LAVORATO, Justice.

Melissa and Adam Williams brought this medical malpractice action against Robert E. Hedican and Ob-Gyn Specialists, P.C. They alleged that the defendants negligently failed to treat Melissa for chicken pox while she was pregnant and as a result her child was born blind in one eye. In a pretrial ruling, the district court excluded expert testimony on causation from the plaintiffs' medical expert. The court concluded the testimony lacked a sufficient foundation under the test suggested in Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993). Employing the same test, we conclude the district court abused its discretion in excluding the testimony. We reverse and remand for further proceedings consistent with this opinion.

I. Background Facts and Proceedings. Melissa and Adam Williams are husband and wife and parents of Maxwell Williams. Robert E. Hedican and Ob-Gyn Specialists, P.C. delivered prenatal care for Melissa in the spring of 1992.

Melissa claims that on April 30, 1992 she reported to one of Hedican's employees that she had been exposed to chicken pox while working at a day care center. At the time, Melissa was seventeen weeks pregnant with Maxwell. Melissa also claims the employee told her that it was not a matter of significance. Melissa received no testing, treatment, or other follow-up and thereafter developed a severe case of chicken pox.

On October 5 Melissa gave birth to Maxwell. Maxwell was born with various maladies including blindness in one eye and skin lesions.

On April 21, 1994, Melissa and Adam brought this action individually and on behalf of Maxwell against Hedican and Ob-Gyn Specialists, P.C. The petition alleged that the defendants were negligent in Melissa's care and treatment when she contracted chicken pox while she was pregnant with Maxwell.

The petition also alleged that because of the defendants' negligence Melissa did not receive medication that could have alleviated her symptoms and substantially improved Maxwell's chance of avoiding harm from the chicken pox virus. The petition further alleged that because of the defendants' failure to provide Melissa with proper antiviral therapy, Maxwell was born with congenital varicella syndrome, which has left him blind in one eye.

Congenital varicella syndrome occurs when mothers have been exposed to the varicella-zoster virus and transmit the virus through the placenta to their unborn children. Varicella-zoster virus is the scientific name for the chicken pox virus. Varicella-zoster immune globulin (VZIG) is an antibody that may destroy the chicken pox virus.

As mentioned, chicken pox is a virus. A virus is a particle smaller than the nucleus of a cell and consists of one or sometimes more strands of genetic material enclosed in a protein shell. The genetic material in the chicken pox virus is deoxyribonucleic acid or DNA.

Because viruses are not living organisms, they cannot reproduce on their own. Once a virus enters the body, the virus' genetic programming directs the virus to healthy cells that make up certain tissues of the body. The viral particles next attach themselves to the walls of the individual cells, penetrate the cell wall, and ultimately enter the cell nucleus. Once inside the cell nucleus, the viral particles take over the reproductive genetic machinery of a cell and reproduce other viral particles. Eventually, the viral particles break out of these cells and spread to other healthy cells, where they repeat the process. Fever, aching, and the outbreak of skin lesions occur after the viral particles have reproduced in sufficient numbers and have invaded enough tissue in a particular part of the body for symptoms to develop.

The body has an elaborate and complex immune system to combat viruses. Antibodies are part of this system and are carried by the blood. Antibodies attack viruses and either destroy or disable them. Once the body produces antibodies in response to a particular type of virus, the antibodies remain part of the body's immune system and are available to attack the same viral type in the event of reinfection. If a person has never been exposed to a particular viral disease, the person may not have as part of the person's immune system the antibodies necessary to respond to that virus. Such persons are inoculated (deliberately injected) with attenuated virus to stimulate the body's own immune response so that if they are exposed to a virulent strain of that virus, they will have antibodies to combat the disease.

A person may also be given an injection of antibodies, or immune globulin, acquired from another person. Once in the body, the immune globulin performs the role that the body's own antibodies would have performed in attacking and destroying or disabling the viral particles. Eventually, the body's immune system manufactures its own antibodies that aid in combating the disease and that will provide protection if there is future exposure to the virus.

The chicken pox virus enters the body through the nose or mouth. For a period of three days following this invasion, the viral particles locate in the respiratory area of the head and upper neck. After four to six days of initial exposure and reproduction, the viral particles spread to the liver, spleen, and other organs of the body. This is the primary viremia stage. Once having spread, the viral particles continue to reproduce. About fourteen days after exposure, the secondary viremia stage begins and the viral particles spread to the skin causing rashes and vesicles that are the outward manifestations of chicken pox.

The chicken pox virus can spread to the fetus of a mother who has been exposed to the virus by crossing the placental barrier to reach the fetus. Fetal infection may occur when the chicken pox disease reaches either the primary viremia or secondary viremia stages but may be more likely to occur in the secondary viremia stage.

Both medical experts — the plaintiffs' and the defendants' — agree that VZIG (the immune globulin for the chicken pox virus) should be given to susceptible pregnant women who are exposed to chicken pox within ninety-six hours of exposure. Doing so can prevent or lessen the effects of chicken pox in the mother. Such treatment is important because pregnant women are at a heightened risk for potentially fatal complications from chicken pox, such as viral pneumonia and encephalitis. The obstetrical literature supports this recommended treatment and the reasons for it.

After ninety-six hours, the viral particles will have become too numerous and diffusely spread for VZIG to be successful in preventing the disease. Early treatment is vital when the viral particles are fewer in number and concentrated near the point of entry in the upper respiratory area.

The two experts, however, disagree on whether VZIG can prevent or lessen the effects of chicken pox in the fetus. According to the plaintiffs' expert, Dr. James Balducci, "it is a matter of common sense and sound scientific reasoning that if VZIG successfully neutralizes the varicella zoster particles in the mother, the fetus will be protected from those particles as well." In reaching that conclusion, he points to several factors. First, a viral particle that is disabled from attacking the mother is disabled from attacking the fetus. Second, the varicella particles do not begin to cross the placental barrier and infect the fetus until four to six days after the mother's exposure, which is the primary viremia stage. Third, if VZIG is administered before this stage it will be able to attack the disease before the viral particles begin to spread to the fetus. Fourth, VZIG can cross the placental barrier to the fetus. Last, by crossing the placental barrier to the fetus, VZIG can provide a direct benefit to the fetus by attacking viral particles that did make it across the placental barrier. Balducci bases this reasoning on "the physiology of the way immune globulin works."

The defendants' expert, Dr. Charles Grose, takes the opposite position, although he acknowledges that it is theoretically possible for VZIG to prevent chicken pox in the fetus. According to Grose, it has never been scientifically demonstrated that VZIG can prevent chicken pox in the fetus.

Balducci is certified by the American Board of Obstetrics and Gynecology. He is the Chief of the Obstetrics Department and the division head of the Maternal Fetal Medicine Unit at Lehigh Valley Hospital in Pennsylvania. He has published numerous articles, including a study of outcomes after first trimester varicella infection.

To support his opinion, Balducci relies on his own experience and a study performed by a well-known researcher in the area, Dr. Gisela Enders.

Balducci's hospital serves an area 120 miles in radius and regularly treats pregnant women who have been exposed to chicken pox. Balducci personally treats as many as four to five women per month who have been exposed to the disease. He has treated or been involved in the treatment of about 150 women who received VZIG while they were pregnant. In his experience, none of these women gave birth to a baby showing signs of congenital varicella syndrome.

In addition, Balducci's hospital has been involved in the delivery of 12,000 to 14,000 babies. He is unaware from his direct clinical experience or from anecdotal reports that any of these children were born with congenital varicella syndrome where the mother had received a timely administration of VZIG.

In his deposition, Balducci acknowledged that his personal study was poor because a scientific study would include two sets of populations where one set receives the VZIG during pregnancy and the other set does not. According to Balducci, such a study would be unethical because some mothers in the untreated set would probably die from severe pneumonia.

Dr. Enders' study was published in the respected British medical journal "The Lancet" in 1994. According to the study it is "the largest reported prospective study of the outcome of varicella in pregnancy and thus provides the best estimate to date of the risk of congenital varicella syndrome following maternal infection." The study stated: "Post-exposure prophylaxis with varicella zoster immune globulin attenuates disease [in women exposed to varicella-zoster] and there was evidence from the present study that it may also reduce the risk of fetal infection." The study made these conclusions after observing that "[n]o cases of congenital varicella syndrome occurred among the 97 pregnancies in which maternal varicella infection followed post-exposure anti-varicella-zoster immune-globulin prophylaxis. . . ."

Grose is board-certified in pediatrics and pediatric infectious diseases. He is a professor in two departments at the University of Iowa College of Medicine: the Department of Pediatrics and the Department of Microbiology. Grose is also the head of the Division of Infectious Diseases. Grose has published over 100 medical articles and chapters related to varicella zoster virus, including several chapters for recent textbooks.

In Grose's opinion "there is no medical data that administration of VZIG to the pregnant woman will modify congenital varicella syndrome."

Grose attacks Balducci's opinion about the beneficial effects of VZIG on the fetus. He points out that Balducci (1) "has no evidence for [his opinion] based on a scientific study that is corroborated by statistical data," (2) cites evidence that is irrelevant to his opinion, and (3) "freely admits [that] there is no published medical literature to support the assertion that VZIG will have beneficial effects on the fetus." In addition, Grose cites the VZIG manufacturer insert and Centers for Disease Control publications that state there is no evidence that VZIG protects the fetus.

Grose discredits the Enders study, claiming that the sample size was too small and the VZIG applications were much stronger than the American-manufactured VZIG. He criticizes Balducci's reliance on his clinical experience, claiming Balducci's clinical results were a "poor study."

In response, Balducci states he based his opinion on reliable medical literature and clinical experience. He claims that Grose overstates the importance of German VZIG-strength relative to American VZIG. Balducci says that "timing" is much more important than dosage and cites Grose's own practice of using lower dosage than that recommended by the manufacturer. Balducci also points to American medical literature that cites Enders' German VZIG studies as proof of VZIG effectiveness for pregnant women. Last, Balducci notes that Grose actually agrees with him on many of Balducci's scientific premises.

The parties presented all of this information from the two experts to the district court by way of depositions, answers to interrogatories, affidavits and medical articles. The district court considered this information in deciding the defendants' Iowa Rule of Evidence 104(a) motion. In the motion, the defendants asked the court to rule Balducci's testimony inadmissible under Iowa Rules of Evidence 702 and 403. Additionally, the defendants asked the district court to employ the test suggested in Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993). More specifically, the defendants asked the court to enter an order barring Balducci's testimony regarding a causal connection between the nonadministration of VZIG to Melissa and the injuries suffered by Maxwell.

The district court granted the motion pursuant to Iowa Rules of Evidence 702 and 403, employing the Daubert test. The court ruled that Balducci's reasoning was not supported by scientific study. Additionally the court ruled that Balducci's theory has never been (1) tested, (2) subjected to peer review or publication, and (3) borne out in his own practice. The court went on to reiterate all of Grose's challenges to Balducci's opinion about the beneficial effects of VZIG if administered in a timely manner. Finally, the court ruled that Balducci's opinion testimony would result in the admission of evidence whose probative value is far outweighed by its unfair prejudicial effect.

Thereafter we granted the plaintiffs' application for interlocutory review. See Iowa R. App. P. 2. We note that the defendants joined in the application and asked that we affirm the district court's ruling.

On appeal, the plaintiffs challenge the district court's exclusion of their expert's testimony. We generally review evidentiary rulings for abuse of discretion and do so here. See Hutchison v. American Family Mut. Ins. Co., 514 N.W.2d 882, 885 (Iowa 1994) (applying abuse of discretion standard to review an Iowa Rule of Evidence 703 ruling).

II. Admissibility of Expert Testimony.

The decisive issue here is whether there was adequate foundation for Balducci's expert opinion that VZIG more probably than not would have prevented or reduced the symptoms and suffering of Maxwell Williams. To decide this issue we must determine the proper test for evaluating the admissibility of this expert opinion.

A. Iowa's Traditional Approach to Admissibility of Expert Testimony. Iowa Rule of Evidence 702 governs the admissibility of expert testimony and uses the same language as Federal Rule of Evidence 702. It provides:

If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education may testify thereto in the form of an opinion or otherwise.

Iowa has a liberal rule on the admission of expert testimony. Ganrud v. Smith, 206 N.W.2d 311, 314 (Iowa 1973). The trend in Iowa has been toward broadening the scope of the admissibility of such testimony. State v. Klindt, 389 N.W.2d 670, 674 (Iowa 1986). Rule 702 has codified that liberality. Hutchison, 514 N.W.2d at 885.

As part of our liberal approach, we abandoned the so-called Frye test. See State v. Hall, 297 N.W.2d 80, 85 (Iowa 1980). In Frye, scientific evidence is admissible if it is based on a scientific technique generally accepted as reliable within the scientific community. Frye v. United States, 293 F. 1013, 1014 (D.C. Cir. 1923). We did, however, retain the Frye rationale that the evidence — whether scientific or otherwise — be reliable. Hall, 297 N.W.2d at 85.

In Hall we implicitly recognized that the proffered evidence must aid the jury to resolve a disputed issue. A threshold finding of reliability was therefore necessary "because unreliable evidence cannot assist a trier of fact." Klindt, 389 N.W.2d at 672. The amount of foundation necessary to establish reliability depended on the complexity of the evidence and the likely impact of the evidence on the fact-finding process. Hall, 297 N.W.2d at 85.

We also recognized that "[t]here is no requirement that the expert be able to express an opinion with absolute certainty: `[A]n expert's lack of absolute certainty goes to the weight of this testimony, not to its admissibility.'" State v. Buller, 517 N.W.2d 711, 713 (Iowa 1994) (quoting United States v. Cyphers, 553 F.2d 1064, 1072-73 (7th Cir. 1977)).

Rule 702 did not, as far as our case law is concerned, abrogate any of these principles in our approach to the admissibility of expert testimony. See State v. Murphy, 451 N.W.2d 154, 156-57 (Iowa 1990).

Enter Daubert.

B. Daubert. The Supreme Court decided Daubert in 1993, some thirteen years after Hall. In Daubert, the question before the Supreme Court was the admissibility of novel scientific evidence. The plaintiffs had sued Merrell Dow for birth defects allegedly caused by ingestion of the drug Bendectin during pregnancy. Daubert, 509 U.S. at 582, 113 S.Ct. at 2791, 125 L.Ed.2d at 476. The plaintiffs tried to introduce the testimony of several experts who were relying on animal rather than human statistical studies in reaching their conclusion that Bendectin caused birth defects. The federal district court rejected the evidence because the experts did not base their testimony on human studies. Daubert v. Merrell Dow Pharms., Inc., 727 F. Supp. 570, 575 (S.D.Cal. 1989). In rejecting the testimony, the district court ruled the proffered evidence did not meet the so-called Frye "generally accepted" test. The Ninth Circuit affirmed the district court's rejection of the expert's testimony regarding the effects of Bendectin. Daubert v. Merrell Dow Pharms., Inc., 951 F.2d 1128 (9th Cir. 1991).

The Supreme Court vacated the circuit court of appeals decision. Daubert, 509 U.S. at 598, 113 S.Ct. at 2799, 125 L.Ed.2d at 485. In doing so, the Court unanimously rejected the Frye "general acceptance" test. Id. at 597-98, 113 S.Ct. at 2799, 125 L.Ed.2d at 485-86. The Court could have stopped there, but, over the dissent of Chief Justice Rehnquist and Justice Stevens, offered dicta characterized as "general observations" on how trial courts should handle purportedly scientific evidence. Id. at 593, 599, 113 S.Ct. at 2796, 2801, 125 L.Ed.2d at 482, 486.

Daubert emphasized that trial judges should act as gatekeepers to ensure that all scientific evidence admitted is both reliable and relevant. Id. at 589 n. 7, 113 S.Ct. at 2795 n. 7, 125 L.Ed.2d at 480 n. 7. In addition to shifting to a more flexible test centered on reliability and relevancy, the Court changed the manner in which evidence is to be admitted. Hereafter, trial judges — rather than the scientific community — will determine the scientific validity of expert testimony. Jane Campbell Moriarity, The First Circuit Experience With Daubert v. Merrell Dow Pharmaceuticals: Requiring the Gatekeeper to be Reliable?, MSL L.Rev. Winter 1996-1997, at 28-29 [hereinafter "Moriarity"].

To perform this gatekeeping task, trial judges should focus primarily on Federal Rule of Evidence 702. Daubert, 509 U.S. at 589, 113 S.Ct. at 2795, 125 L.Ed.2d at 480. The first part of rule 702 permits admission of "scientific knowledge." Id. at 590, 113 S.Ct. at 2795, 125 L.Ed.2d at 480. The term "scientific," according to Daubert, "implies a grounding in the methods and procedures of science." Id. at 590, 113 S.Ct. at 2795, 125 L.Ed.2d at 481. "Knowledge" means more than "subjective or unsupported speculation." Id. The Court explained that to qualify under the scientific evidence prong of rule 702, evidence must meet the following standard:

The term [knowledge] applies to any body of known facts or to any body of ideas inferred from such facts or accepted as truth on good grounds. Of course, it would be unreasonable to conclude that the subject of scientific testimony must be "known" to a certainty; arguably, there are no certainties in science. But, in order to qualify as "scientific knowledge," an inference or assertion must be derived by the scientific method. Proposed testimony must be supported by appropriate validation — i.e., "good grounds," based on what is known. In short, the requirement that an expert's testimony pertain to "scientific knowledge" establishes a standard of evidentiary reliability.

Id. (citations omitted). By this passage, "reliability" became the defining standard for the Supreme Court. Moriarity at 36.

The second part of rule 702 requires that the scientific testimony assist the trier of fact to understand the evidence or to determine a fact in issue. Daubert, 509 U.S. at 591, 113 S.Ct. at 2795, 125 L.Ed.2d at 481. Daubert related this requirement to relevancy. Id. at 591, 113 S.Ct. at 2796, 125 L.Ed.2d at 481. Thus, expert testimony that does not relate to an issue in the case is not relevant and is therefore not helpful. Id. The Court characterized this requirement as "fit." Id.

Following this discussion, Daubert suggested a new framework for determining the admissibility of expert scientific testimony:

[T]he trial judge must determine at the outset, pursuant to Rule 104(a), whether the expert is proposing to testify to (1) scientific knowledge that (2) will assist the trier of fact to understand or determine a fact in issue. This entails a preliminary assessment of whether the reasoning or methodology underlying the testimony is scientifically valid and of whether that reasoning or methodology properly can be applied to the facts in issue.

Id. at 592, 113 S.Ct. at 2796, 125 L.Ed.2d at 482.

The Court provided a nonexhaustive checklist to help trial courts decide whether the underlying reasoning and methodology are both scientifically valid and applicable to the issues in the case. These include whether the theory or technique (1) can be (and has been) tested, (2) has been subjected to peer review and publication, (3) is generally accepted within the relevant scientific community, and (4) has a known or potential rate of error. Id. at 595, 113 S.Ct. at 2797, 125 L.Ed.2d at 483.

Thus, the Frye "general acceptance" standard is but one of several factors trial courts may consider. As to this factor, the Court made clear that trial judges need not insist in the reliability assessment on "explicit identification of a relevant scientific community and an express determination of a particular degree of acceptance within that community." Id. at 594, 113 S.Ct. at 2797, 125 L.Ed.2d at 483.

Considering all of these factors, Daubert emphasized that the inquiry under rule 702 is flexible and focuses solely on principles and methodology, not on the conclusions they generate. Id. at 594-95, 113 S.Ct. at 2797, 125 L.Ed.2d at 483-84. Thus, courts are to consider only whether the process is scientifically sound, not whether the results are persuasive.

Additionally, the Court pointed to other applicable rules that trial judges should be mindful of in assessing a proffer of expert testimony under rule 702. These include Federal Rules of Evidence 703, 706, and 403. Id. at 595, 113 S.Ct. at 2797-98, 125 L.Ed.2d at 484. Rule 703 allows expert testimony based on otherwise inadmissible hearsay if the facts or data are "of a type reasonably relied upon by experts in the particular field in forming opinions or inferences upon the subject." Rule 706 allows a trial court to ask the assistance of an expert. Rule 403 permits the exclusion of relevant evidence "if its probative value is substantially outweighed by the danger of unfair prejudice, confusion of the issues, or misleading the jury. . . ."

In sum, "[i]n its role as the `gatekeeper' of untrustworthy scientific evidence, the district court is required, in a Rule 104(a) hearing, to review the scientific basis for the evidence and determine whether it is sufficiently reliable for admission." Moriarity at 37 (footnote omitted).

Finally, the Court addressed concerns that abandonment of the Frye "general acceptance" requirement might result in "free-for-all" admissions of "absurd and irrational pseudoscientific" assertions. Daubert, 509 U.S. at 595, 113 S.Ct. at 2798, 125 L.Ed.2d at 484. Answering those concerns, the Court pointed to cross-examination, presentation of contrary evidence, and careful instruction on the burden of proof as the traditional and appropriate means of attacking "shaky but admissible evidence." Id. at 596, 113 S.Ct. at 2798, 125 L.Ed.2d at 484. Additionally, the Court noted if the evidence is insufficient to allow a jury to conclude that a position more likely than not is true, summary judgment or directed verdict are additional safeguards. Id. In this part of its opinion, the Court "expressed its faith in the power of the adversary system to test `shaky but admissible evidence' and advanced a bias in favor of admitting evidence short of that solidly and indisputably proven to be reliable." Borawick v. Shay, 68 F.3d 597, 610 (2d Cir. 1995) (discussing Daubert).

One writer concludes that Daubert poses a dilemma. G. Michael Fenner, The Daubert Handbook: The Case, Its Essential Dilemma, and Its Progeny, 29 Creighton L.Rev. 939, 951-52 (1996). The dilemma is that Daubert is both more and less restrictive of expert testimony. Id. As the writer explains,

First: To say that Daubert is less restrictive of expert evidence, to say that it opens the door for the introduction of expert evidence that would not have been admissible under the Frye test, is not to say that Daubert's test is an easier test. It may be more lenient in that it allows more — and more novel — science into evidence, but it can be much more difficult in that the Daubert test can require a more exacting, expensive, and time consuming foundation.

Second: the Daubert rule gives two seemingly contradictory results.

On the one hand, more science comes in. Science does not have to be generally accepted by other scientists to be admissible in court; the universe of admissible science is expanded by doing away with the general acceptance requirement. On the other hand, less science comes in. The trial judge is to act as gatekeeper and is to scrutinize carefully the proffered scientific evidence and to keep out what is not good science. The universe of science actually admitted may be contracted by the close scrutiny judges are supposed to give this evidence. While it may be that most science generally accepted in the relevant scientific community will be good science, it is not necessarily so.

Id. at 953.

C. Daubert's Scope. The dissent in Daubert raised numerous questions about the majority's "general observations." One of those questions asked: "Does all of this dicta apply to an expert seeking to testify on the basis of `technical or other specialized knowledge' — the other types of expert knowledge to which Rule 702 applies — or are the `general observations' limited only to `scientific knowledge'?" Daubert, 509 U.S. at 600, 113 S.Ct. at 2800, 125 L.Ed.2d at 487 (Rehnquist, C.J., and Stevens, J., dissenting).

Since Daubert several federal courts have dealt directly with this question. Most recently, one federal district court held that Daubert applies only to novel scientific testimony and is simply not applicable to "technical or other specialized knowledge." See Thornton v. Caterpillar, Inc., 951 F. Supp. 575, 577 (D.S.C. 1997). To support its position the Thornton court cited to the following language in Daubert:

Rule 702 also applies to technical and other specialized knowledge. Our discussion is limited to the scientific context because that is the nature of the expertise offered here.

Id. (quoting Daubert, 509 U.S. at 588 n. 8, 113 S.Ct. at 2794 n. 8, 125 L.Ed.2d at 481 n. 8).

The court in Thornton defined "technical" as anything "pertaining to or connected with the mechanical or industrial arts and the applied sciences." Id. (quoting Random House Dictionary of the English Language 1831 (2d ed. 1988)). "Technical knowledge," the court continued, is "the knowledge of these mechanical and industrial arts and applied sciences." Id. The Thornton court also defined "specialized knowledge" as referring "to any knowledge focused on a particular area of study, profession, or experience." Id. Differentiating scientific knowledge from technical and specialized knowledge, the court continued:

Scientific knowledge differs from technical and specialized knowledge in that it is a validation. Scientific knowledge is the process of formulating a hypothesis and then engaging in experimentation or observation to verify or falsify that hypothesis. It is this knowledge garnered from experimentation and observation that was offered as evidence in Daubert.

Id.

The Thornton court then concluded that the principles stated in Daubert should be narrowly limited to controversial and novel scientific evidence such as the evidence before the Court in Daubert. Id. at 578. Thus, the court declared, "Daubert has no other application for use in the expert field of engineering, as well as the fields of general medical issues, real estate or other types of technical subjects or those requiring specialized knowledge." Id. The court in Thornton singled out auto mechanics, engineers, accountants, attorneys, DEA agents, IRS agents, real estate appraisers, and the like as those having technical knowledge and therefore not within the ambit of Daubert. Id. The court likewise excluded from Daubert persons having "other specialized knowledge." Id.

Applying this reasoning to the facts before it, the Thornton court concluded that opinion evidence from the plaintiff's expert mechanical engineer as to design and lack of warning was not within the narrowly limited area of unique, untested and novel scientific evidence as described in Daubert. Id.

Several other federal courts have similarly restricted Daubert in favor of a conventional rule 702 analysis. See, e.g., Compton v. Subaru of Am., Inc., 82 F.3d 1513, 1519-20 (10th Cir. 1996) (limiting Daubert to cases involving unique, untested, or controversial methodologies or techniques and applying conventional rule 702 analysis to engineer's expert opinions on design defect in automobile roof because his opinions were based upon general engineering principles and his experience as an automotive engineer); Iacobelli Constr., Inc. v. County of Monroe, 32 F.3d 19, 25 (2d Cir. 1994) (holding that trial court erred in applying Daubert to an expert on construction site conditions, contract documents and project results because such testimony is not scientific); United States v. Muldrow, 19 F.3d 1332, 1337-38 (10th Cir.) (applying conventional rule 702 analysis to expert testimony of police officer with specialized knowledge of drug trafficking while at same time applying Daubert analysis to expert testimony of forensic chemist to determine reliability and relevance of his methodology), cert. denied, 513 U.S. 862, 115 S.Ct. 175, 130 L.Ed.2d 110 (1994); Liriano v. Hobart Corp., 949 F. Supp. 171, 177-79 (S.D.N.Y. 1996) (holding that it is not appropriate to invoke the Daubert test in cases where expert testimony is based solely on experience or training, as opposed to a methodology or technique, and applying conventional rule 702 analysis to safety consultant's expert testimony that meat grinder was defective because it lacked adequate warning).

The Washington supreme court has recently refused to apply the Daubert analysis in a medical malpractice case in which a defendant-doctor failed to treat the plaintiff's emphysema with a protein replacement therapy called Prolastin. See Reese v. Stroh, 128 Wn.2d 300, 308, 907 P.2d 282, 286 (1995) (en banc). The trial court had excluded expert testimony from the plaintiff's medical expert that, based upon reasonable medical probability, the Prolastin therapy would be effective for the plaintiff's condition. Id. at 305, 907 P.2d at 284. The trial court excluded such testimony because the expert did not have a statistically significant basis for his opinion. Id. at 304-05, 907 P.2d at 284.

Rejecting the Daubert analysis, the Reese court held that a conventional analysis under its rules of evidence was more appropriate because the expert's medical opinion was based on practical experience and acquired knowledge and not on some novel scientific procedure. Id. at 307-09, 907 P.2d at 286. The court remanded for such an analysis. Id. at 310, 907 P.2d at 287. The lack of substantial statistical support concerning Prolastin's effectiveness for the treatment of the plaintiff's condition went to the weight of the expert's testimony and not to its admissibility. Id. at 309, 907 P.2d at 287.

We note that the Washington court of appeals in a well-reasoned opinion had previously found the testimony admissible under a Daubert analysis. This was despite the fact that there was no statistical proof for the expert medical testimony on causation. See Reese v. Stroh, 74 Wn. App. 550, 563-64, 874 P.2d 200, 208 (1994).

In Hutchison, a case involving qualifications of an expert under our rule of evidence 702, we allowed a clinical neuropsychologist to give an opinion disputing a causal relationship between the plaintiff's car accident and her alleged head injuries. Hutchison, 514 N.W.2d at 889. We did so because rule 702's criteria for qualifications of experts (knowledge, skill, experience, training or education) "are too broad to allow distinctions based on whether or not a proposed expert belongs to a particular profession or has a particular degree." Id. at 887-88. We refused to impose barriers to expert testimony "other than the basic requirements of Iowa rule of evidence 702 and those described by the Supreme Court in Daubert." Id. at 887.

Such a broad reference to Daubert in Hutchison, however, was not a sweeping acceptance of the Daubert analysis for all expert testimony. While we think the restrictive approach to Daubert taken in Thornton, Compton, Iacobelli, Muldrow, Liriano, and Reese is a reasonable one, we do not now decide under what circumstances a Daubert analysis — which can be time consuming and costly — is appropriate. We refuse to do so because the issue is not before us. The district court in the case before us employed a Daubert analysis with no objection from the plaintiffs that such an analysis was inappropriate under the facts. Nor do the plaintiffs take such a position on appeal. Accordingly, we proceed to consider the merits employing the analysis suggested in Daubert without deciding whether such an analysis is called for under the facts.

D. The merits. Before considering the admissibility of Balducci's expert opinion, we harken back to Daubert's admonition that any inquiry under rule 702 is a flexible one that focuses solely on principles and methodology, not on the conclusions they generate. Daubert, 509 U.S. at 594-95, 113 S.Ct. at 2797, 125 L.Ed.2d at 483-84; see Kenneth J. Chesebro, Taking Daubert's "Focus" Seriously: The Methodology/Conclusion Distinction, 15 Cardozo L.Rev. 1745, 1746 (1994) ("The Court's instructions are quite clear: Rule 702 authorizes courts to scrutinize only the `scientific validity' of the `principles and methodology' used by an expert — not the persuasiveness of the `conclusions' so generated.") (emphasis added).

The district court stated numerous reasons for its ruling that Balducci's testimony was inadmissible: (1) "[a]lthough the reasoning does have some appeal, it is not supported by scientific study"; (2) "Balducci's theory has never been tested"; (3) "Balducci's theory has not been subjected to peer review or publication"; (4) Balducci's theory has not "been borne out in his own practice"; (5) "[b]ecause the doctor's theory is not subject to being tested and has not been subjected to peer review or publication, his known rate of error is indeterminable"; (6) "Balducci's theory is not generally accepted"; (7) Balducci's "reliance upon a German study by Dr. Enders is flawed [because] the VZIG manufactured in the United States and the VZIG manufactured in Europe are not the same product"; (8) "[t]here are no published studies in the medical literature of the United States demonstrating the efficacy of administering VZIG . . . to control the possibility of [congenital varicella syndrome] in the fetus."

The above enumerated reasons seen in terms of the Daubert analysis center on reliability. Relevancy or "fit" is all but conceded.

1. Scientific knowledge: Reliability. The first step under a Daubert analysis is to determine whether Balducci's expert opinion that VZIG has a beneficial effect on the fetus is "scientific knowledge." That entails determining whether the reasoning or methodology he used in reaching his opinion is based on scientifically valid principles. There must therefore be "good grounds" in the record validating that opinion. This is the reliability prong of the analysis.

Grose agrees with Balducci on how the chicken pox virus enters the body, how long it takes for the virus to become symptomatic, how long it takes for the virus to cross the placental barrier and reach the fetus, and the time-frame (within ninety-six hours of exposure) during which VZIG should be given to a pregnant woman to be effective for her. Additionally, the medical research and medical consensus support these views. All of this information is therefore scientific knowledge.

The two experts part ways on whether VZIG can prevent chicken pox in the fetus. Balducci insists that common sense and sound scientific reasoning support his opinion. His reasoning can be stated simply. First, a viral particle that is disabled from attacking the mother is disabled from attacking the fetus. Second, the varicella particles do not begin to cross the placental barrier and infect the fetus until four to six days after the mother's exposure, which is the primary viremia stage. Third, if VZIG is administered before this stage, it will be able to attack the disease before the viral particles begin to spread to the fetus. Fourth, VZIG can cross the placental barrier. Last, by crossing the placental barrier to the fetus, VZIG can also provide a direct benefit to the fetus by attacking viral particles that did make it across the placental barrier. In essence, Balducci reasons from established propositions to reach his conclusion that VZIG is beneficial to the fetus.

Apparently, Grose concedes that this reasoning supports Balducci's opinion because Grose testified in his deposition that, based on this reasoning, it is theoretically possible for VZIG to prevent chicken pox in the fetus. Grose insists, however, that it has never been scientifically demonstrated that VZIG can prevent chicken pox in the fetus. In short, Grose thinks Balducci's opinion or hypothesis that VZIG can prevent chicken pox in the fetus is not scientific knowledge because it has not been scientifically validated through statistically significant testing and is therefore not scientifically reliable.

Balducci counters insisting that his opinion is based on valid scientific methodologies and principles. In short, he believes his theory is scientific knowledge because it is more than subjective belief or unsupported speculation. In support of his position he relies on his own experience and studies done by Enders, Gibbs and Sweet, and Prober.

As mentioned, Balducci's hospital serves an area of 120 miles in radius and regularly treats pregnant women who have been exposed to chicken pox. He himself has treated as many as four to five women per month who have been exposed to the disease. He has treated or been involved in the treatment of about 150 women who received VZIG while they were pregnant. In his experience, none of these women gave birth to a baby showing signs of congenital varicella syndrome.

In addition, Balducci's hospital has been involved in the delivery of 12,000 to 14,000 babies. Whether from his direct clinical experience or from anecdotal reports, Balducci is unaware that any of these babies were born with congenital varicella syndrome where the mother had received a timely administration of VZIG.

While Balducci admits his personal study is less than ideal, the study and the hospital's experience is some evidence tending to support his theory. As Balducci concedes, the ideal study would include control groups involving mothers who did and did not receive VZIG during pregnancy. Obviously, as Balducci testified in his deposition, such a study would be medically unethical because some mothers in the untreated group would probably die from severe pneumonia.

As Daubert points out, "it would be unreasonable to conclude that the subject of scientific testimony must be `known' to a certainty; arguably there are no certainties in science." Daubert, 509 U.S. at 590, 113 S.Ct. at 2795, 125 L.Ed.2d at 481; accord Buller, 517 N.W.2d at 713 (refusing to require "absolute certainty" in expert testimony). We should be no more demanding about Balducci's theory. The quality of Balducci's study and his experience goes to the weight of his opinion, not to its admissibility.

Moreover, Balducci is not the only medical person who believes that VZIG may be beneficial to the fetus. As mentioned, a well-known researcher in the area, Dr. Gisela Enders, thinks so as well. According to Dr. Enders, her study is "the largest reported prospective study of the outcome of varicella in pregnancy and thus provides the best estimate to date of the risk of congenital varicella syndrome following maternal infection." Significantly, the study stated: "Post-exposure prophylaxis [preventive treatment] with varicella zoster immune globulin [VZIG] attenuates disease [in women exposed to the chicken pox virus] and there was evidence from the present study that it may also reduce the risk of fetal infection." The study made these conclusions after noting that "[n]o cases of congenital varicella syndrome . . . in infancy occurred among the ninety-seven pregnancies in which maternal varicella infection followed post-exposure anti-varicella-zoster immune-globulin prophylaxis."

As the study was published in a well-respected medical journal, Balducci's theory that VZIG may be beneficial to the fetus has taken the first steps to thorough peer review. Daubert, 509 U.S. at 593-94, 113 S.Ct. at 2797, 125 L.Ed.2d at 483 (stating publication is one element of peer review).

The most serious problem with the district court's reasoning, and this affects several of the court's stated reasons, is its treatment of the Enders study. The defendants contend the Enders study is "irrelevant" while the court stated that reliance on the Enders study was "flawed." The court then treats this "flaw" as taking away any scientific support for Balducci's theory; the court treats the Enders study as completely irrelevant.

Beyond the deeply suspicious logic that one level of strength and dosage in an experiment can provide absolutely no evidence of the potential effectiveness of other doses and strengths of the same substance, the only reason provided in the record for the court's position is Grose's naked assertion that German VZIG tests are irrelevant evidence of American VZIG effectiveness. The defendants provide no medical or scientific literature in support of this assertion. Balducci, on the other hand, refers to several American scientists who cite to the Enders study as providing relevant proof of VZIG effectiveness in America with respect to pregnant women.

Additionally, Balducci answers Grose's criticism of the Enders study. He maintains that Grose overstates the importance of German VZIG-strength relative to American VZIG. Rather, Balducci contends, timing is much more important than dosage. Balducci cites Grose's own practice of using lower dosage than recommended by the manufacturer of American VZIG.

Given this conflict between the two experts about the significance of the Enders study, we think treating the Enders article as totally irrelevant is an immoderate action. We think it is more reasonable to conclude that the Enders article is relevant but subject to some criticism regarding the certainty of conclusions it dictates as to American VZIG applications. Again, these criticisms go to the weight of the evidence, not to its admissibility.

Two other scientific articles — both mentioned by Balducci — touch on the effectiveness of VZIG on the fetus. In one, Charles Prober expresses "the hope that VZIG ameliorates the course of chicken pox and perhaps even reduces transmission to the fetus." In the other, Ronald Gibbs and Richard Sweet state that "although the risks of the congenital varicella syndrome is small, some authorities have recommended administration of varicella immune globulin to susceptible pregnant women who have not previously had varicella as soon as possible but within three days of exposure in the hope of protecting the fetus during the viremia." Although these statements are not scientific evidence, they do acknowledge that some medical authorities prescribe VZIG not only to protect the mother but also with the idea of protecting the fetus.

The district court's disregard of the Enders study as irrelevant led the court to the following overstatements: (1) Balducci's theory "is not supported by scientific study," (2) "Balducci's theory has never been tested," (3) "Balducci's theory has not been subjected to peer review," and (4) "his known rate of error is indeterminable." Given the somewhat obvious conclusion that the Enders study was (at least in part) a published test of Balducci's theory and was subject to peer review, we easily conclude the court's conclusions were in error. Additionally, we doubt that the "rate of error" is even applicable here. See Sorensen v. Shaklee Corp., 31 F.3d 638, 649 (8th Cir. 1994) (holding that rate of error does not apply where the testimony advanced involves theory and not any particular technique).

The district court portrays Balducci's theory as mere nonscientific conjecture and speculation. The defendants do not question that the subject of Balducci's area of expertise is a "science" or is "testable." The district court, however, seems to have confused concerns over the ethicality of prospective studies with the "impossibility" of ever testing this theory. Both parties appear to agree that if enough interest was generated, a national registry could be developed to garner statistical evidence on the question whether VZIG has beneficial effects on the fetus. Further, the district court seems to have missed the point of "testability" — the inquiry that separates science from nonscience.

The fact that a theory or technique has not been widely peer-reviewed — meaning subject to the scrutiny of the scientific community — does not necessarily brand the theory or technique as scientifically invalid. As Daubert points out, some well-grounded but innovative theories may be "too particular, too new, or of too limited interest to be published," publication being but one element of peer review. Daubert, 509 U.S. at 593, 113 S.Ct. at 2797, 125 L.Ed.2d at 483.

The defendants describe Balducci's theory as without "valid statistical evidence." We do not accept the proposition that statistical proof has to be presented before a medical expert can testify on causation. The fact that no study has statistically proven that VZIG is therapeutically effective in protecting the fetus from chicken pox affects the weight of Balducci's expert opinion, not its admissibility. See Reese, 128 Wn.2d at 308-09, 907 P.2d at 287 (holding that lack of statistical support for medical opinion on causation does not make the opinion inadmissible; there were no statistically significant studies proving the efficacy of Prolastin therapy when used to treat emphysema yet medical doctor was allowed to testify that failure to use Prolastin to treat plaintiff's emphysema was cause of plaintiff's worsening condition). More important, we think the Enders study qualifies as sufficient objective evidence that more probably than not VZIG has some effect in protecting the fetus from chicken pox.

The district court also stated that "Balducci's theory is not generally accepted" and cited research bulletins and the VZIG manufacturer insert. All of this literature stated there was no evidence that administration of VZIG to a susceptible pregnant woman will prevent fetal infection or congenital varicella syndrome. We point out that all of these materials were published several years before the Enders article. There is no evidence of a post-Enders study rejecting the conclusion that VZIG may reduce the risk of fetal infection. Additionally, we note that "a reliability assessment does not require, although it does permit, explicit identification of a relevant scientific community and an express determination of a particular degree of acceptance within that community." Daubert, 509 U.S. at 594, 113 S.Ct. at 2797, 125 L.Ed.2d at 483.

Interestingly, Grose concluded from children's studies that VZIG would be effective for adults and published his conclusions. Balducci through the same kind of scientific reasoning has concluded that VZIG will have a beneficial effect on the fetus if VZIG is administered before the disease process reaches the stage when it can pose a threat of harm to the fetus.

We agree with the plaintiffs that Balducci's opinions are grounded on fundamental principles of virology that have been tested, peer reviewed, and generally accepted for years. As mentioned, the Supreme Court remanded in Daubert. On remand, the Ninth Circuit employing the Daubert analysis again rejected the expert testimony in question. In doing so, the court noted that expert testimony flowing from existing research rather than developed for litigation purposes is an important factor in determining whether the proposed testimony amounts to "good science." Daubert v. Merrell Dow Pharms., Inc., 43 F.3d 1311, 1317 (9th Cir.), cert. denied, ___ U.S. ___, 116 S.Ct. 189, 133 L.Ed.2d 126 (1995). We think Balducci's expert testimony meets this "good science" test.

Significantly, Grose agrees with the basic science underlying Balducci's opinions. Balducci's conclusion about VZIG's beneficial effect on the fetus is based upon those principles, corroborated by a legitimate scientific study by Enders, and supported by Balducci's own experience. Balducci's methodology in reaching this conclusion — reasonable extrapolation from available scientific and medical information — is the same methodology employed by the defendants' own expert when he reached the conclusion from children studies that VZIG administered in a timely manner is an effective treatment for pregnant women exposed to chicken pox. Thus, a logical and reliable basis exists for Balducci's opinion about the beneficial effects of VZIG on the fetus; any weaknesses in that opinion goes to its weight, not to its admissibility. See Compton, 82 F.3d at 1518 ("[A]s long as a logical basis exists for an expert's opinion . . . the weaknesses in the underpinnings of the opinion go to the weight and not to the admissibility of the testimony.") (quoting Jones v. Otis Elevator Co., 861 F.2d 655, 663 (11th Cir. 1988)).

We conclude that the reasoning and methodology underlying Balducci's opinion about the beneficial effects of VZIG on the fetus are grounded in "scientific knowledge," "based upon what is known." Daubert, 509 U.S. at 590, 113 S.Ct. at 2795, 125 L.Ed.2d at 481. That science has not had the time, resources, or inclination to do the statistical studies to test Balducci's opinion about the beneficial effects of VZIG on the fetus does not preclude his opinion. This is because his reasoning and methodology in reaching that opinion are based on scientifically valid principles and are therefore sound and reliable for evidentiary purposes.

2. "The Fit": Relevancy. One of the elements the plaintiffs have to prove is the connection between the failure to administer VZIG to Melissa and the injuries suffered by Maxwell, a causation issue. As mentioned, Balducci's causation opinion boils down to this: Reliable medical literature and clinical experience show that VZIG provides effective protection to pregnant women against the chicken pox virus provided that the VZIG is administered within ninety-six hours of exposure. After ninety-six hours, the chicken pox infection reaches the stage when it is a threat to the fetus. According to scientific reasoning, by destroying the viral particles in the mother before they can be a threat to the fetus, VZIG will benefit the fetus as well.

Melissa's pregnancy, the alleged time of exposure to the chicken pox virus, the non-treatment of Melissa within the ninety-six hour time frame, Maxwell's alleged congenital varicella syndrome, and Balducci's opinion of VZIG's preventative capabilities based on scientifically valid principles all "fit together" to make Balducci's testimony relevant on the causation issue. These facts together with Balducci's opinion of VZIG's preventative capabilities would assist the trier of fact to understand the evidence and determine the causation issue.

E. Rule 403: probative value versus unfair prejudice. Iowa Rule of Evidence 403, which is identical to Federal Rule of Evidence 403, allows a trial court to exclude relevant evidence if its probative value is substantially outweighed by the danger of unfair prejudice, confusion of the issues, or misleading the jury. The district court employed rule 403 and excluded Balducci's opinion for the additional reason that its probative value would be far outweighed by its unfair prejudicial effect because the case involves the birth of a child with significant injuries.

The district court treats substantial injuries as inherently requiring a higher standard of evidence than other cases. Daubert, as well as Iowa case law, indicates that unfair prejudice for scientific evidence is more linked to the complexity of the evidence and the unwillingness or inability of the expert to explain it, rather than to the type of injuries involved. Daubert, 509 U.S. at 595, 113 S.Ct. at 2798, 125 L.Ed.2d at 484 (recognizing that "expert evidence can be both powerful and quite misleading because of the difficulty in evaluating it"); Hall, 297 N.W.2d at 85 ("Obviously the complexity of the subject-matter will influence the foundational showing of reliability.")

Rule 403 allows the trier of fact to exclude relevant evidence. Because it does so, courts should apply the rule sparingly. Hendrix v. Raybestos-Manhattan Inc., 776 F.2d 1492, 1502 (11th Cir. 1985); Villari v. Terminix Int'l, Inc., 692 F. Supp. 568, 572 (E.D.Pa. 1988).

We have set out Balducci's reasoning underlying his opinion about the beneficial effect VZIG has on the fetus. We doubt the jury would have much difficulty in understanding or evaluating his reasoning. As the Reese court said in a case involving a similar evidentiary problem:

[The plaintiff's expert's testimony is] based on the information known to the medical profession at the time of the plaintiff's treatment, [and] is the type of information jurors and their physicians rely on in their everyday lives to make decisions about health care. There is nothing mystical about it, and jurors are perfectly capable of determining what weight to give this kind of expert testimony. A jury can certainly evaluate the foundation for [the medical expert's] opinion that the failure to prescribe Prolastin therapy caused a preventable worsening of the [p]laintiff's condition. Furthermore, the jury can evaluate the [d]efendant's reasons for failing to apply Prolastin as well as the lack of substantial statistical support concerning the therapy's efficacy.

Reese, 128 Wn.2d at 309, 907 P.2d at 287 (citation omitted).

If Balducci's opinion is as shaky as the defendants contend it is, they will have adequate opportunity to vigorously cross-examine him and present contrary evidence from their own expert to convince the jury to reject it. See 3 Jack B. Weinstein Margaret A. Berger, Weinstein's Evidence ¶ 703[03], at 703-42 (1994) ("Allowing extensive cross-examination so as to elicit the expert's assumptions and test the expert's data may be a more appropriate method of enabling the jury to perform its function than excluding information that passes the tests of relevancy and prejudice."). We have every confidence in a jury's ability to listen to all the evidence and reach a reasoned verdict based on proper instructions from the court. We see no danger of unfair prejudice in this case.

III. Disposition.

In sum, we conclude that Balducci's expert testimony satisfies the requirements of Daubert. Additionally, we conclude there is no danger of unfair prejudice because of such testimony. We therefore hold that the district court abused its discretion in excluding the testimony. We remand for further proceedings consistent with this opinion.

We have considered all of the arguments whether or not we have addressed them. Those we did not address are either without merit or not properly preserved.

REVERSED AND REMANDED.