that point to record that information, that it is made as easy as possible, with screens that are guided and menu-driven and there are pick-lists and so on that force the clinician to go down certain lines, to record certain items of data that they might not have thought to record, and that takes a great deal of intelligent design behind the scenes to make it look as natural as possible for the clinician to do that. What falls out the other end, lo and behold, is automatic surveillance information that is transmitted centrally which includes a core data set for infectious diseases, but you start at the back end, and say, "That is what we want to get," and as you say, we want to see that information centrally. That is really the minimum that we have to get hold of, but then you have to work forward to the doctor or nurse in front of the patient and have the screen so designed that they will answer those questions accidentally rather than it being an add-on. Some of the experience with the coronary heart disease data sets, which are very large, 50 or more items to be collected on a patient with chest pain, is that when these are added side by side to what a nurse in A&E is already collecting on a screen, they are not overjoyed, to put it mildly.
  (Mr Spittle) If I could respond to the point you made about the danger of putting into silicon or software old practices that are no longer relevant or perhaps best practice, it is a very valid point, and it is one that I think is not just related to the medical environment, but is one that we see commercially in all computer systems, and it is one that when you go into any enterprise what people traditionally want to do is make things communicate that do not currently communicate. What they then want to do is to liberate data that is marooned in old applications, the applications are now out of date but the data is relevant, you want to take that data and re-combine it in a way that makes sense, and it is only when you have got to that end point—and there are stages of discovery—that you actually have to model what it is that you want to get to, what it is you are trying to do. Do not try and automate exactly what you are doing today. So I think you are absolutely spot on the point. There is a grave danger of going down the route of purely automating what we are doing. We have to determine the desired state. If I can come back to the original question, question one, I would like to echo my colleague's comments. I think this is an amazingly optimistic time for IT. There is a tremendous amount of work that has gone on. We are at a point where IT I think can make a tremendous difference to everybody's lives in a very beneficial way. I personally am very optimistic, and I know the industry is very optimistic about what we can do, particularly in this area of what we call bio-informatics or life sciences. On the one hand, the good news is that we have a lot of experience about providing IT to make systems better; on the other, the bad news is we do not have as much expertise in the medical field as we have in some of the commercial fields. The third piece of good news is I do not actually see them as IT-dissimilar projects. I think a lot of the techniques and a lot of the technology we have today can deliver immediate and real benefit to almost any project that we would like to undertake. The key, however, is to think about the projects in totality and to think about things such as standards and to think about things such as architectures. One of the things I would observe about medical projects is, because of the disparate nature—they are individual islands of excellence or individual islands of automation—I do not see the level of integration, whether at local, regional or national level, that is typical in commercial systems. If that were to change, I think there are amazing benefits that could be had. There is no doubt from the point of view of analysis, the larger the data set, the better the opportunity of getting really good answers out of that information, and I believe that one of the key things that we could do in this country is to think about the infrastructure that we would need to put in place, not just to serve one or two applications, but that actually are of great generic value, which many applications could actually tap into. If we get this wrong, there is a danger that we will have a multiplicity of independent databases, which none of us can query except the person who owns the database. We would not be able to link across borders with other countries, and particularly if we are talking about epidemiology and infectious diseases, that is complete nonsense. So we have a tremendous opportunity that does not require IT innovation. What we have to have is a little bit of innovation in our thinking about how we design this, because the benefits are immense.

Baroness Walmsley

  623. Can I pursue you on the commercial applications and perhaps get some examples? Did you have in mind the sort of thing that supermarkets do with point of sale information, to enable them to order just in time and to find out the buying pattern of the purchaser?
  (Mr Spittle) No, it is not, but it is not a bad analogy, because I can see how that would be useful. I actually had in mind stock exchanges, because I have probably spent a lot of my time talking to stock exchanges and merchant banks, who put a lot of investment into IT. What you have here is a phenomenal flow of information that comes in to the traders, they have to have lots of information come in, they also have to push information out, so you have both push and pull of information. You have to have databases that talk to each other, and you have to run very high-speed analytical tools that can determine what is actually going on from the data that you have collected. In addition, you have to have the ability to send messages 24 hours a day, seven days a week, up to 10,000 or 20,000 messages a second, that absolutely must get there, because the information is particularly critical, in that case from a financial point of view, and obviously from a medical point of view we are talking about equally critical information, but they have exactly the same needs from an IT point of view that I think I could envisage in a medical environment.

Chairman

  624. From where does the Department of Health get its main advice about information technology? Is it from IBM, for example?
  (Mr Spittle) I am absolutely the wrong person to ask. This is the first time anybody in government has asked me, in my development role, about IT for medical. The only thing I would say is that I do believe that there is a lot that the IT industry could do to make our value proposition, i.e. to tell you what we can do; I think we are probably remiss in not showing you what we can do in an abstract, technological way that is understandable.
  (Sir John Pattison) The current programme to develop IT in the NHS stems from a seminar in Number 10 in February 2002, at which two of the major suppliers had representatives present to give us technical advice and specialist advice. Subsequent to that, because it all had to be linked to resources that are only going to become available as part of SR 2002, from April 1 of this year we have talked extensively to the supplier community, and there is a large range of such people, both in terms of where their corporate headquarters are and also the size of the companies involved and their previous experience in the health sector. It is slightly difficult for us at the moment because we have this year published notices in the Official Journal of the European Community to procure consortia who will act as the engines for the installation of IT infrastructure and certain applications in large parts of the NHS. So we are somewhat limited now in what we should say in public whilst that procurement process goes along. We cannot really identify an individual company and take advice from that because we have to run a fair and open procurement process.
  (Dr Catchpole) In respect to question one about awareness of unrealised potential, I think awareness relates to horizons, and picking up Lord Oxburgh's point, I think there may well be a case for re-engineering some of the processes of clinical care, but more importantly, it is a conceptual leap that needs to be made, the idea that information becomes locationally independent. That is the key to unlocking a lot of what we are talking about. If you are using a palm-top at the patient's bedside, the information that you put in there could be available not only to the nurses at the nursing station, to the people ordering the patient's lunch in the kitchens, but also to the local or regional surveillance experts at the same point in time. It is the idea that we do not have to have silos or bunkers of information, and the only way you can get information from here to here is to transfer it from this bit of paper to another bit of paper that goes somewhere else. Having said that, there is a tension, a major tension, between the benefits that come from sharing and greater access to data and the real concerns that exist about protecting confidentiality. The work of the Patient Information Advisory Group—and if I may declare an interest, I sit on that Group—has been very helpful in terms of laying down the statutory instruments that have allowed us to secure information flows for surveillance in that respect, but we have to recognise that tension. It is about freeing up the idea that information is fixed in a particular location.

Lord Oxburgh

  625. Could I follow up a little bit on that? If I could start with Sir John, when would you expect the contracts to be placed for media procurement?
  (Sir John Pattison) Later this calendar year.

  626. With a window for implementation?
  (Sir John Pattison) We have certain targets which are in 2004 and 2005, and of course, there is very much more to do after that as well, but those are the first targets.

  627. My point is that this contract is going to be one of the largest IT contracts placed by any government department anywhere, and it is going to be some time before anything of this kind is done again. One really has to look forward as far as one can, a good 10 or 15 years, at the structure of the NHS, at the evolution of medicine, the advent or not of tele-medicine, for example. Do these thoughts come in?
  (Sir John Pattison) Indeed, and I think one of the four main components of this phase of the national IT programme for the NHS, which lasts until December 2005, is infrastructure, and to put in a broadband network for the NHS and to procure that nationally. There is, of course, a discussion going on at the moment about what is the extent of that network. Currently, the NHS Net is the backbone wide area network, that is the door to door network. Inside the door of the surgery or hospital is a local area network which is left to local procurement. There are two schools of thought about what we should do about that, and that remains an unresolved question at the moment. The NHS Net's current contract comes to an end in 2004, when our aim is to procure a broadband network in a way that is coherent with the general government push towards broadband. That is one area in which we have considerable expertise, at least on a wide area net, the national network, and one in which I am confident that we can secure that up-date. In the mean time, it is quite clear that for heavy users of IT, in trusts or in primary care, the capacity of the current network is a seriously limiting factor, and so we have an interim upgrade over the next 18 months which is going to give 256K to every GP surgery and 2 Megabits to every trust, ahead of going to that new network, which is called N3, imaginatively.

Baroness Walmsley

  628. How are we being protected against the sort of debacle that happened in the Passport Office, when they instituted a new computer system? The consequences of this going wrong are a lot more serious than a queue of people shouting for their passports.
  (Sir John Pattison) These are very risky projects, and the history of them, whether in the public sector, which we always hear about, or the private sector, which we do not always hear about, is that the success rate tends to be low; in fact, the success rate in general terms is 33 per cent, I am told. So only one in three will succeed. What we have done on that is to secure for ourselves Richard Granger, who is the Director General of NHS IT. He comes from the private sector. He has experience of putting in large computer systems. We can look at the experience of the Passport Office as one experience; we can look at the experience of what Richard Granger installed for congestion charging in London as another experience, and say that we may well have somebody who is capable of delivering on time and on price something that works. However, if I may just make a personal comment, I cannot exaggerate the value of Richard Granger to this programme, and the likelihood of its success. These are skills and experience which we simply do not, or have not had up to now in the Department of Health and the NHS. We are good, and we have introduced somewhere in the NHS everything that we want to install, but we have never done it on a scale that is implied as necessary and correct in order to support the National Health Service. So he is bringing in people who we would not automatically have brought in and did not know about, and I think that is increasing the likelihood of the success of this enormous project.

  629. I am aware that this question has been partly answered, I think by Dr Paton in answer to Lord Oxburgh, but I am sure there is a lot more to say. We have had a lot of evidence that suggests that improved use of IT is essential if we are to have a better surveillance system, and certainly Dr Paton has outlined in an earlier answer some of the shortcomings of it, particularly in hospitals. What do you consider to be the key features of an effective IT system to support surveillance, and how important is the data coming from the sentinel practices?
  (Dr Catchpole) Picking up on something that Dr Paton mentioned, I think systems should be integrated as far as possible into clinical operational systems. It should not be that we do surveillance as well as looking after patients, but that in looking after patients we are undertaking surveillance. It is perhaps easy to say that, but apart from avoiding unnecessary duplication of effort, the other point is that we are likely to get greater accuracy and speed of accessibility to surveillance data if we do that; if we do not have to wait for data to be translated from one system to another. So it benefits not only the people providing the data, but it also benefits those who receive the data in terms of speed and accuracy of information. They should be easy to use and easy to support, and as colleagues have said, new technologies greatly increase the chances of that. The idea of using new technology so that we do not have to support 200, 300, 400 systems in 400 trusts, but we can use browser technologies and centralised data warehouses to manage data, offers great opportunities for the future. So I think it is about making best use of new technologies. We need to have robust systems that are secure, so they do not stop working at the moment we are faced with a major outbreak or deliberate release episode, for example. They need to be robust; they need to be secure, and they need to be integrated into operational systems. Those are characteristics of the systems themselves. There are an awful lot of other important issues surrounding that about the need for standards. I am impressed by the documentation coming out of the Centre for Disease Control in the USA. Their national electronic disease surveillance system, NEDSS, shows tremendous vision, but of course, it is entirely based on the fact that they will have a common coding system in SNOMED and a common record structure system in HL7, and we need those standards urgently if we are to be able to harness all the information that is available across the Health Service. I am sure my colleagues have many other points to make about what makes a good IT system for surveillance. On the sentinel side, I think sentinel systems have an important role in surveillance, although they cannot replace the need for broader based surveillance systems. We need to be clear about what we mean by sentinel practices, because they may either be a representative sample of all practices, which is typically what GP spotter practices are based around, and they provide us with a representative picture of what is going on, or they may be truly sentinel in terms of early warning, in which case they are purposefully selected to provide an early warning. In the context of monitoring HIV disease amongst gay men, London genito-urinary medicine clinics are a suitable sentinel network for picking up early warnings of problems with epidemiology. I think sentinel systems are particularly useful for allowing us to collect more data, better data, on specified diseases, but we need to make sure that in the spotter practice systems we can link together the sentinel data coming from the clinical side with matching laboratory data. So sentinel systems are important, they may serve a number of different purposes, but they do not replace broad-based surveillance. Sentinel systems will probably never be as sensitive at picking up outbreaks of less common infections, unusual sub-types of salmonella infection, picking up agents that may be the subject of deliberate release activities, unless your sentinel site just happens by chance to be in the right place, you run the danger of missing those events. So sentinel systems are important, and they certainly add qualitatively and quantitatively to surveillance activities.

Chairman

  630. One of the difficulties with sentinel systems is in fact that the qualitative assessment of much of this is based on symptomatology, is it not, and you can only go so far with that?
  (Dr Catchpole) Absolutely. They need not exclude laboratory data, of course, but it requires investment to link the two together. We talk about looking five to ten years; one of the issues that comes up when you look at that sort of timescale is the issue of near-patient testing, of course, and we have grave concerns within the Public Health Laboratory Service that near-patient testing may potentially undermine laboratory reporting, which is the cornerstone of surveillance in this country. But of course, near-patient testing may also offer an opportunity, particularly in the context of sentinel surveillance in general practice spotter practices, where you could marry together that clinical data with the information that comes from near-patient testing. We desperately need to be looking at technologies and systems that will allow us to capture that information.

  631. Mr Spittle, do you have any comments on that?
  (Mr Spittle) Yes. I echo the comment about the data capture at the end points. Whether that is at a GP surgery or whether it is a public laboratory, it has to be as automated as possible. You are going to have to define a template to some extent. I think you are going to have to be rigorous about the standards. The system designed must be designed around open standards for our own protection in terms of procurement and longevity and scaleability. The scale of the operation must be such that people appreciate the need to have hugely scaleable databases, which we do not have in the Health Service today. They are in use commercially, and people are building them, so this is not invention that is required, but this is the ability to collect vast amounts of data, whether it is text or whether it is digits or whether it is images—because increasingly images are going to be more important. You have to have the ability to be able to run the latest query tools against those databases that are available today. I come back to the point that getting the infrastructure right is so fundamental, because at the end points, the applications that interface with the infrastructure are going to have to be bespoke for the purpose for which you want them to perform, and they have to be as automatic as possible, and as simple as possible. But there is so much you can do with a common infrastructure that is going to give you immediate benefit. The worst thing that could happen is to build new, modern systems as islands of automation and not link them together. I would urge everybody to build a very robust infrastructure, that you can hang systems from and interchange data with. That is going to be the key here. We are going to have to mandate standards, and we are going to have to mandate lots of other things that you need to have successful IT systems. You made a comment earlier, Baroness, about the implementation: project management is the key. That is what makes the projects run or not. Very good, qualified project managers are worth their weight in gold, and that is also part of this.

Baroness Walmsley

  632. Can I just come back on this? Robust systems are all very well, but we also need robust training of all the people that are going to use it, because they are all human beings in the end. We have heard anecdotally that if doctors and nurses washed their hands between all patients, we would need 30 per cent more staff. How many more staff are we going to need in order to allow the time it is going to take for the health care professionals to undergo the appropriate training for these systems?
  (Sir John Pattison) I am not sure we have estimated the precise answer to your question, but the general point is very well taken. There are recent articles indicating that other health care systems, such as Kaiser Permanente, are investing six times the amount in training than they are in the IT systems themselves, and it will have to be of that sort of order if you take the true costs into account. We have published our strategy for the training of information staff, and I stress that in order to indicate that this is a range of different types of staff, all of whom are needed to have accurate, high-quality information entered into whatever system we install. It starts from the data entry clerks and goes right up to high-quality informaticians. We have put into our budget thinking, as it were, the need to provide resources to liberate the staff from their daily work in order for them to be trained, though in a capacity-constrained system that is quite difficult, because there are shortages of many of the key professional staff. Part of the key though is to provide something which many of those front-line professional staff really feel is helping them to do their work. It is not entirely age-related, but there is an age-related gradient, but we can, I think, bank on some familiarity with computer systems that we could not in the past, such as keyboard skills, and if we then provide something which is going to enhance their daily life, it is surprising how relatively easy it is to get that accepted. That is going to be one of the things that we have to pull out.
  (Mr Spittle) Without wishing to disagree with my colleague, because everything he said about training is absolutely true, if we do not implement some of the IT, we are going to have a critical skills shortage that we will not be able to manage in any other way. The project I am currently engaged in with Oxford University on e-Diamond for digital mammography addresses one of those concerns: we do not have enough skilled radiologists, and any particular radiologist does not see enough mammograms in his or her lifetime to be expert enough to diagnose this. The system we are putting in and trialling this year will give the very best diagnosis that it is humanly possible to do, so everybody who is linked into the system can get a much better diagnosis, and therefore a much better prognosis than we have ever done before. So longer term, this is going to help skills, this is going to actually reduce the number of really skilled people that we are going to need. You do have to skill people to use systems, but what you are doing is fundamentally altering the landscape here, if we get this right.

Lord Oxburgh

  633. My question follows on from what Baroness Walmsley asked and really relates to the human interface. You can have the fanciest system in the world, but if people do not feel that it helps them, or if it seems hostile, it is not going to work. I would take it for granted that as far as the people who input the data—who are going to be the physicians and nurses primarily—recording information in a way that is suitable for the system has to be no more onerous, and probably less onerous, than the system that they are working with at the moment. They have to see positive advantages in their day-to-day operation, and that it is going to make their jobs easier rather than more difficult. I take it that this is something which has to be a requirement.
  (Sir John Pattison) Indeed, and I think there are certain things happening which are all helping us with that. Firstly, technology. I will never be able to type as fast as I can write; it is just not possible; I cannot use a keyboard as fast as I can scribble on a piece of paper. But we have tended to concentrate our thinking in the past on desktop devices which have a keyboard in front of them. Technology is moving on. We do not want in the future to be confined to desktop, fixed devices. Portable devices which will take handwriting and convert it into neat typescript are a reality and are on the market, and are of crucial importance to community health staff. So that is helping us. The fact that the rest of the world in a way is encouraging us to put our tax returns in on line, to order from the supermarkets on line, etc, all helps us realise that actually, there are great benefits to filling in an electronic version of something over handwriting something, making a mistake and wondering what on earth you are going to be able to do about it. That is helpful. The other helpful side I think is that the developers are paying particular attention, it seems to me, to that human/machine interface, and it does seem to me that it is getting easier and easier for people to relate to these systems. Primary care, I would have to say, is further advanced than hospital care in terms of the investment that there has been in IT in the past and therefore the use that practitioners are making of it. It is steadily growing. All this is encouraging and helping us with the user/machine interface.

Baroness Finlay of Llanduff

  634. I have to declare an interest, first as a jobbing clinician, but also as Vice Dean of the University of Wales College of Medicine. I am going to sound like a dinosaur. The complexities of clinical practice are enormous, and based on pattern recognition, and there are huge numbers of things currently recorded which are very soft qualitative data, recorded in case notes, which are partly there as defensive practice. There are two aspects: you are recording it because you want to communicate it to somebody else, or you are recording it because you want evidence that you did it or thought it or decided it, which is very individual, soft data. The example Mr Spittle gave of mammography is a lovely example, where one thinks IT really must be the answer, because that is pattern recognition in a single format, always the same format, just as lab results coming out. But I have to say, the jobbing clinician's experience of IT is that the server is down, the system is down, or you just get the hang of it and the whole thing has got to be upgraded, and it is not compatible with your portable, and then you go, as I do, between three places of work and the upgrades are all slightly different, so what happens is that it does not work anywhere. Are we actually at a time with computer technology and systems that we should be spending a huge amount of money on transferring a system which, however imperfect, worked well for a lot of recording of data or should we simply be saying computers should be used for developing things such as the mammography that Mr Spittle outlined or the haemoglobin and the other results that are there on screen on the ward within half an hour from going through the auto analyser because all of that is linked? The bacteria that you have identified, all the viruses, the form is written anyway, the report is written and that can be collected centrally. I have a real concern that IT is being looked at as the all singing all dancing and that the clinicians are just completely drowning and it is far too early to say this is actually going to be robust for the next ten years because, as with anything that needs upgrading between one or two years, you will have lost the cohort of clinicians. I am already completely fed up with upgrades and completely brassed off with fire walls and completely brassed off with emails and poorly communicating systems that are always down when I want to get on them.
  (Dr Paton) I think you have explained very clearly why Sir John said this is a risky project. There is a great deal of money going to be thrown at this and there are some big problems with that. It is easy to buy computers, it is easy to buy broadband networks. It is hard to sort out clinicians' views on systems, cultural change is necessary. It is also hard to design the software even to do the simple things in front of you never mind linking it all together. That is not necessarily there or even written in many situations and that side of things is going to be very hard. We have talked about core sets and how we integrate them into the report, that is actually very complicated. It is right to put the infrastructure in first, but at the same time this other thinking has got to be done about how we address this. You are very welcome to come to my hospital any time and have a look at how people use it. We have a very broad operational system which looks old fashioned, it is 20 years old and the kind of screens that we are using look very old fashioned, but it works with just simple keyboard keys so that every patient care note made by a nurse or a physiotherapist or an infection control nurse or a dietician goes straight into the computer and as a microbiologist I can read absolutely everything that is going on with that patient and I can have all the results to hand, the blood pressures and the temperature and hopefully within the next year or two all the x-rays immediately available to me and all the drugs that they are being prescribed, it is all there right in front and it is being used comfortably and with confidence by clinicians.

  635. I work with a system like that, it is an integrated system.
  (Dr Paton) And it is down to the kind of software. The doctors do not write the notes on it, they are all on paper. Some of that can be structured and some of it is free text and I would be entirely comfortable with that. I do not think we can structure medicine or medical thinking. Everything has a specific term that you have to use for it, so you do not end up writing notes unless you have to. At least to get doctors to say this is what is wrong with my patient from an accepted list of terms, this is what I have just done on this patient from an accepted list of terms, surely that is not beyond something that can be delivered and once you have that you begin to be able to analyse it effectively.

  636. But we have lots of wards where there are not sinks by every bed and there are not individual patient toilets and individual patient bathrooms and I worry that huge amounts of money will be thrown at IT and yet we have the NHS functioning with crumbling buildings and the rudiments of infection control are not there.
  (Dr Paton) I am keen on sinks and computers, we need to buy both.

Chairman

  637. I am conscious of the time and the fact that we may have a Division which will destroy our quorum. I hope Lord Turnberg will excuse me if we go to question four for Sir John about the overall vision for an effective IT strategy for infectious disease within the wider national or international context.
  (Sir John Pattison) I think in some respects this question has been partially answered already. It is our vision that we are installing something for a variety of general elements of which the specialist element requirements will be extracted rather than developing and adding on a whole series of separate specialist requirements. So our vision would be that we would install electronic records in primary care and secondary care and that we would be able to extract from those records—which would be very new in primary care but absolutely essential since 86 per cent of the interactions take place in primary care and a large percentage of those are due to infection—the information that we need in order to populate an integrated care record. Our vision is one of a series of events happening to people during their lives in relation to health services and the very rich data of those individual events will need to be captured and remain in the GP's surgery, if that is what the event is, or the hospital if somebody has been admitted for an operation, if that is what the event is. The vast majority of that data can remain there, it does not need to populate an integrated care record. For example, the recordings of the blood pressure in somebody who has had a period in intensive care are very important during that episode; by and large they do not need to be transmitted to anybody else as the patient recovers, moves to the ward and moves out back to the community. Certain elements of that do need to be placed in the integrated record which will be, in old speak, the electronic health record, that is the cradle to grave concept for an individual citizen or patient. Much of the information for the purposes of infectious disease control can come out of those systems, we can link them to the laboratory systems and at the moment we are towards the end of installing the pathology messaging function between pathology laboratories, haematology, biochemistry and microbiology and primary care. I am aware, since Dr Catchpole told me, that there was some problem about the PHLS laboratories at one time because they were regarded as being outwith the NHS, but I hope we have now solved that and we would be able to link electronically the information that is coming out of the laboratories and going to whoever requested the results with the clinical data and analysed in a way that Mr Spittle was mentioning earlier on. So that is our vision.
  (Dr Catchpole) I welcome that enormously. I think in building on that we need to make sure that we define which are the right data items that need to go into the integrated care record. I was struck by the recent consultation exercise where the data sets were defined for national service frameworks, for diabetes, coronary heart disease and the others and I feel that probably what we need is to do a similar exercise for infectious disease, to define the dataset in the same way as they had done in the other national service frameworks but for infectious disease, perhaps health protection more broadly. I also was struck by recent consultations at the National Patient Record Analysis Service. I suspect that organisations like the Public Health Laboratory Service were not foremost in the minds of anybody when it was thought who should be consulted in that process, but we could gather a lot of useful surveillance data from those sort of analysis processes. I think it is about raising consciousness of infectious disease and surveillance when we approach these activities. The point that probably needs to be made is that surveillance benefits individual clinical care not just public health population-based practices. It is useful to know what the local antimicrobial resistance pattern is when you are prescribing. For example, we know there are regional variations in resistance to gonorrhoea. Similarly, when you are having to treat someone with severe pneumonia it would be useful to know if there had been a local increase in micro plasma, it may make a great difference. You would take a very different approach to treatment and prescribing than if you did not know that. So surveillance benefits direct patient-based clinical care.

Lord Oxburgh

  638. Would you agree with the restatement that it should rather than does?
  (Dr Catchpole) Yes.

  639. One of the complaints which we have regularly picked up from those who supply this kind of information from what they describe as the coal face is that they feel they are talking to an empty room, that is this information goes in centrally and they do not see enough or anything back from it in some cases. So I think the point you make is in fact very well taken, but it is vital that the people who put the information in see a return.
  (Dr Catchpole) I forgot to mention in response to Baroness Walmsley's question about what makes a good surveillance system that the information that goes into it is immediately accessible to all those who need to know locally, regionally as well as nationally.


 
previous page contents next page

House of Lords home page Parliament home page House of Commons home page search page enquiries index

© Parliamentary copyright 2003