In this diversity action, plaintiff appeals from orders entered by Judge Pratt in the Eastern District of New York on July 15, 1977 and March 6, 1978 dismissing at the close of plaintiff’s case her medical malpractice and drug products liability suit against defendants Parke-Davis Company, Dow Chemical Company, Dr. Jack Sherman, the County of Nassau, and Meadowbrook Hospital. We affirm the dismissal of the claims brought against the County of Nassau, Meadowbrook Hospital, and Dow Chemical Company. With respect to the claims brought against Parke-Davis Company and Dr. Sherman, we reverse and remand for a new trial.
The infant decedent Mark Ezagui was born to the plaintiff Elaine Ezagui on September 11, 1960. Defendant Sherman vaccinated Mark, at that time healthy in body and mind, with either Quadrigen, manufactured by Parke-Davis, or Compligen, manufactured by Dow, on January 18, 1961. On January 23, 1961, Mark was admitted to Nassau County’s Meadowbrook Hospital with a very high fever later measured at 108° and subsequently diagnosed as postvaccinal encephalopathy (“PVE”), which is known to cause brain damage, convulsive seizures, blindness, deafness, paralysis, mental retardation, and possibly death. Mark was discharged from Meadowbrook Hospital on February 2, 1961. Thereafter, until his death on April 26,1970, Mark was under the care of various medical personnel, among them Dr. Sherman. The diagnosis of PVE resulting from the inoculation was repeatedly confirmed.
Plaintiff Elaine Ezagui, suing for herself and as the administratrix of the estate of Mark Ezagui, initiated this action by service of-a notice of claim on the County of Nassau on January 28, 1969, and by service of summonses and complaints on the other defendants beginning in October, 1969. Basing her claim on New York law, which the parties concede controls this case, plaintiff demanded damages for personal injuries suffered by the decedent and for his wrongful death, and on her own behalf, for loss of services and medical expenses, all allegedly sustained as the result of the January 18, 1961 vaccination. Plaintiff’s amended complaint against drug manufacturers Dow and Parke-Davis alleged (1) failure to warn of a known defect with respect to the drugs Compligen and Quadrigen, (2) breach of warranty of merchantability and fitness for use with respect to the same drugs, and (3) negligence. Plaintiff’s complaint against Dr. Sherman alleged (1) lack of informed consent and (2) negligence. Plaintiff further alleged that these breaches of defendant’s duties to plaintiff proximately caused the personal injuries and wrongful death of Mark Ezagui.
On July 7, 1977, the district court denied plaintiffs’ application to estop collaterally Parke-Davis from denying that Quadrigen was a defective product as determined in Tinnerholm v. Parke-Davis Company, 411 F.2d 48 (2d Cir. 1969) and in Parke-Davis & Co. v. Stromsodt, 411 F.2d 1390 (8th Cir. 1969). On July 13, 1977, the district court dismissed the complaint against defendants County of Nassau and Meadowbrook Hospital on the ground that the claim was untimely. On January 27, 1978, the trial began.
On February 2, 1978, the district court dismissed the complaint against all of the remaining defendants, on the ground that plaintiff had failed to make out a prima facie case against any. With respect to Dr. Sherman, the district court found that plaintiff had not introduced sufficient evidence to support a finding of medical malpractice. With respect to Dow and ParkeDavis, the district court found that plaintiff had not introduced sufficient evidence to support a finding either that Compligen or *731Quadrigen was defective or that one of them had proximately caused injury- to Mark Ezagui. Our review of the record, however, persuades us that plaintiff did make out a prima facie case against ParkeDavis and against Dr. Sherman.
Defendant Parke-Davis
Parke-Davis developed Quadrigen during the 1950’s as a quadruple antigen product, combining diphtheria toxoids, tetanus toxoids, Salk polio vaccine, and pertussis (whooping cough) vaccine. Vaccines confer protection against diseases by introducing antigens into the body which stimulate the production of immunizing antibodies. This process occurs when lymphocytes, cells contained in the lymph glands, absorb the antigens and produce an antitoxin against the particular disease. With some infectious diseases, such as diphtheria and tetanus, it has been possible to isolate the soluble toxin or poison excreted by these bacteria and to inactivate this toxin with formaldehyde, thereby converting the toxin into what is called a toxoid. This toxoid helps immunize the body against disease by stimulating the production of antibodies, but the toxoid will not cause disease because it has lost its poisonous qualities.
By contrast, the bacterial organism which causes pertussis is so complex as to make impossible the isolation and deactivation of the toxin or poison. Since the ingredient in the pertussis bacteria which stimulates the production of protective antibodies has not been isolated, Parke-Davis and other drug companies have manufactured pertussis vaccine consisting of whole pertussis bacteria, treated to reduce their propensity to cause the disease. Because this treatment cannot completely deactivate the relevant toxin, reactions to pertussis vaccine are more frequent than are reactions to other vaccines.
In the early 1940’s, drug manufacturers developed a method for combining pertussis vaccine with diphtheria and tetanus toxoids in a three-way antigen product known as “DTP” and marketed by Parke-Davis under the trade name “Triogen”. This combination allowed one shot to do the work of three and was regarded as an important advance. This three-in-one combination produced no apparent increase in toxicity or reactivity.
In 1953 Dr. Jonas Salk developed a polio vaccine. Following commercial development of the Salk Vaccine, Parke-Davis decided to add the new polio vaccine to its “Triogen” product in order to develop a four-way antigen product, whereby . one shot would protect against polio as well as diphtheria, tetanus, and pertussis. This new product Parke-Davis marketed under the trade name Quadrigen, beginning in July, 1959.
Combining the older Triogen product with the new Salk polio vaccine, however, required a change in preservative which many investigators later believed caused the marked increase in adverse medical reactions experienced with the use of Quadrigen. All vaccines packed in multidose vials require a preservative to maintain their sterility. Prior to the development of the Salk polio vaccine, the universal preservative was merthiolate. Although originally intended to maintain sterility, merthiolate was later shown to act as a stabilizer of the vaccine, decreasing toxity but maintaining potency. Merthiolate, however, adversely affected the polio vaccine. Accordingly, Parke-Davis selected a different preservative for use in Quadrigen. This preservative was benzethonium chloride, or Phemerol, which was Parke-Davis’ trade name for this product. Later research, however, indicated that use of Phemerol caused certain endotoxins in the pertussis vaccine to leak out from the bacterial cell into the fluid which was injected. One of these endotoxins, the lipopolysaccharide, was known to cause a fever which could lead to convulsions and brain damage, as occurred in this case. From the time when they first began to investigate the marked increase in adverse reactions reported by doctors using Quadrigen, until very recently, Parke-Davis research personnel have been on record as believing that the leakage of these endotoxins was responsible for the measured in*732crease in adverse medical reactions associated with Quadrigen, an increase which finally led to the withdrawal of Quadrigen from the market in November, 1962. (The return in 1962 to the older three-in-one product, administered with a separate polio vaccine, reduced the incidence of adverse medical reactions to their pre-Quadrigen levels).
Two earlier cases explicitly relied upon this “Phemerol causes leakage” theory in affirming judgments against Parke-Davis involving Quadrigen vaccinations which allegedly caused severe personal injuries to the infants vaccinated. See Tinnerholm v. Parke-Davis & Co., 411 F.2d 48 (2d Cir. 1969), Parke-Davis & Co. v. Stromsodt, 411 F.2d 1390 (8th Cir. 1969). These cases both found a defect in Quadrigen because of this leakage problem and found that this defect proximately caused injuries to the plaintiffs involved. Plaintiff in this case argued before trial that these cases should collaterally estop Parke-Davis from denying that Quadrigen was defective. The district court, however, denied plaintiff’s application for collateral estoppel on the ground that new scientific evidence cast doubt on the “Phemerol causes leakage” theory. ,In so denying plaintiff's application, the district court clearly followed the requirements of New York law. In Vincent v. Thompson, 50 A.D.2d 211, 377 N.Y.S.2d 118 (2d Dept. 1975), another Quadrigen case, the Appellate Division denied collateral estoppel based on the Tinnerholm case partially because of the purported “discovery” of the same new scientific evidence at issue here.
The Appellate'Division’s ruling, properly construed, disallows collateral estoppel as to chemical defect whenever the plaintiff relies solely on cases where the “Phemerol causes leakage” theory was an essential step in the court’s finding of chemical defect. Accordingly, plaintiff cannot invoke collateral estoppel as to chemical defect on the basis of Stromsodt, supra, because that case also relied upon the “Phemerol causes leakage” theory in affirming a finding of chemical defect. Nor can plaintiff rely upon the collateral estoppel order affirmed in Grant v. Parke-Davis & Co., 544 F.2d 521 (7th Cir. 1976) (reported in full in Commerce Clearing House, Products Liability Reporter, H 7848), inasmuch as the collateral estoppel order in that case was grounded on the findings of chemical defect in Tinnerholm, supra, and Stromsodt, supra.
Plaintiff, however, sought collateral estoppel not only as to chemical defect, but also as to the inadequacy of the warnings which accompanied the product, relying on Stromsodt, supra and other Quadrigen cases. In Stromsodt, the Eighth Circuit affirmed a district court finding that the package inserts accompanying Quadrigen were inadequate in light of the known risks of harm associated with normal use of the product, risks which had been repeatedly brought to Parke-Davis’ attention. The package insert at issue in both Stromsodt and in this case, as well as in the other Quadrigen cases (where they were also found to be inadequate), states that the incidence of “local and systemic reactions following the administration of Quadrigen are usually mild . . . [and] is usually no greater than is normally experienced with trivalent vaccine.” According to Judge Mehaffy of the Eighth Circuit:
“To tell the practitioner who had successfully used Triogen that reactions from Quadrigen are usually no greater than normally experienced with Triogen simply does not comport with the record facts in this case and the knowledge of the medical experts associated with Parke-Davis and others. Parke-Davis had knowledge that many doctors would not use the product, that other doctors reported to it prior to the inoculation of plaintiff that they were having more severe reactions to the use of Quadrigen than with Triogen, and reports of febrile or higher fever reaction.” 411 F.2d at 1400.
Thus the Eighth Circuit explicitly held that the package insert was inadequate in light of the frequency and severity of adverse reactions associated with Quadrigen and reported to Parke-Davis. This holding in no way depended on the “Phemerol causes *733leakage” theory. Accordingly, Vincent v. Thompson, supra, does not bar reliance on Stromsodt to estop collaterally Parke-Davis from denying that the package inserts supplied with Quadrigen were inadequate in that they failed to apprise doctors administering the vaccine of the known hazards attendant upon such use. Indeed, the Vincent court explicitly approved use of Stromsodt for collateral estoppel purposes independent of the “Phemerol causes leakage” theory. 377 N.Y.S.2d at 128. True, the holding with respect to the inadequacy of the package inserts was an alternative holding in Stromsodt. But even “if a court decides a case on two grounds, each is a good estoppel.” Irving National Bank v. Law, 10 F.2d 721, 724 (2d Cir. 1926) (L. Hand, J.) Stromsodt was not a case where the appellate court explicitly affirmed upon one ground only. Cf. Vincent v. Thompson, 377 N.Y.S.2d 121, n.2 (refusing to apply collateral estoppel to finding of district court upon which appellate court did not rely). Accordingly, on retrial the plaintiff will be entitled to an order collaterally es-topping Parke-Davis from denying the inadequacy of the warnings contained in the Quadrigen package inserts.
New York cases have consistently held that a manufacturer’s knowledge of special risks of harm attendant upon normal use of his product imposes a duty upon the manufacturer to warn adequately those using his product of those risks. See, e. g., Donigi v. American Cyanamid Co., 43 N.Y.2d 935, 403 N.Y.S.2d 894, 374 N.E.2d 1245 (1978), aff’g 57 A.D.2d 760, 394 N.Y. S.2d 422 (1st Dept. 1977). The knowledge of risk available to Parke-Davis in 1960, as shown by the record below, was more than sufficient to impose upon Parke-Davis a duty to warn of that risk. Although failure to warn sounds in negligence, New York courts have also held that inadequate warnings will render a product defective for purposes of warranty and strict products liability. See Alfieri v. Cabot Corp., 17 A.D.2d 455, 235 N.Y.S.2d 753 (1962). Accordingly, the failure to provide adequate warnings here establishes a prima facie case of product defect. On retrial, plaintiff will have the opportunity to show that this defect, the failure to warn adequately, proximately caused the damages suffered by plaintiff.
The failure to provide adequate warnings of known risks associated with normal use of the product, moreover, also violates the Federal Food, Drug and Cosmetics Act, 21 U.S.C. § 352, and the New York Education Law § 6815, both of which state that “A drug or device shall be deemed to be misbranded ... if the labeling is false or misleading in any particular . . The term ‘labeling’ means all labels and other written, printed, or graphic matter . . . accompanying such article.” Since Mark Ezagui was clearly within the class of those people who are the intended beneficiaries of these statutes, under New York law plaintiff is entitled to an instruction that the violation of these statutes shown here is negligence per se. Plaintiff need then show only that this violation proximately caused the injuries and death of Mark Ezagui to make out a prima facie ease of negligence against Parke-Davis.
Even apart from a collateral estoppel order, we believe that the district court erred in concluding that plaintiff’s case against Parke-Davis should be dismissed for failure to make out a prima facie case as to either (1) product defect, or (2) proximate causation. Our review of the record persuades us that the plaintiff introduced enough evidence to go to the jury on both issues.
First, the evidence in the record tending to show that Quadrigen was unreasonably dangerous and therefore defective was more than sufficient. Indeed, the New York Court of Appeals in Stromsodt and the district court in Tinnerholm both concluded that Parke-Davis knew of the defective nature of Quadrigen because of the complaints it was receiving from doctors, many of whom refused to use the drug, and because of the decidedly adverse testing results brought to its attention. Perusal of the record in this case strongly suggests *734that Parke-Davis marketed Quadrigen although fully cognizant that commercial sale of the drug would impose an unreasonable risk of harm on the many infants vaccinated. Evidence in the record includes, inter alia, the following:
Exhibit 12: Report of Dr. Wedgwood to Parke-Davis November 19, 1959, finding-that “5% of the children had quite severe . . . reactions.”
Exhibit' 19: Report to Parke-Davis from Dr. Workman of HEW, dated March 10, 1959, indicating that he was “very much concerned” about an “unstable product”.
Exhibit 20: Letter from Parke-Davis to Dr. Workman, dated March 16, 1960, indicating that one lot of Quadrigen was “substandard” and another “rejected because of high potency.”
Exhibit 22: Letter from Dr. Reichelderfer, of the District of Columbia General Hospital, to Parke-Davis, dated April 8, 1959, indicating that he was “reluctant to continue further trial of this product because I believe there are fewer reactions when the DPT and polio antigens are given separately.”
Exhibit 37: Letter from Dr. Workman to Parke-Davis, dated October 20, 1960, indicating that Quadrigen lots failed seven out of ten tests for freedom from toxicity. Attached letter from R. W. Kolb to Parke-Davis, dated September 29, 1960, indicating that when Quadrigen was injected into thirty mice, four died.
Exhibit 40: Letter from Detroit City Department of Health to Parke-Davis, dated November 17, 1960, indicating “wide differences in pertussis potency from bottle to bottle from the same lot.” The committee of doctors were of the “unanimous opinion” that much more testing of Quadrigen was needed and could not “STRESS TOO MUCH TO THE OFFICIALS OF PARKEDAVIS & COMPANY THE IMPORTANCE AND URGENCY OF MAKING A MAXIMUM EFFORT TO PROVIDE DR. VOLK WITH A NEW SUPPLY OF QUADRIGEN” for further testing (all emphasis in original).
Exhibit 76: Parke-Davis internal memorandum, dated November 6, 1959, noting that “a vial of Quadrigen had been returned . . . with a contaminant growing visibly on the surface thus confirming other complaints along the same line several five gallon lots showing contamination . . . . It was decided that an immediate solution to the contamination problem could not be obtained by any major change in the process, since too much time would be involved.”
Exhibit 79: Parke-Davis internal memorandum, dated August 11, 1960, suggesting conferences with doctors who were “complaining seriously about this problem . . . .” Attached ParkeDavis internal memorandum, dated August 9, 1960, indicating that “Complaints have been noted generally in recent weeks . . . . Efforts to enlist the services of the same practitioners will be made. There may be some difficulty in certain patients where unsavory reactions have been experienced. . . It is difficult to believe that reactions of the magnitude now being reported could have been overlooked in the original extensive field trials with this product.”
Exhibit 81: Quadrigen fails seven of seven toxicity tests.
Exhibits 83-84: Parke-Davis inter-office memo, dated August 22, 1960, confirms high incidence of reactions to Quadrigen: 74% experienced fever and 38% had “significant systemic reactions.”
Exhibit 86: Parke-Davis inter-office memo, dated September 27, 1960, reports that Massachusetts Health Department, by letter dated September 15, 1960, advised all Massachusetts doctors that each of the quadruple antigen products then on the market, including Quadrigen, had failed to meet minimal standards. All physicians were advised “not to use them.”
*735Exhibit 88: State University of Iowa study, dated October 24, 1960, reports adverse reactions in 66.7% of the children inoculated with Quadrigen, and severe adverse reactions in 19.0%.
Exhibit 96: Parke-Davis inter-office memo, dated April 15, 1966, discloses letter from Dr. Chinnock, dated March 13, 1959, reporting vomiting, convulsions, and encephalopathy in a five month old infant following a Quadrigen inoculation. The child later died from the effects of the encephalopathy.
Exhibit 105: Letter from Dr. Jennings to Parke-Davis, dated August 21, 1959, reporting “really bothersome” reactions following Quadrigen inoculations.
Exhibit 110: Letter from Dr. Wishropp to Parke-Davis, dated August 11, 1960, reporting that “we have had some rather severe febrile reactions after giving Quadrigen injections . . . Dr. Jennings ... is not giving Quadrigen any longer because of the frequency of reactions, but is giving another product instead.” This letter typified the response of many doctors. (See, e. g., Exhibit 146a, a Parke-Davis inter-office memo reporting that two doctors, one of whom used an average twenty vials of Quadrigen monthly, had severe reactions in every instance.)
A jury could reasonably conclude from this evidence that the pertussis vaccine is capable of causing encephalopathy such as that experienced by Mark Ezagui, and that the combination in Quadrigen of the pertussis vaccine with other chemicals materially increased this risk.1
Plaintiff further bolstered her ease as to the existence of a chemical defect by:
1. reading into the record, according to the terms of the pre-trial order, Dr. Carson’s testimony which affirmed the defective nature of Quadrigen;
2. reading into the record portions of Dr. McLean’s testimony in the Tinnerholm case, supra, with respect to the causal chain between the pertussis vaccine, fever, convulsions, and encephalopathy and Parke-Davis’ knowledge of same;
3. reading into the record portions of the testimony of Dr. Dethmers, a ParkeDavis attorney, in the Tinnerholm case, supra, with respect to the reports received by Parke-Davis from doctors using Quadrigen; and
4. reading into the record portions of Dr. Gajewski’s testimony in the Tinnerholm case, supra, tending to establish the pertussis component of Quadrigen as the defective part.
Second, having introduced sufficient proof to go to the jury on the issue of chemical defect, the plaintiff also introduced sufficient proof, in addition to the evidence described above, to go to the jury on the issue of proximate causation as to this particular plaintiff. Dr. Sherman testified that there was only one inoculation given, that in his opinion that inoculation caused the PVE, that he did not know of any other possible cause, and that there was no hyperimmune or allergic reaction. Dr. Kaplan corroborated Dr. Sherman’s conclusion that the one inoculation given caused the PVE. Later diagnoses made by the Jewish Chronic Disease Hospital, by the Willowbrook Hospital, and by the Suffolk State Hospital also linked Mark Ezagui’s PVE to the January 18, 1961 inoculation. There was no evidence or testimony tending to prove another cause. Finally, plaintiff *736read into the record portions of Dr. McLean’s testimony in the Tinnerholm case, supra, to the effect that the risk of an encephalopathic reaction caused by the pertussis vaccine subsequent to vaccination was close to zero with the older Triogen product, but much higher with Quadrigen. This testimony further corroborated Dr. Sherman’s conclusion that Mark Ezagui’s reaction was not allergic or hyperimmune and therefore to some degree unavoidable whenever the pertussis vaccine is administered. With this evidence before it, a jury could reasonably find that Quadrigen suffered from a chemical defect when marketed and that this chemical defect proximately cause the death of Mark Ezagui. Cf. Tinnerholm v. Parke-Davis Company, 411 F.2d 48, 51-53 (2d Cir. 1969).
We also find that the plaintiff introduced enough evidence to go to the jury on the theory that the inadequacy of the warnings accompanying the product rendered the product defective and that this defect proximately caused the death of Mark Ezagui. The gravamen of this claim is that, had the warnings been accurate and complete, either Mrs. Ezagui would not have consented to use of the drug, or Dr. Sherman would not have used it. Both Dr. Sherman and Mrs. Ezagui testified at trial that, had they known the true dimensions of the problems associated with Quadrigen, they would not have allowed that drug to be used instead of the older Triogen product. Although interested testimony of this nature should be received with caution, its weight is for the jury to determine. See Parke-Davis & Co. v. Stromsodt, 411 F.2d 1390 (8th Cir. 1969); Tinnerholm v. Parke-Davis & Co., 285 F.Supp. 432 (S.D.N.Y. 1968).
Finally, with respect to plaintiff’s negligence cause of action, we find that the evidence described above in the context of plaintiff’s causes of action based on warranty and strict products liability, also made out a prima facie case as to the negligence of Parke-Davis. Parke-Davis, as a matter of law, owed plaintiff a duty to exercise reasonable care in the design and in the decision to market Quadrigen. The evidence introduced at trial could support a jury finding that Parke-Davis breached this duty by marketing Quadrigen despite its knowledge that Quadrigen caused a significant increase in the proportion of adverse medical reactions, compared with the results obtained with the older Triogen product.
Defendant Dow Chemical Corporation
Plaintiff introduced evidence at trial relevant to defendant Dow on the same theories advanced against defendant ParkeDavis. Similar in quality but not in extent to the evidence introduced against ParkeDavis, the evidence introduced tended to show that the Dow drug, Compligen, was similar in design and effect to the ParkeDavis drug, Quadrigen, and was equally capable of causing the adverse reactions suffered by Mark Ezagui. However, inasmuch as this evidence consisted primarily of two of Dow’s answers to plaintiff’s interrogatories, and of the alleged similarities between the Dow drug and the Parke-Davis drug, we find that the limited evidence introduced here was insufficient to create a jury issue as to the defective nature of the Dow drug. Plaintiff was required to do much more than merely allege that the Dow drug was similar to another allegedly defective drug, whose defective nature was itself an issue for trial.
But even assuming arguendo that plaintiff introduced enough evidence to get to the jury on the question of Compligen’s defective nature, whether it be a chemical defect or a failure to adequately warn of known risks, plaintiff failed to introduce enough evidence to get to the jury on the issue of proximate cause. Although the evidence is conflicting, most of it — including Dr. Sherman’s first notations, his first conversations with Mrs. Ezagui, and contemporaneous hospital records — strongly suggests that Quadrigen and not Compligen was used in the inoculation given to Mark Ezagui January 18, 1961. Where one of two drugs, but not both, has caused a plaintiff injury, and the clear preponderance of *737the evidence points to one drug and not to the other, it would be unreasonable for a jury to conclude that the other drug, in this case Compligen, caused plaintiff’s injuries, especially where the proof of the other drug’s defective nature was as uncertain as it is here. This is not a case where the plaintiff has clearly shown negligence on the part of two defendants, and has introduced evidence showing an equal probability that either of the two defendants caused the injuries suffered, such that a court would be warranted in shifting to the defendants the burden of exonerating themselves if they are not to share liability. Cf. Summers v. Tice, 33 Cal.2d 80, 199 P.2d 1 (1948) (plaintiff proved that two defendants had negligently fired their rifles in his direction, but could not prove whose shot had actually hit him; burden of exoneration shifted to defendants). Accordingly, the district court properly dismissed the complaint brought against defendant Dow.
Defendant Dr. Sherman
The district court dismissed the complaint against Dr. Sherman on the ground that plaintiff had failed to introduce sufficient expert testimony to show that Dr. Sherman had departed from accepted medical practice and that such departure proximately caused the injuries suffered by Mark Ezagui. In fact, plaintiff did not rely upon any of her own experts in her claim against Dr. Sherman, but instead relied upon cross-examination of Dr. Sherman in order to establish the required standard of care. Although § 4401-a of the New York CPLR requires expert testimony “as to any cause of action for medical malpractice based solely on lack of informed consent”, this requirement applies only to causes of action arising after July 1, 1975 and in any event applies only to causes of action based solely on lack of informed consent. Accordingly, while sound litigation strategy would normally include additional expert medical testimony, this plaintiff was not subject to dismissal for failure to introduce such testimony. See generally, McDermott v. Manhattan Eye and Ear Hospital, 15 N.Y.2d 20, 255 N.Y.S.2d 65, 203 N.E.2d 469 (1964).
In order to get to the jury on this cause of action, plaintiff had to introduce enough evidence to support a jury finding that Dr. Sherman departed from accepted medical practice and that this departure proximately caused the injuries suffered by Mark Ezagui. We hold that plaintiff introduced enough evidence to go to the jury on any of three theories, each of which would justify a finding of negligent treatment and proximate causation. First, there was sufficient evidence to support a jury finding that Dr. Sherman had failed to comprehend and act in light of the hazards attendant upon the use of Quadrigen as reported in the medical literature and in the package inserts. Second, there was sufficient evidence to support a jury finding that Dr. Sherman was negligent in vaccinating Mark Ezagui despite the existence of a mucoid condition or runny nose, when the package inserts explicitly warned against vaccinating a patient suffering from such a condition. Third, there was sufficient evidence to support a jury finding that Dr. Sherman did not adequately disclose the hazards associated with Quadrigen which he was charged with knowing and disclosing and that there was therefore a lack of informed consent.
Defendant County of Nassau and Meadow-brook Hospital
The district court dismissed plaintiff’s claims against the County of Nassau and Meadowbrook Hospital on the ground that plaintiff had failed to serve notice of her claim upon the county within ninety days after the cause of action arose, as required by New York General Municipal Law § 50-e. The district court held, and we agree, that subsequent treatment at Meadowbrook Hospital in 1968 did not establish a continuous course of treatment beginning in 1961 and ending in 1968 such as would make plaintiff’s service of notice in 1969 timely under New York law. See, e. g., Davis v. City of New York, 38 N.Y.2d 257, 379 N.Y.S.2d 721, 342 N.E.2d 516 (1975). Accordingly, we affirm the dismissal of the complaint against the County of Nassau and Meadowbrook Hospital.
*738Having found that plaintiff introduced sufficient evidence at trial to withstand a motion to dismiss, we need not deal with the propriety of the trial court’s enforcement of its pretrial order in a manner which ultimately prevented the plaintiff from introducing the expert testimony of Dr. Carson, especially inasmuch as such a ruling is not likely to recur on retrial.
We affirm the orders of the district court appealed from insofar as they dismissed the claims brought against Dow Chemical Company, the County of Nassau, and Meadow-brook Hospital. We reverse the orders insofar as they dismissed the claims against Parke-Davis Company and Dr. Sherman and remand for further proceedings not inconsistent with this opinion.
. In addition, we note that the district court excluded from evidence reports of adverse medical reactions received by Parke-Davis subsequent to the January 18, 1961 vaccination of Mark Ezagui. While such later reports were not admissible to prove Parke-Davis’ knowledge of defect at the time of vaccination, these later reports should have been admitted to prove the existence of a defect, just as subsequent accident reports are admissible to prove the cause of a car accident. Accordingly, on retrial, plaintiff will be entitled to introduce post-1961 letters and memoranda. Parke-Davis will be entitled to an instruction that these post-1961 letters and memoranda are not to be relied upon to show Parke-Davis’ knowledge of defect at the time of vaccination, except to the degree that they disclose facts which Parke-Davis should have, through the exercise of reasonable diligence, known or discovered even before such letters or memoranda were received.