Other Efforts Being Considered To Expand FDA's Postmarketing Risk Assessment
Despite the number and variety of postmarketing risk assessment programs that FDA has initiated, the changing healthcare environment is challenging the Agency's efforts to rapidly identify, quantitate, and understand new risks associated with marketed products. As already discussed, the Agency is addressing some of these challenges. In some cases, areas have been targeted for expansion, but resources have not permitted the desired enhancements. Some of the targeted areas are discussed in the following sections.
Develop and improve automated systems
The FDA has been developing new systems to manage spontaneous reports for drugs, biological products, and devices, but additional work is needed. AERS, which is 14 months into operation, is just now beginning to achieve some of its goals. The MAUDE system will also need the types of analytical enhancements that are underway for AERS.
Expand systems integration
The Agency would benefit from increasing the integration of its current systems for reporting, monitoring, and evaluating adverse events and product defects. The level and types of integration that would provide the needed enhancements and efficiencies need to be evaluated carefully. Promising areas under exploration include sharing data entry, creating a common electronic gateway, and sharing analytical techniques. The Agency is also considering establishing a shared data warehouse that would allow reviewers and researchers from any center to investigate adverse event data using the same software. Database integration would improve the FDA's ability to assess adverse reports across centers. Substantial efforts and resources will be needed to develop the integration proposed in these areas and provide accessibility to all reviewers and researchers.
Increase access to large healthcare databases
FDA's spontaneous reporting systems could benefit from increased access to broad-based health information databases that would allow the rapid exploration of potentially serious problems and more rigorous investigations than are currently possible. For example, a database maintained by a health maintenance organization will include usage data as well as event data, permitting estimations of incidence rates. Such estimations are impossible using only spontaneous reports, which provide no information on usage rates. The Center for Drugs and the Center for Devices have contracts with some health maintenance organizations for access to such databases, but these programs need to be expanded and made accessible to the entire Agency.
Create a network of sentinel sites
The creation of a network of sites (sentinel sites) would help provide optimal surveillance of products that are being used primarily at user facilities such as hospitals or clinics. A representative sample of these facilities could maintain full and accurate reporting of a reasonably high proportion of all adverse events that occur for a given product. For example, products used in organ transplant recipients could be monitored better if some organ transplant centers were identified and supported as sentinel sites. As noted earlier, this type of system is being piloted for medical devices. Under this plan, a network of designated hospitals would report on all device-related adverse events that occur at the sites. The pilot program has been limited by the cost of supporting such sentinel sites. The Center for Devices is now exploring how such as system could be expanded to include a representative sample of facilities.
Integrate pre- and postmarketing collaboration
Close communication and interaction among pre- and postmarketing groups within the Agency would enhance the prospects of effective continuous surveillance as products move into the marketplace. Both the Center for Drugs, through its new Office of Postmarketing Drug Risk Assessment (OPDRA), and the Center for Biologics, through its Managed Review Procedures, have planned programs to encourage this type of interaction. Other centers are considering similar approaches. For example, OPDRA is exploring the use of a tool that will allow some reviewers to view postmarketing safety reports in the aggregate.
Optimally, both premarketing and postmarketing staff should be actively involved in the design of postapproval studies and the analysis of observational studies submitted by sponsors in response to FDA postapproval requests. The availability of postmarketing data to premarketing reviewers considering additional indications for a currently marketed product would provide information on safety considerations for supplemental applications.
Achieving increased integration of pre- and postmarketing information will require additional staff, as it adds to the responsibilities of both premarketing and postmarketing reviewers.
The Agency could enhance its ongoing programs by investing in research efforts designed to increase the understanding of the causes of and factors contributing to product-related injuries. The Agency needs more research on how to identify and report adverse events, how to provide healthcare professionals and consumers the right information to help them recognize and report on product-related problems, and how to improve analytical methods. The Agency also needs to investigate ways to focus more attention on medical products in the immediate postmarketing period.
A second area for research is expanding the Agency's ability to identify signals of potential safety problems from a database of spontaneous reports. Like the proverbial search for a needle in a haystack, the number and variety of reports, together with the number and variety of products and the lack of reliable usage information, make it difficult to distinguish variability and noise from a real concern. Only a small number of external statisticians and epidemiologists are concerned with methods for screening and analyzing such databases. FDA has put some effort into developing improved tools to explore these databases so potential problems can be identified. Such efforts range from identifying new statistical methods, to establishing action thresholds, to developing computer software that primary report reviewers can use to screen the database for potential concerns. More work in this area is needed.
FDA's use of large-scale medical databases from health maintenance organizations and other sources for routine evaluation of product safety is in its infancy. Appropriate methods and software for analyzing these data are also needed, both for identifying signals and for doing follow-up investigations of signals identified elsewhere.
Backgound incidence rates
Another goal that will require large information systems in pharmacoepidemiology is to gain a better understanding of reporting rates by developing background incidence rates for problems in a population. For the relatively modest cost of conducting additional detailed chart reviews, product-related studies of a given syndrome could be extended to the descriptive epidemiology of the condition in the general population.
Performing laboratory studies of approved products will help discover ways to enhance product safety. One example is the gene amplification (polymerase chain reaction) testing that is being performed on biological products to identify the presence of adventitious agents, such as the human immunodeficiency virus (HIV) and the hepatitis C virus. In another example, identifying surrogate biomarkers of toxicity has been identified by an internal review group of FDA scientists as a high-priority issue for the Agency.
The Agency is considering working with members of the healthcare community to create product registries. Discussions already have been initiated with the American College of Cardiology on the potential for a stent registry. FDA and the Centers for Disease Control and Prevention are exploring the possibility of establishing an independent registry center that manufacturers could contract to develop product registries when needed.
Amend existing statutes
FDA may want to work with Congress to change the FD&C Act to improve the Agency's ability to gather data on serious adverse events and move more quickly to mitigate their effects. One example might be to amend the FD&C Act so it contains the same suspension authority that is available under the Public Health Service Act. This would allow the Agency to suspend the marketing of medical products that present a danger to health, but that do not meet the imminent hazard threshold.