Portal de Periódicos da CAPES

Cancer Statistics, 2002: Progress or Cause for Concern?

2002; Wiley; Volume: 52; Issue: 1 Linguagem: Inglês

10.3322/canjclin.52.1.6

1542-4863

Richard Young,

Nutrition, Genetics, and Disease

Albert Einstein once said that "Everything should be made as simple as possible—but not simpler." When enormous amounts of information are packed into a brief report using both estimates and actual data related to endpoints scattered over a five-year span—opportunities for confusion and misinterpretation abound. This year's compilation of data on cancer incidence, mortality, and survival is no exception. The report provides ample data for optimists to proclaim continued progress in cancer control, as well as data for naysayers to be skeptical of these claims of progress. In truth, there is cause for both optimism and concern. Before plunging into the wealth of interesting data in the Cancer Statistics, 2002 report, let me begin with my view of the most important information in the report: the take home message, if you will. For the most recent year available (1998), overall cancer incidence and mortality rates declined once again by 1.1 percent each.1 This age-adjusted cancer decline began back in 1991 and has continued at a rate of approximately 1.1 percent per year since that time (see Figure 2, page 28). Given what we know about these estimates and the reasons for this change, steady overall progress is expected to continue into the foreseeable future. This progressive reduction in cancer incidence and mortality, now almost a decade in duration, is a triumph by any standard and a product of cancer prevention, early diagnosis, screening, and improved cancer treatment programs widely utilized throughout the country. But if the overall take home message is so positive, why then is there cause for concern? First, cancer continues to take a terrible toll on the American people. One in four deaths in the United States is due to cancer. And cancer ranks second only to heart disease as a cause of death in this country. Estimates indicate that sometime within the next 10 to 15 years, cancer will overtake cardiovascular disease and become the most common cause of death here in the United States. Second, our population is both growing in size and increasing in age. As a result of these changing national demographics, the actual number of deaths from cancer continues to increase. Indeed, there were about 5,000 more deaths from cancer in 1999 compared with 1998 as shown in Table 7 (page 37). The increase in cancer deaths as a percentage of all deaths results from the aging of our population coupled with the decline in cardiovascular deaths—a factor which allows more people to survive to ages where cancer risk increases. So are we winning or losing the war on cancer? To some extent, the answer depends on the body of data one chooses to defend one's conclusion. In my view, the most accurate and honest way to express outcomes is to use age-adjusted mortality rates, which provide a more precise way of comparing like groups of Americans when estimating true risk. Seen from this vantage point, we are clearly winning the war on cancer. If progress in cancer control continues this decline in age-adjusted cancer incidence and mortality, and even accelerates it—as many believe it will—then we will actually see both a decline in age-adjusted mortality as well as a decline in actual overall mortality from cancer. Although the overall decline in incidence and mortality is great news, there continue to be disturbing gaps in the progress we have made. Deaths from lung cancer in women continue to increase, although rates have begun to decline in women under 65. Our progress in other areas also lags behind the norm. African Americans have the highest incidence and mortality rates for cancer and a poorer probability of survival once diagnosed with cancer. Nevertheless, here too there is some good news. This year, African-American men showed the largest decline in both incidence and mortality of any subgroup in the analysis; no doubt in part because there was substantial room for improvement. Searching for the best and most accurate way to express this data is not a straightforward task. Incidence and mortality information undergo constant refinement in an effort to make the data more accurate and relevant. This is both necessary and appropriate but it can sometimes be distracting and even somewhat confusing. Two examples are seen in this year's analysis. To enhance accuracy of the lung cancer data, the ICD coding was improved to properly distinguish between primary lung tumors and those metastatic to the lung. As reasonable as this is, it can be confusing if the change is not highlighted and its effects are not explained. In this instance, the change in coding caused a decline of 1.6 percent in total lung cancer deaths while increasing the number of deaths from other cancers. Finally, an appropriate, but nevertheless problematic, change will be the standard year by which incidence and mortality data is adjusted for age. Beginning with the 1999 data, the age distribution of the US population for 2000 will replace the baseline of 1970, which has been used for over two decades. Changing to the year 2000 standard will increase the age-standardized incidence rate for all cancers combined by roughly 20 percent. The purpose of this shift is to provide a more meaningful picture of the average incidence and death rate given the aging US population. While the change is proper and undoubtedly more accurate, it will not make interpretation easier, nor is it likely to lessen the debate and controversy over the magnitude of the progress we have made.