Full opinion text
MEMORANDUM OPINION LACEY, District Judge. THE COURT: In these consolidated actions plaintiffs charged the defendants with patent infringement and unfair competition. The patent issues were resolved by a consent judgment entered against the defendants. Plaintiffs’ claims that defendants violated § 43(a) of the Lanham Act, 15 U.S.C. § 1125(a), and the New Jersey law of unfair competition, were tried before the court without a jury. The following constitutes my findings of fact and conclusions of law. Fed.R.Civ.P. 52. THE PARTIES The Plaintiffs The plaintiffs in both actions are Boehringer G.m.b.H., a West German corporation, and its affiliate, Boehringer Ingelheim Ltd. (Boehringer Ltd.), a Delaware corporation, having a principal place of business in Ridgefield, Connecticut. They will be collectively referred to as “Boehringer.” The Defendants In Civil No. 79-358, Pharmadyne Laboratories, Inc. (Pharmadyne), is a New Jersey corporation with its principal place of business at East Brunswick, New Jersey. Bernard Bedrick (Bedrick) at all pertinent times has been president of Pharmadyne. Pharmadyne and Bedrick will be referred to collectively as “Pharmadyne.” In Civil No. 79-938, Premo Pharmaceutical Laboratories, Inc. (Premo), is a New York corporation having its principal place of business in South Hackensack, New Jersey. Seymour Blackman (Blackman) at all pertinent times has been president and chief executive officer of Premo. Premo and Blackman will be referred to collectively as “Premo.” PROCEDURAL HISTORY On January 31, 1979, plaintiffs filed a three-count complaint against defendant Pharmadyne. In the first count plaintiffs alleged Pharmadyne had infringed their patent. Plaintiffs also alleged that Pharmadyne had committed unfair competition under both the Lanham Act and New Jersey state law by deliberately marketing their dipyridamole tablets with the same arbitrary and distinctive trade dress as plaintiffs’ Persantine-brand dipyridamole. Plaintiffs alleged that their distinctive trade dress had acquired secondary meaning; that Pharmadyne was creating confusion of origin by simulating plaintiffs’ trade dress; and that such conduct was facilitating the passing off by pharmacists of Pharmadyne’s products for plaintiffs’. The plaintiffs sought injunctive relief, damages, and costs. On February 23, 1979, a consent judgment was entered against Pharmadyne on Count 1. Under the terms of that judgment, Pharmadyne was enjoined from selling its generic dipyridamole until the patent expired on April 24, 1979. In March 1979, plaintiffs brought suit against defendant Premo. The complaint made the same allegations as against Pharmadyne and requested the same relief. On March 9, 1979, plaintiffs’ application for a temporary restraining order was heard by one of the other judges in the district. He granted the application. Six days later I heard oral argument on the question of whether a preliminary injunction should issue on plaintiffs’ unfair competition claims and denied plaintiffs’ application for a preliminary injunction on Counts 2 and 3. My denial of injunctive relief rested on the inadequate factual foundation laid by plaintiffs. For example, plaintiffs contended that they had demonstrated secondary meaning. I found “that on the basis of the present record secondary meaning has not been shown.” Opinion of March 15,1979, at 8. I also noted that the plaintiffs had not shown “that the color of their tablet is distinctive.” Id. at 9. Evidence of passing off was missing. Id. at 11. The case can now be decided on a complete record, with all parties having had ample opportunity to adduce evidence addressing the factual issues underscored by my earlier opinion. On May 8,1979, Premo filed a motion for summary judgment on Counts 2 and 3. In an opinion filed on June 19, 1979, I denied the motion, without prejudice to renewal, to give the plaintiffs an opportunity to produce evidence of secondary meaning and passing off. Subsequently, contending that plaintiffs had failed to offer the necessary evidence, Premo again moved for summary judgment. Oral argument took place on the motion on October 9, 1979. At the conclusion of oral argument, I denied Premo’s motion, holding that plaintiffs had shown that genuine issues of material fact existed. In November 1979, plaintiffs amended their complaint against Pharmadyne by adding a fourth count. This additional count alleged that Pharmadyne had violated the Lanham Act through inaccurate statements made in advertisements for its dipyridamole. Shortly thereafter, Pharmadyne sought to file an amended answer which included a counterclaim that plaintiffs had broken the antitrust laws. Neither plaintiffs’ fourth claim against Pharmadyne nor Pharmadyne’s counterclaim are now before this court. On February 19,1980, Pharmadyne withdrew its antitrust counterclaim. Thereafter, I consolidated the Premo and Pharmadyne cases; at the same time I severed Count 4 of plaintiffs’ complaint against Pharmadyne. Thus, the only claims tried were plaintiffs’ claims that defendants had violated the Lanham Act and New Jersey law by copying the Persantine trade dress. This court has jurisdiction over plaintiffs’ claims under 43(a) of the Lanham Act, under 28 U.S.C. §§ 1331, 1332 and 1338(b). The court also has jurisdiction over the New Jersey statutory and common law claims through explicit grants of federal jurisdiction, 28 U.S.C. §§ 1332 and 1338, and on the basis of pendent jurisdiction. I Dipyridamole, a prescription drug, is a coronary vasodilator used by physicians in the long-term treatment of angina pectoris. Dipyridamole treatment usually takes many months, and almost invariably necessitates repeat or refill prescriptions. Dipyridamole is the generic name of this medication. Boehringer, which for many years has engaged in the sale and preparation of pharmaceutical products, has given its brand of dipyridamole the trade name Persantine. In 1966 Persantine became a federally registered trademark. From 1962 until about 1971, Geigy, Inc. (now Ciba-Geigy Corp.), marketed Persantine brand dipyridamole in the United States pursuant to a license agreement with Boehringer G.m.b.H. Since about 1973, Boehringer Ltd. has marketed Persantine brand dipyridamole in the United States under a patent and trademark license from Boehringer G.m.b.H. All Persantine brand dipyridamole marketed in the United States has been manufactured by Geigy under the control and by the authority of one or both of the plaintiffs. Persantine has been a commercial success. One reason for the profitability of Persantine has been that it enjoys the reputation with the medical profession of being a highly useful and safe medication. This reputation is in part based upon the substantial sums of money invested in research and development, clinical testing and quality control with respect to Persantine and other pharmaceutical products. In the case of Persantine, plaintiffs have maintained their reputation for quality by imposing strict standards for Geigy’s manufacture of Persantine. Plaintiffs periodically conduct tests to assure themselves that Geigy is conforming to all manufacturing specifications, including color of tablets. The testing and extensive promotion of Persantine and the satisfactory results obtained from it all play a significant role in the decisions of physicians to prescribe Persantine. Since October 1978, both Premo and Pharmadyne have been marketing dipyridamole tablets in the state of New Jersey and in interstate commerce. Both of these companies market generic pharmaceutical products that duplicate the trade dress of successful brand name drugs on which patents have expired. All parties market their dipyridamole tablets in doses of 25 mg. Plaintiffs have stipulated that the dipyridamole manufactured by defendants is the chemical and therapeutic equivalent of Persantine. Plaintiffs contend, however, that the dissolution rates of the products differ, and they further contend that some significance attaches to the variations in rates. I reject this argument. Variation in dissolution rates, unless linked to some chemical, biological, or physical trait that affects the performance or operation of the drug, is of theoretical interest only. It has no bearing on the resolution of this case. It is noted, however, that the New Jersey Drug Utilization Council has refused to place defendants’ dipyridamole on the New Jersey Formulary of generic drugs approved for substitution, citing “potential bioequivalence problems.” Thus, under New Jersey’s drug substitution law, N.J. S.A. 24:6E-1 et seq., defendants’ products cannot legally be substituted for plaintiffs’ Persantine. Cf. N.J.S.A. 24:6E-8. Dipyridamole itself is a yellow powder. Plaintiffs chose to sell Persantine in the United States in small orange biconvex tablets. This decision was not inevitable; any color could have been used. Indeed, plaintiffs sell a different colored Persantine in Germany; and defendants concede that no technological obstacle exists to selling dipyridamole in a color other than orange. Nonetheless, defendants also sell their dipyridamole in orange biconvex tablets. That defendants, the late-comers to the market, use the same color — and size and shape— tablet as plaintiffs’ is not happenstance. Both defendants admit that they intentionally copied the size, shape, color and overall appearance of plaintiffs’ Persantine brand dipyridamole. See Blackman deposition, Tr. 40. Pharmadyne has made a similar concession of intentional copying. Virtually identical to plaintiffs’ tablet in size, shape, and color, defendants’ tablets do differ in one respect: the logo. Persantine brand dipyridamole tablets bear the logo “BI” on one side and the number “17” in small print on the other side. “BI,” of course, stands for Boehringer Ingelheim. The number “17” aids professional people having access to reference books in identifying the specific product. Tr. 46. Pharmadyne’s dipyridamole has a black “P” printed on the capsule; Premo, which had no logo at all until April 1980, now utilizes a double “P.” These logos, however, cannot be regarded as a realistic means of allowing patients to identify the source of their dipyridamole. A side-by-side comparison of defendants’ dipyridamole products with Persantine reveals their tablets to be identical, for all practical purposes, to plaintiffs’. The logos appear in such fine print that few patients will detect the difference. Premo offered its logo as a distinguishing factor in SK&F v. Premo, 625 F.2d 1055 at 1061 (3d Cir., 1980). Judge Gibbons, writing for a unanimous court, rejected that contention: The district court, examining the capsules, specifically found that the logos were so small that they would be ineffective to prevent confusion as to source or passing off. We have also examined them, and conclude that the district court’s finding is not clearly erroneous. Indeed, an opposite conclusion well might be, for to our over-fifty eyes the logos are decipherable only with great difficulty. This observation holds particular relevance to the facts of this case for, as defendants have stressed throughout this litigation, many of the users of Persantine are elderly. Only those with acute vision are likely to detect differences in logo, and a substantial number of Persantine users will not have sufficiently sharp eyesight. The conclusion that the logos inadequately serve the goal of identifying for the patient the manufacturer of the dipyridamole, and alerting the patient to a change in product, is buttressed by the arguments tendered by defendants for allowing them to sell their dipyridamole in tablets similar to Persantine in appearance. In particular, defendants have vigorously insisted since the institution of this litigation that it is important for generic drugs to look like name brand drugs so as not to cause patient “anxiety.” Yet this argument — one of defendants’ principal justifications for the sale of look-alike medications — irreconcilably conflicts with a contention that logos enable patients to learn of the source of their medication. Cf. Hoffmann LaRoche, Inc. v. Premo Pharmaceutical Laboratories, Inc., Civ. 77-1001 (D.N.J. September 3, 1980), slip op., 20. It is clear that defendants intentionally designed their dipyridamole tablets to be confusingly similar to those sold by plaintiffs. Before turning to defendants’ list of claimed reasons for this copying, it may be helpful first, in placing some of the issues in context, to discuss the law of generic substitution as dealt with at trial. Cf., Hoffmann LaRoche, Inc. v. Premo Pharmaceutical Laboratories, Inc., supra, slip op., 10. In the early 1970s, some states repealed drug antisubstitution statutes and enacted generic drug substitution laws. The idea behind the new laws was to provide lower cost prescription drugs to consumers. Forty-six states have now enacted generic drug substitution statutes. Varying in many particulars, the statutes generally allow a pharmacist to fill a physician’s prescription for a brand name product with a bioequivalent product of another manufacturer. In some states substitution laws may, under certain circumstances, compel such substitution. Usually, the process of substitution begins with the physician. The doctor may prescribe a brand name product but indicate, in the appropriate place on the prescription form, “substitution permitted” or “do not substitute” or other, comparable language. See N.J.S.A. 24:6E-1 et seq. Assuming the filling were to take place in New Jersey, the pharmacist would then review the prescription when presented by the patient. If substitution is permitted by the physician, the pharmacist goes to the New Jersey State Formulary List and can substitute a generic there listed and matched with the brand name product. Most states are like New Jersey, in that substitutions can be made by the pharmacist only if the physician has indicated on the prescription form that substitution may take place. Not all, however, have the additional formulary requirement. In many states, the patient has the right to decide if he or she wants a generic drug. Thus, the pharmacist may say, “Your physician is permitting substitution. Do you want a generic drug instead of the name brand drug?” Sometimes, however, when the prescription is filled through a third-party payment plan, the patient may not have a choice. II Defendants offer a variety of explanations for why their dipyridamole must have the same appearance as that of plaintiffs’ Persantine. The three main reasons are: (1) to avoid causing anxiety in a patient who is “switched” from Persantine to defendants’ products; (2) to enable patients on several medications to identify each by a color, e.g., orange, green, and blue pills; and (3) to enable identification of a drug to be made in an emergency situation, i.e., here by the size, color and shape of the defendants’ products. Cf. SK&F v. Premo, supra; Hoffmann LaRoche, Inc. v. Premo Pharmaceutical Laboratories, Inc., supra (the same reasons were there tendered by Premo). Defendants have stated that if they “had not marketed their dipyridamole product in a form identical in appearance to persantine brand dipyridamole, defendants would have been marketing an inferior product in that a non-look alike generic dipyridamole would have resulted in patient anxiety thus lessening the therapeutic effects of the pharmaceutical.” Premo’s Critique to Plaintiffs’ Proposed Finding of Fact 13 (“Critique”). Defendants’ “anxiety” argument rests on several premises. First, defendants must assume that patients will not know that they have been “switched” from a brand drug to a generic. Copying the appearance of the brand drug so that patients do not know of the new source of their medication serves no purpose if patients learn of the change in some other way. Second, defendants must also, believe that a patient’s learning of a change in drug manufacturer actually generates “anxiety.” Neither premise can withstand scrutiny. In SK&F, Premo made the same “anxiety” argument. Arguing before the Third Circuit that the district court injunction should be lifted, Premo relied on an affidavit by Dr. Nathaniel Shafer: ‘In my opinion, many patients who have been using Dyazide on an extended basis would become uneasy, confused and react adversely if they received a renewal of their prescriptions with a different colored or shaped medication, even though the medication is completely identical. This could hamper the therapeutic effectiveness of the generic medication.’ SK&F at 1061. Judge Gibbons deftly punctured this analysis: The Shafer affidavit makes no reference, however, to the requirement in the law of most states which permit generic substitution that- the purchaser be notified of the substitution. Assuming compliance with those laws it is difficult to credit the assertion that a difference in color would increase the patient’s anxiety, because the patient will be informed of the substitution and will usually have the right to refuse to accept the generic. Id. Thus, the very generic substitution laws which defendants rely upon mandate that the patient learn of the substitution. See N.J.S.A. 24:6E-7-10. Moreover, the evidence in the record in this case shows that patients will ordinarily be told that they will be or are receiving a generic drug. Medical experts from both sides testified that when they permit substitution they will alert the patient of the change; at the same time they will inform the patient of the appearance of the generic. During this discussion the doctor will tell the patient that a non-look-alike generic is the same medication as before and that substitution is being permitted to save the patient money. It further appears that, even if the physician does not advise the patient of the switch, the pharmacist ordinarily will do so. As has been noted, the pharmacist should advise the patient of the proposed substitution, so that the patient can decide whether he or she wants a generic drug. Thus, regardless of whether the generic is identical in appearance to the brand name, the patient ought to be — and usually is — told that a generic medication is now being dispensed. Moreover, if neither the physician nor pharmacist alerts the patient to the substitution of a look-alike, another factor, in most cases, will lead a patient to become aware of the substitution. The generic drug industry has grown precisely because generics generally cost the patient much less money. Thus, in the normal circumstance, i.e., when the price savings are passed on to the consumer, the patient will be alerted to the change in drug by the reduction in price. Questions concerning the drop in price will then be answered by the physician or pharmacist, and the answer will include disclosure of the substitution of the generic drug. See N.J.S.A. 24:6E-7-10. Thus, imitating the trade dress of a brand name to conceal from the patient the fact that he or she is getting a generic, is ultimately futile, if, as is usually the case, the physician and pharmacist fulfill their responsibilities. From one source or another, most patients will become cognizant of the switch. Defendants are no more successful in supporting the second prong of the argument: switching causes an “anxiety” that diminishes the effectiveness of the drug. Dr. Griep, the plaintiffs’ physician whose testimony I credit, testified that patients whom he did not advise about the switch have telephoned him after they have received a generic with an appearance different from that of the accustomed brand name medication. After he explained the situation, his patients accepted the substitution without anxiety. Dr. Griep, Tr. 189-90. See Dr. Staffer, Tr. 91. The testimony of Dr. Bloom, one of the defendants’ witnesses, did not aid the defendants. Dr. Bloom testified that in prescribing drugs, he tells the patient of their appearance and that the generic may not be a look-alike. These explanations almost always allay any doubts the patients might have had. Dr. Bloom testified that in some 13 years of practice, with some 5,000 patients, he had only 1 or 2 patients who would not take a generic equivalent of a different color than the brand name the patient was getting. Tr. 579. The solution employed by Dr. Bloom in this rare case is simple and available to other doctors: maintaining the patient on the same medication. Tr. 568. Moreover, even as to these few cases, defendants did not elicit from Dr. Bloom that the patients would have accepted a generic look-alike. Not one physician or pharmacist testified that he or she would not inform the patient of the switch, irrespective of whether the generic had the same appearance as the brand name. If “anxiety” were such a difficult problem, one would have expected defendants to produce testimony that physicians and pharmacists routinely omit any mention of the substitution when discussing medication with patients. Yet no such testimony was forthcoming. Defendants did introduce testimony that pharmacists would prefer not to discuss the fact of substitution with customers. Mr. Sperber, an employee of Henry Schein, a generic distributor for Premo’s dipyridamole, was disarmingly frank: pharmacists do not want generic non-look-alikes because they always necessitate an explanation to the patient. This testimony, though, did not demonstrate that non-look-alikes promoted “anxiety” — it did show that patients who suddenly are given a non-look-alike may wonder why they have received a drug with a different appearance. It also suggests that at least some pharmacists may not advise a patient of a “switch.” Defendants’ “anxiety” argument is without merit. If legitimate medical and pharmaceutical practices are followed, the patient knows that he or she is having a generic substituted at a lower price. If the patient is alarmed, the fears usually will be relieved by discussion. And if the patient remains “anxious,” the doctor can resume use of the name brand medication. Each of these general propositions holds true for dipyridamole, regardless of whether defendants use orange biconvex tablets or pink- and-green capsules. Defendants’ “anxiety” argument is plausible only under circumstances which do not advance defendants’ position. To eliminate patient “anxiety,” as defendants claim they have done, the patient must not be told of having been “switched.” Under the defendants’ approach, the physician would not alert the patient, and the pharmacist, in filling the prescription, would say absolutely nothing to the patient of the displacing of the name brand by the generic — and would charge roughly the same price as for the trade name drug. The result: the patient suffers no “anxiety” because the patient does not know of the “switch.” Patient “anxiety” is eliminated by pharmacy deception and resultant patient ignorance. I now turn to the second of defendants’ reasons for copying Persantine’s appearance — it enables patients to identify by col- or the medication to be taken for each ailment. Prom the testimony of Dr. Bloom, Tr. 557, and Mr. Sperber, Tr. 288 et seq., the defendants argue that once on an orange tablet for their angina pectoris, patients must remain on an orange tablet when “switched” from a brand name to a generic drug. Yet defendants offered no testimony to support the fundamental, if unarticulated, premise that patients cannot mentally adjust to “switching” from an orange “heart pill” to a green “heart pill.” In the absence of such testimony, I am unwilling to believe it to be beyond the ability of patients to stop associating orange with their heart medicine and start thinking of green as the color of the pill taken for their heart condition. Dr. Bloom’s color chart does not support defendants’ “color” argument. See Tr. 553; 558; D.Pr.Ex. 53. When Dr. Bloom “switches” a patient to a generic of a different color than the brand name, he can simply change the card given the patient, at the same time advising the patient that the “new” drug will be green in color instead of red. Tr. 557; 563. The last of the principal reasons advanced by defendants for their copying Persantine’s appearance is that it will aid physicians in emergency situations in identifying the drug taken by a patient who has overdosed. Dr. Bloom, Tr. 571-74. I do not believe that the goal of assisting physicians played any part in the defendants’ decision to simulate Persantine’s trade dress. Dr. Bloom, testifying for the defendants, completely discredited this rationale in a colloquy with the court. THE COURT: What I’m trying to get at is there . . . any conceivable benefit in an emergency situation to have a generic match in shape, color and size Boehringer Ingelheim’s Persantine? THE WITNESS: Specifically for Persantine you say? THE COURT: Yes. THE WITNESS: Well, since I don’t think it can engender a kind of emergency room visit of the kind we discussed, I don’t think it’s pertinent here .... Tr. 575. See also Dr. Griep, Tr. 181. Additionally, even if Persantine were a drug that could produce overdose situations, defendants undermined their argument that color would be useful in identifying the active ingredient. During cross-examination of plaintiffs’ witness Norman Hacking, Premo’s counsel asked, “I take it you don’t dispute that there are drugs which are identical in appearance to that dipyridamole manufactured by your company?” Hacking answered: “I don’t dispute that?” Tr. 56. Thus, the three principal reasons advanced by defendants to explain their copying are bereft of merit. The defendants also insist that they help to fulfill the purpose and spirit of the generic substitution laws of the various states by mimicking the appearance of Persantine. Actually, the state laws called to my attention do not address the question of lookalikes. If any state.. believed it to be desirable to have generics imitate name brand drugs, the generic drug substitution laws could have authorized look-alikes. Yet, apparently no state has acted to approve look-alikes by generic companies. Therefore, I fail to perceive how defendants’ simulation of Persantine’s trade dress furthers legislative goals in the enactment of generic substitution laws. Indeed, the actions of defendants may conflict with the public policy embodied in most drug substitution legislation. As I have pointed out, generic substitution laws contemplate that the patients will be advised of any “switch” to a generic; consumers are to be informed of a change in drug manufacturers of their medication. As is demonstrated below, defendants’ decision to copy Persantine’s appearance, instead of advancing that goal, may thwart it. See N.J. S.A. 24:6E-7-10. A subsidiary justification tendered by defendants for borrowing Persantine’s appearance is that generic drug manufacturers for at least 30 years have manufactured nonbranded generic products in the same color, size, and shape as the corresponding national branded products. To support this contention, defendants submitted an exhibit of generics that look the same as their name brand equivalent. D.Pr.Ex.4. Yet, this proves little, for plaintiffs have countered with an exhibit of generic drugs looking nothing like the corresponding national brand. P.Ex. 85; 141, 141a-i. In SK&F v. Premo, Judge Gibbons wrote that: [t]he parties conceded at the argument on the application for preliminary injunctive relief that although an exhibit could be devised showing numerous examples of generics that have adopted the form and appearance of the branded drugs, an exhibit of like number of dissimilar generics could be prepared as easily. At 1060 n.3. The record in this case confirms the validity of that concession — both look-alike and non-look-alike generics are currently being marketed. It should also be noted that in his deposition defendant Blackman recognized that branded generics — generally a generic product put out by a larger pharmaceutical firm — have an appearance different from that of the branded drug. Speaking of branded generics, Blackman said their manufacturers “make their product distinguishable in one way or another, be it slight or gross.” Dep.Tr. 26. The defendants produced testimony of doctors and pharmacists that all generics should look like their brand name equivalents. This, of course, would simplify the job of those physicians and pharmacists who wanted to be relieved of the responsibility and necessity of telling patients that there has been a “switch” from a brand name to a generic. Patients kept in the dark about the change will be neither curious nor querulous. Yet this testimony is a two-edged sword. It tends to substantiate plaintiffs’ claim that pharmacists who do not relish explaining the “switch” may be tempted to take advantage of the opportunity afforded by generic look-alikes to change medications without telling their customers. I do not accept defendants’ testimony that generics ought to look like their name brand counterparts. Instead, I credit the contrary testimony offered by plaintiffs. Dr. Stoffer clearly and convincingly explained why he had concluded that generic drugs containing the same ingredients as brand name drugs “should be made in different colors, sizes or shapes.” Tr. 88. I found many patients in my personal practice had been substituted to a generic drug without their knowledge and consent. And when you ask these patients, in fact, what drug they’re taking, they name the brand name drug, and it was no small wonder that they did, since the bottle was labeled by the pharmacist in practically all eases with the brand name, and since the prescription that I wrote for that patient was for the brand name drug, and since they were not advised by the pharmacist that the drug dispensed was anything other than the brand name drug, they came to the conclusion that that’s what they were taking, and since the tablet inside the bottle looked very much, size, color, shape like the brand name tablet, they assumed that they were taking the brand name tablet. Tr. 89. See Tr. 90. From the testimony adduced at trial, it has become abundantly clear to me that defendants deliberately decided to copy the size, shape, and color of plaintiffs’ Persantine. And it has become equally clear that the reasons provided by defendants for copying Persantine’s appearance are utterly specious. This leaves open the question of what actually motivated the defendants to imitate Persantine’s trade dress. My review of the evidence persuades me that the explanation of defendants’ conduct is that such copying increases sales of their generic dipyridamole because unscrupulous pharmacists will buy it to pass it off profitably as Persantine. Several lines of analysis lead ineluctably to this conclusion. Without evidential support, and internally inconsistent, the reasons proffered by defendants for their copying in and of themselves suggest the existence of an improper motive. Since the institution of this lawsuit, defendants have argued that sound pharmacological and medical practice dictated their decision to sell small orange biconvex dipyridamole tablets. However, as I demonstrated above, none of these arguments survives close scrutiny. On the one hand, there is a dearth of credible evidence to substantiate defendants’ explanations for why they copied, e.g., patient anxiety, emergency room identification in case of overdose, et cetera. At the same time, defendants are trapped by their own arguments. For example, Pharmadyne, in its Proposed Findings of Fact 45-50, emphasizes the importance of the placebo effect in giving patients relief. Thus, Pharmadyne contends, “If patients who had been taking Persantine were switched to the generic version which had a different color, they will suffer anxiety which would limit the effectiveness of the drug making it inferior to Boehringer’s Persantine solely because of the placebo effect.” Id. 46. Yet, Pharmadyne also maintains that its logo will enable consumers to know the source of their medication. Thus, plaintiffs’ Proposed Finding of Fact 28 says, “Defendants’ copying of the PERSANTINE® appearance for their generic dipyridamole makes detection of substitution of defendants’ products for PERSANTINE® difficult, if not impossible, for consumers.” Pharmadyne responded: “Pharmadyne’s logo on its tablets and Boehringer’s imprints on its products make it easy to identify the respective products.” Pharmadyne Critique 28. Surely, if patient “anxiety” were a genuine concern, Pharmadyne would not place on its products a mark designed to alert patients they were now getting a new brand of medicine. The fact is that neither defendant company has come forward with a legally acceptable reason for copying Persantine’s appearanee. I am virtually compelled to draw the inference that the unstated, but actual, reason for their having duplicated Persantine’s trade dress is that defendants hoped to gain sales by having pharmacists pass off the look-alike dipyridamole as Persantine. Some direct support for this inference comes from defendants’ advertising. Both defendants clearly set forth in their advertisements that their dipyridamole can be bought in place of plaintiffs’ Persantine. For example, a Premo distributor, Henry Schein, Inc., provides the pharmacist readers of its catalogue with the following message: “NEW GENERIC! Compare to Persantine Dipyridamole Tablets 25 mg. Orange-Sugar Coated” P.Ex. 64a. By drawing attention to the comparability of the products and the identity of colors, defendants have made drug wholesalers and pharmacist buyers of generic dipyridamole aware that the customer can be surreptitiously given their less expensive generic product in the place of the more costly Persantine. Further evidence that defendants expect pharmacists to “pass off” their dipyridamole for Persantine comes from a letter written by Blackman to pharmacists. In his letter of January 10, 1980, Blackman analyzes for his customers the consequences of the district court’s decision in SK&F v. Premo. After warning pharmacists of the “hazardous situations and grave economic consequences which would ensue” — these essentially being a rehash of the justifications offered in this court for copying— Blackman raises the following specter: Equally disconcerting is the heavy economic impact upon the dispensing pharmacist. How does an injunction against look-alike generic drugs affect you economically? As each manufacturer comes into the marketplace with a product which is being made by one of his competitors, the newcomer will have to choose a different color, shape and size from each preceding manufacturer. To enable you to legally and properly fill any and all “no substitute prescriptions,” you will be obliged to inventory every manufacturer’s product, no matter how small the demand. You would literally have tens of thousands of bottles on your shelves, for which you have no room and your dollar inventory would increase 100-fold, which you cannot afford. P.Ex. 70 (emphasis in original). This statement is rather curious. One would think that as manufacturers entered the market, a pharmacist would have to stock the product of each manufacturer to fill properly a “no substitute prescription,” regardless of color, of the new product. Color, therefore, should not affect a pharmacist’s decision whether to carry the new entrant’s drug. The size of an inventory ought to be independent of the colors of the drugs comprising that inventory. Color becomes relevant only if the pharmacist intends to substitute a look-alike drug for the medication actually written on the prescription. If substitution takes place, then the pharmacist need not carry numerous other look-alikes; one or two will do. The implication of Blackman’s statement is clear and unavoidable: The advantage to pharmacists of generic look-alikes is that they allow the pharmacist to buy a smaller number of products. And the only way a pharmacist can obtain that advantage is to dispense the look-alike generic without regard to the brand name actually appearing on the prescription form. As noted earlier, defendants had no legitimate medical reason for imitating the appearance of the Persantine tablet. The reason suggested by the evidence is that defendants hoped to capture quickly a large share of the market and that this could be done only by enabling pharmacists to pass off dipyridamole as the more expensive Persantine. Defendants’ desire to acquire rapidly a sizable market share is reflected by the evidence concerning their marketing practices. For example, both defendants began to sell their dipyridamole several months before plaintiffs’ patent expired. Once plaintiffs brought suit, the defendants readily consented to an injunction on the patent infringement claim. The obvious inference, therefore, is that the defendants began selling dipyridamole before patent expiration, not because they believed the patent invalid, but to gain a running start in penetrating the dipyridamole market. When defendants began producing their generic dipyridamole, they faced substantial barriers in penetrating the market. For over 12 years Persantine had been the sole brand of dipyridamole available. Defendants were aware that physicians, who had been the targets of repeated major promotional campaigns by Boehringer, and who had confidence in Persantine as an effective medicine, would not readily abandon Persantine for generic medication. They also knew that patients on Persantine for years, or at least many months, would not quickly shift to a generic if given a choice through notice. Additionally, they knew their products were not on the New Jersey Formulary. The reality of those fears that doctors would not quickly allow pharmacists to substitute their dipyridamole for Persantine is borne out by the testimony of Bernard Koff, plaintiffs’ witness. Mr. Koff, whose testimony I credit, is president of Market Measures, Inc., a pharmaceutical research agency conducting audits in the prescription drug field. Since being founded in 1968, Market Measures has run more than 100 surveys. In September 1979 Mr. Koff’s firm performed a seven-state audit on behalf of plaintiffs. The seven states — Delaware, Illinois, Missouri, New York, Pennsylvania, Virginia and Washington — were selected because they had double-line signature laws. That is, in each state a doctor, when prescribing medication, must sign either the line on the prescription form authorizing substitution or the line prohibiting substitution. Because doctors are compelled in these states to decide whether substitution should take place, auditing prescription forms provides some measure of the preference, if any, for the brand name drug. The survey undertaken here consisted of reviewing all prescriptions on file at 188 pharmacies. Over 125,000 prescriptions were examined. The researchers found approximately 590 prescriptions written for Persantine; of these, 520 were on a prescription blank where the double line was used. Significantly, of the 520 prescriptions for Persantine where the doctors chose to permit or prohibit substitution, physicians disallowed substitution 78 percent of the time. Id. Mr. Koff testified that the forms uncovered by the researchers were written by over 200 different doctors. Tr. 138. The survey in question was not a probability sample but, rather, a stratified quota sample. In a probability sample, all members of the universe are equally likely to be selected; in a stratified quota sample, the sample is chosen in such a way that it will be composed of a specified proportion of objects with a particular trait. Tr. 214-16. Thus, in the audit of Persantine prescriptions, the pharmacies to be included in the survey were not chosen randomly from a list of all pharmacies but were selected to provide a sample with specified percentages of stores that were rural or urban, independent or chain-owned; statewide geographic distribution was also sought. Tr. 216. Because the survey was not a probability sample, the results cannot be statistically extrapolated to the universe. Tr. 131. Therefore, I will not accord the survey as much weight as I would a probability sample. Nonetheless, the survey deserves some weight. Mr. Koff testified that in at least 25 studies the results obtained from a stratified quota sample were similar to the findings generated by a probability sample. Tr. 217. The technique has been used in the pharmaceutical market research field for a long period of time; arid it is an accepted tool in that field, as well as in other areas of market research. Mr. Koff concluded that stratified quota samples yield results comparable to those of probability samples. I accept this testimony. I also note that the defendants had months to prepare their own study to meet the results of Mr. Koff’s, whose report they had been given. They did not do so. Additionally, the defendants, through the testimony of Drs. Gatty and Abelson, attacked the validity of the Rittenhouse survey; neither, however, challenged the Koff survey. Accepting Mr. Koff’s survey, find that it tends to show that a substantial majority of doctors prescribe Persantine and do not allow substitution. This illustrates the problems confronting the defendants in their drive for prompt market penetration. In most states, if a physician prohibits substitution, the pharmacist may not lawfully substitute. Thus, as defendants correctly anticipated, a large part of the potential market would be closed to them where doctors prohibited substitution for Persantine. The only way to overcome such prohibition is to have pharmacists fill a Persantine prescription with generic dipyridamole without telling the customer. And to avoid arousing suspicion in the patient, the generic product must be of the same appearance as the brand name medicine and sold at the same price. As defendant Blackman said in his letter to pharmacists, if patients are given a different colored medicine, they will become inquisitive: “Are you sure this is the same medication I received last time? It looks different.” “I think you made an error.” “Please call my physician and double check.” “This medication isn’t the same as I previously had taken.” P.Ex. 70. Accordingly, I find that defendants simulated the appearance of plaintiffs’ Persantine-brand dipyridamole tablets in order to rapidly penetrate the market and thereafter increase sales by enabling pharmacists to pass off surreptitiously their dipyridamole as Persantine. I further find that the defendants Blackman and Bedrick were personally responsible for the decision of their respective companies to market a tablet virtually identical with Persantine. Ill Under the general rubric of unfair competition, plaintiffs have alleged that defendants have committed the tort of unprivileged imitation. Defendants disagree, contending that plaintiffs have failed to establish the necessary elements of that tort. To make out a claim of unprivileged imitation, one must show both nonfunctionality and secondary meaning. These are necessary elements both under New Jersey case law and the Lanham Act. SK&F v. Premo, supra, at 1063, 1066. Dipyridamole is a yellow powder. The initial choice of color of the Persantine tablet was arbitrary and unrelated to any therapeutic effect; this is also true with respect to its size and shape. The maintenance of this size, color, and shape has also been arbitrary. In the United States, Persantine is marketed as an orange tablet; in Germany, plaintiffs’ wholly owned subsidiary markets it as a yellow tablet. I accept and credit Dr. Griep’s testimony that the color, size, and shape of Persantine are not functional. Dipyridamole in the dosage of 25 mg. can be marketed in an infinite combination of colors, sizes, and shapes. Pharmadyne has contended that Persantine’s trade dress is actually functional. I reject that claim. Indeed, defendant Black-man testified to the contrary. Q So, for the first-time users of dipyridamole the size, shape, and color of the tablet has no therapeutic function; isn’t that correct? A I would say that is quite correct, in my opinion. Dep.Tr. 46. I turn now to the question of secondary meaning. The field of cardiovascular pharmaceutical preparations, as defined in the 1979 Physicians’ Desk Reference, includes more than 200 products. The overall appearance of Persantine brand dipyridamole in this field is unique. See P.Ex. 136. I find unpersuasive defendants’ evidence purporting to show a lack of distinctiveness. Tr. 55-63. Plaintiffs have expended a good deal of time, money, and energy in developing Persantine’s sales. For more than 15 years Persantine brand dipyridamole has been promoted to physicians and pharmacists for use in treating patients suffering from certain cardiovascular disorders. Throughout that period Persantine has been marketed in its distinctive orange-colored, smoothly rounded, biconvex trade dress. Boehringer has some 450 scientifically trained sales people who call on doctors to inform them of Boehringer’s products. On any given day these sales people will visit some 3,000 doctors in the United States; “a very large percentage of those calls would concern Persantine.” Tr. 17. Boehringer also advertises Persantine in medical journals. At any one time these advertisements will appear in not less than 20 separate journals. Supplementing visits by sales people and advertisements in periodicals, Boehringer also sends mail directly to physicians on a regular schedule. Additional promotion of Persantine takes place at medical conventions. For the last seven or eight years physicians have been given over a thousand samples per month for their personal use. The cost to Boehringer since 1972 of these promotional activities has exceeded $15 million. This excludes costs allocable to sales force efforts. These vigorous sales campaigns have produced excellent results. Since 1972, Boehringer has sold over two billion Persantine tablets. Nonetheless, defendants argue that Boehringer’s trademark is no longer valid. This is simply not so. To protect its trademark, Boehringer has displayed the word Persantine in prominent lettering, as well as the registered trademark “R” designation. The word dipyridamole accompanies the trademark Persantine to inform the consuming public that Persantine is a trademark for dipyridamole. In all of Boehringer’s promotional advertising the generic name dipyridamole has been stated, making it clear to physicians and pharmacists that Persantine is, in fact, the trademark. Moreover, many of Boehringer’s promotional efforts have been designed not only to familiarize doctors and pharmacists with the name “Persantine” but also with Persantine’s appearance. P.Ex. 2 is a typical example of Boehringer’s attempt to link in the mind of physicians the name “Persantine” and its appearance. The advertisement reads, “Persantine brand of dipyridamole is available as round orange sugarcoated tablets.” Similarly P.Ex. 3 shows that plaintiffs, contrary to defendants’ assertions, have endeavored to highlight for doctors Persantine’s appearance: “Persantine brand of dipyridamole is available as round orange-colored orange sugar-coated tablets of 25 milligrams.” See P.Ex. 4 — 6. P.Exs. 11 and 15, among others, refute defendants’ contention that Boehringer did nothing to promote pharmacists’ associating Persantine’s trade dress with the Persantine product. Accordingly, I find that heavy promotion of plaintiffs’ Persantine has taken place, and that this promotion has emphasized both the appearance and origin of the product. I further find that physicians are motivated to prescribe Persantine, and pharmacists purchase it, because of its source. These facts add up to highly persuasive evidence of the existence of secondary meaning. The defendants’ decision to copy Persantine’s appearance also supports the conclusion that Persantine’s trade dress has acquired secondary meaning. Defendants deliberately chose to imitate Persantine’s appearance. As demonstrated above, the decision to simulate plaintiffs’ product was not motivated by legitimate business or medical concerns. Product simulation occurred so that defendants’ generic dipyridamole could be sold to customers without their resisting substitution. Imitating Persantine’s appearance served no purpose other than to eliminate problems defendants anticipated in selling their dipyridamole if it were dissimilar from Persantine. This, then, is the key: Defendants’ marketing practices are convincing evidence of their belief that consumers would not accept a differently colored form of dipyridamole. Persantine’s appearance was nonfunctional when originally selected by Boehringer. It remains nonfunctional today. The sole value to Boehringer of Persantine’s appearance is that it identifies the source of the pills for users. Defendants, recognizing that consumers believed the orange color indicated the source of their medicine, knew that a pink or maroon or green pill, or an orange pill of different size or shape, would not be accepted. Indeed, defendants’ lawyers admitted this point at oral argument: “The reason we’re doing this is to gain patient acceptance so as not to alarm the patient with a different color.” Tr. 41-42. And the reason a different color or shape or size would not be accepted was precisely because the patient would then know that the pill was originating from a new source. Dispensing generic dipyridamole when the prescription calls for Persantine can take place without detection by the patient only if the patient is not alerted to the fact that he or she has been switched to medicine coming from a different source. Persantine’s trade dress has acquired secondary meaning; therefore, only if the substitute has the same trade dress can the buyer be misled into thinking that the medicine comes from the accustomed source. Conversely, if trade dress does not indicate source, varying the appearance would not cause the patient to think the medicine originated from an unfamiliar source, because the patient would not associate appearance with source. Thus, defendants’ selection of a tablet with trade dress virtually indistinguishable from that of the plaintiffs’ betrays their belief that the plaintiffs had established secondary meaning in the Persantine trade dress. Putting together plaintiffs’ advertising of Persantine, the commercial success of Persantine, the efforts to educate doctors and pharmacists as to Persantine’s appearance, the results of the Koff survey, the decision of defendants to simulate Persantine’s trade dress, and the utter lack of justification for that decision, I find that plaintiffs have proved secondary meaning as to Persantine’s trade dress. Since Persantine’s trade dress is nonfunctional, plaintiffs have established the two elements of the tort of unprivileged imitation. At trial, plaintiffs offered into evidence the results of a survey they commissioned Dr. Marguerite Rittenhouse to perform. The purpose of the survey was to determine if patients, pharmacists and doctors would associate the name “Persantine” with a small, orange biconvex tablet. Defendants moved to exclude the survey and Dr. Rittenhouse’s testimony. At trial I decided that both her testimony and the survey should be admitted, with any deficiencies in survey design or execution going merely to the weight of the evidence. Upon further reflection, I have concluded that the survey should not have been admitted. Properly conducted surveys are admissible. The Court of Appeals for the Third Circuit has discussed the attributes common to well-run surveys. A proper universe must be examined and a representative sample must be chosen; the persons conducting the survey must be experts; the data must be properly gathered and accurately reported. It is essential that the sample design, the questionnaires and the manner of interviewing meet the standards of objective surveying and statistical techniques. Just as important, the survey must be conducted independently of the attorneys involved in the litigation. The interviewers or sample designers should, of course, be trained, and ideally should be unaware of the purpose of the survey or the litigation. Pittsburgh Press Club v. United States, 579 F.2d 751, 758 (3d Cir. 1958). Because the Rittenhouse survey did not satisfy several of the requirements set forth in Pittsburgh Press, defendants’ motion to strike the survey and the Rittenhouse survey will now be granted. One problem with the survey was that plaintiffs’ attorneys were involved in designing the questions that were to be asked. See, e.g., Tr. 443-44. This conflicts with the admonition in Pittsburgh Press that surveys be designed without guidance from lawyers. A second weakness of the survey stems from Dr. Rittenhouse’s knowledge of the purpose of the survey. According to Pittsburgh Press, the person conducting the survey ought not to know the goals of the party paying for the survey. Dr. Rittenhouse, however, did know the aims of the survey; indeed, at trial she testified both as to what were, and were not, the survey’s objectives. See Tr. 535-36. Cf. James Burrough Ltd. v. Sign of Beefeater, Inc., 540 F.2d 266, 278 (7th Cir. 1976) (interviewer unaware of survey’s purpose). The most serious defect in the survey arises out of Dr. Rittenhouse’s method of selecting subjects to be interviewed. Initially, I admitted the evidence on the basis of Dr. Rittenhouse’s statement that a sample yielded by her technique would produce results comparable to those from a probability sample. However, upon further reflection, I now believe that the lack of any effort to ensure that representative interviewees were obtained, see, e.g., Tr. 444-46; Dep.Tr. 56-57; 125-27, when combined with the other faults of the survey, requires the survey to be ruled inadmissible. To put it mildly, the selection process used here was haphazard. Doctors were visited because they were high prescribers of Persantine; patients were interviewed because doctors or pharmacists told Dr. Rittenhouse the names of Persantine users. Pittsburgh Press stressed the importance of choosing a representative sample. 579 F.2d at 578. See Bank of Utah v. Commercial Security Bank, 369 F.2d 19, 27 (10th Cir. 1966), cert. denied, 386 U.S. 1018, 87 S.Ct. 1374, 18 L.Ed.2d 456 (1977) (“A survey is inadmissible when the sample is clearly not representative of the universe it is intended to reflect”); Donohue v. Board of Elections of State of New York, 435 F.Supp. 957, 970 (E.D.N.Y.1976); Allen v. Morton, 333 F.Supp. 1088, 1094 (D.D.C.1971); The Manual for Complex Litigation 2.716 (Tentative Draft 1980); Note, Public Opinion Surveys as Evidence: The Pollsters Go To Court, 66 Harv.L.Rev. 499 (1953); Handbook of Recommended Procedures for the Trial of Protracted Litigation, 25 F.R.D. 351, 429 (1960). Yet the sample comprising the survey introduced by plaintiffs simply is not representative. I do not believe prejudice will accrue to any party as a result of my decision to exclude the Rittenhouse survey at this stage of the proceeding. Defendants surely are not harmed, for they have accomplished one of their long-standing objectives: the exclusion of survey done by Dr. Rittenhouse. And plaintiffs will suffer no injury, for the other evidence in the record is sufficient to show that secondary meaning exists. IV Plaintiffs alleged in their complaint that the defendants, by distributing generic dipyridamole in a trade dress that invited confusion and passing off, have violated § 43(a) of the Lanham Act, the New Jersey Unfair Competition statute, and New Jersey common law. I find that plaintiffs have proved their claim by a preponderance of the evidence. As clearly appears, defendants deliberately copied the arbitrary trade dress of plaintiffs’ product. In SK&F v. Premo, supra, at 1062, Judge Gibbons stated that “it has been held that it is actionable under New Jersey law for a drug manufacturer to put a product in the hands of a pharmacist in a form in which the manufacturer can reasonably anticipate that it may be passed off as another product even if the manufacturer does nothing else to encourage passing off. Under this formulation, defendants’ liability is clear. Not only did defendants anticipate passing off; they courted it through advertising and counted on it to boost sales. Plaintiffs also adduced some evidence of actual passing off. While not conclusive, this evidence supports plaintiffs’ contention that defendants have facilitated the passing off of generic dipyridamole as Persantine by copying Persantine’s trade dress. At the request of plaintiffs, Pharmacy Practices Review, Incorporated, undertook a field survey to determine what medicine would actually be dispensed when a physician prescribed Persantine. Robert Taylor, president of Pharmacy Practices Review, and Aubert Schwartz, its field director, detailed the methodology of the survey. Based on their testimony, which I credit, I find the survey to have been conducted properly and to be probative on plaintiffs’ claim of passing off. During the survey, 79 different pharmacies were shopped. Of the 79 pharmacies, 57 dispensed Persantine and labeled the vial correctly. As P.Ex. 66-1A reflects, 22 pharmacies did not dispense Persantine, though the prescription called for Persantine. Plaintiffs do not contend that in each instance the dispensing of generic dipyridamole instead of Persantine constituted passing off. For example, a drugstore in Sunrise, Florida, labeled the bottle “GENERIC FOR PERSANTINE. 25MG”; a pharmacy in Millville, New Jersey, wrote on the label that it was dispensing “Dipyridamole 25mg.” Consumers receiving generic dipyridamole in vials labeled in this fashion will not be misled into believing they have been given Persantine. However, based on the survey, one can conclude that some consumers receiving generic dipyridamole will not know they are receiving Persantine. Eight pharmacists labeled the vial to indicate that the product was generic but the label also used the name “Persantine.” Whether adding the letters “Gen” or the word “generic” actually tells the consumer that he or she is being switched from Persantine is questionable— absent explicit information from the pharmacist. Consumers are unlikely to learn of the change in medication from what is probably, for them, an obscure word or phrase. Moreover, in six instances the pharmacist labeled the medication “Persantine” without any indication on the label that the product was actually a generic. Attempting to weaken the power of plaintiffs’ evidence, defendants, on cross-examination of the buyers used in the survey, brought out the fact that some pharmacists whose labels were misleading orally advised the buyer that he or she was getting a generic drug. Whether this makes clear to a consumer that he or she is not getting Persantine — the name of the drug on the label — is dubious. However, even if orally contradicting the label eliminates consumer deception, there remain ten occasions on which pharmacists dispensed a generic product without telling the buyer and used the name Persantine on the label. And in five of these instances the consumer had no means of knowing that the vial did not contain Persantine, as named by the physician and written on the label. Defendants, in an apparent attempt to minimize the damage done to their position by the survey, contend that pharmacists labeled the generic dipyridamole product “Persantine” because Persantine has become the generic name for dipyridamole. Unsupported by the evidence, that argument must be rejected. As earlier noted, plaintiffs have introduced voluminous amounts of advertising in which the word Persantine is followed by the small “r” in a circle signifying reg