Off the Trach: The Pennsylvania Frye Rule Applied to Medical Causation in Civil Cases

It is unfortunate that in cases such as this, when science enters the courtroom, common sense must leave. Trach vs. J. Fellin and Thrift Drug, Eckerd Store, 2002 Pa. Super. Lexis p. 11, 24-25 (Ford Elliot, J. concurring).

Introduction

You’ve just gotten a multi-million dollar case. Your client, an Internet retailer, has been losing tons of sales due to an unscrupulous competitor’s manipulation of Internet search engines. You do the legal research and can find no cases on all fours with yours. That is not surprising B Internet case law really hasn’t had much chance to develop. Undaunted, you stitch together a tight argument based on threads from cases on unfair competition and tortious interference. After all this, this is what courts do all the time -- take established lines of legal reasoning, apply them to new fact situations and set new precedents, right?

“Wrong,” says the court, “Your legal theory is novel. It has not been generally accepted in the legal community.” “But the Internet is too new,” you answer. “Nothing about it could be generally accepted.” “It doesn’t matter”, intones the court. “We can’t apply any novel legal theory until it’s been generally accepted. You lose. Is this story a Kafkaesque view of the future? No, it is an example of the strait jacket which recent interpretations of Pennsylvania’s Frye rule have placed on scientific evidence.

In an earlier series of articles , this author argued that Pennsylvania appellate decisions since 1996 misinterpreted and misapplied the Frye rule to the issue of medical causation in civil cases. The article looked forward to the then-pending Pennsylvania Supreme Court decision in Blum v. Merrell Dow, which was expected to decide two critical issues. First, would the test for admissibility be the Frye test or the U.S. Supreme Court test in Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993)? Second, and perhaps more significantly, exactly how would the Frye rule be applied? Would the court require that the expert’s scientific conclusions be generally accepted, or simply that the expert’s methodology be so accepted?

Unfortunately, Blum answered neither question. Holding that the expert’s testimony in Blum was so inadequate as to render his conclusion inadmissible under either Frye or Daubert, the court considered it to be “jurisprudentially unsound” to decide these issues in Blum. 564 PA 3,8; 764 A.2d 1,4 (2000). The Frye rule was left to stand without any clarification as to what the Frye standard really meant.

In a stinging dissent, Justice Cappy observed:

This Court has long interpreted Frye as requiring that the methodology employed by the testifying scientist be generally accepted in the scientific community. Yet, we have not stated that the conclusion reached by the scientist regarding causation must also be generally accepted in the scientific community.

. . . [T]his additional step in the Frye test, requiring the conclusion also be generally accepted by the scientific community, was added by the Commonwealth Court in McKenzie v. Westinghouse Electric Corp., 74 A.2d 1167 (Pa. Cmwlth. 1996). 564 Pa. at 9, 764 A.2d at 9.(Emphasis supplied)

Justice Cappy went on to “squarely reject that portion of the Superior Court=s holding which would require a scientist’s conclusions, as well as the methodologies utilized in reaching those conclusions, are generally accepted in the medical community.” Id. at 10, 11.

Since Blum, the Court has issued a new rule, Pa. R.Civ. P. 207.1, to deal with the flood of Frye motions , and the Superior Court has decided two civil cases under the Frye rule; Trach v. Thrift Drug, supra.; and Grady v. Frito Lay, Inc., 2001 Pa. Super. Lexis 3537. The Trach case is especially important, as the court has granted a hearing en banc, and oral argument will be heard in late June. 2002 Pa. Super Lexis 400.

The Pennsylvania Frye Rule

The ostensible goal of the Pennsylvania Frye rule is to allow the scientists to judge the scientists; if the scientists accept the science, then the courts will as well. Pennsylvania considers this superior to the Daubert standard, which “substitutes a judicial evaluation and determination of scientific reliability.” Blum, supra, 564 Pa. at 6, 764 A.2d at 3.

Is that, in fact, how the Superior Court has applied Frye? To put Trach and Grady in their proper perspective, we must review two cases which preceded the Supreme Court’s decision in Blum: Wack v. Farmland Industries, Inc., 744 A.2d 265 (1999) and Thomas v. West Bend Company, 760 A.2d 1174 (1999). In Wack, the plaintiff claimed to have developed adenocarcinoma of the buccal cavity as a result of drinking water contaminated with gasoline. The court reviewed the studies proffered by the plaintiff’s expert and found that they were “not in the form of a firm conclusion.” The court stated “at best, the studies suggest a possible link between exposure to petroleum products and the incidence of buccal carcinoma.” 744 A.2d at 270 (emphasis original). The Wack court went on to say:

Neither Dr. DiGregorio, nor the studies on which he relied, make a distinction between Mrs. Wack’s rare type of adenocarcinoma and the more common squamous cell carcinoma of the buccal cavity. Thus, even if we accept the proposition that the cited studies support a general causal connection between exposure of petroleum products and buccal cancer, the studies plainly do not support a causal link between benzene and Mrs. Wack’s specific and rare of form of cancer. 744 A.2d at 270-71. (emphasis supplied)

This statement is extraordinary for two reasons. First, it suggests that reasonable people may have interpreted the studies as showing a stronger association than the court did, for at another point, the court stated “it is arguable whether (plaintiff’s expert) appropriately analyzed the results of the studies on which he relied.” (Id. at page 271). This implies that it is also arguable that he analyzed the results appropriately. Those are the kinds of arguments, which used to be heard by juries.

The court also admitted that the medical literature draws no distinction between squamous cell and adenocarcinoma of the buccal cavity. A logical conclusion is that this is because epidemiologists who studied buccal cancer could find no valid scientific reason to separate out the two cell types. Instead, the court jumped to its own pseudo B scientific conclusion B that squamous cell carcinoma and adenocarcinoma have different causes -- without any basis on the record to support it.

In reviewing Wack, we must revist critical language in Frye which gave birth to this whole line of case law:

Just when a scientific principle or discovery crosses the line between the experimental and the demonstrable stage is difficult to define. Somewhere in this twilight zone, the evidential force of the principle must be recognized, and while courts will go a long way in admitting expert testimony deduced from a well recognized scientific principle or discovery, the thing from which the deduction is made must be sufficiently established to have gain generally acceptance in a particular field in which it belongs. 293 F. at 1014. (emphasis supplied)

What is the “thing” from which the deduction is made? If studies show that benzene causes buccal cancer generally, can an expert deduce from those studies that benzene caused adenocarcinoma of the buccal cavity as well? The Wack court allowed no such deduction.

In Thomas v. West Bend Co., 760 A.2d 1174 (Pa Super 2000), plaintiff claimed that a voltage shock from an electrical appliance caused cardiomyopathy, a condition which plaintiff’s expert, a cardiologist, examined him, reviewed his medical records and found no other explicable cause for the cardiomyopathy. Citing two articles showing that high voltage electrical current could cause cardiomyopathy, the expert actually prepared a medical article, which he intended to submit for publication.

Although the trial court found that this proposed article “will advance medical knowledge” the court excluded the evidence on the grounds that the doctor’s theory was a “newly described etiology” and not generally accepted. 45 Pa D&C 4th at 485, 486. On appeal to Superior Court, Thomas argued that because the expert used all of the traditional tools of cardiology --- EKG results, biopsy studies, echocardiagram, bloodwork, the patient’s history and films -- his resulting conclusion should be admissible even if it had not yet been generally accepted. 760 A.2d at 1179. The court, citing the Superior Court’s decisions in Blum and Wack stated that:

While Dr. DePace has taken the first step towards establishing general acceptance of that proposition (by drafting an article for publication) it was not generally accepted at the time of the trial court’s decision. The trial court, therefore, did not err by excluding this untested hypothesis. (Id. at 1180 (emphasis supplied).

What did the court mean by this? Would Thomas have been decided differently of Depace’s article had been published before he was involved in litigation? What scientific (let alone legal) sense does that make?

In Grady v. Frito Lay, the Superior Court’s first decision after the Supreme Court’s decision in Blum, the plaintiff claimed that a defectively jagged tortilla chip caused an esophageal tear. The trial court excluded his expert’s testimony as insufficient under Frye. The Superior Court reversed, holding that the expert could reasonably rely on medical literature to support the opinion. In this case, the literature consisted of several case reports (but no epidemiologic studies) associating esophageal tears with corn chips. The court also allowed into evidence tests done by another expert on the subject tortilla chips showing that they could cause esophageal tears.

In admitting the evidence, the court cited U.S. Supreme Court’s decision in Kumho Tire v. Carmichael, 526 U.S. 137 152 (1999) stating that the role of the trial court is “to ensure that an expert, whether basing testimony upon professional studies or personal experience, employs in the courtroom the same level of intellectual rigor that characterizes the practice of an expert in a relevant field.” 2001 Pa. Super Lexis 537. p.16. The court, in its en banc opinion found the tests employed basic calculations and used basic scientific principles, although their conclusions were hotly disputed.

Significantly, Judge (now Justice) Eakin, who wrote the Wack opinion, dissented in Grady. According to Judge Eakin, the “manifest difference between this methodology and basic mastication cry out for more than is found here. [The expert’s] analysis makes a leap of logic in this and other questions, and the trier of fact could not be expected to fill up the gaps.” In short, the majority in Grady allowed the plaintiff’s expert to extrapolate from his mechanical tests to the actions of the mouth. The dissent would not allow this extrapolation.

The Grady court’s quotation of Kumho is especially ironic. Although everyone would agree that experts in the courtroom should apply the same level of “intellectual rigor” that they do in their professional practices, that may not be sufficient to survive the current application of Frye in the Pennsylvania courts. Dr. DePace did this in Thomas, yet his testimony was excluded. And Pennsylvania=s appellate courts have recognized he would not be the only one.

In Tagliati v. Nationwide Insurance Company, 720 A.2d 1051 (1993), the Superior Court had to decide whether thermogram treatment should be reimbursable under the Pennsylvania Motor Vehicle Financial Responsibility Law. It noted that many courts have held thermographs to be inadmissible, but:

The instant appeal does not require us to decide whether the thermograph results have gained scientific acceptance so as to be admissible under Frye. Rather the questions at issue here is whether thermograph is a reasonable and necessary medical treatment, within the meaning of the MVFRL. 720 A.2d at 1055. (emphasis supplied)

The court stated:

Medical technology is advancing more rapidly than the law. It is thus conceivable that a cutting-edge procedure, device or service may fall within the meaning of reasonable and necessary medical treatment, even though it has not gained generally acceptance within the members of the medical community.@ Id. (emphasis supplied)

The court went on to conclude that thermograms were a reasonable and necessary medical treatment, (i.e. that it entirely proper doctors to use them in their private practices) but they still might be inadmissible under Frye. 720 A.2d at 1055 5.

Similarly, the Supreme Court has expressly refused to apply Frye rule to testimony about the standard of care in malpractice cases. A physician cannot be held liable for malpractice where, in the exercise of his judgment, he followed a course of treatment advocated by a considerable number of recognized and respected professionals. In Gala v. Hamilton, 552 Pa. 466, 715 A.2d 1108 (1998) the Supreme Court held that an expert opinion that a medical school of thought existed should be allowed into evidence even if the expert could cite no medical literature to support its existence. The court stated “limiting evidence to medical literature would have the effect of preventing expert witnesses from testifying to the existence of a school of thought based on their experience as practitioners and the information they obtained during the medical training.” Id. at 1111 (emphasis supplied). The Supreme Court thus recognized that experts could apply “the same level of intellectual rigor” in court as in their private practices, even if they had no medical literature to support their testimony.

How Frye Got Things Off The Trach

This dichtomy, between allowing experts to apply their everyday methodologies and requiring their testimony to be generally accepted played out in the recent case of Trach v. Thrift Drug. In Trach, the pharmacist mistakenly filled a prescription with doxepin, an anti-depressant, instead of an antibiotic. Plaintiff took the pills faithfully; since the prescription was wrong, he wound up taking several times the maximum recommended dose of doxepin. Mr. Trach developed severe dizziness, confusion, headaches, nausea, blurred vision, and pain in his right eye. Most of the symptoms disappeared within a month, but his vision problems and some cognitive problems were permanent.

Plaintiff’s expert was a toxicologist who testified in great detail about the mechanism by which doxepin works: causing unequal dilation of the pupil, affecting the ciliary muscle of the eye so as to cause blurred vision which would also lead to closed angle glaucoma. The Superior Court agreed that expert testimony that the short term symptoms that Trach experienced were proximately caused by the drug overdose was admissible under Frye; indeed, they were consistent with recognized adverse reactions to a therapeutic dose of doxepin. The critical question was whether the long term side effects could be appropriate subjects of expert testimony. The courts agreed that because the case involved an unanticipated overdose of the drug, research studies on the effect of the overdose would not be expected. Both courts thus found that the absence of medical literature should not, by itself, bar any expert testimony.

The courts purported to look only at the expert’s methodology and not his conclusion. The trial court found that the expert’s reliance on “references in the PDR (Physicians Desk Reference) to neurological and cognitive adverse reactions do not indicate whether these conditions are permanent or temporary. Since Trach has the burden of proof in this issue, he cannot rely on the ambiguity of the PDR to prove his point. The PDR is not evidence of a scientific consensus that doxepin can cause permanent cognitive difficulties.” 46 D&C 4th 231, 256 (2000). The courts claimed that the expert’s methodology was not generally accepted because he could point to no literature to support his claim that there was no significant difference between open angle and closed angle glaucoma. The PDR, however, also did not distinguish between open angle or closed angle. Even a therapeutic dose would be contra-indicated by any glaucoma. The trial court noted “the fact that glaucoma is identified as a contra-indication suggests a possible biological relationship. However, it is impossible to infer from the PDR that the scientific community generally accepts the proposition that an overdose of doxepin can cause glaucoma.” Id. at 257; also see quoted at 2002 Pa. Super. Lexis at 18. (emphasis supplied).

Importantly, the Superior Court claimed it was deciding the Frye question Aprimarily on the methodology prong and thus consistently with the dissents of Justices Castille and Cappy in Blum.@ 202 Pa. Super. Lexis at 19 4. Yet, its own words show it looked to general acceptance of the proposition that an overdose of doxepin causes glaucoma, and not the methodology.

It is generally accepted in toxicology that something that causes damage at low doses will cause even more damage at higher doses. The plaintiff’s expert in Trach simply did what toxicologists do in their every day practice B looked at the harm doxepin could cause at a therapeutic dose, and extrapolated to conclude that the greater harm that Trach suffered was due to his massive overdose. Although the conclusion was novel, it was only because the facts were novel. The methodology -- use of the dose B response principle -- was generally accepted.

Judge Ford Elliot grasped this in her concurrence. She stated that the Dr. Shane “extrapolated the main side effects of the recommended dosage of Doxepin, which include narrow angle glaucoma, dizziness, confusion and blurred vision, to reach his conclusion that Trach=s massive overdose of doxepin caused his glaucoma and other ongoing symptoms.”

Judge Ford Elliot concluded:

Despite what to the average layman would be an obvious causal link between a massive overdose of a potent drug and development of previously non-existing permanent symptoms, the same or similar to the transient symptoms the drug is known to cause a recommended doses, I must agree with the majority that Dr. Shane=s testimony failed to meet the standards of reliability required by Frye in its progeny in this Commonwealth. I write separately only to note my dismay with this outcome under the facts of this case. It is unfortunate that in cases such as this, when science enters the courtroom, common sense must leave. Id. at page 25 [check all page cites]

Judge Ford Elliot was wrong only about one thing. The Frye rule does not prevent an expert from extrapolating a conclusion from well known scientific principles. The Supreme Court of Illinois found that the Frye rule permits experts to offer testimony based on extrapolation precisely because scientists, like lawyers, doctors and virtually everyone else in everyday life, do extrapolate from known principles to estimate unknown information, as part of their normal professional practices. In doing so, they bring the same intellectual rigor into the courtroom that they use as experts in their relevant fields. And that is precisely why the testimony should be admitted.

The Illinois case is Donaldson v. Central Illinois Public Service Company, 2002 Ill. Lexis 283. The plaintiffs were the parents of four children who suffered from neuroblastoma, a rare form of cancer, which they contended resulted from their exposure to carcinogenic compounds at the defendant=s utility plant site. The Illinois Supreme Court, interpreting the same Frye decision, which has caused so much confusion in Pennsylvania, stated unequivocally that:

First, general acceptance does not concern the ultimate conclusion. Rather, the proper focus of the general acceptance test is on the underlying methodology used to generate the conclusion. If the underlying method use to generate an expert’s opinion are reasonably relied upon by the experts in the field, the fact finder may consider the opinion B despite the novelty of a conclusion rendered by the expert. 2002 Ill. Lexis 283 #21.

The Donaldson court went on:

Second, general acceptance of methodologies does not mean universal acceptance of methodologies. The medical community may entertain diverse opinions regarding causal of relationships, but this diversity of opinion, does not preclude the admission of the testimony that a causal or relationship exists if the expert used generally accepted methodology to develop the conclusion. 2002 Ill. Lexis 283 #21.

Furthermore:

[G]eneral acceptance does not require that the methodology be accepted by unianimity, consensus, or even a majority of experts. A technique is not Agenerally accepted@ if it is experimental or of dubious validity. Thus, the Frye rule is meant to exclude methods new to science that undeservedly create a perception of certainty when the basis for the evidence or opinion is actually invalid. 2002 Ill. Lexis 83 at #22.

Finally, the court noted that Frye does not make the trial judge a “gatekeeper” of all expert opinion testimony, but only:

if the scientific principle, technique or test offered by the expert support his or her conclusion is “new” or “novel”. Only novelty requires the trial court conduct a Frye evidentiary hearing to consider general acceptance. We recognize that a “new” or “novel” scientific technique is not always easy to identify, especially in light of constant scientific advances in a modern era. Generally, however, a scientific technique is “new” or “novel” if it is “original or striking” or “does not resemble something formerly known or used.” Id. at #22-23.

In short, the Illinois Supreme court held that “trial judges decide the general acceptance of the technique; a jury decides whether it will accept the expert’s conclusion which is based on that technique.” Id. at #29.

In Donaldson, the plaintiff’s experts relied upon the technique of “extrapolation” to form the basis of their opinions. The court noted that:

[E]xtrapolation is utilized in the scientific community when the medical inquiry is new or the opportunities to examine a specific cause and effect relationship are limited. . . . For example, in this instance, the small number of a neuroblastoma cases limits the study of the disease. As a result, extrapolation offers those with rare disease the opportunity to seek a remedy for the wrong they have suffered. Thus, in these limited instances, an expert may rely upon scientific literature discussing similar, yet not identical, cause and effect relationships. The fact the an expert must extrapolate, and is unable to produce specific studies that show the exact cause and affect relationship to support his conclusion, affects the weight of his testimony rather than its admissibility. 2002 Ill. Lexis 283 at #34.

The court concluded that because there was such limited study of neuroblastoma and because research is also very limited in cases of environmental exposure, which are not detected until after the offset of the illness, extrapolation was an appropriate scientific technique and the expert should be allowed to be heard.

The Donaldson case avoids the flaws that have become obvious in Pennsylvania’s application of Frye. By excluding scientific testimony that is developed through analysis of new data through accepted techniques, Pennsylvania courts can easily keep the truth out of court.

Indeed, the trial court in Trach noted “medical science is an evolving discipline and etiologies that are not generally accepted in the medical community today, may well be generally accepted tomorrow. For example, there is now a consensus in the medical community that asbestos causes lung disease. As a result, there are now many meritorious asbestos claims. Twenty years ago, these [clearly meritorious] suits would have all been dismissed under the [Pennsylvania] Frye rule. 46 D&C 4th at 264.

Thus, Pennsylvania’s interpretation of Frye undermines three important principles of Pennsylvania law:

1. That plaintiff is entitled to a trial by jury;

2. In deciding whether a plaintiff’s case should go to the jury, the plaintiff is entitled to the benefit of all favorable evidence along with all reasonable inferences arising from that evidence, resolving any conflict in favor of the plaintiff; and

3. The plaintiff’s burden of proof is by a preponderance in the evidence and not beyond a reasonable doubt. Smith v. Grabb, 705 A.2d 894, 899 (1997).

The Superior Court’s decisions do not allow the experts to make an inference or deduce any conclusion that does not appear within the four corners of scientific literature. It requires the scientific equivalent of the Acase on all fours” with the case at issue. Any ambiguity in that literature, such as the reference to glaucoma in the PDR in Trach or the reference to buccal cancer in Wack, is construed against the plaintiff, because the plaintiff has the burden of proving general acceptance, and this is one burden that cannot be met by favorable inference. Second, by requiring that the studies show a “statistically significant@ relationship, i.e. a 95% confidence level, and not merely a Asuggestion,” the courts effectively convert the burden of proof from a preponderance of the evidence to one beyond a reasonable doubt.

Why is the Pennsylvania Frye rule so unfair to plaintiffs?

Injured plaintiffs normally do not have sufficient funds to support epidemiologic research related to their legal claims. In contrast, most defendants in mass toxic tort cases has sufficient funds to choose which relations to study, how to study them, and whether to publicize the results. Finally, there are short-term effects on injured plaintiffs because of the varying interest of repeat and non-repeat players in the justice system. Most plaintiffs have no interest in funding a long-term epidemiologic study, that may have no impact on their particular case, or may take too long to have an impact. Corporations with the expectations of being sued multiple times, have a greater incentive to fund epidemiologic studies, publish those that are favorable to their defense.

The long-term effect of the misguided restrictions on evidence is simple. As more and more relevant evidence is excluded, the outcome in court cases will become less and less consistent with the truth. This, we fear, will concomitantly lead to decreased respect by the public for science and the law. Epidemiology in the Legal Arena and The search for Truth. American Journal of Epidemiology, Volume 154, No. 12, pg. S27, S33 (2001).

This is exactly what concerned Justice Castille in his dissent in Blum. He noted that although plaintiff’s experts in Blum Acertainly were impeachable, and were impeached, because of their Alitigation driven interest@, the same was true of Merrell Dow’s experts. 564 Pa. at 29, 764 A.2d at 20. Much of the science in this area, held up by Merrell Dow as the objective, generally accepted scientific view that requires exclusion of plaintiff’s experts’ >contrary conclusions, itself was a product of Merrill Dow’s litigation driven influence. Id. (emphasis original)

Justice Castille went on to say: The Superior Court recognized that the studies and data relied upon by the appellate=s experts were universally accepted “good science,” but concluded that the way those experts utilized that science to draw conclusion on causation was not. The flaw in this reasoning is that it conclusively determines that no opinion other than the initial researchers may ever be heard. In a situation, as here, where so much of the underlying research and interpretation was in the control of a party driven by a litigation incentive, such a holding would be, and is, absurd. Id. at 563 Pa. at 41 and 1164 A.2d at 26.

Conclusion

There is by now no question that the overuse of an overly restrictive application of Frye has created a wide array of unintended negative consequences. Rather than streamlining litigation, it has generated multiple layers of costly, time-consuming ancillary litigation. Rather than keeping “junk science” out of the courtroom, it has to substituted naive judicial assumptions about science for real scientific analysis. Rather than protecting jurors from the undue influence of scientific pretenders, it has tipped the balance of justice in favor of those with the money to make what passes for “mainstream science.” Rather than facilitating the resolution of cases in the courtroom, it has undermined the ability of jurors even to hear the real issues in controversies. Worst of all, it has prevented way law and science from interacting to advance both justice and truth.

For generations, lawyers, judges and juries were able to present, test and evaluate scientific evidence under the Frye standard, carefully scrutinizing an expert’s methodology, but allowing his conclusions to be accepted or rejected by a jury hearing all the evidence relevant to a case. It is time for the Pennsylvania Supreme Court to restore that ability. On its next occasion to do so, it should take the opportunity it declined in Blum. It should recognize that extrapolation be a valid scientific methods, and reject the sterile and unscientific notion that a scientist=s conclusions, drawn from valid scientific method, must nevertheless be “generally accepted” before they can serve as evidence in a court of law.