BOYLE, J.

We granted leave in these cases to resolve the question whether it is error in a medical malpractice action for the trial court to instruct the jury that a doctor or surgeon is not a guarantor of results. In Dziurlikowski, we hold that the instruction given was erroneous, and we reverse the jury's verdict and remand the case for a new trial. We affirm the jury's verdict in Jones because we find that the guarantor instruction was balanced by the duty of care instruction and was not erroneously given in the context of the case.

We note that the instructions at issue in these two cases are not identical. In Jones, the instruction stated:

No physician can be required to guarantee results, but the law demands that they [sic] bring and apply to the case at hand that degree of skill and care, knowledge and attention ordinarily possessed and exercised by other orthopedic surgeons in the same specialty under like circumstances.

In Dziurlikowski, the instruction was more extensive and included the statement that an adverse result is not, of itself, evidence of negligence:

The difficulties and uncertainties in the practice of medicine and surgery are such that no one can be required to guarantee results and all the law demands is that the individuals involved bring and apply to the case in hand that degree of skill, care, knowledge and attention ordinarily possessed and exercised by practitioners of the medical profession under like circumstances. The mere fact that an adverse result may occur following surgery is not in itself evidence of negligence.

The differences in these instructions require a distinct analysis in each case. Jones raises the issue of the propriety of the "no guarantor of results" language when coupled with amplifying language properly defining in affirmative terms the duty of care of a physician to a patient. The instruction in Dziurlikowski combines the "no guarantor" and affirmative duty of care language used in Jones with an additional phrase: "The mere fact that an adverse result may occur following surgery is not in itself evidence of negligence." Thus, the issue in Jones is simply whether, read as a whole, the instruction correctly described the standard of care in a traditional medical malpractice case. The issue in Dziurlikowski, however, is more complex. Due to the nature of this case, we are required to determine whether, in a case where circumstantial evidence of negligence is raised by the proofs, such an instruction, coupled with a statement that an adverse result is not of itself evidence of negligence, interfered with the factfinder's understanding of permissible inferences.

I

FACTS

In Jones, the plaintiffs claimed that the defendant doctor had breached the standard of care in his diagnosis and treatment of an injury to Mr. Jones' leg and ankle. Mr. Jones, a diabetic, sustained his injury while at work when he slipped on some oil and fell. After an x-ray was taken by the plant doctor, Mr. Jones was advised that he had a sprain and that he should rest for five days. After a week, Mr. Jones, still unable to walk, contacted his personal physician, who took x-rays, diagnosed a fracture to the left leg, and referred the plaintiff to defendant, Charles Porretta, M.D., an orthopedic surgeon.

Dr. Porretta reviewed the x-rays and, concurring in the diagnosis of a fracture, applied a short-leg walking cast to the left foot and told Mr. Jones to remain off the leg for forty-eight hours and then to move about on the leg. Five weeks later, the defendant removed the cast and further x-rayed the leg and foot.

About six weeks later, Mr. Jones was continuing to experience discomfort and again consulted with Dr. Porretta, who ordered more x-rays. These revealed a second fracture, but the leg was not recast. A short time later, the plaintiff returned to the defendant's office, where Dr. Porretta gave Mr. Jones permission to return to work. Six months later, Mr. Jones returned to Dr. Porretta's office after continuing to experience difficulty with his left foot. The defendant discussed surgical alternatives with the plaintiff, but Mr. Jones declined in view of his diabetic condition, which made surgery hazardous. Dr. Porretta then prescribed a corrective shoe.

Mr. Jones consulted with two other doctors about the problems with his left foot and leg, and a brace was prescribed. The problems continued, however.

During this time, Mr. Jones developed a callus on his right foot, which he claimed resulted from the need to bear his weight almost entirely on his right leg. The plaintiff consulted a podiatrist, who found an additional fracture on the left heel bone during routine x-rays of both feet. Treatment was undertaken for the callus, but it later became infected and two toes had to be amputated from Mr. Jones' right foot. Mr. Jones was unable to return to work and is permanently disabled.

The plaintiffs, in their complaint, claimed that the defendant had breached the standard of care in several regards: in failing to take the appropriate x-rays of the injured left ankle and leg, in failing to properly perform a reduction of the fractured bones before applying the cast, in failing to apply a proper cast, in failing to order no weight-bearing, in failing to conduct a proper and thorough examination when the defendant first saw Mr. Jones, and in failing to refer Mr. Jones to a specialist or to request a consultation.

The defense presented by Dr. Porretta was that, even if the second fracture had been discovered, the treatment would have been the same as that prescribed in this case. The defendant further argued that Mr. Jones' diabetes and its complications (diabetic neuropathy and Charcot's joint) were the proximate cause of the plaintiff's injuries and damage.

At the close of proofs (and after giving the standard jury instruction on professional negligence, SJI 30.01), the court, over the plaintiffs' objections, added the instruction at issue in this case.

The jury returned a verdict for the defendant and the plaintiffs moved for a new trial on the basis of the additional instruction. The court denied the motion and plaintiffs appealed in the Court of Appeals.

The Court of Appeals agreed with the plaintiffs that the additional instruction was a deviation from a standard jury instruction and was erroneously given in this case. It then reversed the jury's verdict on the basis of the presumption of error requiring reversal set down by this Court in Javis v Ypsilanti Bd of Ed, 393 Mich. 689; 227 N.W.2d 543 (1975). The Court noted, however, that it would not have so held if an objection had not been raised. The defendants appealed in this Court. The application was held in abeyance pending the Court's decision in Moody v Pulte Homes, 423 Mich. 150; 378 N.W.2d 319 (1985). After that decision, we granted leave to appeal.

In Dziurlikowski, the plaintiffs, Ronald and Mary Ann Dziurlikowski, claimed that the defendants, Dr. Thomas Morley, M.D., South Oakland Anesthesia Associates, P.C., Elizabeth Kucherenko, C.R.N.A., and William Beaumont Hospital had breached the standard of care by failing to follow proper procedures in the giving of anesthetic to the plaintiff husband.

Mr. Dziurlikowski was admitted to William Beaumont Hospital suffering from bleeding ulcers. While in the hospital, he underwent recommended surgery, during which the doctors also found it necessary to remove his gall bladder. This additional complication extended the anticipated length of the surgery from approximately two hours to approximately four hours.

After the surgery, Mr. Dziurlikowski experienced difficulty with movement in his right arm and underwent physical therapy. His condition was diagnosed as a brachial plexus palsy, and at the time of trial he was still unable to use his arm as before.

The plaintiffs claimed that the injury was caused by the anesthesiology team which, they alleged, had failed either to properly position his arm on the operating table or to notice that his arm had been moved during the surgery. The defendants argued that there was no negligence and that brachial plexus palsy was a relatively rare and unfortunate but occasionally unavoidable consequence of extended anesthesia.

At trial, the testimony of several experts was presented, and during the cross-examination of one of the defendants' experts, counsel for the plaintiffs attempted to impeach the testimony by the use of an article in the journal Anesthesiology. The witness testified that the periodical was a reliable authority, but that he did not agree with the author of the particular article being referred to by the plaintiffs' counsel. The defendants objected to the use of the article, and the court sustained the objection, noting:

But let me tell you that I'm not entirely confident of that ruling and I'll be willing to review that if you wanted to review this matter Thursday. I'd like you to show me some authority for your position as to the correct procedure.

Plaintiffs' counsel responded:

Perhaps we could do this. I could finish the cross-examination that I have not touching on the articles, and I will notify Mr. Feringa and Mr. Secrest sometime Wednesday if it's my desire to pursue the matter of the articles further.

The matter was not raised again during trial. At the close of proofs, the trial court instructed the jury with the appropriate Standard Jury Instructions and then added the disputed instruction on "guarantor of results/adverse result not evidence of negligence." The court additionally instructed, "I charge you that you may not arrive at your verdict by the process of guess, conjecture or speculation."

The jury returned a verdict for the defendants, and the plaintiffs moved for a new trial, claiming that the two additional jury instructions and the ruling with regard to the cross-examination of defendants' expert were erroneous and impermissibly prejudiced the plaintiffs' case. The trial court denied the motion, and the plaintiffs appealed the issues in the Court of Appeals.

The Court of Appeals agreed with the defendants that the second jury instruction on "guess, conjecture or speculation" was not inappropriate, when the instructions were considered in their entirety. It found, however, that the plaintiffs were correct in their contention that the "guarantor" instruction was erroneously given and that the plaintiffs' counsel should have been allowed to impeach defendants' expert by use of the article. The Court of Appeals declined to reverse on the basis of the impeachment ruling, but noted, with regard to the "guarantor" instruction:

While we are not convinced that the giving of this instruction critically influenced the jury's verdict in this lengthy trial and are therefore reluctant to reverse on this basis, we are bound by the presumptively prejudicial rule of Javis v Ypsilanti Bd of Ed, 393 Mich. 689; 227 N.W.2d 543 (1975). [ 143 Mich. App. 729, 732; 372 N.W.2d 648 (1985).]

The defendants appealed the reversal in this Court, and the plaintiffs cross-appealed the remaining two issues.

II

STANDARD FOR ADDITIONAL JURY INSTRUCTIONS

As a preliminary matter, we agree with the defendants in both cases that these instructions should not be analyzed as omissions or deviations from the Standard Jury Instructions. The applicable Standard Jury Instructions were given to the jury in both of these cases, after which the contested instruction was given. The instructions were not a modification of the Standard Jury Instructions but, rather, were additional jury instructions which must be analyzed according to the standards set forth in MCR 2.516(D)(4) which provides:

This subrule does not limit the power of the court to give additional instructions on applicable law not covered by the SJI. Additional instructions when given must be modeled as nearly as practicable after the style of the SJI, and must be concise, understandable, conversational, unslanted, and nonargumentative.

The defendants argue that the "guarantor" instructions in these cases satisfy the dictates of MCR 2.516(D)(4) because they are based on Michigan case law. While we may agree that, as a general proposition, it is a correct statement of the law that a doctor or surgeon does not guarantee results, MCR 2.516(D)(4) requires more than a showing that the additional instructions correctly state the law; it requires that the instructions be "concise, understandable, conversational, unslanted, and nonargumentative," as well as based on "applicable law." Further inquiry is thus required as to the theories and proofs presented and the form of the instructions in order to determine if these jury instructions were applicable or appropriate in either or both of these cases.

III

GUARANTOR INSTRUCTION IN STANDARD MEDICAL MALPRACTICE CASE

We first turn to the guarantor instruction given in Jones. We are not convinced that the guarantor instruction in this case was erroneously given. Although it could be viewed as argumentative and prejudicial in a case with different facts, we agree with the defendants that, in this case, it was nothing more than an innocuous recitation of what a reasonable juror already understands. We note that a part of the plaintiffs' argument was that Dr. Porretta made a mistake in judgment and that "[h]e's a physician; he's to heal." The defendant's request for the "no guarantor" charge was apparently focused on the position that a physician is not liable for an error of judgment if the physician has acted within the appropriate standard of care. We find that the instruction, along with a statement which reiterated the correct standard of care applicable to a finding of professional negligence, was not erroneously given. See also Warfield v Wyandotte, 117 Mich. App. 83; 323 N.W.2d 603 (1982).

A physician is not normally held liable in a fault-based system unless it is shown that the conduct violated the standard of professional care. A "no guarantor/bad result" instruction states "well-nigh universally recognized principles of medical malpractice law." Watson v Hockett, 107 Wn.2d 158, 163; 727 P.2d 669 (1986). While use of the term "guarantor" is a somewhat formalistic means of advising a jury as to the standard of care in a professional medical malpractice case, it is not a misleading statement when it is followed, as it was in the Jones case, with a reiteration of the level of care that a physician owes to a patient.

The doctrine is well settled in both the American and the English jurisdictions that the general practitioner of medicine or of surgery does not, in the absence of special contract, impliedly warrant the success of his treatment or operation, but does guarantee to possess and carefully to apply such professional skill and learning as are ordinarily possessed by general medical practitioners in the locality in which he practices.

* * *

[This rule] applies equally to a specialist. But a specialist, by holding himself out as possessing special knowledge and skill in his branch of the medical profession, impliedly guarantees that he possesses learning and skill superior to those of the general practitioner, and that he will carefully apply them in his treatment or operation. [27 ALR 1250, 1250-1255.]

In many other jurisdictions, "guarantor" instructions, when properly balanced by the applicable standard of care have been found to have been appropriate when given in conjunction with proper standard of care instructions. See, e.g., Watson v Hockett, supra; Fall v White, 449 N.E.2d 628 (Ind App, 1983); Roach v Hockey, 53 Or. App. 710; 634 P.2d 249 (1981); Lange v Schultz, 627 F.2d 122 (CA 8, 1980). In addition, several states include in their standard jury instructions, instructions which distinguish and define the standard of care of physicians in medical malpractice cases. See, e.g., California Jury Instructions, Civil, BAJI 6.02; 3 Ohio Jury Instructions 331.01(3), p 183 (1984); 1 New York Pattern Jury Instructions — Civil (2d ed), PJI 2:150, p 390 (1974); 1 Wisconsin Jury Instructions — Civil 1023.

We agree with the plaintiff that there are times when the giving of such an instruction may mislead the jury by focusing the jury's attention on an irrelevant issue unsupported by the defendant's theory of the case or the evidence submitted. We do not find that the jury was distracted by the instruction which, at worst, could be viewed as unnecessary. As the Nebraska court stated in its assessment of an instruction which told the jury that there was no presumption of negligence upon the mere happening of an accident:

Although the giving of instruction No. 15 was unnecessary, there was no prejudicial error since the jury was properly instructed upon the burden of proving the negligence charged as a proximate cause of the injury. [ McKinney v Cass Co, 180 Neb. 685, 695-696; 144 N.W.2d 416 (1966).]

We note, finally, that our decision in Jones does not reflect our approval of the use of this instruction. Where proofs put the significance of an adverse result in issue, it may be more appropriate to explain the physician's duty of care by advising the jury that there are inherent risks in medical treatment which are beyond the physician's control. In all events, like the Standard Jury Instructions, the giving of supplemental instructions are to be determined by the trial court "not in an abstract or theoretical sense, but in the context of the `personality' of the particular case on trial, and with due regard for the adversaries' theories of the case and of counsel's legitimate desire to structure jury argument around anticipated jury instruction." Johnson v Corbet, 423 Mich. 304, 327; 377 N.W.2d 713 (1985).

We reverse the decision of the Court of Appeals in Jones. No presumption of error arises as a result of a deviation from or addition to the Standard Jury Instructions, and, in the context of this case, the instruction given was not erroneous.

IV

GUARANTOR/BAD RESULT NO EVIDENCE OF NEGLIGENCE INSTRUCTION IN RES IPSA LOQUITUR MEDICAL MALPRACTICE CASE

We begin our analysis in Dziurlikowski by noting first that the plaintiffs and the defendants acknowledge that the case went forward at trial on a theory of res ipsa loquitur. Having held in Jones that the "no guarantor" instruction was not erroneous, we now consider whether the "no guarantor/bad result no evidence of negligence" instruction was error in this context.

We are well aware of the longstanding debate as to whether this state recognizes the doctrine of res ipsa loquitur. The discussion by Justice SMITH in his minority opinion in Indiana Lumbermens Mutual Ins Co v Matthew Stores, Inc, 349 Mich. 441, 452-455; 84 N.W.2d 755 (1957), provides an excellent summary of the controversy.

We have thus reached decision without reliance upon the Latin phrase "res ipsa loquitur." ("If that phrase had not been in Latin, nobody would have called it a principle." Lord Shaw in Ballard v North British R Co, [1923] Sess Cas, HL (Scot) 43, 56; Prosser on Torts [2d ed], § 42, p 201, note.) We are aware, of course, of how often we have stated that in this jurisdiction we do not employ the principle. (See observations of BLACK, J., in Higdon v Carlebach, 348 Mich. 363, 374, 375 [ 83 N.W.2d 296 (1957)], quoting, also, Dean King of the Detroit College of Law, to the effect that "The Supreme Court continues to say that the doctrine of res ipsa loquitur does not apply in Michigan and then proceeds to apply it.") We are likewise aware of such cases as Pattinson v Coca-Cola Bottling Company of Port Huron, 333 Mich. 253 [ 52 N.W.2d 688 (1952)], in which we approved the award of damages in a case involving an exploding soft-drink bottle, a situation in which courts even professing to apply res ipsa loquitur have denied recovery on similar facts. Loebig's Guardian v Coca-Cola Bottling Co, 259 Ky. 124 ( 81 S.W.2d 910) [1935]. It is time to attempt some clarification of our position with respect to this so-called "doctrine" or "principle."

As noted above, Dean King says we do apply the doctrine of res ipsa loquitur. Justice CARR feels otherwise. "This Court," he says, "has never approved the res ipsa loquitur rule." Higdon v Carlebach, supra, 378. Which is right? The real difficulty here lies in an embarrassment of riches. Both may be right. For the term res ipsa loquitur has no fixed meaning. Its meaning varies not only from jurisdiction to jurisdiction but it is not always consistently applied even within a single jurisdiction.

* * *

Leaving the general problem and looking particularly at the situation in our own jurisdiction, there can be no doubt whatever that as the doctrine of res ipsa loquitur is defined by one of the great masters of our law, we do have it and we do apply it in this jurisdiction. For purposes of comparison we will quote on the subject, in parallel columns, Holmes, J., in Graham v Badger, 164 Mass. 42, 47 ( 41 N.E. 61) [1895], and MORSE, J., in Barnowsky v Helson, 89 Mich. 523, 524, 525 [ 50 N.W. 989 (1891)]:

Holmes, J. | MORSE, J. | "Res ipsa loquitur, — which is | "In this case the falling of the merely a short way of saying | roof was in and of itself some that, so far as the court can | evidence that the work of raising see, the jury from their | it was not being done with the experience as men of the world | ordinary care and may be | warranted in thinking that an | skill. It is true that the mere fact accident of this particular kind | of an injury does not impute commonly does not happen except | negligence on the part of anyone, but, in consequence of negligence, | where a thing happens which would and that therefore there is | not ordinarily have occurred if due a presumption of fact, in the | care had been used, the fact of such absence of explanation or other | happening raises a presumption of evidence which the jury believe, | negligence in someone." that it happened in consequence | of negligence in this case." |

It is too late, in our opinion, to insist that res ipsa loquitur, as above defined (and this is a classic definition), does not apply in this State.

Justice SMITH failed to persuade a majority that the Court should acknowledge the fact that Michigan had applied a version of res ipsa loquitur. The next year, in Weisenberg v Village of Beulah, 352 Mich. 172; 89 N.W.2d 490 (1958), his opinion was quoted with approval by a four-to-three majority, and, in 1959, in Mitcham v Detroit, 355 Mich. 182, 188; 94 N.W.2d 388 (1959), Justice VOELKER, writing for a five-justice majority, underscored the point with typical delightful imagery:

[W]e seem frequently to apply our version of " res ipsa" much as a disdainful lady might proclaim her profound aversion to strong drink just as she is tossing off a bumper of patent medicine harboring enough alcohol to floor a stevedore.

Ironically enough, the Michigan version of the doctrine of res ipsa loquitur in some respects plainly "out ipsas res ipsa," as it were. . . .

Justice VOELKER concluded that "[b]y whatever euphemisms we may choose to call it, we suspect that res ipsa loquitur is here to stay" and thus appeared to acknowledge the doctrine in this state. Id., p 190.

Despite Justice VOELKER'S pronouncement, as late as 1981, in Wilson v Stilwill, 411 Mich. 587, 607, n 5; 309 N.W.2d 898 (1981), Justice MOODY, writing for the Court, noted in a footnote:

Michigan has not formally adopted the doctrine of res ipsa loquitur. However, the underlying concepts of res ipsa loquitur, which are circumstantial evidence and negligence concepts, have been applied in Michigan.

Whether phrased as res ipsa loquitur or "circumstantial evidence of negligence," Mitcham v Detroit, supra, p 186, it is clear that such concepts have long been accepted in this jurisdiction. The time has come to say so. We, therefore, acknowledge the Michigan version of res ipsa loquitur which entitles a plaintiff to a permissible inference of negligence from circumstantial evidence.

The major purpose of the doctrine of res ipsa loquitur is to create at least an inference of negligence when the plaintiff is unable to prove the actual occurrence of a negligent act. According to Prosser Keeton, Torts (5th ed), § 39, p 244, in order to avail themselves of the doctrine, plaintiffs in their cases in chief must meet the following conditions:

(1) the event must be of a kind which ordinarily does not occur in the absence of someone's negligence;

(2) it must be caused by an agency or instrumentality within the exclusive control of the defendant;

(3) it must not have been due to any voluntary action or contribution on the part of the plaintiff.

In Wilson v Stilwill, supra, p 607, this Court noted these conditions, as well as a fourth criterion: "`Evidence of the true explanation of the event must be more readily accessible to the defendant than to the plaintiff.'"

Application of the rule of res ipsa loquitur is limited in medical malpractice cases. "It is the general rule, in actions for malpractice, that there is no presumption of negligence from the mere failure of judgment on the part of a doctor in the diagnosis or in the treatment he has prescribed, or from the fact that he has been unsuccessful in effecting a remedy, or has failed to bring about as good a result as someone else might have accomplished, or even from the fact that aggravation follows his treatment." Shain, Res ipsa loquitur, 17 S Cal L R 187, 217 (1944). As a result, and because the jury often cannot rely on its own experience in such cases, expert evidence must usually be presented. "What this means is that ordinarily laymen are not qualified to say that a good doctor would not go wrong, and that expert testimony is indispensable before any negligence can be found." Prosser Keeton, supra, § 39, p 256.

The rule that a prima facie res ipsa medical malpractice case requires more than a showing of bad result is well supported in our cases and in the case law of other states:

There appears to be little question that the doctrine of res ipsa loquitur is inapplicable in malpractice actions when its invocation is sought solely upon the fact that the treatment was unsuccessful or terminated with poor or unfortunate results, and this conclusion is but in accord with, or resulting from, the universally recognized propositions that the mere fact of a poor or unsuccessful result does not raise a presumption or inference of negligence, does not constitute evidence in itself of negligence, does not establish a prima facie case, and does not shift to the defendant the necessity of carrying the burden of proof or going forward with the evidence. [162 ALR 1265, 1267.]

In Zoterell v Repp, 187 Mich. 319; 153 N.W. 292 (1915), we discussed the theory of circumstantial evidence of negligence in the malpractice context and concluded there was not competent evidence to raise an issue for the jury. We stated:

The difficulties and uncertainties in the practice of medicine and surgery are such that no practitioner can be required to guarantee results, and all the law demands is that he bring and apply to the case in hand that degree of skill, care, knowledge, and attention ordinarily possessed and exercised by practitioners of the medical profession under like circumstances [citations omitted]; and the bare fact that full recovery does not result, or that a surgical operation is not entirely successful, is not in itself evidence of negligence. [ Zoterell, supra, p 330.]

Fifty years later, in a California case, Justice Roger Traynor expounded on these same concepts:

A physician's duty is to exercise that degree of care and skill ordinarily possessed and exercised by members of his profession under similar circumstances. ( Sinz v Owens [1949] 33 Cal.2d 749, 753 [ 205 P.2d 3; 8 ALR2d 757].) He does not guarantee a cure. The doctrine of res ipsa loquitur cannot properly be invoked to make him an insurer of the recovery of persons he treats. The Latin words cannot obliterate the fact that much of the functioning of the human body remains a mystery to medical science and that risks inherent in a given treatment may occur unexplainably though the treatment is administered skillfully. The occurrence of an injury that is a calculated risk of an approved course of conduct, standing alone, does not permit an inference of negligence.

Such an inference must be based on more than speculation. If it is to be drawn from the happening of an accident, there must be common knowledge or expert testimony that when such an accident occurs, it is more probably than not the result of negligence. [ Clark v Gibbons, 66 Cal.2d 399, 421-422; 58 Cal.Rptr. 125; 426 P.2d 525 (1967).]

Thus, in a normal professional negligence case, a bad result, of itself, is not evidence of negligence sufficient to raise an issue for the jury. This does not mean that a bad result cannot be presented by plaintiffs as part of their evidence of negligence, but, rather, that, standing alone, it is not adequate to create an issue for the jury. Something more is required, be it the common knowledge that the injury does not ordinarily occur without negligence or expert testimony to that effect.

In a case where there is no expert evidence that "but for" negligence this result does not ordinarily occur, and in which the judge finds that such a determination could not be made by the jury as a matter of common understanding, a prima facie case has not been made, and a directed verdict is appropriate. However, if there is such evidence, even if it is disputed, or if such a determination could be made as a matter of common understanding, the jury is to determine whether plaintiff has proven whether it is more likely than not that defendant's negligence caused plaintiff's injury. The court may grant a motion for a directed verdict only if it is determined that reasonable minds could not differ that this result could ordinarily happen without negligence.

In cases where reasonable minds could differ as to whether this result could ordinarily happen "but for" negligence, the plaintiff is entitled to have the case submitted to the jury, which, acting in its traditional factfinding role, will decide the liability issue. In appropriate cases, the defendant is entitled to have the jury advised in some form that there are "risks inherent in . . . treatment," Clark v Gibbons, supra, p 422, and that an unfortunate result, standing alone, does not establish a lack of proper care, skill, or diligence. In this case, however, where the proofs have established a res ipsa case for the jury, such a statement as that given is, without additional guidance for the jury, at best, an incomplete instruction on the function of res ipsa loquitur in creating a permissible inference from circumstantial evidence. A res ipsa case is a circumstantial evidence case. In a proper res ipsa loquitur medical case, a jury is permitted to infer negligence from a result which they conclude would not have been reached unless someone was negligent. See Pederson v Dumouchel, 72 Wn.2d 73; 431 P.2d 973; 31 ALR3d 1100 (1967).

We agree with plaintiffs' observation in Dziurlikowski that, when coupled with the instruction that an adverse result is not evidence of negligence, the instruction had the potential to mislead the jury as to a permissible inference which the jury might draw from the plaintiffs' proofs. While it is correct to state that a bad result alone does not raise an inference of negligence, it is also an accurate statement that the bad result did permit the jury to infer negligence if it credited the testimony of plaintiffs' expert witness that such a result did not ordinarily occur in the absence of someone's negligence. The testimony of that witness was such that the trier of fact could have concluded that the defendant was negligent and the jury should have been instructed in some form to enable it to infer negligence if it found that the proofs supported the conclusion that the injury would not have occurred otherwise, see Miller v Kennedy, 11 Wn. App. 272; 522 P.2d 852 (1974). Moreover, such language would be an incorrect statement of law in those classes of medical malpractice cases where common experience indicates that injury does not occur if there is proper skill and care. Where such an instruction is granted, plaintiffs also have a right to have the jury instructed on all of the aspects of a res ipsa case, not just those favorable to the defendant. See, e.g., Kennedy v Gaskell, 274 Cal.App.2d 244; 78 Cal.Rptr. 753 (1969).

Plaintiffs attempted to prove that "but for" negligence, such a complication does not ordinarily occur after surgery. As the basis for this contention was not in the general knowledge of the lay person, plaintiffs were required to, and did, present expert testimony by Dr. Mervin Jeffries to this effect. Although defendants disputed the testimony of plaintiffs' expert, the jury, if it had believed Dr. Jeffries, could have found that "but for" negligence this result would not ordinarily occur. However, while the jury was given the standard circumstantial evidence instruction, SJI2d 3.10, the jury was not instructed on any of these points, and, rather, was told that it could not find that the injury was evidence of negligence. We agree with plaintiffs that this instruction substantially "interfered with the circumstantial inference of negligence which Plaintiffs were entitled to have the jury consider," a material part of this case. Thus, we agree that plaintiffs have demonstrated that affirmance of the jury's verdict in Dziurlikowski would be "inconsistent with substantial justice," MCR 2.613, and we affirm the decision of the Court of Appeals granting the plaintiffs a new trial.

V

ADDITIONAL ISSUES IN DZIURLIKOWSKI

Having affirmed the decision of the Court of Appeals to grant plaintiffs a new trial in Dziurlikowski, we note that plaintiffs have raised two other issues. First, it is claimed that the trial court's refusal to allow impeachment of the defendant's expert by the use of an article in Anesthesiology was error. While we are not entirely convinced that the ruling was incorrect, we do not address that issue in this case. We agree with the Court of Appeals when it stated:

The trial court expressly stated that it would reconsider its decision if plaintiffs presented some authority in support of their position within the following two days. Plaintiffs' attorney indicated satisfaction with that ruling and remarked that he would notify opposing counsel the next day if he planned to pursue the matter further. The issue was never reintroduced, leading us to conclude that plaintiffs abandoned this issue as a matter of trial strategy. For us to reverse on this basis now might encourage parties to harbor error on the record to be used in the event of an unfavorable verdict. [ 143 Mich. App. 734.]

Under the circumstances of this case, we find that this issue has not been properly preserved for appellate review and thus decline to consider it.

Finally, plaintiffs in Dziurlikowski claim that the jury instruction given in this case that charged that the jurors may not arrive at a verdict "by the process of guess, conjecture or speculation" was erroneous. They further claim that, because one of the defendants' witnesses explained his inclusion of the words "presumed brachial plexus injury from hyperextension" in the discharge summary as speculation, their case was sufficiently prejudiced, on this ground alone, to require reversal. We disagree. The instruction correctly informed the jury as to its responsibility and only charged that it may not use speculation to arrive at the verdict, a correct statement of the law. It did not charge that an opinion by a doctor on a report, which was arrived at entirely or partially by speculation, must be disregarded, and the jurors were instructed that they must determine which witnesses to believe and how much weight to give testimony. The jury was further charged:

In doing so you may take into account each witness' ability and opportunity to observe his or her memory, his or her manner while testifying, any interest, bias or prejudice he or she may have, and the reasonableness of his or her testimony considered in the light of all the evidence in the case.

Under the circumstances, we are not persuaded that this additional instruction was error in this case.

VI

CONCLUSION

We affirm the decision of the Court of Appeals in Dziurlikowski and remand the case to the trial court for a new trial. In Jones, the decision of the Court of Appeals is reversed, and the jury's verdict is affirmed.

RILEY, C.J., and LEVIN, BRICKLEY, CAVANAGH, and ARCHER, JJ., concurred with BOYLE, J.

GRIFFIN, J., took no part in the decision of this case.