Skip to main content

COVID-19 Resources

What people with cancer should know:

Guidance for cancer researchers:

Get the latest public health information from CDC:

Get the latest research information from NIH:

Grant Details

Grant Number: 5R03CA101362-02 Interpret this number
Primary Investigator: Elliott, Michael
Organization: University Of Pennsylvania
Project Title: Combining Data From the Nhis and Brfss
Fiscal Year: 2004


DESCRIPTION (provided by applicant) Cancer surveillance research requires accurate estimates of risk factors at the small area level. These risk factors are often obtained from surveys such as the National Health Interview Survey (NHIS), or the Behavioral Risk Factors Surveillance Survey (BRFSS). Unfortunately, no one population-based survey provides ideal prevalence estimates of such risk factors. One strategy is to combine information from multiple surveys, using the complementary strengths of one survey to compensate for the weakness of the other. The NHIS is a nationally representative, face-to-face survey with a high response rate; however, it cannot produce state or sub-state estimates of risk factor prevalence because sample sizes are too small. The BRFSS is a state-level telephone survey that excluded non-telephone households and has a lower response rate, but does provide reasonable sample sizes in all states and many countries. Several methods are available for constructing small-area estimators that combine information from both the NHIS and the BRFSS, including direct estimators, estimators under hierarchical Bayes models and model-assisted estimators. The project will focus on the latter, which includes constructing generalized regression (GREG) estimators using BRFSS data and control totals derived from the NHIS dataset, re-weighting schemes to reduce distance measures between sampled elements in BRFSS and NHIS datasets, and the use of existing and newly developed techniques to smooth the resulting estimates using weighting and small-area smoothing techniques. This approach is particularly promising because, in contrast to other proposed techniques, it can be applied using existing publicly available data.



Back to Top