Grant Number:	5R03CA115245-02 Interpret this number
Primary Investigator:	Chen, Jarvis
Organization:	Harvard School Of Public Health
Project Title:	Geosocial Health Disparities: Temporal Analyses
Fiscal Year:	2006

DESCRIPTION (provided by applicant): Our proposed project, prepared in response to PAR-04-020, "Small Grants for Behavior Research in Cancer Control," addresses critical temporal issues regarding the use of geocoding and census-derived area-based socioeconomic measures (ABSMs) to enhance US capacity to monitor socioeconomic inequalities in cancer incidence and mortality. In our prior Public Health Disparities Geocoding Project, we employed census data and health data centered around the 1990 census from Massachusetts (MA) and Rhode Island, so as to determine which area-based socioeconomic measures (ABSMs), at which level of geography (census block group, census tract, or ZIP Code) are most appropriate for monitoring US socioeconomic inequalities in health. Our key finding, based on analyses spanning from low birthweight to cancer incidence to mortality, was that the census tract poverty level most consistently detected expected socioeconomic gradients in health across a wide range of outcomes, and offered maximal geocoding and linkage to ABSMs (compared to block group and ZIP Code data). Additionally, this measure was readily interpretable, and could feasibly be used by cancer registries and other health agencies. One critical question left unanswered is the extent to which results are potentially biased, during the latter half of an intercensal period (e.g., 2005-2010), when the only available ABSM data are derived from the last decennial census and may be outdated. To evaluate the extent of this possible bias, we propose to analyze MA mortality and cancer incidence data from 1995-2001 in relation to both 1990 ("best available") and 2000 ("best possible") census-derived ABSMs. We will: (1) create and characterize the analytic data set, by geocoding the public health surveillance data to its 1990 and 2000 census tract geocodes; (2) generate 1990 and 2000 census tract poverty measures; (3) empirically investigate the temporal sensitivity of the analyses, by assessing the extent to differences exist between socioeconomic gradients detected using the 1990 vs 2000 census tract poverty measure; and (4) disseminate recommendations for use of ABSMs. By accomplishing this work, our project has the potential to improve use of ABSMs by US cancer registries and other health agencies for monitoring socioeconomic inequalities in health in the US at the national, state, and local level.





None