I respectfully dissent. My brethren view too narrowly the scope of the duty imposed on McNeil to provide “adequate warnings” with inherently unsafe prescription drugs. Settled law holds McNeil liable for the harm the drugs actually and proximately cause if no “learned intermediary” intervenes or if McNeil knows that the drugs will be utilized in a widespread practice of illegal and dangerous medical activity. McNeil cannot immunize itself from liability simply by providing “adequately” written warnings that as a practical matter will not effect prophylaxis. Rather, such drugs with accompanying but inadequate warnings are defective products unreasonably dangerous to the public. See Restatement (Second) of Torts, Section 402A (1965); State Stove Mfg. Co. v. Hodges, 189 So.2d 113, 118 (Miss.1966), cert. denied sub nom. Yates v. Hodges, 386 U.S. 912, 87 S.Ct. 860, 17 L.Ed.2d 784 (1967) (incorporating Restatement § 402A into Mississippi law); Gordon v. Niagara Machine & Tool Works, 574 F.2d 1182, 1186-95 (5th Cir. 1978) (under Mississippi law, intervening negligence does not insulate a manufacturer from liability for failure to warn ultimate users whom it has reason to know will not receive warnings); cf. Incollingo v. Ewing, 444 Pa. 263, 444 Pa. 299, 282 A.2d 206, 215 (1971) (interpreting inadequate warnings under Restatement § 402A; “[o]n th[e] basis [of evidence] the trial court instructed the jury that if it found that Parke, Davis was on notice that the drug was being used indiscriminately, and yet failed to try to restrict its use to proper situations, then the company could be found negligent. We think the charge was proper. When a required warning is retained unchanged in the face of being widely disregarded, and the supplier knows or has reason to know of such wide disregard, a jury may be permitted to find the warning insufficient.”); Salmon v. Parke, Davis & Co., 520 F.2d 1359, 1362 (4th Cir. 1975) (“when a manufacturer does not change a warning it knows is widely disregarded, a jury may infer that the warning is insufficient.”) (citing Incollingo).
I.
At trial, Swayze introduced the expert testimony of two eminent anesthesiologists, familiar with Mississippi medical practices in 1978, and the testimony of the nurse who administered the anesthetic to Michael Swayze. This testimony was more than sufficient to establish that McNeil knew or should have known — had actual or constructive knowledge — of the widespread and common practice in many Mississippi hospitals that nurses not only were administering fentanyl improperly, but also illegally were choosing the anesthetic, determining the dosage, and administering the drug. Attending surgeons totally failed to provide any supervision regarding anesthesia, and no anesthesiologists were present. See Appendix 1. These facts are abundantly sufficient to convince a reasonable juror that McNeil knew, or should have known, that fentanyl would be prescribed without active or passive supervision by any doctor. Thus, the majority’s and trial court’s observations that doctors acted to direct or to supervise nurse anesthetists and that McNeil could not know the level of supervision, majority at 471; Rec. Vol. 2, District *473Court Opinion, at 14, directly contradict the clear record at trial.1
It is undisputed that such activity violates Mississippi law and sound medical practice, regardless of the training of the C.R.N.A.S. The nurse here and nurses in many Mississippi hospitals not only administered narcotic, but more importantly chose, prescribed, and monitored the anesthesia solely on their own. The choice of anesthetic agent and of dosage, and the actual administration of anesthetic were all proximate medical causes of the harm. See Appendix 2.
II.
The case law makes clear, and the majority concedes, that a duty to warn consumers attaches if no physician adopts the role of “learned intermediary” in prescribing the medicine. See Reyes v. Wyeth Laboratories, 498 F.2d 1264 (5th Cir.1974), cert. denied, 419 U.S. 1096, 95 S.Ct. 687, 42 L.Ed.2d 688 (1974); Givens v. Lederle, 556 F.2d 1341 (5th Cir.1977); Davis v. Wyeth Laboratories, Inc., 399 F.2d 121 (9th Cir. 1968). Despite the majority’s reasoning, the mere physical presence of a physician in an operating room does not in any way distinguish Reyes, Givens, or Davis. The failure of a doctor to exercise his fiduciary duty to supervise care and to prevent unqualified persons (legally and by training) from prescribing fentanyl cannot be consistent with performing a “learned intermediary” role. See Brazzell v. United States, 788 F.2d 1352, 1358 (8th Cir.1986) (on rehearing) (doctor assumed the role of distributor, rather than learned intermediary, invoking manufacturer’s duty to warn ultimate consumer); Reyes, 498 F.2d at 1277 (drug administered “without the sort of individualized medical balancing of the risks to the vaccinee that is contemplated by the prescription drug exception.”); Givens, 556 F.2d at 1345; Williams v. Lederle Laboratories, 591 F.Supp. 381, 389 (S.D. Ohio 1984) (“The appropriate test ... is whether the drug is commonly administered without individualized bal-ancing____”). But see Stanback v. Parke, Davis & Co., 657 F.2d 642, 647 (4th Cir. 1981).
An operating room can become the legal equivalent of a flu clinic when treated as such by doctors. The majority thus errs in implying that the doctors who supervised or performed surgery also supervised or performed anesthesia. Majority at 468. It is common knowledge that doctors and medicine have become highly specialized and compartmentalized; surgeons are not licensed as anesthesiologists. McNeil’s detail men provided information to nurses because surgeons played no role in anesthesia. The nurses illegally and solely chose, prescribed, and administered anesthesia as much as if a doctor never entered the room. Denial of liability because of a surgeon’s mere physical presence exalts form over substance and artificially defers to professionalism that does not exist in this case.
III.
I also disagree with the majority’s implication that subjecting McNeil to a broader scope of duty would in any way infringe or intrude upon doctor-patient relationships. Majority at 471; see Pharmaceutical Mfrs. Ass’n v. Food & Drug Admin., 484 F.Supp. 1179,1186-88 (D.Del.1980) (upholding F.D.A. regulation requiring doctors and pharmacists to provide package warning inserts to patients receiving estrogen and rejecting that the regulation would intrude upon the “physician’s right to exercise professional judgment”). The facts here establish that many surgeons are in total breach of their duty of supervision and care in regard to anesthesia. See Hall v. Hilbun, 466 So.2d 856 (Miss.1985) (Mississippi standard of medical care); Pharr v. Anderson, 436 So.2d 1357, 1361 (Miss.1983) (negligent failure to treat). Because these surgeons rely entirely upon the C.R.N.A.s, no meaningful or real doctor-patient relationship exists in *474regard to anesthesia. Thus, McNeil could not infringe upon these surgeons’ relationship to the patients, because the doctors themselves have confined the relationship to surgical matters. See Carlsen v. Javu-rek, 526 F.2d 202, 206-07 (8th Cir.1975) (mere presence of one of two doctors at surgery, even when that doctor shrugged his shoulders in response to nurse’s decision to administer anesthesia, which nurse interpreted as partial approval, was insufficient to create participatory role in anesthesia which would permit negligence claim against the doctor for breach of duty).
Even were a meaningful doctor-patient relationship to exist, as the majority intones, intrusion on such a relationship is not a valid or recognized legal principle for shielding doctors from liability. On the contrary, the law constantly intrudes on medical relations through the law of malpractice, forcing changes in medical care provision, insuring compliance with legal and medical duties, and subjecting doctors to liability for their inadequate relations with patients. Similarly, manufacturers constantly alter medical standards of care both through generation of new technologies and through marketing and market mechanisms. In such a context, the intervention of a responsible party to force compliance with the law and to protect an unwitting and vulnerable public from the doctors who are supposed to but do not provide adequate care is not merely acceptable, but absolutely necessary.
The failure of nurses, doctors, hospitals, and Mississippi authorities to insure a meaningful doctor-patient relationship does not and should not inure to the benefit of McNeil. The majority and trial court (implicitly), however, find that the manufacturer has no duty to pressure the medical community to properly supervise C.R.N.A.S. Majority at 471; Rec. Vol. 2, District Court Opinion, at 13. I believe, however, that both law and morality require McNeil to monitor and ensure that the products it manufactures and markets are not generally used in an unreasonably dangerous fashion. If this entails some pressure to avoid deaths that McNeil can prevent, McNeil is required to do no less. See Jackson v. Johns-Manville Sales Corp., 727 F.2d 506, 515 (5th Cir.1984), aff'd on rehearing and reinstated in part, 750 F.2d 1314 (5th Cir. 1985) (en banc), questions certified to Mississippi Supreme Court, 757 F.2d 614 (5th Cir.1985) (en banc), certificate for questions dismissed, 469 So.2d 99 (Miss.1985) (en banc), aff'd on certified issues, 781 F.2d 394 (5th Cir.1986) (en banc), cert. denied, — U.S. -, 106 S.Ct. 3339, 92 L.Ed.2d 743 (1986). McNeil cannot simply disclaim any duty when reasonable alternative actions exist. At the very least, the jury should have had the opportunity to pass upon the reasonableness of the manufacturer’s activities.
The plaintiff has suggested numerous actions that McNeil reasonably could have taken to avoid the risks. Thus, the failure to provide adequate warnings or to take any action to change the circumstances constitutes sufficient evidence from which a reasonable juror could infer that McNeil’s manufacture and marketing of the drug were the proximate cause of Michael Swayze’s injuries, and that the doctor, nurse, and/or hospital were not the sole proximate cause as found by the trial court. Rec. Vol. 2, District Court Opinion, at 15; see Jackson, 727 F.2d at 523 (citing New Orleans & N.E.R. Co. v. Burge, 2 So.2d 825, 826 (Miss. 1941)); Van Buskirk v. Carey Canadian Mines, Ltd., 760 F.2d 481, 493 (3rd Cir.1985).
IV.
The general law of products liability from which the learned intermediary law of prescription drugs derives also requires reversal. The unreasonably dangerous determination can be made in myriad ways; Mississippi adopted a “flexible” standard for the jury to determine “unreasonableness” in Hamilton Fixture Co. v. Anderson, 285 So.2d 744 (Miss.1973). See generally J. Beasley, Products Inability and the Unreasonably Dangerous Requirement (1981). It is settled, moreover, that no warning can be considered adequate if the *475warning foreseeably will not be communicated, and thus cannot prevent the concomitant harm. As Judge Wisdom succinctly stated in Borel v. Fibreboard Paper Products Corp., 493 F.2d 1076, 1091, 1092 & n. 29 (5th Cir.1973), cert. denied, 419 U.S. 869, 95 S.Ct. 127, 42 L.Ed.2d 107 (1974) (emphasis in original):
We agree with the Restatement: a seller may be liable to the ultimate consumer or user for the failure to give adequate warnings. The seller’s warning must be reasonably calculated to reach such persons and the presence of an intermediary party will not by itself relieve the seller of this duty. Sterling Drug Co. v. Cornish, 8 Cir.1966, 370 F.2d 82; Yarrow v. Sterling Drug, 8 Cir.1969, 408 F.2d 978; Noel, Products Defective Because of Inadequate Warnings, 23 S.W.L.J. 256 (1969)____ Where a product is extremely dangerous, however, the seller or manufacturer cannot rely upon the intermediate party to transmit a warning. See Prosser, [The Fall of the Citadel, (Strict Liability to the Consumer), 50 Minn.L. Rev. 791, 826-828 (1966)]....
We need not decide today whether fenta-nyl, an “inherently unsafe” product, is “extremely dangerous” or presents an “unreasonable” risk, because the relevant circumstances reveal that McNeil’s reliance on an imaginary intermediary was inadequate to provide warning.2
In this context, the reasoning of the district court is plainly in error. The district court recognize[d] the argument of the Plaintiff that detail men were another method of warning or disseminating information about Fentanyl, and it is stipulated that they visited only with the anesthesiologists and the anesthetists. However, it’s not McNeil’s fault that the surgeons are not fulfilling all of their obligations.
If McNeil had taken no actions to educate the anesthetists along with the an-ethesiologists about this drug, having knowledge that the supervision by surgeons was not what it should have been, we may have an allegation of failure to adequately get the warning to the person who actually needed it, that is [to] the person who would be in the best position to weigh the benefits against the risks, the anesthetists.
But McNeil did both here, in this Court’s opinion. They made this package insert available; they made it accessible. The P.D.R. was available; it was accessible. And this court does not find that this case falls within the rule of Gordon v. Niag[a]ra Machine & Tool Works, 506 F.2d 419 (5th Cir.1974), after remand, 574 F.2d 1182 (5th Cir.1978), reh. denied, 578 F.2d 871 (1978), that they knew on all occasions that the physicians were not knowledgeable of the dangers of this particular drug and knew that they were not fully exercising their supervision.
You cannot generalize in this area. Dr. Attix ... had worked with Mr. Collier for a number of years. Collier had been trained at Mayo Clinic; he had ad*476ministered this drug thousands of times; and there had never been any serious consequences or results from the use of this drug during surgical procedures.
Rec. Vol. 2, District Court Opinion, at 12-13.
First, the testimony established a generalized pattern and commonly known practice in many hospitals that surgeons, the only doctors present, routinely did not supervise anesthesia. If the pattern and practice of illegal and incompetent prescription were known, the harm was foreseeable. Further, Niagara does not stand for the proposition cited; Niagara only holds that foreseeable intervening negligence does not shield manufacturers from liability. Thus, if the manner of harm were foreseeable, it does not matter whether McNeil knew specifically of Dr. Attix and Mr. Collier, or of “all occasions.”
Second, it is precisely because C.R.N.A.s are not qualified to prescribe, regardless of their training or the level of their understanding, that no warnings to the nurses can be considered adequate. It was undisputed at trial that C.R.N.A.s are not in any position to assess the “consequences” to the patients from the dangers of the drug. They are legally and technically unqualified to weigh the benefits against the risks, regardless of whether they fully understand the risks of respiratory complication from fentanyl. Mississippi law and the undisputed medical testimony currently require that the choice of an anesthetic agent and the prescription of an appropriate dosage be performed by a medical doctor. Hence, educating nurses about the risks cannot remedy any failure to provide adequate warnings to the doctors. Cf. Restatement (Second) of Torts, § 390 (1965).
Third, the only warning given to the surgeons, often the only doctors present, was through publication in the Physician’s Desk Reference (P.D.R.). The parties stipulated that sales were made only to anesthetists and anesthesiologists, and testimony established that surgeons likely did not see the warnings accompanying the drug itself. A reasonable juror could conclude that P.D.R. warnings were insufficient. But here it does not matter whether the surgeons ever received or understood the P.D.R. warnings, because McNeil arguably knew that surgeons routinely did not supervise prescription. If surgeons did not supervise prescription, warnings to them could not be passed on to the patient or to anyone who could use the warning; if surgeons would not play any role in anesthesia, no warning to them could ever have any effect.
Thus, the majority improperly relies upon the presumption of Comment j to Section 402A of the Restatement, that McNeil’s warnings would be heeded. Cf. Restatement (Second) of Torts, § 388, Comment g (1965) (presumption of proper use “[sjave in exceptional circumstances, as where ... the person to whom it is supplied is obviously likely to misuse it”); Doss v. Apache Powder Co., 430 F.2d 1317, 1321 (5th Cir.1970) (applying Restatement § 388). We have recently reiterated that the duty to warn can be satisfied through intermediaries only “ ‘by proving that [the] intermediary was ... capable of passing on a warning.’ ... [The] warning was thus evaluated solely on the basis of whether or not it would reach the ultimate consumer; there was no mention of the adequacy of [the] warnings to or training of [the intermediary].” Leonard v. Aluminum Co. of America, 800 F.2d 523 (5th Cir.1986) (quoting Aim v. Aluminum Co. of America, 717 S.W.2d 588 (Tex.1986)). If the intermediary makes clear from his actions that he effectively will not pass on or will not use the warning, the manufacturer cannot rely upon him to do so. See Restatement (Second) of Torts, § 389, Comment f (1965) (“Again, if the seller has reason to anticipate the possibility that the warning which he conveys to an intermediate dealer will not reach the ultimate user ... the seller is not relieved of liability when the user does not receive the warning.”).
V.
It is also subject to question whether a warning to patients could be fully “adequate.” Surgery patients may not be able *477to evaluate or to consent to an illegal and dangerous manner of prescribing and administering anesthesia, as the majority recognizes, majority at 471. Even a warning explicitly stating that, at risk of death, patients should not consent to any administration of fentanyl under these conditions might not be adequately comprehensible. Patients may have no means of evaluating the likelihood of danger from illegal prescription nor of weighing the danger against the benefits; this is precisely why a learned intermediary is required for prescription drugs. See Canterbury v. Spence, 464 F.2d 772, 780 (D.C.Cir.1972), cert. denied, 409 U.S. 1064, 93 S.Ct. 560, 34 L.Ed.2d 518 (1972) (“The average patient has little or no understanding of the medical arts, and ordinarily has only his physician to whom he can look for enlightenment with which to reach an intelligent decision.”). Similarly, patients may be unlikely to believe that doctors would willingly subject them to such risks, and thus foreseeably would not heed the warnings. Cf. Brazzell, 788 F.2d at 1359 (listing three factors for the validity of presumptions relating to causation: warnings to the public without the advice of a doctor would not be helpful; public policy favors placing the risks of loss on manufacturers not consumers; and the publicity surrounding the “swine flu program minimized the impact of any warning.”); Salmon, 520 F.2d at 1363 (overpromotion nullifies effect of valid warnings). But to say that no warning could be adequate requires neither absolution from liability, nor prohibition of the product. This case does not raise a regulatory question, appropriate for the legislature or administrative agencies, but rather the judicial question of who should bear the costs of avoidable risks.
McNeil in this case chose to supply the drug without further effort. If McNeil had preferred not to run the foreseeable risk of liability, McNeil could have changed the circumstances under which it provided the drug. I thus feel compelled to disagree with the district court’s determination that to remove a socially useful anesthetic from the market is “preposterous” or that we would have to revert to “the old ether” or “a shot of whiskey.” Rec. Vol. 2, District Court Opinion, at 14. As the plaintiff suggested, McNeil could have prevented liability by removing, selectively, the drug from hospitals that could not ensure that qualified doctors would prescribe.
McNeil would not have had to police the operating room by engaging in selective withdrawal or by conducting other activities suggested by the plaintiff. Enforcing compliance would have remained the task of the hospitals, doctors, and Mississippi authorities. McNeil’s only task would have been to obtain adequate assurances of compliance on which it could reasonably rely. Moreover, competitive or media pressures resulting from selective withdrawal would likely have forced hospitals themselves to abandon the illegal practice and to insist that anesthesiologists were hired or that surgeons were required meaningfully to supervise anesthesia. At that point, McNeil might have boosted its market share through the beneficial publicity of responsible action. The precedent of temporarily and entirely removing Tylenol (after some lots were laced with cyanide) from the market is one example of how effective withdrawal and restoration can be to prevent harm and to maintain or boost market share.
Finally, I do not believe that holding McNeil accountable for the foreseeable harm would be “preposterous” even were fentanyl the only anesthetic, were McNeil the sole provider of fentanyl, and were McNeil required (to avoid liability) to remove all fentanyl from the market. In that event, Restatement § 402A requires a careful balancing of the social benefits and risks, with equal regard for the successes and the failures of modern medicine in the particular circumstances, to determine whether the particular product’s risks are “reasonable.” See Reyes, 498 F.2d at 1274. The conclusion is not foregone, as the district court would imply; the health benefits from surgery — both necessary and elective, and with regard to life expectancy and quality of life valuations — would have to be *478weighed against the empirical likelihood of respiratory failure and death under the circumstances.
If liability were found, McNeil could still provide a useful anesthetic, and spread the risk of loss through its pricing mechanisms. This would comport with public policy and settled tort law, and would equitably and efficiently expand the segment of the public that shoulders the costs of permissible, regulated risks, from those who are directly harmed to purchasers (including those harmed), shareholders, and the insurance industry’s customers. See generally G. Calabresi, The Costs of Accidents: A Legal and Economic Analysis (1970); Abraham, Cost Internalization, Insurance, and Toxic Tort Compensation Funds, 2 Va.J.Nat.Res.L. 123 (1982).
Conclusion
The majority would permit the problem of illegal prescription of anesthesia by C.R. N.A.s to rest solely with the individual physician, who “assumes the burden of presiding over the patient’s best interests.” Majority at 472. As stated above, this is a situation where no doctor presides over the patients’ interests. The majority’s incantation of the talismanic phrase “learned intermediary” cannot transfigure the reality of this tragedy.
Moreover, this is not merely a doctor’s problem, but a social problem; doctors, nurses, and hospitals are clearly at fault and state authorities are not enforcing compliance with law. In such a case, McNeil cannot simply continue to sell its product, knowingly provide an ineffective warning, and let the public confront substantial, illegal, and potentially unreasonable risks.
Like others, drug companies must bear the burden of compensating the avoidable harms that they have caused; like others, drug companies will still be praised for the social benefits that they convey. We must always permit two questions to be asked and meaningfully analyzed: Whether the benefits in fact exceed the harms, and whether the harms can be avoided. Here, McNeil reasonably could have avoided the harms. I thus believe that both the law and common sense require the directed verdict to be reversed and the case remanded for a new trial.
APPENDIX I
Danny Collier, the C.R.N.A. who prescribed fentanyl for Michael Swayze, testified as follows:
Q. Isn’t it a fair statement that Dr. Edward Attix [(the surgeon in the Swayze operation)] usually relied upon you, the C.R.N.A. that was working with him, to handle all aspects of anesthesia for a particular case?
A. On most cases yes.
Q. Is it fair to say that you worked with most surgeons in the Hattiesburg area there at the hospital?
A. Correct.
Q. Is it also a fair statement to make that the other surgeons did not get actively involved with anesthesia questions in the surgeries in which you and the other surgeons worked with?
A. Yes.
Q. So, basically, the other surgeons left all questions of anesthesia up to you, Danny Collier, is that correct?
A. Yes.
Q. In your seven years while you were practicing your trade of C.R.N.A. in Hat-tiesburg, do you ever have a recollection of having pulled a package insert such as Exhibit A [the fentanyl warning] from the box [of anesthetic ampules mailed directly to nurse Collier, not to surgeons, which nurse Collier would bring into the operating room from the hospital pharmacy at his own initiative] and discussing it with a surgeon?
A. Not to my recollection.
*479APPENDIX I — Continued
Q. I understand you kept [the box with the warning] in the drawer of the cart so it was available in the operating theater, but would the surgeon as a matter of course ever come across Exhibit A in a standard operation?
A. Not to my knowledge.
Q. Did you perform an induction on Michael Swayze?
A. Yes, I did.
Q. When and how did you do that?
A. When Dr. Attix was in the theater and was ready to go, he told me he was ready to start. That’s when I performed the induction.
Q. How did you do the induction?
A. I did a crash induction.
Q. Who made the determination to utilize a crash induction?
A. I did.
Q. Did Dr. Attix make that decision?
A. No.
Q. Did you and Dr. Attix have any discussion of Sublimaze before making that induction?
A. No.
Q. Now, did Dr. Attix instruct you when to administer [incrementally] another cc. of Sublimaze?
A. No.
Q. Was it entirely in your judgment when to administer and if you needed to administer another cc. of Sublimaze?
A. Yes.
Q. Is that true for the entire 15 cc. of Sublimaze that you administered?
A. Yes.
Q. Did Dr. Attix have any discussion with you during the surgical procedure about your anesthesia method or the drug you used or the amount?
A. No.
Q. Did Dr. Attix know what you were administering and in what increments and amounts you were administering during the surgical procedure?
A. Not to my knowledge.
Q. In other words, you didn’t call it out to him, Doctor, I’m fixing to do X, Y, Z?
A. That’s correct.
Q. Now, following the surgery or close to the end of surgery, you administered to Michael Swayze 1 cc. of Narcan [an antagonist to fentanyl to combat respiratory depression], is that correct?
A. Correct.
Q. And that’s all of the Narcan that Mr. Swayze received that day?
A. As far as I know.
Q. Whose decision was it to administer 1 cc. of Narcan?
A. It was my decision.
Q. Did you exercise the judgment that Michael Swayze needed that 1 cc. of Nar-can?
A. Yes.
Q. Did Dr. Attix have any discussion with you about that administration of 1 cc. of Narcan?
A. No.
Q. Did you ever inform Dr. Attix either before or after you administered that 1 cc. of Narcan as to what you did?
A. No.
Q. To your knowledge, did Dr. Attix know that the man ever received any Narcan during the surgical procedure and immediately thereafter?
A. Not that I know of.
Q. The interchange between you and Dr. Attix prior to surgery, during surgery, and immediately after the surgery of Michael Swayze, was it similar to the interchange between you and Dr. Attix and other surgeons during surgery in other surgeries?
A. Yes.
*480APPENDIX I — Continued
Q. But to your knowledge, [Dr. Attix] didn’t know what you were using, when, or how much you were using at that particular time, is that correct?
A. Not to my knowledge.
Q. Is that last statement that I made and you affirmatively answered true of most surgeries that you performed at Methodist Hospital, that the physician would not actively direct you by verbal comments to you and that the physician would be unfamiliar during the time he was performing the surgery of what you were using, when you were using it, and how much? Is that true?
A. Well, the surgeon has confidence in me to do the anesthesia; and the surgeon at that particular time has his hands full with the operation. To take time out to discuss the particulars would take him away from his surgery, what he is there for.
Q. My statement is generally true?
A. Yes.
Q. Is it fair to say that a large number of hospitals in Mississippi of the approximate size of Methodist Hospital utilized C.R.N. A.’s in the same manner that Methodist Hospital utilized your services for the seven years that you were there?
A. To my knowledge, yes.
Rec. Vol. 3, at 95, 97-100, 106-10, 113-14. Dr. James Arens testified similarly as to the pattern and practice:
Q. And in these particular hospitals that we’ve been talking about, there were CRNA’s there that were practicing their trade of anesthesiology [ (sic) ] on a day-today basis without the supervision of an anesthesiologist?
A. Yes.
Q. From your review of the depositions in this case that you’ve described, that’s basically what happended in the Michael Swayze surgery?
A. That’s correct.
Q. And the young man, Danny Collier, had had, if I remember correctly, ten or eleven years of experience at day-to-day practicing anesthesiology [ (sic) ], so I don’t think we could find fault with his experience, could we?
A. I haven’t reviewed his cases, but in generic terms, no.
Q. So, I think that’s what lets you come to your conlusion that Danny Collier just had lack of understanding of the product that he was utilizing; is that correct?
A. Yes, the lack of understanding of the consequences.
Q. I understand what the law says, but is it a fact, Doctor, that generally surgeons in Mississippi, and including these hospitals where CRNA’s practice their trade without supervision of an anestheologist (sic), that these surgeons generally were unfamiliar with anesthesia agents, their propensities and their side effects, similar to Dr. Attix’s unfamiliarity that you testified to — is that a fair statement?
A. That’s a fair statement.
Q. So, my question is: Who was actually exercising the independant medical judgment in these communities where CRNA’s were practicing their trade without the supervision of an anesthesiologist if the surgeons were generally unfamiliar with the agents, the propensities and the side effects?
A. No one.
Q. So what I am saying is: It was pretty common knowledge in the anesthesia community that there were certain hospitals that CRNA’s were practicing their trade [in] without supervision and the surgeons were basically unfamiliar with the anesthesia agents, the propensities and the side effects; and, therefore, nobody was really exercising independant medical judgment for the patient with regard to anesthesia?
*481APPENDIX I — Continued A. That’s correct.
Q. And that was pretty common knowledge in the anesthesia community?
A. It was for that reason that I had the surgery residents from the University of Mississippi rotate through anesthesia, because I don’t agree with it, I think it’s wrong. But it currently is no different in Mississippi than it is in any other state in the union in 1985. And this is one of the problems we have in which CRNA’s at the present time, in 1985, as a matter of fact are — in some states have a license to practice anesthesia unsupervised.
Record, Vol. 3, at 146, 150, 152-54. Finally, Dr. Roy Wilson testified as follows:
Q. Doctor, were you familiar and are you now familiar with the commonly recognized medical standards in Mississippi in 1977 and ’78 as to the qualifications that a person must possess in order to prescribe and administer a narcotic anesthesia drug such as fentanyl?
A. Yes, sir.
Q. What were those standards?
A. A person must be a licensed physician, dentist, or in the case of veterinarian, portions of the drugs of a veterinarian [ (sic) ] in order to administer these drugs or to dispense the drugs.
Q. Under these commonly recognized medical standards that we’re referring here, did nurse anesthetists possess the professional qualifications to prescribe and administer fentanyl?
A. To administer, but not to prescribe.
Q. All right, sir. Under the standards that we’re talking about, Doctor, what does the term, and I quote, under the direction and supervision, end quote, of a physician mean?
A. This means that the drug, or whatever method of anesthetizing an individual, if it were an inhalation agent, is being administered essentially on the license of a physician.
Q. All right, sir. How does the requirement that we’ve just talked about under prescription and administration and under the direction and supervision of a physician limit, if at all, a nurse anesthetist in his or her independent, number one, selection of the anesthesia agent, the dosage to be administered, and the administration itself of the drug?
A. This can probably only be answered by saying that there is a dichotomy of care, because there is practice and propriety: Practice being the way that it is commonly done; propriety being the way that it is legally constituted and ethically constituted to administer these drugs.
Q. Under medical propriety, Doctor, how would that limit the nurse anesthetist in that category? You mentioned the dichotomy of the situation?
A. This would limit the nurse, because the nurse has no license nor is able to obtain a license to administer narcotics.
Q. Were these standards of practice that we’ve just discussed, known and recognized in the medical community in Mississippi and among manufacturers and sellers of ethical [prescription] drugs in that state in 1977 and ’78?
A. To the best of my knowledge, yes.
Q. Can the instances which you’ve discussed where anesthesia was being performed by nurse anesthetists in Mississippi in this time period that we’re talking about without the supervision of an anesthesiologist, who, if anyone, actually supplied the, quote, physician direction and supervision, end quote, requirement of the standard which you previously mentioned?
A. In the circumstances where there was supervision of any type, it was by the operating surgeon.
Q. In actuality, as a matter of routine practice, who made these decisions in those situations?
A. In most circumstances, the nurse anesthetist made this decision.
Q. Was that fact commonly known in the medical community in Mississippi in 1977, ’78?
*482APPENDIX I — Continued A. It was.
Q. Who generally, and as a matter of routine practice in these situations, selected the drugs and determined the dosage to be administered?
A. Under most circumstances, the nurse anesthetist.
Q. What part, as a matter of routine practice, if you know, Dr. Wilson, did the operating room surgeon or his assisting physician generally play in selecting the particular anesthesia drugs to be used in such a case and/or the dosage to be administered to the patient?
A. Unfortunately, very little.
Q. To what extent — what was the extent of the knowledge within the medical community within Mississippi and the drug community in Mississippi in 1978 that this practice which we’ve just discussed was prevalent and being done?
A. It was absolute.
Q. [D]o you have an opinion that you can state, based on reasonble medical probability, as to whether or not a nurse anesthetist in general or Mr. Collier, the nurse anesthetist in Michael’s case on July 13th, had the medical expertise to select the drug, to determine its dosage and to administer it without the intervention of an anesthesiologist based on what he read from Plaintiff's Exhibit 8 [the McNeil warning]?
A. Mr. Collier has no medical license or qualifications, and so, therefore, the answer on that basis alone would be, no, he did not have.
Rec. Yol. 4, at 268-72.
APPENDIX II
Dr. Wilson also testified:
I feel that the selection of the anesthetic fentanyl for a child with this type of surgery was totally inappropriate at the time it was selected; second, the amount of anesthesia given to the child was extremely large to the point of being dangerous in itself; third, the administration of additional doses of the fentanyl to the child were done with no documentation as to any necessity on the part of the nurse anesthetist.
Rec. Vol. 4, at 282.
. Cf. Talking Heads, Psycho Killer, Stop Making Sense (movie or record version 1985) (“Can’t seem to face up to the facts.”).
. Williams, 591 F.Supp. at 389, implies that McNeil can be held liable regardless of whether McNeil actually or constructively knew of the absence of learned intermediary intervention, if the illegal administration is a common practice. Williams (implicitly) and other cases (explicitly) abandon the old view that negligent warning and defective warning pose the same inquiry. Contrary to the majority’s and district court’s assertions, majority at 467 n. 3; Rec. Vol. 2, District Court Opinion, at 9, foreseeability concepts derived from negligence may be inapposite to strict liability warning claims, because the underlying rationale of strict liability focuses on the product, not on conduct, in relation to the circumstances. See Jackson, 727 F.2d at 512 (criticizing incorporation of negligence concepts of duty into strict liability); Hamilton, 285 So.2d at 747-48; cf. Bridges v. Chemrex Specialty Coatings, Inc., 704 F.2d 175 (5th Cir.1983). But see Walton v. Chrysler Motor Corp., 229 So.2d 568, 573 (Miss. 1969), overruled, Toliver v. General Motors Corp., 482 So.2d 213 (Miss.1985) (en banc). See generally J. Beasley, Products Liability and the Unreasonably Dangerous Requirement 421-43 (1981) (“Failure to Provide Adequate Warnings”) (discussion citing cases). The majority should not decide that negligent and defective warning are equivalent issues under Mississippi law or otherwise; under either view, the facts are sufficient to hold McNeil liable.