National Cancer Statistics (NKS) - Methods

The National Cancer Statistics provide information on the status of and trends in cancer mortality (deaths) and incidence (new cases) in Switzerland and in the different language regions, which are defined on a cantonal basis. Cantons are allocated to one of the language regions according to the language spoken by the majority of their population. The cantons of Vaud, Valais, Neuchâtel, Geneva, Jura, Fribourg and Ticino form the "French and Italian-speaking" language region, while the remaining 19 cantons form the "German-speaking" language region.

The National Cancer Statistics include the annual cancer monitoring (data updated on the FSO website) and a national report on cancer in Switzerland, which is published every five years. Various indicators are presented by cancer site to describe the burden of cancer.

These are composite statistics based on data from existing sources: the Federal Statistical Office's (FSO) Cause of death statistics (COD), data from the National Agency for Cancer Registration (NACR), data from the Childhood Cancer Registry (ChCR) and the FSO's population statistics.

Mortality data

Data source
Cancer mortality data are taken from the FSO’s Cause of death statistics (COD), which provide a comprehensive record of the causes of death of people living in Switzerland. Based on the medical cause of death certificate, the FSO codes the causes of death (primary cause, secondary disease, concomitant disease) and defines a main cause of death which is used for the compilation of statistics, including the national cancer statistics. As each death has to be recorded, the survey is nearly complete, although the causes of death of people residing permanently in Switzerland but who die abroad, are rarely transmitted to the FSO.

Coding
Causes of death are coded according to the International Classification of Diseases (ICD), and the standardisation rules established by the World Health Organisation (WHO) allow good international comparability.

Quality of mortality data
Changes in medical diagnoses over time necessitate the regular adaptation of the International Classification of Diseases.
Until 1994, causes of death in Switzerland were coded on the basis of ICD-8. There were also coding rules specific to Switzerland. When several causes appeared on the medical certificate of death, these rules automatically gave priority to some causes of death, regardless of the order in which the causes had been indicated by the doctor. Some causes of death, in particular cancers, were given preference. Since changing to ICD-10, Switzerland has applied international coding rules, abandoning its own prioritisation rules. This change to the methodology should be borne in mind when comparing years prior to 1995 with later years.

To show how the number of new cancer deaths recorded before 1995 may have been influenced by the coding rules specific to Switzerland (and by the use of ICD-8 before the introduction of ICD-10), the table below shows the ratio between the number of cancer deaths that would have been recorded using ICD-10 and the number of cancer deaths actually recorded using ICD-8. In the older age groups, multimorbidity is relatively common, with the result that several causes of death may be recorded on the cause of death certificate. For this reason, a main cause of death has to be selected in the older age groups more often than in the other groups. The change in method was therefore more pronounced in the older age groups. The data recorded prior to 1995 are not corrected, which needs to be taken into account when interpreting trends.

Value of correction factor of data prior to 1995 by sex and age group

Sex	ICD-10	0-59	60-69	70-79	80-84	85+	Total
Men	C00-C97: All cancers	0.98	0.97	0.94	0.89	0.85	0.93
Women	C00-C97: All cancers	0.98	0.98	0.96	0.93	0.88	0.94

Note for interpretation: if the deaths recorded before 1995 had been coded according to the ICD-10 and to international standards their number would have been equal to = (number of deaths registered before 1995) * (correction factor) Source: Lutz JM, Pury P, Fioretta G, Raymond L. The impact of coding process on observed cancer mortality trends in Switzerland. European journal of cancer prevention 2004; 13: 77-81

Data on new cancer cases (incidence)

Data sources
In Switzerland, new cases of cancer are registered at cantonal level in cancer registries (CCR), which record all new cases of cancer diagnosed among residents of their canton every year. Data from the registries are then collated and checked by the National Agency for Cancer Registration (NACR) before being forwarded to the FSO.

Coding
Data are recorded on the basis of the International Classification of Diseases for Oncology (ICD-O-3) and the corresponding ICD 10 codes are assigned to each case so that the statistics correspond to international standards.
All cases of primary malignant tumours are included in the statistics, with the exception of non-melanotic skin cancers (C00-43, C45-97, ICD-10). Primary tumours are defined according to the rules of the International Agency for Research on Cancer (IARC) / International Association of Cancer Registries (IACR) / European Network of Cancer Registries (ENCR): http://www.iacr.com.fr/images/doc/MPrules_july2004.pdf.

Cantonal registration and national incidence estimates
Before the Federal Act and Ordinance on the Registration of Cancerous Diseases (CRA and CRO) came into force in 2020, legal provisions governing tumour registration varied considerably between cantons, occasionally limiting acces to data. Registers have been set up gradually since 1970 (see map) covering the whole of French-speaking Switzerland and Ticino since 2006, and all German-speaking cantons since 2020. The entry into force of the Act and Ordinance has led to harmonisation of the legislation and the data collected, as well as to changes in the data collection processes.

Two exceptions should be noted in the coverage rates;

1. The registry for the two cantons of Basel (Basel-Stadt: BS and Basel-Land: BL) has only included the district of Laufen (BL) since 2013, which until 1994 had belonged to the canton of Bern. This district's population (approximately 7% of the population of the canton of Basel-Landschaft) was deducted from the cantonal population to calculate incidence rates.
2. Data from 2018 and 2019 for the canton of Vaud and from 2020 for the canton of Aargau have as of yet not been included in the statistical analysis.

Share of the population (in %) covered by a cantonal cancer registry

	German-speaking Switzerland	French- and Italian-speaking Switzerland	Total Switzerland
1991-1995	46.6	70.8	53.5
1996-2000	46.6	85.4	57.8
2001-2005	46.7	85.9	58.1
2006-2010	48.2	100.0	63.5
2011-2015	75.5	100.0	82.9
2016-2020	91.8	87.5	90.5

Where data collection is not complete throughout the country, the data collected (observed data) has to be extrapolated to the whole of Switzerland in order to obtain national estimates. These estimates are important for tracking changes in incidence over time and making international comparisons. The estimates are obtained by applying an extrapolation weight (population with registration/total population) to the observed numbers to offset the lack of registration in a given geographical area or year/period. The weighting is calculated for different strata of variables known to be associated with a cancer risk, such as age, sex and language region. In Switzerland, the prevalence of putative risk factors, such as smoking or dietary exposure, vary between language regions. Statistical Methods for Cancer Reporting in Switzerland, Version 1.0.

Data quality
Data on cancer incidence are usually compared with data on cancer mortality, allowing missing registrations to be detected. In some registries, these comparisons have not been fully and systematically conducted for all the years covered by the registry. This was the case for the Basel-Stadt/Basel-Landschaft registry for 1981-2001 and 2010-2012. For the canton of Zurich, the Zurich/Zug registry confirmed that comparisons have been made systematically since 1997; it was not possible to determine whether complete comparisons were made for the years 1980-1996.

The quality of data from a cancer registry depends, among other things, on the completeness of the data collected, i.e. the number of cases registered in relation to the total number of new cases in the population covered by the registry. Completeness should be close to 100% to enable the comparison of incidence rates (between different periods or geographic areas) to conclude that there are real differences in the risk of developing cancer. A publication* based on methods for assessing completeness appeared in 2017. Furthermore, the IARC publishes other quality indicators for certain Swiss registries**.

The NACR publishes eight indicators as part of its standard work to assess data quality***: 1) number of observed and expected cases; 2) the mortality/incidence ratio; 3) proportion of cases initially identified based on information on the cause of death certificate ( Death Certificate Notification, DCN), 4) proportion of cases registered based on information on the cause of death certificate only (death certificate only, DCO); 5) proportion of cases with histological examination; 6) Proportion of cases without information on cancer stage; 7) Heterogeneity of cancer registries in the proportion of cases registered with DCO; 8) Heterogeneity of cancer registries in the proportion of cases that were histologically examined..

*Lorez M., Bordoni A., Bouchardy C. et al. Evaluation of completeness of case ascertainment in Swiss cancer registration. Eur J Cancer Prev.2017; 00: 000-000

**Cancer incidence in five continents, Vol IX. Lyon: IARC/WHO, 2007. IARC/WHO Scientific Publications No. 160

*** "Annual Data Quality Reports (aDQR)", "Swiss annual Data Quality Report (aDQR)", published on the website of NACR under Downloads and Links, Statistics and reports, Data quality reports: https://www.nkrs.ch/en/downloads

Data on new cancer cases in children

Cancers in children are registered for the whole country by the Childhood Cancer Registry (ChCR). Aggregated data on new cases of cancer and deaths due to cancer in children are then transmitted to the FSO.

Coding
Childhood cancers differ from those arising in adulthood and are classified, according to tissue type, on the basis of the International Classificaiton of Childhood Cancers, version 3 (ICCC-3).

Data quality
The data quality indicators are published on the ChCR website: https://www.childhoodcancerregistry.ch/

Population data

Data sources
Mid-year data on the permanent resident population (persons residing permanently in Switzerland) are used to calculate various indicators. Up to and including 2010, the annual population statistics (STATPOP) were used to calculate the mid-year population (average of the population at the beginning and end of the year). Since 2011, the population and household statistics are used.
The data can be accessed here: https://www.pxweb.bfs.admin.ch/pxweb/de/px-x-0102020000_104/-/px-x-0102020000_104.px/

Data quality
See: Statistik des jährlichen Bevölkerungsstandes (1981-2010) (not available in English) Steckbrief (ESPOP)

Population and Household Statistics (STATPOP)

Definition of Cancer Sites

In Switzerland the classification of the causes of death and of tumour incidence was based on the 8th revision (ICD-8) until 1994, and since 1995 on the 10th revision (ICD-10) of the International Statistical Classification of Diseases and Related Health Problems.

Definition of Cancer Sites

		ICD-8	ICD-10
Mouth and digestive organs	Oral cavity and pharynx	140-149	C00-14
	Oesophagus	150	C15
	Stomach	151	C16
	Colon-rectum	153 + 154.0 + 154.1	C18-20
	Liver	155	C22
	Gallbladder and extrahepatic bile ducts	156	C23-24
	Pancreas	157	C25
Respiratory organs	Larynx	161	C32
	Lung, Bronchi, Trachea	162	C33-34
	Pleura	163	C38.4 + C45.0
Bones	Bones, Joints, Cartilage	170	C40-41
Skin	Melanoma	172	C43
Reproductive organs	Breast	174	C50
	Cervix uteri	180	C53
	Corpus uteri	182	C54-55
	Ovary	183.0	C56
	Prostate	185	C61
	Testis	186	C62
Urinary organs	Kidney	189.0	C64
	Bladder	188	C67
Brain and CNS	Brain and central nervous system	191 + 192	C70-72
Thyroid	Thyroid	193	C73
Blood and lymphatic vessels	Hodgkin lymphoma	201	C81
	Non-Hodgkin lymphoma	200 + 202	C82-86 + C96
	Multiple Myeloma	203	C90
	Leukaemia	204-207	C91-95
	of which lymphocytic leukaemia	204	C91
	of which myeloid leukaemia	205 + 206 + 207.2	C92-94
	of which others and unspecified	207.0 + 207.1 + 207.9	C95

Links