Johnson v. Blue Cross & Blue Shield of Alabama, Inc. (457 F. Supp. 2d 1288, 2006)

MEMORANDUM OPINION

SMITH, District Judge.

Plaintiff, LeeAnn D. Johnson, filed this action under the Employee Retirement Income Security Act of 1974 (“ERISA”), 29 U.S.C. § 1001 et seq., against defendant Blue Cross and Blue Shield of Alabama, Inc. Plaintiff claims that defendant, acting as claims administrator of the group health insurance policy under which she is a beneficiary, erroneously characterized a surgical procedure recommended to her by her treating physicians as “investigational.” The effect of that label was to exclude the procedure requested from the scope of the policy’s coverage. Plaintiff seeks a declaratory judgment that defendant violated ERISA, and an order requiring defendant to cover the costs attendant to performance of the procedure.

The action now is before the court on cross motions for summary judgment filed by plaintiff and defendant. In disposing of these motions, the court must also address defendant’s motion to strike certain of plaintiffs evidentiary submissions. For the reasons set forth below, plaintiffs motion for summary judgment is due to be granted. Accordingly, defendant’s motion for summary judgment will be denied. Moreover, because the court need not consider any of plaintiffs evidentiary submissions to resolve the case in her favor, defendant’s motion to strike is due to be denied as moot.

Part One

Summary Judgment Standards

Federal Rule of Civil Procedure 56 provides, in part, that summary judgment not only is proper, but “shall be rendered forthwith if the pleadings, depositions, answers to interrogatories, and admissions on file, together with the affidavits, if any, show that there is no genuine issue as to any material fact and that the moving party is entitled to judgment as a matter of law.” Fed.R.Civ.P. 56(c) (emphasis supplied). “[T]he plain language of Rule 56(c) mandates the entry of summary judgment, after adequate time for discovery and upon motion, against a party who fails to make a showing sufficient to establish the existence of an element essential to that party’s case, and on which that party will bear the burden of proof at trial.” Celotex Corp. v. Catrett, 477 U.S. 317, 322, 106 S.Ct. 2548, 91 L.Ed.2d 265 (1986). Likewise, “summary judgment should be granted where the evidence is such that it ‘would require a directed verdict for the moving party’ [if the case proceeded to trial].” Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 251, 106 S.Ct. 2505, 91 L.Ed.2d 202 (1986) (internal citations omitted).

In either situation, the relevant question is whether the admissible evidence on file “show[s] that there is no genuine issue as to any material fact and that the moving party is entitled to judgment as a matter of law.” Fed.R.Civ.P. 56(c).

In making this determination, the court must review all evidence and make all reasonable inferences in favor of the party opposing summary judgment.

The mere existence of some factual dispute will not defeat summary judgment unless that factual dispute is material to an issue affecting the outcome of the case. The relevant rules of substantive law dictate the materiality of a disputed fact. A genuine issue of material fact does not exist unless there is sufficient evidence favoring the nonmoving party for a reasonable jury to return a verdict in its favor.

Chapman v. AI Transport, 229 F.3d 1012, 1023 (11th Cir.2000) (en banc) (quoting Haves v. City of Miami, 52 F.3d 918, 921 (11th Cir.1995)). See also Anderson, 477 U.S. at 251-52, 106 S.Ct. 2505 (asking “whether the evidence presents a sufficient disagreement to require submission to a jury or whether it is so one-sided that one party must prevail as a matter of law”).

Because cross motions for summary judgment are presented, “[t]he court must rule on each party’s motion on an individual and separate basis, determining, for each side, whether a judgment may be entered in accordance with the Rule 56 standard.” 10A Charles A. Wright, Arthur R. Miller & Mary Kay Kane, Federal Practice and Procedure: Civil 3d § 2720, at 335-36 (1998) (footnote omitted). See also, e.g., Arnold v. United States Postal Service, 649 F.Supp. 676, 678 (D.D.C.1986). Further, the court is required to “relate all material facts in genuine dispute in the light most favorable to the party resisting summary judgment.” Serrano-Cruz v. DFI Puerto Rico, Inc., 109 F.3d 23, 24 (1st Cir.1997) (citing Sanchez v. Alvarado, 101 F.3d 223, 225 n. 1 (1st Cir.1996)).

Part Two

Summary of Relevant Facts

Plaintiff is a 41-year-old female who suffers from morbid obesity. She is also a beneficiary under an ERISA-governed group medical insurance plan established by her husband’s employer, Book Systems, Inc. Book Systems, Inc. is the plan sponsor and plan administrator. Defendant, Blue Cross and Blue Shield of Alabama, Inc., is the claims administrator and insurer under the Book Systems plan. In that capacity, defendant fields authorization requests, exercises discretion to determine what benefits are available to plan members, and pays claims.

A. The Book Systems Policy

The contours of plaintiffs health insurance policy are outlined in a document called the “Summary Plan Description.” Relevant here are the following three explicit benefit exclusions, which are listed under the heading “Health Benefit Exclusions” within the Summary Plan Description:

We will not provide benefits for the following:

1. Services or expenses we determine are not medically necessary.

18. Investigational treatment [or] procedures. ...

22. Services or expenses for treatment of any condition including, but not limited to, obesity, diabetes, or heart disease, which is based upon weight reduction or dietary control or services or expenses of any kind to treat obesity, weight reduction or dietary control. This exclusion does not apply to surgery for morbid obesity if medically necessary and in compliance with guidelines of the Claims Administrator. Benefits will only be provided for one surgical procedure for obesity (morbid) in a lifetime. Benefits will not be provided for subsequent surgery for complications related to a covered surgical procedure for obesity (morbid) if the complications arise from non-compliance with medical recommendations regarding patient activity and lifestyle following the procedure.

The appellations “investigational” and “medically necessary” are clearly terms of art, and they are defined in the “Definitions” portion of the Summary Plan Description. The pertinent definitions are reproduced in material part below.

Medically Necessary or Medical Necessity: We use these terms to help us determine whether a particular service or supply will be covered. When possible, we develop written criteria (called medical criteria) that we use to determine medical necessity. We base these criteria on peer-reviewed literature, recognized standards of medical practice, and technology assessments. We put these medical criteria in policies that we make available to the medical community and our members. We do this so that you and your providers will know in advance, when possible, what we will pay for. If a service is not medically necessary according to one of our published medical criteria policies, we will not pay for it. If a service or supply is not addressed by one of our published medical criteria policies, we will consider it to be medically necessary only if we determine that it is:

• appropriate and necessary for the symptoms, diagnosis, or treatment of your medical condition;

• provided for the diagnosis or direct care and treatment of your medical condition;

• in accordance with standards of good medical practice accepted by the organized medical community;

• not primarily for the convenience and/or comfort of you, your family, your physician, or another provider of services;

• not “investigational;”

Investigational: Any treatment, procedure, equipment, drugs, drug usage, or supplies that either we have not recognized as having scientifically established medical value, or that does not meet generally accepted standards of medical practice. When possible, we develop written criteria (called medical criteria) concerning services or supplies that we consider to be investigational. We base these criteria on peer-reviewed literature, recognized standards of medical practice, and technology assessments. We put these medical criteria in policies that we make available to the medical community and our members. We do this so that you and your providers will know in advance, when possible, what we will pay for. If a service or supply is considered investigational according to one of our published medical criteria policies, we will consider it to be non-investigational only if the following requirements are met:

• The technology must have final approval from the appropriate government regulatory bodies;

• The scientific evidence must permit conclusions concerning the effect of the technology on health outcomes;

• The technology must improve the net health outcome;

• The technology must be as beneficial as any established alternatives; and,

• The improvement must be attainable outside the investigational setting.

Dr. Patrick Ryce, Senior Vice President and Medical Director for defendant Blue Cross and Blue Shield of Alabama, explained defendant’s interpretation of these provisions in a declaration attached to defendant’s motion for summary judgment. According to Dr. Ryce, “if one of Blue Cross’s -written medical policies states that a particular procedure or technology is considered ‘investigational,’ and therefore, not medically necessary, that directive controls and Blue Cross will not cover the procedure,” even though it meets the other requirements for being characterized as “medically necessary.” As should be clear from the definitional language, however, the initial characterization of a particular procedure or technology as “inves-tigational” in a written policy does not necessarily close the door on the issue. Rather, the determination in the written directive will endure only until such time as the procedure in question “meets the five criteria for non-investigational [procedures] set forth in the definition of ‘inves-tigational.’” Upon satisfying those five criteria, the procedure will no longer be subject to exclusion under the policy— assmning, of course, that it also satisfies all of the additional requirements for characterization as a “medical necessity.”

The definitions also refer to the protocol used in developing the written medical eri-teria. Specifically, “peer-reviewed literature, recognized standards of medical practice, and [internally created] technology assessments” are all relevant to deciding whether a given procedure should be classified as “investigational” or “noninves-tigational.” Dr. Ryce stated in his declaration that “[r]andomized, double-blind, controlled studies are considered the gold standard of evidence-based decision making,” but conceded that such studies are often not available. In such a situation, other evidence, including solicited opinions from practitioners in the relevant discipline, are utilized. As for who weighs the evidence, Dr. Ryce stated that he personally reviews the literature, but that defendant’s Medical Advisory Committee — a group of independent, board-certified doctors of various specializations — is consulted as well.

B. Plaintiffs Requests for Coverage

Because plaintiffs morbid obesity is potentially life-threatening, her primary care physician, Dr. Barbara Bennett, and her consulting bariatric surgeon, Dr. John D. Husted, recommended that she undergo a laparoscopic surgical procedure known as a Biliopancreatic Diversion with Duodenal Switch (“BPD/DS”). As discussed more fully below, BPD/DS is just one of several available weight loss surgeries, but it is the one that plaintiff and her doctors felt best suited her.

By letter dated December 21, 2003, Dr. Husted submitted to defendant a request for a “Medical Necessity and Benefits Predetermination,” asking that plaintiff receive coverage for the proposed BPD/DS surgery. In his letter, Dr. Husted described plaintiffs medical condition and her previous unsuccessful attempts to lose weight through less drastic measures, and generically recommended bariatric surgery as the best option for safe and effective treatment of morbid obesity. On January 7, 2004, defendant summarily denied the request, stating in a letter to Dr. Hust-ed that “laparoscopic duodenal switch for treatment of morbid obesity is considered investigational by Blue Cross and Blue Shield of Alabama and is not covered.” The denial of coverage was dictated by defendant’s written medical policy, in effect at the time of the request, which deems BPD/DS surgery “investigational.”

Upon receiving the denial, Dr. Husted submitted a second letter, this time asking for a predetermination that defendant would cover a different surgical procedure, the Roux-en-Y gastric bypass (“RNY”). Defendant notified Dr. Husted on February 6, 2004 that plaintiff met the medical criteria for the RNY, and that coverage would accordingly be provided for that procedure. That same day, Dr. Husted requested that defendant reconsider its previous denial of coverage for the BPD/DS surgery, and supported his request with a copy of Blue Cross and Blue Shield of Tennessee’s policy on bariatric surgery, which apparently authorizes the BPD/DS procedure. Again, defendant denied the request for BPD/DS coverage, concluding that “[under] the terms of this patient’s contract, this procedure would be considered investigational and not covered.” The unnamed customer service representative who authored the denial letter explained: “This is not a reflection on the type of service recommended, but simply a service that is excluded under the terms of the patient’s contract.”

In the months that followed, plaintiff and/or her attorneys wrote numerous additional letters requesting reconsideration of the coverage decision. Along with many of these letters, plaintiff or her attorneys included medical literature and myriad other materials discussing the merits of the BPD/DS procedure. On two occasions, plaintiff requested copies of all documents relied upon by defendant in reviewing her previous appeals, and a statement explaining how the principles contained in any such materials justified the policy determination that the BPD/DS procedure was “investigational.” Throughout this process, defendant continually reviewed the medical literature and remained steadfast in the conclusion that the BPD/DS is “investigational,” and therefore not covered, even though its representatives were not very specific in explaining this position: they simply stated that the literature indicated “significant outcome problems” with the BPD/DS. Dr. Ryce has been more forthcoming at the summary judgment stage, however, and a declaration containing his gloss on the medical evidence in the Administrative Record is the primary document upon which defendant now relies.

C. The Medical Evidence

The Administrative Record in this case is heavily laden with medical literature, much of which is devoted to explaining the techniques for performing various types of surgical procedures for morbid obesity, and distinguishing one variety from the others. The court is pleased to announce that its central concern is not the distinctions in the methods employed, but the differences, if any, in the results obtained and complications suffered. Therefore, a brief comparison of the BPD/DS and RNY procedures will suffice.

1. Overview of the BPD/DS and RNY surgeries

Both of the gastric surgeries at issue in this case involve tinkering with the arrangement of the small intestine, and adjusting the size of the stomach itself. In an unaltered system, food travels down the esophagus and is digested in the stomach. The stomach has an opening at the bottom, commonly called the “gastric outlet.” During digestion, however, the gastric outlet is sealed by the “pylorus,” a muscular valve that opens and closes as necessary. When the pylorus does open, it allows food to enter into the upper portion of the small intestine, or the “duodenum.” From there, the food continues to move through the small intestine, first reaching the middle section (called the “jejunum”), and finally arriving at the terminal portion (called the “ileum”). Subsequently, the mostly digested food pours into the large intestine through the “cecum,” a large pouch-like formation.

An assortment of obesity surgeries currently on the market change this process with the goal of inducing significant weight loss. One document in the Administrative Record divides the available procedures into three categories:

• Gastric restrictive surgical procedures create a small gastric pouch, resulting in weight loss for early satiety and decreased dietary intake. The decreased capacity of the stomach reduces the volume of food an individual consumes before feeling full.

• Combination surgical procedures include decreasing the stomach capacity and bypassing part of the digestive tract. They combine food malabsorption and [a reduction in] the volume of food an individual can consume.

• Malabsorptive surgical procedures bypass a section of the small intestines. Weight loss results from intestinal malabsorption without dietary modification.

Both the RNY and BPD/DS fall into the category of “combination surgical procedures,” because they are intended to make use of both the malabsorptive and restrictive approaches.

The RNY procedure is an older technique than the BPD/DS, tracing its roots to 1969; at present, it is by far the most commonly performed obesity surgery. The procedure involves “a horizontal or vertical partitioning of the stomach, which results in a 90% restriction [in stomach size and capacity]. It is followed by a Roux Y procedure in which the small intestine is reconfigured into a Y consisting of two limbs and a common channel.” In the end, the gastric outlet of the stomach is no longer connected to the duodenum, but instead is fused with the jejunum, resulting in a shorter intestinal pathway. This sort of gastric bypass

not only prevents the ability to ingest larger volumes [of food] at any one meal, but also induces a “dumping syndrome” if the patient ingests too much food or a high-sugar liquid meal. This unpleasant “dumping syndrome,” occurs when a large amount of partially digested food is delivered directly to the part of the small intestine from the stomach and can cause nausea, weakness, sweating, faintness, abdominal pain and vomiting. ... Since a major portion of digestion occurs in the stomach' — specifically in the process of breaking down food into nutrients — the amount of nutrients available for absorption is also reduced.

Like the RNY, the BPD/DS dates back many years, but it is not as old as the RNY, and it was modified several times before reaching its present form. In 1979, an article by Dr. Nichola Scopinaro described the Biliopancreatic Diversion without duodenal switch (“BPD”) as “a mixed or hybrid surgery consisting of restrictive ... and malabsorptive ... components.” The BPD turned out to be somewhat of a dangerous procedure, due primarily to a high incidence of post-operative protein malnutrition (one estimate at 11.9%) and anemia (one estimate at 35%). In 1988, Drs. Douglas W. Hess and Douglas S. Hess used “a combination of Dr. Scopinaro’s biliopancreatic diversion and Dr. DeMeester’s duodenal switch procedures ... [and] developed a hybrid which has the advantages of the [BPD] without some of the associated problems.” This BPD/DS is now regularly performed at over thirty university and non-academic medical centers. The procedure calls for the creation of a “sleeve-type” gastrectomy to limit oral intake of food and encourage early satiety, followed by construction of an alimentary limb that bypasses a portion of the small intestine called the ileum, in order to induce malabsorption. One effect of this design is to preserve the pylorus valve at the lower end of the stomach that controls the entrance of food into the upper portion of the small intestine. The continued presence of the pylorus reportedly “affords normal gastric functioning,” and eliminates “dumping syndrome.”

Although more technically demanding than an “open” surgery, both the RNY and the BPD/DS can be performed laparoscop-ieally as well. Using a laparoscope “improves recovery time and reduces periop-erative morbidity.” Other “[advantages of the laparoscopic-assisted approaches include decreased pain, improved pulmonary function in the earlier postoperative period, reduced hospital stay, and a more pleasant cosmetic result.”

2. Potential for weight loss: BPD/DS vs. RNY

Both the RNY and BPD/DS procedures result in considerable weight loss. The Administrative Record does not contain much weight loss data specific to the RNY, but one article reports that “[l]ong-term weight loss at 5 to 15 years seems to be 49% to 77% of excess body weight.” Several studies of patients undergoing BPD/DS documented similar or better results with that procedure: “Hess had a [excess body weight reduction] of 80[%] at 2 years and 70[%] at 8 years. Marceau had a 73% [reduction] at 4.5 years, and 87% of his patients had [an excess body weight reduction greater than] 50[%].” At 10 years, “Hess has previously reported mean excess weight loss of 73% ... following open [BPD/DS].” With respect to laparoscopic BPD/DS, “[p]ublished 10-year data with 93% follow-up demonstrate sustained average loss of 76% of excess weight.”

Comparing the two procedures, one publication on file notes that the RNY “can produce an approximate 35% total weight reduction with a 40% to 85% decreased excess body weight,” while “[t]he [BPD/DS] procedure results in a 70% to 80% excess weight loss.” This data has led many of the authors of the literature in the Administrative Record to conclude that “BPD-DS is the most effective operation to lose weight.” Indeed, defendant concedes that the BPD/DS yields superior weight loss when compared with the RNY.

3. Complications: BPD/DS vs. RNY

A review of the literature in the Administrative Record makes plain that surgery to alleviate morbid obesity is not foolproof. Given the risks associated with that condition, however, neither can it be characterized as foolhardy. While the morbidly obese are much more likely to develop, among other things, diabetes and life-threatening heart conditions, “it appears that the reduction in incidence of diabetes and cardiovascular risk factors is excellent with either a gastric restrictive or malab-sorptive procedure.” To some extent, though, the complications associated with the various procedures differ.

a. Death

At the outset, it must be acknowledged that both the RNY and BPD/DS carry at least a marginal risk of death. Significantly, although each of the articles in the Administrative Record express the view that any loss of life is regrettable, none seem to single out the death rate associated with either surgical procedure as a major surprise. Dr. Ryce, on the other hand, appears to view the loss of life as a particularly shocking potentiality with the BPD/DS. Citing a BPD/DS study by Dr. Baltasar, Dr. Ryce points out that “[s]even of his patients suffered serious complications. Some died early, and some died later on; one from liver failure and another from attempts to treat the surgically-induced protein malnutrition.” Of course, Dr. Ryce omits mention of the fact that these two deaths occurred in patients who were undergoing conversions from different types of previously performed obesity surgeries. In fact, “[tjhere was no mortality in the 102 primary operated patients” who received BPD/DS without having previously submitted to different sorts of weight loss surgeries. Moreover, one of the two patients was a super-morbidly obese individual with a BMI of 94, whereas plaintiff here has a BMI around 43. On the whole, “[t]he mortality rate for all 125 patients [studied by Dr. Baltasar was] 1.6%.” This seems to be consistent with the statement in one document in the file that “[mortality [following the BPD/DS] is similar to the Roux Y procedure.”

b.Liver failure

Beyond the risk of death, there are additional side effects common to both procedures. According to one article in the Administrative Record, “[approximately 95% of all bariatric surgical patients have some [preexisting] liver pathology” — ie., “[t]he anatomical or functional manifestations of a disease.” It has been suggested that obesity surgery — especially BPD without DS — may exacerbate or cause liver problems, and Dr. Ryce mentions this possibility as a potential complication with BPD with DS as well. Indeed, in one study of 440 patients undergoing BPD/DS procedures, one individual died ten months following the operation from what was described as “fatty liver, liver failure, renal failure ... multiple organ failure.” Another study — the one cited by Dr. Ryce to substantiate his concerns about liver failure with BPD/DS — reported that one patient, out of the 125 who had BPD/DS procedures, passed away due to liver failure. Liver failure is thus a possibility, but “is rarely reported” with BPD/DS, and is apparently not a significant concern with RNY either. In fact, one article suggests that “[t]he correction of the dsymeta-bolic state associated with obesity also improves liver function and morphology.... Even when cirrhosis is present, surgery can reverse the liver condition.”

c. Gallbladder removal

Dr. Ryce noted emphatically in his declaration that “Dr. Hess finds it necessary to perform a removal of the gallbladder even if normal, in every patient he does [the BPD/DS] procedure on.” Neither Dr. Hess nor Dr. Ryce explain why this is significant, but other evidence on file suggests that concomitant removal of the gallbladder is not unique to BPD/DS surgeries: “The gallbladder, draining into the bypassed duodenum following [RNY procedures], becomes not only a useless structure but a liability. Patients have an approximately 45% chance of developing gallstones or sludge during the 6 months of massive weight loss following bariatric surgery.”

d. Anemia, protein deficiency, and other metabolic concerns

Anemia is a concern with both RNY and BPD/DS surgeries. In his declaration, Dr. Ryce asserts that anemia following BPD/DS soars to 35%, and cites an article that is included in the Administrative Record to support that statistic. Dr. Ryce is mistaken in this claim, however: the statistic cited actually refers to anemia levels following BPD procedures without the accompanying duodenal switch. Truth be told, the incidence of post-operative anemia with BPD/DS surgery appears to hover closer to 9%, and even in the BPD without DS procedures, where anemia was a more pervasive problem, oral supplementation with iron brought the number of sufferers down to around 5% in at least one study. Although there is no data on file relating the incidence of anemia with RNY procedures, evidence in the Administrative Record shows that oral supplementation with iron is absolutely essential following RNY procedures as well as BPD and BPD/DS operations.

According to Dr. Ryce, there is an “11.9% incidence of protein malnutrition” and a “30% incidence of bone loss” following BPD/DS surgery. These data are drawn from the same article cited to support Dr. Ryce’s discredited claim concerning the high rate of anemia following BPD/DS operations. Again, both the 11.9% protein malnutrition statistic, and the 30% bone loss number, refer to patients who submitted to BPD surgery without DS. As far as the court can tell, that article does not even list bone loss or protein malnutrition as complications associated with the BPD/DS procedure, which is discussed on a separate page. Even so, other articles do support the claim that these complications occur in BPD/DS patients — -just not to a dramatic degree. In fact, “BPD-DS has almost halved the prevalence of most side effects [of the BPD].” While only limited bone loss data is included in the Administrative Record, one study reports that, “[i]n a review of 747 patients after BPD [without duodenal switch] or BPD-DS and a mean followup [sic] of eight years, the incidence of fracture was 2% per year, which is within the normal limits for the general population.” Moreover, another author calls attention to the fact that “[t]he bone problems caused by BPD [without DS] do not seem to differ substantially from those reported after ... gastric bypass for obesity.” With both RNY and BPD/DS operations, lifelong calcium supplementation is considered “mandatory.”

Protein deficiency is a very serious complication that is occasionally confronted in BPD procedures where the patient does not also have the duodenal switch. However, “[t]he yearly hospitalization rate for protein deficiency is about 1% after BPD-DS' .... The yearly revision rate because of hypoalbuminemia [a condition caused by a lack of protein] is 0.1% after BPD-DS'.” In the Blue Cross and Blue Shield Association’s own Medical Policy Reference Manual, it is observed that, “in the duodenal switch group [of one study], there was a lower incidence of metabolic abnormalities such as protein malnutrition, which prompted reversal of the procedure in 1.7% of those undergoing biliopancreatic bypass versus only 0.1 % after the duodenal switch procedure.” ^ Finally, an abstract in the Administrative Record concludes that “[n]o statistically significant ... difference in the occurrence of deficiency was observed [in patients undergoing BPD as opposed to RNY] for any of the nutritional parameters studied,” and that “[t]he incidence of hypoalbuminemia was negligible in both groups.”

e. Intestinal bacterial overgrowth and diarrhea

Dr. Ryce points to the possibility of intestinal bacterial overgrowth as another “major risk” of the BPD/DS. In fact, the only article cited to confirm this claim appears to consider intestinal bacterial overgrowth a possibility with all “BPD” procedures, which the authors defines as all operations that “divert[] bile and pancreatic juice from their normal entrance and mov[e] them farther down the alimentary tract.” Among those operations are the BPD without DS, the BPD with DS, and the RNY. Only after prefacing the discussion with that critical clarification does the article move on to state that “[t]he risks involved with BPD must be emphasized,” including the risk of “intestinal bacterial overgrowth.” What is more, the authors conclude their substantive analysis of this risk by stating that “[w]e know that the serious form [of intestinal bacterial overgrowth] is rare, not life-threatening, and usually managed by antibiotics.”

Another side effect that is often shared by patients who undergo the BPD/DS and those who choose the RNY is diarrhea. In the RNY patients, this is often due to the “dumping syndrome” mentioned previously. The Administrative Record does not reveal the percentage of RNY patients who experience dumping syndrome or post-operative diarrhea, but it does clearly show that those who undergo the BPD/DS do not suffer from dumping syndrome! One article includes statistics indicating that, of 909 patients who underwent the BPD/DS operation, 9.4% suffered from diarrhea. In a different study of BPD/DS patients, revisions were performed on 2 patients out of 440 — or around 0.4%' — due to excess diarrhea. Yet another article reports that 1 patient out of 125 studied required a revision due to diarrhea and anal leakage, and concluded that, overall, “[f]requency of stool was not a major problem. Most patients have 1-2 stools very early in the morning and 1-2 throughout the rest of the day.”

A central difficulty in evaluating all of these studies is the fact that none of the articles in the Administrative Record define what constitutes “diarrhea.” One medical dictionary merely states that it is “[ejxcessive and frequent evacuation of watery feces.” Nonetheless, after reading the abstract of a published study by Dr. Anthone which reported that the “mean” or average number of bowel movements in patients who underwent the BPD/DS operation was three per day, Dr. Ryce concluded that “[ajbout one half of his patients developed diarrhea post-op.” Dr. Ryce does not offer for comparison the average number of daily bowel movements in RNY patients, or the percentage of such patients who suffer from diarrhea. It is noteworthy, however, that the converse of diarrhea — vomiting—is apparently a postoperative side effect in about 25% of RNY patients.

f. Reoperation rates

One measure of operational success is the rate of revision, or reoperation. Dr. Ryce notes that two prominent supporters of the BPD/DS procedure — Dr. Douglas W. Hess and Dr. Douglas S. Hess — had to perform revisions on 17 of the 440 patients in their study, for a revision rate of approximately 3.8%. Charts in the Technology Assessment published by the national Blue Cross and Blue Shield Association reveal reoperation rates as high as 35% and as low as 0% in studies of the RNY procedure.

g. Tendency of BPD/DS to reduce certain RNY-specific complications

Although much of the foregoing discussion has confronted complications endemic to the BPD/DS, it should be noted that the BPD/DS actually reduces or eliminates a number of complications associated with the RNY. One documented difficulty following RNY procedures is painful ulceration in the stomach area, near the anastomotic site. The evidence indicates, however, that “the addition of the duodenal switch procedure [renders] the possibility of a marginal ulcer ... remote,” and may actually prevent them altogether. As mentioned above, another source of concern for RNY patients is so-called “dumping syndrome,” an unpleasant and dangerous condition, which, to be avoided, requires severe dietary limitations. The articles in the Administrative Record are all in agreement that BPD/DS not only reduces, but totally eliminates, dumping syndrome. Thus, the vomiting and dietary precautions associated with the RNY are generally not seen with the BPD/DS.

The RNY also presents the risk of chronic gastritis, an inflammation of the mucous membrane that lines the stom-aeh. “Because of the fragile mucosa at the gastrojejunostomy, gastric irritants such as aspirin and non-steroidal anti-inflammatory drugs (NSAIDs) are contraindicated in [RNY] and [BPD without DS]. In contrast, [BPD/]DS patients have available a full range of food and medication choices.” This is particularly important to the plaintiff in the present action, because she takes NSAIDs for pain emanating from a foot condition known as plantar fasciitis. Undergoing RNY surgery would preclude further reliance on this form of treatment for her co-morbidities, assuming those conditions persisted following the procedure.

One condition that would almost certainly remain after a weight loss operation would be plaintiffs alleged predisposition to stomach and intestinal cancer. According to plaintiff, visualization of the stomach through an endoscope is an excellent method of early detection of these cancers. Although the Administrative Record does not appear to contain any medical literature confirming (or disproving) that endoscopy is useful in this regard, both parties agree that the RNY procedure makes endoscopic visualization of the entire stomach and duodenum impossible. Conversely, endoscopic visualization remains feasible following the BPD/DS operation, because there is no division of the stomach into upper and lower pouches.

D. Defendant’s Reasons for Classifying the BPD/DS as “Investigational”

As discussed above, defendant was somewhat tight-lipped as to its reasons for refusing to cover the BPD/DS procedure during the internal appeals process, and the Administrative Record in this case reflects that stance. Dr. Ryee, however, has guided the court through his reasoning in a declaration that includes citations to materials in the Administrative Record. In addition to attempting to poke holes in the medical literature provided to the court—which, with only one equivocation, wholly endorses the BPD/DS for patients like plaintiff—Dr. Ryce points to the conclusions of his company and other insurers as justification for the continued “investigational” status of the BPD/ DS.

Dr. Ryce states in his declaration that he found defendant’s own written medical policy on gastric surgery for morbid obesity persuasive. That policy contains the following discussion of the BPD/DS operation:

Biliopancreatic Bypass with or without Duodenal Switch has malabsorptive properties and eventual metabolic complications have been demonstrated. Techniques of gastric bypass that produce clinical manifestation of malabsorption should be used only in selected patients, in whom conventional Roux-en-Y is expected to fail and who are committed to vigilant long-term follow-up needed to minimize the incidence of serious metabolic complications.

Dr. Ryce also cited to a Technology Evaluation Assessment created by the national Blue Cross and Blue Shield Technology Evaluation Center. According to Dr. Ryce, “[t]he TEC Assessment concludes that there is not sufficient peer-reviewed literature to warrant the increased risks to the patient from a malabsorptive procedure, such as the BPD/DS, versus a gastric restrictive procedure such as the Roux-en-Y procedure that Blue Cross approved.” This characterization of the conclusions in the TEC is only partially correct. The TEC does in fact conclude that there is insufficient medical evidence to authorize the BPD/DS. However, the entire purpose of the TEC in the Administrative Record, insofar as evaluation of the BPD/DS and BPD surgeries is concerned, was the possibility of utilizing those surgeries for the treatment of super-morbid obesity. Indeed, on the very page cited by Dr. Ryce, the authors of the TEC pose the question: “Are outcomes of bilio-pancreatic diversion and/or long-limb gastric bypass as good as outcomes of open gastric bypass for patients with super-obesity

Another source of authority relied upon by Dr. Ryce was the gastric surgery policy of Aetna, a large medical insurance company and third-party payor. That policy, which is contained in the Administrative Record, provides in pertinent part as follows:

Although it appears that open or laparo-scopic DS is a feasible and promising operation resulting in effective weight loss with an acceptable morbidity, long-term studies are needed to confirm safety of the procedure, the durability of results, and proper patient selection criteria. At present there is inadequate scientific data to support the performance of this procedure. Studies of the DS procedure (open or laparoscopic) published to date have been limited to case reports and reports of uncontrolled case series. Tthere [sic] are no published clinical trials directly comparing the DS procedure to established surgical procedures. Thus, in a recently published review of the evidence, Gentlieschi, et al. (2002) concluded that, based on currently available evidence, the DS procedure appears feasible, but that randomized controlled clinical trials are needed comparing malabsorptive procedures (including the duodenal switch) to other bariatric operations. The authors concluded that “further studies are needed to determine the safety and effectiveness of this procedure.!”’] [sic] An assessment conducted by the BlueCross BlueShield Association Technology Evaluation Center (2003) stated: “There are limited data on outcomes of bilio-pancreatic diversion and/or long-limb gastric bypass for patients with super-obesity. There are no high-quality comparative trials and only limited clinical series data for these indications. These limited data do not establish that these or other variants (e.g., duodenal switch) have any additional benefit for patients with super-obesity, as compared to gastric bypass.”

The Blue Cross and Blue Shield Medical Policy Reference Manual, published by the national Blue Cross and Blue Shield Association, ostensibly provides further support for Dr. Ryce’s decision:

As noted in the Policy section, this policy suggests that malabsorptive procedures for treatment of morbid obesity remain investigational. This interpretation of the term investigational may be questioned by those who would point out the procedure, particularly the Scopina-ro procedure, has been performed for some 20 years with results of large case series reported in the peer-reviewed literature. The percent of excess weight loss, typically at or above 70%, may be higher than that reported with gastric restrictive procedures, reported at around 60%, but higher among those patients who maintain intact stomas. However, one of the criteria used to define the term investigational, as defined in the Introduction to the Medical Policy Reference Manual, is whether the malabsorptive procedures are at least as good as the alternatives; i.e., gastric restrictive procedures. This involves a judgment as to whether the acknowledged increased metabolic risks associated with malabsorptive procedures are more than outweighed by an increased benefit associated with potentially greater weight loss. While most of the studies of bariatric surgeries report results in terms of weight loss, the degree of weight loss is essentially an intermediate outcome. For questions that ask whether surgery improves health outcome, and/or how much surgery improves outcomes, weight loss by itself is useful only if the relationship between the amount of weight loss and the degree of improvement in health outcomes has been established. The underlying medical rationale for the surgery, and thus the basis for its coverage eligibility, is not the degree of weight loss, but the decreased risk of the morbid complications of obesity, i.e. a decreasing incidence of diabetes and cardiac risk factors, among others.... A 2003 TEC special report focused on the relationship between weight loss and changes in morbidity after weight loss surgery. While this report concluded that the evidence was sufficient to conclude that surgery improves health outcomes for patients with morbid obesity as compared to nonsurgical treatment, there were inadequate data to draw conclusions as to the relation of increment of weight loss to increment of improvement in health outcome measures. There were also insufficient data to identify a weight loss threshold for success of a surgical procedure.

Ideally, one would like to compare the incidence of morbidities in gastric restrictive versus malabsorptive procedures. However, there is no report of a head-to-head comparison among similar patients. It is difficult to compare results between case series due to variations in surgical procedures and different outcome measurements. In addition, the literature focuses on the degree of weight loss and not the incidence of obesity-related morbidities. However, it appears that the reduction in incidence of diabetes and cardiovascular risk factors is excellent with either a gastric restrictive or malabsorptive procedure. Therefore, this policy regarding the investigational status of malabsorptive procedures is based on the judgment that there is insufficient evidence to demonstrate that the increased risks of malabsorp-tive procedures compared to restrictive procedures are outweighed by a significantly greater reduction in obesity-related morbidities.

Based on these sources of information, and his own take on the medical literature, Dr. Ryce and defendant continue to assert that the classification of the BPD/DS as “inves-tigational” should be upheld.

On the other hand, plaintiff argues that not only is defendant’s classification incorrect, it is also suspect, because the effect of the determination not to authorize the BPD/DS is to save defendant money. In support of this position, plaintiff submitted along with her summary judgment brief the declaration of Ellen Bell, the Insurance Administrator at the Metabolic Surgery Center in Nashville, Tennessee. According to Bell, the RNY procedure would be billed out at a total price of $30,075, whereas the cost of the laparo-scopic BPD/DS operation would be $41,-450 — a difference of $11,375. Of that, Bell states that $8,200 would go to the surgeon performing the procedure for an RNY, and $10,200 would go to the surgeon for a BPD/DS. Defendant responds, through Dr. Ryce, by stating that the actual amount billed by the hospital has little or nothing to do with the amount that defendant would pay for a particular procedure. Rather, Dr. Ryce submits that, “regardless of whether one compares reimbursement rates for the Roux-en-Y and the BPD/DS for 2004 or 2005, Blue Cross would have .provided higher reimbursement in the amount of $542 to a nonparticipating surgeon (like Dr. Husted) for the approved Roux-en-Y procedure.”

Part Three

Standard of Review Under ERISA

As a threshold matter, the court must determine the appropriate standard for reviewing defendant’s interpretation of the policy, because ERISA itself does not specify the standard applicable to the decisions of a claims administrator or other fiduciary. See Jordan v. Metropolitan Life Insurance Co., 205 F.Supp.2d 1302, 1305 (M.D.Fla.2002) (citing, e.g., Marecek v. BellSouth Telecommunications, 49 F.3d 702, 705 (11th Cir.1995) (other citation omitted)).

The Supreme Court held in Firestone Tire & Rubber Co. v. Bruch, 489 U.S. 101, 109 S.Ct. 948, 103 L.Ed.2d 80 (1989), that “a denial of benefits ... is to be reviewed under a de novo standard unless the benefit plan gives the administrator or fiduciary discretionary authority to determine eligibility for benefits or to construe the terms of the plan.” Id. at 115, 109 S.Ct. 948. Pivoting off the Bruch decision, the Eleventh Circuit has promulgated three standards of review applicable to the decisions of a claims administrator: “(1) de novo where the plan does not grant the administrator discretion; (2) arbitrary and capricious [where] the plan grants the administrator discretion; and (3) heightened arbitrary and capricious where there is a conflict of interests.” Buckley v. Metropolitan Life, 115 F.3d 936, 939 (11th Cir.1997). Here, the plan grants the claims administrator the discretion to construe the terms of the plan. Further, because defendant both serves as claims administrator and pays claims from its own funds, it has a conflict of interest. See Brown v. Blue Cross & Blue Shield of Alabama, 898 F.2d 1556, 1568 (11th Cir.1990) (holding that “a strong conflict of interest exists when the fiduciary making a discretionary decision is also the insurance company responsible for paying the claims”); see also Fick v. Metropolitan Life Insurance Co., 347 F.Supp.2d 1271, 1285 (S.D.Fla.2004) (“Under Eleventh Circuit authority, the plaintiff need not prove that the defendant’s self-interest was in fad a motivating factor in the decision to deny benefits.”) (emphasis supplied). Defendant accordingly concedes that “[t]he heightened arbitrary and capricious standard of review applies in this case because Blue Cross serves both as the insurer and the claims administrator of the Book Systems Plan.”

The applicability of the heightened arbitrary and capricious standard does not mean, however, that the conflict of interests is the court’s sole (or even primary) focus. Instead, the court must first determine — on de novo review — whether defendant’s interpretation of the policy provision at issue was “wrong.” See HCA Health Services of Georgia v. Employers Health Insurance Co., 240 F.3d 982, 993 n. 23 (11th Cir.2001) (“ ‘Wrong’ is the label used by our precedent to describe the conclusion a court reaches when, after reviewing the plan documents and disputed terms de novo, the court disagrees with the claim administrator’s plan interpretation.”). Given the nature of de novo review, the court is obliged at this stage to consider all the evidence in the administrative record, plus “[relevant] factual opinions and evidence bearing on the claimant’s entitlement to benefits that were not part of the administrative record at the time the administrator denied coverage.” Wise v. Hartford Life & Accident Insurance Co., 360 F.Supp.2d 1310, 1325 (N.D.Ga.2005) (citing Moon v. American Home Assurance Co., 888 F.2d 86, 89 (11th Cir.1989) (concluding that limiting review to the facts available at the time of the administrator’s denial “is contrary to the concept of a de novo review”)).

Only if the court deems the administrator’s interpretation to be erroneous on de novo review must it proceed further with the analysis. See, e.g., Brown, 898 F.2d at 1566 n. 12 (“It is fundamental that the fiduciary’s interpretation first must be ‘wrong’ from the perspective of de novo review before a reviewing court is concerned with the self-interest of the fiduciary.”) (italics supplied); Freling v. Reliance Standard Life Insurance Co., 315 F.Supp.2d 1277, 1286 (S.D.Fla.2004) (“[A] court need only consider whether the administrator’s decision was ‘tainted by self-interest,’ if it first finds that the administrator’s determination is wrong.”) (citation omitted). Simply put, if defendant here correctly concluded that the BPD/DS is “investigational” under the policy held by plaintiff, the inquiry is at an end, and summary judgment is due to be granted in defendant’s favor.

However, “[i]f the court determines that the claims administrator’s decision is ‘wrong,’ the court then proceeds to decide whether ‘the claimant has proposed a “reasonable” interpretation of the plan.’ ” HCA Health Services of Georgia, 240 F.3d at 994 (quoting Lee v. Blue Cross/Blue Shield, 10 F.3d 1547, 1550 (11th Cir.1994)). Assuming claimant’s position is held to be untenable, summary judgment should be granted to the administrator on that basis alone. See id. Alternatively, if the claimant’s reading of the provision at issue is reasonable, the court must again test the administrator’s wrong interpretation, this time through the less discerning lense of “arbitrary and capricious” review. See, e.g., id.; Florence Nightingale Nursing Service, Inc. v. Blue Cross/Blue Shield, 41 F.3d 1476, 1481 (11th Cir.1995). An administrator’s incorrect interpretation is not arbitrary and capricious if it is “reasonable”: i.e., if “there was a reasonable basis for the decision, based upon the facts known to the administrator at the time the decision was made. ” Jett v. Blue Cross & Blue Shield of Alabama, 890 F.2d 1137, 1139 (11th Cir.1989) (emphasis supplied).

When heightened review is in place, if both the claimant’s interpretation and the administrator’s wrong interpretation are reasonable, there is yet another level of scrutiny to be applied. See HCA Health Services of Georgia, 240 F.3d at 994 (noting that, “[t]he reason the claimant’s reasonable interpretation does not [automatically] trump the claims administrator’s wrong interpretation is because the plan documents explicitly grant the claims administrator discretion to interpret the plan”).

Under the heightened arbitrary and capricious standard of review, the burden shifts to the claims administrator to prove that its [erroneous, but reasonable] interpretation of the plan is not tainted by self-interest. The claims administrator satisfies this burden by showing that its wrong but reasonable interpretation of the plan benefits the class of participants and beneficiaries. Even when the administrator satisfies this burden, the claimant may still be successful if he can show by other measures that the administrator’s decision was arbitrary and capricious. If the court finds that the claims administrator fails to show that its plan interpretation benefits the class of participants and beneficiaries, the claims administrator’s plan interpretation is not entitled to deference [and the claimant may be entitled to summary judgment].

Id. at 994-95 (internal citations omitted). The court unfolds and applies each step in this somewhat convoluted analysis below.

Part Four

Was Defendant’s Decision ‘Wrong”?

The initial task before the court is de novo interpretation of the policy provisions at issue. See Denton v. First National Bank of Waco, 765 F.2d 1295, 1304 (5th Cir.1985) (“First, the court must determine the correct interpretation of the Plan’s provisions.”). Next, the correct legal interpretation of the policy provisions must be applied to the facts at hand. See, e.g., Freling v. Reliance Standard Life Insurance Co., 315 F.Supp.2d 1277, 1287 (S.D.Fla.2004). At the outset, the court notes that the policy in question does clearly exclude “investigational” procedures from coverage. The issue here is simply whether the BPD/DS is properly considered “investigational.”

A. The Correct Interpretation of the Policy Provisions

The Summary Plan Description defines “investigational” as “[a]ny treatment, procedure ... that either we have not recognized as having scientifically established medical value, or that does not meet generally accepted standards of medical practice.” Although it might at first appear that this is the definition to be applied by the court on de novo review, the terms of the Summary Plan Description indicate otherwise. The policy contemplates the that initial construction of the “investigational” definition often will be promulgated in internally-produced written medical criteria. Here, defendant produced the Gastric Restrictive Procedures Policy, which concludes that the BPD/DS falls within the category of “in-vestigational” operations. Under the terms of the definitional section, since a written policy labeling the BPD/DS as “in-vestigational” has been created, that policy endures until such time as a series of enumerated elements is met. In this case, Dr. Ryce and defendant’s other representatives repeatedly reviewed the materials submitted by plaintiff and determined that the elements listed were not met. Thus, the court must review not the initial application of the “investigational” definition in the Gastric Restrictive Procedures Policy, but defendant’s determination that the BPD/DS did not satisfy the elements necessary to permit removal of the “inves-tigational” designation.

The Summary Plan Description lists the following five elements as prerequisites to revision of an “investigational” determination:

• The technology must have final approval from the appropriate government regulatory bodies;

• The scientific evidence must permit conclusions concerning the effect of the technology on health outcomes;

• The technology must improve the net health outcome;

• The technology must be as beneficial as any established alternatives; and,

• The improvement must be attainable outside the investigational setting.

Having determined the proper focus of the inquiry, the court must move to the next step in the analysis.

B. The Correct Application of the Policy Provisions to the Facts

Significantly, the court need not consider three of the elements listed in the Summary Plan Description. Defendant admits that there is no “government regulatory bod[y]” in the business of approving surgical procedures. Thus, this element is not a barrier to lifting the “investigational” label; it is simply irrelevant in this context. Further, it is apparently conceded that “[t]he scientific evidence ... permit[s] conclusions concerning the effect of the technology on health outcomes.” Defendant’s main argument — that the BPD/DS procedure is not as safe or as effective as the RNY procedure — presupposes that there is sufficient data to “permit conclusions” concerning the procedure. Finally, defendant admits that “[t]he improvement [is] attainable outside the investigational setting,” and that, if defendant approved the BPD/DS, plaintiff would have the operation at a non-academic facility.

Ignoring these superfluous elements, the court must evaluate defendant’s answers to the two contested questions: (1) does “[t]he [BPD/DS] ... improve the net health outcome;” and (2) is “[t]he [BPD/DS] ... as beneficial as any established alternatives?”

1. Does the BPD/DS improve net health outcome?

It appears from the briefs that there is some dispute as to the correct application of the term “net health outcome.” Neither party, however, bothers to offer anything other than conclusory arguments that the standard has or has not been met. The policy does not define the term; and, as far as the court is concerned, there is no need to do so. Requiring that a procedure improve “net health outcome” is an eminently clear method of describing the need for an analysis of overall results, with an eye toward determining whether a profit was garnered, or a loss was incurred. Although the standard itself implies a patient-by-patient inquiry, its context (a health insurance policy designed to govern the plans of countless insureds) connotes a universal assessment of patient well-being. The term is therefore not ambiguous, and no one claims otherwise. Cf. Florence Nightingale Nursing Service, Inc. v. Blue Cross/Blue Shield, 41 F.3d 1476, 1481 n. 4 (11th Cir.1995). Thus, taking account of all patients in the proper category who have undergone the BPD/DS, the question the court must ask is: did the procedure more often than not leave the patient better off than she was before?

The evidence here is more than sufficient to clear this relatively low hurdle. As recounted exhaustively above, virtually every publication on file wholeheartedly vouches for the efficacy of the BPD/DS procedure. One article notes that “[t]he [BPD/DS] procedure results in a 70% to 80% excess weight loss, with 93% good or excellent results.” Another study reveals that, five years after undergoing the BPD/DS, 97% of patients observed had lost greater than 50% of their excess weight. In the face of these numbers, defendant rightly concedes that the BPD/DS yields superior weight loss when compared with the RNY. At the same time, however, defendant adds the caveat that the attainment of weight loss alone is not enough to call an operation successful. Indeed, one insurance article aptly observes that “[t]he underlying medical rationale for the surgery, and thus the basis for its coverage eligibility, is not the degree of weight loss, but the decreased risk of the morbid complications of obesity, i.e. a decreasing incidence of diabetes and cardiac risk factors, among others.” Fair enough, but even viewing the studies in the light most favorable to defendant, they reveal that Type II diabetes is cured in around 98% of BPD/DS patients, sleep apnea disappears, and the reduction in cardiac risk factors is “excellent.”

Although it might seem that this evidence alone is sufficient to support a finding of improved health, the benefits must still be reconciled with the complications in order to arrive at a “net” outcome. The most serious complication of the BPD/ DS — and the one principally, if not exclusively, relied upon in defendant’s briefs — is protein malnutrition. If protein malnutrition actually occurred more often than not in BPD/DS patients, the question whether BPD/DS “improves health outcomes” would be a difficult one. The court would probably be obliged to conclude that net health outcome was not improved, because, while the weight loss would ameliorate or eliminate certain morbidities, the life-threatening protein malnutrition would fill their void.

But protein malnutrition is not common by any means. In fact, the evidence in the Administrative Record shows that, even in those who undergo BPD ivithout DS, protein malnutrition occurs in only 11.9%. This is a significant percentage, no doubt, but still not sufficient to establish that net health outcome is not improved in most patients. More importantly, studies verify that this complication is even more rare in BPD/DS patients. Cf. Heasley v. Bel-den & Blake Corp., 2 F.3d 1249, 1259 (3d Cir.1993) (holding that “the term ‘experimental procedure’ must be defined in terms of