Medical Organizations Can’t Shape the Rules for Admitting Expert Testimony

96-well plateA little more than a year ago I wrote a post about the then-new Recommendations for Reporting of Incidental Findings in Clinical Exome and Genome Sequencing from the American College of Medical Genetics and Genomics (ACMG). Those Recommendations (since modified somewhat) proposed that whenever a patient undergoes whole-genome or whole-exome sequencing (WES) for any purpose, the laboratory doing the testing should always sequence and report to the ordering physician the results for 57 (now 56) genes on the ACMG’s list. Among the questions I addressed in that post was this one: “Do those Recommendations become by definition the standard of care for the specialty, immediately or in the near future?” I wondered specifically about a future case in which a doctor ordered WES, the lab analyzed only the genes the doctor was interested in, leaving out some of the ACMG’s 56, and the patient subsequently suffered a bad medical outcome linked to an omitted gene. Would failure to follow the ACMG Recommendation be evidence—maybe even conclusive evidence—of malpractice?

A July 29, 2014 decision by the U.S. Court of Appeals for the Eleventh Circuit (which covers Georgia, Florida, and Alabama), Adams v. Laboratory Corporation of America, may shed some light on the question, at least indirectly. That case held that a court should not defer to medical organizations’ standards for what constitutes competent evidence of professional negligence.

Christina Adams had five pap smears during 2006-08. Her doctor sent them to LabCorp for analysis, and was told that there were no abnormal cells. LabCorp’s protocol was for test slides to be examined under a microscope by a trained technician called a cytotechnologist, who would send the slides to a pathologist only if abnormal cells were detected. Ms. Adams experienced vaginal bleeding in 2009 and was then diagnosed with advanced cervical cancer. She and her husband sued LabCorp for negligence, alleging that the cytotechnologists who reviewed her slides breached the standard of care that applies to the practice of cytotechnology.

To prove the breach, the Adamses planned to rely on the testimony of Dr. Dorothy Rosenthal, a professor of pathology at Johns Hopkins with 40 years of experience. The federal district (trial) court in Georgia, having reviewed Dr. Rosenthal’s pre-trial deposition, ruled that her testimony would not be admissible at trial. With her testimony excluded, the Adamses had no expert opinion to support their allegations of negligence, so the district court granted LabCorp’s motion for summary judgment (that is, it dismissed the Adamses’ case before trial on the basis that they could no longer possibly win). The Eleventh Circuit reversed, ruling that Dr. Rosenthal’s testimony is admissible, vacating (undoing) the grant of summary judgment, and remanding (sending back) the case to the district for further proceedings, which will presumably include a trial. All three judges on the Eleventh Circuit panel concurred in the judgment, with one of the three writing a separate opinion.

The general requirements for expert testimony in the federal courts are set out in Federal Rule of Evidence 702, which is in turn derived from three Supreme Court cases decided in the 1990s—the so-called Daubert trilogy, named after the first of the three, Daubert v. Merrell Dow Pharmaceuticals. As summarized by the Eleventh Circuit, Rule 702 and the Daubert trilogy require three things: the expert must be qualified to testify competently about the matter at hand, the expert’s methodology must be reliable, and the testimony must be helpful to the jury. The district court rejected Dr. Rosenthal on the basis that her methodology was unreliable in the extreme: “an ipse dixit assessment that is devoid of any methodology that would allow another expert to challenge it in any objective sense.”

So what did Dr. Rosenthal do that was so unreliable? She went to LabCorp’s Atlanta facility and examined all of Ms. Adams’ slides under the same model microscope that the LabCorp cytotechs had used. She then applied the long-standard Bethesda System (which she had helped to develop, and which LabCorp also used) to classify the cells, and pointed to abnormal cells that, in her opinion, should have been identified by LabCorp. As the district court emphasized, she knew in advance that the patient had cervical cancer, she didn’t mix Ms. Adams’ slides with those of other patients (in other words, she didn’t do a “blinded review”), and she testified in her deposition that she had a “philosophical bent” in favor of patients who are later diagnosed with cancer.

The district court gave four grounds for its decision, and the Eleventh Circuit found all four to be “manifestly erroneous.” I want to focus on just one, so I’ll discuss the other three very briefly. First, the Eleventh Circuit held, Dr. Rosenthal’s opinion was not a mere ipse dixit (she said it). On the contrary, she “formed her opinion by using reliable tools, applying an established body of medical knowledge, and drawing on her extensive experience in the field.” In fact, the court emphasized, LabCorp’s own expert used exactly the same non-blinded approach in developing her opinion that the LabCorp cytotechs were not negligent.

Second, the Eleventh Circuit rejected the district court’s reliance on Dr. Rosenthal’s admitted pro-patient “philosophical bent.” Any such bias—which could be probed during cross-examination—would be a factor for the jury to consider in judging her credibility, but could not provide a basis for a court to exclude her testimony entirely. And third, it was not necessary for Dr. Rosenthal to have conducted her evaluation of the slides under “similar conditions and surrounding circumstances” as those under which the LabCorp cytotechs worked. That, the court of appeals held, is simply not required by the law. In fact it couldn’t be, because such a rule “would bar all expert medical testimony unless the expert has somehow recreated the same conditions that the defendant was under.”

The Eleventh Circuit devoted most of its attention to a fourth issue: the district court’s criticism of Dr. Rosenthal for failing to follow “litigation guidelines” promulgated by the College of American Pathologists (CAP) and the American Society of Cytopathology (ASC). Under those guidelines, pathologists offering evidence in court should use only “blind review” in which “the plaintiff’s slides are mixed in with other normal and abnormal slides,” with the process repeated by multiple reviewers. Dr. Rosenthal didn’t do that, looking only at the plaintiff’s slides in isolation.

The Eleventh Circuit rejected the district court’s reliance on the CAP/ASC guidelines as the “generally accepted standards in the area of pathology or cytotechnology.” On the contrary, it characterized the guidelines as “policy proposals to limit how courts can find the members of the organizations liable for professional negligence when they are sued.” The organizations have thus “moved away from disinterested scientific inquiry and into litigation policy to serve their members own interests.” The law does “not permit delegating to industry groups the gatekeeping duties of the courts.” If such industry efforts succeeded, “why couldn’t pharmaceutical companies adopt guidelines setting high standards of proof for establishing that a plaintiff’s injury was caused by a given drug and justify doing so based on their experience with the complex nature of pharmacology?” It is clear that the court viewed the CAP and ASC as self-interested trade associations protecting their members’ economic interests, and gave no deference to their status as independent professional organizations.

What does this tell us about the potential role of the ACMG Recommendations in future litigation? The AMCG Recommendations and the CAP/ASC guidelines share one significant feature: they are both recommendations for professional conduct promulgated by learned medical societies. But there are also some critical differences. The ACMG Recommendations are intended to guide the conduct of clinicians and laboratories in the day-to-day practice of medicine. They reflect no apparent intent to influence litigation. To the extent they were to come into play in litigation, they would be neutral in the sense of potentially useful to either party. A plaintiff might charge a defendant physician with negligence for failing to test for one the 56 genes. But a defendant physician might also defend a claim by showing that he or she did follow the Recommendations.

The CAP/ASC guidelines, in stark contrast, are written explicitly to shape litigation, to tell courts what methods plaintiffs’ experts must use. In the Eleventh Circuit’s view, they are transparently pro-defendant, a self-serving effort to create a very high methodological bar for plaintiffs’ experts. A plaintiff’s expert might sometimes benefit from the guidelines by showing that he or she met the standard, but the more frequent outcome is likely to be what happened in Adams, with a defendant attacking a plaintiff’s expert for failing to follow the CAP/ASC prescription.

Adams can reasonably be read as supporting several propositions. First, prior cases have held that a medical organization’s practice standards can be relevant in determining the standard of care, and Adams doesn’t undermine that basic principle. However, practice standards are not binding on the courts; at most, they are evidence of what the standard of care should be. Next, courts are likely to follow the lead of Adams and ask whether the standard is a neutral, scientifically-based effort to improve the quality of daily practice (as the ACMG standards clearly are, regardless of whether you agree with them), or an effort to protect the organization’s members from litigation. In the latter case, the result is likely to be just what happened in Adams: rejection, accompanied by scathing rhetoric about a trade association’s attempt to usurp the authority of the courts. In other words, the effort will backfire.