Estate of Aitken v. Shalala, 986 F. Supp. 57 (D. Mass. 1997)

*58 Craig A. Hoover, Jeffrey G. Schneider, Elizabeth D. Smith, Hogan & Hartson, Washington, DC, Scott P. Lewis, Anne Robbins, for Plaintiffs Lillian Aitken, Albert Noecker, Norma Heyen, American Physical Therapy Ass'n.

Anne Robbins, Palmer & Dodge, Boston, MA, for Plaintiffs Ruth Sharpe, Thelma Turner, Etta Langill.

David O. Buchholz, Sheila M. Lieber, U.S. Dept. of Justice, Civ. Div., Washington, DC, for Defendants.

O'TOOLE, District Judge.

The Health Care Financing Administration ("HCFA") of the United States Department of Health and Human Services ("HHS") has issued a "national coverage determination" that the cost of treatments to promote the healing of open wounds by means of various techniques of electrical stimulation therapy ("ES") will not qualify for reimbursement under Medicare. The plaintiffs herein claim that the national coverage determination was unsupported by the administrative record and that it resulted from the agency's adoption, sub silentio and in disregard of applicable rulemaking procedures, of a new rule or regulation governing such determinations. They sue the Secretary of HHS and the Administrator of HCFA for declaratory and injunctive relief and have applied for a preliminary injunction precluding the defendants from giving effect to the determination until the matter has been further considered by the Secretary.

For the reasons set forth below, the Court grants the plaintiffs' motion for a preliminary injunction and remands the matter to the Secretary pursuant to 42 U.S.C. § 1395ff(b) (3) (C).

The Medicare program, Title XVIII of the Social Security Act, 42 U.S.C. § 1395 et seq. (the "Medicare Act") was established in 1965 as a federally funded health care insurance program for the elderly and disabled. Recognizing that questions would inevitably arise about the appropriateness of reimbursement for particular medical services, Congress charged the Secretary with the task of deciding which procedures would be *59 covered by Medicare. See 42 U.S.C. § 1395ff(a). The Secretary, in turn, has delegated the responsibility of making coverage decisions to HCFA. See, e.g., 46 Fed.Reg. 56911, 56929 (Nov. 18, 1981).

The Medicare Act provides a controlling standard against which all coverage determinations must be measured:

Notwithstanding any other provision of this subchapter, no payment may be made ... for any expenses incurred for items or services

(1) which ... are not reasonable and necessary for the diagnosis or treatment of illness or injury or to improve the functioning of a malformed body member.

42 U.S.C. § 1395y(a) (1) (A) (emphasis added)

HCFA typically contracts with public or private "carriers" to assist in the administration of the Medicare programs. See 42 U.S.C. §§ 1395h, 1395u, and 1395kk. Although coverage decisions are commonly made by such carriers on a case-by-case basis, HCFA occasionally issues "national coverage determinations" in situations where there is substantial disagreement in the medical community about the effectiveness or safety of a procedure or where there is inconsistency in coverage by various Medicare carriers.

In making coverage decisions, HCFA historically has interpreted the "reasonable and necessary" limitation on payment to mean that, to qualify for reimbursement, a procedure or device must be "safe, demonstrated as effective, generally accepted in the medical community, and appropriate." See, e.g., Proposed Rule, Criteria and Procedures for Making Medical Services Coverage Decisions, 54 Fed.Reg. 4302, 4307-08 (Jan. 30, 1989); Notice, Procedures for Medical Services Coverage Decisions, 52 Fed.Reg. 15560, 15561-62 (Apr. 29, 1987); Part A Intermediary Letter, No. 77-4; Part B Intermediary Letter, No. 77-5 (January 1977), Medicare & Medicaid Guide (CCH) ¶ 28,152, at 10,601 (1977). The standard has been the same whether the coverage decision is made by a carrier on a case-by-case basis or by HCFA as part of a national coverage determination.

Judicial review of national coverage determinations, however, differs substantially from judicial review of individual Medicare coverage decisions. First, national coverage determinations may not be held unlawful or set aside on the ground that HCFA failed to abide by the procedural requirements of the Administrative Procedure Act ("APA"), 5 U.S.C. § 553, or by the notice and comment provisions of the Medicare Act. See 42 U.S.C. § 1395ff(b) (3) (B). Moreover, a court reviewing a national coverage determination must uphold that determination unless it finds that the record on which it is based "is incomplete or otherwise lacks adequate information to support the validity of the determination," in which case it may not invalidate the determination or direct that coverage be granted, but may only remand the matter to the Secretary for additional proceedings. See 42 U.S.C. § 1395ff(b) (3) (C).

Each of the individual plaintiffs suffers from one or more chronic wounds or sores that are resistant to healing.^[1] In the first instance such wounds are treated by a combination of "basic therapies," such as cleaning and dressing the wound, debridement (removing dead skin or tissue around the wound), hydrotherapy (such as whirlpool baths), maintenance of a healthy diet, and decreasing pressure on affected areas by moving the patient and through the use of air mattresses. Pls' Mem. Supp. Prel. Inj, Ex. 1, Unger Decl. at ¶ 13. Many patients, however, experience little benefit from these therapies. For such patients with "recalcitrant" or intractable wounds, some form of ES may be used to try to advance the healing process. Id., Ex. 4, Thomas Decl. at ¶ 5.

ES involves the application of an electrical current to wound sites. There are different techniques: direct current, alternating current, *60 pulsed current, pulsed electromagnetic induction, and spinal cord stimulation. ES, like the other therapies, is customarily administered by physical therapists, including the members of the plaintiff American Physical Therapy Association ("APTA").

In 1980 HCFA began to permit Medicare carriers to reimburse the cost of some forms of ES on a case-by-case basis. However, in April, 1997, HCFA published a notice that, effective May 14, 1997, it would no longer authorize reimbursement for the cost of ES for wound healing in any case. The notice stated:

There is insufficient evidence to determine any clinically significant differences in healing rates. Therefore, ES cannot be covered by Medicare because its effectiveness has not been adequately demonstrated.

Id., Ex. 6, Medicare Coverage Issues Manual ("MCIM") § 35-98, 5 Medicare and Medicaid Guide (CCH) ¶ 27,201 (1997).

In response to this announcement, APTA requested a postponement of the effective date of the coverage determination and an opportunity to present to HCFA additional information supporting the effectiveness of ES. HCFA allowed APTA to present further evidence of ES's effectiveness in the healing of recalcitrant chronic wounds.

HCFA was not persuaded to change its view, however, and sent a letter to APTA on July 7, 1997, advising that the agency had decided not to modify its determination to deny coverage for ES. AR at 1. In the letter, HCFA noted that "scientific data demonstrated that ES does not appear to be markedly superior or inferior to conventional or alternative therapies for chronic wound healing." It further noted that "no electrical stimulation units are specifically approved for the treatment of wound healing" and urged ES device manufacturers to seek "FDA approval of the devices specifically for wound healing." Id. (emphasis in original).

According to the amended complaint, the individual plaintiffs suffer from chronic wounds and sores which have not responded to conventional medical treatment. Some of the plaintiffs experienced significantly increased healing rates when their physical therapists employed ES as an adjunctive therapy. Other plaintiffs, who have wounds which are unresponsive to conventional therapy, say that HCFA's blanket refusal to cover ES has prevented them from receiving treatment which their therapists believe would accelerate the healing process of their wounds.

"[A] party seeking preliminary injunctive relief must prove: (1) a substantial likelihood of success on the merits; (2) a significant risk of irreparable harm if the injunction is withheld; (3) a favorable balance of hardships; and (4) a fit (or, at least, a lack of friction) between the injunction and the public interest." Equal Employment Opportunity Comm'n v. Astra USA, Inc., 94 F.3d 738, 742 (1st Cir. 1996).

The defendants argue that the Medicare Act's provision limiting judicial review of national coverage determinations precludes the court from enjoining enforcement of the decision not to reimburse ES.^[2] As the plaintiffs rightly point out, however, their motion does not seek an order finding ES to be reimbursable by Medicare; instead it simply seeks to suspend the implementation of the national coverage determination pending its reconsideration by the Secretary. The result would be a return to case-by-case evaluation for the short term. This degree of preliminary injunctive relief is not foreclosed by the statutory provision the defendants cite. In other contexts, such relief is often given. See Illinois Public Telecommunications Ass'n v. Federal Communications Comm'n, 123 F.3d 693, 693-94 (D.C.Cir.1997) (vacating an agency rule which it remanded to the *61 agency for further consideration); Petroleum Communications, Inc. v. Federal Communications Comm'n, 22 F.3d 1164, 1173 (D.C.Cir.1994) (vacating FCC rule on grounds that it was arbitrary and capricious while not foreclosing the possibility that the agency may develop a convincing rationale for re-adopting the same rule on remand).

"The sine qua non of [the preliminary injunction] formulation is whether the plaintiffs are likely to succeed on the merits." Weaver v. Henderson, 984 F.2d 11, 12 (1st Cir.1993). The first inquiry, then, concerns the likely validity of the plaintiffs' substantive arguments against the implementation of HCFA's coverage determination for ES.

The plaintiffs make two basic arguments. First, they contend that HCFA has adopted a new substantive "rule" regarding the eligibility of a procedure for coverage. According to this argument, the new "rule" requires ES to be shown to be superior to existing treatments and, because it involves use of a "medical device," that it be approved by the Federal Drug Administration ("FDA"). Such a rule may not be adopted, they say, without adherence to the formal notice-and-comment procedures of the APA, 5 U.S.C. § 553(b), and the Medicare Act, 42 U.S.C. § 1395hh(b) (1), which did not happen.^[3]

The plaintiffs have not shown a likelihood of success on this claim. Their argument is founded to a large extent on statements in HCFA's July 7 letter to APTA that explain its decision to issue the national coverage determination. These statements are more properly characterized as an articulation of the agency's reasoning than a change in the rules. The letter explained why HCFA had concluded that ES did not meet the agency's customary interpretation of the statutory standard, that is, why ES could not be determined to be "safe and effective," so that it was not, therefore, "reasonable and necessary." Contrary to the plaintiffs' argument, it appears the agency was not changing the standard, but applying it.

Whether HCFA erred in its application of the standard is more or less the point of the plaintiffs' second contention: that the administrative record does not adequately support HCFA's determination. They maintain that the record, rather than warranting a general coverage determination, actually supports the status quo ante, when individual decisions were made case-by-case and ES was not excluded in all cases from coverage.

While judicial review of national coverage determinations is limited, it is not nonexistent. A national coverage determination may be remanded to the Secretary for further consideration if a court determines that "the record is incomplete or otherwise lacks adequate information to support the validity of the determination." 42 U.S.C. § 1395ff(b) (3) (C).^[4]

Testing whether an administrative record contains information adequate to support the agency's conclusion is a common task in the judicial oversight of administrative action. The court's evaluation must, of course, recognize and defer appropriately to the agency's experience and expertise in dealing with the subject matter of the determination.

An administrative decision might be condemned as arbitrary if the agency has "offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise." Motor Vehicle Mfrs. Ass'n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43, 103 S. Ct. 2856, 2866, 77 L. Ed. 2d 443 (1983). See also Dubois v. United States Dep't of Agric., 102 F.3d 1273, 1283-85 (1st Cir.1996). The agency must examine "the relevant data" and articulate "a satisfactory explanation for its action including *62 a `rational connection between the facts found and the choice made.'" State Farm, 463 U.S. at 43, 103 S. Ct. at 2866 (quoting Burlington Truck Lines v. United States, 371 U.S. 156, 168, 83 S. Ct. 239, 245, 9 L. Ed. 2d 207 (1962)). If there does not appear to be a relationship between the agency's findings and the applicable criteria of judgment, the case may be remanded for a "more adequate application" of the criteria. State of New York on Behalf of Stein v. Secretary of Health & Human Servs., 924 F.2d 431, 433 (2d Cir.1991).

In this case, there needs to be, at the least, a better explanation by HCFA of its coverage determination and, perhaps, a revision of that determination. Examination of the record raises questions about whether the study upon which HCFA apparently based its decision, properly interpreted, provides "adequate" support for the coverage determination. To answer these questions a remand is necessary.

The administrative record comprises principally the following categories of evidence: a number of published papers concerning the value of ES in wound treatment, a good deal of anecdotal support for the use of ES in clinical settings, including many letters from physical therapists and others in the field urging HCFA not to issue a blanket coverage determination, some publications issued by the Agency for Health Care Policy and Research ("AHCPR"), a body existing under the aegis of the Public Health Service within HHS, and a report by a "health technology assessment consultant" to HCFA, identified in the record only as ECRI. In terms of quantity, the great bulk of the materials in the record tends to support the effectiveness of ES, contrary to HCFA's conclusion. The sheer quantity of material, of course, cannot decide the question, especially when a good deal of the material seems to have been prompted by an organized effort by APTA to get its members to write to HCFA on this issue of concern to them. Rather, what matters is the quality of the evidenceits validity, meaning and significance. HCFA appears to have based its decision on the ECRI report, and an examination of the reasonableness of its reliance on that report is in order.

In brief, the ECRI report represents a survey and analysis of a number of published studies concerning the effectiveness of ES in wound healing. The report concluded that there were significant flaws in the methodologies of most of the studies, although those flaws did not always "confound" the results. AR at 629. It also concluded that the quality of the ES studies was roughly equivalent to the quality of similar published studies of other wound healing therapies. Id. ECRI also made the following "Quantitative Conclusions":

• ES facilitates the healing rate of chronic ulcers.

• ES facilitates the complete healing of chronic ulcers.

• The relationship between outcomes and ES can be affected by wound size and type of stimulator.

• Decubitus ulcers are more likely to heal completely in response to ES than venous ulcers.

Id.

ECRI then reported the following "General Findings":

• ES devices are safe when used appropriately.

• Most types of ES are more effective than minimal treatment (e.g., salinesoaked gauze).

• ES is not markedly superior to or inferior to conventional or alternative therapies.... There is insufficient evidence to determine if clinically significant differences exist.

AR at 631.

One reasonably wonders how and why HCFA relied on a report that found ES treatment (1) "safe when used appropriately," (2) "more effective than minimal treatment," and (3) effective to about the same degree as other conventional therapies, to conclude that ES does not satisfy the agency's purported standard: that the treatment be "safe and effective."

The record contains the minutes of HCFA's Technology Assessment Committee's consideration of the issue, along with what appear to be materials used in a staff *63 presentation to the Committee. The minutes record what the Committee was told about the ECRI report, including the following summary:

ES is not markedly superior or inferior to conventional or alternative therapies for chronic wound healing. However, there is insufficient evidence at this time to determine whether any clinically significant differences in healing rates [exist]. The assessment found that studies conducted on ES as a therapeutic modality for wounds were Poorly done and demonstrated no evidence of effectiveness.

AR at 446D (emphasis added).

The first two sentences of this paragraph accurately reflect some of ECRI's findings. The emphasized sentence, however, does not. It would have been more accurate to qualify the statement that ECRI had found and described a number of flaws in the various published studies by noting that ECRI had also concluded that the flaws did not always make the studies meaningless and that similar flaws existed in the literature with respect to "conventional" therapies as well (for which Medicare coverage is available).^[5]

More fundamentally, ECRI's conclusion that ES is about as effective as other therapies ("not markedly superior or inferior") does not support the conclusion that ES is not effective. To say that the effectiveness of "ES plus no therapy" is "indistinguishable" from that of "conventional therapy plus no therapy" does not provide an adequate basis for choosing to cover the latter and exclude the former. See AR at 827, 828. ECRI reported that the clinical studies of ES and conventional therapies were generally similar in quality and reliability.^[6] Based on its evaluation of those studies, ECRI concluded that there seemed to be no significant difference in effectiveness between ES and conventional therapies. See AR at 796-807. If the evidence was of similar qualityas ECRI said it wasthen HCFA's asserted reason for the distinction between ES and conventional therapies cannot be accepted. Ordinarily, it would be legitimate for the agency to insist on stronger or more reliable evidence from randomized clinical trials before accepting that a given therapy is effective, but its preference in that respect must not be inconsistent with the agency's approval of other therapies supported by similar evidence. In other words, the agency may not play favorites.^[7]

Moreover, the agency must be careful not to transform an understandable preference for one kind of evidence into an impassable barrier. As some of the critics of the coverage determination have pointed out, clinical *64 patients with intractable wounds are not experimental animals, and it is perhaps an impossible expectation to demand that studies involving real patients have all the controlled perfection of laboratory studies.^[8] Such an expectation seems even less defensible here because the analysts of the data have said, in effect, that the ES studies are about as good as one usually gets in this area.

It is not as if the published studies were the only evidence in the record that supported the usefulness of ES in treating intractable wounds. In 1994 the HHS's own Agency for Health Care Policy and Research issued a "Clinical Practice Guideline" on "Treatment of Pressure Ulcers" in which it advised physicians,

At this time, electrotherapy is the only adjunctive therapy with sufficient supporting evidence to warrant recommendation by the panel. The panel recommends that clinicians consider a course of treatment with electrical stimulation for Stage III and IV pressure ulcers that have proved unresponsive to conventional therapy. Electrical stimulation may also be useful for recalcitrant Stage II ulcers.

AR at 500.

The record contains a number of submissions from therapists testifying to their own clinical experiences with the ES as a successful treatment. Some included actual patient charts. There are very good reasons for being cautious about such anecdotal clinical evidence, and HCFA is certainly not bound to accept it, but neither should it be free to disregard it entirely, without explanation.

In sum, this is what emerges from the administrative record: There is a good deal of anecdotal support from clinicians for the use of ES in treating intractable wounds. The evidence from published studies of randomized clinical trials is somewhat ambiguous in its support of ES, although it appears that ES is effective in some applications. Because of flaws in the way some of the reported studies were designed and conducted, these studies cannot be relied on as proof of the effectiveness of ES, although they do not disprove its effectiveness either. Despite their flaws, the studies were sufficiently persuasive to lead the AHCPR recently to recommend the use of ES in certain cases. The published studies tend to support the conclusion that ES is about as effective as "conventional" therapy.

It is hard to see how this record supports the agency's determination that ES should never be covered in any case. The record "lacks adequate information to support the validity of the determination." 42 U.S.C. § 1395ff(b) (3) (C). Rather, the explanation currently offered by HCFA for its decision "runs counter to the evidence before the agency." State Farm, 463 U.S. at 43, 103 S. Ct. at 2866.

On this claim, the plaintiffs have shown a substantial likelihood of success on the merits, thereby satisfying the first of the criteria for a preliminary injunction. Indeed, the plaintiff has sufficiently demonstrated that the determination is so inadequately supported by the record as to justify a remand of the matter pursuant to 42 U.S.C. § 1395ff(b) (3) (C).

When the likelihood of success on the merits is great, as it is here, a movant can show somewhat less in the way of irreparable harm and still obtain preliminary injunctive relief. See Astra, 94 F.3d at 743. See also, Michigan Coalition of Radioactive Material Users, Inc. v. Griepentrog, 945 F.2d 150, 153 (6th Cir.1991) ("Simply stated, more of one excuses less of the other."); Maram v. Universidad Interamericana De Puerto Rico, Inc., 722 F.2d 953, 958 (1st Cir.1983) (stating that "these interests must be weighed inter sese"). Nevertheless, the importance of showing irreparable harm in the calculus of *65 injunctive relief is of great importance and should not be lightly set aside. Astra, 94 F.3d at 743.

The individual plaintiffs have adequately shown that the blanket exclusion of ES from Medicare reimbursement pending a remand to the Secretary will cause them irreparable harm. "Termination of benefits that causes individuals to forego such necessary medical care is clearly irreparable injury." Massachusetts Ass'n of Older Ams. v. Sharp, 700 F.2d 749, 753 (1st Cir.1983). The defendants' argument that no patient who truly requires ES therapy will be deprived of such treatment because the ethical code of medical professionals requires them to provide indicated care regardless of ability to pay is entirely unpersuasive. Taking that argument to its logical conclusion, the defendants could declare any procedure "not covered" and then maintain that there is no "irreparable injury" because medical professionals must render such free care as is necessary in any event. Both good sense and precedent, happily congruent, compel the rejection of that proposition.

The defendants' attempt to distinguish termination of medical benefits from a determination not to reimburse the cost of a particular procedure is likewise unpersuasive. There is no practical difference in predicament between a patient who is denied reimbursement for all possible treatments and one who is denied reimbursement only for the specific treatment she needs. In both cases, the person suffers the same irreparable harm. See Sharp, 700 F.2d at 753.

The record on the present motion demonstrates amply that without a preliminary injunction the individual plaintiffs will be forced to forego appropriate treatment for their wounds with potentially serious consequences. That is a sufficient showing of irreparable injury to support the issuance of a preliminary injunction.

The plaintiffs have also shown that the balance of the harm weighs significantly in their favor. The harm to be suffered by defendants if a preliminary injunction is granted would be minimal. As both parties agree, for years prior to last July, Medicare carriers have made determinations about whether ES was "reasonable and necessary" on a case-by-case basis, and Medicare has reimbursed patients and their doctors for such treatment. Contrary to the defendants' argument that a preliminary injunction "would upset the Secretary's administration of the Medicare program," all that will happen is that there will be some administrative inconvenience as the defendants suspend implementation of a coverage decision pending further consideration and action by HCFA and the Secretary.

In this case, it is sufficient to say that the public interest is greatly served by assuring that the agency will perform its statutory duties carefully and fairly.

For the foregoing reasons, the plaintiffs' motion for a preliminary injunction is GRANTED.

Pursuant to 42 U.S.C. § 1395ff(b) (3) (C), the national coverage determination regarding reimbursement for ES is remanded to the Secretary for additional proceedings to supplement the record with evidence adequate to support or modify its decision.

It is SO ORDERED.

For the reasons stated in the accompanying Memorandum and Order, it is hereby ORDERED that defendants Donna E. Shalala, Secretary of Health and Human Services, Bruce C. Vladeck, Administrator, Health Care Financing Administration ("HCFA"), and their agents, representatives, and employees be, and the same hereby are, enjoined and restrained from enforcing or giving any effect to the national coverage decision issued by HCFA denying Medicare reimbursement for the use of electric stimulation for the treatment of wounds published at MCIM § 35-98, 5 Medicare and Medicaid Guide (CCH) ¶ 27,201 (1997).

*66 The preliminary injunction shall remain in effect until such time as the Secretary shall have supplemented the record as required by 42 U.S.C. § 1395ff(b) (3) (C) and this court shall have issued a ruling granting or denying a permanent injunction or until it shall be otherwise ordered.

[1] The types of wounds involved may include decubitus ulcers (commonly referred to as "bedsores") as well as venous ulcers and ischemic ulcers, which affect veins and arteries. See Administrative Record ("AR") at 635-36.

[2] 42 U.S.C. § 1395ff(b) (3) (C) provides:

In any case in which a court determines that the record is incomplete or otherwise lacks adequate information to support the validity of the determination, it shall remand the matter to the Secretary for additional proceedings to supplement the record and the court may not determine that an item or service is covered except upon review of the supplemented record.

[3] The defendants do not agree that the rulemaking procedures apply to HCFA's actions here. Because the larger question is resolved on other grounds, that particular dispute need not be settled at this juncture.

[4] This section permits a court to "review coverage determinations to determine whether the Secretary arrived at the result in [an] arbitrary manner or without an adequate basis." H.R.Rep. No. 99-1012, 99th Cong., 2d Sess. 351, reprinted in 1986 U.S.C.C.A.N. 3607, 3996.

[5] The Committee's misunderstanding may have resulted from what appears to have been a highly tendentious staff presentation. See generally AR at 447-462.

[6] For example, the ECRI report states, "ES controlled studies for venous ulcers are about equal to or slightly inferior in quality compared to other controlled studies for venous ulcers," and that "ES controlled studies for decubitus ulcers are about equal to or slightly superior in quality compared to other controlled studies for decubitus ulcers." AR at 629.

[7] There is another problem. The record upon which HCFA based its decision also suffers from an apparent definitional inconsistency. According to ECRI, none of the studies of ES compared ES treatments to treatments by conventional therapy. AR at 253A, 822. The ECRI report used the term "conventional therapy" to include, variously, debridement, cleaning agents, topical agents, dressings, bandages, systemic or local antibiotics, compression therapies, systemic medications, and nutritional supplements. AR at 797. Based on this definition, ECRI's statement that there are no studies which compare ES to conventional treatment appears simply wrong. For example, the control group described in the Kloth & Feedar (1988) study analyzed by ECRI (AR at 826) were treated by pressure-relieving devices, debridement, enzymes, nutritional supplements, and dressings treatments within ECRI's definition of "conventional therapy." Indeed, most of the studies reviewed by ECRI compared ES to some form of conventional treatment. Nonetheless, the ECRI report asserts that these studies compared ES to "no therapy." AR at 822, 845.

The information provided to the Technology Advisory Committee ("TAC") defined conventional therapy otherwise. In the HCFA staff presentation to TAC, conventional treatment was defined in a more limited way as "debridement, cleansing, and/or wound dressing." AR at 446F. It is not important (or proper) here to decide what the most appropriate definition of conventional therapy should be; that kind of decision falls squarely within the purview of agency expertise. What is important, however, is that the agency define its critical terms with clarity and apply the definitions with consistency.

[8] One external reviewer of the ECRI report wrote: "You found much to criticize in many studies, but I know you would not volunteer to be treated by placebo for six months while your leg stinks and deteriorates. We have found that, when the FDA is truly pleased with a protocol, we can't afford to do the study, there aren't enough patients of some exact description, and no patient in his right mind would participate. Intellectual analysis misses the basic question: Is the treatment reasonably safe and is there `valid' but not necessarily unassailable evidence that it works"? Dr. Harris letter, dated March 8, 1996, to Dr. Erlick, AR at 900-901.