10.1     We have heard from a succession of clinical academics and other scientists that little funding is available for applied research into the problems of antimicrobial resistance, or for the collection of the data necessary for systematic and effective surveillance. The position appears to be in marked contrast to the vast resources devoted, largely by the pharmaceutical companies, to basic research and drug development.

Research into different ways of using antibiotics

  10.2     According to the RCGP, research funding for evaluation of different antibiotic strategies is hard to come by. The College has even taken the unusual step of providing funds itself (Q 310). Professor Finch likewise called for more research into antibiotic strategies. "We need to understand the best regimen for particular infections, not relying on those which were tested at the time of licensing in the US where the practice of medicine differs" (Q 386; cp Griffin p 548).

Research into strategies of professional and public education

Box 11

UK PHARMACEUTICAL INDUSTRY

EXAMPLES OF SURVEILLANCE ACTIVITIES

Glaxo Wellcome (p 407)

Glaxo Wellcome tell us that they "co-operate with local health authorities" to provide information on resistance patterns to guide prescribing. They are also "in contact with the WHO surveillance programme with the objective of both contributing to, and reacting to, knowledge of resistance emergence and spread".

SmithKline Beecham (SKB) (p 476)

According to SKB, the pharmaceutical industry has been involved in surveillance "for some time", with the aims of evaluating the medical need and commercial value of new products, and of informing appropriate use of existing ones. Industry provides funding; the work is done by clinical microbiology laboratories and research institutions; data from most such studies are published. "Overall, SKB is funding microbial susceptibility surveys in over 50 countries."

SKB claim to have been "at the forefront of a movement to address the lack of more global, longer-term studies". SKB funded the Alexander Project, to produce and publish high-quality susceptibility data for respiratory pathogens from Europe and the USA. The project began in 1992, and was extended in 1996 to countries in Asia and Africa. Other similar studies supported by industry "are now emerging".

SKB are among the companies supporting the WHO Antimicrobial Resistance Monitoring Programme. SKB contributed to the funding of WHONET (see paragraph 9.3), and support a Centre-of-Excellence Laboratory Advisory Board to help developing countries' laboratories to produce reliable surveillance data.

According to SKB, industry is "continuing to explore ways of making surveillance data available...to aid local prescribing decisions"; to aid the interpretation of the data, so as to "reduce the potential for recurrence, relapse, transmission and increase in resistance"; and to use surveillance data, prescribing data and clinical outcome data to model and predict the causes, trends and effects of resistance.

SKB acknowledge that the industry's resources put it in "a unique position" to fund data-generation. It does so "willingly, and with the goal of an increasingly evidence-based approach to antibiotic usage in mind".

Zeneca Pharmaceuticals (p 544)

"In support of our marketed anti-infectives", Zeneca recently conducted a 56-centre surveillance programme in the United Kingdom. They also have studies in Europe, and in the USA, in collaboration with CDC.

  10.3     Professor Peter Borriello of the PHLS, Director of the Central Public Health Laboratory, observed that, if prescribing practice does indeed affect resistance, then research is needed into what affects prescribing practice, and how best to influence it. "It is not usually the easiest thing...because you are dealing with individuals" (Q 107). According to Dr Grimshaw, this is in fact a flourishing field, with "an increasing body of international literature from rigorous studies", and a "substantive" programme of research and development funded by the NHS. "Up until now most of the evidence is North American, but the United Kingdom is rapidly catching up" (Q 672). Dr Williams of WHO likewise called for evaluation of programmes of patient/public education (QQ 119, 138).

Collection of prescribing data

  10.4     The PHLS are not in a position to match patterns of prescribing with patterns of resistance (QQ 71-80, 107). Their resistance data are subject to acknowledged limitations; and they have no prescribing data at all, though some data sets exist for parts of the United Kingdom (Wales, and Tayside as described below). The PHLS have plans to begin to remedy this deficiency: in a development of TB surveillance (MYCOBNET), "it is planned to analyse treatment regimens in relation to acquired resistance"; and they are undertaking a GP research initiative by which they hope to acquire a network of sentinel practices. We were surprised to hear that they have not used PPA data for this purpose, though Professor Duerden said, "I think they should be made use of".

  10.5     The Department of Health propose to collaborate with the Oxford University Unit of Health-Care Epidemiology to compare the wide regional variations in use of antibiotics noted in Chapter 2 with variations in resistance (p 344). This "will provide a useful first step in determining whether the level of community antibiotic utilisation affects the level of antibiotic resistance".

  10.6     Dr Davey told us about his work at the micro-level in Tayside. With some difficulty, the Tayside Medicines Monitoring Unit have induced GPs and hospitals to use the same identifying number for each patient, so that individual records can be linked (Q 237). To improve the usefulness of the records themselves, the Infectious Diseases Unit at Ninewells Hospital has produced a stamp or sticker which prompts the doctor to write essential information in the patient's notes; but they are having difficulty exporting the concept to other units in the hospital, let alone elsewhere (p 151, Q 242). Dr Davey drew our attention to the possible implications, for use of this sort of information, of the Data Protection Bill [HL] currently before Parliament, which implements the EU Data Protection Directive (95/46/EC).

  10.7     We have been told often that, whereas GPs are generally equipped to quite a high level with information systems suitable for recording and analysing prescription data, British hospitals are not (DH p 344; AMM QQ 45, 49; Davey p 150, Q 251). Dr Davey said, "Probably the single biggest fault...is that the information systems are not being designed to provide clinically useful information; they have been designed to provide information on accounting and throughput" (Q 259). Dr Winyard of the NHS Executive (Q 784) acknowledged this as "a real issue"; a few hospitals have now installed computer prescribing systems, and the Department of Health has commissioned the National Prescribing Centre to conduct a pilot study with ten of them.

Research into infection control

  10.8     The preface to the Cooke Report says, "The group has been very aware of the continuing need for, and importance of, applied research into the surveillance, prevention and control of hospital-acquired infection". Professor Percival agreed (Q 108), and complained in colourful terms of the difficulty of getting grants for "pragmatic research" into, for instance, control of MRSA. Dr Mayon-White would like to see more formal trials of infection control policies; this would not be easy, but would be an improvement on "best advice and existing practice" (Q 173).

Collection of denominator data for disease surveillance

  10.9     The Public Health Laboratory Service (PHLS) are the first to admit that their surveillance data have shortcomings, and one of the biggest is in the area of denominators (PHLS p 38, cp DH p 335). Generally speaking, to know that in a certain period laboratories in a certain area found X isolates of a certain organism to exhibit a certain resistance is to know little of value, beyond the clinical importance of each individual result. The laboratories can generally express X as a percentage of all isolates of that organism tested for susceptibility; the denominator is the total number of isolates tested, the number of resistant isolates being the numerator.

  10.10     However this is still defective information, since laboratory isolates are only a fraction, and an unknown fraction, of all the examples of the organism at large in the area. What is worse, they are not normally a representative fraction; generally speaking, laboratories only see what doctors choose to send them (Q 77). This will tend to consist of problematical samples, including probably a higher proportion of resistant strains than is found in the general population (Q 111). It will consist largely of samples taken in hospitals (QQ 81, 383): according to the AMM, "Typically some 40 per cent of the microbiology specimens processed by a district general hospital laboratory come from the community" (p 7). It may contain multiple specimens from the same person, or specimens from persons who have infected each other (AMM p 5). It will exclude conditions such as pneumonia which do not often give rise to specimens suitable for susceptibility testing in vitro (i.e. in the laboratory) (Q 77). Finally, it will not be randomised for factors such as age and gender (BSAC p 78). The problem is compounded because, at present, PHLS reference laboratories only see what peripheral laboratories choose to send them. Studies funded by the pharmaceutical industry are subject to the same constraints (SKB p 477). Better denominator data can be achieved, but only with extra effort and expense.

Surveillance beyond the hospital

  10.11     The PHLS admit that their data are largely confined to hospital isolates, and to isolates from invasive disease. They do not know "the wider pattern of resistance or susceptibility in the community at large, where we do not necessarily get the specimens into the laboratories. We have not conducted special surveys through the PHLS for those" (Q 71). The ICNA are particularly aware of the lack of surveillance of "community reservoirs of MRSA, especially in nursing homes" (p 126).

  10.12     Dr Mayon-White similarly told us, "The study of antimicrobial resistance outside hospitals in Britain has been highly selective, typically focused on a single species or group of micro-organisms, often in one geographical area" (p 107). He advocates "the collection of comparable data on antimicrobial resistance in a number of different settings across the country" (Q 173). He told us, however, that it is difficult to find funding for such work, since it goes beyond standard service provision but falls short of research. He suggested that it might find a home in the NHS R&D programme, as "health services research" (Q 175).

  10.13     In general practice, sentinel or spotter practices can be recruited, and invited to submit more specimens than would be necessary for purely clinical reasons (Q 78). Professor Finch considers that a network of such practices would provide "valuable information" (Q 384).

Research into systems of surveillance

  10.14     According to Dr MacGowan of the BSAC, "We are still not very certain what the correct method of surveillance is" (Q 109). He would look to the NHS Directors of R&D to fund such work; but he fears it is not "fashionable". Likewise Dr Bartlett of PHLS calls for careful evaluation of new surveillance systems (Q 107).

A funding gap?

  10.15     In short, across the range of our enquiry, there appear to be research needs, and a lack of public resources to meet them. The research in question would be highly applied, whether into better ways to use existing antibiotics, or ways to educate doctors and patients to use antibiotics more wisely, or means to prevent and control infection, or systems of disease surveillance. Some of it would amount to sheer data-collection, whether to collect the denominator data needed to put reported cases of disease in perspective, or to survey prescribing practice in order to inform the campaign for prudent use.

  10.16     We put this matter to Professor Roy Anderson, a researcher in his own right in the basic but highly relevant field of population genetics, and a Governor of the Wellcome Trust. Speaking for the Trust, he acknowledged the problem: since 1992 the Trust has put around £90m into bacteriology (p 306), but only around £4.5m of that into the area of resistance. He was robustly defensive of the Trust's position: it was for the Government, not the Trust, to fund the work of public sector research establishments such as the PHLS (Q 689), and to support research to improve the operations of the NHS (Q 710). In any case, the Trust received few applications of high quality for support for work of this kind (QQ 706, 711). He acknowledged, however, that the Trust's panels of reviewers for grant applications tended to be composed of the best fundamental researchers, who might not fully appreciate the merit of more down-stream work (Q 710).

  10.17     Professor Anderson indicated that the Wellcome Trust is moving to help to fill the funding gap. Scientific advances in genome sequencing (see Chapter 6) and population genetics are attracting researchers and funding to this area; and the Trust may be going to do more in future for epidemiology (QQ 686, 706). The Trust attaches importance to clinical research, and is funding research fellowships in medical microbiology (QQ 709, 713-5). Although the Trust will not support the PHLS directly, they are hoping to support university-based projects involving the PHLS in close collaboration (Q 689).

  10.18     According to Professor S Amyes of Edinburgh University, whose Department of Medical Microbiology has a large research section devoted to the study of resistance, the Medical Research Council (MRC) has been notably reluctant to support bacteriology in general and resistance-related research in particular (p 544; cp Ayliffe, p 377). The MRC puts over £60m per year into immunology, infection and inflammatory disease. But most of this is basic molecular biology; expenditure specifically addressing resistance came to only £0.3m in 1995-96. The MRC say (p 429), "The current scale of research in part reflects the capacity of the academic science base to develop high quality research proposals. MRC's view is that there is potential to develop more high-quality work on resistance in the United Kingdom". In January, the MRC held an interdisciplinary workshop with the Department of Health and a number of independent experts on resistance, to explore ways in which academia might complement the efforts of industry; this recommended that clinical microbiology and antibiotic resistance should be priority areas for MRC training schemes and LINK (academic-industrial research partnership) awards.[76]

  10.19     The apparent lack of high-quality proposals may be deceptive: the Royal College of Pathologists (p 455) identifies a vicious circle in operation. "The climate is such that many who would apply, decide not to do so because they are reluctant to waste their own and funders' time—and this despite funders' stated wish, at least from time to time, to support such work."

  10.20     We put the matter to the Minister for Public Health and the Chief Medical Officer, and they acknowledged the problem. The Minister said (Q 780), "The way in which research budgets are presently constructed does not necessarily mean that public health issues are given a proper opportunity to have a bid for resources". She regards this as a "fundamental problem about the structure and nature of public health research". She has already left the MRC "under absolutely no illusion" as to her view of the matter, and she intends to pursue it further. The Chief Medical Officer is similarly anxious (Q 782) that development should be supported, as well as research (e.g. resistance breakpoints, as well as basic molecular biology); he considers that this is appropriate for support from the NHS R&D budget.