Challenge

The environment in which one lives can affect the risk of developing cancer. Risks and exposures vary by region and by neighborhood. Epidemiologists concerned with describing the totality of environmental risks might be able to improve estimates by accounting for the risk associated with each place a person has lived: where they lived and how long they lived there. However, improving these estimates requires accurate information on residential histories.

Westat conducted a study for the National Cancer Institute (NCI)/SEER to assess the potential sources of residential addresses and the accuracy of the data they offer.

Solution

The study consisted of 2 parts. Part 1 encompassed a scan of more than 100 possible data sources, including social media and open access data providers, for residential and mobility information. Our analysts also looked at the information available from companies that sell data as a product. The identified sources were classified and ranked based on availability of the needed data and an assessment of feasibility of using the source for cancer research studies.

Part 2 evaluated the top-ranked sources identified in Part 1 for the accuracy and completeness of the residential history data. Westat collected a sample of residential histories from employees at NCI and the National Institute of Environmental Health Sciences. Study participants were asked about their residences dating back to birth, and defined as spending 6 or more months in one location. This information was compared against vendor-supplied data about the participants’ residential histories.

Results

The results from Part 1 indicated that the most readily available information on residential history is maintained by private “data-as-a-service” providers, such as business intelligence services, marketing companies, or credit reporting agencies.

Findings for Part 2 indicate that commercial data can be used to generate residential histories for cases in cancer registries. The resulting residential histories provide sufficiently reliable residential history and mobility information for assessing environmental or residential cancer risk. For further information on the results of the study, see Tools for using commercial sources of residential histories for cancer research.