Epidemiologists can gain valuable information from looking at carefully planned descriptions of the patterns of cancer within populations, usually national or regional populations. Because cancers do not occur uniformly in all populations, such studies have provided many important clues about the causes of cancer. Later, we will describe the distributions of different cancers in the European Community to illustrate how much variation can exist even between rather similar countries. Much greater variations exist between countries in different continents and between groups with very different social and economic standing. Some cancers, notably breast cancer, are strongly associated with the pattern of fife found in developed societies in Europe and North America. Others, notably cancer of the liver, are found much more frequently in the developing world. Clues can therefore be sought by seeing what factors might be most closely linked to the individual cancers under study. Not only can national, social and economic differences yield information. The incidence of cancer at different ages and in different sexes and different races and at different times in history can yield valuable facts and dues. By studying the pattern of cancers in these groups the epidemiologists will seek to identify factors in the environment or factors in the host (the person with cancer) which put that person at a greater risk of cancer and, when possible, to provide an accurate measurement of the size of that risk.
In any population the pattern of cancer can be described in a number of ways:
- Incidence - the number of new cases in each year per head of the population or usually, to make comparisons easier, per 100,000 or per million heads.
- Mortality - the number of people who die of that cancer in each year per head of the population or per 100,000 people.
- Prevalence - the number of cancers that exist at any one time in a particular population, which will depend both on the incidence (new cases developing) and the mortality.
At first glance it might seem easy to produce this information. In fact it can be surprisingly difficult. Death certificates ought to provide accurate information about the mortality from cancer in a whole population when records are kept. In fact death-certificate information can be quite inaccurate and is not always collected well.
Many countries have developed a system of cancer registration in order to collect information about cancer incidence. The information is collected from hospital records, death certificates and hospital laboratories, and great attention is now given to cross-checking and comparing information to ensure its reliability. Most European countries collect information on mortality. However, even within Europe, we find differences in national commitment to collecting such data (Luxembourg, for example, does not have cancer registration). Perhaps the countries with the most outstanding record of accurate collection of valuable information are Denmark and Scotland. In England and Wales the regional cancer registries are broadly accurate, but some registries arc better than others. Cancer registration is improving rapidly in the new southern European members of the European Community, such as Portugal and Greece.
IB the United States national cancer registration began towards the end of the nineteenth century, but it was not until 1933 that information was collected in all states and only as late as 1979 was a national registry of all cancer deaths established.