The displays the age, gender, and lifetime smoking history of lung cancer cases from the KCR. We identified a total of 54,532 lung cancer cases from 1995 to 2007. Approximately 59% were men, and two-thirds of cases were 60–79 years of age. As mentioned previously, 20.3% of cases had no smoking data, but these were assumed to be lifetime smokers for our primary analysis. The also displays statewide population characteristics from the BRFSS and the U.S. Census 2000 for comparison.
Characteristics of lung cancer cases in Kentucky (1995–2007) compared with the general Kentucky population (2000)a
Analysis in SaTScan indicated three high-rate lung cancer clusters with significant p-values (p<0.05), each comprising multiple counties. displays a map of all three high-rate clusters and their circular scan windows, as well as crude lung cancer incidence rates. Cluster 1 (p<0.01) is located in southeastern, or Appalachian, Kentucky. It includes Knott, Perry, Letcher, Floyd, Breathitt, Magoffin, Leslie, Pike, Johnson, Harlan, Martin, and Wolfe counties, with a total of 5,040 cases during the study period. The total population of the counties in this cluster was a little more than 300,000 in 2000. For residents of this area, the RR of developing lung cancer was 1.21 compared with the rest of the state.
Unadjusted lung cancer incidence rates in Kentucky by county (1995–2007), with high-rate clustersa identified by spatial scan statistic
Cluster 2 (p<0.01), comprising 1,428 cases, is also in southeastern Kentucky. Located near Cluster 1, it includes Whitley, Knox, and McCreary counties and had a combined population of 84,740 in 2000. Lung cancer incidence for this cluster (RR=1.17) was very similar to that of Cluster 1.
Cluster 3 (p=0.01), with a total of 16,444 cases during the 13-year study period, is located in north-central Kentucky. It includes 15 counties: Carroll, Trimble, Henry, Gallatin, Owen, Oldham, Grant, Boone, Shelby, Franklin, Kenton, Scott, Pendleton, Jefferson, and Campbell. This region, home to about 1.27 million residents in 2000, includes several counties in the Louisville, Kentucky, and Cincinnati, Ohio, metropolitan areas, as well as the capital of Kentucky (Frankfort). Within this region, the RR of developing lung cancer was 1.04 when compared with the rest of Kentucky.
All three alternative analyses produced results similar to those from the primary analysis, with significant or near-significant high-rate clusters in the same regions, though sometimes including fewer counties. In the Appalachian region, Breathitt, Floyd, Knott, Letcher, and Perry counties were part of a significant cluster in every analysis. These are five of the six counties wholly within the boundaries of the circular scan window for Cluster 1 ().
We did not identify significant clusters of high lung cancer incidence centered in south-central or central Kentucky, the regions associated with high radon levels,12
or in McCracken County, where the Paducah Gaseous Diffusion Plant is located.
displays coal production (quintiles) per square mile by county, as well as the three lung cancer clusters we identified in the primary analysis. All of the counties in Clusters 1 and 2 produced coal, and most of them were in the two highest quintiles. Of the 15 counties in these clusters, only two were not in the top two quintiles of coal production. Similarly, 13 of 17 southeastern Kentucky counties with high coal production (top two quintiles) were included in Cluster 1 or Cluster 2. Eight of nine counties with very high coal production (top quintile) were in Cluster 1. None of the coal-mining counties in western Kentucky were part of a lung cancer cluster, and none of the counties in Cluster 3 mined any coal.
Coal production among Kentucky counties (1983–2007), with high-rate clustersa identified by spatial scan statistic
Relative to the rest of the state, access to public water service was low in the Appalachian region, and ranged from 10.4% to 61.9% among counties in Cluster 1, with <50% of households in nine of 12 counties having access to public water service. In Cluster 2, 59.9%–77.9% of households had access to public water. All counties in the Western Coal Field had public water service for at least 75% of households, with most counties offering public water service for at least 85% of households.