A drug recall often begins with a company determining that it has a product in the market place that is an actual or potential
health hazard or otherwise is in violation of the US Code of Federal Regulations (CFR). To protect the public health, a company can initiate a voluntary recall at any time. Firms also may initiate a voluntary
recall in response to a request by the US Food and Drug Administration that is triggered by a significant violation or concerns
of a potential health hazard. Recalls may also occur due to anticipation of hazards violation or a state agency finding. The
current FDA recall program, described in Chapter 7 of the FDA Regulatory Procedures Manual (RPM), and in other related procedures such as the Investigations Operations Manual (IOM), implements 21 CFR Part 7, Subpart C Recalls. FDA's Guidance for Industry: Product Recalls, Including Removals and Corrections also defines the agency's current thinking on recall programs and recall management (1–4).
When a company believes that one of its marketed drug products poses a potential or confirmed health hazard to the public,
the company must notify the FDA recall coordinator for its district (there are 19 districts in the United States) and collect
information for the coordinator. FDA's product recall guidance is specific as to the information, including the reason for
the recall, to be submitted: "Please explain how the problem occurred and the date it occurred. Explain how the problem was
discovered and the date discovered" (4).
The guidance also requests information about the root cause of the problem: "We recommend that you provide this information
to your local District Recall Coordinator once the root cause has been established. It is important to establish the root
cause of the problem so that appropriate preventive measures can be taken" (4).
If the investigation is inconclusive in determining a root cause, that information should be reported to the district coordinator.
The district coordinator will gather the information, including the recall strategy provided by the firm. FDA's Center for
Drug Evaluation and Research (CDER) then assesses the potential health hazard and scope of the problem. The goal of the firm
and FDA is to remove the hazardous or violative product from the market as rapidly as possible to protect the public health.
For this reason, all efforts are directed toward learning as much as possible (and as soon as possible) about the product
and the potential danger to take appropriate action to resolve the situation.
To date, this recall program has focused predominantly on individual recall events. FDA's efforts to look for and examine
drug-recall trends during the past several years have been limited in their depth of data mining and analysis. There is a
need for a "big picture" overview of major recall categories, accompanied by an analysis of the associated fundamental root
causes. To gain this overview, FDA initiated the Recall Root Cause Research (RRCR) project in September 2007.
The recall root cause project
Background.
The RRCR project began in the Division of Manufacturing and Product Quality (DMPQ), which falls within CDER's Office of
Compliance (OC). Product quality scientists in OC and FDA's Office of Pharmaceutical Science were interviewed to set priority
areas of greatest interest for the research project. Identified recall categories include, but are not limited to, subpotency,
dissolution, and microbial contamination. The RRCR project is meant to provide a robust assessment of available drug-product
recall information using knowledge of pharmaceutical manufacturing and applied statistics to identify important recall patterns
and trends. The four questions to be answered by this research project are:
- What recall categories have the greatest impact on patients relative to safety, efficacy, and availability?
- What can we learn by looking at recalls in the aggregate?
- What are the manufacturing root causes of the drug recalls?
- How can FDA systems be improved to better identify emerging trends?
Classes of recalls
|
The project will provide a retrospective assessment of drug recalls that reviews information from related available data sources.
Specifically, the project will find and review text information, archived data, metadata for Class I and Class II recalls
(see sidebar, "Classes of recalls"), CGMP issues such as labeling, and selected focus areas using FDA records. The timeframe
and sampling frame are defined for the information to be assessed. Independent factors extracted from the archived records
are defined as well, and data sets have been developed and populated. These sets are summarized to establish the scope and
range of the recalls; the summaries will be used to explore potential cause-and-effect relationships.
Benefits and goals.
The integral relationship between designing quality into a drug product and process, as well as future defect prevention,
has not always been evident in the pharmaceutical development process. Understanding true root causes can lead to a mechanistic
understanding of how manufacturing flaws and quality defects can be prevented during the design and process qualification
stages, and through continual life-cycle improvements. Anticipating, monitoring, and detecting signals of process drift will
help maintain a state of control, reduce variation, and improve process performance. Rational risk assessment derived from
root-cause analysis can be applied to identify and address areas of potential manufacturing vulnerability that should receive
special attention to prevent quality problems.
In addition to its retrospective analysis of recall root causes, the RRCR project will assess FDA's overall system for capturing
certain drug-recall information. Currently, the information available is used to administratively manage recalls; it is less
used to understand the recall process itself or to improve it. This project will address working on the process so that CDER
staff are able to generate more accurate root-cause information and function more efficiently. The end result will refocus
part of the FDA recall program on taking a more systematic approach to assist in defect prevention action in addition to taking
action on individual recalls.
Ultimately, as a result of the RRCR project, it is hoped that problem solvers will recognize that enhanced collection and
analysis of recall root cause information prompts the need for further investigation and attention to the origins of product
variability.
As the current FDA recall program transitions to one that augments the intense review of one recall at a time to one that
incorporates robust trending mechanisms, it will be more common for FDA to identify meaningful patterns. The RRCR project,
by contrast, will provide the time and resources needed to see the big picture of similar recall cases and their causes.
Currently, recalls are summarized by class and by specific defect category over a defined time period. The RRCR project's
thorough datamining of various related cases will provide insights into current practices that lead to variability. It is
hoped that the study findings will help the industry focus on common recall problems and solutions because FDA will be able
to perceive and share with industry major trends that individual firms cannot. In addition, the RRCR project's description
and quantification of complex issues can help to provide feedback to investigators, compliance staff, reviewers, and industry
for further attention and action. For example, many recalls are initiated for failure of dissolution while on a stability
program. The project will examine, What can be learned from these recalls that can be summarized and will be useful to both
FDA and industry?
It is also hoped that the project will find alternative ways of looking at recall data and generate new ideas and hypotheses
for further assessment. Information gathered should illustrate how process design and knowledge management programs can be
enhanced to anticipate potential problems in production.
Summary
The operational goals of the RRCR project are to review and summarize classes of recalls and other focus topics to determine
potential root causes. These root-cause findings can then be used to identify major trends. The information and understanding
gained through the project can be used not only at FDA to improve programs, but also to support industrial process development
and qualification activities and to assist investigations into product quality problems. Ultimately, application of the RRCR
project's enhanced practical knowledge of root causes can lead to reduced risks to the public health.
The RRCR project will bring attention to recalls from an analytical viewpoint. Qualitative and quantitative summaries will
present the critical information in different formats yielding new insights into the root causes. Trends and patterns will
be identified and evaluated to assist the agency and industry.
Project leaders expect that recommendations from the project will lead to suggestions for improving drug manufacturing and
product quality. These recommendations will be communicated internally at FDA as well as to industry, through journal articles
and public presentations at conferences (5).
For more information on CGMPs, please contact CDER's Division of Manufacturing and Product Quality. See the following website
for more information, including subject contacts: http://www.fda.gov/Drugs/DevelopmentApprovalProcess/Manufacturing/default.htm.
NOTE: Minor editorial changes per the authors have been made to the online version of this article after original publication
(as of Nov. 13, 2009).
Lynn Torbeck* is a principal statistician at Torbeck and Associates and a member of Pharmaceutical Technology's editorial advisory board, tel. 847.424.1314, Lynn@Torbeck.org
. Richard Friedman is the division director of the Division of Manufacturing and Product Quality in the Office of Compliance at the US Food
and Drug Administration's Center for Drug Evaluation and Research (CDER). Michael Smedley is the branch chief for the Recalls, Shortages and Certificates Branch of CDER's Office of Compliance's Division of Manufacturing
and Product Quality.
*To whom all correspondence should be addressed.
Submitted: June 24, 2009. Accepted: July 9, 2009.
References
1. Code of Federal Regulations, Title 21, Food and Drugs (Government Printing Office, Washington, DC), Part 7, Subpart C, Recalls.
2. FDA, Chapter 7: Recall Procedures, Regulatory Procedures Manual, March 2007,
http://www.fda.gov/ora/compliance_ref/rpm/pdf/ch7.pdf, accessed July 9, 2009.
3. FDA, Investigations Operations Manual 2009,
http://www.fda.gov/ora/inspect_ref/iom/default.htm, accessed July 9, 2009.
4. FDA, Guidance for Industry: Product Recalls, Including Removals and Corrections (Rockville, MD, 2003),
http://www.fda.gov/ora/compliance_ref/recalls/ggp_recall.htm, accessed July 9, 2009.
5. L.D. Torbeck, "Recall Root Cause Analysis," presented at the FDA/PDA Joint Regulatory Conference, Washington, DC, Sept.
9, 2008.
This article represents the views of the authors and not of FDA.
