Elsevier

European Journal of Cancer

Volume 42, Issue 13, September 2006, Pages 1915-1951
European Journal of Cancer

Quality, comparability and methods of analysis of data on childhood cancer in Europe (1978–1997): Report from the Automated Childhood Cancer Information System project

https://doi.org/10.1016/j.ejca.2006.05.007Get rights and content

Abstract

In collaboration with 62 population-based cancer registries contributing to the Automated Childhood Cancer Information System (ACCIS), we built a database to study incidence and survival of children and adolescents with cancer in Europe. We describe the methods and evaluate the quality and internal comparability of the database, by geographical region, period of registration, type of registry and other characteristics. Data on 88,465 childhood and 15,369 adolescent tumours registered during 1978–1997 were available. Geographical differences in incidence are caused partly by differences in definition of eligible cases. The observed increase in incidence rates cannot be explained by biases due to the selection of datasets for analyses, and only partially by the registration of non-malignant or multiple primary tumours. Part of the observed differences in survival between the regions may be due to variable completeness of follow-up, but most is probably explained by resource availability and organisation of care. Further standardisation of data and collection of additional variables are required so that this study may continue to yield valuable results with reliable interpretation.

Introduction

In European populations, about 1% of all malignant neoplasms occur in patients aged less than 20 years.1 This low frequency represents a major difficulty for studies of putative risk factors and clinical management, and the problem is further accentuated when the number of cases is split into a wide variety of tumour types, most of which are uncommon in adults. As a result, international data on childhood and adolescent cancer are sparse.2, 3, 4, 5, 6

The Automated Childhood Cancer Information System (ACCIS) is a European project aimed at collection, analysis, interpretation and dissemination of data on cancer incidence and survival of children and adolescents in Europe.7 To date, the ACCIS database contains data from some 80 population-based cancer registries, which cover about 50 % of the population aged 0–14 years and about 25% of the population aged 15–19 years living in the 35 participating countries. Over 160,000 childhood and adolescent cancers were diagnosed during the period 1970–2001 in the 1300 million person-years of observation. The wide coverage and large size of this study permit identification of geographic and temporal differences in incidence across Europe, and so provide useful information for generating aetiological hypotheses.8 Simultaneously, differences in population-based survival of various groups of patients help to identify areas for improvement in the management of childhood and adolescent cancer patients.5, 6, 9

However, valid conclusions about differences in cancer incidence or survival between populations and over time can be drawn only from comparable datasets of high quality with complete registration and follow-up. This paper describes the methods of data collection and analysis and evaluates internal quality and comparability of the part of the ACCIS database that was selected for detailed analyses of incidence and survival of individual and combined tumour types. It also discusses the extent to which the comparison of incidence rates and survival may be influenced by the differences in methods of registration and follow-up and identifies priorities for further improvement of data comparability across Europe.

Section snippets

Setting up the ACCIS database

Some 100 European population-based cancer registries were invited to participate in the ACCIS study. Acceptable data were received from 78. To participate, the registries submitted a file of records for all cancer cases incident over a specified registration period (from about 1970 onwards) in children (aged 0–14 years at diagnosis) and adolescents (age 15–19 years) resident in the defined registration area at the time of diagnosis. All malignant and non-malignant tumours of the central nervous

Study size

Overall, 88,465 tumours in children and 15,369 in adolescents were included in various analyses (excluding the overlapping registrations). In addition, 2199 childhood leukaemia cases from DCOG were included in leukaemia-specific analyses in replacement of 863 leukaemia cases of the two other Dutch registries [Coebergh and colleagues, leukaemias, this issue]. The precise sample size varied between the different analyses.

The underlying person years at risk included in the analyses of time trends

Discussion

Sixty-two population-based cancer registries contributed to the analyses. They differed by size, geographical location, administrative area, age-range and other criteria for registration, time period available, methods of registration and follow-up of the patients, data management systems, availability of data sources, data protection legislation, etc. These different settings contributed to the complexity of the ACCIS database, its exploitation and interpretation of results. We have chosen to

Conclusion

Despite common recommendations and standardisation of data, the subsets of the ACCIS database are not strictly comparable. However, we have identified a number of artefacts and quantified their possible impact on the comparisons. Geographical differences between the regions are relatively small and removal of the detected artefacts would reduce them to some extent. Further standardisation of data collection and processing is therefore necessary, especially in children and adolescents.

Conflict of interest statement

None declared.

Acknowledgement

The ACCIS project was funded by the European Commission from the Europe Against Cancer programme (contracts SI2.126875, SI2.321970 and SPC.2002303), jointly with the International Agency for Research on Cancer. Data analyses were partly funded by the French Ligue Nationale Contre le Cancer, Comité du Rhône. ESF was partially supported by the Communauté urbaine de Lyon and Federal Ministry of Health of the Federal German Government.

The authors wish to thank Mr Nicolas Mitton (France) for his

References (47)

  • Ries LAG, Smith MA, Gurney JG, et al., editors. Cancer incidence and survival among children and adolescents: United...
  • Capocaccia R, Gatta G, Magnani C, Stiller C, Coebergh JW, editors. Childhood cancer survival in Europe 1978–1992: the...
  • J. Little

    Epidemiology of childhood cancer

    (1999)
  • E. Kramárová et al.

    International classification of childhood cancer 1996

    (1996)
  • L. Teppo et al.

    Finland, Finnish Cancer Registry, 1980–1989

  • E. Kramárová et al.

    The International Classification of Childhood Cancer

    Int J Cancer

    (1996)
  • J. Berkson et al.

    Calculations of survival rates for cancer

    Proc Staff Meet Mayo Clin

    (1950)
  • Ferlay J, Burkhard C, Whelan S, Parkin DM. Check and conversion programs for cancer registries (IARC/IACR Tools for...
  • Intercooled Stata 8.2 for...
  • Microsoft® Excel...
  • D.M. Parkin et al.

    Comparability and quality control in cancer registration

    (1994)
  • Cited by (49)

    • International incidence of childhood cancer, 2001–10: a population-based registry study

      2017, The Lancet Oncology
      Citation Excerpt :

      Inclusion of the 15–19 years age group was motivated by the shortage of internationally comparable data in this transition age between childhood and adulthood. An age range of 0–19 years was also chosen in previous US and European studies of childhood cancer incidence and survival.3,4 Evidence before this study

    • Registration of childhood cancer: Moving towards pan-European coverage?

      2015, European Journal of Cancer
      Citation Excerpt :

      This point could be checked indirectly by comparing region-specific cancer mortality (if available) between areas that are covered by registration and areas that are not, although this validation has limited value in childhood population due to relatively high survival [92]. In a previous study, regional cancer registration (then concentrated in Southern and Western Europe) yielded higher incidence rates than national systems [93]. In some countries, the regional registries exchange and complete their databases among themselves or with a paediatric registry, although this is not the rule in all countries with multiple registries.

    • Uses of cancer registries for public health and clinical research in Europe: Results of the European Network of Cancer Registries survey among 161 population-based cancer registries during 2010-2012

      2015, European Journal of Cancer
      Citation Excerpt :

      These studies were: the ECO web site [22,23], Cancer Incidence in 5 Continents, volume IX and X [7], and EUROCARE 4 and 5 with its subsidiary database [8] i.e. the Rarecare project [31]. Participation to other European projects is also considered as a sign of adequacy, e.g. the EUNICE project on methodology of cancer survival [32], Automated Childhood Cancer Information System (ACCIS II), the Epiderm study on Basal cell carcinoma of the skin [33], the International cancer Benchmarking Partnership (ICBP) [34] and the European Prospective Investigation into Cancer and Nutrition (EPIC) project on diet and cancer [13]. The responding CRs are referred to as general, if they registered most if not all tumours occurring in the entire population of their registration area.

    View all citing articles on Scopus
    View full text