Jennie R. Lefevre, Sally M. Hill, Frederick L. Rada and Mary Christina Veldman v. Secretary, Department of Veterans Affairs

FRIEDMAN, Senior Circuit Judge. The issue on the merits is the validity of the determination of the Secretary of Veterans Affairs, pursuant to the Agent Orange Act of 1991, 38 U.S.C. § 1116 (Supp.V.1993) (the 1991 Act), not to create a presumption that prostate cancer, liver cancer, and nose cancer are connected to exposure to herbicides in Vietnam, which would be applied in determining eligibility for disability and survivor’s benefits. The Secretary contends that under the 1991 Act we do not have jurisdiction to review his determination. We *1193 hold that we have jurisdiction, and affirm his determination. I. A. In 1984, Congress recognized that it would be virtually impossible to determine on a case-by-case basis whether exposure to herbicides in Vietnam caused a disease in a particular veteran. 130 Cong.Rec. 13,159 (remarks of Senator Simpson). It therefore decided to require the Secretary to create or reject a presumption-of-service connection for particular diseases, based upon the statistical probability of such connection, as reflected in scientific studies of the relationship between those diseases and exposure to herbicides and the incidence of those diseases in persons and animals subject to herbicide exposure. Id. at 13,157-59; See Veterans’ Dioxin and Radiation Exposure Compensation Act of 1984, Pub.L. No. 98-542, 98 Stat. 2725 (1984 Act). By 1991 Congress had concluded that the Secretary’s administration of the 1984 Act had created doubt about the way the Act was being applied. A Senate Committee report on the 1991 Act stated: A number of reviews of the scientific literature on the effects of exposure to dioxin have been carried out under contract with VA and published by VA pursuant to the mandate in [the 1984 Act].... General acceptance of these reviews has been impaired because of concern that VA may have exerted influence on their content. Although the Committee does not share such a concern, it nevertheless recognizes that the perception of a possibility of some taint does exist and cannot be dismissed out of hand. Other than these reviews, the Committee is unaware of any other unified analysis of the results obtained from studies on the effects of dioxin exposure or of any up-to-date analysis. S.Rep. No. 101-82, 101st Cong., 1st Sess. 41 (1989). Congress therefore enacted the 1991 Act to provide a review, by an entity completely independent of VA, that will yield unified compilation and analysis of the results from the various scientific studies. In order to accomplish this result, the Committee bill would provide for an independent contract scientific organization— the National Academy of Sciences unless NAS is unwilling to undertake this effort— to undertake a comprehensive review and evaluation of all of the scientific evidence, literature, and studies — including the Selected Cancers (SC) Study being carried out by CDC — pertaining to the adverse health effects in humans or other animals of exposure to dioxin and other substances in herbicides used in Vietnam. Id. at 42. The 1991 Act directed the Secretary to seek to enter into an agreement with the National Academy of Science (the Academy or NAS), an independent non-profit, nongovernmental scientific organization, under which the Academy would “review and summarize the scientific evidence and assess the strength thereof, concerning the association between exposure to an herbicide used in support of military operations in Vietnam,” and “each disease suspected to be associated with such exposure.” 38 U.S.C. § 1116, Note, § 3(c). The agreement was to require the Academy to transmit to the Secretary written reports over ten years, the first within 18 months and the remaining ones at least every two years. Id., Note § 3(g)(1). Under the 1991 Act the Academy is directed, “[f]or each disease reviewed,” to determine (to the extent that available scientific data permit meaningful determinations)— (A) whether a statistical association with herbicide exposure exists, taking into account the strength of the scientific evidence and the appropriateness of the statistical and epidemiological methods used to detect the association; (B) the increased risk of the disease among those exposed to herbicides during service in the Republic of Vietnam during the Vietnam era; and (C) whether there exists a plausible biological mechanism or other evidence of a *1194 causal relationship between herbicide exposure and the disease. (2) the Academy shall include in its reports under subsection (g) a full discussion of the scientific evidence and reasoning that led to its conclusions under this subsection. Id., Note. § 3(d). Within 60 days of receiving the Academy’s report, the Secretary is to determine whether a “positive association existed between (A) the exposure of humans to an herbicide agent, and (B) the occurrence of a disease in humans....” 38 U.S.C. § 1116(b)(1). There is such a positive association “if the credible evidence for the association is equal to or outweighs the credible evidence against the association.” 38 U.S.C. § 1116(b)(3). If the Secretary determines, “on the basis of sound medical and scientific evidence,” that such positive association exists, he is required to establish a presumption of service connection for the disease. 38 U.S.C. § 1116(b)(1). In making his determinations, the Secretary is required to “take into account” the Academy’s report and “all other sound medical and scientific information and analysis available to the Secretary.” 38 U.S.C. § 1116(b)(2). “In evaluating any study for the purpose of making such determinations, the Secretary shall take into consideration whether the results are statistically significant, are capable of replication, and withstand peer review.” Id. B. Pursuant to its agreement with the Secretary, the Academy conducted an extensive investigation in which it “review[ed] and summarize[d] the strength of the scientific evidence concerning the association between herbicide exposure during Vietnam service and each disease or condition suspected to be associated with such exposure.” The result was a 764 page Report, which exhaustively reviewed the scientific evidence regarding the association between exposure to herbicides and various diseases. The Academy made no new studies, but reviewed existing evidence including various epidemiological studies and studies of biologic plausibility. The latter is biological evidence, usually from animal testing, that a particular agent is associated with a particular disease. The Report was based on the Academy’s review of more than 6000 abstracts of scientific or medical articles, approximately 230 of which were given detañed analysis. The Academy summarized the extent of the evidence of association between herbicide exposure and 31 specific health problems by placing each disease in one of four categories. The first category, diseases for which the scientific evidence constituted “Sufficient Evidence of an Association” “Between Specific Health Problems and Exposure to Herbicides,” contained five diseases, none of which is at issue in this ease. The second category was: Limited/Suggestive Evidence of an Association: Evidence is suggestive of an association between herbicides and the outcome but is limited because chance, bias, and confounding could not be ruled out with confidence. For example, at least one high-quality study shows a positive association, but the results of other studies are inconsistent. The Report placed three types of cancer in this category, one of which, prostate cancer, is at issue here. The third category was: Inadequate/Insufficient Evidence to Determine Whether an Association Exists: The avaüable studies are insufficient quality, consistency, or statistical power to permit a conclusion regarding the presence or absence of an association. For example, studies faü to control for confounding, have inadequate exposure assessment, or fafl to address latency. The Report placed 20 diseases in this category. Two of them — hepatobiliary (liver) and nasaJ/nasopharyngeal (nose) cancer — are at issue here. The fourth category, “Limited/Suggested Evidence of No Association,” contains 4 types of cancer. None of these is here involved. The Report noted that the Academy “was charged with reviewing the scientific evidence, rather than making recommendations regarding DVA policy, and [placement of the *1195 diseases in categories] is not intended to imply or suggest any policy decisions, which must rest with the Secretary.” The Report also considered separately whether exposure to herbicides increased risk for Vietnam veterans for each of the diseases. It concluded that “it is not possible for the Academy to quantify the degree of risk likely to have been experienced by Vietnam veterans because of their exposure to herbicides in Vietnam,” “[g]iven the large uncertainties that remain about the magnitude of potential risk from exposure to herbicides in the occupational, environmental, and veterans studies that have been reviewed, inadequate control for important confounders in these studies, and the lack of information needed to extrapolate from the level of exposure in the studies reviewed to that of individual Vietnam veterans....” Id. In assessing the evidence of biologic plausibility, the Report noted: “Several studies ... have been performed in laboratory animals. In general they have produced negative results, although some were not performed using rigorous criteria for the study of cancer in animals, and some produced equivocal results that could be interpreted as either positive or negative.... [T]here is as yet no convincing evidence of, or mechanistic basis for, the carcinogenicity of any of the herbicides used in Vietnam.” The Report also addressed the chemical compound known as TCDD or dioxin, which was a contaminant “found in varying levels in different batches of Agents Orange, Pink, Purple, and Green.” The Academy reported that there was extensive evidence that dioxin caused cancer in laboratory animals. C. Upon receipt of the Academy’s Report, the Secretary announced that he would recognize as service connected the five diseases the Report had placed in the first category (“Sufficient Evidence of an Association”). The Department appointed a task force, “an internal, high-level panel” of experts, “to carefully analyze the report from a medical/seientific perspective and to solicit comments from representatives of veterans service organizations and other interested parties.” In a detailed report to the Secretary, the task force “attempt[ed] to translate the Academy’s evaluation of the scientific literature and other available information into terms compatible with the legal mandate of P.L. 102-4, i.e. whether the credible evidence for an association between exposure to herbicides and any specific disease was equal to or outweighed the credible evidence against an association. This is not a simple task. The Task Force also tried to evaluate additional evidence presented by the veterans service organizations and a recently published follow-up on the Seveso incident published in Epidemiology.” The task force concluded that the Academy “had done a thorough review and evaluation of the available scientific and medical information regarding health effects of the exposure to Agent Orange and other herbicides used during the Vietnam conflict. There were no obvious gaps in the information sources cited,” and “the approach taken by the NAS Committee was reasonable and scientifically sound and the organization of the report was understandable.” The task force made recommendations based on the Academy’s conclusions. It endorsed the Secretary’s decision that the five diseases that the Academy placed in its first category (“Sufficient Evidence of an Association”) “also meet the standard for a positive finding under the statute, i.e., the credible evidence for an association is equal to or outweighs the credible evidence against an association.” The task force spent substantial time on the three types of cancer that the Academy placed in the second category (“Limited Suggestive Evidence of an Association”). The task force concluded that two of those cancers (respiratory cancers and multiple myeloma) met the statutory standard and recommended the establishment of a presumption of service connection for them. It concluded, however, that prostate cancer did not meet the statutory standard and recommended not establishing a presumption of service connection for it. The task force concluded that none of the diseases that the Academy placed in the third category (“Inadequate/Insufficient Evidence of an Association”), which included liver and nose cancer, met the statutory standard, and recom *1196 mended not to establish a presumption of service connection for them. The Secretary adopted the task force’s recommendations, and published a detailed explanation of his decision in the Federal Register. 59 Fed.Reg. 341-46 (1994). He explained his reasons for denying a presumption of service connection for most of the diseases, including the three here at issue. The Secretary discussed the scientific reports, both pro and con, upon which he based his decision. The Secretary requested that the Academy “focus particularly on the evidence regarding prostate cancer and peripheral neuropathy in its next review.” D. The four petitioners are a Vietnam veteran who has prostate cancer and three widows of Vietnam veterans who died of liver or nose cancer. They allege that the veterans’ exposure to herbicides in Vietnam caused the cancers. The veteran has a claim for disability compensation pending before the Department of Veterans Affairs. The three widows have pending their claims for dependency and indemnity compensation. The petition for review asserts that the Secretary’s denial of a presumption for the three types of cancer “will cause the VA imminently to deny their claims for compensation.” The petition for review challenges the Secretary’s determination “insofar as it determines not to allow veterans presumptive service connection, based on herbicide exposure in Vietnam, for prostate cancer, hepatobiliary cancers and nasal/nasopharyngeal cancer” (emphasis in original). The petitioners’ brief asserts that the Secretary’s denial of a “positive association” between those three types of cancers and exposure to herbicides in Vietnam was arbitrary, capricious and contrary to law. II. The jurisdiction of this court directly to review acts of the Secretary is set forth in 38 U.S.C. § 502 (Supp.V.1993): An action of the Secretary to which section 552(a)(1) or 553 of title 5 (or both) refers, other than an action relating to the adoption or revision of the schedule of ratings for disabilities under section 355 of title 38, is subject to judicial review. Such review shall be in accordance with [the Administrative Procedure Act] and may be sought only in the United States Court of Appeals for the Federal Circuit. Section 552(a)(1) requires each agency to publish in the Federal Register, among other things, “substantive rules of general applicability ... and statements of general policy or interpretations of general applicability....” Section 553, captioned “Rule making,” specifies the procedure for rulemaking. Thus, we may directly review rules promulgated by the Department of Veteran’s Affairs, including substantive rules of general applicability, statements of general policy and interpretations of general applicability because these are all actions “to which section 552(a)(1) refers....” Section 551(4) of title 5 defines a rule as the whole or part of an agency statement of general or particular applicability and future effect designed to implement, interpret, or prescribe law or policy or describing the organization, procedure, or practice requirements of an agency. As our predecessor court explained: “rule making is legislative in nature, is primarily concerned with policy considerations for the future rather than the evaluation of past conduct, and looks not to the evidentiary facts but to policy-making conclusions to be drawn from the facts.” American Express Co. v. United States, 472 F.2d 1050 (C.C.P.A. 1973) (citations omitted). These definitions cover the Secretary’s determination not to establish a presumption-of-service connection with respect to prostate, liver, and nose cancer and the veteran’s exposure to herbicides in Vietnam, and therefore we have jurisdiction to review it. The determination was a rule because it was a “statement of general ... applicability and future effect designed to implement ... or prescribe ... law or policy_” It prescribed the basis on which the Department would adjudicate every claim seeking disability or survivor benefits for specified diseases allegedly caused by exposure to herbicides in Vietnam. It reflects the result of a process that was “legislative in nature, [was] primari *1197 ly concerned with policy considerations for the future ..., and look[ed] ... to policy-making conclusions to be drawn from the facts.” Congress delegated to the Secretary the authority to determine whether or not to create a presumption of service connection between certain diseases and military service in Vietnam, and that determination would control the decisions in all subsequent cases involving the issue. The Secretary contends that his determination is not a substantive rule because it will not have binding effect, since a claimant may establish service connection for any disease or injury by showing that it was incurred in service. Not only is this position unrealistic, but it is inconsistent with the reasons that led Congress in the 1984 Act to adopt the procedure of the Secretary creating presumptions of service connection for various diseases associated with herbicide exposure. Congress did so because it recognized that ordinarily it would be impossible for an individual veteran to establish that his disease resulted from exposure to herbicides in Vietnam. Relying on the statement in Animal Legal Defense Fund v. Quigg, 982 F.2d 920, 927 (Fed.Cir.1991) that “a rule is substantive when it effects a change in existing law or policy which affects individual rights and obligations,” the Secretary contends that his determination was merely an announcement “that there would be no change in the existing law and policy regarding service connection for those diseases” (emphasis in original). The policy decision that the 1991 Act required the Secretary to make was whether or not, under the Act’s standards, to establish a presumption of service connection for various diseases. The Secretary’s determination not to establish a presumption for a particular disease was just as much a change in policy as his determination to establish a presumption for another disease. Until the Secretary had acted, there was no existing law regarding presumptions for the Department to follow. The Secretary changed existing law by establishing or refusing to establish a presumption. The Supreme Court dealt with a similar issue in Frozen Food Express v. United States, 351 U.S. 40, 76 S.Ct. 569, 100 L.Ed. 910 (1956). That ease involved the exception in § 203(b)(6) of the Interstate Commerce Act, 49 Stat. 543 (1935), from the requirement that motor carriers have certificates of public convenience and necessity, to carry “agricultural commodities.” “After an investigation instituted on its own motion [which included an evidentiary public hearing], the Commission issued an order that specified commodities are not ‘agricultural’ within the meaning of § 203(b)(6).” Id. at 41, 76 S.Ct. at 570. The order incorporated findings from an accompanying report that “list[ed] those commodities that the Commission [found] exempted under § 203(b)(6) and those that are not.” Id. at 42, 76 S.Ct. at 570. The order did not, however, direct carriers hauling the commodities declared non-exempt to comply with the certificate requirement. Id. at 45, 76 S.Ct. at 572 (Harlan, J., dissenting). Frozen Food Express, a motor carrier that was transporting commodities that the Commission had ruled were not “agricultural commodities” (and therefore required a certificate), filed suit before a three-judge district court challenging the Commission’s ruling. The district court “dismissed the action, saying that the ‘order’ of the Commission was not subject to judicial review,” Id. at 43, 76 S.Ct. at 570, because “the ‘order’ was no more than a report of an investigation ... Vhich does not command the carrier to do, or to refrain from doing anything.’” Id. (quoting United States v. L.A & SLR Co., 273 U.S. 299, 47 S.Ct. 413, 71 L.Ed. 651 (1927)). The Supreme Court reversed, holding that the Commission’s action was reviewable. It stated: The determination by the Commission that a commodity is not an exempt agricultural product has an immediate and practical impact on carriers who are transporting the commodities, and on shippers as well. The “order” of the Commission warns every carrier, who does not have authority from the Commission to transport those commodities, that it does so at the risk of incurring criminal penalties. § 222(a). *1198 Where unauthorized operations occur, the Commission may proceed administratively and issue a cease and desist order. § 204(c). The determination made by the Commission is not therefore abstract, theoretical, or academic. The “order” of the Commission which classifies commodities as exempt or nonexempt is, indeed, the basis for carriers in ordering and arranging their affairs. The “order” of the Commission is in substance a “declaratory” one, see 60 Stat. 240, 5 U.S.C. § 1004(d), which touches vital interests of carriers and shippers alike and sets the standard for shaping the manner in which an important segment of the trucking business will be done. Id. at 43-44, 76 S.Ct. at 571 (citations omitted). That reasoning covers the present case. The Secretary’s determination not to establish a presumption of service connection for prostate, liver and nose cancers “has an immediate and practical impact” on Vietnam veterans and their survivors who claim benefits on the ground that the veterans’ cancers resulted from exposure to herbicides, was not “abstract, theoretical, or academic,” “touches vital interests of’ veterans and their survivors, and “sets the standard for shaping the manner in which an important segment” of the Department’s activities “will be done.” “[JJudicial review of executive action “will not be cut off unless there is persuasive reason to believe that such was the purpose of Congress.’” Gutierrez de Martinez v. Lamagno, — U.S. -, -, 115 S.Ct. 2227, 2231, 132 L.Ed.2d 375 (1995) (citation omitted). Not only is there no showing that Congress so intended, but the few references to the subject in the legislative history of the 1991 Act support reviewability. During the debate on the 1991 Act, Senator Daschle stated that the Secretary’s determination that a particular disease does not warrant a presumption of service connection would be subject to judicial review. For example, the Secretary’s published determination that a presumption of service connection is not warranted for a particular disease clearly seems to be a “statement of general policy ... subject to judicial review under the Veterans Judicial Review Act.” 137 Cong.Rec. S1271 (daily ed. January 30, 1991). Senator DeConcini, the floor manager, agreed “that such a determination would be reviewable.” Id. Accord 137 Cong.Rec. H731 (daily ed. Jan. 29, 1991) (statement of Rep. Evans). Moreover, in the 1984 Act, which first required the Secretary to determine whether a presumption was warranted, one of the bill’s primary sponsors stated: The Administrator’s compliance with the required process, as well as any regulations issued by the Administrator would be subject to judicial review. I cannot overstate the importance of judicial review ... It is only just that ... Vietnam veterans have their day in court. 130 Cong.Rec. 13,155 (remarks of Senator Specter). There is a well-settled presumption that agency actions are reviewable. See McNary v. Haitian Refugee Ctr., Inc., 498 U.S. 479, 496, 111 S.Ct. 888, 898, 112 L.Ed.2d 1005 (1991); Wheeler v. United States, 11 F.3d 156, 159 (Fed.Cir.1993). Nothing the Secretary relies on refutes that presumption. The Secretary’s determination is reviewable. III. A. The 1991 Act created an unusual administrative scheme by which the Secretary is to determine whether there is a service connection between particular diseases and herbicide exposure in Vietnam. Unlike the typical responsibility assigned to a government agency, the Secretary was directed to enter into an agreement with the Academy, a private non-government scientific agency, for the latter to summarize and assess the scientific evidence dealing with the issue and report its findings to him. The major work involved thus was assigned to a non-government entity; the Academy was given 18 months to submit its report, but the Secretary was to render his decision within 60 days of receiving the report. Following the receipt of the Academy’s report, and obviously giving it great weight, the Secretary was to determine, for each of the diseases involved, whether there was a *1199 “positive association” between exposure of people to a herbicide agent and the existence of the disease. The standard for determining whether a positive association exists is whether the “credible evidence” for the association equals or outweighs the credible evidence against it. If the Secretary finds a credible association for a particular disease, he must establish a presumption of service connection for it. The statutory scheme contemplates that the Academy’s findings on these medical and scientific issues reflected in the Report would be the key element in the Secretary’s decision. Congress necessarily intended that the Secretary, although not bound by the Academy’s findings, would place great reliance on them. Under the Administrative Procedure Act, we review the Secretary’s decision to determine whether it was “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law....” 5 U.S.C. § 706(2)(A). This is a “highly deferential” standard of review. Ethyl Corp. v. EPA, 541 F.2d 1, 34 (D.C.Cir.), cert. denied sub. nom., 426 U.S. 941, 96 S.Ct. 2662, 2663, 49 L.Ed.2d 394 (1976). Under the unusual statutory scheme here involved, with the function of reviewing and evaluating the scientific evidence given to a non-governmental, independent scientific entity, an extremely strong showing of error would be required before we properly could reverse the Secretary’s determination. B. The petitioners make a sweeping attack on the evidentiary basis for the Secretary’s findings and determinations. Although the argument is framed mainly in terms of the Secretary’s alleged failure to consider the evidence in the Academy Report that indicated a positive association between the three cancers involved and herbicide exposure, it really asserts that the Secretary erred because he did not give sufficient weight to that evidence. In essence the petitioners urge us to reweigh the evidence and come to a different conclusion than the Secretary reached. Our role, however, “is not to substitute our judgment for that of the” Secretary. Motor Vehicle Mfrs. Ass’n. v. State Farm Mutual Auto Ins. Co., 463 U.S. 29, 43, 103 S.Ct. 2856, 2866-67, 77 L.Ed.2d 443 (1983). Instead, we consider whether the Secretary has “examine[d] the relevant facts and articulate[d] an adequate explanation for [his] action including a rational connection between the facts found and the choice made-[and] whether the decision was based on a consideration of the relevant factors and whether there has been a clear error of judgment.” Id. As we show in Part IV, the Secretary’s decision refusing to establish a presumption of service connection for the three cancers must stand. The Secretary did not uncritically accept or merely rubber-stamp the Academy’s factual findings. Instead, he carefully reviewed the evidentiary bases for them, recognizing the strengths and weaknesses of the underlying scientific studies that the Academy discussed. As the Academy had urged, he did not view the report’s categorization of the diseases according to the strength of statistical association “as recommendations regarding DVA policy.” Instead of merely applying the presumption to whole categories, the DVA task force, and then the Secretary, examined each disease to determine if it met the statutory standard. “The task force spent a great deal of time focusing on” the diseases in the Academy’s second category, Limited/Suggestive Evidence of an Association, and determined that two of those diseases, respiratory cancers and multiple mye-loma met the standard, whereas prostate cancer did not. The Secretary followed this recommendation, thus demonstrating independent judgment by recognizing that the Academy’s factual conclusions did not equate to the ultimate decision entrusted to him. The petitioners rely to a considerable extent upon two declarations by experts in biostatistics and toxicology (Drs. Levin and Legator), which they filed with their opening brief. Those declarations, however, were not part of the record on which the Secretary based his decision, and are not part of the record before this court. Indeed, this court denied the petitioners’ motion to supplement the administrative record with those declarations. We therefore do not consider them. *1200 C. A major ground of the petitioners’ challenge to the Secretary’s evaluation of the evidence is that the Secretary improperly failed to recognize that the number of scientific studies showing a positive association between herbicide exposure and each of the three cancers was greater than the number showing no association or a negative association. According to the petitioners, the numerical superiority of this evidence required the Secretary to treat it as credible evidence of an association. They point out that the statutory evidentiary standard the Secretary is to apply in making his findings on whether a positive association exists is whether “credible evidence” for an association equals or outweighs the credible evidence against it. The statute does not define “credible evidence.” In evaluating evidence, however, the critical question is its quality, not its quantity. Just as “numerical superiority of witnesses alone could not necessarily entitle the plaintiff to relief,” Banks Construction Company v. United States, 364 F.2d 367, 362, 176 Ct.Cl. 1302 (1966), so numerical superiority of scientific studies does not require the trier of fact to accept them. The Department task force that reviewed the Academy Report anticipated and explained the reasons for rejecting the numerically-superior documents: it is likely that there will be controversy surrounding any conclusion about which evidence is credible and which evidence is less so or not credible. This is clearly demonstrated by the statements of Mark Venuti of the National Veterans Legal Services Project which have been endorsed by the Vietnam Veterans of America. Many readers of the NAS report will consider all positive evidence equally “credible” despite the confidence limits of the studies or the statistical significance of the findings. There will be a tendency to look at the charts prepared by the NAS and count up those studies which could be interpreted as positive and balancing them against cited studies showing no association or a negative association. This was not the purpose of those charts and the NAS staff and representatives of the Committee responsible for the report findings refused to interpret their findings that way either in the report or in face-to-face meetings with VA staff or at Congressional hearings. The Secretary explained why he rejected the petitioners’ approach: Simply comparing the number of studies which report a negative relative risk for a particular condition is not a valid method for determining whether the weight of evidence overall supports a finding that there is or is not a positive association between herbicide exposure and the subsequent development of the particular condition. Because of difference in statistical significance, confidence levels, control for confounding factors, etc., some studies are clearly more credible than others, and the Secretary has given them more weight in evaluating the overall credibility of the evidence concerning specific diseases. 59 F.R. 342. There is no basis for assuming that, by using the term “credible evidence” in the 1991 Act, Congress intended to preclude the Secretary from applying the traditional principle that the critical issue in evaluating evidence is its quality, not its quantity. The Secretary cannot be faulted for refusing to follow the more numerous studies that support the petitioners’ contentions. D. The petitioners challenge the Secretary’s refusal to establish a presumption of service connection for liver and nose cancer because he did not discuss the biologic plausibility of an association between those diseases and herbicide exposure. The 1991 Act provides that “[f]or each disease reviewed, the Academy shall determine ... whether there exists a plausible biological mechanism or the evidence of a causal relationship between herbicide exposure and the disease.” 38 U.S.C. § 1116, note d. Although the Act did not require the Secretary to address this factor, it required him to consider the Academy report, which he admittedly did. The petitioners contend that by implication the Act required him also to discuss biologic plausibility. The fact that the 1991 Act expressly required the Academy to consider biologic *1201 plausibility but did not impose a similar requirement on the Secretary is a strong indication that the latter was not required to do so. Russello v. United States, 464 U.S. 16, 23, 104 S.Ct. 296, 300-01, 78 L.Ed.2d 17 (1983) (“ ‘[W]here Congress includes particular language in one section of a statute but omits it in another section of the same Act, it is generally presumed that Congress acts intentionally and purposely in the disparate inclusion or exclusion’” (citation omitted)). This is particularly so because the Act specified in detail exactly what were the precise duties of the Academy and the Secretary, and provided specific standards for each of them to follow. If Congress had intended the Secretary specifically to consider biologic plausibility, it is a reasonable assumption that it would have told him to do so — as it did for the Academy. The Academy Report discussed biologic plausibility, but concluded that “there is as yet no convincing evidence of, or mechanistic basis for, the carcinogenicity of any of the herbicides used in Vietnam.” The Report also addressed the carcinogenicity of dioxin, a contaminant “found in varying levels in different batches of Agents Orange, Pink, Purple, and Green,” and reported extensive evidence that dioxin caused cancer in laboratory animals. Although the carcinogenicity of an herbicide contaminant could be viewed as credible evidence of a positive association, the Secretary could reasonably conclude otherwise. This is especially so because Congress directed that the Academy was specifically to address “whether there exists a plausible biological mechanism or other evidence of a causal relationship between herbicide exposure and the disease,” and the presence of dioxin in an herbicide does not compel the conclusion that there was such a relationship caused by exposure to the herbicide. IV. We turn to the Secretary’s refusal to establish a presumption of service connection for the three cancers involved. 1. The Secretary stated the following with regard to prostate cancer: The NAS report assigns prostate cancer to a category labeled limited/suggestive evidence of an association, which it defined as meaning there is evidence suggestive of an association between herbicide exposure and a particular health outcome, but that evidence is limited because chance, bias, and confounding could not be ruled out with confidence. Prostate cancer is a very common male genitourinary cancer which shows marked increased prevalence with age. There are statistically significant occupational studies which show no association between prostate cancer and herbicide exposure (e.g., Fingerhut ...; Manz ...; Saracci_). Some occupational studies have shown a slight elevated risk for prostate cancer among farm and forestry workers (e.g. Burmeister ...; Alavanja ...); however, only one study concerning a small sub-set of farmers (Morrison ...) associated the increased risk of prostate cancer among farmers specifically with herbicide exposure. The Morrison study is so recent that it is too early to determine whether its results will be replicated by other research. Accordingly, the Secretary has found that the credible evidence against an association between prostate cancer and herbicide exposure outweighs the credible evidence for such an association, and he has determined that a positive association does not exist. 59 F.R. 242 (1994). The Secretary’s decision accurately analyzes and summarizes the Academy’s findings and is reasonable in light of that analysis. In placing prostate cancer in the “limited/suggestive evidence of an association” category, the Academy pointed out: Most of the agricultural studies indicate some elevation in risk of prostate cancer. One large well-done study in farmers showed an increased risk, and subanalyses in this study indicate that the increased risk is specifically associated with herbicide exposure. The three major production worker studies ... all show a small, but not státistically significant elevation in risk.... It should be noted, however, that most of the associations are relatively weak (< 1.5). *1202 The petitioners point out, and the Secretary concedes, that he erred in citing the Fingerhut, Manz, and Saracci studies as supporting the statement that “[t]here are statistically significant occupational studies which show no association between prostate cancer and herbicide exposure.” Instead, the three studies actually show a positive association, but are not statistically significant. The quoted portion of the statement, however, is not inaccurate. It would have been accurately supported if the Secretary had cited to the Ronco and Wicklund studies, which were statistically significant and showed a negative association and none at all, respectively. We will “not reverse simply because there are uncertainties, analytical imperfections, or even mistakes in the pieces of the picture petitioners have chosen to bring to [our] attention ... but only when there is such an absence of overall rational support as to warrant the description arbitrary and capricious.” Center for Auto Safety v. Peck, 751 F.2d 1336, 1370 (D.C.Cir.1985) (citation omitted). In this ease, the Secretary’s miscitation of the three studies does not so undermine the Secretary’s rationale as to render it arbitrary and capricious. The petitioners also contend that the Secretary erred in characterizing the Morrison study as a small subset. The Academy Report shows that the Secretary’s statement is accurate. His statement referred to a subset of the study that covered only 20 cases of exposure to herbicides. The Secretary reasonably could have concluded that because the entire study covered 1,138 cases of exposure, this was but a small segment of the study. The petitioners further object to the Secretary’s statement that it is “too early to determine whether [Morrison’s] results will be replicated by other research.” They contend that this statement indicates that the Secretary misconstrued his statutory mandate to consider whether the studies are capable of replication. The Secretary’s statement merely reflects the Morrison study’s warning that the results “should be considered tentative because of the relatively low increases in risk and because an association has not been noted previously.” The Secretary reasonably recognized that future research on the issue might undermine or contradict the conclusions of the Morrison study. Indeed, a further Academy report on prostate cancer was due in July 1995. The petitioners make various arguments concerning the relative merits of the studies the Academy considered. As noted, however, it is not our function to reweigh the evidence. The Secretary’s decision is reasonable based on the Academy’s factual determinations, particularly its finding that almost all studies, certainly all statistically significant ones, show a weak positive association at best. B. The Secretary stated with regard to hepatobiliary (liver) cancers The NAS report assigns ... hepatobili-ary cancers ... to a category labeled inadequate/insufficient evidence to determine whether an association exists, which is defined as meaning that the available studies are insufficient quality, consistency, or statistical power to permit a conclusion regarding the presence or absence of an association with herbicide exposure. Hepatobiliary cancers are cancers of the liver and bile duct. There are a variety of risk factors that should be considered by a credible study, including hepatitis B and C, alcohol abuse, cirrhosis, exposure to ... PCB, and smoking. The relevant studies are few and have not adequately controlled for these risk factors. A Swedish case control study (Hardell ... 1984 ...) showed a relationship between herbicide exposure and the subsequent development of hepatobiliary cancer; however, other studies of similar size (Ronco ... 1992 ...; Wiklund ... 1983 ...) indicated no relationship. A large occupational study (Fingerhut et al, 1991) and a study of farmers in Denmark and Italy (Ronco 1992) found no relationship. Accordingly, the Secretary has found that the credible evidence against an association between hepatobiliary cancer and herbicide exposure outweighs the credible evidence for *1203 such an association, and he has determined that a positive association does not exist. 59 F.R. 343 (1994). The Academy Report summarized the various studies as follows: There are relatively few occupational, environmental, or veterans studies of liver cancer (Table 8-6), and most of these are small in size and have not controlled for life-style-related risk factors. One of the largest studies (Hardell et al., 1984) indicates an increased risk for liver cancer and exposure to herbicides, but another study of Swedish agricultural workers (Wiklund, 1983) estimates a relative risk that is significantly less than 1.0 [i.e. a protective association]. The estimated relative risks from other studies are both positive and negative. As a whole, given the methodological difficulties associated with most of the few existing studies, the evidence regarding liver cancer is not convincing with regard to either an association with herbicides/TCDD or the lack of an association. The petitioners contend that the Secretary erred in stating that there were few studies. This statement, however, reflects the Academy’s statement based on the 18 studies that it examined. The petitioners cite no other studies that they contend the Secretary also should have considered. According to the petitioners, the Secretary erred in failing to consider that nine studies are in favor of association, that eight are against it and that one finds no association. Instead, the Secretary relied on the studies with comparably sized samples, all of which were among the largest. Of the four he discussed, three do not show a positive association. Moreover, none of the 18 studies are statistically significant. Given this evidence, the Secretary reasonably concluded that the credible evidence against an association outweighed the credible evidence for one. C. The Secretary stated with regard to nasal/nasopharyngeal cancer: The NAS report assigns ... nasal/naso-pharyngeal cancer to a category labeled inadequate/insufficient evidence to determine whether an association exists, which is defined as meaning that the available studies are insufficient quality, consistency, or statistical power to permit a conclusion regarding the presence or absence of an association with herbicide exposure. NAS noted an association between nasal cancers and occupational exposure to nickel and to chromates. Exposure to wood dust is also a risk factor for nasal cancers; smoking and exposure to formaldehyde may increase the risk associated with wood dust. There is also evidence that leather workers have an increased risk for nasal cancers and that there is an association between chronic nasal diseases and consumption of salt preserved foods. Most studies (e.g. Wiklund, 1983; Roneo et al. 1992) showed inconclusive results, and often did not control for confounding variables. Two other epidemiological studies based on the same three cases (Saraeci et al., 1991. Coggon D., Pannett B., Winter P.D., Achedson E.D., Bonsall J., 1986. Mortality of workers exposed to 2 methyl-4-chlorophenoxyacetic acid. Scandinavian Journal of Work, Environment, and Health 12:448-454) and one case-control study (Hardell L., Johansson B. Axelson 0., 1982. Epidemiological study of nasal and nasopharyngeal cancer and their relation to phenoxy acid or chlorophenol exposure. American Journal of Industrial Medicine 3 247-257) showed increased risk associated with herbicide exposure. However, that risk was not statistically significant, which diminishes the importance of these studies. Accordingly, the Secretary has found that the credible evidence against an association between nasal/nasopharyngeal cancers and herbicide exposure outweighs the credible evidence for such an association, and he has determined that a positive association does not exist. 59 F.R. 346 (1994). The Secretary’s discussion again echoed the Academy’s Report, which discussed numerous factors and exposures to things other than herbicides that have been associated with an increased risk of nasal cancer. It summarized the epidemiological results much as the Secretary did, noting that the Saracci *1204 and Coggon studies were based on the same three cases and that the Hardell study “found [a positive association] for those exposed to phenoxy acids, based on eight exposed cases.” It noted that a Center for Disease Control study of Vietnam veterans found no “significant associations for Vietnam service,” and that “[o]ther studies show[ed] inconclusive results.” It summarized the epidemiologic evidence as “inadequate or insufficient ... to determine whether an association exists between exposure to herbicides and nasal/nasopharyngeal cancer.” Aside from the biologic plausibility and numerical arguments previously discussed, the petitioners contend that the Secretary made two key errors with regard to nasal cancers. They contend that the Secretary (as did the Academy) mistakenly stated that the Sarraci and Coggon studies were based on the same three eases. Even if this statement was wrong, however, it did not render the Secretary’s decision arbitrary or capricious. Neither he nor the Academy attached any significance to the supposedly-shared samples, but only noted the fact. It does not appear to have influenced the Secretary’s conclusion. Instead, he focused on the inconclusive nature of the studies discussed, citing their lack of significance. The petitioners also contend that the Secretary’s decision was arbitrary and capricious because it did not discuss a subset of the Hardell study that showed a positive association between chlorophenol exposure and nasal cancer. The Secretary’s decision discusses the Hardell study in terms of what it shows regarding “the risk associated with herbicide exposure.” This shows that he was aware of the Hardell study, but considered only the part dealing with herbicide exposure. Other than the inclusion of the subset in a table in the Academy’s report, the petitioners offer no convincing reason why the chlorophenol subset was a relevant factor that the Secretary should have deemed credible evidence of a relationship between herbicide exposure and nasal cancer. They have thus failed to demonstrate that this was relevant evidence, and thus the Secretary’s alleged failure to consider it would not be arbitrary and capricious. Citizens to Preserve Overton Park v. Volpe, 401 U.S. 402, 416, 91 S.Ct. 814, 823-24, 28 L.Ed.2d 136 (1971). AFFIRMED.