Mining the CDC's New Environmental Health Tracking Database for Stories
I recently wrote about the new National Environmental Public Health Tracking Network launched by the Centers for Disease Control. A fascinating resource for reporters, but molasses-slow at its debut.
I'm happy to report that after playing with the network again, the online database has recovered from its torpor, which might be explained by an estimated 10,000 hits upon its launch.
CDC epidemiologist Judith Qualters, who oversees the Environmental Health Tracking Branch, was kind enough to walk me through the database and talk about how journalists can mine it for stories in their communities.
First, a bit of background on the data that's in this network. The network, which costs about $23 million yearly to operate, receives data from 16 states and New York City. The data come from the state and local public health departments, water quality agencies, air quality agencies, pesticide monitoring programs, hospitals, poison control agencies, and other sources. The network also incorporates national data from cancer registries, the Environmental Protection Agency, and the U.S. Census.
The CDC hopes to add data from five more states in September, although they haven't been chosen yet, Qualters said.
Reporters can search for data on health conditions, including asthma, heart attacks, birth defects (coming soon), cancer, childhood lead poisoning, carbon monoxide poisoning and birth outcomes.
You can also examine data for environmental factors related to health, including lead levels in homes, air quality, and water quality.
"If you're interested in asthma, you can go on the site and look at county-level data for hospitalizations in your area," Qualters said. "We tried to provide a lot of information - what we know about preventing asthma, how we track asthma."
As an example, I looked at asthma hospitalizations by county, to find that Contra Costa County, a suburban county east of San Francisco, had the state's highest rate of asthma hospitalization, with 13.9 age-adjusted hospitalizations per 10,000 people. Rural Nevada County, in contrast, had 6.1 hospitalizations per 10,000 people. This page explains more about the indicators used and the sources of the data.
Why might that be? Is there more pollution in Contra Costa County? Do people in Contra Costa County receive less preventive asthma care than those in Nevada County, raising the possibility of an asthma attack so severe it requires a stay in the hospital? Or does Contra Costa County simply do a better job of tracking hospital admissions for asthma, causing their rates to appear higher?
Qualters hopes journalists will use the tracking network as a starting point for questions that will lead to stories, rather than a definitive answer. A look at childhood lead poisoning data in California, for example, reveals that in 2004, rural Tulare County had a higher percentage of children under three with lead poisoning than any other county.
"It drives you to think about what could be the factors influencing this," Qualters said. "Are more children ... being screened for lead poisoning? Why? Is the housing stock older?
She does caution reporters against drawing cause and effect conclusions solely from the surveillance data in the tracking network. "You shouldn't be jumping to conclusions. You don't want to look at Column A and Column B and say, 'aha!'''
Reporters also should not make comparisons between states because of differences in how states report their health and environmental data, she said.
"It helps to talk to someone who knows the data," she says. And to that end, she recommends talking - before you publish or broadcast - to the state-level officials who know the most about the health data they're reporting to the national network. Here's a list of the 16 states participating so far in the network.