New policy tools for studying SES disparities in colorectal cancer mortality and other outcomes

Researchers and policy planners tend to use a single indicator to analyze and draw conclusions about social inequalities.  However, health disparities are complex social phenomena.   A more rigorous approach is to compute results from various indices, and graphically compare them.  The Health Disparities Calculator (HD*Calc) tool is designed to facilitate this approach.  We use HD*Calc to extend previous analysis of mortality from colorectal cancer.  We test for changes in disparities in mortality by SES for men and women ages 50-74 from 1980-2010, the latest year for which data are available.  

Colorectal cancer (CRC) is the second leading cause of cancer death in the United States.  Survival rates are much higher for early (70-90%) than for late (20%) stage disease.  Since 1950, rates of CRC mortality have steadily dropped.  Screening and treatment for CRC improved and became more widespread starting about 1980.  We test the fundamental cause hypothesis that medical interventions will create patterns of disparities that favor higher SES patients. 

We used HD*Calc to extend the analysis of colorectal cancer mortality to 2011.  We linked areal socio-economic quintiles from the Census to CRC mortality rates from the National Death Index (NDI) to address two research questions.  First, did disparities in CRC mortality rates decrease between 1981 and 2011?   Second, did the socioeconomic gradient in CRC mortality rates shift such that disparities worsened for lower SES groups between 1981 and 2011?   Because different indicators may yield different results in terms of magnitude or even direction and, if viewed independently, can generate misleading findings, we examine CRC mortality data using two absolute summary measures, and two relative summary measures. 

We found an absolute decline in CRC mortality from 1980-2010.  A cross-over began in 1990:  the high SES quintile, which initially had higher CRC mortality rates in 1980, had lower mortality by 2010.  All indices showed significant differences in disparity, confirming a reduction, followed by an increase, over the period.  The concentration indices, which are sensitive to change in direction, showed that the change and crossover in disparities favored those with high SES, providing evidence for the fundamental cause hypothesis that medical interventions will favor high SES patients.   HD*Calc creates efficiencies by providing instantaneous statistical calculations and graphical representations.  This tool can be used with any time series data to measure disparities and, as such, contributes to measuring disparities in health in all policies.