Research
 
RESEARCH GUIDE TO THE N.IRELAND CANCER REGISTRY


 
RESEARCH GUIDE TO THE N.IRELAND CANCER REGISTRY

APPENDIX A

Back to Top
    2   What is a population based cancer registry?

    A population based cancer registry aims to collect data on all malignant and certain non-malignant tumours diagnosed in its catchment population.  The data includes details on the patient, the tumour and treatment, and deaths.  Procedures for cancer registration are widely established throughout the world and generally follow guidelines established by bodies such as the International Union Against Cancer (UICC), the International Agency for Research on Cancer (IARC), the International Association of Cancer Registries (IACR), and the World Health Organisation (WHO).

    Population based cancer registries are an important tool for the monitoring of time trends and geographical variations in cancer incidence.  In 1989 a working group for the Registrar General’s Medical Advisory Committee noted that in addition to these traditional uses of cancer registration the system had become vital in several other areas:

    These included:

    • the management of the substantial resources required for the preventative, curative and laboratory services for cancer;
    • the planning and evaluation of services, particularly the screening programmes for breast and cervical cancer;
    • the planning and evaluation of clinical management and treatment based on accurate and unbiased survival data and clinical trials;
    • research into causes of cancer, involving case control studies and the flagging of cohorts and information for health education and health promotion for both professionals and the public.


    Future uses of cancer registration (especially if linked with other databases) were identified including evaluating programmes of care, quality assurance and relating costs to clinical outcome.  Cancer registration is also essential for participation in international research into the aetiology and epidemiology of cancer.

Back to Top
    3  Data held by the N. Ireland Cancer Registry

    Population based incidence data on cancers in N. Ireland is available from 1993.  The first incidence report of the N. Ireland Cancer Registry, published February 1999, covered the period 1993-95 (ref: 1).  This builds on an in-depth analysis of cancer deaths examining 25 year trends and patterns published by the Registry in 1995 (ref: 3).  The data held by the Registry include information on the patient, the disease and the treatment.  Appendix A contains details of this data and the completeness of various key fields.

Back to Top
    4  Data Security

    The N. Ireland Cancer Registry has a policy regarding security, confidentiality and issue of data, which is available on request.  The N. Ireland Cancer Registry maintains a strict respect for the confidentiality of the data held.  Names, addresses and other personal information are collected to prevent multiple registrations of the same patient and to help identify individuals for follow-up and survival analysis.  The Registry is independent of all other agencies and does not share confidential information except for ethically approved studies or to assist doctors in reviewing their own work.  Standard provisions for ethical confidentiality apply to the Registry which is also registered with the Data Protection Registrar under the Data Protection Act (1984).  All confidential information within the Registry is encoded, protected by security systems and destroyed when no longer needed.

Back to Top
    5  Accessing the Data

    Doctors, consultants and General Practitioners have access to information on patients they have treated.  The data in the Registry is passed from doctors to the Medical Director of the Cancer Registry who is then responsible for the data.  Named data can only be released to medically qualified persons.  Researchers not in that position must have a medically qualified person involved in their research who will take responsibility for the medical confidentiality of the data.  In most occasions ethical approval will be required before named data is released for purposes of research.

Back to Top
    6  Costs

    Data produced in reports and summary tables of anonymised data are freely available on the internet (http://quis.qub.ac.uk/nicr/intro.htm).  Studies which require new data analysis within the N. Ireland Cancer Registry will use resources additional to that for which the Registry is funded.  Researchers must pay for the analysis to extract the data they require.  The charges will reflect the time involved in preparation of the data.

Back to Top
    7  Facilitating Analysis

    Registry staff can provide pointers to facilitate the analysis of  data and recommend specific text books and statistical packages to assist researchers with their analysis.
     

    ACQUIRING DATA
     

  • General Data:  General data is available in aggregated format (unnamed) with basic epidemiological analysis in the Incidence Report 1993-95 or online at: http://quis.qub.ac.uk/nicr/intro.htm
  • Named Data:  Access to named data is restricted to maintain confidentiality.  Researchers must comply with the Registry’s guidelines on confidentiality and issue of data which is available from the Registry on request.  Researchers requiring named data will normally be required to provide proof of ethical approval for their study.

    • - Doctors, GPs and consultants may however have access to named data on patients they have treated.
    - Doctors may also have access to information on patients treated by their colleagues after they have obtained written agreement from the relevant clinicians.
Back to Top
     8  Investigation of Alleged Clusters

    Vigilant individuals will often be concerned that a population has a higher rate of disease than they would expect.  Part of this suspicion may reflect the increasing frequency with which cancer is diagnosed in our population.

    This increase has several causes:
     

  • Firstly an aging population.  As we live longer we are more likely to develop diseases of old age including cancer.
  • Secondly, competing causes of deaths, such as infection, has largely been controlled.
  • Thirdly, technological advances have improved the accuracy of diagnosis and so we are better at diagnosing cancer when it occurs.
  • Fourthly, lifestyle changes including the use of tobacco, a high fat, low fibre diet combined with increased alcohol consumption and reduced levels of exercise have increased our risk of developing cancer.
Back to Top
     
    Definitions

    Cluster – a number of persons, gathered or situated close together.

    Cluster in general epidemiology – a population that has a high rate of disease.

    Points to note:
     

    • It is unlikely that analysis will identify a causal factor : the results can only suggest areas for further investigation.
    • Multiple analyses may generate apparent excess risks by chance.


    Small expected numbers may give spuriously high relative risks.

    It should be remembered that virtually every disease, infectious or not, shows spatial clustering which may represent a coincidental occurrence of causal factors for the disease or it may reflect a cluster of cases each caused by an unrelated mechanism.  It is important to study alleged clusters as this may contribute to the investigation  of previously undetected causes of cancer.  Information, however, from alleged clusters which have been scientifically investigated indicate that the discovery of an unknown cause of cancer is unlikely.

    The Cancer Registry will follow procedures for cluster investigation as outlined by the Ontario Cancer Treatment and Research Foundation (ref: 4).

    The study of individual clusters of disease do not offer many prospects for scientific advance for the following reasons:

    a) Clusters are usually too small to contribute a useful epidemiology study with adequate control of confounding variables.

    b) Reported clusters often use vague definitions of disease with cases too heterogeneous for useful study.

    c) Difficulties defining population at risk.

    d) Poorly characterised, heterogeneous or low concentration of alleged exposure.

    e) Publicity makes unbiased data collection difficult.

Back to Top
     
    STEPS TO TAKE

    1. Assessing the Inquiry
     

      - Reported cases must be of sufficient number, minimum 5 cases per unit of analysis, and of the same type or body system.

      - A plausible biological association must exist with a suspected exposure.


    If it is not possible to analyse in a meaningful way, the informant is provided with general information discussing cancer clusters and specific information on cancer rates and known risk factors which may help them understand what they have observed.

    2. Verification

    If it is possible to proceed the Registry will verify the data including cases known to the Registry.  It will also check any reported cases additional to that known to the Registry.  This reduces reporting bias.

    3. Analysis

    Statistical analysis will then be undertaken taking account of differing age distributions using:

       
      - age and sex standardised incidence rates,

      - comparisons of geographical areas under study with similar areas and

      - analysis using special statistical methods designed for the study of spatial, temporal or space-time disease clusters in conjunction with the Small Area Health Statistics Unit (SAHSU) in London.


    4. The next step is assessment in detail of the biological plausibility using previous epidemiological, experimental and animal studies remembering not to stifle the investigation of previously undetected causes of cancer.

    Recommendations based on the steps above may be:
     

      i) No further action.

      ii) Future surveillance or

      iii) Detailed studies depending on satisfaction of criteria on statistical significance, biological plausibility and documentation of exposure.

Back to Top
    9  References and Additional Useful Resources
     
    1. Cancer Incidence in Northern Ireland 1993-95.  Editors:  Gavin A T, Reid J R.  Published by HMSO, February 1999.
    2. The Expert Advisory Group on Cancer to the Chief Medical Officers of England and Wales.  A Policy Framework for Commissioning Cancer Services.  London:  Department of Health, 1995 (EL (95) 51).
    3. N. Ireland Cancer Registry.  Cancer Deaths in N. Ireland – An Analysis of Patterns and Trends.  Editors:  O’Reilly D, Gavin A T.  1995.
    4. King WD, Darlington G A, Kreiger N.  Response of a Cancer Registry to Reports of Disease Clusters, Br J Cancer 1993; Vol. 29: No. 10 1418-25).
    Additional Useful Resources
     
  • Cancer Registration Principles & Methods, IARC Scientific Publication No 95, 1991; Jenson OM, Parkin DM et al
  • Report of the Standardisation Workshop, Cancer Surveillance Group, Edinburgh, April 1995.
  • Methods for Investigating Localised Clustering of Disease, F E Alexander & P Boyle, IARC Scientific Publication No. 135, Lyon
  • Cancer Incidence in Five Continents Vol. VI, IARC Scientific Publication No. 120, Lyon 1992.
  • Survival of Cancer Patients in Europe, IARC Scientific Publication, No. 132, Lyon
  • Guidelines on Confidentiality and Issue of Data.  Available from N. Ireland Cancer Registry.
  • Directory of Ongoing Research IARC.
Back to Top Back to Top