Reed v. Wal-Mart Stores, Inc.

197 F. Supp. 2d 883 (2002)

Linda REED, Plaintiff,
v.
WAL-MART STORES, INC., Associates Health and Welfare Plan and Administrative Committee, Associates Health and Welfare Plan, Defendants.

No. 98-CV-70950-DT.

United States District Court, E.D. Michigan, Southern Division.

March 28, 2002.

Mark Granzotto, Elizabeth L. Gleicher, Detroit, MI, for plaintiff.

Michael A. Alaimo, Miller, Canfield, Detroit, MI, William E. Pilchak, Charfoos, Reiter, Farmington Hills, MI, Charles S. Mishkind, Miller, Canfield, Grand Rapids, MI, Daniel G. Cohen, Pilchak & Cohen, Farmington Hills, MI, Sheryl A. Laughren, Berry Moorman, Detroit, MI, Garrett M. Jones, Jackson, Lewis, Orlando, FL, for defendants.

MEMORANDUM OPINION AND ORDER

HOOD, District Judge.

This matter is before the Court on the parties' Cross Motions for Entry of Judgment. For the following reasons, Defendants' Motion for Entry of Judgment is DENIED, and Plaintiff's Motion for Entry of Judgment is GRANTED. Judgment will be entered in favor of Plaintiff and against Defendants.

I. FACTS

Plaintiff Linda Reed, a former employee of Wal-Mart Stores, Inc. ("Wal-Mart"), participated in the Associates' Health and Welfare Plan ("the Plan"), an employer-sponsored, self-funded health and welfare employee benefit plan governed by the Employment Retirement Income Security Act ("ERISA"). In August of 1997, Plaintiff was diagnosed with breast cancer. She underwent a left radical mastectomy with lymph node dissection. The surgery revealed that the cancer had affected six lymph nodes. Because of her condition, Plaintiff was referred Dr. Roy Baynes, an oncologist at the Karmanos Cancer Institute. Dr. Baynes recommended high dose chemotherapy ("HDCT") with peripheral stem cell rescue ("PSCR") or autologous bone marrow transplant ("ABMT"), and requested that the Plan's Administrative Committee pre-authorize coverage to permit Plaintiff to undergo these treatments.

The Plan has a provision which states that "[c]harges for treatment or service that [are] determined by the Plan Administrator to be experimental, investigational, unnecessary, not medically necessary, and/or inappropriate for the condition, even if prescribed and/or ordered by a doctor" are excluded from coverage. (Summary Plan Description at D-12, Exhibit 3 to Howard Affidavit.) Based on this provision, Defendants assert that the Plan consistently denies payment for HDCT with PSCR or AMBT for patients with Stage II cancer where fewer than 10 nodes are affected because such procedures are experimental and/or investigational. Conversely, "[p]atients who are at high risk of relapse who could benefit from this therapeutic approach" would likely be covered. See M-TECS 96-35 at 4 (under Protective Seal). This would include women at Stages II B and III A with 10 or more positive nodes, Stage III B (except inflammatory disease), and Stage IV (in no more than 2 metastatic organ sites).

On January 9, 1998, Wal-Mart sent Plaintiff a denial letter. In pertinent part, the letter read: "Based on the medical information provided, the proposed procedure for this patient's diagnosis of Stage II breast cancer with only six positive nodes is considered experimental/investigational and is therefore not covered under her medical plan." On January 19, 1998, Plaintiff appealed the Plan's denial, but submitted no "medical information, studies, surveys, peer review articles or the like bearing upon the question of whether [HDCT] treatment with PSCR or ABMT had become more than an experimental/investigational procedure for individuals with her specific diagnosis." On February 9, 1998, the Administrative Committee, after reviewing the Administrative Record, affirmed the denial of coverage on the basis that the procedure was experimental/investigational. Plaintiff filed an ERISA action in federal court in March 1998 challenging the claim denial.

After a half day of testimony during a preliminary injunction hearing, the parties stipulated that Plaintiff's request for coverage would be resubmitted to the Administrative Committee. The Stipulated Order of Remand allowed Plaintiff time to present the Committee with documentation in support of her position that she was entitled to HDCT treatment with PSCR or ABMT under the Plan. The re-submission was to be presented to the Committee "in the same manner as if she were appealing a decision of the Committee under proper notification of the grounds for denial of a claim." (See Stipulated Order of Remand).

Plaintiff supplemented the Administrative Record with a feasibility study, performed by one of her experts, of the treatment sought and over 40 articles discussing HDCT treatment with PSCR or ABMT. The Administrative Committee also had before it the transcript from the preliminary injunction hearing which included the testimony of Plaintiff's experts. In the course of its review, the Committee submitted the relevant information to the Medical Care Ombudsman Program, an independent medical review service provided by Prudential Insurance Company of America ("Ombudsman Program").

The Ombudsman Program provides consultation services to the Administrative Committee and reviews clinical information to evaluate whether a proposed treatment should be eligible for coverage by the Plan. The Ombudsman Program did this by independently selecting two experts to review Plaintiff's claim. Both experts, after reviewing the claim, recommended that the treatment be denied as being experimental and/or investigational. Accordingly, the Plan sent a letter to Plaintiff explaining that her "request for an autologous stem cell transplant for the diagnosis of Stage II breast cancer with less that 10 positive lymph nodes remain[ed] denied as being experimental/investigational." After being denied a second time, Plaintiff returned to this Court seeking relief.

II. STANDARD OF REVIEW

Section 502(a)(1)(B) of ERISA, 29 U.S.C. § 1132(a)(1)(B), provides that a participant may bring a civil action "to recover benefits due to him under the terms of the plan, or to clarify his right to future benefits under the terms of the plan." In determining whether benefits under a plan were properly denied, a court reviews, under the appropriate standard, the record presented to the plan administrator. Wilkins v. Baptist Healthcare System, Inc., 150 F.3d 609, 619 (6th Cir. 1998).

A court reviews a denial of benefits under an ERISA plan de novo, unless the plan expressly gives the administrator discretionary authority to determine eligibility benefits or to construe the plan. Firestone Tire & Rubber v. Bruch, 489 U.S. 101, 115, 109 S. Ct. 948, 103 L. Ed. 2d 80 (1989); Univ. Hosp. of Cleveland v. Emerson Elec. Co., 202 F.3d 839, 845 (6th Cir. 2000); Peruzzi v. Summa Medical Plan, 137 F.3d 431, 433 (6th Cir.1998). When the plan so provides, a court reviews the administrator's decision under an "arbitrary and capricious" standard of review. Univ. Hosp. of Cleveland, 202 F.3d at 845; Peruzzi, 137 F.3d at 433; Yeager v. Reliance Standard, 88 F.3d 376, 380 (6th Cir. 1996). Both parties agree that the arbitrary and capricious standard applies in this case.^[1]

When a court applies this very deferential standard, it must determine whether the plan administrator's decision was "rational in light of the plan's provisions." Univ. Hosp. of Cleveland, 202 F.3d at 846. If, based on the evidence, it is possible to offer a reasoned explanation for a particular outcome, that outcome is not arbitrary and capricious. Id.; see also Bartling v. Fruehauf Corp., 29 F.3d 1062, 1071 (6th Cir.1994). In making such a determination, the court is confined to only those facts known to the plan administrator at the time he made his decision. Peruzzi, 137 F.3d at 433-34; Yeager, 88 F.3d at 381. The fact that a contrary conclusion could have been reached on the basis of the evidence does not afford a basis to override the plan administrator's decision. Brandon v. Metropolitan Life Ins. Co., 678 F. Supp. 650, 654-55 (E.D.Mich.1988). When operating under this standard, a court is not to substitute its judgment for that of the plan administrator. Motor Vehicle Mfrs. Ass'n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43, 103 S. Ct. 2856, 77 L. Ed. 2d 443 (1983).

While both parties agree that the arbitrary and capricious standard of review applies to this case, the parties disagree as to whether there is a possible conflict of interest that this Court should take into account in applying this standard. Plaintiff argues that: (1) Defendants' plan is self funded; (2) Defendants bear all of the risk of paying Plaintiff's claim; and (3) Defendants appoint all members of the Administrative Committee that review the claim at issue in this case. "This fact," according to Plaintiff, "requires careful scrutiny of the defendants' decision as well as its decision-making process." Defendants reply that "[j]ust because Wal-Mart Stores, Inc. may appoint members to the Administrative Committee does not reflect a conflict of interest is present which should be taken into account under arbitrary and capricious review."

While a conflict of interest does not change the standard of review, Peruzzi, 137 F.3d at 433, the Sixth Circuit has noted that "[c]ourts should be particularly vigilant in situations where ... the plan sponsor bears all or most of the risk of paying claims, and also appoints the body designated as the final arbiter of such claims. Under these circumstances, the potential for self-interested decision-making is evident." Univ. Hosp. of Cleveland v. Emerson Elec. Co., 202 F.3d 839, 846 n. 4 (6th Cir.2000). This Court will not overlook the possible conflict of interest in this case. To the contrary, this Court must consider it in reviewing the denial of the claim in this case.

III. ANALYSIS

A. Breast Cancer

As Plaintiff recognizes, "[b]reast cancer prognosis is dependent upon several factors, including the size of the tumor, the number of lymph nodes involved with the tumor, the microscopic characteristics (also known as the grade) of the cancer, and whether the tumor is receptive to estrogen." This court has also recognized that:

Breast cancer is classified into four stages based on several factors, including the size of the tumor, how far the disease has spread, and whether the tumors are operable. Stage I cancer is characterized by small tumors within the breast, with no lymph nodes involvement. Patients with Stage I breast cancer have a high five year disease free survival rate, as high as eighty percent. Stage II breast cancer involves larger tumors, not to exceed 5 cm, and may include maligning of the lymph nodes. The five year disease free survival rate for Stage II breast cancer varies from 40 to 80 percent depending on the size of the tumor and the number of nodes involved. A patient is classified as having Stage III breast cancer if the tumor is more than five centimeters in size. A patient is diagnosed as having Stage IV breast cancer once the cancer has spread outside of the breast.

Sluiter v. Blue Cross and Blue Shield of Michigan, 979 F. Supp. 1131, 1133-34 (E.D.Mich.1997).

Plaintiff in this case has Stage II breast cancer with six lymph nodes affected.

B. The Terms "Experimental" and "Investigational" are Ambiguous

As indicated earlier, the Plan has a provision which states that "[c]harges for treatment or service that [are] determined by the Plan Administrator to be experimental, investigational, unnecessary, not medically necessary, and/or inappropriate for the condition, even if prescribed and/or ordered by a doctor" are excluded from coverage. (Summary Plan Description at D-12, Exhibit 3 to Howard Affidavit.) Based on this provision, Defendants assert that the Administrative Committee's decision to deny Plaintiff's claim was not arbitrary and capricious. Specifically, Defendants relied upon the "experimental" and "investigational" terms within this provision to deny the claim.^[2]

Plaintiff argues that the terms "experimental" and "investigational" are ambiguous, and that the doctrine of contra proferentim thus applies. That is, if the Plan's language is susceptible to more than one interpretation, the Court should construe the ambiguities against the drafting party. Univ. Hosp. of Cleveland, 202 F.3d at 846.

Though Defendants assert that the terms are unambiguous, their own experts undercut this position. When asked whether the proposed treatment is considered experimental and/or investigational for Plaintiff's diagnosis, an expert of Defendants' prefaced his answer with the statement: "I do not know how experimental/investigational is defined." (Administrative Record 59-1037). He continued by assuming two different definitions of the terms and analyzing Plaintiff's claim thereunder. This would support Plaintiff's claim that the terms are ambiguous.

Defendants' expert first stated that "[i]nvestigational therapy often is considered to be therapy under study in a clinical trial to evaluate its toxicity, safety, and efficacy." Id. He concluded that "[i]n all of these circumstances, the proposed clinical trial for this plaintiff would be considered investigational." Id. The second definition used by the Defendants' expert was whether "high dose therapy for the treatment of Stage II breast cancer and 4-9 positive axillary nodes requires further evaluation before it is used outside of the clinical trial setting." He categorized the proposed therapy as investigational under this definition, as well. Id.

The Defendants' second expert apparently defined the terms "investigational" and "experimental" as whether a sufficient number of clinical trials had been performed to demonstrate that the therapy is either inferior, equivalent or superior to standard treatment. Because this expert felt that researchers had theretofore been unable to ascertain whether the possible benefit that might be associated with HDCT/ASCR compared with standard therapy, he opined that the procedure was experimental/investigational.

Defendants' experts reveal that there are at least three possible definitions for the terms "experimental" and "investigational." Because the terms are susceptible to more than one interpretation, they are ambiguous. The doctrine of contra proferentim therefore applies, and the Court construes the ambiguities against the drafting party — the Defendants in this case. See Univ. Hosp. of Cleveland, 202 F.3d at 846.

C. The Proposed Treatment is Experimental and/or Investigational Across the Board

Defendants cannot escape the conclusion that the terms are, in fact, ambiguous. When a Plan gives the administrator the discretion to interpret its terms, however, "the administrator's interpretation [of ambiguous terms] must be upheld unless it is arbitrary and capricious, or ``unreasonable.'" Peruzzi, 137 F.3d at 433. Defendants turn to the dictionary to construe the terms, and conclude that "experimental" means, in essence, "pertaining to, derived from or founded on a test, trial, act, or operation for the purpose of investigation as a searching inquiry for ascertaining facts; thorough attempt to learn the facts about something complex or hidden." See Defs.' Mem. in Opp'n to Pl.'s Mot. for Entry of J. and Reply in Supp. of J. in favor of Defs. at 3 (referring to RANDOM HOUSE DICTIONARY 465).^[3]

The statements of some of Defendants' experts imply that the use of high dose therapy for the treatment of Stage II breast cancer and 4-9 positive axillary nodes is currently only used in a clinical trial setting. (Administrative Record 59-1037) (emphasis added). According to Defendants, "Plaintiff's [own] doctors testified that the treatment at issue was a ``trial' and ``study.'" The mere use of the terms "trial" or "study" does not require the conclusion that a particular treatment is "experimental" or "investigational." See, e.g., Pirozzi v. Blue Cross-Blue Shield of Virginia, 741 F. Supp. 586, 593 (E.D.Va. 1990) ("Use of a protocol does not, by itself, indicate that a procedure is experimental.").

It is worth noting that the literature, as well as Defendants' experts, indicate that high dose therapy for the treatment of Stage II breast cancer and ten or more lymph nodes is also experimental and/or investigational. As one of Defendants' experts noted: "In the case of Stage II breast cancer with ten or more positive nodes, the role of high dose therapy remains to be determined." Furthermore researchers "do not yet know whether high dose therapy is any better than standard adjuvant therapy for patients with Stage II breast cancer and 10 or more positive lymph nodes."

Construing the ambiguous terms against the Defendants, it appears that the requested treatment would be excluded for patients with ten or more lymph nodes affected for the same reasons that the Plan denies claims for women with less than ten positive lymph nodes, since its effectiveness across the board has yet to be proven. The terms could also be construed against the drafters in a way that the therapy would be covered regardless of the number of lymph nodes affected. Defendants have not opted for either of these options. In light of the Defendants' application of this definition in the present case, this Court concludes that the Plan's interpretation of the terms is unreasonable and that it arbitrarily and capriciously denied Plaintiff's claim. See Firestone Tire & Rubber, 489 U.S. at 110, 109 S. Ct. 948 (stating that a decision to deny benefits can be held arbitrary and capricious if the plan administrator's interpretation is unreasonable).

D. The Administrative Committee's Decision was Arbitrary and Capricious

Plaintiff argues that the Administrative Committee arbitrarily and capriciously denied her request for treatment under the Plan because, in Plaintiff's estimation, her six node disease "is equivalent to that of a patient with ten positive nodes." Defendants have not directly contested Plaintiff's claim that breast cancer patients with six lymph nodes who have bad prognostic factors can sometimes behave exactly the same as greater-than-ten-node patients. R. Baynes, Administrative Record at 59-250. Instead, they rely on the fact that "nearly all of the published data contained in the Administrative Record involves studies of ten or more nodes or metastatic disease," which makes them irrelevant to Plaintiff's condition.

While it is true that most of the studies in the Administrative Record only involved ten or more nodes, it is also true that the treatment, across the board, has not been proven to be more effective than standard adjuvant chemotherapy. Since the treatment, across the board, has not proven its efficacy, it would appear that the Plan would offer it to no one, regardless of the number of nodes affected. Yet the Plan chooses to cover only women with Stage II breast cancer with ten or more lymph nodes affected. The absence of treatment data for six node disease may, as Defendants assert, indicate its experimental and investigational nature. However, the data for ten node disease is just as uncertain. Indeed, Defendants' own experts recognize that "we do not yet know whether high dose therapy is more effective than conventional dose therapy for patients with more than 10 positive axillary nodes." Pl.'s Exhibit D, p. 2.

During the first appeals process, Defendants claimed that Plaintiff submitted no "medical information, studies, surveys, peer review articles or the like bearing upon the question of whether [HDCT] treatment with PSCR or ABMT had become more than an experimental/investigational procedure for individuals with her specific diagnosis." It is not at all apparent, however, that Defendants actually considered her specific diagnosis once she did submit the requested materials. One of Plaintiff's experts opined that Plaintiff's six node disease, "was biologically exactly the same as the patients that [the Defendants] would cover under the plan." W. Peters, M.D., Administrative Record 59-347. Defendants did not respond to this assertion.

Plaintiff argues that, because Defendants would allow coverage for the requested treatment were Plaintiff to have ten or more lymph nodes affected, Defendants' denial of the claim based on the fact that Plaintiff only has six lymph nodes affected is arbitrary and capricious. In essence, Plaintiff argues that if Defendants are willing to provide coverage for one class of patients, then they must offer coverage to all classes of patients. While at first blush, this contention appears to lack merit, not the least of which could be public policy, a deeper inquiry proves otherwise.^[4]

Defendants assert that "there is substantial evidence in the Record supporting the distinction in determining whether the progress of the cancer warrants the invasive and life threatening treatment at issue." As examples, Defendants cite the Administrative Record at pages 59-217, 59-484, 59-492, 59-638, 59-648-9, 59-725.

As to the first citation to the Administrative Record, the study did, in fact, limit its participants to those who had ten or more lymph nodes affected. It noted that the utility of the treatment was dependent upon the number of nodes involved. In this vein, the article stated that "[t]he prognosis for patients with extensive axillary lymph node involvement at the time of presentation with primary breast cancer is poor." Administrative Record at 59-217 (emphasis added). The term "extensive," as used in this article, obviously contemplates the involvement of more that ten lymph nodes. As Defendants readily emphasize, Plaintiff has only six lymph nodes affected, which is less extensive than the ten node disease. The article's caveat, therefore, cannot apply to Plaintiff's situation, and does not support the six-node/ten-node distinction advocated by Defendants.

Defendants' second and third citations neither advocate nor detract from its position. They simply use patients with ten lymph nodes affected with no discussion whatsoever as to why. See Administrative Record at 59-484. Defendants' fourth citation makes no reference at all to how many lymph nodes were affected in its patients. The fifth citation is the only one which even remotely supports Defendants' position. It states that, because 55% to 87% of women with primary breast cancer involving 10 or more axillary lymph nodes will relapse within 5 years of diagnosis, "efforts to utilize high-dose chemotherapy with ABMT in these patients have been undertaken despite the recognized risks of high-dose therapy," Administrative Record at XX-XXX-XX, the inference being that such treatment is not offered for patients with fewer affected lymph nodes.

Defendants' final citation actually supports Plaintiff's position more than it does Defendants' argument. It involves the requested procedure being performed on women with more than five lymph nodes involved. It, too, fails to describe why it used the lymph node cut-off number of five, as opposed to some other number, in the study. As such, its utility is also limited.

Defendants further claim that even Plaintiff's doctor recognizes that "he and the medical community make a distinction based on the total number of nodes affected — he just disagrees about the number used." See, e.g., Administrative Record 59-347, 59-367 to 59-368. While this is true, Plaintiff's expert bases his opinion on the biology of the cancer. He states that a three node disease has a different biology than the more-than-three-node-disease, but maintains that six-node disease and ten-node are biologically equivalent. As mentioned earlier, Defendants have not rebutted this assertion, and a review of the record reveals nothing to the contrary.

Defendants denied Plaintiff's claim because "... nearly all of the published data contained in the Administrative Record involves studies of ten or more nodes or metastatic disease" and Plaintiff only had six nodes affected. This rationale further demonstrates that the Administrative Committee has determined which forms of breast cancer to treat via high dose chemotherapy in an arbitrary and capricious manner. While it is true that the two main factors affecting survival and recovery rates are the size of the tumor and the number of lymph nodes involved, see Sluiter, 979 F.Supp. at 1134 (emphasis added), it is also true that HDCT with PSCR or ABMT has not proven to be more effective that standard adjuvant chemotherapy, regardless of the number of nodes affected. Moreover, "there are patients with six lymph nodes who have bad prognostic factors . . . who behave exactly the same as greater-than-ten-node patients." It is irrational for this treatment, provided to those with ten or more lymph nodes affected, not to be provided to Plaintiff in this case.

Whereas Plaintiff's six node disease is biologically equivalent to a ten node disease, and Defendants would have provided the treatment to a woman with ten nodes affected, this Court concludes that the Committee acted arbitrarily and capriciously in denying Plaintiff's request for treatment.

IV. CONCLUSION

At least from the prospective of these Defendants, the absence of data cannot be what automatically categorizes a treatment as investigational or experimental because, although there is data on the efficacy of the treatment on ten-or-more-node-disease, it is inconclusive.^[5] Yet Defendants offer coverage in such circumstances.^[6] Nor can efficacy alone be the gauge; HDCT with PSCR or ABMT has not been proven to be more effective than standard chemotherapy in any Stage II cancer case.^[7] Stripped of these rationales for providing the treatment for ten node disease, but not for six node disease, Defendants are left with none.

The evidence presented to the Committee indicates that "it was medically irrational to distinguish between ten node disease and six node disease with aggressive pathology." Moreover, the decision-making process appears to have completely ignored the extensive information Plaintiff provided the Committee in this regard. While Defendants claim that "[t]he Plan's denial letter to Plaintiff specifically states that the Administrative Committee based its decision on the medical information provided . . . encompassing over 1,000 pages," the Committee's decision completely belies, even seems to ignore, the evidence in the Administrative Record, which contains over 40 articles detailing how the therapy is experimental across the board. This constitutes arbitrary and capricious decision making. See Motor Vehicle Mfrs. Ass'n, 463 U.S. at 43, 103 S. Ct. 2856 (stating that a decision can be overturned if the administrator "entirely failed to consider an important aspect of the problem [or] offered an explanation for its decision that runs counter to the evidence.").

Construing the ambiguous terms, "investigational" and "experimental," against Defendants, Univ. Hosp. of Cleveland, 202 F.3d at 846, and taking into consideration the potential for self-interested decision-making inherent in this case, id. at 846 n. 4, this Court concludes that the Administrative Committee has not demonstrated a "deliberate, principled reasoning process . . . supported by substantial evidence," Baker v. United Mine Workers of Am. Health & Retirement Funds, 929 F.2d 1140, 1144 (6th Cir.1991), and that its decision was not "rational in light of the plan's provisions." See Univ. Hosp. of Cleveland, 202 F.3d at 846. Consequently, Plaintiff is entitled to Entry of Judgment in this matter.

Accordingly,

IT IS HEREBY ORDERED that Defendants' Motion for Entry of Judgment (Docket Number 89, filed June 28, 2001) is DENIED and that Plaintiff's Motion for Entry of Judgment (Docket Number 102, filed July 5, 2001) is GRANTED.

NOTES

[1] The Plan states that "[t]he Plan Administrator shall have sole discretion and authority to control and manage the operation and administration of the Plan ... The Plan Administrator shall have complete discretion to interpret the provisions of the Plan, make findings of fact, correct errors, and supply omissions. ..." See Defs.' B. Exhibit 4, § 4 at pp. 4-5.

[2] The denial letter only delineated these two reasons as rationale for the denial. When asked under oath by this Court whether these two reasons were the only cause for denial, Prudential representative, Dr. Chernov, answered, "Yes."

[3] Defendants did not provide the edition of the Random House Dictionary to which they cited.

[4] Public policy may militate against giving this argument credence. Were this Court to enter judgment for Plaintiff based on this logic, insurance companies could deny coverage for all classes of patients simply to avoid allowing coverage for any group.

[5] The General Accounting Office ("GAO") report, for example, states that "[d]espite its increased coverage and use, most experts say they do not yet know whether ABMT for breast cancer is effective." Administrative Record at 59-168.

[6] In fact, it is rather illogical to offer coverage for a treatment whose effectiveness proves to be inclusive.

[7] The GAO report further states that "[m]ost experts say that more research is needed before definitive conclusions can be reached about the treatment's effectiveness compared with conventional chemotherapy." Administrative Record 59-167. Moreover, "[w]hile there were some differences of opinion, the consensus of most of the experts and the literature was that current data indicate ABMT may be beneficial for some breast cancer patients but that there is not yet enough information to establish that it is more effective than standard chemotherapy." Id. at 59-172.