Article Text

Download PDFPDF
41 Correlations between county-level incidence and mortality for commonly diagnosed cancers in the United States
  1. Adewole Adamson1,
  2. H Gilbert Welch2
  1. 1Dell Medical School at the University of Texas at Austin, Austin, USA
  2. 2Brigham and Women’s Hospital, Boston, USA


Importance Incidence and mortality are fundamental epidemiologic measures of cancer burden. Few studies, however, have examined individual cancers to determine how these fundamental measures correlate across place.

Objective To assess the relationship between cancer incidence (in situ and invasive) and mortality for the 12 most commonly diagnosed nonhematologic cancers in the United States.

Design Cross-sectional study in which the unit of analysis was the county.

Setting County-level incidence (2000-2016) and mortality (2002-2018) were obtained from the Surveillance, Epidemiology, and End Results (SEER) 21 database.

Participants Counties with at least 10 deaths from each cancer; range: 171 counties (thyroid cancer) – 726 counties (lung cancer).

Main Outcome(s) and Measure(s) The correlation between incidence and mortality. The 12 cancers were grouped into terciles based on their correlation coefficient (i.e. high, moderate, and low correlation). For each cancer we also determined the 10-year risk of death, both from the diagnosed cancer and from other causes. Analyses were conducted between June 2022 and September 2022.

Results County-level incidence and mortality were strongly correlated in some cancers, yet uncorrelated in others. The high correlation tercile included cancers of the lung (r=0.96), pancreas (r=0.76), liver (r=0.75) and stomach (r= 0.72). For patients with these cancers, the risk of death from the diagnosed cancer was more than 4 times the risk of death from other causes. The moderate correlation tercile included cancers of the colon (r=0.70), kidney (r=0.52), uterus (r=0.46), and bladder (r=0.43). There was little or no relationship between incidence and mortality for cancers in the low correlation tercile: prostate (r=0.21), melanoma (r=0.15), breast (r=0.06) and thyroid (r=-0.03). The risk of death from the diagnosed cancer for these patients was either lower or no different than their risk of death from other causes.

Conclusions and Relevance For some cancers in the United States, the fundamental epidemiologic measure of disease frequency – incidence – now has little relationship with the feared outcome of cancer death. While this was expected for melanoma, thyroid cancer, and prostate, the lack of correlation between breast cancer incidence and mortality was unexpected. We surmise that low correlation is most likely explained by differences in diagnostic practice leading to variable amounts of cancer overdiagnosis.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.