Aaron Bravman ("Bravman") and his wife, Muriel Bravman, appeal from a judgment of the United States District Court for the Southern District of New York, Robert W. Sweet, Judge, granting Baxter Healthcare Corporation ("Baxter") summary judgment and dismissing Bravman's amended complaint in a diversity case. The district court determined as a matter of law that Bravman had suffered no physical harm from the implant of a noisy artificial heart valve manufactured by Baxter and did not have a cause of action for the noise and ensuing nervousness and sleeplessness nor for the fear of heart valve failure that the noise engendered. For the reasons set forth below, we affirm in part, reverse in part, and remand with directions. We hold that Bravman does not have a claim for relief under a product or design defect theory of liability. We hold, however, that Bravman is entitled to a jury determination whether the noise emitted by the heart valve caused a cognizable injury and whether Baxter violated a duty to warn Bravman's doctor of the potential noise.

On April 5, 1988, Bravman, a Pennsylvania resident, underwent surgery at New York University Hospital to replace his natural mitral heart valve with an artificial Edwards-Duromedics Heart Valve made by Baxter, an Illinois corporation. Bravman's surgery was made necessary by the serious deterioration of his natural heart valve. Dr. Frank Spencer of New York University Hospital performed the surgery. Bravman met once with Dr. Spencer before his operation. At this meeting, Dr. Spencer recommended that Bravman use a mechanical valve rather than a porcine valve, made from natural pig tissue, because of the mechanical valve's greater longevity. However, Dr. Spencer did not discuss with Bravman the variety of mechanical valves available nor did he warn Bravman about the noise associated with the Edwards-Duromedics valve.

Bravman's surgery was successful and his artificial valve continues properly to control the flow of blood into his heart. Nonetheless, Bravman complains of a complication--his valve makes a loud and disruptive noise. Although everyone concedes that all mechanical valves make some noise, Bravman asserts, and is supported in his assertions, that his Edwards-Duromedics valve makes an unacceptable amount of noise. In fact, Bravman contends that, in a quiet environment, his heart valve can be heard from a distance of thirty feet. As a consequence of the valve noise, Bravman claims that he cannot sleep, that he is despondent, and that he was obliged to take an early retirement. Bravman states he has even explored the possibility of undergoing additional surgery to replace the valve to eliminate the noise.

Bravman raised four claims for relief before the district court: breach of express warranty, breach of implied warranty, negligence, and strict product liability. After discovery was closed, Baxter filed a motion for summary judgment, requesting that Bravman's complaint be dismissed with prejudice. The district court granted the motion on the basis that Mr. Bravman's fear that his mechanical heart valve was going to fail was legally insufficient to support a claim that he was injured and that the excessive heart valve sounds Bravman had experienced since his operation did not support a legally cognizable product defect claim under New York law. Bravman v. Baxter Healthcare Corp., 794 F. Supp. 96 (S.D.N.Y.1992).

On appeal, Bravman raises three issues: first, he contends that the district court erred in concluding as a matter of law that the heart valve noise caused him no physical harm, and, therefore, that he has a valid failure-to-warn claim against Baxter; second, he claims that the district court erred in finding that there was no legally cognizable product defect in his Edwards-Duromedics heart valve; and third, he argues that the district court should have allowed a jury to determine whether a design defect existed in light of evidence that alternative valve designs did not create the same level of noise as the Edwards-Duromedics heart valve. We find that the determination whether Bravman sustained a physical injury is a disputed question of fact that should not have been resolved on summary judgment; given the disputed evidence in the record on the question of Baxter's awareness of the noise level, the excessiveness of the noise, and the availability of other mechanical valves with similar functionality and operation at a lower noise level, Bravman's failure to warn claim is viable. The district court, however, correctly determined that as a matter of law Bravman could not recover under either a product defect or design defect theory of liability.

We begin first with Bravman's claim that Baxter had a duty to warn him that the Edwards-Duromedics heart valve produced excessive noise and that the company violated its duty to warn the relevant medical community about this noise. As the manufacturer of the heart valve, Baxter was obliged "to warn of all potential dangers which it knew, or in the exercise of reasonable care should have known, to exist." Lindsay v. Ortho Pharmaceutical Corp., 637 F.2d 87, 91 (2d Cir.1980).

Evidence in the record shows that Baxter knew that its Edwards-Duromedics valve tended to be noisy and that the noise generated concern among doctors and complaints among patients. In response, Baxter argues that all mechanical valves necessarily make noise and that the noise actually serves an important diagnostic role by enabling cardiologists to monitor the valve sounds and determine whether the valves are functioning properly. This argument, however, does not rebut the complaint of excessive noise. In particular, Bravman points to two types of noise problems of which Baxter was aware when Bravman received the Edwards-Duromedics valve.

First, there is evidence that at the time of Bravman's surgery Baxter knew that the Edwards-Duromedics valve made more noise than some other valves that were available on the market at the time of Bravman's surgery. For example, Baxter had received information from the German Heart Center that the Duromedics valve created more noise than did a rival valve.1 The record also includes some evidence that cardiologists familiar with the Edwards-Duromedics valve told Baxter that patients complained about noise from the valve. In 1987, Baxter sponsored a symposium on the Edwards-Duromedics valve at which 22 cardiologists from around the world discussed the valve. At the symposium, at least two cardiologists presented information concerning noise problems with the Edwards-Duromedics valve that doctors had observed in the hospitals where they worked. Both cardiologists suggested that prospective users of the Edwards-Duromedics heart valve should be warned about these potential problems.

Bravman also claims that Baxter knew of a second, apparently infrequent, but more severe noise problem: that the Edwards-Duromedics valve creates extraordinary noise in some patients. For example, there is a suggestion in the record that one patient who had received an Edwards-Duromedics valve attempted to commit suicide because his valve noises were disconcertingly loud--the valve noises could be heard, in a quiet room, from as far as twenty feet.

The district court found that Bravman did not present a legally cognizable failure to warn claim despite these allegations. Although there is no dispute that Baxter had at least some information concerning potential noise problems with the Edwards-Duromedics valve, the court determined that the "tonitruous jangle" of the valve did not inflict any physical harm upon Bravman and that there was no evidence that the failure to warn placed Bravman in any physical danger. In support of its findings, the district court cited another failure to warn case, Gerardi v. Nuclear Utility Services, Inc., 149 Misc. 2d 657, 660, 566 N.Y.S.2d 1002, 1004-05 (Sup.Ct.West.Co.1991), in which a New York court allowed a plaintiff who had not suffered a physical injury to recover for emotional harm only because there was a real danger to the plaintiff's physical safety.

We do not think that the question whether Bravman has suffered a physical harm can be resolved as a matter of law. Unlike the purely psychological terror suffered by the protagonist in Edgar Allan Poe's "The Tell-Tale Heart,"2 Bravman's complaint, that his artificial heart valve creates excessive noise that prevents him from sleeping, among other things, is objectively verifiable. While Bravman's claim is unusual, it is by no means nonsensical. At some level, noise can constitute a cognizable harm.3 If for example, Bravman's heart valve was as loud as a lawn mower, and he (or his surgeon) had been given no warning to this effect, we would expect general agreement that Bravman suffered from a cognizable injury. While it is undisputed that mechanical heart valves need to make some noise, this case presents a material factual dispute as to whether Bravman's heart valve creates noise at a level which could reasonably lead to physical harm.4

Baxter suggests that even if Bravman has a cognizable injury, the evidence precludes a finding that the failure to warn proximately caused this alleged injury. Baxter argues that its duty to warn, if any, was to give information to Dr. Spencer as a "learned intermediary," and that while it had not warned Dr. Spencer, Dr. Spencer testified that he would not have warned Bravman even if he had known about the noise problems.

Although Baxter correctly states the law, Baxter errs in concluding that Dr. Spencer's deposition testimony necessarily resolves this question in its favor. A plaintiff proceeding under a failure-to-warn theory in New York must demonstrate that the failure to warn adequately of the dangers of a product was a proximate cause of his or her injuries. See Glucksman v. Halsey Drug Co., Inc., 160 A.D.2d 305, 307, 553 N.Y.S.2d 724, 726 (1st Dept.1990). In cases such as this one, the duty to warn is owed to the medical community and, more specifically, to the treating physician who is to act as an "informed intermediary" between the manufacturer and patient. Id., 553 N.Y.S.2d at 726. Although the apparently highly qualified Dr. Spencer testified that he would not have passed on the noise information to Bravman even if he had received it, that testimony is insufficient to resolve the proximate cause question. It is up to the trier of fact to determine whether, and the extent to which, Dr. Spencer's testimony on this point is credible, or even if it is, whether it would be found by a jury to be material. In Hoffman-Rattet v. Ortho Pharmaceutical Corp., 135 Misc. 2d 750, 516 N.Y.S.2d 856 (Sup.Ct.N.Y.Co.1987), the court stated that "unless the physician's statement is self-disserving, the issue of the credibility of the physician's affidavit should ordinarily be left for the trier of the facts and should not be resolved by the court on a motion for summary judgment." Id. at 751, 516 N.Y.S.2d at 858. Here, there is nothing conclusive to show that Dr. Spencer's statement was self-disserving, and, therefore, we do not think that the question of proximate cause can be determined properly on summary judgment.

The district court, however, was correct in dismissing Bravman's product and design defect claims. Bravman contends that the district court dismissed these claims because it had erroneously concluded that the Edwards-Duromedics heart valve is an "unavoidably unsafe" product. Unavoidably unsafe products are those that "in the present state of human knowledge, are quite incapable of being made safe for their intended ordinary use." Restatement (Second) of Torts § 402A cmt. k (1979). Under New York law, unavoidably unsafe products "are not deemed defective or unreasonably dangerous so long as they are accompanied by proper directions for use and adequate warnings as to potential side effects." Lindsay, 637 F.2d at 90.

We believe that despite the Edwards-Duromedics valve's noise potential, it may be treated, at least at the time of Bravman's surgery, as an unavoidably unsafe product. See McPheron v. Searle Lab., Inc., 888 F.2d 31, 33 (5th Cir.1989) (listing courts following majority rule that medical devices that must be prescribed and inserted by a physician are unavoidably unsafe products). However, we note that the district court findings on the product and design defect claims did not depend upon this determination.

Bravman's product defect claim fails because he has offered no evidence to suggest that the excessive valve noise he has experienced is the result of a mistake in the manufacturing process. See Voss v. Black & Decker Mfg. Co., 59 N.Y.2d 102, 106-7, 463 N.Y.S.2d 398, 401, 450 N.E.2d 204, 207 (1983). In fact, Bravman's complaint appears to be directed at the general configuration of the Edwards-Duromedics valve and, therefore, is better characterized as a design defect claim.

Design defect claims in New York are determined under a utility/risk analysis--that is, a plaintiff must show that "if the design defect were known at the time of manufacture, a reasonable person would conclude that the utility of the product did not outweigh the risk inherent in marketing a product designed in that manner." Id. at 108, 463 N.Y.S.2d at 402, 450 N.E.2d at 208. Despite the alleged noise problems experienced by some patients, the Edwards-Duromedics heart valve has an impressive record of prolonging the lives of its recipients. Given the success of Edwards-Duromedics heart valves in this regard, we affirm the district court's determination that "it is reasonable to conclude as a matter of law that the utility of their design outweighs isolated instances of excessively noisy valves." Bravman, 794 F. Supp. at 102.

Accordingly, the judgment of the district court is affirmed on Bravman's product and design defect claims and reversed and remanded on his claim that Baxter violated its duty to warn the cardiological community about the alleged noise problems with the Edwards-Duromedics heart valve. Mrs. Bravman's derivative loss of consortium claim is also remanded to the district court.