Select Committee on Science and Technology Minutes of Evidence

Examination of Witnesses (160-178)



Lord Patel

  160. I want to go back to the original question Lord Oxburgh raised about standards—and to a degree Dr Crook answered it—in microbiological methodology. The other side of the question was about the standard range of diagnostic pathogens that could be tested for. Are you saying that that would be a good thing to establish—that you test for a standard range of pathogens—because if you are then that clearly must have cost benefit analysis?
  (Dr Kelsey) Yes, but to some extent in the past we have seen these as political decisions. They are not purely taken for medical reasons sometimes. That may be a little bit too harsh but the cryptosporidium, for example, is one—it is waterborne and occurs in outbreaks of contaminated water supply or swimming pools and in severely immuno compromised patients. Again, the ability for laboratories to take on that additional work was limited, and it will require additional resources. I think these decisions need to be made and this is something the Health Protection Agency needs to do a scientific, medical and perhaps cost benefit analysis on.

  161. So is it a good idea to test for a full range of pathogens?
  (Dr Kelsey) Yes, but within limited resources. We have to ration and we cannot test for every known pathogen on every patient. We already ration. This is why microbiologists are out on the wards doing diagnostic ward rounds; this is why general practitioners do not necessarily send in pathogens—because they know the chances of a reasonable yield are low and syndromic management, such as for community viral pneumonia, as Dr Spencer said, is perfectly adequate.

Lord Oxburgh

  162. It seems to me that the new element in surveillance is NHS Direct. Do any of you have experience or have you found any value in interpreting patterns of reporting of symptoms through NHS Direct, because this is quite a new element in this public health arena?
  (Dr Spencer) This avenue is being explored, I believe. I have heard presentations from NHS Direct in relation to deliberate release scenarios and this is something being taken forward by the emergency response division to capture this information, if there is suddenly a rapid increase in people ringing NHS Direct with certain symptoms. We are trying to produce a very simple A4 sheet with tick boxes for various presentations because many of these diseases associated with deliberate release start with a flu-like illness, so we are pursuing that.

  163. So in principle there seems to be an opportunity for feedback to the GPs rather rapidly?
  (Dr Spencer) Indeed.

  164. "Look out for so-and-so"?
  (Dr Spencer) Yes, and there are other software programmes out there as well that have been trialled certainly in Bristol demonstrating the spread of winter respiratory viruses. Again, you have to collect the data, of course, and feed it in but you can see the pattern associated with wind direction.
  (Dr Crook) If I may respond to Lord Patel's question whether there is standardised surveillance that would address generally the surveillance needs of the country, I do not believe there is. There are different types of surveillance for different types of questions. If you are looking for the general endemic disease like meningococcal disease you have on-going passive surveillance as a very reasonable way of doing that. If you need surveillance for the very unusual or new and emerging disease, the schemes we have at the moment are rather poorer than that, and you cannot have a standard set of organisms that you are testing for that eventuality because it will not be on your list. You have to have a different plan for doing that entirely, and there are ways of capturing those new and emerging diseases, or old diseases that we have forgotten about that may re-emerge, and that needs some thought and some planning in order to conceive how to capture it. The extent to which laboratories play a role in that is probably ancillary to the syndromic recognition of those conditions appearing and in a sense the emergence of HIV is a lovely example of that because there was no test for HIV. It had to be recognised syndromically as a condition, and case definitions determined, and that data gradually accrued until laboratory tests caught up. Then there is the last form of surveillance which is how much burden of disease does the pneumococcus cause in the UK at any one time, and you cannot use the passive surveillance that we have at the moment—you have to define a study where you know the size of the population you are studying, you look for the disease in that population determined syndromically as well as with laboratory support, and from that you are able to determine the burden. So you need different types of surveillance and there needs to be a menu, and there is not a standard one you can just impose in the country and expect it to work.

Baroness Warwick of Undercliffe

  165. The next question is about IT, and has come up in just about every single one of the answers you have given. It has come up certainly in the AMM's submission and I think Dr Spencer exhibits in his brief written evidence to us some scepticism about it. We have had a lot of evidence that improving IT in laboratories could help to integrate surveillance and not put undue pressure on existing staff. Could it? How could it? And what are the barriers to upgrading IT? Presumably money is one of them but what are the other barriers?
  (Dr Spencer) Money is one; availability of suitable software is another—

  166. Could you say something about new systems that might be available?
  (Dr Spencer)—and whether we should wait for the full implementation of the electronic patient records that we are supposed to have up and running in the next n number of years, where n is a very small number. I do not know much about it—Dr Kelsey is more of an expert—but it is a problem.
  (Dr Kelsey) Can I correct something, if I may? I do not think it is laboratory systems that are the problem; I think laboratory systems work very well. There are large numbers of them throughout the world really quite similar. The problems are with the capturing of epidemiological data, and this goes back to my points about order coms and capturing from the person requesting what the real reasons for the inquiry are and what the risk factors are. We do not seem to have that; there are some systems going in but it is not standardised. Nobody has written the standard data set to say, "We need to capture this epidemiological information on these specimens or from these patients with these syndromes". We then have the problem of what is available in terms of output, in terms of epidemiological results, and I do not think in the laboratory level these exist. They may at higher levels where people are doing modelling, where CDSC is doing its data warehousing, and that again is outside my experience, but I think there is a great need for this concept of data warehousing and data mining and looking at inputs from all the different systems, and they are not going to just be laboratory systems. They are going to have to come in from spotter practices and from veterinary and other areas, and whether you need more than one regional centre or even one national centre to do it I doubt but, as I said before, you need to be able to feed that information back to the people who sent it in or else there is going to be little validity in it, because what has happened with a lot of information on management data which has left trusts to go to the Leeds clearing house is that the information that goes up is unreliable, so if there is no feedback, there is nobody there who can say, "I never said this; although you tell me I have an outbreak, this is patently not true". So there has to be better information systems that hand data back down again. On coding and standardisation, we have a problem with coding what comes out of the hospital notes because the coding is now done by coders who work very hard with a very poor resource which is the medical note. So the accuracy of data that goes through under syndromic purposes, ie you are admitted with diabetes mellitus and you have a urinary tract infection or you have a community viral pneumonia is very inaccurate, and it is inaccurate because medical notes are very poor and the coding is done from what is in the medical notes. Also, people write things in the medical notes which are not true. One of our coders came to me the other day and said, "They have written that this patient is septicemic but if I look at ICD10 I have to have an organism to go with that", and the answer is medical terminology is changing more rapidly than is the ICD10 system or the other coding systems, so we talk about in medicine now sepsis syndromes where people appear septic but do not necessarily have an organism where you can isolate microbiology from their blood, but there is no sepsis syndrome or similar in ICD10 at the moment so the coders get very confused. So we are putting out a lot of very poor syndromic information.

Lord Haskel

  167. Introducing IT is not only being done in the medical world but everywhere. What determines the success of introducing IT is quite often how accepted it is by the people who are using it, what their attitude is towards it, how convinced they are of the need for it. What is the attitude in a world of microbiology towards IT?
  (Dr Kelsey) I would say it is very positive. If you go back to the early days of the NHS computer, and some of the first systems were put into laboratories, the data that goes into laboratory systems is very accurate because it is our record and we are judged by the quality of our output so information captured by laboratory systems is accurate and reliable. What happens to it after that is always the problem, of course, so I think there is a very positive attitude to IT because it is the daily tool of all microbiologists and biomedical scientists.
  (Dr Crook) I have taken the opportunity to speak to a friend of mine at the CDC, Clare Broom, who has responsibility for implementing for the United States the accrual of electronic data for the CDC, and it is not simple. It has been a highly complex process and I think Lord Haskel rightly points out people. One of the biggest problems has been the resistance of people, and although I agree with Mike Kelsey that for working within laboratories microbiologists are very positive about IT, providing the means to accrue that data from those laboratories has been one of the difficulties and was faced by the Coserve system that was mentioned. A lot of the problem with implementing Coserve, which was not necessarily an ideal system, related to the unwillingness of the people to work with that system and relay the data and it is not a surprise, as you rightly point out it is a common feature of implementing IT—the resistance of people. What Clare Broom has found equally problematic is the multiplicity of systems that you have to face with implementing an IT system. Every laboratory has its own wrinkles and its own need for running a laboratory system, and to get some system that is open so that electronic traffic can flow is extremely difficult. To interface is said to be easy but in reality it has not been easy for the CDC where they have some of the better people doing it. This is a very difficult area and a hugely expensive one, from what I understand.


  168. We are going to visit CDC so we will make a note of that.
  (Dr Crook) Clare Broom is the right person, if I may suggest.

Lord McColl of Dulwich

  169. I think this question certainly has been dealt with pretty fully but could I just say that as most clinicians, like me, do not send stool cultures when we are presenting with patients with acute diarrhea, the surveillance of food poisoning must be pretty warped, must it not, pretty biased, if swabs are not coming in from the world at large but only from hospitals?
  (Dr Spencer) Yes. I think it goes without saying that the majority of food poisoning incidents go unreported. However, if you look at chicken carcases, the incidence of salmonella since the introduction of vaccination of chicken flocks in the UK has declined dramatically whilst campylobacter, which most newspapers find difficulty in spelling so we never hear about it, has continued to increase.
  (Dr Crook) May I say that you are absolutely right not to send in a stool as a basic principle because it is known that every year people have about two episodes of diarrhoea and if one filled laboratories with the entire UK population's stools twice a year we would have grave difficulties! It would be the wrong approach to surveillance. I know I am making light of it but I do not mean to. There needs again to be some thinking about the case definition and the circumstances in which one chooses to send a stool, and it is less in the laboratory and much more amongst the clinicians like yourself to determine that. I believe that were you to have three or four cases, somehow linked either through a wedding party or otherwise who had diarrhoea, I have no doubt that in general clinicians in the UK are extremely diligent at submitting stools in those cases because the culture of recognising what is of public health importance is extremely high in this country by contrast with many others. I feel I am in somewhat of a position to say that—it is superb in the UK by comparison with many other places.


  170. You made an earlier comment, and for want of time I did not bring it up, which was that it is a question more of reliance on clinical judgment than just sending everything off to the microbiology laboratory. I think you were trying to get that point over. Is that so?
  (Dr Crook) That is very much the case I am making. One of the principal reasons we are very much in support of joint training is to acquire the clinical skills to discern the severity of the illness so that you can make a judgment about which samples to send when, and furthermore—which is not well supported in training both through the Royal College of Pathology where I sit and endeavour to do my best—to understand the quantitation of testing and how to deal with that. So who to test on, why, and how to interpret that can be quantitated. The euphemistic phrase for this is evidence-based medicine and to get that right is terribly important in determining how to send samples and more particularly when not to, so that you can have a view about the risk of not sending samples. To have a view about when not to send stool samples is terribly important otherwise you flood microbiology laboratories with unnecessary work which has been studied. The Bandolier which is part of the evidence-based medicine (Cochrane) group that is very active and grew out of work in a number of centres including Oxford and was strongly supported by the Department of Health publishes once a month and looked at laboratory tests, and most laboratories are testing in the order of 30-50 per cent of samples which, when you explore the need for them, there is no need. They are unnecessarily being sent to laboratories and that is not the laboratory's fault per se—it is proximal in the process of sending a sample. So it is clinical training which microbiologists can play a great role in, and do, by training and teaching and advising clinicians on when to send samples. That is for clinical diagnosis but the same applies to surveillance. You can define when to send samples for surveillance purposes.

Chairman: Moving on to the final question and Lord Haskel, it may be that we will have to ask you to respond to part of this in writing.

Lord Haskel

  171. Part of this question has already been responded to in the discussion on IT, but this is really a question for you, Dr Kelsey, because in your written evidence you say that "it is not clear whether the different models envisaged in the current pathology modernisation schemes are appropriate for microbiology", and we wondered if you could expand on this?
  (Dr Kelsey) Yes. The term "pathology modernisation" is one which is not terribly well-defined and the word "modernisation" is being dragged through the mud every day this week. The problem arises that somebody somewhere has seen the need to change the way we manage pathology laboratories, and this has not been clearly defined. There are several models—at least two, probably more, at least three I can think of in the UK—where people have said, "We will change managing pathology because we think it is a good thing to do". There are no clear outcomes for measuring it. We have examples such as the Lincolnshire model where they moved everybody's contracts from the hospitals to a lead trust and managing all the staff and all the laboratories for one trust. That is one model but is it appropriate to London where the laboratories on the whole tend to be big with heavy workloads, and if you were to move my sector of London all into one budget you would have a budget of £50-60 million, and I do not know of a chief executive who could keep his hands off that budget, so there are dangers in that. Then there are other models of co-operation or factory building. There is the Leeds-type model where they have built a large factory off a motorway so they can bus in huge amounts of samples. I think what we need to remember is that pathology is always best performed closest to the patient, whether they are in primary care or general practice or in hospitals. You need to add the clinical element to the patient. Doing a test can be misleading without the appropriate interpretation: you cannot interpret things without the appropriate clinical skills, which means you have to be close to the doctors or close to the patient. So there is a little bit of anxiety as to what modernisation means in pathology. We then have the next problem which is whether microbiology is like every other pathology specialty. Is it the same as histopathology or chemical pathology? Therefore the question arises whether microbiologists should be in the same networks as the other pathology specialities or in separate networks, and this is difficult and there is not one preferred model over another. I think it is vitally important that we talk to other pathologists because we share a common reference work, common IT, and we exchange information. Histopathologists are constantly in and out of each other's rooms because we add to each other's diagnoses, and that sort of interchange of information is exceptionally valuable. But we do have a role in microbiology which is somewhat greater than the trust or the immediate population we have—a responsibility to the public health of our population—and therefore somebody has said that we now need to create microbiology networks separate from pathology networks but all within the frame of getting ahead of the curve of pathology modernisation; we do not know what the role of the director of regional microbiology is going to be, or the inspector of microbiology. We are told he will be an inspector of anatomy, but we do not really understand what that person's role is going to be. Is it going to be one of accreditation? We already have an organisation that deals with accreditation. So the situation is unclear, and what we do not want to be told is that we all have to conform to one model of modernisation because the pattern of co-operation and of networking may be different in a county like Lincolnshire than from north London, or south London for that matter, so we are anxious we are not forced into something that we are going to regret.

Chairman: Lord Haskel, have you any supplementary?

Lord Haskel

  172. Not really. Obviously there are many ways of dealing with it and presumably it is the outcome that will matter?
  (Dr Kelsey) Absolutely, and what we are frightened of is that somebody will say, "We will form you into networks", and the frightening words are "managed networks" because the PHLS was a managed network which to some extent may have been responsible for its downfall because it was to some extent management that made the PHLS not as acceptable to the NHS as it should have been. NHS laboratories to some extent felt there was not quite the closeness of working relationships with the PHLS as there should have been, or that existed some years before. So, again, if a management structure is forced on us we do not necessarily think it will be appropriate. A co-operative structure existed I think after the Sainsbury's re-organisation of the Health Service back in 1974, I cannot be sure, where we formed regional committee structures, and there was a regional microbiology committee where we looked after the funding of large parts of the equipment, and this was done on a co-operative basis not by a management structure forced down below, and it was one we all accepted and contributed to. So we are a little bit anxious about the nature of managed networks.

Lord Turnberg

  173. That answer runs the risk of getting the message across of protecting the patch while really what modernisation should be all about is trying to integrate for the patient's sake and the public interest again all the laboratory services, whether they be microbiological or general What can be against that?
  (Dr Spencer) If we talk about modernisation of pathology I do not think microbiology can exist without our pathology colleagues or our clinical colleagues. The problem is modernisation of pathology is based on strategic health authorities, whereas many microbiological networks, however you define them, are regional and it is a question of marrying the two. I do not see there is a problem there and I still think you can be part of laboratory medicine or whatever we call our committee, which I think is important because we do share IT and common interests; we may help the histologists make a diagnosis and they may help us so I think we have to be in there with them. But there is nothing to stop us, and certainly speaking from the south west we have an active group of medical microbiologists as well as infection control personnel, so I would rather be inside the tent.
  (Dr Kelsey) The real problem is not networks. We believe networks are good; we believe modernisation is good; we believe pointing our activities to the benefit of the patient is right and appropriate. The problem is what some people believe modernisation does, or networking, or management, which means taking money out of the system because there are going to be savings, and that is what we do not believe is true. We believe in co-operation: we believe in aiming activities towards the benefit of the patient: but what we do not believe is that there are savings to be taken out of the system at this point.

  174. I was quite disturbed by what you said—
  (Dr Kelsey) I thought you might be.

  175. What I want to know is whether the microbiology community will be ready by April 1 to take on the new activities relevant to the HPA? There are so many uncertainties. Are you ready and capable?
  (Dr Kelsey) I think a large number of the uncertainties we have had to talk about today exist whether the PHLS goes on or not because they are about the changing pattern of delivery of care and the changing nature of microbial disease, so I do not think things will get worse. That would be my view, and I think we will do the best we can.

  176. What about the management arrangements the HPA will have to deal with?
  (Dr Kelsey) The majority of laboratories will continue to care for the patient and will continue to do surveillance to the best of their abilities.

  177. We are safe in your hands?
  (Dr Kelsey) I would not go that far!


  178. On those words we must draw this session to a close. May I thank you very much for coming along and for giving such full and informative answers to our questions. There are a couple of questions we would like to put to you for you to answer in writing about training of epidemiologists and microbiological training but, again, thank you so much for coming along and, indeed, you will get the transcript and be able to correct factual errors—although not errors of judgment!
  (Dr Spencer) Thank you for the opportunity.

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