Researcher Tries to Accurately Measure Socioeconomic Differences and Health Disparities
Perhaps never before has the discussion of health disparities among Americans been so prominent on the national agenda than it is today. Yet there is little information on how differences in socioeconomic position affect health. Nancy Krieger, associate professor in the Department of Health and Social Behavior, is working to fill those gaps through a special four-year project.
"Most public health databases maintained by health departments give little or no socioeconomic data, so you can't determine if socioeconomic inequalities in health are getting better or worse," said Krieger. Health department records, however, do have addresses, permitting records to be linked to data about the neighborhoods in which people live. The basic aim of the project, said Krieger, is to come up with measures of socioeconomic position of residents using data from the census. The researchers then link the data to geographical locations that can be readily used by any health department to track ongoing trends.
For more than two years, Krieger and her colleagues have worked with approximately 1 million records from databases of the Massachusetts and Rhode Island Departments of Public Health as well as US census data. They want to develop rigorous methods to survey socioeconomic inequalities in health that will permit tracking the extent to which people living in more affluent areas have lower disease and mortality rates, as some studies have suggested.
Currently there is no agreement about how to best define a geographical area in terms of socioeconomic position or about which area-based measures of socioeconomic position are most informative, especially across multiple kinds of health outcomes.
Many researchers have used postal codes as the boundaries to define an area, said Krieger, but these codes may not be very scientifically informative. They were created to ease the work of postal carriers and do not necessarily represent a group of people similar in financial or social position.
Other scientists have used census-defined areas, measured at several levels of geography. The smallest is a block, but much of the socioeconomic data at the block level is censored to guard the privacy of the residents and therefore unusable.
What could be more useful, said Krieger, is a census block group--made of about 1,000 people--or a census tract, which includes at most nine block groups and on average contains about 4,000 people.
"What our project is trying to study is which level of geography makes the most sense and most accurately captures what is going on: the zip code, block group, or census tract," said Krieger. "We are evaluating these area-based socioeconomic measures at each level of geography in relation to diverse outcomes: mortality data, birth data, and population-based registry data for cancer, tuberculosis, sexually transmitted disease, and domestic violence."
The work is painstaking. Krieger and her colleagues are systematically going through diverse socioeconomic variables measuring income, poverty, social class and education for every health outcome at each geographic level in the 1 million records. By the end of 2001, Krieger hopes to have completed analyses demonstrating which of the socioeconomic variables produce the most informative picture of health and socioeconomic position.
Krieger said that she was motivated to undertake the project by a desire to learn more about class inequalities and health, including their contribution to racial and ethnic disparities in health. She said she felt there had to be a well-defined way to establish the information.
"People don't live randomly in neighborhoods," said Krieger. "They are incredibly economically segregated. That's powerful information."
Nancy Krieger
is using geography and census data to determine which definition of location
is most telling about socioeconomic position and health.
