FDA's PREDICT Targeting System 'Working as Intended' Five Years After Implementation, GAO Says
The Food and Drug Administration’s PREDICT screening system is “working as intended,” resulting in higher targeting rates for unsafe food shipments in the five years since implementation, according to a recent Government Accountability Office report (here). Though some problems persist, including data collection issues, shipments with higher risk scores are being screened more often and FDA inspectors are finding more violations in connection with these shipments, according to data from the report.
Sign up for a free preview to unlock the rest of this article
If your job depends on informed compliance, you need International Trade Today. Delivered every business day and available any time online, only International Trade Today helps you stay current on the increasingly complex international trade regulatory environment.
According to GAO, FDA officials said they monitor PREDICT on an ongoing basis to ensure that PREDICT, which was implemented in 2011, is “focusing entry reviewers’ attention on items determined to be of higher risk.” In fiscal year 2014, for example, FDA examined 3.51% of entry lines with a risk score of 91 or higher on a scale of 1-100, finding violations 11.88% of the time. On the other hand, entry lines rated 30 or lower were examined around 1% of the time, with less than 1% of lines found in violation, the report said. “For fiscal years 2012 through 2014, entry lines that received higher PREDICT scores generally were more often selected for a field examination, for a label examination, or for sampling, and entry lines with higher PREDICT scores were more often found violative,” GAO said.
Ongoing implementation of the Food Safety Modernization Act will provide FDA with more data that the agency can use to further improve its PREDICT targeting, GAO said. Among other FSMA mandates, the Foreign Supplier Verification Program and Voluntary Qualified Importer Program will provide FDA data on importers that the agency will use to refine importer risk scores, the report said. Data from increased foreign food facility inspections “would be provided to PREDICT, and PREDICT would use that information in assigning risk scores to imports from that facility,” GAO said.
The report did find issues, pointing out that FDA collects some open source data for targeting purposes but has no procedures for doing so. GAO also criticized FDA’s uneven implementation of recommendations resulting from the agency’s own evaluation of PREDICT in 2013, including one unimplemented recommendation that said FDA should identify sources of delays associated with the sampling process.
In a response attached to the GAO report, FDA said it agrees that it needs formal standards for open source data collection, and said it would work “across the agency” to develop such a document. While agreeing in principle that it should complete recommendations of its suggested improvements to PREDICT, it said remaining recommendations do not have a direct impact on the targeting system’s effectiveness. “Establishing an implementation timeline is based on consideration of the feasibility and impact of implementation relative to available FDA resources and other FDA priorities such as implementation of” ACE and FSMA, FDA said.