A few days ago I received a plain white envelope in the mail from Abbott, the infant formula manufacturer. Inside was a form letter, telling me that the company was voluntarily recalling some of its cans of Similac powdered formula because they may contain beetle parts. The news was jarring; while I had just started my six-month-old son on solids, I wasn’t planning on including beetle bits on the menu.
The letter was dated September 27, five days after the company had announced the recall; while there were earlier news reports, somehow I’d missed them.
Coincidentally, I’d also just seen that that U.S. Food and Drug Administration (FDA), had launched a new data set on recalls, which we had called for earlier this year and which the agency later confirmed plans to do. Sure enough, this new data set, which is available in programmer-friendly XML format, contains information about the Similac recall.
I also found the April report when McNeil Consumer Healthcare voluntarily recalled some batches of infant Tylenol and children’s Benadryl, among other drugs, which at the time set me on a medicine cabinet emptying frenzy to find all the half empty bottles we used with my three kids.
The FDA’s release of these data in this format ultimately should be a boon to parents who would like to get this information in real time, once some enterprising programmers (hello Sunlight Labs!) start mashing it into some useful interfaces that makes it easier to search. It should also help public health, consumer advocates, journalists and others more easily analyze recall data to find out if there are patterns with particular companies or outbreaks. This can only help in a time that has seen recent massive recalls because of salmonella infestation in peanut butter and eggs.
But the FDA is the first to admit that the data are still lacking. The agency does not have the power to require companies doing recalls of drugs, medical devices, or food to provide a standard list of specific information. This includes such basics as an estimate of how many items are affected, the reason for the recall, and the geographic distribution of the product. (While some companies provide this information voluntarily, they’re not required to do so.) So last May, the agency issued a draft proposal–one of 21 transparency enhancing changes suggested by the agency–that it should seek such authority. Getting standardized information from companies would enhance the usefulness of this tool tremendously.
Last week the FDA also released another new data set, this one a database where the public can explore medical device inspections.However, the data only include 25 percent of inspections, and leaves out clinical trial inspections “because information about new medical device development is confidential.”
The trade journal Medical Device Daily (subscription required) calls the data “all bones and no meat,” because a user would still need to file a Freedom of Information Act (FOIA) request to get copies of actual inspection forms that give more detail.
For example, a search of St. Jude Medical, the company that we investigated in our award-winning “Heart of the Matter,” shows a list of five records; clicking on a record, however, shows us that the company was asked to take voluntary action to fix something, without saying why or what. These are data most useful to experts or investigative reporters with plenty of time and resources to dig in, not for ordinary consumers.
Incidentally, the new FDA recall databases appear to be available at Data.gov, the government’s data clearing house, but the medical device inspection data are not. All are now listed on the Sunlight Foundation’s National Data Catalog, here and here.
Next time there’s a recall of food or drugs affecting my kids I don’t want to wait to get an envelope in the mail. I want to get the scoop as soon as the government has it, with all the relevant details. I can’t act on what I don’t know.