JAMA Study Discovers Regional Trends in Cancer-Related Mortality

An analysis of cancer death records across the country, by researchers at the University of Washington, has identified clusters of counties that had a particularly high rate of mortality depending on the cancer type.

One of the leading causes of death in the United States, cancer also places a significant economic burden on the country. An annual report by the CDC, the National Cancer Institute, and the American Cancer Society in early 2016 showed that cancer-related deaths have been on the decline. One can attribute the decline to a combination of factors:

• Increased awareness of, and access to, preventive screening
• An improved understanding of the molecular pathology of cancer
• Better, more targeted drugs
• A personalized approach to treatment

The current report in JAMA was based on the hypothesis that information derived at the county-level can help researchers gain more insight on the impact of public health programs, which are designed and implemented locally. The authors write that such data can help clinicians understand the needs of their local community and also help identify cancer hotspots, which can lead to the primary cause of spikes in incidence or death.

The study included death records from the National Center for Health Statistics (NCHS) and population counts from the Census Bureau, the NCHS, and the Human Mortality Database from 1980 to 2014 were used. The study focused on 29 cancers—including stomach, colon, liver, lung, melanoma, brain, kidney, Hodgkin and non-Hodgkin lymphoma, and multiple myeloma—to estimate county-level mortality rates. The primary outcomes were age-adjusted county, year, sex, and cancer type.

The following table shows the recorded death rates across the country during the period between 1980 and 2014, based on cancer type:

Cancer Type

Death Rate

Tracheal, bronchus, and lung

5,656,423

Colon and rectum

2,484,476

Breast

1,573,593

Pancreatic cancer

1,157,878

Uterine cancer

209,314

Kidney cancer

421,628

Liver cancer

487,518

Testicular cancer

13,927

Non-Hodgkin lymphoma

829,396

Overall, a 20.1% decrease in the cancer death rate was observed for the study period. However, the authors noted that the death rate varied substantially across counties. In 1980, mortality ranged from 130.6 per 100,000 in Summit County, Colorado, to 386.9 in North Slope Borough, Alaska. In 2014, the numbers ranged from 70.7 per 100,000 for Summit County, Colorado, to 503.1 per 100,000 for Union County, Florida.

Further, the authors observed a cancer type—dependent variation in the death rate, with specific clusters of counties having a high mortality rate. Clusters of breast cancer were documented in southern states, along the Mississippi river; incidence of liver cancer was high along the Texas–Mexico border; and kidney cancer rates spiked in several states, including North and South Dakota and counties in West Virginia, Ohio, Indiana, Louisiana, Oklahoma, Texas, Alaska, and Illinois.

The authors emphasize that curating county-level data highlights local patterns that would otherwise be lost in state or national numbers. This data, they write, can inform decisions on preventive efforts, access to care, and appropriate treatment; they can also help identify worsening incidence, inadequate access to quality treatment, or other etiological factors involved.

Reference

Mokdad AH, Dwyer-Lindgren L, Fitzmaurice C, et al. Trends and patterns of disparities in cancer mortality among US counties, 1980-2014 [published online January 24, 2017]. JAMA. 2017;317(4):388-406. doi: 10.1001/jama.2016.20324.