FDA Centralized Review Pilot Showed Promise, Data Quality Issues, Says Agency Report

The Food and Drug Administration's pilot program for a centralized entry review produced some useful insights into the agency's review process and industry's submission practices, FDA said in a report on the pilot (here). The agency began testing centralized entry review in 2013 to consider possible improvements to document processing (see 13061915). The pilot "quantified problems that slow down entry review, costing both importers and [Office of Regulatory Affairs (ORA)] time and money," the FDA said in the report. "Findings suggested opportunities for improvement and led to recommendations for a renewed emphasis on more frequent training for both industry and ORA entry reviewers, standardization of entry decisions,and improved communications."

The report didn't provide any concrete changes that result from the pilot, but said it will "inform future decisions of whether to pursue centralized entry review" and "help improve the process as it is currently performed." Personnel from the FDA's ORA worked on the 49-day pilot, during which they reviewed certain import entries arriving at one of 14 ports in six districts, the report said. Usually, the entry review is handled at the district-level, it said. The centralized review pilot was the result of a set of industry recommendations to the agency (see 12110702). Originally expected to last from Aug. 12 through Nov. 2, 2013, the test was cut short six weeks early due to last year's government shutdown (see 13102323).

The test exposed some issues with electronic data submission, said FDA. The submissions often lack the necessary information that would maximize use of the FDA's electronic system, and during the pilot "document review was necessary to verify that the transmitted information matched the entry documentation, or to determine if the documents contained additional information not submitted in the original electronic submission." FDA's "quality audits found that brokers/filers submitted an incorrect product code to FDA for 40 percent of the entries in both pilot and non-pilot ports," the agency said in the report. Also, "there was insufficient information supplied in the electronic entry to make an initial admissibility decision in 70 percent of non-pilot entries and 17 percent of pilot entries," the agency said. The pilot also showed minimal use of the FDA's Import Trade Auxiliary Communication System, with less than 1 percent of couriers in the pilot using ITACS, it said.

Pilot program reviewers sent a large number of detention requests, perhaps due to instructions to base admissibility "strictly on what was transmitted in the original electronic submission," the agency said. When compliance couldn't be verified on that basis, the reviewer would request further review, it said. The approach significantly increased compliance work and time for product clearance, and the instructions were eventually revised "to find a suitable equilibrium," said FDA. Communication was also problematic, the agency said. "Even with the extensive pre-planning, communication during the pilot proved to be challenging, and survey respondents noted a significant decrease in routine industry communications."

The entry review process and speed of review varied greatly among FDA reviewers in the pilot program, the agency said. "Differences included the placement of remarks, use of non-standardized verbiage, and steps taken prior to forwarding entries to the 'home' district," said FDA. Some of those issues could be addressed through national guidance and routine training, it said. Review time also differed based on the "complexity of admissibility requirements, accuracy and depth of information submitted with the electronic entry, auto lookup functionality, and the reviewer’s familiarity with a product," it said. Center for Drug Evaluation and Research and Center for Veterinary Medicine products took the longest time -- about 30 hours on average -- while Center for Biologics Evaluation and Research goods took the least time, at around 10 hours.

The beginning of the pilot resulted in a significant slowdown in reviews -- more than twice as long than at ports not involved -- but that did improve over time, said FDA. By around September, pilot ports began "generally reaching admissibility decisions faster than non-pilot ports," the agency said. Industry reported a difference once adjustments were made to the pilot, unanticipated problems like the entry review application outage were solved, and the process became more familiar to ORA personnel, said FDA. "Overall, industry liked expanded weekend coverage and noted better freight movements and, at some pilot ports, improved service."