Cancer burden in the year 2000. The global picture

Abstract

Although the general idea of ‘burden’ of a disease to a community seems fairly straightforward, there are multiple dimensions in which it may be expressed, either in terms of disease frequency (the ‘need’ for services) or the demand which it places upon them. In this review, we confine ourselves to three elementary measures of cancer frequency: incidence, mortality and prevalence. Incidence is the number of new cases occurring. It can be expressed as an absolute number of cases per year (the volume of new patients presenting for treatment) or as a rate per 100 000 persons per year. The latter provides an approximation to the average risk of developing a cancer, and is necessary if we wish to compare the risk of disease between populations (countries, ethnic groups, or different time periods within a country, for example). When considering the impact of primary prevention strategies, a reduction in incidence (occurrence of new cases) is the appropriate statistic to use. Mortality is the number of deaths occurring, and the mortality rate the number of deaths per 100 000 persons per year. The number of deaths provides one measure (and a rather unambiguous one) of the outcome or impact of cancer. It is the product of the incidence and the fatality of a given cancer. Fatality, the inverse of survival, is the proportion of cancer cases that die and this is generally assumed to be the most severe sequel of the disease. Mortality rates therefore measure the average risk to the population of dying from a specific cancer, while fatality (1-survival) represents the probability that an individual with cancer will die from it. Mortality rates are sometimes used as a convenient proxy measure of the risk of acquiring the disease (incidence) when comparing different groups, since they may be more generally available (as described below). However, when used in this way, an assumption of equal survival/fatality in the populations being compared is introduced. Since this is rarely correct—there are, for example, quite large differences between countries—it is safer to use mortality as a measure of outcome rather than occurrence. Prevalence: There is no agreed definition of ‘prevalence’ of cancer. Strictly speaking, it is the number of persons in a defined population alive at a given time who have had cancer diagnosed at some time in the past. However, the resource requirements for treating newly diagnosed patients are very different from those for supporting long-term survivors. Thus, overall prevalence is not particularly useful for healthcare planning purposes, especially as a large proportion of long-term survivors can be considered cured. Partial prevalence, which limits the number of patients to those diagnosed during a fixed time in the past, is therefore a more useful measure of cancer burden. Prevalence for cases diagnosed within 1, 3 and 5 years are likely to be of relevance to the different stages of cancer therapy, namely, initial treatment (1 year), clinical follow-up (3 years) and cure (5 years). Patients who are still alive 5 years after diagnosis are usually considered cured since the death rates of such patients are similar to those in the general population. There are some exceptions, primarily that of female breast cancer, for which the risk of death remains higher than the general population for many more years. Several other more complex statistics have been used to measure the impact of disease, particularly in health economics. They include person-years of life lost (how many years of normal lifespan are lost due to deaths from cancer). This measurement may be refined by giving different values to life-years at different ages, so that a year saved at, for example, age 20 years, is valued