In 1981, Doll and Peto published a well-known paper estimating that roughly 75-80% of cancer mortality was preventable. Forty years later, cancer mortality has declined some overall – but we still see vast disparities. Some of these disparities have gotten worse over time.
Rate of Cancer Deaths in the United States, 2014-2018, All Types of Cancer, All Ages, All Races and Ethnicities, Male and Female. Courtesy of the U.S. Cancer Statistics Working Group. U.S. Cancer Statistics Data Visualizations Tool, based on 2020 submission data (1999-2018): U.S. Department of Health and Human Services, Centers for Disease Control and Prevention and National Cancer Institute; www.cdc.gov/cancer/dataviz, released in June 2021.
With the Biden administration’s reignited Cancer Moonshot initiative, combined with a renewed emphasis on equity, we might wonder: How much lower could cancer mortality drop if we addressed social drivers of health (SDoH)?
Social Drivers of Cancer Mortality
Mokdad and colleagues, in their 2017 article on mortality trends over time, mentioned 4 drivers of cancer mortality disparities:
- Socioeconomic status: people with higher income and education levels are more likely to seek care and be able to access it
- Access to health care: many people in the United States are under- or uninsured
- Quality of care: whether patients are seen and treated promptly and followed appropriately
- Lifestyle: smoking, obesity, and lack of physical activity
Of course, there are strong feedback mechanisms among these drivers. For example, people with lower incomes are more likely to smoke cigarettes. Similarly, socioeconomic status is a driver of access to quality healthcare.
Recent Evidence
Two recent studies provide more context on the social drivers of cancer mortality. Scott et al used a large dataset of 99 variables drawn from publicly available data to create a multifactor index — the burden index. They were able to explain 23% of the variance in overall cancer death rates and 34% of the variance in the county-level cancer death rate using principal components analysis.
Pinheiro and colleagues used data from a longitudinal study with a 10-year follow-up. They considered 8 SDoH, retaining 6 in the final analysis. Low education, low income, area-level poverty, poor public-health infrastructure, lack of health insurance, and social isolation were all associated with cancer mortality. In multivariable models, individuals with more SDoH factors had higher mortality, even after adjusting for risk factors like smoking.
Composite Indices
We previously discussed associations between composite SDoH indices and life expectancy disparities. With strong evidence of the social drivers of cancer mortality, how much variation can publicly available composite indices explain?
To recap: the Area Deprivation Index (ADI) includes 17 measures in 4 categories. The Social Deprivation Index (SDI) is based on 7 variables. The Social Vulnerability Index (SVI) incorporates 15 variables. All of them capture aspects of income, education, employment, and housing.
For this analysis, we assigned county-level cancer mortality rates (2015-19, age-adjusted) to 72,700 Census tracts across the entire US. We then linked these rates with the above composite indices. We ran basic linear regressions (Y = cancer mortality, X = each SDoH index) and interpreted the adjusted R2 as the percent variance explained.
Interestingly, the SDI and SVI perform nearly the same, explaining only 2% of the variance in local cancer mortality. The ADI does better, explaining 31%. Notably, incidence rates from 2018 also explain 31% of the variation. Combined, the ADI, SDI, SVI, and incidence rates explain 50% of the variation. The linked data are available for readers to explore at their leisure.
Now, these SDoH indices were not developed to predict cancer mortality, unlike Scott et al’s burden index. Yet, wouldn’t we hope that a composite index of SDoH would have some ability to explain disparities in cancer mortality, given the evidence that rates are (at least partly) driven by social factors?
Next Steps
Three currently available composite measures of SDoH don’t explain much variation in area-level cancer mortality. New approaches are needed to meet this moment.
Interest in SDoH has never been higher, nor has the level of investment available in addressing cancer mortality. Forty years after Doll and Peto’s landmark analysis, perhaps the time has finally come to stop guessing at the disparities and start addressing them. Since we cannot improve what we don’t measure, let’s put some efforts into proper measurement.
Lisa M. Lines