House of Commons - Science and Technology

Select Committee on Science and Technology Sixth Report

CANCER RESEARCH

Epidemiology and Public Health Research

96. In order to evaluate progress in the prevention and treatment of cancer, it is vital to know the number of people who develop the different forms of the disease each year (incidence), and their survival rates. We also need to know the number of cancer deaths, and how these figures change over time.

97. Cancer registries provide this information. Registries collect data for all cancer patients in a defined population, and they provide a detailed and evolving picture of the public health burden of cancer. The UK has been covered since 1962 (since 1993 for Northern Ireland) by a system of population-based cancer registries which record a standard NHS minimum data set for all cancer patients living in their region. It is the largest national cancer registration system in the world. The registries are funded by the Department of Health, and data are published regularly on cancer incidence, survival and mortality, both at regional and national level. National data are published by the Office for National Statistics, and contributed to the World Health Organisation for the purpose of international comparison.

98. A review of the cancer registration system in England and Wales was carried out in 1999 by a panel chaired by Professor Charles Gillis, and his report is due for publication later this year. The cancer registration system has been criticised for the inadequate quality, completeness and timeliness of its data collection and publication. The priority accorded to cancer registration within the NHS is reflected in low investment - an estimated £15-20 million per year. The London School of Hygiene and Tropical Medicine told us that "Chronic under-funding and the lack of strategic vision from the Department of Health are largely responsible for the fact that both regional and national cancer registries have had great difficulty producing timely and accurate data on cancer incidence and survival". A similar point was made by the Director of the Northern Ireland Cancer Registry.[142] We agree with the London School of Hygiene and Tropical Medicine: "Cancer registration in the UK is too valuable as a public health resource for its organisation and funding to continue in their current haphazard fashion: this is the result of historical accident rather than strategic planning."[143] We return to this issue in paras 103-112.

THE DATA PROTECTION ACT 1998

99. Serious concerns have been expressed by public health researchers over the possible impact of the new Data Protection Act 1998 on their ability to gather and process complete data.[144] In pursuit of their public health functions, cancer registries have for 40 years relied on the concept of implied consent. That is, people are registered unless they explicitly say that they do not want to be.[145]

100. The Data Protection Act incorporates into UK law the European Directive of 1995 (Directive 95/46/EC) on the protection of individuals with regard to the processing of personal data. It provides important safeguards for individuals about whom data are stored, including a requirement that they consent to storage and processing of personal data about them. Exemptions to this provision include certain uses of data for statistics and research purposes, but these are less precisely defined in the UK legislation than in the European Directive, which exempts data "required for the purposes of preventive medicine, medical diagnosis, the provision of care or treatment or the management of health-care services, and where those data are processed by a health professional subject under national law or rules established by national competent bodies to the obligation of professional secrecy or by another person also subject to an equivalent obligation of secrecy".[146]

101. Local interpretations of the Data Protection Act have already led to anomalies in cancer registration practice. The UK Association of Cancer Registries told us that some hospitals are requesting that the service level agreements being established with cancer registries include a clause requiring patients' consent for inclusion of their data.[147] The Institute of Cancer Research told us that enforcement of the 1998 Act has already seriously held up their research into mesothelioma. Whereas previously, medical and other patient records were available to bona fide medical researchers, access now appears to require explicit patient permission in each case.[148] The London School of Hygiene and Tropical Medicine believes that if this interpretation of the Act is correct then "it would make both cancer registration and almost all cancer epidemiological research effectively impossible".[149] Professor Ponder is concerned that this could "destroy the cancer registration system".[150] He reports having received advice from a MREC that "in effect, all of the genetic epidemiology studies which form a large part of our work, are in their opinion illegal... because of doubts whether cancer registration itself is legal".[151]

102. We agree that access to high quality data on incidence and outcome of cancer is an essential tool in the study of cancer and that it would be a great loss if implementation of the Data Protection Act 1998 were to compromise the integrity and comprehensiveness of these data or hinder researchers' access to them. We believe that a requirement for informed consent for cancer registration would be difficult to administer and costly, and would almost certainly lead to unquantifiable selection bias, making the data far less useful. In some circumstances, for example the investigation of childhood leukaemia in the vicinity of nuclear installations, it might become difficult to retain any confidence in the data.

CANCER — STATUTORY REGISTRATION

103. NHS Trusts and laboratories are under a legal obligation to notify the authorities of all patients who contract notifiable diseases for public health protection purposes. Under legislation dating back to the 1880s, 30 diseases or groups of diseases are notifiable, including smallpox, plague and cholera;[152] similar legislation is in force in Scotland.[153] Notification of these diseases is exempt from the requirement for the subject's consent under the Data Protection Act 1998.

104. One solution put to us to ensure that cancer registries can continue to collect comprehensive data and analyse them, and that registry data remain available to bona fide researchers, is to make cancer a notifiable disease.[154] The London School of Hygiene and Tropical Medicine told us that making cancer notifiable would have three benefits: it would emphasise the political importance attached to cancer; registration would become exempt from the 1998 Act requirement for patient consent; and it would place a legal obligation on NHS Trusts and laboratories to co-operate in registration, thus improving the quality of public health data on cancer available to Government and researchers.[155]

105. The Data Protection Commissioner would appear to agree. Ian Bourne, the Compliance Manager (Health) in the Commissioner's Office, told us that "cancer registration could be put on a statutory basis, as is the case with notifiable diseases. This means that practitioners must or could disclose regardless of the wishes of the individual data subjects. Although it is not for the Commissioner to take a view as to whether cancer registration should be put on a statutory basis, by virtue of [Section] 35 of the 1998 Act personal data are exempt from the non-disclosure provisions where the disclosure is required by or under any enactment or by any rule of law.".[156]

106. Legislative or regulatory measures for compulsory cancer registration are in force in almost half the 50 countries reporting data to the World Health Organisation's international compendium of cancer statistics.[157] In the USA, the Cancer Registries Amendment Act 1992 invoked a National Program of Cancer Registries, noting that statewide cancer incidence and mortality data can be used to examine the ethnic and geographical variations and to direct cancer control interventions.[158] The Act enables the US Health Secretary "to support the operation of population-based statewide cancer registries to collect, for each form of in-situ and invasive cancer (with the exception of basal cell and squamous cell carcinoma of the skin)", demographic, occupational and administrative data on each person with cancer and the data of the diagnosis, pathology and treatment of cancer.

107. When we asked the Minister if she had considered the option of making cancer notifiable she told us that this was "not a proposal which I have looked at before" and that there was "potential for cancer registries to make a big difference without needing to do that".[159] The National Cancer Director suggested that this issue should be "revisited if we had problems actually getting 100 per cent information or near 100 per cent information on cancer patients".[160] The Chief Medical Officer for England, Dr Liam Donaldson, has previously described the national cancer registry as "the best example of a disease register that attempts to provide comprehensive, routinely available data for the whole of Britain on an important group of diseases".[161] A recent British Medical Journal editorial on cancer trends in England and Wales points out that "Tough but achievable targets cannot be set, services planned, and performance audited without high quality, timely, and complete statistics on incidence, mortality and survival by tumour type; analysis of their trends and of the factors explaining these trends; and predictions of the effects of these factors as the age structure of the population changes. This is an enormous but under-funded task: it currently takes six years to report the number of cancers which occur in a specific year.".[162] The cancer registries covering Great Britain have been in operation for 40 years and comprise the largest national collection of such data in the world; they have been the source of a very large volume of research on the causes, prevention and treatment of cancer and survival from cancer.

108. We believe that this is not an area where there is any room for complacency. Operation of the cancer registry in Saarland, Germany, was suspended for more than a year due to enactment of a federal data protection law in 1977. On reopening in 1979 access to data for research was still restricted. The former East German cancer registry, in operation since 1953, was forced to close under the same law after the re-unification with West Germany in 1991.[163] Accurate and comprehensive data sets on cancer incidence and outcome are crucial for public health research into the causes and effective treatment of cancer. Some research projects are already under threat, even where there is no threat to patient confidentiality.[164]

109. We note that an advisory group has been established to examine the issue of patient confidentiality and data protection across the NHS. We recommend that, as a matter of urgency, the advisory group on patient confidentiality should address the concerns posed by the 1998 Data Protection Act regarding the registration of cancer.

110. Making cancer a notifiable disease would be a departure from the established practice of applying notification solely to infectious diseases. Some cancers may be caused by the long-term effects of persistent viral infections (for example, liver cancer and hepatitis B), but we feel there is a risk that allowing cancer to become associated with infectious notifiable diseases by using the same legal mechanisms to ensure its registration could give rise to the unwelcome impression that cancer patients should be isolated. Making the registration of cancer a legal requirement through separate legislation would, however, ensure that cancer registries have comprehensive and useful data. Accessing those data for research purposes would continue to require approval from a research ethics committee (REC) and patient confidentiality issues could, as now, be dealt with at that stage. Ethics committees properly require that patient records are only used for worthwhile research purposes and we accept that they represent a rigorous safeguard.[165]

111. One concern about patient confidentiality is that in some research it is necessary for researchers to approach patients for whom a cancer diagnosis has been recorded in the register. This is usually done through the patient's GP who advises whether or not such an approach would cause distress to the patient.[166] Such safeguards can be and already are routinely imposed by ethics committees in order to protect patients' interests without unnecessarily obstructing important research.

112. We recommend that the Government should introduce legislation to make the registration of cancer a legal requirement, both to ensure the completeness of cancer registry data and to ensure access to those data for legitimate research purposes.

SOCIOECONOMIC AND ETHNIC DIFFERENCES IN CANCER INCIDENCE AND OUTCOME

113. Mrs Markhandoo, a breast surgeon working in Hackney in East London, told us that cancer survival rates in East London were 22 per cent lower than in the country as a whole. Hackney has higher than average socioeconomic deprivation and relatively low literacy levels, and many women do not take part in breast screening. Mrs Markhandoo reports that with current data it is difficult to identify whether the poor breast cancer survival rate in black women is due to socioeconomic factors or some ethnic differences in the nature of the cancers.[167] The link between poverty and cancer outcomes is well documented. Challenging Cancer, for instance, states that "Too many of the poorest in our society die from cancer. If the survival rates among the poorest matched those among the richest in England and Wales, 12,700 untimely deaths could have been prevented among those diagnosed between 1986 and 1990.".[168] A study by the Cancer and Public Health Unit at the London School of Hygiene and Tropical Medicine and the Office for National Statistics also found that "survival is lower in deprived areas for most adult cancers".[169] Cancer Black Care, a support group aiming to address the cultural and emotional needs of black cancer patients and their families, drew attention to a link between prostate cancer and race. We heard similar concerns from the American Cancer Society: indeed it is noticeable that much of the research upon which these findings are based has been performed in the USA. We are not aware of any equivalent research on cancer outcome by ethnic group in the UK. Better information is necessary if these factors are to be fully understood and appropriate action taken.

114. Like the rest of the NHS, cancer registries have collected ethnicity data since 1993 and, over time, better quality data should become available on which to base further investigations into links between race or ethnic group and cancer. It is important to understand such links since they affect the way in which cancers should be treated and the way in which public health education is best delivered. We would welcome initiatives to extend research on the links between ethnicity, socio-economic conditions and cancer incidence and survival.

Research into Improving Patients' Quality of Life

115. Treating cancer is not only an attempt to cure or slow the progress of the disease. For many patients, improving their quality of life while living with cancer, or coping with the side-effects of treatment, is just as important as the treatment of the disease itself and, in some cases, more important. Cancerlink told us that "Quality of life must represent a more significant position in cancer research issues, which often seem heavily dominated by marginal gains in 'survival'".[170] CancerBACUP told us that "a cure may not always be possible and research efforts should also focus on maximising quality of life".[171] Mrs Newton of the Cancer Resource Store told us "there should be more research on the impact of living with cancer".[172] We have already discussed the various complementary therapies sought out by patients to help them cope with cancer and its effects (see para 40). Beyond complementary therapies, there are other research areas which address these quality of life issues.

PSYCHOSOCIAL RESEARCH

116. There is ample evidence of the psychological impact which a diagnosis of cancer can have on patients and, indeed, the Calman-Hine report recommended more psychological support for patients, their families and carers. The majority of cancer centres, however, still have no consultant-led psychological service. Macmillan Cancer Relief told us that "With greater numbers of people living, and living longer, with cancer, it is increasingly clear that patient care goes beyond medical intervention. Evidence suggests that survival and recovery rates may be significantly enhanced when treatment is delivered via a multi-disciplinary team and encompasses care and practical and emotional support".[173] Psychosocial research attempts to understand this impact and to find approaches to managing patients which help them to cope with their diagnoses. The CRC told us that they will spend £2.5 million in the coming year on communications projects, helping to understand the doctor-patient communications issues which are so important in enabling patients and their families to come to terms with the condition.[174] The ICRF spends about £1.5 million each year on research in the psychosocial area.[175] We would welcome a psychosocial service for cancer patients and their families and carers in all cancer centres and units and would encourage research into the beneficial effects of these services.

PALLIATIVE CARE RESEARCH

117. Palliative care seeks to ease the impact of living with cancer and its treatment. Patients' quality of life when receiving toxic drugs or when in the untreatable stages of cancer can be greatly improved with good palliative care.[176] Macmillan Cancer Relief argued that "this is particularly true of vulnerable patients in an advanced stage of the disease for whom invasive treatment may be inappropriate".[177] Nevertheless, research into service development and patient care is accorded far less priority for funding and attracts far less academic interest. Macmillan Cancer Relief, which told us of the positive results of its research into the benefits of care-based research on patient outcomes, appears to be one of the few organisations investing in research in this area.[178] The Royal Pharmaceutical Society told us that research priorities may be distorted because "there are no clear national targets relating to this area of opportunity, and the health economic and other evaluative techniques currently in use may not attribute an appropriate value to care which is not life extending".[179] We recommend that palliative care research should be conducted as a complement to research into cancer treatments and should receive Government funding.

Patient Involvement in Cancer Research

118. The central objective of most cancer research is to improve outcomes for cancer patients, be that in the short or long term. Several witnesses suggested that patients themselves should therefore be more closely involved in setting the research agenda. Cancerlink, for instance, said that "cancer patients and their carers should be involved in decisions about research at every level—from deciding on research priorities to application and dissemination".[180] CancerBACUP argues that "research is primarily driven by scientists and oncologists, and does not always reflect the priorities of patients. Patients are interested not only in cures for cancer but also in cancer prevention, especially for future generations, and in quality of life issues".[181] Mr Fred Walker, a cancer patient and secretary of the Castleford and District Self Help Group, agreed that patients and their carers should be involved with research decisions, stating that this involvement should be more than "a tokenistic gesture".[182] That such a large proportion of the funds for research are raised by patients and their carers through the research charities makes the case for their involvement in setting research priorities even stronger. The ICRF told us that it has to be accountable to its donors, who are kept fully up to date with directions and progress and that, together with the CRC, it is seeking ways of improving accountability to donors.[183]

119. Scientific judgements on the timeliness, feasibility and promise of various research approaches are matters properly left to experts. We have been persuaded, however, that decisions over long-term strategies for tackling cancer can and should involve patients, carers and their representatives. In some instances this already happens. Some research committees have patients representatives. Indeed, one of our witnesses, Gillian Speed, a cancer patient from Sheffield, is a patient representative on the North Trent Cancer Network Culyer Funding Team.[184] Research Ethics Committees also have lay members. Patients should also be consulted in the design and delivery of clinical trials. Moreover, as Cancerlink told us, patient involvement of this kind requires investment in training and support to build the capacity of patients and their representatives to take part in the research decision-making process.[185] We recommend that all Government-supported cancer research committees include patients' representatives as full members and that the Government provides resources to ensure proper training and support for patients' representatives.

142 See Annex 4. Back

143 Ev. p. 311. Back

144 See, for example, Ev. p. 155. Back

145 Q. 453. Back

146 EU Directive 95/46/EC, Article 9(3) See also http://europa.eu.int/eur-lex/lif/dat/1995/en_395L0046.html. Back

147 Q. 563. Back

148 Ev. p. 271. See also Ev. p. 376. Back

149 Ev. p. 311. Back

150 Q. 453. Back

151 Ev. p. 387. Back

152 Public Health (Control of Disease) Act 1984, 1984c22; The Public Health, England and Wales (Infectious Disease) Regulations 1988, SI1988/1546. Back

153 Public Health (Notification of Infectious Diseases) (Scotland) Regulations 1988, SI1988/1550 (S155). Back