731 F.2d 1253

UNITED STATES of America, Plaintiff-Appellee,
v.
AN ARTICLE OF DEVICE ... "Toftness Radiation Detector,"
Toftness Post-Graduate School of Chiropractic,
Inc., a corporation, and Irwing N.
Toftness, an individual,
Defendants-Appellants.

No. 83-1404.

United States Court of Appeals,
Seventh Circuit.

Argued Dec. 7, 1983.
Decided April 4, 1984.

E. Campion Kersten, Kersten & McKinnon, Milwaukee, Wis., for defendants-appellants.

Don O. Burley, U.S. Dept. of Justice, Washington, D.C., for plaintiff-appellee.

Before CUDAHY, ESCHBACH and COFFEY, Circuit Judges.

CUDAHY, Circuit Judge.

1

Under the Food, Drug, and Cosmetic Act, 21 U.S.C. Secs. 301 et seq. (the "Act"), a prescription medical device is "misbranded" if it cannot be used safely and effectively for its intended purposes. The Toftness Radiation Detector ("TRD") is a chiropractic instrument which purportedly detects electromagnetic radiation from the human body and focuses that radiation so that a chiropractor can detect conditions which require chiropractic treatment. The government sued to condemn the TRD as "misbranded" under the Act and to enjoin the TRD's inventor from producing, distributing or using the TRD. After a trial in the district court, a jury rendered a verdict for the government, and this appeal followed.

2

* Defendant-appellant Irwing N. Toftness is a licensed chiropractor who practices in Cumberland, Wisconsin, and who invented the TRD. The TRD, which is patented, consists of a plastic cylinder containing a series of plastic lenses. According to Toftness, low levels of electromagnetic radiation emanate from the human body, and that radiation emanates most strongly from areas of neurological disturbance. The TRD is supposed to be capable of detecting and focusing this radiation through the plastic lenses. When a trained user holds the TRD close to a patient's skin, the user is supposed to detect this radiation by rubbing his or her fingers on the detection plate and feeling resistance to the movement of the fingers. After locating the points of disturbance, the chiropractor can then make adjustments to the body to alleviate these neurological disturbances.

3

Toftness is also the president of defendant-appellant Toftness Post-Graduate School of Chiropractic, Inc. Only licensed chiropractors who have completed a training course at the Toftness School may use the TRD. The course lasts several weeks and costs $400. At the time of the trial in this case, approximately 700 chiropractors had attended the TRD course and had signed leases for the TRD. The fifteen year leases of the instrument call for payment of $700 for the first year and $100 for each of the next fourteen years. The leases also provide that the use of the instrument should be limited to research purposes and that the user should keep careful research records and forward those records to the school.

4

The government brought this action under the Food, Drug, and Cosmetic Act of 1938, 21 U.S.C. Secs. 301 et seq., contending that the TRD is a "misbranded" device under 21 U.S.C. Sec. 334(a)(1). We shall explain the statutory and regulatory framework in more detail below, but the heart of the government's case is its contention that the TRD simply does not work and is therefore "misbranded." Government witnesses testified at trial that the TRD is incapable of detecting any radiation coming from the human body. Defense witnesses testified in turn about their experimental use of the TRD and their purported success in treating patients while using the TRD. The jury rendered a verdict for the government, and the district court denied the defendant's motion for a new trial.1

5

This appeal followed, and appellants here argue that the district court erred with respect to three issues. First, appellants contend that the district court improperly instructed the jury that the TRD was, as a matter of law, a "device" as defined in 21 U.S.C. Sec. 321(h) and thus subject to the provisions of the Act. Second, appellants argue that the district court erroneously instructed the jury that the burden of proof was on the defendants to show that the TRD was properly labeled as a "prescription device" under 21 C.F.R. Sec. 801.109 (1983). Third, they argue that the court erred by instructing the jury not to "pile inference on inference." For the following reasons, we conclude that the district court did not err in these three matters and we affirm its judgment.

II

6

The first issue on appeal is whether the TRD is a "device" as defined in section 201(h) of the Act, 21 U.S.C. Sec. 321(h), and thus subject to the Act's misbranding provisions. The district court instructed the jury that the TRD is a "device," and appellants contend that the district court erred in directing a verdict on this point. According to appellants, the TRD is not subject to the misbranding provisions because its intended uses are limited to research purposes. In our view the district court properly directed a verdict on this issue because the appellants' attempted rebuttal was based on a misreading of the Act. The Act's definition of "device" includes instruments used for research so long as the intended uses of the instrument in question include the diagnosis and treatment of diseases or other conditions.

7

To determine whether the TRD is a "device" under the Act, we begin with section 304(a)(1) of the Act, 21 U.S.C. Sec. 334(a)(1), which permits the government to seize any misbranded "article of ... device" in interstate commerce. Section 201(h) of the Act, 21 U.S.C. Sec. 321(h), defines "device" to include:

8

instruments, apparatus, and contrivances, including their components, parts, and accessories, intended (1) for use in the diagnosis, cure, mitigation, treatment, or prevention of disease in man or other animals; or (2) to affect the structure or any function of the body of man or other animals.

9

The FDA has issued regulations which further elaborate on this definition by defining "intended uses." The regulations provide in relevant part:

10

The words "intended uses" or words of similar import in Secs. 801.5, 801.119, and 801.122 refer to the objective intent of the persons legally responsible for the labeling of devices. The intent is determined by such persons' expressions or may be shown by the circumstances surrounding the distribution of the article. This objective intent may, for example, be shown by labeling claims, advertising matter, or oral or written statements by such persons or their representatives. It may be shown by the circumstances that the article is, with the knowledge of such persons or their representatives, offered and used for a purpose for which it is neither labeled nor advertised.

11

21 C.F.R. Sec. 801.4 (1983). The dispute here concerns the language: "use in the diagnosis, cure, mitigation, treatment, or prevention of disease in man." 21 U.S.C. Sec. 321(h)(1). Defendants appear to offer two distinct but closely related theories supporting their contention that the TRD does not fall within that definition of "device." First, they introduced evidence showing that the TRD was not intended to be used as the sole means of diagnosing patients or of evaluating the success of their treatments. Second, they introduced evidence tending to show that the TRD was intended only for research use instead of for diagnosis or treatment.

12

In response, the government refers us to evidence showing that the TRD was intended for use in the diagnosis and treatment of patients by chiropractors. That evidence includes the instructions that Dr. Toftness prepared for use of the TRD, the financial arrangements through which chiropractors were trained in the use of the TRD and permitted to use it and the testimony of the appellants' witnesses showing that they had made chiropractic adjustments to patients on the basis of TRD readings. The instructions for the use of the TRD contain, in various versions, directions for sensing points of maximum radiation, for interpreting those radiation readings and for making appropriate chiropractic adjustments. The financial arrangements include the tuition for instruction in the proper use of the TRD and the fifteen year leases for the instrument itself. Further, several of appellants' witnesses testified that they had actually made chiropractic adjustments to patients based upon readings from the TRD and that they had measured the effectiveness of treatment through use of the TRD. That evidence supported appellants' contention that the TRD is "effective," see infra Part III, but it also supported the government's argument that the TRD is a "device."

13

On the other hand, the appellants point out that the instructions stated that the TRD was to be used only in conjunction with standard chiropractic techniques such as "Chiropractic Palpation" or "Line of Drive." The instructions also directed the user to keep careful records of all adjustments on "Daily Research Cards." There was, in addition, evidence showing that TRD users have not advertised that they use the TRD, that they have not charged patients more money for use of the TRD in diagnosis or treatment and that the users have informed patients that they were testing this new instrument. Appellants, based on this evidence, argue that the intended use of the TRD until now has been solely for research and that users have simply tested the device by using it and comparing their findings from its use against findings made using established techniques.

14

To avoid the directed verdict, appellants needed either to cast doubt on the government's evidence of intended use or to introduce other evidence rebutting the government's evidence. However, the appellants stipulated to much of the evidence supporting the government's position, and they introduced some of it themselves.2 It would have been impossible for any reasonable jury to have rejected this evidence. Therefore, to avoid the directed verdict, appellants rely on the evidence tending to show that the TRD was used only in conjunction with established chiropractic techniques and that the TRD was being used only for research. On appeal we must view that evidence in the light most favorable to the appellants, and if the evidence would have permitted a reasonable jury to find in appellants' favor, then the directed verdict on the issue was improper. Hohmann v. Packard Instrument Co., 471 F.2d 815, 819 (7th Cir.1973). See Brady v. Southern Railway Co., 320 U.S. 476, 479-80, 64 S. Ct. 232, 234-35, 88 L. Ed. 239 (1943) (standard for directed verdict).

15

However, even in the light most favorable to appellants, their evidence does not rebut the showing that the TRD was a "device," for their attempted rebuttal is based on an incorrect reading of the Act. First, the fact that the TRD readings were not the sole basis for diagnosis or treatment does not mean that the TRD was not intended for use in the diagnosis and treatment of disease. Even if used in conjunction with other techniques, the TRD was still intended to be a basis for diagnosis and treatment. Appellants' interpretation of the term "device" would exclude from the definition instruments used in connection with other procedures or agents. This interpretation would leave precious few medical instruments within the ambit of the Act, for certainly few instruments are used alone in diagnosis or treatment. An instrument "need not be the only agent in an allegedly curative process to be a device within the definition." United States v. Article of Device ... "Hubbard Electrometer," 333 F. Supp. 357, 360 (D.D.C.1971).

16

Second, evidence showing that the TRD was used solely for research does not rebut the evidence that the device was intended for use in the treatment and diagnosis of disease. The research in the instant case necessarily involved the use of the TRD for diagnosis and treatment, for it was the effectiveness of the instrument for precisely those uses which was being tested. The research simply could not be carried out without using the TRD in the diagnosis and treatment of disease. The appellants' own evidence regarding effectiveness of the TRD showed that they intended it for use, and in fact used it, in diagnosis and treatment. We cannot disentangle the "research" from the research methods used. Under appellants' theory, a medical device in the research stage of development could be completely exempt from the Act's regulatory provisions even when the device was being used in the clinical diagnosis and treatment of patients for research purposes. But the Act and its regulations do not except instruments involved in research from the definition of "device," for those instruments may also pose a threat to public health in the research stage. Instead, special, less restrictive labeling requirements apply to investigational devices. See 21 U.S.C. Sec. 360j(g) (exemption for devices for investigational use); 21 C.F.R. Secs. 812.1--.150 (1983) (regulations for investigational use of devices); 21 C.F.R. Sec. 801.122 (1983) (exemption for research "not involving clinical use"); H.R.Conf.R. 1090, 94th Cong., 2d Sess. 64, reprinted in 1976 U.S.Code Cong. & Ad.News 1070, 1116-17.

17

We recognize that the appellants were obliged to come forward with evidence of the TRD's effectiveness, and such proof ordinarily involves research into the effectiveness of the device in its intended clinical uses. However, appellants were not placed in a Catch-22. It is possible first to investigate an instrument in a way which will not be subject to the full panoply of labeling requirements for devices on the open market, and then to use the results of the investigation to meet the law's requirements for devices on the market. See 21 C.F.R. Secs. 801.122 and 812.1--.150 (1983). Appellants have not argued that the TRD falls within an exception for investigational devices.

18

Because the government introduced substantial and convincing evidence showing that the TRD was intended for use in the diagnosis and treatment of disease, and because the appellants' evidence did not, under the structure of the Act, rebut that showing, the district court correctly instructed the jury that the TRD was a "device" within the meaning of 21 U.S.C. Sec. 321(h). A reasonable jury could not have found otherwise.

III

19

We next address the burden of proof on the misbranding of prescription devices. To reach this issue, we must follow an elaborate trail through both the United States Code and the Code of Federal Regulations. We begin our trek with section 304(a)(1) of the Food, Drug, and Cosmetic Act of 1938, 21 U.S.C. Sec. 334(a)(1), which provides for the seizure and condemnation of articles of drug or device which are "adulterated or misbranded" in interstate commerce. The parties stipulated that the TRD has moved in interstate commerce, and we have shown in Part II that the TRD is a "device." To learn whether the TRD is "misbranded," we turn to section 502 of the Act which provides in relevant part:

20

A drug or device shall be deemed to be misbranded--

21

* * *

22

* * *

23

(f) Unless its labeling bears (1) adequate directions for use; and (2) such adequate warnings against use in those pathological conditions or by children where its use may be dangerous to health, or against unsafe dosage or methods or duration of administration or application, in such manner and form, as are necessary for the protection of user: Provided, that where any requirement of clause (1) of this subsection, as applied to any drug or device, is not necessary for the protection of the public health, the Secretary shall promulgate regulations exempting such drug or device from such requirement.

24

21 U.S.C. Sec. 352(f). In this case the government has argued that the TRD did not bear "adequate directions for use" as required in section 502(f)(1). For a definition of "adequate directions for use," we leave the United States Code behind and journey into the Code of Federal Regulations. In the section dealing with medical devices, we read that: " 'Adequate directions for use' means directions under which the layman can use a device safely and for the purposes for which it is intended." 21 C.F.R. Sec. 801.5 (1983). Obviously there are many medical devices which would be ineffective at best, and dangerous at worst, if left in the hands of a layman, and section 801.5 appears to deem any such devices "misbranded" and thus subject to seizure. However, the regulations provide several exemptions from the "adequate directions for use" requirement. See 21 C.F.R. Secs. 801.109--.127 (1983). The broadest of these exemptions, and the one at issue in this case, is the exemption for "prescription devices," that is, those devices which require the supervision of a licensed practitioner for their safe and effective use. 21 C.F.R. Sec. 801.109 (1983).

25

In this case the parties have stipulated that the TRD cannot be used by laymen; therefore, it would be impossible for the TRD to be labeled with "adequate directions for use" as defined in 21 C.F.R. Sec. 801.5. Under the statute and regulations, the device is thus misbranded unless it falls within one of the exemptions from the requirement, and the appellants in this case have always contended that the TRD falls within the "prescription device" exemption in 21 C.F.R. Sec. 801.109.

26

In examining section 801.109 of the regulations, we note that the device must meet each of the conditions set forth in the section. The issue here centers upon subsection (c) of the regulation, which requires that:

27

(c) Labeling on or within the package from which the device is to be dispensed bears information for use, including indications, effects, routes, methods, and frequency and duration of administration, and any relevant hazards, contraindications, side effects, and precautions under which practitioners licensed by law to administer the device can use the device safely and for the purpose for which it is intended, including all purposes for which it is advertised or represented: Provided, however, that such information may be omitted from the dispensing package if, but only if, the article is a device for which directions, hazards, warnings, and other information are commonly known to practitioners licensed by law to use the device. Upon written request, stating reasonable grounds therefor, the Commissioner will offer an opinion on a proposal to omit such information from the dispensing package under this proviso.

28

21 C.F.R. Sec. 801.109(c) (1983) (emphasis supplied). The consequence of the italicized language is that a prescription device is misbranded unless it can be used safely and effectively for the purposes for which it is intended. That is, the device has to work--if it does not work, it is misbranded. And the main issue at trial here was whether the TRD in fact works. The government introduced evidence to prove it does not, and the appellants introduced evidence to show it does. The district court instructed the jury that appellants had the burden of proving that the TRD works, and the main issue on appeal is whether this instruction was proper.

29

The appellants contend that the government should have had the burden of proving that the prescription device exemption did not apply because the government generally has the burden of proving misbranding. See United States v. Four Cases ... Slim-Mint Chewing Gum, 300 F.2d 144, 148 (7th Cir.1962). The government responds with the general principle that a party claiming entitlement to a statutory exemption bears the burden of proving the entitlement. United States v. First City National Bank, 386 U.S. 361, 366, 87 S. Ct. 1088, 1092, 18 L. Ed. 2d 151 (1967); Federal Trade Commission v. Morton Salt Co., 334 U.S. 37, 44-45, 68 S. Ct. 822, 827-28, 92 L. Ed. 1196 (1948).3 Apart from these two general principles and the structure of the statute and regulations, there is very little to guide our consideration of this issue. The statute and regulations are silent, and there appear to be no indications in the history of the legislation or the regulations which could assist us. The parties have cited a number of cases to us, but only three address this issue even tangentially, and they are inconclusive.4

30

Our only significant guides, therefore, are the structure of the statute and regulations, and the two general rules already noted. We agree with appellants that the government generally bears the burden of proving misbranding. But as the statute and regulations are structured, the government met its burden by proving that the TRD (1) was a "device" under the Act; (2) moved in interstate commerce; and (3) did not bear directions for use adequate to permit a layman to use it safely and effectively. The fact that the government got a directed verdict on the first issue and that the parties stipulated to the second and third does not affect the burden of proof. Under the logic of the statute and regulations, we are persuaded that the prescription device provisions of 21 C.F.R. Sec. 801.109 are framed as an exemption from the more general labeling requirements.

31

In accordance with our analysis, a device is "misbranded" unless it bears "adequate directions for use." 21 U.S.C. Sec. 352(f)(1). The regulations define "adequate directions for use" as directions appropriate for a layman. 21 C.F.R. Sec. 801.5. Then, Subpart D of the labeling regulations establishes various exemptions from the general labeling requirements, including the prescription device exemption. 21 C.F.R. Sec. 801.109. The FDA has framed the prescription device labeling requirement as an exemption, and appellants claim they qualify under it. Thus, appellants would appear to bear the burden of proof under the reasoning of United States v. First City National Bank, supra, 386 U.S. at 366, 87 S. Ct. at 1092, unless placing the burden of proof on them would be contrary to the Act.

32

Appellants tell us that it would be absurd to treat the prescription device exemption as a true exemption, for it makes no sense to suggest that sophisticated medical devices--which cannot be used by laymen--are presumptively misbranded merely because they do not include directions for laymen. Certainly the FDA could have promulgated regulations which would have established two categories of devices--prescription and over-the-counter--with two separate sets of labeling requirements. In proceeding against a device under such regulations, the FDA would presumably choose whether to proceed under either the prescription or the over-the-counter provisions of the regulations. Under such a hypothetical regulatory scheme, we would have little difficulty in holding that the government bore the burden of proving that the device did not satisfy the prescription device requirements. But that scheme is, of course, hypothetical. Instead, the FDA promulgated the regulations actually before us, and those regulations make prescription devices one of several exemptions to the more general labeling requirements. No purpose justifying this odd structure occurs to us other than the purpose of shifting the burden of proof. By treating the large category of prescription devices under an exemption to the more general requirements, the FDA appears to have wanted to make its task somewhat easier by placing on claimants the burden of proving that their device is safe and is actually effective for its intended purposes.

33

Although this regulatory arrangement may seem strange insofar as it makes prescription devices presumptively misbranded, the device is not contrary to either the letter or intent of the statute. See United States v. Articles of Drug, 625 F.2d 665, 674-75 (5th Cir.1980) (exemption for prescription drugs). The exemption framework requires the makers of prescription devices to be able to prove that their devices do in fact work safely for their intended purposes when they are put on the market.5

34

The government also argues that it was appropriate to cast the burden of proof on the appellants because "the information pertinent to an exemption is often peculiarly within the knowledge of the defendant." Brief for the United States at 22. We agree that in cases such as the one before us, the government may be at something of a disadvantage in its access to proof of the degree of effectiveness of the device. For example, in this case several government experts testified that, in their opinion, the TRD was completely worthless in diagnosis. The appellants attacked that testimony on the basis that the government's experts had never received the special training at the Toftness School which appellants contend is necessary for the proper use of the TRD. Where the government's access to the necessary information may be limited, as it was in this case, it seems not inappropriate to put the burden of persuasion on the party who claims that the device works and who presumably has better access to the relevant information. The appellants must do more than merely show the government is wrong; it is not unfair that they be expected to come forward with affirmative evidence showing that the device is effective and to bear the burden of persuasion.

35

Because the appellants are claiming the application of an exemption to the general labeling requirements, and because they appear to be in the better position to come forward with evidence that the TRD works safely and effectively, we conclude that the district court properly instructed the jury that claimants had the burden of proving that the prescription device exemption applied to the TRD.

IV

36

Finally, appellants contend that the district court erred when it instructed the jury not to "pile an inference on an inference." Appellants say their proof that the TRD worked was necessarily circumstantial, requiring the jury to draw inferences from the evidence presented and then to draw the further reasonable inference that the TRD was effective. According to appellants, the "inference on inference" instruction unduly confined the jury's consideration of the evidence.

37

Jury instructions which tell the jury not to "pyramid" inferences or pile them on top of one another have long been controversial. See 1A J. Wigmore, Evidence Sec. 41 (Tillers rev. 1983). The "inference on inference" instruction cannot be taken literally, for the reasoning process normally begins with known facts which form the basis for inferred facts from which further inferences can be drawn. See id. Sec. 41 at 1111. So long as the finder of fact is reasonably certain of a preliminary inference, it is not unreasonable to use that inference as the basis for further reasoning. See Fenner v. General Motors Corp., 657 F.2d 647, 650-51 (5th Cir.1981), cert. denied, 455 U.S. 942, 102 S. Ct. 1435, 71 L. Ed. 2d 653 (1982); Prudential Insurance Co. v. Glasgow, 208 F.2d 908, 912 (2d Cir.1953). Nevertheless, the process of inferential or circumstantial reasoning can, in some cases, reach far beyond the reasonable scope of the evidence, arriving at conclusions based more upon speculation or conjecture than upon the evidence at trial. See Daniels v. Twin Oaks Nursing Home, 692 F.2d 1321, 1324-26 (11th Cir.1982). For that reason it is entirely appropriate for the trial judge to warn the jury not to get carried away in long, speculative chains of inferences.

38

In our view the "inference on inference" instruction is clumsy and unnecessarily controversial. As Judge Wisdom wrote:

39

The so-called rule against pyramiding inferences, if there really is such a "rule" and if it is anything more than an empty pejorative, is simply legalese fustian to cover a clumsy exclusion of evidence having little or no probative value.

40

N.L.R.B. v. Camco, Inc., 340 F.2d 803, 811 (5th Cir.1965). However, the instruction given in this case, when read in context with the other instructions, appears to have been a rhetorical device aimed more against guesswork and speculation than against the normal process of inferential reasoning.6 The instruction should not have prevented the jury from engaging in the normal process of reasoning from proven facts to an inference and from there to a further reasonable inference.

41

We deal here with matters of common sense. It is always possible for lawyers, judges or logicians to examine what appears to be a reasonable inference and to show how that inference is actually the sum of several shorter inferential steps. See, e.g., 1 Weinstein's Evidence p 401 (1983) (describing inferential steps in several examples). The district court's language, if read strictly and without regard for the context, could be construed to prohibit any steps except those which cannot be broken down any further. Such a reading, however, would ignore common sense. As an example of the kind of reasoning which appellants contend the instruction might have interfered with, appellants point to their evidence showing high statistical correlations between diagnoses based on the TRD and those based on more familiar chiropractic techniques. The appellants asked the jury to infer from those correlations that the TRD, when used properly, produced findings similar to those of proven techniques. The appellants then asked the jury to reach the further inference that the TRD did in fact work. Such a conclusion would not be speculative, even if we can with hindsight break the process down into two or more steps. We do not believe that the jury would have understood the "inference upon inference" instruction to prohibit them from taking any logical step which might be broken down into two or more smaller steps. Such an interpretation would be contrary to common understanding and inconsistent with the remainder of the district court's instructions. We think it would be the better practice to avoid the "inference on inference" language and to concentrate instead on the jury's duty not to engage in speculation that is beyond the scope of the evidence. However, we find no error here in light of the district court's other language on guesswork and speculation.

42

For the foregoing reasons the judgment of the district court is

43

AFFIRMED.

1

The instant action originated in the District of Oregon and was transferred by stipulation of the parties first to the Eastern District of Wisconsin and then to the Western District of Wisconsin. Judge Robert W. Warren of the Eastern District of Wisconsin presided at trial

2

Appellants stipulated that the TRD instructions in the record were accurate. Dr. Toftness described the financial arrangements in his testimony, and defense witnesses described their use of the TRD in research

3

The courts have applied this general principle to actions under the Food, Drug, and Cosmetic Act. See, e.g., Durovic v. Richardson, 479 F.2d 242, 250 n. 6 (7th Cir.) (exemption for drugs generally recognized as safe and effective), cert. denied, 414 U.S. 944, 94 S. Ct. 232, 38 L. Ed. 2d 168 (1973); United States v. An Article of Drug ... "Bentex Ulcerine," 469 F.2d 875, 878 (5th Cir.1972) (burden on claimant to prove drug is entitled to exemption), cert. denied, 412 U.S. 938, 93 S. Ct. 2772, 37 L. Ed. 2d 397 (1973)

4

Appellants rely most heavily on United States v. Evers, 643 F.2d 1043 (5th Cir.1981), but that case does not support their position. In Evers there was no dispute at all over the burden of proof; the facts were undisputed and the case was decided on summary judgment. The case addressed only the application of 21 U.S.C. Sec. 331(k) to the unusual situation in which a physician held a prescription drug for sale only to his patients--not to any other physicians. The court held that the law could not "reasonably be read to require a physician who is holding a drug for sale only to patients to provide adequate information to physicians to whom he is not distributing the drug." 643 F.2d at 1053. The appellants here emphasize the court's statement:

Since Calcium EDTA is a prescription drug, the FDA can establish an act of misbranding under section 502(f)(1) of the Act only by proving that Dr. Evers did not provide adequate information for use by physicians, as is required by the exceptions to that section.

643 F.2d at 1052. The quoted sentence appears to assume that the FDA bears the burden of proof with regard to any exceptions to the misbranding rules. But as we have noted, there was no factual dispute in Evers, and the court did not consider the burden of proof issue apart from that passing comment. Therefore, we do not think Evers is dispositive on this issue.

The government in turn relies most heavily on United States v. Articles of Device Consisting of Three Devices ... "Diapulse," 527 F.2d 1008 (6th Cir.1976) ("Diapulse"). At issue in Diapulse was the application of the prescription device exemption to the labeling requirements. The claimants argued that the device fell within the proviso that labeling need not contain certain information which is "commonly known to practitioners licensed by law to use the device." 21 C.F.R. Sec. 1.106(d)(3) (1976) (predecessor to current 21 C.F.R. Sec. 801.109(c) (1983)). The government had presented affidavits from physicians saying the relevant information was not commonly known, and the claimant presented no evidence on the issue. The court of appeals therefore held that the government was entitled to summary judgment. In a footnote, the court said it believed that the claimant had the burden of proving the elements of the exception for commonly known information and that in the absence of any evidence from the claimant, the government should prevail. 527 F.2d at 1012 n. 6. However, the footnote in Diapulse dealt with an exception within an exception, thus making more clearly applicable the principle that the burden of proof is on the party claiming entitlement to a statutory exception. In addition, the footnote was dictum, for the court decided the case based on the claimant's failure to rebut the government's affidavits.

The Fifth Circuit approved a parallel regulatory structure for prescription drugs in United States v. Articles of Drug, 625 F.2d 665 (5th Cir.1980). There the district court had found the "layman" standard for adequate directions for use to be unreasonable as applied to prescription drugs. The Fifth Circuit reversed and held that the regulatory scheme employing the general "layman" standard in conjunction with exemptions for prescription drugs was a reasonable interpretation of the Act and of Congress' purpose. 625 F.2d at 674-75. The court in Articles of Drug did not expressly consider the burden of proof issue, but it did approve a parallel regulatory structure with a general rule and exemptions which appear to impose on proponents of prescription drugs the burden of proving that an exemption applies.

5

By way of comparison, before a "new drug" goes on the market, the drug's proponent must submit to the FDA, among other information, "full reports of investigations which have been made to show whether or not such drug is safe for use and whether such drug is effective in use...." 21 U.S.C. Sec. 355(b)(1)

6

The relevant portion of the instructions reads as follows:

Before a fact sought to be established can be said to have been proved by circumstantial evidence alone, it is necessary not only that the circumstances proved by the evidence shall give rise to a reasonable inference of such fact, but also that no other inconsistent equally reasonable inference can be drawn from those same circumstances.

A "reasonable inference" is defined as a process of reasoning whereby from the facts otherwise admitted or established by the evidence in light of your common knowledge and experience, a reasonable and logical conclusion may be drawn that a certain fact is true. A reasonable inference is therefore said to be clearly distinguished from a mere guess or conjecture.

The law does not permit speculation; moreover, on the basis solely of one inference so drawn, a further inference may not be drawn. In other words, you can't pile an inference on an inference, but you may draw an inference only from the facts or circumstances which you find have been established by a preponderance of the evidence.

Tr. at 1010-11. The district judge who presided in this case denied a motion for a new trial based in part on a similar instruction in Juneau Square Corp. v. First Wisconsin Nat'l Bank, 475 F. Supp. 451, 460 (E.D.Wis.1979), aff'd, 624 F.2d 798 (7th Cir.), cert. denied, 449 U.S. 1013, 101 S. Ct. 571, 66 L. Ed. 2d 472 (1980). There he explained:

Taken in context, this instruction was not erroneous. It explained to the jury the distinction between reasonable inferences and mere speculation and correctly stated that inferences must be drawn from a factual basis.

475

F.Supp. at 460