Evidence-based medicine in court: if you don't know what evidence-based medicine is, read on. Chances are it will soon play a part in one of your cases.

Seven years ago, Alan Trach went to see his dentist for pain in his jaw. The dentist prescribed the antibiotic Amoxil, and Trach took the prescription to Thrift Drug to be filled. The pharmacist's assistant on duty mistakenly filled the prescription with Doxepin, an antidepressant.

After taking the Doxepin at the dosage the dentist had prescribed for Amoxil, Trach began to experience strange sensations, including shakiness, confusion, nausea, and headaches. He also had difficulty walking and talking. The next day, Trach began having vision problems in his right eye.

Trach had taken 23 of the 40 pills prescribed when he learned that he had been given the wrong medicine and had been taking it in an amount that was four times the maximum daily dosage. He suffers permanent neurological injury and glaucoma as a result of the overdose.

At trial in a lawsuit against the pharmacy and others, Trach's medical expert, John Shane, admitted that there were no references in the medical literature showing that overdoses of Doxepin could cause brain and eye injury. For obvious reasons, no one had given overdoses of Doxepin to human subjects in an empirical study.

Instead, Shane extrapolated from the elements of anatomy, physiology, and biochemistry, testifying that Doxepin interferes with the transmission of the neurotransmitter acetylcholine, causing an anticholinergic effect (the blocking of impulses through the parasympathetic nerves). Shane also testified that Trach's symptoms were consistent with the adverse reactions identified in the manufacturer's package insert for Doxepin and in the Physicians Desk Reference (PDR). The jury found for Trach, and the defendants appealed.

Pennsylvania Superior Court Judge Phyllis Beck reversed the judgment and ordered a new trial on damages (liability had been conceded). (1) She found that it was "impossible to infer from the PDR that the scientific community generally accepts the proposition that an overdose of Doxepin can cause glaucoma."

The judge noted that "the absence of medical literature or published studies relating the effects of an overdose should not in itself bar the testimony. However, in the absence of systematic studies, the trial court must examine the methodology underlying the expert's testimony." Accepting the testimony of the defense experts, Beck found that Shane's methodology was "flawed and not generally accepted."

Experts who practice in fields that have traditionally involved both art and science--including doctors, lawyers, and educators--and who offer opinion testimony must be prepared to back it up with scientific proof. Expert opinions that are based on logic or fundamental laws of science, like Shane's, are increasingly disallowed by post-Daubert courts acting as gatekeepers. (2)

Consequently, plaintiffs are being cut off at the pass; juries never even get to consider the value of their experts' testimony. Instead, courts are increasingly allowing expert testimony that merely parrots epidemiologic studies, even poorly designed ones. The substance of this testimony is generally referred to as "evidence-based medicine," and it is supplanting the traditional battle of the experts in litigation.

Coined by researchers in the early 1990s, (3) the term "evidence-based medicine" describes a clinical strategy that uses current research findings as the basis for medical treatment decisions. It has been defined as the "conscientious, explicit, and judicious use of the current best evidence in making decisions about the care of individual patients." (4)

Medical professionals, who have traditionally had wide latitude in making clinical decisions, are increasingly turning to this accumulated "best evidence," which narrows the choice of treatments to established norms. (5) Rooted in epidemiology, evidence-based medicine can be used to make decisions regarding a patient's diagnosis, prognosis, or ongoing care. (6)

Developments in clinical research over the last 40 years have led to this paradigm shift. In 1960, the randomized clinical trial was a rarity. Today, no drug can enter the market without a clinical demonstration of its safety and efficacy. Similar trials are increasingly being applied to surgical therapies and diagnostic tests. In essence, these studies have become the basis for the practice of medicine--evidence-based medicine.

Many courts view evidence-based medicine as a "science" that can establish medical standards of care--called clinical practice guidelines (CPGs)--and resolve causation questions. Honestly developed evidence-based medicine and clinical practice guidelines can be a boon to health care and help ensure just outcomes in civil litigation. But when these guidelines have been corrupted, either directly or as a result of manipulated medical studies, they can be used to gain unjust results. (7)

Evidentiary guidelines

American courts often have struggled with the problems of accepting or rejecting evidence produced by emerging technologies (for example, DNA analysis, fingerprints, and lie detectors). These struggles have produced a string of decisions guiding and empowering judges in making evidentiary decisions. How evidence-based medicine and CPGs will fare in light of these decisions remains to be seen, but we can make some predictions.

Frye v. United States held that evidence is admissible only if it is sufficiently established to have gained general acceptance in the pertinent field. (8) The general acceptance standard would be irrelevant to a court that looks to evidence-based medicine as the criterion for accepting or rejecting expert testimony. If an expert's opinion could not be supported by empirical evidence and study replicability, it is foreseeable that mere expert opinion alone will not suffice.

United States v. Barker held that trial judges had the right and responsibility to determine "whether or not to allow the jury to consider a proffered expert's opinions ... [and whether the opinion] will most likely assist the trier of fact in arriving at the truth." (9) This case clearly empowered judges, in their role as gatekeepers, to exclude expert opinion testimony. As evidence-based medicine gains acceptance in the courts, judges probably will exclude opinion testimony that lacks statistical/ epidemiological foundation. Unfortunately, much of the medical-expert opinion testimony offered today does not have epidemiological support.

The purpose of our court system is to reach just outcomes, not necessarily to determine the scientific truth of every underlying issue. We hope judges will recognize this and allow opinion testimony when there is consensual, if not statistical, support for it, especially when the subject of the testimony has not yet been subjected to epidemiological study. If judges, on the other hand, disallow all expert opinion testimony lacking epidemiological proof, many parties in both civil and criminal cases will be severely restricted in the evidence they can offer to support their positions.

In Daubert v. Merrell Dow Pharmaceuticals, Inc., the U.S. Supreme Court inched closer to creating an objective standard for weighing expert testimony: It determined that the trial judge, acting as gatekeeper, is responsible for insuring that an expert's testimony rests on a reliable foundation and is relevant to the task at hand. (10) The question is, does a "reliable foundation" require epidemiological proof?

If the answer to that question is yes, consider the impact of General Electric Co. v. Joiner. (11) In that case, the Supreme Court affirmed a trial judge's decision to exclude evidence, based on animal studies, that supported the plaintiff's claim that PCBs cause cancer. This presents a problem for plaintiffs in similar cases. Because medical ethics and the Geneva Convention limit human experimentation, many epidemiological studies on humans are not possible.

In Kumho Tire Co. v. Carmichael, the Supreme Court said that a trial judge has broad discretion in applying Daubert standards for assessing the admissibility of expert evidence. (12) Because plaintiffs in civil cases and prosecutors in criminal cases bear the burden of proof, they are the most likely to be adversely affected by the convergence of evidence-based medicine, CPGs, and the high bar of admissibility implicit in Daubert. That decision, coupled with Kumho Tire, puts science--and the judges who judge it--in the position to replace juries in weighing expert testimony.

In Weisgram v. Marley Co., the Supreme Court held that federal appellate courts can direct a district court to enter judgment notwithstanding the verdict against a winning plaintiff if the reviewing court finds that testimony admitted at the trial level is unreliable and inadmissible under Daubert. (13)

To date, evidence-based medicine has been used most to resolve questions of causation rather than standards of care. As more CPGs are developed, however, opinion testimony as to both causation and standard of care will be subject to Daubert or Frye challenges based on evidence-based medicine and CPGs. (14)

Lawyers and judges will increasingly discover that this new "science" may determine ultimate questions of fact. The down side of this trend is illustrated by recent examples of how epidemiological studies can be manipulated to produce certain findings.

After years of litigation, the science supporting claims that silicone-gel breast implants cause connective tissue and other systemic diseases was undercut by the findings of a so-called science panel appointed by U.S. District Judge Sam Pointer under Federal Rule of Evidence 706. Although the panel did no research on its own, it reviewed "scientific/epidemiologic and other studies," the vast majority of which were funded by defendants. It determined that the plaintiffs lacked "scientific" proof that silicone-gel breast implants caused disease.

Looking at how this "peer-reviewed scientific evidence" on silicone was developed, published, and evaluated highlights a major problem with evidence-based medicine and nonjury "science panels."

When the first breast implant cases emerged, Dow Chemical--the parent of Merrell Dow and Dow Corning and a party in the Daubert litigation--hired a number of academic and physician medical consultants. Many of them were also on the editorial boards of peer-reviewed medical journals and would ultimately determine which articles submitted by silicone researchers would be published.

The company also retained epidemiologists at some of the most prestigious universities in the country and commissioned studies that critics assert were designed, in part, by Dow and were structured to find results that would exculpate rather than implicate silicone gel and the implant manufacturers. Using the results of these studies, the company put together a packet of "scientific" research supporting the defendants' position.

This packet was delivered to the editorial offices of every major U.S. newspaper. It is not surprising that a glut of editorials appeared declaring that silicone-gel breast implants were safe and that the plaintiffs' claims were based on "junk" science.

This "Dow science" also served as the key reference for Judge Pointer's science panel. It determined the panel's findings, which were, not surprisingly, adverse to the claimants. After this, the litigation of individual breast implant cases collapsed.

Another example of how evidence-based medicine could affect the outcome of litigation involves the publication of misleading research in the Journal of the American Medical Association. The journal's editor, Catherine DeAngelis, published the results of a six-month study of the popular arthritis drug Celebrex. This study showed that Celebrex caused fewer gastrointestinal problems than comparable medications. Later, DeAngelis learned that the study had actually lasted a year and that the complete data pointed to a very different result.

U.S. News & World Report reported:

   The editors from the other major medical journals shared DeAngelis's anger
   and frustration. In a joint editorial published this week, the heads of 12
   international journals ... warn that pharmaceutical companies have gained
   too much control over both medical research and the publication of
   experimental results. Clinical investigators are often excluded from study
   design and data interpretation ... and in some cases unfavorable results
   are intentionally buried to avoid public dissemination. In an effort to win
   back some control over the scientific process, the editors declared a
   radical change in policy.... "We will not review or publish articles based
   on studies that are conducted under conditions that allow the sponsor [drug
   company] to have sole control of the data or to withhold publication." (15)

The same article quotes Harold Sox, editor of the Annals of Internal Medicine: "In the worst cases, the drug firms design the trial [study], explain to physicians how to carry it out, analyze the study, do not let the researchers see all of the data, and then control the publication."

Dow Chemical also designed and funded the early epidemiological studies establishing that Dioxin did not cause cancer in veterans exposed to Agent Orange. After Congress created the Agent Orange Compensation Fund, which in effect relieved Dow Chemical of any liability, the company stopped funding "scientific research" in this area.

Independent studies were then published showing that Dioxin is a human carcinogen. These studies turned the science of Dioxin 180 degrees. Today, the Environmental Protection Agency lists Dioxin as a human carcinogen.

The silicone-gel breast implant litigation, the Celebrex drug study, and the Dioxin studies illustrate the greatest danger of evidence-based medicine. Where the financial stakes are high, major corporations--more so than individual litigants--have the resources to design, define, and fund science that supports their case. These case illustrations show how the rich and powerful can manipulate evidence-based medicine to subvert both the civil and criminal justice systems.

Unresolved questions

Because much medical research consists of case reports or small case studies with a limited number of patients, quality data outcomes are sparse. Treatments for some conditions or diseases have no empirical proof of efficacy or safety. Regardless, these treatments may be recognized by a particular medical specialty as the standard of care.

What happens in a world where CPGs become the accepted standards of care? In a case where treatment for a specific condition or disease is not supported by empirically tested data outcomes, is there no admissible standard of care? Does it also mean that insurance will not cover the treatments because there is no proof of medical efficacy?

Courts typically do not have the luxury of holding their decisions in abeyance until a body of research develops. Although there may be expert opinions and animal studies, will the absence of human population epidemiological studies prevent claimants from proving their cases? Indeed, would a proliferation of evidence-based medicine prevent courts from resolving a significant number of disputes? Would this shift favor criminal and civil wrongdoers?

In a civil case, verdicts are determined by the "great weight of the evidence." This is often defined as 51 percent. Yet in science, a finding is not deemed reliable unless it is proven to have at least 95 percent reliability. Does this mean that the civil proof requirement will de facto shift from 51 percent to 95 percent? And would such a standard, if adopted, deny access to the courts and prevent the courts from performing their duties to resolve disputes and maintain order? At least in criminal cases, where the burden Of proof is "beyond a reasonable doubt," a 95 percent standard may be acceptable.

Whether CPGs will be admissible as standards of care in the absence of expert testimony has not yet been determined. CPGs and other published standards can be used to cross-examine or serve as a foundation for expert testimony in civil cases. If CPGs are adopted by federal agencies and used as standards of care and treatment for Medicare and Medicaid patients, they may constitute admissible "public records" under Federal Rule of Evidence 803(8).

The largest single source of CPGs is the National Guideline Clearinghouse, which currently contains 988 clinical practice guideline summaries. Panels of physicians preparing the CPGs evaluate and grade the great mass of scientific studies on a given subject. Guidelines based on powerful evidence receive the highest grades.

Some of the grading scales refer to levels of evidence. Level I evidence is usually the most powerful and is derived from randomized controlled trials. Level III evidence is based on studies that provide limited information about the relationship between a treatment and an outcome. This category includes descriptive studies--such as case reports and case series--and expert opinion that is often based on clinical experience.

Practice points

As evidence-based medicine continues to emerge, trial lawyers should expect motions in limine and Daubert challenges to a broad range of experts and components of their testimony. (16) AS a result, litigators will face increasing costs as these challenges cause bifurcation and may increase the time it takes before the case comes to trial. Proponents claim that when the trial does arrive, it is shorter because many of the disputed science issues have already been resolved.

Before one of your experts is deposed, make sure he or she has human studies, preferably case-controlled ones, that support the proffered opinions. Tangential human or animal studies may not be sufficient to support the testimony.

In anticipation of motions in limine and Daubert challenges, when you are drafting interrogatories, consider asking, "Do you intend to file a motion in limine or make a Daubert challenge to Dr. Smith or any of the components of his opinion?" If the answer is yes, you can prepare for this before trial. If there is no answer, or if it's no, and the opposing party challenges your expert at trial, in most courts the failure to give you notice will empower you to substitute experts or adjourn for more time.

Trial lawyers will have to learn how study design and power variables are analyzed. Until each study carries a grade with it, lawyers will have to evaluate them on their own. Some factors to consider include whether a study is published in a peer-reviewed, as opposed to a nonpeer-reviewed, journal; whether the number of subjects in a study is 10 or 10,000; and whether a study is funded by industry defendants or by a respected and independent organization like the National Institutes of Health.

Evidence-based medicine and CPGs are not all bad or all good for civil plaintiffs. Because plaintiff attorneys frequently have trouble mustering medical expertise to support their clients' claims, CPGs can help provide the necessary proof of standards of care.

Applying evidence-based medicine to causation issues will probably help defendants more than plaintiffs. Because it has not addressed causation questions in many human subjects, plaintiffs, who bear the burden of proof, may be unable to establish causation for lack of empirical data.

The best advice, as always, is be prepared. Read up on evidence-based medicine and its progeny, CPGs. They are probably here to stay, so we'll just have to learn to work with them.

Notes

(1.) Trach v. Fellin, Nos. 01921EDA00, 01949EDA00 (Pa. Super. Ct. Jan. 18, 2002).

(2.) See Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579 (1993).

(3.) Sharon Staus, Evidence-Based Medicine: A Commentary on Common Criticisms, 163 CAN. MED. ASS'N J. 837-41 (2000).

(4.) Consensus Statement, Evidence-Based Medicine: A New Approach to Teaching the Practice of Medicine, 268 JAMA 2420-25 (1992).

(5.) Arnold J. Rosoff, Evidence-Based Medicine and the Law: The Courts Confront Clinical Practice Guidelines, 26 J. HEALTH POLITICS & L. 327-68 (2002).

(6.) Epidemiology is the emerging "science" of studying factors that influence the distribution, occurrence, and prevention of health-related events--like disease, injury, and pharmacological effects--in human populations.

(7.) Concerned that plaintiffs could use clinical practice guidelines to their advantage, defense interests--including the insurance industry--have worked to pass statutes limiting their use to defendants in civil litigation. These efforts were initially successful in Maine and Minnesota, but pro-defense laws in those states have since been repealed. 24 ME. REV. STAT. ANN. tit. 24 [section] 2971 (1993); MINN. STAT. [section] 62J-34 (1995).

(8.) 293 F. 1013, 1014 (D.C. Cir. 1923).

(9.) 553 F.2d 1013, 1024 (6th Cir. 1977) (quoting Holmgren v. Massey-Ferguson, Inc., 516 F.2d 856 (8th Cir. 1975)).

(10.) Daubert, 509 U.S. 579, 594-95.

(11.) 522 U.S. 136 (1997).

(12.) 526 U.S. 137, 152-53 (1999).

(13.) 528 U.S. 440, 454-56 (2000).

(14.) Two state courts rejected the application of Daubert to standard-of-care issues in medical malpractice cases. Gilkey v. Schweitzer, 983 P.2d 869, 872 (Mont. 1999); Reese v. Stroh, 907 P.2d 282, 306-08 (Wash. 1995). In two other states, appellate courts determined that trial judges had abused their discretion in excluding expert testimony on causation under Daubert. Williams v. Hedican, 561 N.W.2d 817, 827-32 (Iowa 1997); Bunting v. Jamieson, 984 P.2d 467, 474 (Wyo. 1999). An equal number of decisions from state appellate courts have reached contrary conclusions on both of these issues.

(15.) Stacey Schultz, True, False, Whatever, U.S. NEWS & WORLD RER, Sept. 17, 2001, at 72.

(16.) Some of the problems touched on in this article are more specifically addressed in Martha K. Wivell, When Can Science Go to Court? TRIAL, Mar. 2001, at 27; Stuart Ollanik, Defeating Daubert Challenges in Auto Defect Cases, TRIAL, Sept. 2001, at 28; Ira H. Leesfield & Mark A. Sylvester, Admissibility of Expert Testimony: Whats Next? TRIAL, Dec. 2000, at 64.

J. Douglas Peters is a shareholder with Charfoos and Christensen in Detroit and an adjunct associate professor at the Wayne State University School of Medicine.