The opinion of the court was delivered by STEIN, J.

The primary issue in this appeal concerns whether it was error for the trial court, following a three-day jury trial, to dismiss plaintiffs' complaint. That complaint alleged that the treating obstetrician had negligently failed to perform certain diagnostic tests, thereby increasing the risk of harm from a preexistent condition, which resulted in the ultimate demise of a nearly fullterm fetus. The trial court acknowledged that plaintiffs had met their burden of proof concerning negligence, preexistent condition, and the potential for survival of the fetus had an early induction of labor been performed by the treating obstetrician. However, that court dismissed plaintiffs' complaint on the basis that plaintiffs had failed to prove a proximate causal relationship between defendant's alleged malpractice and the death of the fetus.

The Appellate Division affirmed the trial court's dismissal of the complaint. 285 N.J. Super. 113, 666 A.2d 592 (1995). That court concluded that plaintiffs had presented inadequate evidence on the issue of causation because their medical expert failed to express an opinion regarding (1) the fetus's condition on the day the obstetrician failed to perform certain diagnostic tests and (2) the probability or possibility that those tests would have revealed any abnormalities as of that date. We granted certification. 146 N.J. 496, 683 A.2d 199 (1996). We hold that the trial court erroneously precluded the jury from determining whether the obstetrician's failure to perform diagnostic tests increased the risk that plaintiffs' fetus would not survive and whether that increased risk was a substantial factor in causing the fetus's death.

I

On June 1, 1988, plaintiff Linda Gardner, six weeks pregnant, consulted defendant Dr. Myron Pawliw, a board-certified obstetrician and gynecologist. Gardner had sought a recommendation from the Princeton Medical Center for an obstetrician with experience treating high-risk pregnancies and infertility problems. She was referred to Dr. Pawliw. During the initial meeting, Gardner explained in detail her prior gynecological and obstetrical history, which included two first-trimester miscarriages and ongoing treatment for a hormonal insufficiency known as luteal phase defect. She expressed her desire for "special care" to help maintain this pregnancy, emphasizing that she was quite anxious that she might suffer another miscarriage. Dr. Pawliw assured her that he had had experience treating previously infertile and high-risk patients. He stated that he could help her achieve a successful pregnancy.

Gardner received continued supervision and hormonal monitoring by a fertility clinic in Philadelphia during the first trimester of her pregnancy. That supervision was necessary because Gardner had taken fertility drugs to help achieve the pregnancy. Her use of fertility drugs had required close monitoring by the fertility clinic of her hormonal levels and follicle development before she became pregnant. Following Gardner's discovery that she was pregnant, the clinic continued to monitor her progesterone level to help support the pregnancy, as required by her luteal phase defect. Gardner travelled twice weekly to the Philadelphia clinic for testing of her hormone progesterone serum levels. In the first trimester of pregnancy progesterone is normally produced by the corpus luteum, the follicle that remains following the release and fertilization of an ovum. Women with luteal phase defect fail to produce sufficient progesterone to maintain a pregnancy, and thereby require a progesterone supplement, which can be ingested orally, inserted vaginally or injected by syringe. Throughout the first trimester, Gardner was required to use all three methods to sustain sufficient levels of progesterone.

Gardner continued to receive close monitoring of her progesterone levels beyond the first trimester, when the placenta normally takes over the production of the progesterone necessary to sustain the pregnancy. The clinic continued to monitor Gardner's progesterone levels by performing blood tests two times a week and supplemented her progesterone levels throughout the pregnancy. The parties dispute whether standard obstetrical protocol for luteal phase defect includes continued supplementation of progesterone beyond the first trimester, as was provided to Gardner. Dr. Pawliw was aware of the clinic's treatment of Gardner, but did not assume responsibility for its management.

From June through November, Dr. Pawliw saw Gardner on a monthly basis. Her due date was January 20, 1989. Defendant signed orders requiring Gardner to take a medical disability leave for the duration of the pregnancy. Reports from the fertility clinic following its performance of two ultrasound evaluations and an amniocentesis indicated normal fetal development. Early in the third trimester, Gardner saw Dr. Pawliw's associate after she experienced contractions. The contractions dissipated and Gardner's pregnancy continued normally. Beginning in the third trimester, Gardner saw Dr. Pawliw every two weeks.

On December 21, 1988, Gardner noticed a significant decrease in the frequency of fetal movement and a watery vaginal discharge. She had a regularly scheduled appointment with Dr. Pawliw later that day. Gardner testified that she had felt concerned during the appointment and that defendant had acted impatiently towards her. Dr. Pawliw performed an internal examination and determined that Gardner's cervix was partially effaced and dilated. He measured the height of the fundus, which serves as an indicator of fetal growth based on the growth of the uterus. Defendant also took Gardner's blood pressure and performed a urinalysis, which appeared normal. Gardner testified that Dr. Pawliw had had difficulty finding the fetal heartbeat, and that the fetal heartbeat had sounded different than at earlier visits. Dr. Pawliw indicated in his notes that he had found the fetal heartbeat and that it was normal. When Gardner voiced her concerns about the amount of time it took Dr. Pawliw to find the fetal heartbeat, Dr. Pawliw responded curtly, asking if she had always counted the time it took him to find the heartbeat. He told Gardner not to get too worried, warning her that she could develop high blood pressure if she was not careful. When Gardner continued to voice her concerns about the decreased movement, defendant attempted to reassure her by explaining that the fetus was sleeping and that there was nothing wrong.

Defendant did not perform any other diagnostic tests during that visit. The parties dispute whether defendant provided Gardner with any verbal instructions to count the episodes of movement that she experienced from morning until noon, and to call him if she failed to perceive at least ten episodes during that time period. Gardner contends that she received no such instructions, emphasizing that she would have followed any instructions to the letter given her anxiety concerning the fetus's well-being. Gardner acknowledged that defendant had given her written instructions concerning when to call him (if she had frequent contractions, started bleeding, her water broke or if she did "not know what was going on."). At the end of the visit, Gardner was reassured by defendant's remark that she might deliver the fetus as early as Christmas.

From December 21 through December 24, 1988, the same level of decreased fetal movement continued. On December 25, Gardner noticed increased fetal activity. On December 26, Gardner did not notice any fetal movement and thought, based on the explanation previously offered by Dr. Pawliw, that the fetus was sleeping. On the morning of December 27, Gardner and her husband perceived a complete lack of fetal activity. She immediately called defendant's office and was told to come in to be examined. Dr. Pawliw failed to find the fetal heartbeat, although he did not so inform the Gardners. When Gardner suggested that the heartbeat that he had found was hers, not the fetus's, Dr. Pawliw told her that she was wrong. He sent the couple to a nearby facility for an ultrasound scan and biophysical profile. The results were not provided for several hours, although both Gardner and her husband made numerous phone calls to defendant's office. Defendant eventually informed the Gardners by telephone that the fetus had died and asked them both to come to his office later that evening to discuss the various options available for delivering the dead fetus. The Gardners elected to induce labor that evening by the use of prostaglandin suppositories. Dr. Pawliw told them that he would be there to help them through the delivery, but he missed the delivery. Instead, Gardner's husband and a nurse delivered the stillborn fetus.

On March 20, 1990, the Gardners filed a medical malpractice action against Dr. Pawliw, alleging that his failure to perform diagnostic tests, specifically a Non-Stress Test (NST) and Biophysical Profile (BPP), on December 21, 1988, after Gardner informed him of the decreased fetal movement and watery discharge, had caused an increased risk that their fetus would not survive and that the increased risk was a substantial factor in the fetus's death. Plaintiffs also alleged that Dr. Pawliw's negligent failure to deliver the fetus had caused Linda Gardner's husband emotional distress. A jury trial was held between June 7 and June 10, 1994.

At trial, the Gardners presented testimony by two expert witnesses, Dr. James Lewis, a forensic pathologist, and Dr. Marvin Kalafer, a board-certified obstetrician and gynecologist. Dr. Lewis testified that the autopsy results had indicated that the fetus had died because of umbilical cord and placenta abnormalities. The umbilical cord had been shorter than normal and had attached on the edge, or margin, of the placenta, not in the center where it normally attaches. Additionally, the cord had been in a "filamentous" state, which had resulted in a superficial attachment to the placenta. As a result, the fetus had received an insufficient amount of nutrients and oxygen from the placenta, which eventually had resulted in its death. See 4C Roscoe N. Gray Louise J. Gordy, Attorney's Textbook of Medicine § 305.51 (3d ed. 1990).

On cross-examination, Dr. Lewis noted that several of the fetus's internal organs, namely her heart, adrenal glands and kidneys had been smaller than average, falling within the lowest percentiles of fetal development at that stage of gestation. The fetus itself was of normal size. Dr. Lewis opined that the fetus had died on December 24 or 25, 1988. He concluded that the fetus "had to be" alive on December 21, 1988, due to the state of tissue samples analyzed during the autopsy.

Dr. Kalafer's expert testimony was presented by videotaped deposition. The Appellate Division described in detail the content of Dr. Kalafer's testimony and the issues that it generated:

Dr. Kalafer testified that Mrs. Gardner's "marked decrease in fetal activity" was "very significant" because "[i]t is something that often alerts the physician that there might be a potential problem." He also expressed the opinion that defendant's failure to perform any tests when Mrs. Gardner reported a decrease in fetal activity constituted a deviation from the accepted standard of prenatal care:

I feel that . . . in this setting, a patient who is considered a high-risk individual who has had miscarriages in the past who has received extra medication during pregnancy, that it was important to follow-up on that particular complaint of a patient and perhaps perform a study which would help determine if the baby was in an environment that was normal. Specifically, I feel it would be important to perform a nonstress test along with the biophysical profile at that time.

When asked whether the baby would have survived if it had been delivered on December 21, Dr. Kalafer testified:

I don't know. However, if one would have performed a study and if that study indicated that there was a smoldering in utero environment, one would have acted on that and most likely the baby would have been able to survive on the outside being given the fact that even though this is not a term baby, it is very close to term; and in the setting such as a neonatal intensive care unit, one could monitor the baby, and the biggest concern really at that point might be lung maturation.

Plaintiffs' counsel then asked the following question:

Do you have an opinion held to a reasonable degree of medical probability as to whether it was probable that if Dr. Pawliw performed tests on the day of December 21, that it would have shown any effects of the placental and cord abnormalities?

After an objection by defendant's counsel and colloquy among counsel and the witness, Dr. Kalafer responded:

My answer is yes.

. . . .

I feel that if a study would have been performed, one could utilize that knowledge to help better manage the pregnancy. If the test would have been normal, it would have been reassuring. If the test would have been abnormal, then I believe one would have been pushed to deliver this baby."

When asked for his opinion as to whether defendant's failure to conduct testing on December 21 increased the risk that the baby would die, Dr. Kalafer stated:

If a study would have been performed such as the nonstress test and biophysical profile, one could utilize that information; and if it was abnormal, the baby could have been delivered most likely in a setting more appropriately, and the baby could most likely live in the outside world.

And when asked whether the baby's death was substantially caused by defendant's failure to conduct testing on that date, Dr. Kalafer answered:

[I]f these studies were performed, it would have helped the doctor to continue the in utero life or basically either decide that this baby does not have to be delivered or just continue close monitoring.

On cross-examination, Dr. Kalafer gave the following testimony regarding the tests that he asserts defendant should have conducted on December 21:

Q. Now, a nonstress test is either reactive or nonreactive, is that correct, sir?

A. That is how it is qualified, yes.

Q. Would it be fair to say, sir, that you cannot state within a reasonable degree of medical probability that a nonstress test, had one been done on December 21, 1988, would have been nonreactive?

A. I cannot state that, that is right, since one wasn't done.

. . . .

Q. Correct me if I am wrong, doctor, but it is my understanding that a biophysical profile is either reassuring or non-reassuring, is that the terminology that you used, sir?

A. That is one way to describe it, but you can also quantify it perhaps a little bit better.

Q. Would it be fair to say, sir, that you cannot state within a reasonable degree of medical probability that had a biophysical profile been performed on December 21, 1988 it would have been non-reassuring?

A. I can't state that because one wasn't done.

On re-direct examination, Dr. Kalafer testified:

Q. Doctor, you said that you cannot state with certainty or probability what a nonstress test would have shown if performed on December 21. I want to ask you: Can you say to a reasonable degree of medical probability as to whether by failing to do a nonstress test there was an increased risk that a condition which could cause the baby's death would not be recognized?

A. Yes.

. . . .

Q. Similarly you said that you cannot state with medical probability what a biophysical profile would have shown if performed on December 21. I want to ask you: Can you state with a reasonable degree of medical probability as to whether by failing to do a biophysical profile there was an increased risk that a condition which could cause the baby's death would not be recognized?

. . . .

A. Yes.

[ 285 N.J. Super. at 116-18, 666 A.2d 592 (emphasis added).]

Dr. Kalafer also stated that although there were no discernable fetal abnormalities, the autopsy report revealed that there were placental abnormalities and a short cord. When asked whether those abnormalities were significant he stated, "I think this could help explain perhaps why the patient felt decreased movement. It might have been indicative of the fact that this placenta was not nourishing the baby appropriately."

Defendant presented expert testimony from Dr. Stanley Wilchins, a board-certified obstetrician and gynecologist. Dr. Wilchins testified that Gardner's complaints concerning a decrease in fetal activity had not been significant because "as all pregnancies approach toward the end of the third trimester, there normally is a decrease in fetal movement. But in association with other medical conditions, it can be of significance." Dr. Wilchins also testified that the content of any watery discharge at this point in a pregnancy should be analyzed to determine its source. However, he concluded that Dr. Pawliw had not deviated from accepted standards of medical care by not ordering the NST or BPP tests on December 21, 1998, "because there was absolutely nothing going on which warranted it."

When asked whether, if the tests had been performed on December 21, 1988, the results would have been abnormal, Dr. Wilchins testified that "not as a probability but as a medical certainty" the test results would have been "perfectly normal." He based this conclusion on two findings. First, because the ultrasound performed on December 27, 1988, showed a normal amount of amniotic fluid, he opined that the fetus could not have been compromised on December 21, 1988. Second, based on the autopsy finding that the fetus and placenta exhibited only faint meconium staining, he opined that the in utero environment could not have been very stressed on December 21, 1988. Had it been stressed, the meconium staining would have been markedly darker. Accordingly, he concluded that the fetus had been severely stressed only twenty-four to forty-eight hours before death, that is on December 25 or 26, "not a week earlier on the 21st."

Dr. Wilchins further testified that neither an NST nor a BPP test would have revealed the presence of an abnormally short cord or of a marginal insertion of the cord into the placenta. He testified that those tests merely reveal whether a fetus is experiencing stress at the moment the test is given, not whether stress may occur at some point in the future:

THE COURT: Well what if the baby was showing stress, would that be — sometimes be reflected?

THE WITNESS: Stress does turn up sometimes when tests are performed, that is correct. But that's an ongoing present event occurring now. In other words, the fetus has to be experiencing stress now in order for that test to reveal stresses going on. It doesn't say that stresses will occur three or four or five days from now.

When asked whether the tests would have revealed fetal distress if they had been administered on December 21, Dr. Wilchins replied: "I think to the contrary. Based upon the findings of the normal amniotic fluid and the faint meconium, I think that that testing would have showed up perfectly normal or one could not have had a normal amniotic fluid index and the just faint staining of meconium." When asked by plaintiffs' counsel whether a Level II ultrasound would have revealed smaller-sized internal organs or other abnormalities, Dr. Wilchins stated, "[I]t may, it may not."

At the close of defendant's case, defense counsel moved for dismissal, pursuant to Rule 4:40-1, which the trial court granted. The trial court found that plaintiffs could reach the jury on the issue of negligence, because their expert had testified that a failure to order the tests had deviated from the prevailing standard of care. Additionally, the trial court found that it was unrefuted that the fetus's death resulted from the pre-existing condition of the placenta and umbilical cord abnormalities. Furthermore, it accepted Dr. Kalafer's testimony as sufficient to reach the jury on whether the fetus would have survived had Dr. Pawliw induced labor on December 21, 1988.

However, the trial court held as a matter of law that plaintiffs had failed to show a causal connection between the failure to administer the tests and the death of the fetus, noting the limitations in Dr. Kalafer's testimony:

Now the question, was there any reason to have any intervention. And the testimony is unrefuted that unless there was a negative test there was no reason for medical intervention. So the critical question [is], . . . "Do you have an opinion held to a reasonable degree of medical probability as to whether it was probable that if Dr. Pawliw performed tests on the day of December 21st that it would have shown any effects of the placental and cord abnormalities?"

[Defendant's attorney] objects. Why he objects, I don't know. But he decides to interrupt, because it's a critical question to the answer. . . .

But Ms. Rhoads — the next thing that happens is [plaintiffs' attorney] says, "Can you answer that?" And [Dr. Kalafer] says, "No, I can't answer that question."

The court noted that in response to this answer plaintiffs' counsel had argued that both counsel and the witness were confused, and that counsel had asked the question again to the witness:

"Can you say to a degree of medical certainty, . ..

"My answer is yes."

"Why is that?"

"I feel that if the study would have been performed one could utilize that knowledge to help manage the pregnancy." Not the question. We're not managing the pregnancy. It is a clear issue, rather like taking the X-ray on the day in question. "If the test would have been normal it would have been reassuring." Of course. "If the test would have been abnormal, then I believe one would have been pushed to deliver the baby." Agreed. Unrefuted.

In effect, the trial court found there was an insufficient showing of probability that the tests in fact would have shown abnormalities, even though there was a possibility of such a result:

You need an abnormal test, Dr. Kalafer has clearly said, to push for the delivery. That's the critical issue. Had there been an abnormal test, and there was a possibility — unrefuted. Had this doctor listened to Mrs. Gardner, had he determined at that point to have a test, there is a possibility that it would have been abnormal, and that would have caused this doctor to push for delivery of the baby. And Mr. and Mrs. Gardner have every right to want every possibility explored. If there was a one percent chance that that test would have shown this baby in stress and there could have been a careful intervention, I understand exactly why they would have wanted it done. . . . That's not the standard the law requires of the doctor. . . . That's not the test for the law. The law says there has to be a degree of medical probability that the tests not taken would have provided guidance to the physician and that his failure to take that test then is a proximate cause of the injury which follows, the death of the child.

[Emphasis added.]

Although the court did not state what minimum threshold would have established a medical probability in the present case, it found that Dr. Kalafer had not reached the threshold because he had declined to state definitively what the tests would have shown:

[H]e refuses to answer the question. Now it is a very interesting answer. We don't know what the result would have been, therefore I won't say whether to a reasonable degree of medical probability an abnormal test would have refer [sic]. There might have been a one percent chance or a 10 — by the way, the doctor's never asked to quantify it. Let's say it's 50-50. Is that to a degree of medical probability? I don't know. I don't reach it. It is not enough for a doctor to say, I cannot state since the test wasn't done whether or not there would have been an abnormal test. He is refusing to answer the question, and thus in not answering is not providing the basis that is necessary for the plaintiff to get to the Jury.

. . . .

So the doctor refuses to give the critical — that there is to a degree of medical probability, not 100 percent — and by the way, I don't know that the law has ever required — what is probability? Fifty-one percent? Maybe I find it enough to be 40 percent. I don't know. I'm not confronted with that. Some probability that this test, not some possibility the test.

The court therefore concluded that Dr. Kalafer's testimony was insufficient to establish the necessary causative link:

So we have one doctor who will not answer the question in deposition as to whether this test, to a degree of medical probability. It might have turned out abnormal. That's the best we have. There is a possibility it would have been abnormal. Possibility in the law is not enough. There must be a probability that this test would have been normal. And I don't know how to quantify that. Maybe 30 percent is enough. Maybe 40 percent. We don't have that testimony. But a possibility is not enough. Unless there was an abnormal test there would be no reason for intervention by this gynecologist in an early delivery. That is clear from all of the testimony. And that's not defeated. It becomes critical, then, as a matter of proximate cause, a causative link under Scaffiti [sic] that the doctor's failure to take this test had to increase the risk that the child's stress would be undiagnosed and timely intervention not had. And that's the testimony Dr. Kalafer will not give.

[Emphasis added.]

The court indicated that the only evidence on the possible outcome of the tests was Dr. Wilchins's testimony "to a medical certainty" that they would have been negative. On June 17, 1994, the trial court entered judgment in favor of defendant.

The Appellate Division affirmed the trial court's dismissal of plaintiffs' complaint, concluding that plaintiffs had failed to present adequate evidence that defendant's failure to perform the tests on December 21, 1988, had increased the risk of harm posed by the preexistent condition. 285 N.J. Super. at 122, 666 A.2d 592. The holding reflected the Appellate Division's application of the modified standard of causation, adopted by this Court in Evers v. Dollinger, 95 N.J. 399 , 471 A.2d 405 (1984), and Scafidi v. Seiler, 119 N.J. 93 , 574 A.2d 398 (1990), that controls cases where a physician's medical malpractice combines with a patient's preexistent condition to cause harm. The Appellate Division articulated the plaintiff's burden as follows:

Under this modified causation standard, plaintiff must present "[e]vidence demonstrating within a reasonable degree of medical probability that negligent treatment increased the risk of harm posed by a preexistent condition." If plaintiff satisfies this burden, it "raises a jury question whether the increased risk was a substantial factor in producing the ultimate result."

[ 285 N.J. Super. at 121, 666 A.2d 592 (quoting Scafidi, supra, 119 N.J. at 108, 574 A.2d 398).]

The Appellate Division found that plaintiffs had failed to satisfy their burden of proof because they had not presented "any evidence" to show the required causal connection between defendant's alleged malpractice and the death of the fetus. Id. at 122, 666 A.2d 592. The Appellate Division noted that the only expert opinion offered at trial on the issue of the fetus's condition on December 21, 1988, was that of defendant's expert, Dr. Wilchins, who had opined that the tests would not have revealed any abnormality in the fetus. In contrast, plaintiffs' medical expert, Dr. Kalafer, had declined to offer any opinion concerning the state of the fetus on December 21, 1998. Therefore, the panel concluded that plaintiffs could not prove that, had defendant performed the NST and BPP tests on December 21, 1988, a stressful environment would have been revealed, thereby necessitating induction of labor to save the fetus. Ibid. The court acknowledged Dr. Kalafer's affirmative answer to the question concerning whether the failure to perform the tests increased the risk that the doctor would fail to recognize a condition that could cause the fetus's death. However, the court characterized that answer as a "bare conclusion," which was unsubstantiated because of Dr. Kalafer's inability to address the fetus's condition on December 21, 1988, due to the lack of the NST and BPP test results. Id. at 122-23, 666 A.2d 592. The Appellate Division determined that a remote possibility or slight chance that the failure to order the tests increased the risk of ultimate harm was insufficient to meet plaintiffs' burden of proof of causation. Id. at 123, 666 A.2d 592. It concluded that if the tests would not have revealed an abnormality, then the failure to perform them would not have resulted in an increased risk of harm to the fetus.

II

To establish a prima facie case of negligence in a medical-malpractice action, a plaintiff must present expert testimony establishing (1) the applicable standard of care, see Rosenberg v. Cahill, 99 N.J. 318 , 325, 492 A.2d 371 (1985); (2) a deviation from that standard of care, see Clark v. Wichman, 72 N.J. Super. 486 , 496, 179 A.2d 38 (App.Div. 196 2); and (3) that the deviation proximately caused the injury, see Germann v. Matriss, 55 N.J. 193 , 205, 260 A.2d 825 (1970). This Court has lessened the traditional burden of proof on a plaintiff asserting a medical-malpractice claim for establishing proximate cause in the case of a plaintiff suffering from a preexistent condition. See Anderson v. Picciotti, 144 N.J. 195, 205-07, 676 A.2d 127 (1996); Scafidi, supra, 119 N.J. at 108-09, 574 A.2d 398; Evers v. Dollinger, 95 N.J. 399, 417, 471 A.2d 405 (1984). A plaintiff suffering from a preexistent condition must prove that, as a result of a defendant's negligence, she experienced an increased risk of harm from that condition, and that the increased risk of harm was a substantial factor in causing the injury ultimately sustained. Anderson, supra, 144 N.J. at 210, 676 A.2d 127.

The majority of jurisdictions has similarly modified the traditional "but for" causation standard of proof in cases where the injury allegedly resulted in part from a defendant's negligence and in part from a preexistent condition to permit such plaintiffs to submit for jury consideration the questions of whether the defendant's deviation from standard medical practice increased a patient's risk of harm or diminished a patient's chance of survival and whether such increased risk was a substantial factor in producing the ultimate harm. See, e.g., McBride v. United States, 462 F.2d 72, 75 (9th Cir. 1972) (applying Hawaii law); O'Brien v. Stover, 443 F.2d 1013, 1017-18 (8th Cir. 1971) (applying Iowa law); Hicks v. United States, 368 F.2d 626, 632-33 (4th Cir. 1966) (applying Virginia law); James v. United States, 483 F. Supp. 581 , 585-87 (N.D.Cal. 1980); Thompson v. Sun City Community Hosp., Inc., 141 Ariz. 597, 688 P.2d 605, 613-16 (1984); Northern Trust Co. v. Louis A. Weiss Mem'l Hosp., 143 Ill. App.3d 479, 97 Ill.Dec. 524, 529-30, 493 N.E.2d 6, 11-12 (1986); Mayhue v. Sparkman, 653 N.E.2d 1384, 1389 (Ind. 1995); DeBurkarte v. Louvar, 393 N.W.2d 131, 135-38 (Iowa 1986); Roberson v. Counselman, 235 Kan. 1006, 686 P.2d 149, 159-60 (1984); Aasheim v. Humberger, 215 Mont. 127, 695 P.2d 824, 827-28 (1985); Roberts v. Ohio Permanente Med. Group, Inc., 76 Ohio St.3d 483, 668 N.E.2d 480, 482-84 (1996); McKellips v. Saint Francis Hosp., Inc., 741 P.2d 467, 475 (Okla. 1987); Hamil v. Bashline, 481 Pa. 256, 392 A.2d 1280, 1284-89 (1978); Herskovits v. Group Health Coop. of Puget Sound, 99 Wn.2d 609, 664 P.2d 474, 477-79 (1983). See generally John D. Hodson, Annotation, Medical Malpractice: "Loss of Chance" Causality, 54 A.L.R.4th 10 (1987) (listing various rules applied by jurisdictions allowing substantial factor standard of causation).

This Court first adopted the doctrine of increased risk in medical-malpractice litigation in Evers v. Dollinger, supra, in which the plaintiff alleged that a physician negligently failed to perform diagnostic tests that would have revealed a malignant tumor in her breast. 95 N.J. at 404, 471 A.2d 405. As a result of the physician's alleged negligence, the plaintiff's tumor grew unchecked for a period of seven months. The plaintiff claimed that, because of the delay in detection and treatment, she suffered an increased risk that the cancer would spread. After receiving treatment, the cancer returned and metastasized throughout her body. Ibid.

In a routine tort claim, the law requires proof by a preponderance of the evidence that the injury complained of probably would not have occurred "but for" the negligent conduct of the defendant. Id. at 415, 471 A.2d 405; see W. Page Keeton, et al., Prosser and Keeton on the Law of Torts § 41 (5th ed. 1984). The Evers Court recognized that proximate cause can be difficult to prove in the context of harm that results from a combination of a plaintiff's preexistent condition and a defendant's negligent discharge of a duty related to that condition, because the preexistent condition itself serves as a "but for" cause of the ultimate injury. However, the physician's duty encompasses the protection of the patient against harm from preexistent conditions. See Evers, supra, 95 N.J. at 414, 471 A.2d 405. In Evers, therefore, the Court adopted the doctrine of increased risk, establishing a more flexible standard for proving causation in cases concerning an increased risk of injury from a preexistent condition than that used in conventional tort claims, recognizing the "difficulties of identifying, defining and proving injury" in such cases. Id. at 413, 471 A.2d 405.

In adopting a more flexible burden of proof for a plaintiff suffering from a preexistent condition, this Court relied in part on section 323(a) of the Restatement (Second) of Torts (1965), which provides in relevant part:

One who undertakes, gratuitously or for consideration, to render services to another which he should recognize as necessary for the protection of the other's person or things, is subject to liability to the other for physical harm resulting from his failure to exercise reasonable care to perform his undertaking, if

(a) his failure to exercise such care increased the risk of such harm.

The Evers Court also relied on the reasoning expressed by the Pennsylvania Supreme Court in Hamil, supra, which approved the application of the rule of law stated in the Restatement (Second) of Torts § 323(a) to medical-malpractice cases:

[D]efendant was charged with having failed in a duty to protect against harm from another source; hence the fact-finder must consider not only what did occur but also what might have occurred:

Such cases by their very nature elude the degree of certainty one would prefer and upon which the law normally insists before a person may be held liable. Nevertheless, in order that an actor is not completely insulated because of uncertainties as to the consequence of his negligent conduct, Section 323(a) [of Restatement (Second) of Torts] tacitly acknowledges this difficulty and permits the issue to go to the jury upon a less than normal threshold of proof.

[ Evers, supra, 95 N.J. at 415, 471 A.2d 405 (quoting Hamil, supra, 392 A.2d at 1287-88 (footnote omitted)).]

The Evers Court ordered a remand to the trial court to permit the plaintiff to

demonstrate, within a reasonable degree of medical probability, that the seven months delay resulting from defendant's failure to have made an accurate diagnosis and to have rendered proper treatment increased the risk of recurrence or of distant spread of plaintiff's cancer, and that such increased risk was a substantial factor in producing the condition from which plaintiff currently suffers.

[ Id. at 417, 471 A.2d 405]

In the companion cases of Scafidi, supra, and Olah v. Slobodian, 119 N.J. 119 , 574 A.2d 411 (1990), the Court further developed the doctrine of increased risk, establishing a two-prong test for proving causation in cases where a plaintiff's injuries are due in part to a preexistent condition. The Court in Scafidi, supra, held that "[e]vidence demonstrating within a reasonable degree of medical probability that negligent treatment increased the risk of harm posed by a preexistent condition raises a jury question whether the increased risk was a substantial factor in producing the ultimate result." 119 N.J. at 108, 574 A.2d 398. The Court explained:

The rationale underlying the use of a two-pronged jury instruction bears elaboration. Because this modified standard of proximate causation is limited to that class of cases in which a defendant's negligence combines with a preexistent condition to cause harm — as distinguished from cases in which the deviation alone is the cause of harm — the jury is first asked to verify, as a matter of reasonable medical probability, that the deviation is within the class, i.e., that it increased the risk of harm from the preexistent condition. Assuming that the jury determines that the deviation increased the risk of harm from the preexistent condition, we use the "substantial factor" test of causation because of the inapplicability of "but for" causation to cases where the harm is produced by concurrent causes. The "substantial factor" standard requires the jury to determine whether the deviation, in the context of the preexistent condition, was sufficiently significant in relation to the eventual harm to satisfy the requirement of proximate cause.

[ Id. at 108-09, 574 A.2d 398 (citations omitted) (emphasis added).]

When applying the Scafidi test, trial courts must closely analyze the evidence presented to determine whether

it is sufficient to permit a jury to decide, as a matter of reasonable medical probability, that both prongs of the two-part test are satisfied. First, the evidence must permit a jury to find that defendant was negligent and that defendant's negligence increased plaintiff's risk of harm from an established preexistent condition. If that prong is satisfied, then there are concurrent causes of the harm . . . [and] the "but for causation" standard may not be charged to a jury.

[ Anderson, supra, 144 N.J. at 206, 676 A.2d 127.]

If the first prong is satisfied, then the jury will be instructed by the trial court to apply the "substantial factor" standard of causation. Id. at 206-07, 676 A.2d 127. Additionally, the trial court will instruct the jury when it considers damages to apportion the defendant's liability in accordance with comparative negligence principles, on the basis of the percentage of the ultimate harm that directly resulted from the defendant's conduct. Id. at 208, 676 A.2d 127. The tortfeasor will be "`charged only with the value of the interest he [or she] destroyed.'" Scafidi, supra, 119 N.J. at 112, 574 A.2d 398 (quoting Joseph H. King, Jr., Causation, Valuation, and Chance in Personal Injury Torts Involving Preexisting Conditions and Future Consequences, 90 Yale L.J. 1353, 1356 (1981)).

To satisfy the threshold of the first prong of the Scafidi test, a plaintiff must prove to a reasonable degree of medical probability that a physician's deviation from the standard of care increased the risk of harm from the preexistent condition. Id. at 109, 574 A.2d 398. A plaintiff may present proof of "not only what did occur, but what might have occurred." Evers, supra, 95 N.J. at 415, 471 A.2d 405. A plaintiff will typically use expert medical testimony. See Rosenberg, supra, 99 N.J. at 325, 492 A.2d 371. A plaintiff necessarily must establish that a chance of avoiding the harm existed. See Olah, supra, 119 N.J. at 133, 574 A.2d 411.

A complication is presented when a physician's deviation from the prevailing standard of care consists of the failure to perform a diagnostic test, because that very failure to perform the test may eliminate a source of proof necessary to enable a medical expert to testify to a degree of reasonable medical probability concerning what might have occurred had the test been performed. Defendant argues, and the lower courts held, that the failure of plaintiffs' expert to testify that there was a reasonable probability that the tests given on December 21, 1988, would have disclosed the fetus's preexistent condition resulted in a failure to establish the first prong of Scafidi. Plaintiff contends that the lower courts' analysis is too stringent and deviates from the policy underlying Scafidi and Evers. Plaintiff argues that when a physician fails to perform a standard or appropriate medical test, that physician should not benefit from the plaintiff's expert's inability to prove what the unperformed test would have revealed. The question is whether proof that the failure to perform a diagnostic test was a deviation from standard medical practice and that the unperformed test might have yielded information resulting in the avoidance of harm from the preexistent condition is sufficient to demonstrate, as a matter of reasonable medical probability, that the defendant's deviation increased the risk of harm to the plaintiff, thereby satisfying the first prong of Scafidi.

The Supreme Court of Pennsylvania in Hamil, supra, considered the question of the degree of certainty required of expert medical testimony to establish the causal relation between the harm suffered by a patient and the alleged negligence of a health care provider who fails to properly diagnose and treat the plaintiff's preexistent condition in a manner that might have prevented the harm. 392 A.2d at 1282-83. In Hamil, the plaintiff's decedent-husband was suffering from severe chest pains, and the plaintiff brought him to the defendant-hospital's emergency room. Following the hospital's inability to locate the emergency room physician, another physician ordered the performance of an electrocardiogram (EKG). Id. at 1283. When the EKG machine failed to function, the physician ordered that another machine be brought to the emergency room and then proceeded to leave the hospital. Ibid. However, another machine could not be located. Without receiving any treatment or diagnosis from the defendant-hospital, his wife took decedent to a private physician's office, where while undergoing an EKG decedent died. The cause of death was a myocardial infarction. Ibid. The plaintiff's medical expert testified that with proper diagnosis and treatment by the hospital, the decedent would have had a seventy-five percent chance of surviving the attack. Ibid. The defendant's expert testified that decedent would have died irrespective of any treatment provided by the defendant hospital. Ibid. The Pennsylvania Supreme Court held that the plaintiff's expert testimony of a seventy-five percent chance of recovery with prompt treatment was sufficient for the plaintiff to reach the jury for its consideration of the causation question. Id. at 1289.

The Hamil court stated that

Once a plaintiff has introduced evidence that a defendant's negligent act or omission increased the risk of harm to a person in plaintiff's position, and that the harm was in fact sustained, it becomes a question for the jury as to whether or not that increased risk was a substantial factor in producing the harm.

[ Id. at 1286.]

It relied in part on a policy argument articulated by the Fourth Circuit in Hicks, supra:

When a defendant's negligent action or inaction has effectively terminated a person's chance of survival, it does not lie in the defendant's mouth to raise conjectures as to the measure of the chances that he has put beyond the possibility of realization. If there was any substantial possibility of survival and the defendant has destroyed it, he is answerable. Rarely is it possible to demonstrate to an absolute certainty what would have happened in circumstances that the wrongdoer did not allow to come to pass. The law does not in the existing circumstances require the plaintiff to show to a certainty that the patient would have lived had she been hospitalized and operated on promptly.

[ Hamil, supra, 392 A.2d at 1288 (quoting Hicks, supra, 368 F.2d at 632).]

The Hamil court reiterated that the accepted standard for expert medical opinions on causation in Pennsylvania required a "reasonable degree of medical certainty." Id., 392 A.2d at 1288. However, it concluded that "it would be unreasonable and unrealistic in this type of case to expect a physician to state with a `reasonable degree of medical certainty' what might have happened when the [ Restatement (Second) of Torts] (Section 323(a)) recognizes the contingencies involved." Id. at 1288-89.

The Hamil court emphasized that it was the jury's task to balance the probabilities and resolve conflicting testimony, not the medical experts' responsibility. Id. at 1286-88. The Court stated:

Thus where it appears how an accident happened and also that the victim might have saved himself by taking advantage of a precaution which it has been shown defendant negligently failed to afford, courts have generally let a jury find the failure caused the harm, though it is often a pretty speculative matter whether the precaution would in fact have saved the victim.

[ Id. at 1287 (quoting 2 Fowler V. Harper Fleming James, Jr., The Law of Torts § 20.2 at 1113 (1956)).]

In Thompson, supra, the Supreme Court of Arizona considered a claim asserted by a mother on behalf of her thirteen-year-old son, who suffered permanent damage to his leg following a delay in surgery to correct a tear in his femoral artery. 688 P.2d at 608. The delay allegedly was caused by the defendant hospital's transfer of the boy to another hospital because of his family's economic status. Ibid. The plaintiff's expert was unwilling or unable to quantify the chance of complete recovery if surgery had been more promptly performed, but indicated that there would have been a "`substantially better chance' of full recovery had surgery been performed at once." Id. at 615. The defendants argued that the transfer of the patient had caused no damage because his injury might have led to exactly the same residual injury sustained. The defendants contended that even if treated immediately, the plaintiff had at most a five to ten percent chance of complete recovery. Ibid. Like the Hamil court, the Thompson court concluded that flexibility in a plaintiff's burden of proof on the causation question was necessary, emphasizing that the alleged negligence itself had prevented the plaintiff from offering proof of what earlier intervention would have achieved:

Though unwilling or unable to quantify the chance of complete recovery with prompt surgery, plaintiff's experts testified that there would have been a "substantially better chance" of full recovery had surgery been performed at once. . . . [The Restatement rule] permits the case to go to the jury on the issue of causation with less definite evidence of probability than the ordinary tort case. To this extent, no doubt, it permits the jury to engage in some speculation with regard to cause and effect. . . . Defendant's negligent act or omission made it impossible to find with certainty what would have happened and thus forced the court to look at the proverbial crystal ball in order to decide what might have been. Such determinations, of course, have traditionally been the province of the jury rather than the judge.

[ Id. at 615-16 (citing Keeton, et al., supra, § 42).]

The court held that the jury should be permitted to consider whether the delay in surgery increased the risk of harm to the boy and, if it so determined, to "find a `probability' that defendant's negligence was the cause of the damage." Id. at 616.

Other jurisdictions have permitted recovery without requiring either medical certainty or probability concerning what might have happened if a physician had complied with the standard of care. In McKellips, supra, the Supreme Court of Oklahoma observed:

We think in those situations where a health care provider deprives a patient of a significant chance for recovery by negligently failing to provide medical treatment, the health care professional should not be allowed to come in after the fact and allege that the result was inevitable inasmuch as that person put the patient's chance beyond the possibility of realization. Health care providers should not be given the benefit of the uncertainty created by their own negligent conduct. To hold otherwise would in effect allow care providers to evade liability for their negligent actions or inactions in situations in which patients would not necessarily have survived or recovered, but still would have a significant chance of survival or recovery.

[ 741 P.2d at 474 (emphasis added).]

See also, e.g., Mayhue, supra, 653 N.E.2d at 1388-89 (adopting Restatement (Second) of Torts § 323(a) approach based in part on reasoning that health care providers should not benefit from uncertainty created by their own negligent conduct); Roberts, supra, 668 N.E.2d at 484 (adopting rule stated in Restatement (Second) of Torts § 323(a) and holding that "[i]n order to maintain an action for the loss of a less-than-even chance of recovery or survival, the plaintiff must present expert medical testimony showing that the health care provider's negligent act or omission increased the risk of harm to the plaintiff. . . . The plaintiff is not required to establish the lost chance of recovery or survival in an exact percentage in order for the matter to be submitted to the jury."); King, Jr., supra, at 1354 ("Preexisting conditions must, of course, be taken into account in valuing the interest destroyed. When those preexisting conditions have not absolutely preordained an adverse outcome, however, the chance of avoiding it should be appropriately compensated even if that chance is not better than even."). The prevailing view appears to be that if a physician deprives a plaintiff of the ability to quantify the effect of the physician's negligence, the patient should not lose the opportunity for the jury to consider and weigh the proofs presented if there is expert testimony that the physician's deviation increased the patient's risk of harm.

Several courts have found that plaintiffs have satisfied the requisite burden of proving increased risk of harm or lost chance of survival if an expert has testified that manifesting symptoms should have been detected and referred for diagnosis and treatment, even when the exact increase in risk or decrease in a chance for survival could not be quantified. See, e.g., James, supra, 483 F. Supp. at 586-87 (finding plaintiff sustained burden of proof that plaintiff would have benefited from early treatment, "even if plaintiff failed to establish [that, if tumor had been diagnosed, it would have been operable] to a degree of certainty so as to bring into play statistically measurable chances of survival."); Hernandez v. Clinica Pasteur, Inc., 293 So.2d 747, 750 (Fla.Dist.Ct.App. 1974) (finding jury question presented where doctor testified that "probably and possibly" decedent would have had a better chance to live if treatment for cardiac arrest had been promptly administered instead of sending patient home with diagnosis of gastritis and instructions to exercise); McKellips, supra, 741 P.2d at 477 (holding that, where plaintiff who suffered from heart attack but was diagnosed with gastritis had offered proof that defendant's conduct caused substantial reduction of patient's chance of recovery or survival, question of proximate cause was for jury, irrespective of statistical evidence of likelihood of successful intervention).

In Olah, supra, this Court considered issues involving the alleged negligent failure to administer a diagnostic test that might have revealed a preexistent condition, which increased the risk to the patient, who ultimately died. 119 N.J. at 124, 574 A.2d 411. The patient, Lilliann Olah, had suffered from a pancreatic pseudocyst, which was removed surgically on November 29, 1983. On December 14, 1983, Olah collapsed at home, suffering from nausea, gastric and back pain, and was taken by ambulance to the hospital. Id. at 122, 574 A.2d 411. Between December 14 and 19, 1983, Olah's condition stabilized and she was tentatively diagnosed with bleeding of the upper gastrointestinal tract. Ibid. Olah saw her treating physician, Dr. Slobodian, on December 16, 1983. Slobodian consulted with his associate and agreed that Olah should receive a gastrointestinal workup to determine the source of her bleeding, but decided to wait to perform the tests until Olah felt better. Id. at 122-23, 574 A.2d 411. Olah returned home on December 19, 1983. Id. at 123, 574 A.2d 411. On December 20, 1983, Olah collapsed and vomited blood. Ibid. She was rushed to the hospital and arrived without a palpable blood pressure. An emergency room physician gave her intravenous fluids. Slobodian examined Olah, consulted with two specialists and decided that she should receive an endoscopy. The gastroenterologist suggested and Slobodian agreed that they should delay the endoscopy, because emergency surgery was not being considered and the delay would allow Olah's condition to stabilize. Ibid. Olah's condition improved slightly, but within three hours of the consultation with the gastroenterologist Olah began to hemorrhage. Id. at 123-24, 574 A.2d 411. Olah died that evening from a blood-clotting disorder that results from blood loss. Id. at 124, 574 A.2d 411.

Olah's family filed suit for wrongful death, alleging claims for negligence against Slobodian, the gastroenterologist and Slobodian's associate. Ibid. The claims against the associate were tried separately. Ibid. The trial focused on Olah's second allegation that Slobodian and the gastroenterologist had been negligent by failing to order an endoscopy for Olah on December 20, 1983, during the three hours when Olah had stabilized. Id. at 124-25, 574 A.2d 411. The "[p]laintiffs' expert testified that a patient with Mrs. Olah's immediate medical history might start hemorrhaging again at any moment, making it appropriate to perform an endoscopic examination during any period of stability." Id. at 125, 574 A.2d 411. The defendants argued that the plaintiffs had failed to present sufficient "evidence for a jury to find that any of the alleged acts of negligence had, within a reasonable medical probability, caused Mrs. Olah's death." Ibid.

The trial court instructed the jury using the Evers standard of causation. Id. at 125-26, 574 A.2d 411. After receiving several communications from the jury, the trial court concluded that the jury had failed to understand its instructions. The court modified and clarified the instructions, and eventually the jury found that Slobodian's negligence had been a proximate cause only of Olah's pain and suffering, awarding $50,000 damages to plaintiffs. Initially, the trial court vacated the damages verdict. On reconsideration, it ordered a new trial on all issues. Id. at 128, 574 A.2d 411. On appeal, the Appellate Division concluded that the causation issue required retrial because the trial court's instruction had been incomplete. Id. at 129, 574 A.2d 411. The panel held that the trial court should have instructed the jury that "plaintiffs may recover if they demonstrate that there was a `substantial possibility that the decedent could have survived had she been properly attended to by defendant in a timely manner.'" Ibid. The panel reinstated the jury award and remanded the matter for a new trial limited to the causal connection between Slobodian's negligence and Olah's death. Id. at 128, 574 A.2d 411.

This Court granted Slobodian's petition for certification challenging the Appellate Division's reinstatement of the jury verdict and the modification of the causation charge. Id. at 129, 574 A.2d 411. We disapproved of the Appellate Division's imposition of a requirement that plaintiff's proofs quantify the likelihood that the physician's failure to perform diagnostic tests on December 20 would have avoided Olah's death, explaining that the possibility of avoiding harm was subsumed in the jury's conclusion that the failure to test deviated from accepted practice and increased the risk of harm. Id. at 133, 574 A.2d 411. We observed that as long as there was proof to a reasonable degree of medical probability that the failure to test increased the risk of harm, the significance of that increased risk should be determined by the jury under the second prong of Scafidi. Ibid. We stated:

The question whether a patient has a "substantial possibility" of avoiding harm would ordinarily be subsumed in the jury's determination whether a defendant's deviation "increased the risk of harm from the preexistent condition." It is unlikely that a defendant's negligence would increase the risk of harm in the absence of a possibility of avoiding harm. Finally, the "substantial factor" segment of the Evers standard . . . is well designed to focus the jury's attention on "whether the deviation, in the context of the preexistent condition, was sufficiently significant in relation to the eventual harm to satisfy the requirement of proximate cause."

[ Id. at 133, 574 A.2d 411 (quoting Scafidi, supra, 119 N.J. at 113-14, 574 A.2d 398).]

III

When the prevailing standard of care indicates that a diagnostic test should be performed and that it is a deviation not to perform it, but it is unknown whether performing the test would have helped to diagnose or treat a preexistent condition, the first prong of Scafidi does not require that the plaintiff demonstrate a reasonable medical probability that the test would have resulted in avoiding the harm. Rather, the plaintiff must demonstrate to a reasonable degree of medical probability that the failure to give the test increased the risk of harm from the preexistent condition. A plaintiff may demonstrate an increased risk of harm even if such tests are helpful in a small proportion of cases. We reach that conclusion to avoid the unacceptable result that would accrue if trial courts in such circumstances invariably denied plaintiffs the right to reach the jury, thereby permitting defendants to benefit from the negligent failure to test and the evidentiary uncertainties that the failure to test created. See Scafidi, supra, 119 N.J. at 108, 574 A.2d 398; Evers, supra, 95 N.J. at 417, 471 A.2d 405.

The trial court concluded that plaintiffs had produced sufficient evidence that the fetus's death was caused in part by a preexistent condition of an abnormal cord and placenta and that, had Dr. Pawliw diagnosed and treated the condition by inducing labor on December 21, 1988, the fetus would have lived. The trial court also concluded that plaintiffs produced sufficient proof that Dr. Pawliw had deviated from the accepted standard of medical care by not performing the NST and BPP tests on December 21, 1988. However, it dismissed plaintiffs' cause of action, concluding that plaintiffs failed to prove to a reasonable degree of medical probability that the failure to administer the NST and BPP tests increased the risk of harm to the fetus, because plaintiffs' expert failed to testify to a reasonable medical probability that those tests would have revealed the fetus's placenta and umbilical cord abnormalities.

The lower court erred in its conclusion that reasonable medical probability, under the first prong of Scafidi, applied to the likely result of the test. The requirement concerning reasonable medical probability relates to whether the physician's failure to give the test increased the risk of harm from the preexistent condition. As Justice Handler observed in Scafidi, supra:

[F]or purposes of awarding compensatory damages based on the increased risk of future harm caused by the tortious conduct of others, we need not insist on a quantification of the risk of future harm with mathematical exactitude. What should suffice is the reality confirmed by common sense and ordinary experience that a plaintiff has actually incurred the increased risk of future harm.

[ 119 N.J. at 118, 574 A.2d 398 (Handler, J., concurring).]

The trial court and Appellate Division panel erred in concluding that plaintiffs had failed to offer sufficient proof of increased risk to reach the jury. Plaintiffs' burden was not to show as a matter of reasonable medical probability that the tests would have revealed the placenta and umbilical cord abnormalities. Plaintiffs' burden was to show that Dr. Pawliw's failure to perform the NST and BPP tests increased the risk that the fetus would die in utero. Dr. Kalafer testified that defendant had deviated from the standard of care and had acted negligently in failing to administer the NST and BPP tests; that those tests should have been performed because of Gardner's symptoms, her high-risk pregnancy status and her history of miscarriages; that the marked decrease in fetal activity had been very significant; that the complaints should have been followed up; that the fetus's environment should have been checked; and specifically that the NST and BPP tests should have been performed. Dr. Kalafer's interrupted response to the questions concerning whether the test would have revealed any abnormalities of the cord or placenta indicated that, if the test had been performed, the results would have helped to better manage the pregnancy. Additionally, he responded that had the results been normal, then the physician and patient would have been reassured; but if the results were abnormal, then the physician would have been encouraged either to more closely monitor the patient or to deliver the fetus. Dr. Kalafer answered affirmatively when asked whether he could say to a reasonable degree of medical probability that because Dr. Pawliw failed to perform either an NST or a BPP test there had been an increased risk that a condition that could cause the fetus's death would not be recognized. Accordingly, Dr. Kalafer's testimony was sufficient for plaintiffs to satisfy their requisite threshold burden of proof that to a reasonable medical probability the failure to perform those two tests increased the risk of harm from the preexistent condition. Plaintiffs should have been permitted to submit for the jury's determination the questions of whether, based on the parties' expert testimony, the failure to give the NST or BPP tests had increased the risk that the fetus's condition would not be detected, treated or corrected and whether that increased risk had been a substantial factor in causing her death. See Olah, supra, 119 N.J. at 133, 574 A.2d 411; Scafidi, supra, 119 N.J. at 108, 574 A.2d 398.

In holding that plaintiffs' proofs were sufficient to present a jury question on causation, we do not imply that the likely significance of the test results is irrelevant to the outcome of the litigation. Under the second prong of Scafidi, it is the jury's responsibility to determine whether the increased risk of harm resulting from the failure to perform the tests was or was not a substantial factor in causing the ultimate harm sustained. The jury will weigh the testimony and other evidence presented at trial and determine whether any increased risk due to the failure to test was a substantial factor in causing the fetus's death. See Scafidi, supra, 119 N.J. at 108-09, 574 A.2d 398; See also, Jones v. Montefiore Hosp., 494 Pa. 410, 431 A.2d 920, 924 (1981); Herskovits, supra, 664 P.2d at 478 ("It is not necessary for a plaintiff to introduce evidence to establish that the negligence resulted in the injury or death, but simply that the negligence increased the risk of injury or death. The step from the increased risk to causation is one for the jury to make."). On this record, a jury properly could conclude that the "substantial factor" prong of Scafidi was or was not established by the evidence. Additionally, under Scafidi, assuming that the jury found that the failure to perform diagnostic tests was a substantial factor causing the fetus's death, it would also be required to apportion fault according to comparative fault principles. See Anderson, supra, 144 N.J. at 207, 676 A.2d 127. The doctrine of increased risk reduces the burden of proof imposed on a plaintiff suffering from a preexistent condition, but it also has the likely effect of reducing the percentage of fault and resultant damages attributable to a negligent physician. See Scafidi, supra, 119 N.J. at 112-13, 574 A.2d 398.

Finally, the Appellate Division erred in its conclusion that Dr. Kalafer's testimony was a "bare conclusion" insufficient to establish the requisite legal standard of causation. The Appellate Division's characterization of Dr. Kalafer's testimony was influenced by its misperception of the relevant burden of proof. Because the Appellate Division required proof to a reasonable degree of medical probability that the fetus's preexisting condition would have been detected if defendant had ordered the NST and BPP tests, the court determined that Dr. Kalafer's inability to determine the fetus's specific condition on December 21, 1988, precluded him from providing an expert opinion that met that standard. 285 N.J. Super. at 122-25, 666 A.2d 592.

As we have explained, the adequacy of Dr. Kalafer's testimony should be tested against the increased risk standard that we have articulated. Sufficient factual evidence was presented by Dr. Kalafer to the effect that if the tests had been performed a stressed in utero environment might have been observed. Dr. Kalafer testified that the placental and umbilical cord abnormalities noted in the autopsy report were consistent with the patient's complaint about the decrease in fetal movement. He stated that "it might have been indicative of the fact that this placenta was not nourishing the baby appropriately," resulting in stress to the fetus. Dr. Kalafer stated that the NST and BPP tests might have detected a "smoldering" uterine environment, which would have indicated either the need for continued closer monitoring by the obstetrician or an early induction of labor. Dr. Kalafer also concluded that although several of the fetus's organs were smaller than the average fetus of this gestational age, the fetus would likely have survived if she had been delivered on December 21, 1988. Based on the autopsy findings, medical records and deposition testimony provided to him, Dr. Kalafer's expert opinion was adequately supported by factual evidence in the record and sufficient to raise a jury question concerning defendant's liability.

On the record below, when viewing the facts in the light most favorable to plaintiffs, we find that sufficient factual evidence was presented for the trial court to have allowed the jury to resolve the questions whether Dr. Pawliw's failure to perform the NST and BPP tests on December 21, 1988, increased the risk of harm to the fetus and whether such increased risk was a substantial factor in causing the ultimate harm of the fetus's death. See Lanzet v. Greenberg, supra, 126 N.J. at 172, 594 A.2d 1309; Dolson v. Anastasia, 55 N.J. 2 , 5-6, 258 A.2d 706 (1969); R. 4:40-1.

IV

We reverse the judgment of the Appellate Division, and remand the matter to the Law Division for a new trial.