BIO Comments on FDA's Proposed Rule for Safety Reporting Requirements
October 13, 2003
Dockets Management Branch (HFA-305)
Food and Drug Administration
5630 Fishers Lane, Rm. 1061
Rockville, Maryland 20852
Re: Docket No. 00N-1484, Federal Register: March 14, 2003 (Volume 68, Number 50, Pages 12406-12497)
The following comments are provided by the Biotechnology Industry Organization (BIO). BIO represents more than 1,000 biotechnology companies, academic institutions, state biotechnology centers, and related organizations in all 50 U.S. states and 33 other nations. BIO members are involved in the research and development of health-care, agricultural, industrial, and environmental biotechnology products. BIO appreciates the opportunity to comment on the Food and Drug Administration's (FDA's) Proposed Rule for Safety Reporting Requirements for Human Drug and Biological Products.
BIO agrees with FDA's goal of reducing or eliminating duplicative or redundant safety-reporting requirements in New Drug Application (NDA) and Biologics License Application (BLA) annual reports and with the agency's intent to reduce disparities between requirements in the United States and those in other developed nations of the world, in accord with the goals of the International Conference on Harmonization (ICH). In that regard, we applaud efforts to ensure consistency between FDA requirements and definitions and those of other federal agencies such as the National Institutes of Health (NIH), as well as those included in the ICH E2C Guideline and Addendum (we particularly note FDA's endorsement of the data lock point and International Birth Date (IBD) in ICH E2C) and to adopt MedDRA terminology related to medical terms in clinical safety. We also support requiring only a minimum data set for all non-serious suspected adverse drug reactions and eliminating the requirement for expediting reporting of cases from Class Action lawsuits.
While BIO shares FDA's purpose of ensuring the highest quality of safety reporting, in the interest of maintaining and improving public health, we believe the net effect of the agency's proposals will increase the cost and complexity of safety reporting for regulated industry without a commensurate benefit to the public health. In key areas, FDA's proposals appear inconsistent with the agency's forward-thinking approaches rooted in science-based risk analysis and management and, rather than streamlining safety reporting to ensure its usefulness, will complicate and slow the process. We further are concerned that in several areas FDA's proposals will result in the U.S. having requirements and definitions that differ in fundamental ways from those of ICH and the Council for International Organizations of Medical Sciences (CIOMS), thus thwarting the objective of international harmonization.
The proposed rule requests comments on whether the agency should implement the recommendations contained in the Addendum to ICH E2C. We strongly endorse the adoption of the E2C Addendum, in its entirety, as well as full adoption of the ICH E2C Guideline. Full adoption of the ICH E2C Guideline is essential for true global harmonization. However, we believe that global harmonization will not be achieved if the final rule requires extensive new additions and appendices to the Periodic Safety Update Reports (PSUR), as well as preparation of additional periodic reports at slightly different time intervals and with different format and content than required in other regulatory jurisdictions. We believe these requirements will lead to apparent data discrepancies and variances in interpretation of the data. Indeed, this would likely lead to certain inconsistencies in the information provided to different regulators, and resources required for the preparation of these reports would not be available for other pharmacovigilance activities.
BIO enthusiastically endorses FDA's approach of working with stakeholders to ensure that its safety reporting requirements appropriately harmonize with those of international regulatory bodies and, importantly, are in the best interest of consumers. We look forward to a continuing dialogue with the agency to achieve this goal.
BIO is particularly concerned about the agency's assumptions and conclusions regarding the financial impact of these proposals on the regulated industry. The agency assumes, for example, that industry will realize substantial savings (nearly $40 million annually beginning in year 4; shown in Table 12) from implementing the requirements of the proposal. We question both the source of these numbers and the assumptions on which the bottom line is based, including the assumption of substantial savings from efficiencies in database management. We estimate, to the contrary, that industry will need to establish and maintain entirely separate databases to continue to comply with existing global requirements and to comply with the new FDA requirements. We also believe the estimated cost of additional and/or specialized personnel is both seriously under-estimated and based on a flawed assumption that personnel costs, including costs of medical professionals, can be measured accurately by calculating a per-hour compensation (as shown in Table 16). Therefore, BIO recommends that the agency, working with industry, re-estimate these costs and, based on an improved and more accurate estimate, re-propose the rule.
II. B. 2. Quality of Postmarketing Safety Reports
BIO agrees that surveillance depends on the quality of adverse event reports. We believe FDA can ensure quality by requiring that the physician responsible for the content of adverse event reports be a graduate of an accredited medical training program. BIO suggests that the agency remove the additional requirement for licensure because it is logistically more complex and adds little to the requirement for certified medical training.
III. A. 1. Suspected Adverse Drug Reaction
BIO agrees that international consistency is appropriate and desirable, but is concerned that neither the acronym SADR nor the definition proposed is consistent with ICH E2A guidance. For example, there is ambiguity as to whether the abbreviation SADR is to be taken to mean "serious," as opposed to "suspected," adverse drug reaction. We believe the discordance between the proposed definition and the ICH definition may create confusion that would complicate maintenance of exemplary compliance. However, BIO recognizes the merits of FDA's proposal to make explicit the uncertainty inherent in causality assessment for individual case reports. We recommend that the FDA work with ICH to resolve this discrepancy.
The ICH E2A guidance defines "adverse drug reaction" as "all noxious and unintended responses to a medicinal product related to any dose." The phrase "response to medicinal products" means that a causal relationship between a medicinal product and an adverse event is at least a reasonable possibility, i.e., "the relationship cannot be ruled out."
Further, the EU Clinical Trial Directive Final Guidance on Adverse Events (AEs) [section 6.2.2 Assessment of Causality, April 2003] states "causality shall be determined according to the definition of an adverse reaction as given in Article 2 of the Directive 2001/20/EC . . . All adverse events judged by either the investigator or the sponsor as having a reasonable suspected causal relationship to an investigational medicinal product qualify as adverse reactions." In the spirit of global harmonization, we recommend that FDA note that the international definitions stress "reasonable" and "suspected" in their definition of causal relationship. While perhaps technically consistent with ICH E2A, we believe FDA's proposed approach does not incorporate all of the concepts in that document or in the EU Clinical Trials Directive. In particular, for example, the approach proposed by FDA does not include the clearly delineated concept of "reasonable causal relationship," which prompts an assessment of actual evidence or arguments to support causality.
In a number of sections of the proposed rule, the agency defines "reasonable possibility" as synonymous with "the relationship cannot be ruled out." Thus, a causal relationship becomes the default assessment, not only for spontaneous adverse experience reports, but also for clinical trials and other organized data collections. We believe that, should this emphasis remain in the final published rule, sponsors and investigators participating in premarketing and postmarketing clinical trials necessarily, interpreting "the relationship cannot be ruled out" strictly, will report virtually every event, as nothing can ever be completely ruled out. BIO believes this curtailing of clinical judgment in assessment of causality would result in a substantial increase in the volume of expedited reports (if the relationship to study medication "cannot be ruled out" in clinical trial cases, we calculate that the number of Investigational New Drug Application (IND) safety reports will increase by 10- to 12-fold) and could also have an adverse impact on the integrity of studies, given the need to break the blind on an increased number of serious, unexpected, potentially related events. In studies of serious chronic diseases such as cancer and AIDS, it is not at all uncommon for most patients to have at least one serious adverse event. The impact of breaking the blind for all study therapy would reduce expected aggregate time-on-treatment in clinical trials and have a negative impact on the efficiency of clinical trials (a larger sample size would be needed to produce the required statistical power), without a corresponding reduction of risk to study subjects. Indeed, it would be difficult to estimate an appropriate sample size at the outset of a Phase III clinical trial, when the "losses" due to breaking the randomization code would be unknown.
Since the power of many trials depends less on total sample size than it does on the expected number of outcome events, there will be a significant risk of reducing power below nominal levels (particularly in trials with failure-time endpoints). More importantly, knowledge of treatment (including the post-treatment observation period) introduces potential bias that could undermine interpretation of the efficacy/effectiveness analysis (the potential for bias alone can undermine a trial let alone confirmation of bias).
This risk (bias and power-reduction) may be difficult to measure during the trial, let alone anticipate in the design stage. If the proposal is implemented, we believe it will place a significant burden on both the agency and companies to negotiate and manage exemptions to expedited reporting and necessitate the creation of an FDA structure and process around these negotiations to ensure consistency across products.
FDA invited comment on alternative ways to handle serious suspected adverse drug reactions that occur during clinical studies to minimize "over-reporting." While there may be certain clinical situations in which it would be appropriate for FDA and companies to consider alternative reporting methods for specific suspected adverse drug reactions (e.g., progression of a disease under clinical trials in HIV/AIDS, oncology, or specific cardiovascular endpoints for specific protocols for which there is an independent safety data monitoring board), we remain concerned that such "alternative reporting" would not have the intended effect of reducing over-reporting and could exacerbate the previously mentioned problems with the definition "cannot be ruled out."
Absent other mechanisms for dealing with over-reporting, companies routinely will seek alternatives to unblinding and expedited reporting. BIO believes that negotiation of alternatives that are comprehensive enough to protect study integrity will be difficult and resource-intensive because the spirit of the proposed reporting is antithetical to sufficiently restrictive alternatives. We believe the FDA does not have the resources needed to deal with this eventuality.
However, if the agency, notwithstanding these concerns, decides to move forward with the use of alternative ways to handle serious suspected adverse drug reactions from clinical studies to "minimize over-reporting," we recommend that any such approaches only be considered in late stage development (Phase III and IV studies).
BIO also is concerned about the potential effect of this proposal on signal identification and ultimately on drug labeling. Certainly the rule change will result in many more adverse events being considered "drug-related." We believe that, without the investigator being advised to use reason and best medical judgment in determining causality, many events will be listed as "drug-related" even though the likelihood of a true causal relationship is minimal. We therefore are concerned that should many of these events -- where the causal relationship is unproven or, at best, tenuous -- find their way into the product labeling, this could result in a significant and detrimental effect on the utility of the labeling information and potentially of the product itself. The consequence of expediting numerous reports would be to obscure potential signals. Institutional Review Boards (IRBs), already facing difficulties, would be faced with an order of magnitude increase in the number of expedited reports they receive. Since only "active" treatment cases would be unblinded and expedited, these would be virtually uninterpretable but may raise concerns and introduce delay without improving risk assessment.
We strongly recommend the FDA reconsider this proposed change and not define "reasonable possibility" as meaning "the relationship cannot be ruled out." We believe the ICH definition of "reasonable causal relationship," i.e., "…there are facts (evidence) or arguments to suggest a causal relationship" is more appropriate. The essence of the ICH definition has been implemented successfully in the European Directive on Good Clinical Practices (GCP) in Clinical Trials (2001/20/EC) and specifically in the final wording of the "detailed guidance on the collection, verification and presentation of adverse reaction reports arising from clinical trials on medicinal products for human use" (Annex 1, page 12) which states "The expression reasonable causal relationship means to convey in general that there is evidence or argument to suggest a causal relationship." We suggest that this is a more appropriate choice than the definition "the relationship cannot be ruled out."
The implications of "the relationship to product cannot be ruled out" for IND safety can be summarized as threefold:
increased number of IND Safety Reports: as noted above, BIO believes there will be a significant increase in expedited reports because of the proposed change in the definition of causality. Since it is rarely feasible to "rule out" the possibility that an exposure contributed to a specific event, the overwhelming majority of events will be classified as related. Companies following CIOMS III/V will base expectedness on Developmental Core Safety Information (DCSI), as it is reflected in the Reference Document for the trial. Accordingly, they will classify reports as unexpected until the threshold for adding the event to the DCSI is crossed and continue to expedite those events that are serious and possibly related. As such, the majority of serious suspected adverse drug events in IND studies will be unexpected and possibly related, even in large phase III studies.
excessive unblinding in IND studies: as noted, the majority of events in IND studies will be subject to expediting under the new rule. In addition, the rule retains the existing requirement to unblind serious unexpected SADRs prior to expediting. In combination, the requirement to unblind and the "cannot rule out" clause will cause excessive unblinding in many programs. This is particularly true for studies of longer duration, indications associated with significant morbidity and studies of less common outcomes. It will increase sample size requirements and the risk of under-powering studies, as well as threaten conclusions by raising the specter of co-intervention and ascertainment bias.
FDA acknowledges the risk of over-reporting in section III.A.1. However, BIO is concerned that the list of diseases and events that will lead to over-reporting is much longer than the FDA anticipates. The extent and impact of unblinding will be difficult to predict in the design stage. Companies will go out of their way to ensure that they minimize the chance of running uninformative studies. This frequently will lead to iterative negotiations with the FDA, tax agency resources, and result in long lists of proposed reporting exclusions. We believe that in the end, it will create significantly more work, to gain little incremental information that may get lost in the attendant information overload.
IRB/investigator information overload resulting from over-reporting: Investigators and IRBs have stressed the difficulty they have interpreting the significance of Individual Case Safety Reports (ICSR). There is no reason to believe that the signal-to-noise ratio will improve with increased IND safety reporting. As such, the proposal will increase the volume of potentially confusing reports that are distributed to investigators and IRBs. Absent a better framework for investigator/IRB interpretation of IND Safety Reports, it will exacerbate investigator/IRB information overload and divert site resources from the task of reliably reporting high-quality safety information to sponsors.
BIO recommends elimination of this proposal. However, if the proposal is implemented, alternatives to extensive expediting and unblinding will be required in blinded studies. There is no single alternative that mitigates BIO's concern about the risk to study integrity that this proposal represents. Some possibilities that might be considered, with analysis of their potential viability, include:
Expediting blinded safety reports - We do not believe this is a good alternative, as it results in confusion and delays decision-making about potential safety issues. Moreover, it is not consistent with ICH.
Expediting a subset of cases that undergo unblinded evaluation: this would involve a two-pronged approach. First, restrict expedited reporting/unblinding to unexpected cases for which there is a reasonable possibility of a relationship to the study drug (based on the current definition of reasonable possibility; i.e., that there are facts, evidence, or arguments to suggest a causal relationship). Second, use a small internal or external safety committee to monitor adverse events, by treatment arm (without knowledge of the actual treatment assignment). The exact makeup and function of the committee would require delineation; however, a small group of professionals who are not in a position to influence the outcome of the trial potentially could monitor evolving treatment arm imbalances and use a priori guidelines to inform the agency and investigators of noteworthy safety findings. The approach would permit ongoing safety analysis, without interfering with efficient trial management or the validity of study conclusions. While too cumbersome to use in all blinded studies, such an approach, when feasible, could be relatively efficient and protect study validity. For the sponsor, the approach could reduce the burden of over-reporting and eliminate power and bias issues. For FDA, it could reduce resources needed for data management and analysis and increase the likelihood of obtaining timely synthesis of potential safety issues. For the investigator, it could reduce information overload. In addition, it could provide an opportunity to increase the uniformity of ongoing safety evaluation in clinical trials.
Increase the scope of reporting to FDA and leave the scope of reporting to investigators unchanged: this alternative also could limit expedited reporting/unblinding to unexpected cases for which there is a reasonable possibility of a relationship to the study drug (based on the current definition of reasonable possibility; i.e., that there are facts, evidence, or arguments to suggest a causal relationship). In addition, it could allow for blinded expedited reporting of a broader range of cases to the agency only.
We believe this approach could reduce investigator information overload, as well as ensure that the agency receives a more complete picture of the developing safety profile, albeit blinded, in a timely manner.
III.A.2. A Life-Threatening SADR
BIO agrees with the FDA proposal concerning life-threatening suspected adverse drug events, as it applies to pre-approval safety. However, we note that it will be difficult to obtain the investigator's view consistently, when (s)he does not feel the event is life- threatening. We believe sponsors should not be required to report the investigator's reasons for classifying an event as non-life-threatening, when the sponsor deems the event to be life-threatening.
III. A. 4. Contractor
BIO believes that, given the array of possible safety reporting arrangements between "contractors" and "applicants," the rule should not mandate that the applicant is always responsible for safety reporting. However, recognizing the importance of this responsibility, the final rule might suggest that applicants and contractors are responsible for an agreement that specifies responsibilities for safety reporting. In this regard, we note that draft ICH E2D guidance language may be useful: "When companies co-develop, co-market, or co-promote products, it is considered very important that explicit contractual agreements specify the processes for exchange of safety information, including timelines and regulatory reporting responsibilities."
There are certain situations where the applicant (e.g., NDA or license holder) is a small company and the contractor is the large company. In such a situation, the applicant and contractor often have detailed written agreements whereby the contractor is responsible for safety reporting. Indeed, many such agreements are already in place. BIO suggests (1) that the rule take account of the large variability in contractual arrangements, under some of which it would not be workable for the applicant to be required to be the reporter and (2) that any regulatory changes apply only prospectively, since various business partners already have a wide range of safety reporting agreements in place.
III. A. 6. Active Query
Under the proposal, FDA would amend its postmarketing safety reporting regulations and define the term "active query" to mean direct verbal contact … with the initial reporter of a suspected adverse drug reaction or medication error by a health care professional representing the manufacturer (applicant). The proposal would mandate such active query in a number of situations, based on FDA's belief that that in many cases, use of active query during initial contact with these reporters would provide manufacturers and applicants with adequate safety information and could eliminate or decrease follow-up time expended.
However, in the postmarket setting, health care professionals are not required to call the manufacturer to report an adverse event. It is purely a voluntary system. In most instances, when a health care provider contacts a manufacturer/applicant, (s)he is calling to gather information. Most seem unaware of the need to provide any information regarding an event/suspected event to the manufacturer, much less the FDA. For this reason, health care professionals rarely have comprehensive patient information available at the time they contact the manufacturer. Health care professionals also have concerns about liability and patient privacy, questioning, for example, whether they should provide information to a third party without the patient's express consent. For this and other reasons, we believe active query will be difficult. Mandatory active query will be particularly challenging with foreign-U.S. situations, as data privacy laws may act to inhibit such activities and their transborder exchange.
Further, BIO believes direct query could pose an enormous burden on the health care professional. Health care professionals are busy with patient management during daytime hours and prefer not to be contacted at those times when their primary concern is patient care. Many health care professionals request communication from the applicant/manufacturer via fax or by mail rather than by phone. Written communication allows health care professionals or their staffs the freedom to collect the required information and respond to manufacturer's queries as time allows. BIO believes that to pursue direct verbal contact repeatedly to obtain additional information could strain any working relationship between the health care professionals and/or their staffs and the applicant/manufacturer. Ultimately, the health care professional could refuse any communication with the manufacturer/applicant, which would decrease the amount of reporting overall.
When consumers contact manufacturers/applicants, they generally provide as much information as they have available during the initial contact. They can, however, be distraught regarding their diagnosis, treatment, and/or the event, and refuse to give any substantial information concerning the event. They can be reluctant to provide information regarding their health care providers. They do not want the manufacturer/applicant to "bother" their physicians. Or perhaps they have not contacted their physicians for whatever problems they are reporting. If the consumer works, (s)he is not available during the workday hours.
While BIO agrees that information received via direct verbal contact is valuable, we believe written follow-up, especially medical records, can be more accurate. We recommend active query be focused on cases where specific information from the health care professional is required to clarify important elements in case characterization. If active query is used excessively, we believe it could discourage voluntary reporting. Years of experience in data collection in the postmarket setting suggest that there is no single "best" means of communication with consumers and health care professionals.
BIO suggests, in the spirit of an efficient application of resources, that the focus of any active query (that is, a particular line of questioning designed to capture clinically relevant information associated with the drug product), as a possible tool among many to gain full information, be concentrated on follow-up of selected adverse events (expected and unexpected) for those products for which a sponsor and the agency, as well as other regulatory authorities, have formulated Risk Management Programs as the approach to manage safety-related risks for a product through its life cycle, from first in human studies to product senescence. Applicants/manufacturers strive to provide full data sets on such serious and non-serious suspected adverse drug reactions and thereby improve the quality of the ICSR for all stakeholders. This also facilitates analysis of such reports when reviewed in an aggregate setting. It would be desirable that the final rule specify that the intention is to apply the term "full data set" to the existing 3500A form and not to the data elements specified in the ICH E2B(M) Guideline. Certain other clinical events on older products (i.e., those more than 5 years old), such as "Death, Cause Unknown" reports or sudden death in a young patient, or reporting of an unlisted Drug Interaction, and certain serious unexpected events, may also meet criteria for active query follow-up.
The concept of active query already exists, both in the context of clinical trials and spontaneous cases. However, if active query beyond the follow-up that currently exists is contemplated in the final rule, BIO believes there may be an undesired impact on the human and financial resources that must be applied to these activities.
We disagree that a sponsor may have difficulty in making an initial determination of "serious" or "non-serious" criteria regardless of outcome. Much of the decision making regarding "serious" and "non-serious" decisions was dealt with in the final regulations on important medical events. In the rare instance (less than 1% of cases) that an initial determination of "serious" or "non-serious" cannot be made, we suggest that the default in such a setting would be to process the report as "serious." The sponsor then would employ a variety of mechanisms, which could include active query, to obtain the needed information.
BIO also believes that submission of an autopsy report, hospital discharge summary, or death certificate for reports of death and hospitalization should not be required in the final rule. This may infringe on or breach mandates of the Health Insurance Portability and Accountability Act (HIPAA) and raise other confidentiality issues regarding both U.S. and foreign data. Further, clarification as to the ability to meet this requirement within electronic reporting capability is requested. As we strive for quality data, we believe it should be up to the applicant to have appropriate processes in place to obtain source documentation and to analyze it, as necessary, to complete a case. In most cases, applicants include the relevant information in the 3500A form. BIO believes separate, systematic submission of such documentation to the agency is not necessary and would create an enormous burden for FDA. In cases where the information has been supplied to the applicant, it could be provided to FDA to fulfill a specific need upon request.
We also note that the definition of health care professional differs from that used in the EU. This means that it would be impossible to have the same listings of health care professional cases in the various ICH regions, if this requirement remains as currently proposed.
III. A. 7. Spontaneous Reports
BIO appreciates the agency's clarification that spontaneous reports do not include cases identified from information solicited by the applicant or contractor, such as individual case reports of findings derived from a study, company-sponsored patient support program, patient registries, etc. However, we are concerned about the definition that a spontaneous report always must be assumed to indicate a reasonable possibility of causality. In the context of medical information queries and other contact with health professionals, particularly in the case of chronic drug treatments for life-threatening illnesses (e.g., enzyme replacement therapy), it is not uncommon for sponsors to identify an "adverse event case report," although the reporting health professional unequivocally states that in his/her professional opinion there is no reasonable possibility that the drug product caused the event. In such cases, it would seem that FDA's proposal would result in reporting that runs contrary to the considered medical judgment of the health professional.
III. A. 8. Medication Error
BIO is very concerned with FDA's proposed way to handle medication errors. We suggest that a differentiation be made between clinical trials, in which setting this provision is unnecessary, and spontaneous sources. We believe that non-prescription products should be considered separately because medications that have been approved for over the counter use generally have a broader therapeutic index and higher safety margin than most prescription medications. Further, many reported cases of medication errors either result in no adverse event(s) or in events(s) that are non-serious and self-limiting. (Although product defects that cause adverse events should be appropriately reported.) We also suggest that before imposing requirements for reporting medication errors, FDA reach an international consensus on how to address this issue on a global basis; ICH is best equipped to spearhead this effort.
According to the Institute for Safe Medication Practices, there are four potential causes of medication error: (1) failed communication (handwriting or oral communication, drugs with similar names or packaging, missing or misplaced zeros and decimal points, confusion between metric and apothecary systems of measure, use of non-standard abbreviations, or ambiguous or incomplete orders); (2) poor distribution practices; (3) complex or poorly designed technology; and (4) access to drugs by non-pharmacy personnel. As many of these causes reach far beyond control of the pharmaceutical industry, there must be greater emphasis on individuals and settings directly associated with dispensing medications. In addition, when patients are responsible for self-administration, i.e. prescription or non-prescription products taken in an outpatient setting, we believe that efforts to engage pharmacies, healthcare professionals, and patients would be a more direct means to prevent or reduce medication errors.
The agency proposes to define medication error as "any preventable event that may cause or lead to inappropriate medication use or patient harm while the medication is in the control of the health care professional, patient, or consumer. Such events may be related to professional practice, health care products, procedures, and systems including: prescribing, order communication, product labeling, packaging, and nomenclature, compounding, dispensing, distribution, administration, education, monitoring, and use."
The proposal also has additional separable definitions for medication errors. An actual medication error is one that involves a patient, whether or not the product was administered. A potential medication error is actually a product complaint, referring to the product name, labeling, or similarities in product packaging that could result in a medication error in the future, but does not involve a patient.
BIO recommends the agency develop one definition for the term "medication error." As the definition is written now, it could be interpreted as meaning that the agency is requiring industry to monitor prescribing practices of physicians, dispensing practices of pharmacists, and use/compliance of the patient taking the product.
FDA proposes to require that each domestic report of an actual medication error be submitted as an expedited report, whether a suspected adverse drug reaction has occurred or not; the minimum data set would be required to be reported as an expedited report.
Most medication errors are related to the practice and dispensing of medication. If a medication error is made by a health care provider, the provider should be responsible for reporting it. The agency should be informed of medication errors, but the justification to expedite these reports is unclear. BIO recommends changes in the proposal to require that expedited reports be submitted if the medication error was reported because of a misunderstanding of the product name or labeling or package instructions and to require that medication error reports received by the sponsor that currently are not subject to expedited reporting requirements (serious and unlabeled reactions) be reported and discussed in the aggregate in the next periodic report.
In general, BIO believes that a significant portion of the proposal on medication errors is unnecessary in that much of it duplicates and carries over from the agency's proposed Bar Code Label Requirements for Human Drug Products and Blood. Much of the current proposal would deal with post-manufacturer concerns (e.g., mis-communication by a health care professional, etc.) that will be addressed with the implementation of the Bar Code rule. We suggest a considerable portion of this proposal be eliminated, thus focusing it strictly on those matters that will not be dealt with under the Bar Code rule.
III. A. 10. Data Lock Point and International Birth Date
We agree with the proposal to use data lock point and International Birth Date to define reporting period of postmarketing periodic safety reports. This is consistent with ICH E2C. However, we seek clarification as to why the reporting period as defined per IBD would preclude a company from changing the periodicity of a report prior to what would be dictated by the U.S. approval. We recommend the revised wording state that the IBD should be utilized for determination of the periodicity globally. We believe alternative reporting should be considered only in rare instances. BIO suggests the final rule allow for discussion and agreement with FDA on a case-by-case basis.
III.B.2.a. Minimum data set
FDA proposes to amend § 312.32(c) to state that sponsors must not submit an IND safety report for a suspected adverse drug event if the report does not contain a minimum data set (i.e., identifiable patient, identifiable reporter, suspect drug or biological product, and suspected adverse drug event). BIO requests clarification regarding the meaning of "identifiable." In at least one public discussion of the proposed rule, the agency's representative appeared to suggest that the potential for contacting an individual is central to the notion of identifiableness. However, this is inconsistent with the lower threshold for "identifiable" used by many companies, as well as guidance provided by some regulators. The latter interpretation holds that an individual is identifiable if the available information pertains to a specific individual and does not depend on the potential to establish contact with that individual; e.g., a patient is identifiable if age or gender is available. Clearly, age and gender alone do not permit establishing contact.
III.B.2.b. Serious and unexpected SADRs
BIO sees value in requiring that sponsors expedite serious and unexpected suspected adverse drug events, if either the investigator or the sponsor determines there is reasonable possibility that an event is attributable to an investigational drug or biologic.
BIO agrees with FDA's proposal to require that companies maintain records delineating due diligence in investigating outcome and expectedness. However, requiring inclusion of this information in IND safety reports will not clarify the safety profile and will tax resources that would be applied to doing so. While the minimum dataset proposal is consistent with the ICH E2A guidance (60 FR 11284 at 11286), the need to report due diligence efforts, on an expedited basis, is not.
III. C. Postmarketing Safety Reporting
Clarification is needed regarding what constitutes a human subject significant risk deriving from animal and in vitro studies. These non-human studies often are done at different dosing regimes, in specialized models, and, especially in biological products, with other antibodies than those under development (e.g., murine antibodies perhaps with slightly different biological activity). We also believe it is important to clarify when the clock starts with animal and in vitro studies.
III. C. 3. Reporting Requirements
BIO believes that for Traditional Periodic Safety Reports (TPSRs) and Interim Periodic Safety Reports (IPSRs), reporting cycles of 7.5 and 12.5 years should not be required, as 5 and 10 years post-approval should provide an adequate safety profile. If this recommendation is retained, consideration should be given, in line with ICH E2C, to require reports at these time periods only if there has been a new indication/dosage form approved, there is use in new population, or there has been a prescription status change, such that the safety profile of the drug is expected to be significantly different.
We also seek a rationale for the use of January 1, 1998, as a demarcation point for preparation of TPSRs vs. PSURs. It would be extremely difficult, if not impossible, reliably to identify cases that should be included in subgroups (e.g., suspected adverse drug reactions from Class Action lawsuits, medication errors, etc.), as proposed in TPSR and PSUR/IPSR format. Difficulty in identifying cases for subgroups retrospectively is primarily a practical matter, owing to the challenges of retrospective application of coding conventions and necessary changes in system capabilities.
III. C. 5. Determination of Outcome, Minimum Data Set, and Full Data Set
The proposal would amend current rules to require immediate determination of the outcome of a suspected drug reaction (whether serious or not). BIO suggests that the language be changed from "determine the outcome" to "determine the serious nature of the suspected drug reaction." We believe it is confusing to use "outcome" interchangeably with "seriousness."
FDA is proposing that, where the needed information is not immediately available, the sponsor use active query. As noted above, BIO believes that active query should be reserved for specific situations in which no other information collecting method is likely to yield adequate results and should be at the discretion of the sponsor. For example, a report of poison ivy should not require a telephone call to the physician even if the "seriousness" was not specified in the initial report.
The agency proposes that manufacturers and applicants immediately obtain data regarding medication errors that do not result in an adverse reaction and for potential medication errors. Potential medication errors are already covered under product complaint regulations. We recommend they not be included in the safety sections.
It would be more appropriate to update the product complaint regulations to require expedited reporting of certain medication errors which do not lead to product recalls. (Currently, reports under product complaint rules are only expedited when there is a product recall.) BIO believes that forcing this issue to be handled via the adverse drug reaction reporting system introduces redundancies as well as inconsistencies with ICH practices. Rule changes along these lines are even less justifiable with potential medication errors.
By proposing that manufacturers and applicants continue to use active query to attempt to determine the outcome of a suspected drug reaction, the agency is introducing a third classification of "seriousness;" i.e., "outcome unknown." We believe this will introduce inconsistencies with other regulatory agencies and will require substantial changes to all systems and business processes. It will make harmonization essentially impossible. It is not apparent that this classification has inherent value in assessment of patient safety.
FDA's focus on active query, in a large number of situations, neglects to acknowledge that active query is not the only way to contact the health professional and obtain important information. Written inquiries may sometimes be more successful or appropriate (for example, if hospital discharge summaries can be obtained). Again, BIO recommends that active query not be a default but rather a tool that is used selectively at the discretion of the sponsor, not as a regulatory mandate.
Finally, FDA proposes that accounting of a manufacturer or sponsor's due diligence in obtaining data be included in the report of the suspected drug reaction. BIO believes that use of the 3500A to document due diligence distorts the usefulness of this document. It is already sometimes difficult to provide a medically coherent narrative while separating the information into initial and follow-up sections. If, in addition, a log of attempts to reach the health care provider by telephone must be included, the character of these reports is changed. We recommend due diligence efforts be documented in company files and not in the 3500A. If included in the 3500A, the due diligence records are then also open to disclosure under the Freedom of Information Act, which could lead to deleterious consequences for both the sponsor and the health care provider.
III. D. 4. Always Expedited Reports
FDA proposes to require manufacturers and applicants to submit to FDA ICSR for suspected adverse drug reactions received or otherwise obtained, whether foreign or domestic, that are the subject of an Always Expedited Report.
We believe that Always Expedited Reports should have clearly defined medical conditions (described using included MedDRA terms) that are subjected to the rulemaking process. FDA proposes to change the definition of "medically significant" event from "jeopardy and intervention," to "jeopardy and/or intervention." For example, the agency raises the issue of whether placement of an intravenous line - which is consistent with intervention for the underlying condition - would alone meet the proposed "Important Medical Event" (IME) criteria. Moreover, in addition to the proposed list of Always Expedited Reports, FDA proposes also to include "[a]ny other medically significant SADR FDA determines to be the subject of an Always Expedited Report."
BIO recommends that the final rule retain "jeopardy and intervention" for the IME criteria and any addition to the list be subject to notice-and-comment rulemaking. We also suggest that the MedDRA terms that should be used to define each of the medical conditions listed be clarified. Because of the complexity of MedDRA, an international consensus process (such as CIOMS or ICH, in conjunction with the Maintenance and Support Services Organization, MSSO) should develop and maintain the MedDRA groupings that describe these conditions. Use of included MedDRA term groupings will be especially important for those medical conditions that are often reported in vague terms or are otherwise poorly described by reporters in spontaneous reports. We recommend these groupings of included MedDRA terms be reviewed by FDA (preferably in cooperation with an international consensus body) with each new release of MedDRA and updated, as appropriate, to maintain the intended relevance. Timing of the required maintenance activities will be very important to keep "current" with international agreements regarding the implementation of each version of MedDRA. Finally, we believe that the list of conditions that trigger Always Expedited Reports should be negotiated on a product-by-product basis as part of an overall approach to risk management.
The agency has identified a list of adverse events it considers medically significant owing to the possibility they could jeopardize the patient or subject and/or require medical or surgical intervention to treat the patient or subject. These events would be submitted to the agency on an expedited basis whether expected, unexpected, serious, or non-serious. At this time, the agency receives all of these reports, but not in an expedited manner.
These events include congenital anomalies, acute respiratory failure, ventricular fibrillation, torsades de pointe, malignant hypertension, seizure, agranulocytosis, aplastic anemia, toxic epidermal necrolysis, liver necrosis, acute liver failure, anaphylaxis, acute renal failure, sclerosing syndromes, pulmonary hypertension, and pulmonary fibrosis.
Further, confirmed or suspected transmission of an infectious agent by a marketed drug or biologic would be reported to the agency on an expedited basis as this type of event would be considered a public health problem. The agency retains the right to add additional events to this list if it determines an event to be medically significant.
BIO strongly recommends that the FDA reconsider its proposed rule change and delete this section. We believe that the value of company analyses and/or medical interpretation of events is not appropriately weighted in this proposal. The proposed rule does not take into consideration the indication for use or the information contained in the product Investigator's Brochure or Package Insert. This will increase the number of expedited reports sent from the manufacturer to the agency for any given drug or biologic and will place an undue burden on both the agency and industry as potentially every suspected adverse drug reaction could be expedited.
III. D. 6. Follow-up Reports
The proposal would amend FDA's postmarketing safety reporting regulations to define the term "full data set," require manufacturers and applicants to submit a 30 day follow-up report to FDA for any expedited report that does not contain a full data set, and require submission of unexpected suspected adverse drug reactions with unknown outcomes in an expedited manner.
30-Day Follow-up -- Under the proposal, a complete data set would be required for postmarketing events that meet the following criteria: serious and unexpected events, "always expedited" events, and medication errors. If the complete data set is not available with the initial submission, a 30-day follow-up would be required even if new information has not been obtained and efforts to acquire the missing data are required.
45-Day Follow-up -- The proposal would require that manufacturers and applicants reporting an unexpected suspected adverse drug reaction with unknown outcome include in the expedited safety report the reasons for their inability to classify an suspected adverse drug reaction as either serious or non-serious (i.e., unknown outcome). For this purpose, manufacturers and applicants should include in the expedited report a chronological history of their efforts to determine the outcome of the suspected adverse drug reaction.
In the postmarket setting there are inherent difficulties in data collection and often tenuous willingness of "reporters" to provide information to sponsors. Included among data collection challenges from the non-professional public are expressed "privacy" issues, unwillingness to involve their physician -- often coupled with lack of sufficiently detailed medical information -- and non-response to either phone or written requests for information. Additional challenges from medically trained professionals include non-response, lack of interest in providing information, lack of time (either themselves or office staff), and concern that the report might potentially have unintended legal consequences. Given the above constraints, regardless of the number and type of attempts made, additional detail may never be obtained. We believe the added burden to sponsors of providing the proposed 30- and 45-day follow-ups will yield useful follow-up in only a very small number of cases. Efforts to improve the quality of data might better be spent in education of health care professionals and the general public about the value of adverse event reporting and their responsibility to report. Although regulatory agencies have more immediate control of sponsor/manufacturer activity, the ultimate gain may be significantly less than if change were initiated earlier in the process (i.e. public education). Both FDA and sponsor resources might be put to better use by focusing attention on this earlier point of contact. We recommend FDA focus on the adequacy of sponsor Standard Operating Procedures (SOPS) and Guidelines during audits rather than request real-time proof of due diligence.
III. D. 7. Supporting Documentation
The proposed rule would require manufacturers/applicants to include a chronological history of all attempts to secure a minimum data set and determine seriousness and expectedness of suspected adverse drug reactions in postmarketing expedited reports. In addition, the agency is requesting copies of documents such as discharge summaries and autopsy reports/death certificates. The agency is proposing this action to clarify the importance of acquiring complete information for serious suspected adverse drug reactions.
BIO believes the rationale to include information on the FDA form 3500A that applicants/manufacturers maintain in their corporate drug or biologic product safety files is unclear. Manufacturers/applicants have maintained records of their attempts to secure additional information for many years. Submitting copies of hospital discharge summaries and autopsy reports/death certificates for serious, unexpected adverse drug reactions creates patient record privacy issues both in the U.S. and overseas. In addition, we believe such requirements will impose additional burden and cost for purging these records to remove patient-identifiable information. Relevant information from these documents already is summarized and included in the appropriate boxes on FDA form 3500A. These records should be on file, but not included as part of the narrative summary. A field on the 3500A form indicating that the manufacturer/applicant has received a copy of the discharge summary or autopsy report/death certificate would be sufficient. BIO believes the narrative section should be limited to pertinent clinical details only. We recommend a written request be submitted to the manufacturer/applicant from the agency if a copy of any of the documentation retained in corporate files is required. Requiring submission of these documents is inconsistent with the Paperwork Reduction Act and ICH E2B.
Many manufacturers have worldwide reporting requirements, and it is appropriate to have one report satisfy both domestic and international requirements. No other regulatory authority worldwide has requested this type of information for inclusion in any adverse event report or aggregate safety report. BIO believes that preparing different reports for different regulators is redundant, resource wasteful, and does not advance public health.
III. E. 1-4. Traditional Periodic Safety Reports, Periodic Safety Update, Interim Safety Reports, Semiannual Submission of Individual Case Safety Reports
The FDA is proposing that industry submit TPSRs, PSURs or IPSRs depending on when the drug was first approved. If the approval date were prior to January 1, 1998, then a TPSR would be required. However, if the approval date were after January 1, 1998, then the PSUR format would be required. In addition, the proposed rule would require that a number of U.S.-only appendices be submitted with the PSUR. The suggested time frame for submission of these reports over the lifetime of the product is also different from the schedule set out in the ICH guidance.
It would seem prudent to add the requirements for IPSRs to the PSUR and submit all the information at one time. Moreover, the requirement that U.S. ICSRs be submitted using the MedWatch form is retained. BIO recommends the agency consider allowing industry to submit ICSRs using either the MedWatch or the CIOMS form.
FDA also proposes to remove the current requirement for submission of ICSRs in the TPSR. Instead, ICSRs would be included in a separate report submitted on a semiannual basis (ICSRs - semiannual submission). BIO recommends that the previous requirement remain in place - that ICSRs continue to be included in TPSRs and/or PSURs/IPSRs and that a new semiannual report not be required. We believe semi-annual submission of ICSRs would place undue burden on industry, since this would require a similar amount of effort as preparation of TPSR, PSURs, and IPSRs and would require retrieval of cases meeting TPSR/PSUR/IPSR criteria. The associated review and processing activities would essentially be equivalent to preparing a TPSR/PSUR/IPSR on a semiannual basis in addition to the proposed reporting schedule.
We also seek clarification as to the methodology suggested for increased frequency and the assessment of whether frequency of lack of efficacy reports is greater than what would be predicated by the premarketing clinical trials for the products, as an accurate denominator is not available. Finally, BIO believes that the addition of an increased frequency requirement is not in keeping with international harmonization.
BIO believes that public health is best served by safety staff with appropriate training and experience. We disagree with the proposal that only a licensed physician can be responsible for the content and medical interpretation of the data and information contained within the safety periodic reports. BIO believes manufacturers should be responsible for the medical content and allowed to determine appropriate health care professional qualifications required for the preparation of such reports.
In the spirit of embracing ICH E2C principles, we suggest updating section III.E.2.d. in accordance with the recently published Addendum to ICH E2C regarding Company Core Safety Information (CCSI). Overall, in an effort to provide global harmonization, line listings requirements should fully endorse ICH E2C; similarly the agency could provide more flexibility regarding the content of the appendices such that relevant information is contained in the body of the report as opposed to appendices.
Further, again in a move toward international harmonization, we suggest that the Core Data Sheet (CDS) requirement be aligned with that of the ICH E2C and Addendum. The applicant would be responsible for ensuring that all changes to the CCSI made during the reporting period are described in Section 4 of the PSUR. For 6-month and 1-year reports, the version of the CCSI in effect at the beginning of the period covered by the report should be used as the reference. For 5-year reports, the version of the CCSI in effect at the end of the period covered by the report should be used as the reference.
We seek clarification regarding the implementation of the TPSRs (as well as PSURs and IPSRs) as we believe that reporting of specific suspected adverse drug reactions such as "always expedited reports," "actual or potential medication errors," and "reports of class action lawsuits" will be captured in most vendor-designed company safety systems as such only after the effective date of the final rule. We therefore suggest that the content of summary tabulations as specified in the proposed rule be prospective.
BIO also suggests that the final rule include a recommended method for the determination of increased frequency as well as a recommended method for the assessment of changes in frequency of lack of efficacy, as an accurate denominator is unavailable.
Additional new sections are proposed to include details on location of safety records and contact information for the "licensed physician or licensed physicians responsible for the content and medical interpretation of the data and information contained within the TPSR" (or PSUR). BIO suggests (1) this information not be required as additional sections and (2) the report include the contact information of a health care professional responsible for the content of the document. We believe the manufacturer should be responsible for ensuring that the assigned personnel are qualified and that TPSRs, PSURs, and other safety-related records on a product be readily available to FDA upon request.
Periodic reporting for products approved for pediatric use should be treated as those for a special population, according to the ICH E2C Guideline. FDA proposes that applicants submit separate PSURs and IPSRs regardless of application approval date. However, for products that otherwise would be subject to periodic reporting, i.e., products approved for use in adults, we believe that it is appropriate to provide pediatric safety data and analyses in one report. Further, BIO believes that the requirement to provide data for special populations, such as pediatric age groups, should be applied prospectively because, as a practical matter, it will be extremely difficult (if not impossible) to identify the relevant patient subset (e.g., Class Action lawsuits, medication errors, etc.) because of the challenges of applying coding conventions retrospectively and making necessary changes to system capabilities. We believe that, in general, reporting periods should not be different simply because special populations are included.
III. E. 5. Reporting Requirements
Several sections of the proposed rule state that the submission of the single core report (excluding appendices) to multiple regulatory authorities would significantly reduce time spent preparing reports, thereby permitting more time for evaluation of the medical significance of any safety information reported. However, given the proposed additional requirements for PSURs beyond what is specified in the ICH guidelines as well as additional types of reports that the FDA proposes to require, the spirit and focus of the ICH E2C -- avoiding duplication of effort and ensuring that data are submitted to authorities with consistency (one standard, format/content), are minimized. We believe cases meeting PSUR criteria (with line-listing, summary tabulations) should be accepted throughout the three regions.
As currently drafted, BIO believes the proposed rule on PSURs does not allow for true international harmonization (for example, non-serious listed suspected adverse drug reactions are proposed to be included in this rule; they are not specified in ICH E2C). Additionally, the appendices are extensive and would contain a subset of the information already included in the core document (or unattainable information, such as information on U.S. microbial in vitro susceptibility data and the relationship of any observed changes to clinical outcomes for the reporting period). BIO recommends that the final rule not include a requirement for these appendices. However, if such a requirement is retained, we recommend that the time period allotted to prepare PSURs be extended from 60 to at least 90 days.
We agree with the possibility of alternative reporting frequency. However, we believe it is essential that this section of the rule also state that the Marketing Authorization Holder be able to negotiate the existing reporting cycle or frequency for the periodic safety reports, should the IBD be different from the U.S. birth date.
Finally, the proposed rule indicates that the PSUR would allow applicants to submit a single core document for products that have an approved application (i.e., BLA, NDA, ANDA). We seek clarification on how to handle products with multiple formulations or multiple active ingredient combinations, some of which are only approved outside the U.S.
V. Analysis of Impacts
BIO believes that FDA's impact analysis underestimates the resources that are needed to implement the proposed rule. In particular, some of the data upon which analyses related to pre-approval reporting are based appear to be erroneous. Further, as we stated in our introductory comments, many of the estimates - including those that calculate substantial "savings" to the industry - appear to be based on flawed assumptions or on data from unknown or unverifiable sources. We recommend the agency work with the affected industry to achieve numbers that more correctly reflect the true costs to industry, before it moves to finalize the proposal. We most strongly urge that the agency re-propose this rule, based on a re-calculation of costs and burdens, a re-evaluation of the harmonization of many of these proposals with those of other regulatory bodies, and a refined re-assessment of the true cost-benefit of the proposed requirements.
FDA proposes an effective date of 1 year after publication of the final rule for requirements related to coding suspected adverse drug reactions in individual case safety reports using MedDRA but that the effective date for all other components of the proposal become effective 180 days after publication in the Federal Register.
We believe that if the final rule is not significantly modified from the Proposed Rule, there will be an extensive impact on systems (modification of the existing safety databases to meet the specific requirements of the rule, requiring extensive and time-consuming validation) and processes (which, if implemented, will require training of employees, suppliers, investigators, etc.). Therefore, BIO strongly urges, first, that FDA re-propose the rule, as discussed above, and, second, that FDA set the effective date of the entire final rule at 18 months after publication in the Federal Register. We believe this amount of time is needed by industry to evaluate and implement the required changes. Such a delay in implementation is consistent with several recent examples of implementation of new federal regulations including an 18-month delay in implementing changes in FDA labeling requirements concerning aluminum in small volume parenterals and pharmacy bulk packages, "to give industry sufficient time to comply" with the regulations. (68 Fed. Reg. 32979; June 3, 2003)
We recommend the impact on systems be carefully evaluated to provide a clearer picture of the global resource requirements and the timing needed for full implementation of the final changes. In addition, full evaluation of the needed personnel and their qualifications and training requirements must be completed. Finally, BIO believes FDA's estimates of reasonableness and burden are low and recommend that these numbers be evaluated carefully to gain better insight into the implications of implementing the proposed rule throughout the pharmaceutical industry. BIO strongly urges the agency to re-evaluate components of the rule after reviewing all comments and having an opportunity to consider fully the total burden of new requirements, balanced against gains for the public health, and to re-propose the rule based on this finer tuned evaluation.
Throughout the proposed rule, the FDA emphasizes the need for manufacturers/ applicants to exercise due diligence, to continue to seek additional information from adverse event reporters, to document efforts to obtain additional information, to send numerous follow-up reports, and to send to the agency a myriad of documents formerly kept in manufacturers' files and summarized on FDA form 3500A. However, BIO does not believe that industry is likely to gather any more information or better quality information unless the agency (1) educates physicians, pharmacists, and other health care personnel as to the type and quality of information the agency is seeking and why this information is important and (2) makes adverse event reporting to manufacturers mandatory.
In conclusion, BIO believes that although the new proposals would reduce reporting obligation for certain non-serious events, this would not balance other additional efforts that would be required. FDA states that "Many of the postmarketing safety reports that FDA receives are complete and of high quality. Others are incomplete, of mediocre or poor quality or both, making it difficult to ascertain the significance of these reports."
BIO urges the agency to focus on those manufacturers who consistently do not use due diligence and who, upon FDA audit, have not demonstrated that processes are in place to address FDA concerns. We believe that those who adhere to recommendations will be penalized unduly under the new proposals.
BIO appreciates the opportunity to comment on this proposed rule. We are in full agreement with FDA on the need for quality safety reporting and the desirability of international harmonization of such reporting. These, we agree, are in the best interest of the industry, the agency, and the public. However, we are concerned that in a number of areas, this proposed rule over-reaches, frustrates the goal of international harmonization, and - most importantly - will not result in better or more useful safety information. As we stated earlier in our comments, we welcome the opportunity to continue to work with the agency to achieve the best quality and most useful system for safety reporting for the products our industry develops and manufactures.
Gillian R. Woollett, MA, DPhil
Vice President Science and Regulatory Affairs