This is an archive page. The links are no longer being updated.

Before the Senate Committee on Health, Education, Labor, and Pensions
February 1, 2000

INTRODUCTION

Mr. Chairman and Members of the Committee, I am Janet Woodcock, M.D., Director, Center for Drug Evaluation and Research (CDER), Food and Drug Administration (FDA or the Agency). I am pleased to have the opportunity to participate in the discussion of drug related adverse events, FDA's role in addressing these events, and what the Agency is doing and can do to reduce their occurrence. As you may know, one of Dr. Jane E. Henney's first initiatives after being sworn in as Commissioner of Food and Drugs was to establish a Task Force¹ to evaluate the system for managing the various risks of FDA-approved medical products. This Task Force focused particularly on FDA's part in that system. As a result of this evaluation, the Task Force prepared a report for the Commissioner in May 1999, entitled "Managing the Risks from Medical Product Use".² The Agency has been proactive in addressing this issue and recognizes that FDA is but one of many players that can and must have an impact on the safety of healthcare in the United States (U.S.). As you may know, in addition to FDA, there are many other Federal agencies within the Department of Health and Human Services, such as the Agency for Healthcare Research and Quality, the Centers for Disease Control and Prevention, and the Healthcare Financing Administration, that are concerned with the issue of medical errors and have relevant reporting systems in place.

Today I would like to address how FDA sees its role in this complex system of managing medical risks. I will describe the problem as we see it, explain FDA's historic role and how that role has evolved in the changing circumstances of healthcare, particularly the increasing use of pharmaceuticals. Finally, I would like to present recommendations made by our Task Force on Risk Management, some of which the Agency is in the process of adopting. While today's hearing is focusing on adverse events associated with human drugs, I would like to note that use of other medical products such as medical devices and biologicals, also involves risk.

In addition, it is important to note that the President's Fiscal Year (FY) 2001 budget, which will be released on Monday, February 7, 2000, will aggressively deal with the issue of medical errors and adverse events.

The Problem

There is no doubt that the availability of medications to treat, diagnose, and prevent disease has brought great benefit. Many previously fatal diseases - such as tuberculosis and certain cancers--are now curable. Other potentially fatal conditions, such as diabetes and severe high blood pressure are controllable, as are other serious conditions such as epilepsy and rheumatoid arthritis. Medicines are available to help prevent deaths from stroke or heart attacks, improve the quality and longevity of life in patients infected with HIV, and even to reduce the risk of certain cancers. Finally, a whole array of drugs are available to prevent or ease the symptoms of the many less serious problems that affect us all from stomach upsets and sunburn to colds and headache. But all these benefits come at a cost.

No drug is 100 percent safe; no pharmacologically active medicine exists that does not have side effects. These side effects can, and often do, lead to harm in certain people, and thus decrease the overall personal and societal benefit from medicines. Today's discussion involves both understanding the current system for managing the risks of these side effects, as well as estimating their impact on the public health.

A major obstacle to our understanding of these important issues is the lack of a common understanding and terminology for the various types of bad outcomes that can occur during medical care. As the General Accounting Office (GAO) points out in its report being released today, "Adverse Drug Events: The Magnitude of Health Risk is Uncertain Because of Limited Incidence Data," and as was also discussed by the Institute of Medicine (IOM) in their recent document on medical errors, "To Err is Human," the lack of agreed upon definitions makes comparisons between reports and studies quite difficult. I would like to discuss some of the definitions I will use as part of this testimony.

I will use the term "adverse event" to refer to a bad or undesired outcome that occurs during healthcare. An adverse event can be caused by many things, including a patient's disease process, a side effect of a drug, poor hospital care or improper surgical technique, or even some random happening unrelated to healthcare. As such, the term "adverse event" does not itself have any cause-and-effect implication -- it simply describes some sort of harm that occurs during healthcare. In contrast, "adverse drug reaction" (ADR or side effect) connotes some potential relationship between the undesired outcome and a medication. One of the tasks of FDA's safety evaluators is to determine the probability that a reported adverse event is actually an adverse drug reaction. Often, a definite cause-and-effect relationship cannot be established. Often the best one can do with the available data is conclude that the adverse event is "possibly" or "probably" related to the drug (i.e., a suspected ADR). For example, some drugs can cause liver failure; however, very ill patients can develop liver failure for other reasons, even if they are taking a drug known to cause liver failure in some people. Sometimes it is simply impossible to determine with certainty the cause of harm.

If an ADR does occur, a significant question that must be asked relates to preventability. Some ADRs are simply not avoidable, even with the best medical care. For example, some patients undergoing chemotherapy for cancer will develop lowered blood counts, causing them to be susceptible to overwhelming and, perhaps, lethal infections. This is a very real risk of such drugs. This very serious risk does not mean that people with cancer should not receive chemotherapy. Rather, it is simply the reality of the state of pharmaceutical science at the present time. Only the further advance of biomedical science will foster the development of safer, more targeted therapies that will lessen the incidence of such unavoidable side effects.

In contrast, many ADRs are avoidable if the best standards of medical care and available medical knowledge are communicated and utilized. Preventable ADRs can be classified into a number of categories depending on their cause.

A defective drug product can cause ADRs. Currently, in the U.S., few injuries or fatalities are attributable to this cause, due to pharmaceutical manufacturers' compliance with and FDA enforcement of FDA quality standards for the manufacture, holding, and shipping of drugs.

A second preventable cause of ADRs is "medication error." In this testimony, "medication error" refers to mistakes in the medication prescribing and dispensing process -- for example, in calculating dosage, reading a prescriber's handwriting, understanding a verbal medication order, dispensing the proper medication, or administering the medication. (Some groups, including the GAO in its report, use a broader definition of "medication error.") Many studies in the U.S. have documented that a large number of medication errors occur annually. Many of these cause no harm, either because they are caught before the patient gets the medicine, or because the mistake is harmless, in the end, to the patient. Nevertheless, a substantial number lead to ADRs.

A third cause of preventable ADRs was referred to in the recent IOM report as "medical error" (note: the IOM report also included medication errors in this category). Broadly, this means an outcome that could have been avoided had optimal medical care been provided. Medical errors include misdiagnosis, improper choice of treatment, failure to avoid drug interactions with other products, failure to monitor therapy properly, inadequate recognition of or response to side effects, and so forth. Categorization of an ADR as due to a medical error can be straightforward, but in many cases is a matter of judgement.

A final cause of preventable ADRs is patient or consumer error. This may be due to failure to understand instructions (failure to fully instruct the patient is a medical error) or failure to follow instructions (for example, giving one's own prescription medications to a friend or family member).

Whether an ADR is preventable or not is one way of looking at these events. Another important way of analyzing ADRs is whether or not they were expected (that is, whether or not the ADR was previously identified). This categorization of ADRs relates to scientific uncertainty and the limits of our ability to predict.

A degree of uncertainty always exists about the benefits as well as the risks of all drugs. As the GAO report points out, medical scientists may find major new uses for drugs that have been marketed for decades. Likewise, unexpected side effects (those not included in the product's label) may not be detected for many years after a product is first marketed, particularly if the side effect is rare or unusual. More frequent adverse reactions (both serious and trivial) are usually detected in clinical trials before drugs go on the market. However, these premarket studies conducted for drug approval typically expose only 2-5,000 patients, many of them for fairly short durations of time. These are not enough subjects to reliably detect a type of ADR that occurs only once in 5,000 or more exposures. Once a drug is approved, it may be administered to millions of people the first year, and thus previously unknown, rare types of ADRs may occur. In addition, despite a strong attempt by FDA to broaden the patient population in clinical trials, there will inevitably be many circumstances of use once a drug is marketed that are quite different from the premarket testing. Drugs may be used in different populations (children, pregnant women), or with combinations of different drugs or other products, or for different conditions (so-called "off-label" use) than those evaluated in the clinical trials. There is more uncertainty about both risks and benefits in these less-studied circumstances, and unexpected ADRs may arise. Although serious, unexpected adverse reactions represent only a very small proportion of the overall scope of ADRs, they are historically particularly unacceptable to the public and, if not properly understood, could undermine confidence in the drug approval process.

The definitions discussed above are not standardized and different meanings have been used in various studies and analyses of the problem. Understanding of the epidemiology of the problem has similar challenges that limit our knowledge of the scope of harm caused by ADRs.

Scope of the Problem

The GAO report describes what is known about the scope of ADRs. I believe the report accurately describes current knowledge, including the major gaps and deficiencies. Although the published estimates of drug-related deaths and injuries vary widely, all point to a national problem of very serious magnitude. It is generally agreed that tens of thousands of people die every year from adverse effects of the medicines they are taking. It is estimated that 7,000 people die yearly from medication errors (as defined in this testimony) alone. Because most ADRs are not fatal, the overall personal and economic cost to society is much larger than the mortality figures alone would indicate. For example, it is estimated that up to ten percent of hospital admissions to medical wards are drug-related, and 50 percent of those are preventable. Using a consensus panel method, Bootman et al estimated that the cost in the U.S. from ADRs incurred in the outpatient setting exceeds 75 billion dollars annually. However, as noted by the GAO, all the existing literature is fraught with methodological problems. It is difficult to compare rates among studies because of differing definitions. There is a paucity of information about ADRs in outpatient or nursing home settings.

Very little is known about what proportion of all ADRs are preventable. However, it is certain - and it is important to note - that the vast majority of ADRs referred to in the various studies cited by GAO are side effects that have already been identified, recognized and are described in the product label, that is, they are expected ADRs. Given the scope of the problem, and its potential to increase with the ever-growing use of pharmaceuticals and the aging of the population, a mechanism for systematic data collection is urgently needed to gain a comprehensive understanding of the incidence and scope of ADRs. Current data systems are not adequate to define the scope of the problem.

FDA's Role in Preventing ADRs

As described in FDA's Report "Managing the Risks of Medical Product Use," the U.S. has an extensive system for managing drug risks and maximizing benefits. Participants in the system include FDA, other government agencies, healthcare practitioners, patients, State accrediting boards, healthcare facilities, professional societies, academic centers, and increasingly, healthcare plans and payers.

Traditionally, one of FDA's primary roles within this system has been in the premarket phase. FDA sets requirements for the quality, safety, efficacy, and labeling of pharmaceuticals and reviews marketing applications against these requirements. FDA only approves a drug for marketing if the product's manufacturing quality is reliable and if the scientific data from adequate and well-controlled clinical trials demonstrate the safety and efficacy of the product.

After a drug is approved, the prescriber (the "learned intermediary") assumes primary responsibility for managing the product risks (and benefits) for the individual patient because of his/her specific knowledge of the unique circumstances surrounding each individual patient. In this situation, FDA's role has been to assist the prescriber by requiring that all risks and benefits are honestly described in the label and promotional materials, and to assure, through postmarketing surveillance of reports of potential new safety information, that this new information about risks is relayed promptly to clinicians. To minimize risks, product labeling often describes how to select patients, how to select and modify the dose schedule for individual patients, how to avoid interacting treatments, how to monitor for drug toxicity, and what measures to use to avoid or mitigate drug toxicity. FDA and manufacturers rely on practitioners to prescribe products with full knowledge of the prescribing information and limitations detailed in the product labeling. Likewise, practitioners presume their patients will use their medications according to directions given. However, we know this does not always happen.

FDA's other traditional primary role within this overall system is its postmarketing surveillance systems for detecting rare, serious, unexpected ADRs (i.e., serious ADRs not described in the label). These include, for example, adverse reactions not previously observed, increased severity of previously anticipated risks, side effects resulting from previously unknown drug interactions or drug effects in particular populations.

FDA postmarketing surveillance programs have not routinely addressed the epidemiology of known (expected) drug side effect. In addition, ADRs resulting from serious errors in the practice of pharmacy, medicine, or nursing have been considered the purview of the State regulatory boards, not the FDA. However, medication errors resulting from similarities in product names or confusing packaging of products have been areas of increased FDA oversight in recent years. There is no national program devoted to the study and understanding of known adverse drug reactions (expected). The absence of such a program has led to the dearth of standard definitions and reliable data as described above and in the GAO report.

Current FDA Systems

Although FDA's current systems are designed to focus primarily on rare, unexpected adverse events, this activity, nonetheless, represents a significant undertaking and resource commitment. The following is a brief description of the highlights of the postmarketing medical products safety surveillance programs currently operated by FDA.

Reporting Systems  FDA maintains several databanks that are populated with reports from consumers and healthcare professionals of individual cases of suspected adverse reactions involving medical products. A report does not mean that the medical product definitely caused the reaction; rather, is simply means that the reporter (whomever that may be) believes that there is at least the possibility that the medical product caused the reaction. It is not uncommon to have duplicate reports involving the same "case" as it may be reported by the patient, by a pharmacist, and by the patient's prescriber. Thus, in these systems, the number of "reports" does not necessarily equal the number of "cases." FDA must carefully scrutinize the reports and often calls reporters and practitioners back for further information in order to most fully understand the report and in order to determine as accurately as possible whether a new report is actually a duplicate of an old report.

These reports come to FDA in one of two ways. First, individuals (both consumers and healthcare professionals) report directly to FDA via letters, faxes, telephone calls, emails, or directly over computer modems. These are called "direct reports" and can include anything from a report of a serious, unexpected ADR to a report of a well-known, trivial ADR. Approximately 6 percent (approx. 16,000) of the reports FDA presently receives annually are obtained in this manner.

Second, manufacturers are required to submit to FDA certain voluntary reports they receive from consumers and healthcare professionals. Manufacturer reports account for approximately 90 percent of reports received by FDA. The requirements for these submissions are delineated in FDA regulations (Title 21, Code of Federal Regulations section 314.80) and are generally consistent with reporting requirements agreed to in various international forums. Reports of cases that represent serious, unexpected ADRs (approximately 30 percent of the reports received annually) must be submitted to FDA within 15 calendar days of the manufacturer becoming aware of it. As illustrated in Attachment 1 (enclosed), the number of serious, unexpected reports has risen dramatically over the last several years. FDA has worked closely with counterpart regulatory agencies in the European Union, Japan, and Canada and with the innovator pharmaceutical industry to develop reporting standards (for both content and electronic formatting) so that these reports can be submitted more expeditiously in the very near future.

The Adverse Event Reporting System (AERS) is FDA's new adverse event database for drugs and therapeutic biological products described above. It includes about two million reports to date, including approximately 261,000 reports received in Calendar Year 1999 and we expect to receive 345,000 reports annually by 2001. FDA safety evaluators, epidemiologists, and medical and statistical review officers use the data from AERS to identify serious, rare, unexpected adverse events or increased incidence of various reactions. When a "signal" of a potential adverse reaction is detected, the review team reviews available data from AERS and other sources (other databases, other study reports, counterpart agencies in other countries, academic centers) to try to determine more definitively the relationship between the harm and a medical product. Based on the results of the evaluations, FDA has several tools for managing the new risk including requesting further definitive studies from the manufacturer of the product; disseminating new risk information via Internet, Dear Healthcare Professional letters, Public Health Advisories; revising labeling to add new information or make information more prominent via so-called "black boxes"; requiring formal risk management plans from companies; restricted distribution schemes; required patient registries; required written patient informed consent prior to use of the product; MedGuides or other patient oriented information leaflets; or withdrawal of the product from the market.

As with its predecessor system, AERS was never intended or designed to be a registry of all suspected adverse reactions that occur to all Americans with all pharmaceutical products. The system is designed to detect rare, serious, unexpected drug reactions that cannot be detected in routine premarketing clinical trials. Since 1961, FDA's surveillance systems have performed this function extremely well. They are, however, by design, not able to calculate the incidence of ADRs or provide a national overview of the scope of the ADR problem.

MedWatch.  In June 1993, FDA initiated a new program called MedWatch. Through the MedWatch program, FDA encourages consumers and healthcare practitioners to report suspected ADRs that are serious and unexpected either directly to FDA or to the manufacturers. A major message from MedWatch continues to be that such reporting, although currently voluntary, is a responsibility of all healthcare professionals. In addition, the MedWatch program has established MedWatch Partner collaborations with several hundred consumer and healthcare professional organizations in this country and internationally. Through these collaborations, FDA is able to quickly provide feedback to specific organizations and their members when new risk information is learned about drugs that are of particular interest to those in that organization. This intake and feedback loop on new risk information is an integral component of FDA's overall management scheme for communicating new information about risks associated with FDA-regulated products.

Clear and effective communication about risk is a key component in successful risk management. FDA is increasing our activities in this area to promote an effective exchange of information with all responsible partners.

The Drug Quality Reporting System (DQRS).  The DQRS receives reports of deviations from good manufacturing practices (GMPs) that occur during the manufacturing, shipping, or storage of prescription or over-the-counter (OTC) drug products. Drug quality concerns include a number of hazards, which may be due to improper formulation, packaging, or labeling. In FY1999, 2,050 reports were received resulting in the initiation of eight recalls. Most of the recalls were due to inappropriate labels being placed on products or other specific problems with the manufacturing of the product. Manufacturers of prescription medical products are required to submit reports when certain problems in manufacturing occur. FDA field personnel inspect the manufacturers to ensure that they file these reports as required by regulations.

Medication Error Reports.  FDA also receives reports of medication errors from consumers and healthcare practitioners directly or via the manufacturers involving prescription drugs, generic drugs, and OTC drugs. Medication errors can occur when prescribing, dispensing, or administering a product. Common causes of medication errors include confusion in the labeling of products, poor communication, patient misunderstanding, and ambiguities in product names or directions for use. Medication error reports are also voluntarily forwarded to FDA from the United States Pharmacopoeia and the Institute for Safe Medication Practices. The Agency maintains a central database for all reports involving an actual or potential medication error. The database currently contains some 15,000 reports.

Additional data sources.   The "signals" detected from the reports described above are often not definitive enough for drawing conclusions and taking action. FDA needs access to other data sources to aid in analysis and support regulatory decisions. Currently, under a grant program, FDA has cooperative agreements with several investigators who have access to large healthcare administrative databases wherein product use is linked to subsequent diagnoses, hospitalizations, and other adverse events. Through this arrangement, FDA can conduct a limited number of cohort and case-control studies conducted as needed to investigate a specific safety issue. In addition, the Agency can obtain data on the extent of use of pharmaceuticals through contracts. The Agency is a long-time user of the National Disease and Therapeutic Index (NDTI), the National Prescription Audit Plus (NPA), Provider Perspective (PP), and Retail Perspective (RP) for postmarketing surveillance activities.³ FDA must rely on multiple approaches because no single approach is sufficiently comprehensive to permit full evaluation of all-important problems.

Agency Infrastructure.  In the winter of 1998, CDER expanded an existing division to create a new of Office of Postmarketing Drug Risk Assessment (OPDRA), to reflect the increasing importance of postmarketing surveillance and risk assessment. The purpose of this office is to plan, direct, and collaborate in the conduct of epidemiological studies to detect, explore and confirm safety signals, assess risk, and provide oversight for the monitoring of medication errors and other drug surveillance strategies.

OPDRA involvement with the reviewing divisions begins even prior to the approval of the drug product. In meetings with the review division several months prior to planned approval of the product, OPDRA staff and staff from the review division share information on the product and concerns about outstanding, unresolved safety issues. Plans for the oversight and surveillance of the product during its first year of marketing are made and specific postmarketing studies to be requested of the manufacturer are agreed. Once the product is marketed, OPDRA and the reviewing division meets on an ad hoc basis for emergent safety issue and often on a routine basis to discuss emerging safety concerns. Formal consults are provided the reviewing division on safety concerns with suggestions on how the new information might be conveyed and any new risk managed.

For serious and life-threatening concerns, a CDER-wide team with representation from the reviewing division, OPDRA, the appropriate Office of Drug Evaluation and the Office of the Center Director is usually assembled to decide what regulatory action, if any, is appropriate. In addition to OPDRA personnel, individuals in FDA's CDER Office of Compliance operate an inspection program to assure manufacturers' compliance with reporting requirements.

Responsibilities of Other Parts of the Healthcare System.

FDA must constantly balance the public desire for access to as many promising therapies as possible with the risks associated with both unapproved therapies, as well as products that have been approved. In addition to tools the Agency may use in allowing a product on the market, we also depend on other parts of the healthcare delivery system to monitor and prevent risks as well. For example, in July 1998, FDA approved thalidomide for marketing as a treatment for erythema nodosum leprosum (ENL), a serious inflammatory condition in patients with Hansen's disease (also known as leprosy). At the same time FDA imposed unprecedented restrictions on the drug's distribution. Because of its well-known potential for causing birth defects, thalidomide is among the most tightly restricted drugs ever to be marketed in the U.S. Thus far we believe this system of restrictions has worked as planned. However, it is a resource intense system that requires the ongoing vigilance and commitment of not only FDA and the manufacturer, but also patients, pharmacies, the national patient registry program, payers, and the prescribers.

FDA Task Force on Risk Management.

In its May 10, 1999, FDA's Task Force on Risk Management issued its report. The Task Force found that FDA's postmarketing surveillance and risk assessment programs are, for the most part, accomplishing the purposes for which they were designed. However, it noted that these systems were not aimed at understanding the epidemiology and root causes of known side effects. The Task Force issued a number of findings and recommendations directed at improving the current systems for preventing adverse events. These included:

Premarket Review Programs

The Task Force found no evidence that drugs reviewed under the Prescription Drug User Fee Act (PDUFA) has resulted in higher rates of postmarketing serious, unexpected adverse reactions. In fact, they found that rates of serious adverse reactions identified after approval were lower for drugs reviewed under PDUFA. Regardless of that positive finding, they did offer recommendations to enhance the application of FDA's quality control system.

1. Initiate steps to have each Center establish separate QA/QC units to support the QA/QC system as a normal part of all activities;
2. Ensure and document ongoing professional education and current core competency training for all reviewers;
3. Complete the Good Review Practice (GRP) documents and keep them current; and
4. Systematically analyze significant postmarketing events and incorporate them into GRP as lessons learned.

Postmarket Surveillance Programs

The Task Force found that FDA's postmarket surveillance programs were performing as intended. A number of enhancements to the current system were recommended.

1. Rapidly complete the AERS system;
2. Integrate existing postmarketing information systems so analytic tools, data entry, and editing can be uniformly applied, and all information is readily available to every reviewer;
3. Enhance and intensify surveillance of newly marketed products;
4. Enhance clinical and laboratory studies to develop new methods to improve product safety; and
5. Develop new methodological tools for improving inference from available databases.

CDER is in the process of implementing these recommendations, which will also be responsive to the recommendations of the Department of Health and Human Services' Inspector General in her December 1999 report.

FDA's future role in pharmaceutical risk management.

If the findings of the IOM's report on medical errors are accepted, it must be concluded that the current U.S. system for pharmaceutical risk management is not achieving a level of safety acceptable to our society. Unless targeted improvements are made, the situation is likely to worsen over the next decade. The increasing prominence of pharmaceuticals in healthcare, the increasing complexity of drug products and of the healthcare delivery system, and the aging of our population will all combine to increase the difficulties of safe drug prescribing, dispensing, and consumption and increase the complexities of an effective risk management and risk communication system.

While policymakers will undoubtedly feel spurred to action by these reports, all would do well to remember a cardinal principle of medicine - diagnosis before treatment. As pointed out by the National Patient Safety Foundation (NPSF), official reaction to medical errors has long been a counterproductive course of "blaming and shaming" (finding some person or entity who could be held accountable, and holding them up to public criticism). Such actions mislead the public and others into believing the problem is "fixed" when the root causes have not even been recognized, much less appropriately addressed. FDA's drug approval process has been part of this dynamic for many years. For example, a drug withdrawal or relabeling most often indicates a safety net that is working properly. A realistic understanding of the pharmaceutical risk management system, including recognition that all drugs will have adverse effects, is essential to formulating effective policy.

Pharmaceutical safety is best understood within a systems perspective. While FDA's premarket and review processes play an important part in the overall system, they are by no means the only factors in pharmaceutical risk management. FDA's safety assessment is predicated upon an understanding that the healthcare delivery system will be able to manage identified risks, provided those risks are reasonable. If the Agency misjudges the ability of delivery systems to manage a particular problem, then the safety of the product will be compromised. The lack of adequate data on ADRs in various settings means that FDA has to rely on assertions rather than data in determining how well the healthcare delivery system can manage all types of pharmaceutical risks.

Improving the safety of pharmaceuticals will require a concerted and systematic effort by all participants in the medical care delivery system -- including manufacturers, FDA, State and other Federal agencies, healthcare professionals, healthcare facilities, patients, consumers, and care delivery systems. In its report on "Managing the Risk of Medical Product Use," FDA put forth a number of options that the Agency could pursue to improve pharmaceutical safety and risk management. As a first step in evaluating these options, FDA is engaging in extensive dialogue with the stakeholders in this process. In March, FDA, in conjunction with the NPSF, will hold a national workshop for representatives of patient and consumer groups. We believe that the voices of patients and consumers must be heard as policy in this area is shaped. FDA is also participating actively in the pharmaceutical safety initiative of the NPSF. This activity includes representatives of all groups involved in or impacted by the safety of medicines. Finally, we feel the discussions and activities resulting from the publication of the IOM report on medical errors will do much to inform the debate. FDA stands ready to play an even more active role in the assurance of drug safety through its risk detection, analysis, management, and communication programs. However, it must be recognized that our resources are severely constrained even within the context of our current activities.

Conclusion.

Mr. Chairman, medical products provide great benefit to the public, but they can also cause injury - sometimes preventable, sometimes not, sometimes expected, sometimes not. FDA and the many other participants in healthcare delivery act to maximize benefit and minimize risk, but often the actions of each of the participants are not effectively integrated. FDA looks forward to working collaboratively with all other parties involved in the healthcare delivery system to address this very critical issue. And, of course, FDA will conduct its activities to reduce medical errors in a way that protects patient privacy. We are proud that FDA-approved medicines are considered to be among the safest in the world and to the extent we can, we want to fulfill our mandate to assure that drugs marketed in the U.S. offer patients real benefits in as safe a manner as possible.