The opinion
TOBBINER, J.
I concur in the judgment, but I am unable to join either the majority opinion or the opinion of the Chief Justice. I propose here to explain my dissatisfaction with the present definition and application of the doctrine of res ipsa loquitur in that limited number of cases in which rare and inexplicable accidents occur in the operating room. In pursuing the laudable goal of shifting the losses occasioned by such accidents to the parties best able to protect against them through insurance, we have imposed the onus of negligence and malpractice upon capable and dedicated members of the medical profession, burdening the law of res ipsa loquitur with a sweep that is inaccurate, inefficient, and inequitable. I propose a redefinition of the doctrine governing these eases which seems to me more candid, more certain, and more consistent with our underlying objectives.
Initially, I set forth my reasons for joining the majority in affirming the judgments against both defendants. Given the evidence from which the jury could have found that the *415 specific acts and omissions alleged by plaintiff were negligent and proximately caused her injury, I am unwilling to assume that the verdict against the defendants rested upon the trial court’s instruction on res ipsa loquitur. Although I believe that the instruction should not have been given under the circumstances of this case, I would hold that the defendants waived any right to demand a new trial on that ground when they failed to request a special verdict to reveal the theory upon which the jury found them liable. (Code Civ. Proc., § 625.) 1
I cannot agree, however, with the route by which the majority reaches its result. As the Chief Justice demonstrates in his separate opinion herein, neither common knowledge nor expert testimony supported an inference in this case that accidents such as befell the plaintiff ordinarily bespeak a negligent cause. 2 To give a res ipsa instruction under such *416 circumstances invites a purely speculative leap and entrusts the jury with unreviewable power to impose or withhold liability as it sees fit. If public policy demands that defendants be held responsible for unexplained accidents without a reasoned finding of fault, such responsibility should be fixed openly and uniformly, not under the guise of negligence and at the discretion of a jury. (Cf. Escola v. Coca Cola Bottling Co. (1944) 24 Cal.2d 453, 463 [ 150 P.2d 436 ] (Traynor, J., concurring).)
I am likewise disturbed by the conclusion of the Chief Justice. that the victims of accidents which do not truly “speak for themselves” should be required to present evidence that the kinds of accidents they suffered are ordinarily caused by negligence. Even if expert medical testimony were readily available to plaintiffs in malpractice eases, 3 such a rule would unfairly penalize the surgical patient who is *417 injured by an accident of a type too rare or too little understood to permit meaningful statistic analysis of its probable cause. 4 Although I agree with the Chief Justice that the above requirement for application of the res ipsa doctrine follows from its basic premises, I submit that use of the doctrine itself fails to serve the ends of justice in eases such as this. Indeed, even the expanded version of res ipsa loquitur espoused by the majority falls considerably short of truly protecting the victims of. unfamiliar and unexplained surgical mishaps, since the majority would deny plaintiffs the benefit of a res ipsa instruction unless they could produce the kind of testimony which the Chief Justice would require, or could persuade a medical expert to characterize as substandard the conduct of those entrusted with their care. 5
Upon reexamining what seem t.o me the grave shortcomings of these varying formulations of res ipsa loquitur in surgical accident cases, I have concluded that the basic error lies in primary reliance upon the concept of negligence and that the courts should undertake a fundamental reassessment of the largely fictitious and often futile search for fault which presently characterizes medical injury litigation of the kind here involved.
At the outset we must recognize that, in the present state of medical knowledge, risks which even the most cautious physician could not have prevented may lead to accidents which even the most expert cannot explain. Although the vast majority of medical practitioners are protected financially by *418 liability insurance covering such accidents, 6 and although doctors and hospitals can readily transfer the cost of this insurance protection to their patients through higher medical fees, no technique yet devised can protect a doctor from the devastating impact which an adjudication of malpractice can have upon his professional standing. 7 Fearing that his competence may thus be impugned whenever he adopts a procedure difficult to justify to a lay jury, a surgeon may feel compelled to forego an unorthodox technique in order to protect his reputation from ruin. 8 Any system which thus diverts the doctor’s attention from the operating room to the courtroom leaves much to be desired. 9
In light of the expansion of res ipsa loquitur undertaken by such decisions as Quintal v. Laurel Grove Hospital, supra, 62 Cal.2d 154 , and by the majority opinion in the present case, there can be little doubt that the net effect of the doctrine is to shift from plaintiffs to defendants the cost of a certain number of unexplainable accidents in which no meaningful basis exists for finding the defendants at fault. 10 Thus the *419 concept of negligence as a prerequisite to medical liability-now provides only sporadic and illusory protection for the physician. At the same time, insistence under all circumstances upon a nominal finding of fault frustrates the risk-shifting purpose of the res ipsa doctrine as currently applied since it stands as an occasionally insuperable obstacle to the financial protection of inexplicably injured patients.
A system openly imposing liability without any pretense of negligence in this narrow range of cases can avoid unwarranted imputations of fault while permitting the rational development of badly needed doctrine. Simultaneously, such a system can insure that the burdens of unexplained accidents will not fall primarily upon the helpless but will be borne instead by those best able to spread their cost among all who benefit from the surgical operations in which these misfortunes occur. 11
The record in this case supports the conclusion that the plaintiff’s arthritic condition resulted from the premature termination of anesthesia, bringing the operation to an untimely halt. We deal here neither with a complication flowing from an undetectable idiosyncrasy of the patient 12 nor with a risk which the patient voluntarily assumed in electing to undergo this type of surgery; we deal instead with a failure of the operation to accomplish the result that the patient, in light of her own physical condition, reasonably expected it to achieve. 13
*420 If this failure could have been traced to the anesthetic itself, or to some mechanical inadequacy in the hospital’s surgical equipment, the plaintiff would not have been required to establish negligence as a prerequisite to recovery. 14 The wholly fortuitous circumstance that this plaintiff’s injury resulted instead from some undetermined mishap in the operating room should make no difference: in neither case should the patient’s right to recover turn on her ability to isolate a negligent cause for her surgical injury.
In such situations, the jury should be instructed that, if it finds that the plaintiff was injured in the course of an operation within the collective control of the defendants 15 and that this type of injury rarely occurs in such operations, 16 then it must return a verdict for the plaintiff unless the defendants establish that the injury resulted from an idiosyncrasy of the patient 17 or that the patient knowingly and voluntarily assumed the risk of incurring such an injury. 18
*421 Once the elusive and destructive search for an act or omission of “malpractice” has been restricted to those cases in which a negligent cause may actually be demonstrated, 19 a far higher percentage of all medical controversies will be settled out of court, without the “economic and emotional strain of protracted litigation requiring difficult or impossible proof.” (Ehrenzweig, op. cit., supra, 31 U.Chi.L.Rev. at 288.) In the relatively few eases which reach trial, the imposition of financial liability will not be aggravated by the ruinous consequences of a determination of malpractice unless the evidence points logically to such a finding.
We should not impose the stigma of negligence upon a doctor merely because an operation yields an uncommon and inexplicable result; in the present state of the medical art, the rarity of an event may well bear no relationship to negligence. Courts which ignore that fact in formulating the law of res ipsa loquitur unjustly penalize physicians and plunge the legal process into an abyss of uncertainty and obfuscation. Our proper concern for the financial protection of the patient gives us no warrant for faulting the doctor.
I must conclude that, in this limited category of eases, the attempt to fix liability exclusively in terms of traditional notions of fault has outlived its utility. Once it appears that an unexplained surgical accident has caused an unexpected injury, no useful end is advanced by rehearsing the ancient ritual of assessing blame.
An appellate court should not disturb a general verdict merely because the trial court gave the jury an abstractly correct instruction which the facts before it did not warrant, provided that another theory on which the ease was submitted to the jury finds substantial support in the evidence and is unaffected by error. (See Estate of Hellier (1914) 169 Cal. 77, 83 [ 145 P. 1008 ]; Posz v. Burchell (1962) 209 Cal.App.2d 324, 335-337 [ 25 Cal.Rptr. 896 ], and eases there cited; see also Tucker v. Landucci (1962) 57 Cal.2d 762, 766 [ 22 Cal.Rptr. 10 , 371 P.2d 754 ] ; Gillespic v. Rawlings (1957) 49 Cal.2d 359 , 368-369 [ 317 P.2d 601 ]; Edwards v. Gullick (1931) 213 Cal. 86, 88 [ 1 P.2d 11 ]; Verdelli v. Gray’s Harbor etc. Co. (1896) 115 Cal. 517, 525 [ 47 P. 364, 778 ]; Crosett v. Whelan (1872) 44 Cal. 200, 203 ; Moss v. Coca Cola Bottling Co. (1951) 103 Cal.App.2d 380, 384-385 [ 229 P.2d 802 ]; Shields v. Oxnard Harbor Dist. (1941) 46 Cal.App.2d 477, 491 [ 116 P.2d 121 ] (McComb, J.) ; Hume v. Fresno Irr. Dist. (1937) 21 Cal.App.2d 348, 356-357 [ 69 P.2d 483 ]; cf. Gordon v. Aztec Brewing Co. (1949) 33 Cal.2d 514, 520 [ 203 P.2d 522 ] ; Blanton v. Curry (1942) 20 Cal.2d 793, 799-800 [ 129 P.2d 1 ] (per curiam); Gerdes v. Pacific Gas & Electric Co. (1933) 219 Cal. 459, 471-473 [ 27 P.2d 365 , 90 A.L.R. 1071 ]; Christensen v. Malkin (1965) 236 Cal.App.2d 114, 123 [ 45 Cal.Rptr. 836 ]; Rather v. City & County of San Francisco (1947) 81 Cal.App.2d 625, 636 [ 184 P.2d 727 ].) Although our courts have not always taken this approach (see, e.g., Burks v. Blackman (1959) 52 Cal.2d 715, 719 [ 344 P.2d 301 ]; Edwards v. Freeman (1949) 34 Cal.2d 589, 594 [ 212 P.2d 883 ] ; Huebotter v. Follett (1946) 27 Cal.2d 765, 770-771 [ 167 P.2d 193 ]; Oettinger v. Stewart (1944) 24 Cal.2d 133, 139-140 [ 148 P.2d 19 , 156 A.L.R. 1221 ]; Christensen v. Bocian (1959) 169 Cal.App.2d 223 [ 336 P.2d 1018 ]; Schaffer v. Claremont Country Club (1959) 168 Cal.App.2d 351, 358 [ 336 P.2d 254 , 337 P.2d 139 ], reh. den. 168 Cal.App.2d 358 -359), consistent adherence to the rule stated herein would prevent needless appeals and retrials without injustice to either party.
Plaintiff adduced expert testimony to show that, when due care is used, premature termination of anesthetic is rare. The record contains no evidence, however, indicating that in those rare cases in which an anesthetic does terminate prematurely, a negligent cause is more prohahle than a non-negligent one. Although plaintiff presented evidence of specific *416 negligent acts which could have caused premature termination, such evidence provided no rational basis for a conclusion that, of the various possible causes, a negligent one was probably responsible.
The majority asserts: “ [I]f the low incidence of accidents when due care is used is combined with proof of specific acts of negligence of a type which could have caused the occurrence complained of. . . . the jury may properly conclude that the accident was more probably than not the result of someone’s negligence.” (Ante, p. 413.) I cannot agree.
Suppose, for example, that in 5 percent of all operations in which due care is used, a certain spinal anesthetic inevitably terminates prematurely because of an undetectable excess of myelin on the patient’s nerves; suppose further that a specific technique for administering the anesthetic does not alter the likelihood of premature termination in patients with an excess of myelin but creates a 2 percent risk of premature termination in normal patients, whereas another available technique, equally desirable in all other relevant respects, creates only a 1 percent risk of premature termination in normal patients. Under these circumstances, the technique which creates twice as high a risk in normal patients and yields no compensating benefit would presumably be considered negligent.
If one were to examine 100 operations in which this negligent technique had been employed, one would expect to find 2 operations in which such negligence caused premature termination, compared with 5 in which an overabundance of myelin caused premature termination. Yet, in every one of these hypothetical operations, the majority would invite the jury to infer a negligent cause without further guidance from the evidence before it; I find it disturbing to note that in 5 out of every 7 cases of premature termination coupled with a specific negligent act, this inference would blame the doctor for an accident he did not cause.
The strong reluctance of doctors to testify against each other has frequently been noted (see, e.g., Huffman v. Lindquist (1951) 37 Cal.2d 465, 484 [ 234 P.2d 34 , 29 A.L.R.2d 425 ] (Carter, J., dissenting); Belli An Ancient Therapy Still Applied: The Silent Medical Treatment (1956) 1 Vill.L.Rev. 250, 259) and numerous corrective measures have been sug gested (see, e.g., Note, Malpractice and Medical Testimony (1963) 77 Harv.L.Rev. 333, 338-350), but the problem apparently remains (see Note, Medical Malpractice — Expert Testimony (1966) 60 Nw.U.L.Rev. 834, 835 837).
Since the accidents with which we are here concerned by hypothesis occur rarely, there is little hope oí obtaining broadly based statistics of the sort hypothesized for the computations in footnote 2, supra. The complexity of the concept of negligence as applied to medical techniques, coupled with the difficulties of determining the cause of the few accidents which might be included in any purported sample, render suspect the claim of any expert who asserts that in a representative group of eases he was able to determine the relative proportion of negligent and non-negligent causes.
The majority reaffirms the holding of Siverson v. Weber (1962) 57 Cal.2d 834, 839 [ 22 Cal.Rptr. 337 , 372 P.2d 97 ], that rarity alone does not warrant a conditional res ipsa instruction, and limits the holding of Quintal v. Laurel Grove Sospital (1964) 62 Cal.2d 154, 164-166 [ 41 Cal.Rptr. 577 , 397 P.2d 161 ], to cases in which rarity is coupled with “proof of specific acts of negligence of a type which could have caused the occurrence complained of.” (Ante, p. 413.) Plaintiffs who cannot qualify under Quintal by obtaining such proof are thus relegated to the basic rule of Siverson that res ipsa is applicable only if common knowledge or expert witnesses establish that accidents of the sort which befell the plaintiff are “more likely the result of negligence than some cause for which the defendant is not responsible.” (57 Cal.2d at 839.)
A 1959 estimate showed that more than 92 percent of American doctors carried professional liability insurance, with an average coverage ranging from $25,000 for general practitioners to $100,000 for specialists. (Silverman, Medicine’s Legal Nightmare, Saturday Evening Post, April 25, 1959, pp. 36, 120.)
Indeed, many doctors genuinely fear that even if they win a malpractice ease, they will be “all but destroyed professionally.” (Shindell, Medicine versus Law: A Proposal for Settlement (1953) 151 A.M.A.J. 1078, 1079.)
See Cohn, Medical Malpractice Litigation: A Plague on Both Houses (1966) 52 A.B.A.J. 32; McCoid, The Care Required of Medical Practitioners (1959) 12 Vand.L.Rev. 549, 608; Silverman, op. cit. supra, April 11, 1959, p. 48; The Urge To Sue, Time, Nov. 28, 1960, pp. 69, 70. A number of hospitals, for example, are said to have prohibited the use of spinal anesthetics, purportedly reacting to cases adjudicating that physicians employing their facilities were guilty of malpractice because of unfortunate results following the use of such anesthetics. (Silverman, ibid.)
When every patient is viewed largely as a potential plaintiff, the method of treatment chosen by the physician may well be that which appears easiest to justify in court rather than that which seems best from a purely medical standpoint. (See Siverson v. Weber, supra, 57 Cal.2d at p. 839; Rubsamen, Res Ipsa Loquitur in California Medical Malpractice Law — Expansion of a Doctrine to the Bursting Point (1962) 14 Stan.L.Rev. 251, 282.) The probable victim of such litigation-oriented medical practice is oí course the patient, who suffers first when he receives less than the best available care, and second when the doctor whom he decides to sue understandably appeals to the jury’s inclination to protect a physician’s professional standing. (See Fleming, Developments in the English Law of Medical Liability (1959) 12 Vand.L.Rev. 633, 634.)
See generally 2 Harper and James, Torts (1956) § 19.6, p. 1081; see also id., § 19.5, pp. 1080-1081 & fns. 16-18; § 19.7, p. 1089 & fn. 17; Ehrenzweig, Compulsory ‘‘Hospital-Accident” Insurance: A Needed First Step Toward, the Displacement of Liability for ‘‘Medical Malprac *419 tice” (1964) 31 U.Chi.L.Rev. 279, 281-282 & fns. 8-9; Morris, Res Ipsa Loquitur — Liability Without Fault (1958) 25 Ins. Counsel J. 97.
See Ehrenzweig, op. cit. supra, passim; Calabresi, Some Thoughts on Risk Distribution and the Law of Torts (1961) 70 Yale L.J. 499 , 548-549; James, Accident Liability Reconsidered: The Impact of Liability Insurance (1948) 57 Yale L.J. 549 , 550 & fn. 1, 553 & fn. 8; cf. Greenman v. Yuba Power Products, Inc. (1963) 59 Cal.2d 57, 63-64 [ 27 Cal.Rptr. 687 , 377 P.2d 897 ],
As the Chief Justice points out, expert testimony in this case supports the view that a certain number of patients are afflicted with a condition involving an overabundance of myelin surrounding their nerves. This rare condition, known as rachiresistanee, apparently cannot be detected in advance and either prevents the deposit of an adequate quantity of the anesthetizing agent on the patient’s nerves or accelerates the rate at which the agent disappears. One of the defendants testified that the patient “had good and profound anesthesia for the prescribed time before she did feel the pain” and that, for this reason, he concluded that the patient probably ‘ ‘ detoxified faster than normal ’ ’ because of raehiresistanee. In light of the trial court’s instructions and the jury’s verdict, the jury evidently rejected this explanation, and I see no basis on which an appellate court could disturb the jury’s conclusion in this regard.
I note in this connection that some courts have permitted injured patients to sue for breach of a warranty that surgery would not aggra *420 vate their malady. (See Recent Decisions (1962) 37 Notre Dame Law. 725 .) The transition from express to implied warranty, and thence to a legally imposed Lability without fault, is too familiar to require detailed elucidation here. (See Greenman v. Yuba Power Products, Inc., supra, 59 Cal.2d 57, 61-63 .)
See, e.g., Greenman v. Yuba Power Products, Inc., supra, 59 Cal.2d 57 ; see also Bowles v. Zimmer Manufacturing Co. (7th Cir. 1960) 277 F.2d 868, 874 (breach of warranty by manufacturer of surgical pin) ; cf. Note, The Medical Profession and Strict Liability for Defective Products —A Limited Extension (1965) 17 Hastings L.J. 359 .
Compare Ybarra v. Spangard (1944) 25 Cal.2d 486 [ 154 P.2d 687 , 162 A.L.R. 1258 ].
Compare Quintal v. Laurel Grove Hospital, supra, 62 Cal.2d 154 . Plaintiffs who suffer from injuries of a type which commonly accompany a given medical procedure could of course proceed against defendants on an ordinary negligence theory. (See fn. 19, infra.)
Compare Prosser, The Fall of the Citadel (Strict Liability to the Consumer) (1966) 50 Minn.L.Rev. 791, 810-811 & fns. 104-106. A doctor who knew or should have learned of the patient’s peculiarity might theoretically be held liable if his negligence could be shown to have caused the injury.
Compare Farber v. Olkon (1953) 40 Cal.2d 503, 511 [ 254 P.2d 520 ], in which we concluded that a malpractice plaintiff was not entitled to an instruction on res ipsa loquitur since undisputed testimony established that the bone fractures of which the plaintiff complained constituted “a. calculated and even an expected risk of the [electro-shock] treatment.” In determining which risks a patient may voluntarily assume in submitting to a given medical procedure, the controlling consideration must of course be the reasonable expectations of the patient arising out of his relationship with the doctor, not the precise language of any prior agreement or understanding. (See Tunkl v. Regents of University of California (1963) 60 Cal.2d 92 [ 32 Cal.Rptr. 33 , 383 P.2d 441 , 6 A.L.R.3d 693 ]; Darling v. Charleston etc. Hospital (1965) 33 Ill.2d 326 [ 211 N.E.2d 253 ]; cf. Gray v. Zurich Insurance Co. (1966) 65 Cal.2d 263, 270-271 [ 54 Cal.Rptr. 105 , 419 P.2d 168 ].)
Nothing in this opinion should be construed to suggest a change in the means by which a patient might prove actual negligence or in the defenses which a doctor might properly interpose to a negligence claim. Thus, for example, instructions on res ipsa loquitur would remain available when warranted by the evidence; there would no longer be any justification, however, for giving such instructions simply because rarity and specific acts of negligence might both be present in a given case. (See fn. 2, supra.) A verdict predicated upon inferred negligence under a res ipsa instruction would henceforth be sustained only under the conditions set forth in the separate opinion of the Chief Justice.