The Government Accountability Office has a new report out on the Food and Drug Administration: Apparently the agency doesn’t have enough information about the ingredients in dietary supplements to decide if they’re safe.
..a lack of information is one of the most significant factors that limit the FDAâ€™s ability to identify and properly act on safety concerns regarding dietary supplements and foods with added dietary ingredients.
I think there’s an important point to draw from this report. We talk a lot about the FDA’s budget and staffing woes. And the agency is certainly underfunded â€” you can’t expect the FDA to properly regulate foreign food producers when it has less than a half-dozen foreign offices.
But I think we’re also learning that the FDA does a poor job gathering and analyzing data.
We saw this during the peanut recall last month: The FDA didn’t even know Peanut Corporation of America had a facilityÂ in Texas, nor did it know the company’s products had repeatedly tested positive for salmonella.
We saw it during last year’s salmonella outbreak, when the FDA struggled to trace tainted produce back to farms in Mexico. And we’re seeing it again now with GAO’s report on dietary supplements.
The FDA has to regulate two huge global industries. It can’t possibly hope to inspect every food/drug producer on an annual basis â€”Â not unless its staff grows exponentially. But if the agency is going to build a risk-based inspection system, it needs to do a much better job with data collection.
Google, for example, can identify a flu outbreak up to two weeks before the Centers for Disease Control identifies it. How? By aggregating search data: When lots of people in the same city or state start searching for “flu” and other related keywords, there’s an outbreak starting.
That’s not to say ensuring food and drug safety is as simple as analyzing search keywords. Obviously it’s not. But the FDA needs to think about its own creative ways to analyze data. Are certain food producers riskier than others â€” based on what they produce, where they’re located, how they’ve performed in the past? Are clusters of prescription drug users reporting similar symptoms?
The Internet makes it incredibly easy to gather and analyze huge volumes of data â€” but only if agencies decide what they want to analyze.
May 21st, 2009 at 2:43 pm
Gregg: well done. gathering the right data is key. And not just aggregating Google data, but “live” search (e.g., Scoopler for example).
I happen to think that we can combine aggregating data from sources like Google with supply-chain information such as barcodes to manage the risk.
The fact that the FDA includes barcodes in their outreach makes it possible for consumers to scan barcodes and search the FDA site (see http://www.agorasys.com/products.html).