FDA — Campylobacter Risk Assessment: Request for Reconsideration (RFR)

08/01/2004

April 16, 2003

VIA HAND DELIVERY and E-MAIL (w/o Attachments)

Office of the Ombudsman 
Food and Drug Administration 
5600 Fishers Lane 
Room 14B03, HF-7 
Rockville, MD 20857

Re: INFORMATION QUALITY APPEAL

Dear Sir/Madame:

Enclosed please find an appeal of the Center for Veterinary Medicine's ("CVM") denial ("CVM Denial") of a request ("Request") to FDA to correct certain information being disseminated by CVM ("Appeal"). A copy of the Request is contained in Tab B. The CVM Denial was signed by CVM Director Stephen F. Sundlof, dated March 22, 2003, but not received by the Animal Health Institute ("AHI" or "Requestor") until March 25, 2003. A copy of the CVM Denial is contained in Tab A.

AHI also requests a meeting with the person at FDA, or such other person as FDA may designate, who will be deciding the Appeal.

The Appeal is made pursuant to §515, P.L. Law 106-55 (1), and in accordance with FDA's "Guidelines for Ensuring the Quality of Information Disseminated to the Public" ("FDA Guidelines"). (2) The Request concerns a Campylobacter risk assessment (3), including specifically the model and data used in the risk assessment ("Vose Risk Assessment").

The transmittal letter accompanying the Request specifically asked that the Request be handled by the Office of the Ombudsman ("Ombudsman"), or alternatively by the Secretary of the Department of Human Services, or the Office of Management and Budget. This request was necessitated, in part because of CVM's repeated failures to correct the Vose Risk Assessment. Notwithstanding this request the Ombudsman forwarded the Request to CVM for reply, purportedly in compliance with 21 CFR § 10.75. It is from this denial that AHI appeals.

In summary, this appeal to correct the Vose Risk Assessment is both timely and appropriate and warrants that the requested relief be immediately considered and granted for the following reasons. First, the deferral of a decision on AHI's Request until after the conclusion of the Baytril Hearing constitutes an effective denial of the Request for correction under the Data Quality Act. Secondly, the decision on the Request by the employee who made the initial decision (i.e., Dr. Sundlof) also constitutes a denial of the Request, as well as a failure to follow FDA's regulations at 21 CFR § 10.75. Thirdly, in light of the representations made by CVM in the Baytril Hearing regarding the inapplicability of the FDA Guidelines to the Vose Risk Assessment, it is evident that CVM's deferral is in fact a denial, is improper, is at variance with FDA's procedures and constitutes little more than a thinly veiled attempt to avoid the requirements of the Data Quality Act.

In the event the Ombudsman concludes for any reason that FDA should not or cannot review the Appeal, including because of the pendency of the administrative hearing covered by FDA Docket 00N-1571, Requestor requests that the Appeal be considered by the HHS or OMB. Either would appear appropriate in light of the fact that §515, P.L.106-55 directs OMB to issues guidance to provide policy and procedural guidance to Federal agencies for ensuring and maximizing the quality, objectivity, utility, and integrity of information (including statistical information) disseminated by Federal agencies.(4) OMB's has issued guidelines (5)  and both FDA's and HHS's Guidelines (6)were issued pursuant to the OMB guidance.

Should you have any questions or need any further information please feel free to contact me or our counsel, Robert Nicholas. I would appreciate if copies of all correspondence would be sent to Mr. Nicholas as well.

Sincerely,

Kent D. McClure, DVM, JD

cc: Robert B. Nicholas, Esq.

McDermott, Will & Emery 
600 13th Street, N.W. 
Washington, DC 20005

________________________________________

1. Section 515 of the Treasury and General Government Appropriations Act for Fiscal Year 2001; P.L. 106-554, § 515 (2000).

2. The Guideline can be found on CVM's website at: http://www.fda.gov/infoquality/fda.html.

3. "The Human Health Impact of Fluoroquinolone Resistant Campylobacter Attributed to the Consumption of Chicken," Food and Drug Administration, Center for Veterinary Medicine, (October 18, 2000, Revised January 5, 2001) ("Vose Risk Assessment"). The Vose Risk Assessment was finalized, beginning in about January 2001. Since that time the assessment has been widely relied upon, quoted, and otherwise referenced and disseminated many times in additional publications and presentations by CVM, the Centers for Disease Control and Prevention ("CDC"), other federal agencies, and others. The Vose Risk Assessment has been since at least October 1, 2002, and continues presently to be available on CVM's Website at: http://www.fda.gov/cvm/antimicrobial/Risk_asses

4. Section 515 directs OMB to formulate and issue guidelines that "provide policy and procedural guidance to Federal agencies for ensuring and maximizing the quality, objectivity, utility, and integrity of information (including statistical information) disseminated by Federal agencies." P.L. 106-554, § 515 (2001).

5. Guidelines for Ensuring and Maximizing the Quality, Objectivity, Utility, and Integrity of Information Disseminated by Federal Agencies (the "OMB Guidelines"), 67 Fed. Reg. 8452-8460 (February 22, 2002).

6. "Guidelines for Ensuring and Maximizing the Quality, Objectivity, Utility, and Integrity of Information Disseminated to the Public", HHS, (October 1, 2002), http://www.hhs.gov/infoquality.

 

April 16, 2003

VIA HAND DELIVERY and E-MAIL (w/o Attachments)

Office of the Ombudsman 
Food and Drug Administration 
5600 Fishers Lane 
Room 14B03, HF-7 
Rockville, MD 20857

Information Quality Appeal

This letter appeals the Center for Veterinary Medicine's ("CVM") March 20, 2003 denial of a request by the Animal Health Institute ("AHI" or "Requestor") to correct certain information circulated or otherwise disseminated by CVM ("CVM Denial").(1) This appeal is filed under Section 515 of Public Law 105-554 (2)and pursuant to the FDA "Guidelines for Ensuring the Quality of Information Disseminated to the Public" (the "FDA Guidelines").(3) In accordance with the FDA Guidelines, please find attached a copy of the CVM Denial (Tab A) together with a copy of the original request (Tab B). The reasons for this appeal and for the requested corrections are set forth in this letter and in the original AHI request of January 23, 2003 ("AHI Request" or "Request").

This letter is also for the purpose of again requesting a meeting with FDA. Subsequent to the filing of the AHI Request, AHI requested a meeting with FDA to discuss its details. In the CVM Denial, CVM declined to meet with AHI. AHI renews its request for a meeting to discuss the nature of its complaint with the FDA official responsible for the appeal.

Introduction and Background

AHI Request.

The AHI Request seeks the immediate correction or deletion of information disseminated by CVM that is in error, including generally information related to the potential estimated risk to human health from the use of enrofloxacin in poultry. More specifically, and as set forth in the AHI Request, the information that is in error and that needs to be corrected is the methodology, exposition (which currently inaccurately represents what was actually done), data, and conclusions of the Campylobacter risk assessment conducted for CVM by David Vose, and subsequently adopted by CVM. The risk assessment purports to be a quantitative risk assessment. It states that:

To evaluate the human health impact of antimicrobial use in animals, the FDA Center for Veterinary Medicine (CVM) developed a quantitative risk assessment model. The risk assessment was intended to estimate the risk to human health from antibiotic resistant food borne pathogens associated with the domestic use of antimicrobials in food producing animals. Specifically, a mathematical model was derived to relate the prevalence of fluoroquinolone resistant Campylobacter infections in humans associated with the consumption of chicken to the prevalence of fluoroquinolone resistant Campylobacter in chickens.(4)

Title:

"The Human Health Impact of Fluoroquinolone Resistant Campylobacter Attributed to the Consumption of Chicken," Food and Drug Administration, Center for Veterinary Medicine. ("Vose Risk Assessment").

Document Date:

October 18, 2000, Revised January 5, 2001.

Availability:

The Vose Risk Assessment was initially available on CVM's website in final form, beginning in about January 2001. Since that time the Vose Risk Assessment has been widely relied upon, quoted, and otherwise referenced and disseminated many times in additional publications and presentations by CVM, the Centers for Disease Control and Prevention ("CDC"), other federal agencies, and others. The Vose Risk Assessment has been since at least October 1, 2002, and continues presently, to be available on CVM's Website at: http://www.fda.gov/cvm/antimicrobial/Risk_asses. It has also been discussed and otherwise disseminated by various CVM employees, including CVM Director Sundlof.

The AHI Request sought correction of the Vose Risk Assessment because the risk assessment fails to meet the standards for information quality as set forth in the FDA Guidelines in terms of utility, objectivity and integrity. In addition, AHI argued that the Vose Risk Assessment does not meet the required higher standards set forth in the FDA Guidelines for "Influential Information" and "Risk Assessment."

The AHI Request was made pursuant to section 515 of the Treasury and General Government Appropriations Act for Fiscal Year 2001 (Public Law 106-554), which directs the Office of Management and Budget ("OMB") to issue government-wide regulations that "provide policy and procedural guidance to Federal agencies for ensuring and maximizing the quality, objectivity, utility, and integrity of information (including statistical information) disseminated by Federal agencies," (5)as well as in accordance with the FDA Guidelines, which (in part) explain the administrative mechanisms that are in place to enable persons to seek and obtain correction of information maintained and disseminated by FDA that they believe does not comply with the OMB and FDA Guidelines.

The administrative mechanisms (or "Procedures for Submitting Complaints") specifically identified in the FDA Guidelines include: (i) 21 C.F.R. § 10.75, which allows any interested person to obtain formal review of any agency decision or action by raising the matter with the supervisor of the employee who made the decision; (ii) FDA's regulations for dispute resolution during the application review process (21 C.F.R. § 312.48, 21 C.F.R. § 314.103, 21 C.F.R. § 814.42(d), 21 C.F.R. § 814.46(c), and 21 C.F.R. § 808.25(e)); and (iii) 21 C.F.R. § 5.200, which provides for the establishment of an agency ombudsman to help mediate agency-related disputes.

In light of CVM's history of unwillingness to modify the risk assessment or to provide any constructive response to many criticisms, (6)the AHI Request was submitted directly to the Office of the Ombudsman (FDA), rather than to CVM. Additionally, since the CVM Director Sundlof "initially approved and disseminated the risk assessment at issue,"(7) and also because of the pendency of the Baytril hearing, AHI further requested that, "[i]n the event the Ombudsman concludes for any reason that FDA should not or cannot review the Request, including because of the pendency of the administrative hearing covered by FDA Docket 00N-1571, Requestor requests that the Request be considered by the Office of the Secretary of the Department of Health and Human Services ("HHS") or the Office of Management and Budget ("OMB")." (8)

CVM Denial.

Notwithstanding a specific request that the AHI Request be handled by the FDA Ombudsman or else referred to the Secretary of HHS or OMB, the AHI Request was referred by the Ombudsman to, and ultimately decided (and denied) by, the CVM Director. The CVM Denial, signed by Stephen F. Sundlof, DVM, Ph.D., generally deferred a decision on the AHI Request until "within 60 days of the final decision in the hearing on the proposal to withdraw the approval of the new animal drug application (NADA) 140-828 for enrofloxacin (Baytril 3.23% Concentrate Antimicrobial Solution) ["Baytril Hearing"]." Dr. Sundlof explained that "[s]ince, as Center Director, I initially approved and disseminated the risk assessment at issue, I am responding to your request for correction and will reevaluate that decision based on the information submitted as part of your request as well as other information in the administrative file."(9) He also stated that his evaluation of AHI's request would be based on the procedure outlined in 21 C.F.R. § 10.75.

Dr. Sundlof declined to correct the Vose Risk Assessment, as requested by AHI, on the grounds that to render a decision on AHI's Request within 60 days of the Request (10) and before conclusion of the Baytril Hearing would be improper, because: (i) the agency's processes might be disputed, as numerous overlapping issues are already before the Administrative Law Judge ("ALJ"); (ii) waiting until the Commissioner has rendered a final decision with respect to the proposal to withdraw approval of the NADA would promote consistency overall; (iii) processing the request for correction while the hearing is ongoing would make inefficient use of resources as the validity and reliability of the risk assessment are at issue in the hearing; and (iv) AHI will not be unduly prejudiced by the proposed delay.

Appeal

For the reasons set forth herein, the deferral of a decision on AHI's Request until after the conclusion of the Baytril Hearing - a time that is at best six months from now, and possibly significantly later - constitutes a denial of the Request for correction under the Data Quality Act, and makes this appeal both timely and appropriate. The decision on the Request by the employee who made the initial decision (i.e., Dr. Sundlof), and not by the supervisor of such employee, also constitutes a denial of the Request, as well as a failure to follow FDA's regulations at 21 CFR § 10.75. In light of the representations made by CVM in the Baytril Hearing regarding the inapplicability of the FDA Guidelines to the Vose Risk Assessment, it is evident that CVM's deferral is in fact a denial, is improper, is at variance with FDA's procedures and constitutes little more than a thinly veiled attempt to avoid the requirements of the Data Quality Act, as specified below. Therefore, AHI renews its Request to correct the Vose Risk Assessment and asks that the requested relief be immediately considered and granted at this time, either as an initial request for correction by Dr. Sundlof's supervisor, or as an appeal of Dr. Sundlof's denial of the AHI Request.

1. Because Review of an Agency Decision Under § 10.75 Must Be Confined to the "Supervisor" of the Employee Who Made the Decision, CVM's Denial of AHI's Request Was Improper

The CVM Denial states that Dr. Sundlof's evaluation of the AHI Request would be based on the procedure outlined in 21 C.F.R. § 10.75. 21 C.F.R. § 10.75 provides that any interested person(11)may obtain review of any agency decision by raising the matter with the supervisor of the employee who made the decision.(12)If the issue is not resolved at the supervisor's level, then the interested person may request that the matter be reviewed at the next higher supervisory level. This process may continue through the agency's entire supervisory chain of command through the Centers to the Deputy Commissioner for Operations, and then to the Commissioner of Food and Drugs.(13)

The key to FDA's compliance with this regulatory provision is that review of the agency decision be by the actual supervisor of the employee who made the decision. In general, the question of who qualifies as a supervisor under the regulation "follow[s] the established agency channels of supervision or review for the specific matter involved."(14) Thus, if the review of AHI's request is by someone other than the supervisor of the employee who made the decision (e.g., a lower level employee, or the same employee who originally made the decision), then the agency review of the decision will be out of compliance.

In the present matter, CVM's consideration and Denial of AHI's Request was improper because the same person (i.e., Dr. Sundlof) who was given responsibility for deciding whether to correct the Vose Risk Assessment was the very person who, in his words, "initially approved and disseminated the Vose Risk Assessment."(15) This exercise of authority clearly is not permitted under the regulation.

In the current matter, and in light of the goal of the Data Quality Act to provide a person with an opportunity for an objective assessment of allegedly inaccurate data disseminated by a federal agency, the impropriety of referring the Request to CVM for the initial decision cannot be more clear. CVM is the FDA party to the Baytril Hearing and in that capacity seeks to defend the reliability of the Vose Risk Assessment. In this circumstance there is no reasonable possibility that CVM can render an impartial opinion on the reliability of the Vose Risk Assessment under the FDA Guidelines.

One need only look to the CVM Denial to see that the handling of the Request is consistent with CVM's prior failures to respond scientifically to what we believe are legitimate scientific criticism of the Vose Risk Assessment. One need only look to CVM's posture in the Baytril Hearing, regarding the irrelevance of the FDA Guidelines to the Baytril Hearing, to see that the CVM Denial is yet another effort to avoid addressing legitimate scientific criticisms of the Vose Risk Assessment, as further discussed below. Additionally, one need only look to CVM's current draft risk assessment guideline(16) to see that CVM no longer, except in the Baytril Hearing, supports the Vose Risk Assessment framework. In fact it was just for these reasons, and to avoid a result that would be a foregone conclusion, that the Request was initially filed with the FDA Ombudsman.

The purpose of the establishment of FDA's Ombudsman was to assure aggrieved parties an opportunity for a fair hearing, not to require meaningless (and expensive) gestures.(17)When a party such as CVM has demonstrated repeatedly that it will not respond to legitimate scientific criticism, then requiring meaningless gestures, such as the referral of the Request to CVM, is neither consistent with the spirit of the law establishing the role of the Ombudsman, with 21 CFR § 10.75, or with providing an objective review under the FDA Guidelines. Accordingly, and because the "agency has a substantial interest in retaining the ability to review decisions by subordinate officials" under § 10.75, Stauffer Chemical Co. v. FDA, 670 F.2d 106, 108 (9th Cir. 1982), the denial of AHI's Request must be reconsidered by Dr. Sundlof's supervisor on scientific and policy matters, the FDA Ombudsman, the Secretary of HHS, or OMB on this basis alone.

2. CVM's Deferral of the Decision Under 21 CFR § 10.75 is Tantamount to a "Refusal to Act" Which Constitutes a Violation of Sections 555(b) and 706(1) of the Administrative Procedure Act

The Administrative Procedure Act ("APA") requires that each agency "proceed to conclude a matter presented to it within a reasonable time." 5 U.S.C. § 555(b). Under section 706(1), reviewing courts are vested with the authority to "compel agency action unlawfully withheld or unreasonably delayed." Although an administrative agency has considerable deference in establishing a timetable for completing its proceedings, such discretion is not unbounded "since the consequences of dilatoriness may be great." Cutler v. Hayes, 818 F.2d 879, 896 (D.C. Cir. 1987). That is,

There must be a 'rule of reason' to govern the time limit to administrative proceedings. Quite simply, excessive delay saps the public confidence in an agency's ability to discharge its responsibilities and creates uncertainty for the parties, who must incorporate the potential effect of possible agency decisionmaking into future plans.

Public Citizen Health Research Group v. FDA, 724 F.Supp. 1013, 1018 (D.D.C. 1989) (citing Potomac Elec. Power Co. v. ICC, 702 F.2d 1026, 1034 (D.C. Cir. 1983)).

In determining whether agency action has been unreasonably delayed, the court should "ascertain the length of time that has elapsed since the agency came under a duty to act…[consider] the reasonableness of the delay…and…examine the consequences of the delay. Id. (quotingCutler, 818 F.2d at 897-98). Indeed, delays which might be reasonable in the context of economic regulation are "less tolerable when human lives are at stake." Id. (quoting Public Citizen Health Research Group v. FDA, 702 F.2d 1150, 1157 (D.C. Cir. 1983)).

The fact that an agency proceeding may be ongoing does not affect the possibility of judicial intervention to address the delay. Sections 555(b) and 706(1) of the APA give courts the clear authority to review ongoing agency proceedings to ensure that they resolve the questions in issue within a reasonable time. Public Citizen, 702 F.2d at 1158; Potomac Elec. Power Co., 702 F.2d at 1034; MCI Telecommunications Corp. v. FCC, 627 F.2d 322, 340 (D.C. Cir. 1980). Indeed, "[a]t some point administrative delay amounts to a refusal to act, with sufficient finality and ripeness to permit judicial review." Environmental Defense Fund, Inc. v. Hardin, 428 F.2d 1093, 1100 (D.C. Cir. 1970).

The facts here speak for themselves. As noted, supra n. 6, CVM has been aware of the substantial criticism and deficiencies in the Vose Risk Assessment at least since December 1999, when it held a workshop on a draft of the Vose Risk Assessment. CVM did not respond, or did not fully respond, to the many comments and criticisms from the workshop. The final Vose Risk Assessment remains largely unchanged from the draft risk assessment. Between October 31, 2000 and February 21, 2001, many additional comments were submitted to FDA Docket 00N-1571 in reply to the NOOH, including extensive comments on why the final Vose Risk Assessment does not comport with the NAS paradigm and comments raising questions about the validity, objectivity, data integrity, and usefulness of the Vose Risk Assessment. These comments also remain largely unresponded to date, including comments on why the Vose Risk Assessment does not comport with the NAS paradigm. CVM's action under 21 CFR § 10.75 to "defer" a response to the AHI Request for a further six months to a year or more, particularly given the long history of CVM's failure to respond to requests for correction of the Vose Risk Assessment, constitutes an effective denial of the AHI Request.

3. CVM's Stated Position That the Baytril Hearing Does Not Afford Any Opportunity to Use the FDA Guidelines to Challenge The Vose Risk Assessment Effectively Precludes Consideration of the Vose Risk Assessment Under the Data Quality Act      

CVM's Denial seeks to justify waiting until 60 days after the conclusion of the Baytril Hearing to respond substantively to the AHI Request. In light of CVM's position in the Baytril hearing that the FDA Guidelines are to be "disregarded" by Judge Davidson, the CVM Denial effectively means that there will not be an evaluation of the AHI Request under the Data Quality Act.

In reply opposition to Bayer and AHI's joint motion to strike the Vose Risk Assessment as unreliable, CVM asserted,

[AHI's] Motion, pages 14 to 19, seeks to strike testimony and evidence submitted by CVM through FDA's Guidelines for Ensuring the Quality of Information Disseminated to the Public ("FDA Guidelines" or "data quality guidelines"), 67 Fed. Reg. 61343 (Sept. 30, 2002). CVM believes that this administrative hearing does not afford [AHI] any opportunity to use the FDA Guidelines to challenge CVM's testimony or evidence.

Center for Veterinary Medicine's Response to Bayer and AHI's Motions to Strike Written Direct Testimony and Evidence, FDA Docket 00N-1571, February 10, 2003, ("CVM Opposition") at 6 (emphasis supplied).

The CVM Opposition further makes clear CVM's intent to deny an evaluation of the Vose Risk Assessment under the FDA Guidelines and to compel AHI to utilize only the procedure in 21 CFR § 10.75. CVM states further:

Third, should [AHI] wish to seek relief under the FDA Guidelines, [AHI] must follow the procedures set forth in the FDA Guidelines. The FDA Guidelines explain the administrative mechanisms that are in place to enable members of the public to seek correction of information maintained and disseminated by FDA that they believe does not comply with the data quality guidelines. The administrative procedure contained in the FDA Guidelines describes FDA's intent to use existing mechanisms to address requests for information correction ("complaints") from the public. The procedures for submitting complaints, FDA's response to complaints, and requests for reconsideration are outlined in the data quality guidelines. Any agency response that may be afforded to [AHI] pursuant to the FDA Guidelines is limited by the guidance itself.

* * * *

Fourth, …. Assuming the propriety of AHI's submission, FDA's review of AHI's complaint will conform to the administrative procedure outlined in the FDA Guidelines.

CVM Opposition at 7-8 (emphasis supplied). CVM concludes it argument with the following statement:

CVM's position is that the ALJ's evaluation of expert testimony in this hearing should not be based on the FDA Guidelines. Any argument put forth by [AHI] relying on the FDA Guidelines should be disregarded…"

CVM Opposition at 9 (emphasis supplied).

Having taken the position in the Baytril Hearing that Judge Davidson cannot consider the FDA Guidelines in evaluating the reliability of the Vose Risk Assessment, and, further, that the proper avenue for evaluation of the AHI Request is in accordance with 21 CFR § 10.75, Dr. Sundlof then seeks to shut off any evaluation of the assessment under the FDA Guidelines invoking § 10.75, the very same procedure offered by Dr. Sundlof as the exclusive avenue for resolution of the Request . Dr Sundlof states:

Second, the Commissioner's decision in the hearing will be binding on the agency. Our sequence for issuing a decision on your request for correction will enable me to ensure that the decision is consistent with the Commissioner's decision at the hearing. Because most of the issues raised in your request for correction have also been raised in the hearing, that will address, inform, and, perhaps, partially or fully resolve those issues.

CVM Denial at 5.

In other words, only after the ALJ has decided on the validity of the Vose Risk Assessment (assuming he does), without considering reliability under the standards in the FDA Guidelines, and only after the Commissioner affirms or modifies the ALJ's initial decision (assuming a request for reconsideration is filed by a party), only then will CVM consider the Request, bound by the decision already made on the Vose Risk Assessment in the Baytril Hearing.

Quite plainly, CVM's position in the Baytril Hearing, coupled with its March 20 Denial, all but assures that there will be no evaluation of the Vose Risk Assessment under the standards of the Data Quality Act, as implemented by the OMB Guidelines and the FDA Guidelines. Additionally, to the extent that there is any possibility of such an evaluation, CVM's position makes impossible that the consideration of the Request will be timely and, therefore, CVM's actions effectively constitute a denial of the Request.

Nothing in the OMB Guidelines or underlying statute even suggests that such an interpretation and proceeding - viz, requiring that a complaint be ignored until after a decision has been made, and then requiring that that decision be binding in responding to the complaint - is contemplated or allowed. Indeed, if allowed to stand, CVM's above position in the Baytril Hearing, coupled with the CVM Denial, effectively undercuts the Data Quality Act's purpose of providing a set of standards to evaluate the reliability of information disseminated by a federal agency, including information intended to be used in a regulatory proceeding.

he FDA Guidelines recognize explicitly the applicability of the FDA Guidelines to data intended to be used or actually used in a regulatory process. They require, except in limited circumstances, resolution of a complaint under the Data Quality Act before final agency action. For example:

In cases where the agency disseminates a study, analysis, or other information prior to the final agency action or information product,requests for correction will be considered prior to the final agency action or information product in those cases where in the agency's judgment issuing an earlier response would not unduly delay issuance of the agency action or information product and the complainant has shown a reasonable likelihood of suffering actual harm from the agency's dissemination if the agency does not resolve the complaint prior to the final agency action or information product.

FDA Guidelines at 16 (emphasis supplied). (18)

The requirement for resolution under the standards for quality of data disseminated or used by an agency before final agency action is clearly consistent with the purpose of the Data Quality Act. To have the standards of the FDA Guidelines applicable only after final agency action, rather than to have the standards guide agency action, would likely vitiate the Data Quality Act in most instances. It also would appear to run counter to the plain meaning requirement in the OMB Guidelines that "agencies…establish administrative mechanisms allowing affected persons to seek and obtain…correction of information…"(19) In the instant matter, where CVM has taken the position that the FDA Guidelines do not apply to the Baytril Hearing, the CVM approach will all but guarantee that the Data Quality Act will be of no effect.

As discussed infra §§ 4-5, a decision at this time is appropriate since evaluation of the qualifications to consideration of the Request prior to final agency action strongly favor immediate consideration. A reply to the Request is not reasonably likely to delay decision on the Baytril Hearing. There has been actual harm to Requestor by a failure to resolve the Request to date and proper adherence to § 10.75, and there continues to be a reasonable likelihood of further actual harm to the Requestor by the deferral of correction of the Vose Risk Assessment.

4. Undertaking a Proper Review of AHI's Request Under § 10.75 Will Not Disrupt the Hearing or Waste Agency Resources      

In the CVM Denial, Dr. Sundlof posits that currently undertaking a review of AHI's request to correct the risk assessment will disrupt "agency processes" and result in "an inefficient use of resources" because:

1) "Acting on [AHI's] request for correction while the hearing is in progress could interfere with Judge Davidson's adjudication of the issues that overlap [AHI's] request for correction"; and

2) "It is best to take advantage of [the] existing [formal evidentiary hearing] process, especially since some of the staff members who would be needed to assist with the review of [AHI's] request for correction [sic] also devoting much of their time to the proceedings of the hearing."

In light of AHI's position, supra § 1, that Dr. Sundlof is not the appropriate person to evaluate and decide the AHI Request, and in any event is not the person to decide this appeal, Dr. Sundlof's reasons for "deferring" decision on the AHI Request are arguably moot. There is no indication that a substantive reply to the AHI request by FDA, HHS or OMB would be subject to the same assertions. Indeed, given the sixty-day timeframe for a reply, there is no reason to believe that appropriate consideration of the Request would delay or otherwise interfere with the hearing, since conclusion of the evidentiary part of the hearing is almost certainly more than sixty days away. Additionally, and as envisioned by the FDA Guidelines, it would be entirely appropriate for the ALJ's decision on the Vose Risk Assessment to be informed by FDA's evaluation under the FDA Guidelines. Clearly, it makes greater sense, and is a more efficient use of FDA resources for the ALJ to have before him FDA's evaluation of the Vose Risk Assessment under the Data Quality Act when the ALJ makes the initial decision than to have the Commissioner weigh the risk assessment under the FDA Guidelines together with the ALJ's initial decision when the Commissioner makes the final decision in the Baytril matter.

Notwithstanding the above AHI position, even if one were to assume that CVM is the appropriate party to review the AHI Request, there is no reason to believe that a decision on the Request need delay the conclusion of the Baytril Hearing, otherwise interfere with the Baytril Hearing, or make an inefficient use of CVM resources.

Beyond the above-quoted conclusory statements, Dr. Sundlof fails to elucidate why resolution of the 'Vose Risk Assessment issue' pursuant to AHI's Request (versus deferral until after the Baytril Hearing) will disrupt the hearing or result in an imprudent allocation of agency resources, particularly where (as here) the same CVM staff members, the same book of evidence/testimony, and the same issues would be involved in both. (20) Nor does he clarify why, in the past, CVM has expressly recognized the importance of resolving disputes under § 10.75prior to the conclusion of an animal drug withdrawal hearing. (21)

Since CVM has been aware of the scientific criticisms of the Vose Risk Assessment, in most instances for years, as acknowledged in the CVM Denial, and believes it has responded to the criticisms in the Baytril Hearing as well as in the public and other meetings, it is hard to imagine that CVM can seriously contend, as it does that "[d]uplicating the processing of these issues in the hearing and in your request for correction, especially when they are so closely related, would go against CVM's duty to use its resources prudently." Indeed, it might have been more prudent for CVM to have responded openly and directly to the many scientific concerns when they were first raised, and resources of everyone would have been more prudently used had CVM done so. Also, it would still be prudent to do so, since CVM after these many years surely and readily has developed its answers to the scientific concerns raised in most, if not all instances years ago.

In addition, Dr. Sundlof over-dramatizes the impact a review of AHI's Request will have on the steadiness of the hearing as a whole. Clearly, much remains to be accomplished before a final decision will be reached in the present matter. As Dr. Sundlof observed:

Cross-examinations [are] scheduled to begin April 28, 2003 and briefs will be scheduled and filed after the testimony is all on the record. Judge Daniel J. Davidson will issue a decision as soon as possible after the filing of briefs and oral argument. This initial decision is subject to review by the Commissioner of Food and Drugs, either upon request of any party to the hearing or upon the Commissioner's own initiative. The Commissioner is to issue a final decision as soon as possible after the filing of briefs and any oral argument. If there is no appeal from or review of the initial decision, it becomes the agency's final decision by operation of law. (CVM Denial at 2.)

Therefore, any risk of significant delay or other problems caused by a re-allocation of agency resources at this juncture is highly unlikely. In fact, only CVM's decision to wait sixty days to defer consideration of the Request opens any question about the possible impact on the timing of the Baytril Hearing. However, in light of CVM's position before Judge Davidson that the FDA Guidelines are to be disregarded by Judge Davidson, it is difficult to understand Dr. Sundlof's statement that such a CVM evaluation of the Vose Risk Assessment could delay or otherwise impact the Baytril Hearing. If CVM is correct and Judge Davidson cannot consider the FDA Guidelines, then a substantive reply to the AHI Request will not be considered by Judge Davidson and the reply could not possibly impact the Baytril Hearing. Conversely, if Judge Davidson is ultimately to consider the FDA Guidelines, then it would be desirable and constitute efficient use of Agency resources for hmimn to be able to do so soon.

Since FDA's procedural rules governing the Baytril Hearing contemplate admission of new evidence at any time prior to final decision, (22) an evaluation of the Vose Risk Assessment under the Data Quality Act would appear to be relevant and admissible. Given the fact that FDA will have conducted the evaluation, it is equally hard to imagine that CVM would be in a position to disclaim the reply. In light of this, CVM's claim that undertaking a proper review now of AHI's Request under § 10.75 would delay or interfere with the hearing cannot be taken as more than another attempt to prevent an evaluation of the Vose Risk Assessment.

4. CVM's Denial of AHI's Request Will Prejudice AHI

Dr. Sundlof concludes that "the fact that CVM will not issue a decision within 60 days will not prejudice AHI" since "AHI and others have already used multiple opportunities to raise the same issues for CVM's consideration" and "almost all the arguments and information [AHI] has presented are being actively considered by FDA in the context of the hearing…" The stark reality, however, as detailed above, is that CVM's handling of AHI's Request, if left standing, effectively precludes AHI's only opportunity to correct information in the Vose Risk Assessment by the standards established under the Data Quality Act before the conclusion of the hearing. This chain of events could well lead to the Vose Risk Assessment being used to support withdrawal of the NADA for Baytril even though it fails to meet the requirements of the Data Quality Act and the implementing FDA Guideline. This presents a reasonable likelihood of actual harm to AHI, a hearing participant opposing the withdrawal.

CVM dismisses the need for prompt evaluation under the Data Quality Act and implementing FDA Guideline of scientific criticisms of its risk assessment on the grounds that there will be no prejudice if a decision is delayed "because [the] concerns raised are being actively considered." As detailed in the Request, CVM's actions have already had and will continue to have a significant adverse impact on many parties, including AHI, drug manufacturers of antibiotics for veterinary use (including both members and non-members of AHI), veterinarians, food producers, and CVM's scientific credibility.

Recently, an important effort by the International Cooperation on Harmonisation of Technical Requirements for Registration of Veterinary Medicinal Products ("VICH") to harmonize international standards for pre-approval information for food producing animals with respect to antimicrobial resistance was essentially derailed when CVM precluded the publication of the proposal, despite CVM's having previously agreed to the publication in 2001 (Tab C).(23) Although the proposal was successfully published in the European Union and Japan, CVM withheld its publication in this country, purportedly because CVM was in the process of developing a guidance on the pre-approval evaluation of the safety of antimicrobial animal drugs with regard to their microbiological effects on bacteria of human health concern, and first needed to verify the consistency of the proposal with this guidance. In addition, CVM later identified that its "legal counsel, who are also leading the hearing process to withdraw the poultry fluoroquinolone, are not in agreement with how the resistance issue is portrayed in the guidance introduction."(24) CVM sought deletion of already agreed-to language, including reference to limits on quantification of the extent to which animals are a source of human exposure to antimicrobial-resistant microorganisms. The failure to resolve the Vose Risk Assessment quality and methodologic issues is causing CVM to delay action on the VICH Guideline out of fear that it would be used to undermine the utility of the Vose document, leading to the conclusion that CVM simply does not want to resolve questions about the Vose Risk Assessment. (Tab C)

Additionally, CVM has continued to attempt to disseminate its flawed risk assessment model, even by encouraging other countries such as Japan to adopt the model. For example, CVM Director Sundlof recently expressed a willingness to visit Japan and assist members of the Ministry of Agriculture, Forestry and Fisheries (MAFF) to better understand the risk assessment process currently in use in the United States relative to the evaluation of the safety of anti-microbials (Tab D). Clearly, such promotional efforts, particularly by U.S. government officials overseas, can cause actual harm to AHI and its members. Exportation of the Vose Risk Assessment model, that we believe fails to follow the NAS paradigm, fails to meet the requirements under the Data Quality Act, and contains egregious errors and omissions, has a reasonable likelihood of causing direct irreparable harm. AHI and its members are placed in the position of asserting the value of the NAS paradigm and the standards under the Data Quality Act against CVM before a foreign government. Surely, this is better resolved expeditiously within FDA.

The dissemination of the flawed model and data in the Vose Risk Assessment creates disincentives for development of veterinary antibiotics. It also makes impossible the ability of AHI to effectively represent its members. Dissemination of the Vose Risk Assessment provides inaccurate methods, data, and conclusions, which opponents of antimicrobial use in veterinary medicine then use to put pressure on producers of food products to stop using veterinary antibiotics. An example of the impact and ongoing effect of the impact and misuse of the Vose Risk Assessment, known to CVM, is contained in the sworn testimony of CVM's witness Koziol submitted in the Baytril Hearing. (Tab D). Such actions adversely impact AHI and its member companies directly. AHI and its members are placed in the position of asserting the value of the NAS paradigm and the standards under the Data Quality Act against CVM with the media and food producers. Surely, this is better resolved expeditiously within FDA.

It is strange for an agency to take the position that it may disseminate inaccurate information for two years without responding to legitimate scientific criticism, and then conclude (perhaps because the damage has been done already by the incorrect information) that waiting an additional six months or more will not matter "because [the] concerns raised are [still] being actively considered."

Therefore, unless AHI's Request is reconsidered by the Dr. Sundlof's supervisor, the Ombudsman, the Secretary of HHS, or OMB, AHI's ability to seek any remedy under 21 C.F.R. § 10.75 will be effectively nullified and there will be a direct and substantial injury to AHI and its members.

7. The Information In the Vose Risk Assessment Does Not Meet the Applicable OMB, HHS, or FDA Guidelines and is In Error        

As set out more fully in the AHI Request, the Vose "Risk Assessment" (which, in reality, our experts consider to be based on a causally irrelevant ratio of aggregate illnesses to aggregate consumption of chicken, rather than being a true risk assessment) does not actually quantify risks, establish a causal relation between exposures and adverse health effects, demonstrate or use any dose-response relation, or establish that any ill effects in humans are specifically associated with the consumption of chicken. Rather, the Vose Risk Assessment and related materials are inaccurate and should be corrected.

The Vose Risk Assessment is methodologically flawed. It contains critical errors and omissions that render its conclusions incorrect and misleading and that make its use for risk management decision-making and for publicizing the alleged (but incorrectly calculated) human health hazards from use of enrofloxacin in poultry inappropriate. In particular, its flaws, errors, and omissions result in an estimate of risk to human health from the use and/or continued use of enrofloxacin in poultry that is not validated, is inaccurate, conflicts with available data, and constitutes a vast overestimation of the potential risk.

Additionally, by failing to consider the risks (especially the human health risks) from withdrawing enrofloxacin for use in poultry, the Vose Risk Assessment provides an incomplete, incorrect, and misleading analysis not useful for risk assessment or for rational risk management decision-making protective of public health and the public interest. For these and other reasons set out in the AHI Request, the Vose Risk Assessment does not meet the standards for information quality set forth in the FDA Guidelines in terms of utility, objectivity and integrity. In addition, the Vose Risk Assessment fails to meet the required higher standards set forth in the FDA Guidelines for "Influential Information" and "Risk Assessment."

For the reasons set out in the Request, AHI seeks the corrections specified in the Request.

Conclusion

In summary, this appeal to correct the Vose Risk Assessment is both timely and appropriate and warrants that the requested relief be immediately considered and granted for the following reasons. First, the deferral of a decision on AHI's Request until after the conclusion of the Baytril Hearing constitutes an effective denial of the Request for correction under the Data Quality Act. Secondly, the decision on the Request by the employee who made the initial decision (i.e., Dr. Sundlof) also constitutes a denial of the Request, as well as a failure to follow FDA's regulations at 21 CFR § 10.75. Thirdly, in light of the representations made by CVM in the Baytril Hearing regarding the inapplicability of the FDA Guidelines to the Vose Risk Assessment, it is evident that CVM's deferral is in fact a denial, is improper, is at variance with FDA's procedures and constitutes little more than a thinly veiled attempt to avoid the requirements of the Data Quality Act.

 

Respectfully Submitted

Animal Health Institute

Kent McClure, General Counsel 
Animal Health Institute 
13 24 G Street, N.W 
Suite 700 
Washington, D.C. 20005 
(202) 637-2400 
kmcclure@ahi.org

Robert B. Nicholas 
Counsel for AHI 
McDermott, Will & Emery 
600 13th Street, N.W. 
Suite 1200 
Washington, D.C. 20005 
(202) 756-8170 
rnicholas@mwe.com

 


1Though dated March 20, 2003 the CVM Reply was not received by AHI until March 25.

2Section 515 of the Treasury and General Government Appropriations Act for Fiscal Year 2001; P.L. 106-554, § 515 (2001)

3The FDA Guidelines may be found at http://www.hhs.gov/infoquality/fda.html.

4"The Human Health Impact of Fluoroquinolone Resistant Campylobacter Attributed to the Consumption of Chicken," Food and Drug Administration, Center for Veterinary Medicine. (2001), p. 1-2. As used herein "Vose Risk Assessment" includes the methodology or model used in the risk assessment, including selection of parameters and their probability distributions; the exposition and use of the model; the data used in the model; and the conclusions; whether disseminated as part of the Vose Risk Assessment or in any other form.

5Guidelines for Ensuring and Maximizing the Quality, Objectivity, Utility, and Integrity of Information Disseminated to the Public; Republication, 67 Fed. Reg. 8452 (February 22, 2002) ("OMB Guidelines").

6As explained by CVM Director Sundlof:

This risk assessment, the Risk Assessment on the Human Health Impact of Fluoroquinolone Resistant Campylobacter Associated with the Consumption of Chicken, was first issued October 18, 2000, and last revised January 5, 2001. CVM developed the risk assessment to assist in establishing the extent of the adverse human health impact of fluoroquinolone use in poultry. Its purpose is to estimate the extent of the risk to human health from resistant Campylobacter pathogens attributed to the use of fluoroquinolones in chickens in the United States. While CVM was developing the risk assessment, it held a two-day public workshop in December 1999, in which it invited experts and others to discuss a fist draft of the risk assessment. Following the public workshop, there was an open comment period during which comments could be submitted to Docket 98D-0969. As part of the NOOH, CVM referenced the risk assessment. During the public comment period on the NOOH, CVM received extensive feedback on the risk assessment. In addition, CVM held a public meeting in January 2001. Although the focus of this meeting was the establishment of regulatory thresholds on antimicrobial resistance, CVM received additional public comments about the risk assessment. CVM Denial at 1-2.

7Id. at 4.

8AHI Request Cover Letter at 2.

9CVM Denial at 4.

10FDA Guidelines provide that "[w]e will respond in accordance with the time frame specified in the complaint mechanism you use. Where a procedure does not specify a time frame for a response, we will respond within 60 days, in accordance with the OMB and HHS Guidelines."Id. at 16.

11An "interested person" means "a person who submits a petition or comment or objection or otherwise asks to participate in an informal or formal administrative proceeding or court action." 21 C.F.R. § 10.3(a).

12 The regulation states (in pertinent part; emphasis supplied) that:

(a) A decision of an FDA employee, other than the Commissioner, on a matter, is subject to review by the employee's supervisor under the following circumstances:…

(3) At the request of an interested person outside the agency.

21 C.F.R. § 10.75(a) and (a)(3). See also FDA Guidelines at 15; Administrative Functions, Practices, and Procedures, 42 Fed. Reg. 4680, 4692 (Jan. 25, 1977); Procedures for Resolving Scientific/Data Disagreements Within CVM (CVM Guide 1240.2110).

13See Administrative Practices and Procedures; Internal Review of Decisions, 63 Fed. Reg. 63978 (November 18, 1998).

14The role of higher level supervisors would essentially be that of an appellate court. That is, "bureau directors and the office of the Commissioner would act in basically the same way as a reviewing court." Administrative Practices and Procedures, Notice of Proposed Rulemaking, 40 Fed. Reg. 40682, 40693 (Sept. 3, 1975).

15Infra n. 9.

16Draft guidance document ( 152) entitled ``Guidance for Industry: Evaluating the Safety of Antimicrobial New Animal Drugs with Regard to Their Microbiological Effects on Bacteria of Human Health Concern.'' This draft guidance document discusses a recommended approach for assessing the safety of antimicrobial new animal drugs with regard to their microbiological effects on bacteria of human health concern. 67 FR 58058-58060 (September 13, 2002).

17According to FDA:

The Office of the Ombudsman has two main functions: (1) to investigate complaints and resolve disputes between companies and individuals and agency offices; and (2) to determine the appropriate classification and regulatory pathway for combination products and for drug, device, and biological products when the jurisdiction of the product is unclear or in dispute. .http://www.fda.gov/oc/ombudsman/homepage.htm

18See also Management of Formal Evidentiary Hearings at 6 (CVM Guide 1240.3670) (one of the responsibilities of the Hearing Manager is to document "significant decisions in the administrative file including revelation of significant controversies or differences of opinion and their resolution as provided by 21 CFR 10.70 and 10.75").

19OMB Guidelines, 67 Fed. Reg. at 8453 (emphasis supplied).

20That is, if the same issue must be addressed sooner or later by the same configuration of people based on the same book of evidence/testimony, then it is unclear why the considerations highlighted by Dr. Sundlof are all that relevant.

21Supra n. 18.

22The ALJ may take notice of an FDA action on the Request pursuant to 21 CFR § 12.95

23The proposal, entitled "Guidance on Pre-approval Information on Registration of New Veterinary Medicinal Products for Food Producing Animals with Respect to Antimicrobial Resistance" (VICH GL27), was to provide harmonized technical guidance in the European Union, Japan and the U.S. for registration of therapeutic antimicrobial veterinary medicinal products intended for use in food producing animals with regard to characterization of antimicrobial resistance selection in bacteria of human health concern.

24From an email message sent from CVM's expert to the VICH working group, Dr. Bill Flynn, to the working group Chair, Dr. Dik Mevius, on January 23, 2003

Last Revised:  August, 2004