The areas where patients are at higher risk of cancer mortality, just on observation, seem to overlap with areas where there are higher risks of future climate change impacts. So our goal was to see to what extent one could tease out these relationships. What we did was we evaluated the counties that represented the top 25% in the US of cancer survivors in the high cancer mortality areas. This is about a little bit more that half of all counties.

Our aim was to evaluate just what kind of future climate change risks those patients are going to be vulnerable to, and to identify and compare the vulnerability between those counties and those counties where there was not high cancer mortality.

What was the methodology, and what were the findings?

We built a risk model. We did a test validation approach with some cross-validation. First of all, we evaluated a number of different factors, about forty different factors, to see which climate factors might be especially prevalent for patients in high-risk counties. We identified six of them, predominantly, and we validated them. Those factors are extreme weather events like tornados, hurricanes, high precipitation; they are economic impacts, climate cost impacts, as well as air pollution.

So these six factors together, when you build a model you can identify that the counties in which patients are at high risk of cancer mortality are going to experience substantial impact from these factors in the future.

What are the clinical implications of these findings?

The implications are this really points the way for future evaluations of how future climate change risks in vulnerable cancer populations can impact patient adherence to treatment, patient tolerability, patient morbidity, and even potentially patient mortality. It sort of points the way to future evaluations, analyses, and hypotheses.

Is there anything else you would like to add?

I just want to thank my dad, who is a climate change advocate, for this idea in part, and credit him.