# Materials and Methods

## Sources of Data

Data on all deaths among Americans during 1950-94 with cancer as the
underlying cause according to age, sex, and race were provided for
the 3,053 contiguous U.S. counties by the National Center for Health
Statistics. Annual county-, age-, sex-, and race-specific midyear
population estimates based on data from the Bureau of the Census
were aggregated over the 25-year period 1970-94 and the 20-year
period 1950-69 to form the person-years at risk. Data for the recent
time period were produced at the county level for whites and at the
state economic area (SEA) level for blacks and whites. The 506
contiguous SEAs are individual counties or groups of counties that
are relatively homogeneous with respect to various demographic,
economic, and cultural factors; they do not cross state lines. For
1950-69, when detailed county-level population estimates were not
available specifically for blacks, maps are produced for whites
only. Data for all 50 states and the District of Columbia are
presented. Since mortality data were available only at the state
level for Alaska and Hawaii, each was considered a single unit for
county and SEA purposes, resulting in data for 3,055 counties and
508 SEAs. Data for Alaska and Hawaii are available only beginning in
1959 and 1960, respectively, when they became states. Appendix Map 1 of the United States displays the state, SEA, and county boundaries.
Appendix Table 1 lists the state economic areas by state, and
Appendix Table 2 lists the counties within state economic areas by
state. A total of more than 40 cancers (including all forms
combined) were considered, as shown in Appendix Table 3 (which also indicates the sixth to ninth revisions of the International
Classification of Diseases^{7-10}
[ICD] codes used for each form of cancer).

## Calculation of age-adjusted cancer mortality rates

For each form of cancer, the age-adjusted (direct method, 1970 U.S. population standard—see Appendix Table 4) mortality rate R per 100,000 person-years was calculated by race, sex, and geographic area for each of the time periods 1970–94 and 1950–69, as follows:

^{3,11,12}R = 100,000*SUM(w

_{i}r_{i}) = 100,000*SUM(w_{i}d_{i}/n_{i})where

i = the 18 age groups 0–4, 5–9, ..., 85+,

w_{i}= the proportion of the standard population in age group i,

r_{i}= the age-specific rate di/ni,

d_{i}= the number of deaths in age group i, and

n_{i}= the person-years in age group iThe binomial approximation to the variance of the age-adjusted rate was calculated as:

var(R) = 100,000

^{2}*SUM [w_{i}^{2}r_{i}(1–r_{i})/n_{i}]The 95 percent confidence limits of R were calculated from the square root of the variance as:

R ±1.96 [(var(R)]

^{½}A negative lower confidence limit was replaced by zero. For an area with zero deaths, the rate R was zero and the var(R) was estimated using the national rate. An areaspecific age-adjusted rate was deemed significantly different statistically from the U.S. age-adjusted rate if their confidence limits did not overlap. Detailed area-specific data are not presented in this Atlas but are available from the NCI and from the NCI Atlas Static Web site.

Expected numbers of deaths from cancer for each geographic area by race and sex were the sums over age groups of the corresponding national age-specific rates times the agespecific person-years for each area by race and sex.

Male/female and black/white rate ratios (R

_{1}/R_{2}) were calculated using the nationalage adjusted rates rounded to two digits after the decimal point. The 95 percent confidence limits on the rate ratios are not presented but may be calculated as:^{13}exp[ ln(R

_{1}/R_{2}) ±1.96 (1/D_{1}+1/D_{2})^{½}]where

exp[x] = e

^{x},

ln[x] = the natural logarithm of x, and

D_{1},D_{2 }= the deaths associated with R_{1}/R_{2}respectivelyMaximum likelihood estimates of the relative risk standard deviations (RRSDs) and their standard errors (SEs) were calculated under the assumption of a mixed effects model. The RRSDs provide a measure of the standard deviation of the underlying areaspecific relative risks and serve as quantitative indices to compare geographic variation across maps.

^{14}

## Map production

Maps displaying cancer mortality rates by county and SEA were produced using Atlas Pro and Atlas GIS for Windows software (Strategic Mapping, Inc., now part of ESRI) on a Pentium personal computer. For ease of comparisons and readability, the size of Alaska was reduced and, for the SEA maps, the size of the District of Columbia was increased and separated from adjacent areas. Prior to mapping, rates for an area based on sparse data were deemed unstable if (a) the observed number of deaths was less than 6; (b) the observed number of deaths was less than 12 and the rate was not significantly different statistically from the U.S. rate; or (c) the expected number of deaths was less than 6 and the rate was not significantly different statistically from the U.S. rate. The stable rates were then ranked and partitioned into 10 deciles. The legend for each map portrays the national rate and the range of rates for each decile. Rates per 100,000 person-years are generally presented with 2 digits to the right of the decimal, although for some of the rarer cancers 3 digits are shown. In each map and legend, five shades of red (deciles at the median and above) and five shades of blue (deciles below the median) are used, with the intensity of color reflecting the ranked distance from the median. Gray is used for areas with sparse data (i.e., unstable rates as defined above).

------------------

- Pickle LW, Mason TJ, Howard N, Hoover R, Fraumeni JF Jr. Atlas of U.S. cancer mortality among whites: 1950-1980. Washington, DC: U.S. Gov. Printing Office; 1987. DHHS Publ. No. (NIH) 87-2900.

- World Health Organization. Manual of the international statistical classification of diseases, injuries, and causes of death. Vol. 1, 6th rev. Geneva: WHO; 1948.
- World Health Organization. Manual of the international statistical classification of diseases, injuries, and causes of death. Vol. 1, 7th rev. Geneva: WHO; 1957.
- Public Health Service. International statistical classification of diseases, adapted for use in the United States. Vol. 1, 8th rev. Washington, DC: U.S. Gov. Printing Office; 1967. PHS Publ. No. 1693.
- World Health Organization. Manual of the international statistical classification of diseases, injuries, and causes of death. Vol. 1, 9th rev. Geneva: WHO; 1977.

- Chiang CL. Standard error of the age-adjusted death rate. Vital Stat Selected Rep 1961;47:275-85.
- Boyle P, Parkin DM. Statistical methods for registries. In: Jensen OM, Parkin DM, MacLennan R, et al., editors. Cancer registration: Principles and methods. Lyon, France: International Agency for Research on Cancer; 1991. p. 126-58. IARC Scientific Publ. No. 95.
- Rothman KJ. Modern epidemiology. Boston: Little, Brown and Company; 1986.
- Pennello GA, Devesa SS, Gail MH. Using a mixed model to estimate geographic variation in cancer rates. Biometrics 1999; 55:774-81.