Treating Cancer as a Public Health Problem
As a result of the growth and aging of the U.S. population, more Americans died of cancer in 1999 than in any previous year. Yet mere numbers do not by any means reflect the true state of our war against cancer. I have long argued that cancer statistics can be quite misleading and that only an age-adjusted mortality rate--one that takes into account the growth and aging of the population--allows for fair comparisons between different populations and different years.
According to National Cancer Institute data that have been age-adjusted to reflect the composition of the U.S. population in 2000, mortality rates in the United States peaked in 1991. They then declined, albeit at less than 1 percent per year, throughout the 1990s. (The most recent available data are from 1999.)
Most of the recent declines have been for men. In 1984, age-adjusted mortality for all cancers in the United States combined was 210 per 100,000 population (275.1 in men and 170.5 in women). The rate reached a high of 215.4 in 1991 (279.1 in men and 175.6 for women) and then dipped to 202.8 in 1999 (252.6 in men and 169.6 in women). Very little further change occurred between 1998 and 1999, so it may be several years before we know if whether the tide of cancer deaths has truly turned.
Despite signs of progress, however, it is clear that we could be doing more to stem death rates from cancer. For many types of cancer, the reasons for the drop in mortality have yet to be uncovered, meaning that it will be difficult to accelerate the downward trend. For others, such as lung cancer, gains are distributed unevenly throughout the population. Here is a glimpse of how we are doing on some important cancer fronts, focusing on data from 1984, 1991, and 1999:
Lung cancer. Of all the changes in cancer mortality rates, the biggest is for in lung cancer mortality in men. It stood at 88.2 per 100,000 population in 1984, reached 89.9 in 1991, and then fell to 77.2 in 1999, a more than 14 percent decrease. This is quite clearly a result of the long anti-tobacco campaign, which caused many men to give up smoking cigarettes (or not to start) in the 1970s and 1980s. Unfortunately, many women began to smoke at about the same time; their increase in deaths from lung cancer (28.9 in 1984, 37.7 in 1991, and 40.7 in 1999) erases much of the gain in men. A worldwide commitment to reduce smoking in adults and to prevent children from starting would pay great dividends in many aspects of health, among them the fight against cancer.
Childhood cancers. Cancer in children and young adults (up to age 20) fell from 3.9 to 3.4 to 2.9. The drop has been substantial for every form of childhood cancer. The decline in cancer mortality in the U.S. population as a whole first began in persons under 20 and has been largest in young adults and children, with progressively smaller decreases in older and older age groups. (Indeed, cancer risks for the oldest Americans might actually be rising.) Of course, cancer in children is uncommon and has little effect on death rates in the population as a whole. But even so, despite much discussion and investigation, reasons for the greater success in treating cancer in children have not been firmly established.
Breast cancer. Mortality from breast cancer among women fell slightly, from 32.9 in 1984 to 31.7 in 1992, then more rapidly to 27 in 1999, perhaps from a combination of better treatment, especially prevention of recurrences, and screening for early disease. These figures reflect a 10 percent decrease in breast cancer among women over age 50 and a 24 percent decrease among younger women, whose risk of breast cancer is much smaller.
Prostate cancer. A rise, then a fall, in mortality from cancer of the prostate (from 34.1 to 39.2 to 31.1) is not fully understood. The recent decrease may be a result of screening with the prostate-specific antigen test, but a decline is also seen in countries where such screening is not widely used.
Colorectal cancer. Reasons for a 23 percent drop in mortality from colorectal cancer (from 27.3 to 21.1) are not clear, but may include changes in incidence, screening, and treatment.
Cervical cancer. Cervical cancer continues to decline as a result of screening and lower incidence, possibly related to changes in lifestyle).
Stomach cancer. Mortality from stomach cancer is declining worldwide as a result of changes in incidence, though reasons for those changes are not known.
The good news is that mortality rates are mostly down. The bad news is that we might have done much better. In 1987, I argued that several issues needed urgent attention, and most of them still do. One was the need for a thorough, tough, fully independent review of where we are, how we got here, and where we are going. Despite considerable interest in that proposal, we still lack full, and independent review. Another urgent need was for a reorientation of personnel and projects to place more emphasis on prevention, which has hardly occurred at all. I also pointed out the need to encourage senior program managers to report progress more objectively. Yet we continue to be bombarded with stories of dramatic new treatments, each eclipsing the one before. Remember interferon, interleukin, angiostatin, and all the rest?
The most important point I tried to make in 1987 was that the emphasis in cancer research must be shifted from treatment to prevention. Recent mortality trends bear this out. Although treatment has produced some benefits, especially for cancer in children, its part in reducing deaths has been relatively minor. The biggest changes in cancer mortality come from changes in incidence, including the results of efforts to prevent the disease and to screen for it early on.
Since 1987, the proportion of cancer research funding devoted to prevention has grown only slightly. Meanwhile, Herculean efforts at developing and publicizing treatments have produced only modest clinical results. Imagine where we might be today if we had put that effort into research on prevention and screening, which have already produced benefits far out of proportion to our investment. We have treated cancer largely as a medical problem; we should now attack it more as a public health problem.
John C. Bailar III (firstname.lastname@example.org) is professor emeritus in the Department of Health Studies at the University of Chicago.