For more than 80 years, the U.S. courts have used the Frye rule to make determinations about the admissibility of expert witness testimony in the area of science or medicine. The Frye rule (Frye v.
In 1993, a critical ruling was handed down that was destined to redefine the landscape of medicolegal testimony of experts (Daubert v. Merrell Dow Pharmaceuticals, Inc 509 U.S. 579 [1993]). As a result of this watershed case, the Supreme Court effectively made trial judges the gatekeepers of scientific expert testimony on the basis of four criteria:
- Whether the theory used by the expert can be and has been tested.
- Whether the theory or technique has been subjected to peer review.
- The known or potential rate of error of the method used.
- The degree of the methods or conclusion's acceptance within the relevant scientific community.
More recent interpretations of Daubert hold that at least a "significant minority" of the scientific community in which the theory resides must accept the theory or method. The first criterion, testability, is often considered the most important. While the courts (i.e., the judges) will ultimately decide how best to consider testability, it is an important distinction. As for Daubert, it is not likely that testimony will be rejected on the basis of one criterion alone. From a logical standpoint, however, a good scientific investigation is likely to satisfy the first three components of Daubert, while the fourth is likely to be debated in many instances, regardless. Nevertheless, if one cannot pass the Daubert threshold, he or she might be limited in what he or she can say at trial, or might not be permitted to testify at all. Thus, it is important to be prepared for such challenges by becoming acquainted with the studies that can potentially save the day. Conversely, an expert might also be called upon to assist the attorney calling him or her to help prepare a Daubert challenge for the opposing experts. Familiarity of the literature (or lack thereof) will once again be critical. One should bear in mind, of course, that judges will approach this issue variably, since the four criteria are only vaguely defined. For example, what level of proof, accuracy or reliability is dispositive? How many published studies do we need to have? What shall we do when the literature is split on an issue? Can animal research be used in cases involving human disease?
In another case, the 9th Circuit Court of Appeals added a fifth element to Daubert: whether the scientific theory or method existed prior to the present cause of action. This question asks whether the theory or method might have been developed specifically for the purpose of testifying in this particular case. Daubert challenges are usually raised in motions in limine. (See Croft AC: Motions in limine: how they can work for or against you. Dynamic Chiropractic 2001; 19(20):30-38.)
Joiner and Kumho Tire
Two more recent and important cases, G.E. v. Joiner (General Electric v. Joiner 522 U.S. 136, 139, 139 L. Ed. 2d 508, 118 S. Ct. 512, 1997 U.S. LEXIS 7503) and Kumho Tire Co. v. Carmichael. (Kumho Tire Co. v. Carmichael 143 L. Ed. 2d 238, 119 S. Ct. 1167, 1999 U.S. LEXIS 2189, 67 U.S. L. W. 4179 [1999]), provide interesting background and further define the current standard. The rulings should be understood by anyone offering expert testimony in a court of law, since opposing counsel may challenge the proffered testimony, either by motion in limine or in a 402 hearing or voir dire outside the presence of the jury. This is the essence of the Daubert challenge. (Note that several states continue to use Frye, but the basis of the challenges is essentially the same.)
In GE v. Joiner, Robert Joiner worked for General Electric for a number of years and claimed that he developed lung cancer from industrial exposure to polychlorinated biphenyls (PCB), despite his history of cigarette smoking and a family history of lung cancer. He hired two experts. One was a toxicologist and the other a preventive medicine specialist. They relied on published studies to argue that his lung cancer was, in fact, related to PCB exposure. However, the trial judge threw out their testimony on the basis of Daubert, declaring that it was based only on a small number of animal studies. The animals, moreover, had been exposed to much higher concentrations of PCBs than Joiner had. The court then entered a summary judgment in the defendant's favor. The plaintiffs appealed to the 11th Circuit Court of Appeals.
The 11th Circuit Court of Appeals examined the case and the evidence offered by both specialists. It also considered the fact that both had personally interviewed Joiner and his family and co-workers, reviewed medical records, and relied upon scientific literature in forming their opinions. The appellate court criticized the lower court's decision to exclude the testimony on the basis of the small number of studies and the fact that they were animal studies, as being irrelevant and improper. It reversed the lower court. The case then went to the U.S. Supreme Court. The Supreme Court subsequently reversed the appellate court, ruling that the proper standard for the appellate court to apply in such a case is "abuse of discretion." (Note that in any appeals court, the courts do not retry the case or review evidence for its probative value, and only reverse a lower court for important errors in process.) The Supreme Court stated that, in its view, the judge did not abuse his discretion in excluding Joiner's experts' testimony.
In the Kumho Tire v. Carmichael case, one person was killed and the other passengers injured in a rollover crash. At the heart of the issue of this products liability case was the possibility of a tire defect. Plaintiffs called a tire failure analyst, Dennis Carlson Jr., who intended to testify that the tire had a defect. Kumho Tire moved to exclude his testimony and the court applied the Daubert theory, ruling that the four criteria of Daubert did not support the reliability of the expert's methodology. Once again, the 11th Circuit Court of Appeals heard the case and overturned the lower court, arguing that the judge had erred in applying Daubert to the case, since Daubert was to be used only in the context of scientific testimony. Once again, the case then went to the Supreme Court, which again reversed the appellate court's ruling. The Supreme Court wrote that Daubert could be applied to engineers and other experts who are not scientists.
These two cases are important because they affirm the lower court's gatekeeper function in the contest of Daubert and more broadly apply Daubert's scope to nonscientific expert testimony.
How Does Daubert Affect Expert Witnesses?
So, what kinds of testimony are likely to be scrutinized under a Daubert test? First and foremost would be medical testimony: specifically, opinions about critical issues such as injury causation. Medical technology that is used to support opinions about causation, injury extent or severity, or long-term prognosis might also be the subject of a Daubert challenge, if it appears to be new, cutting edge, or obscure. Any so-called "black box technology" is likely to be challenged, including computer software; the "black box," in this case, being the actual computer code. And indeed, at least in the early stages of development, many diagnostic technologies might not make a favorable showing under a vigorous Daubert challenge if it has not yet been subjected to rigorous scientific testing. It has not been uncommon in the past for medical devices to go from the inventor's workbench directly into the clinic without being subject to validation testing. In that event, perhaps none of the four components of Daubert could be defended. In other instances, the technology may be widely accepted and well-developed for some uses, but not for others.
A not-uncommon example of this would be off-label use of medication, in which the drug is used to treat a condition it had not been tested and approved for. Surface EMG might be another example. While the technology has been used for many years for biofeedback training, exercise physiology experiments, incontinence training, and other applications, its use as a spinal diagnostic procedure is a recent development. Many claims have been advanced for its ability to distinguish various spinal pathologies or conditions, yet this is an application for which it has not been rigorously examined in scientific investigations. In recent years, though, sEMG has been used in a number of research studies, which does provide a foundation for some clinical applications. Clinicians should be thoroughly familiar with the breadth and scope of the existing research in order to weather a Daubert challenge.
Diagnostic ultrasound provides another example. For many years, this technology has been widely used in obstetrics, internal medicine, and in the evaluation of extremity joints. But its application in spinal pathology has been slow to catch on. Would diagnostic ultrasound withstand a Daubert challenge when used to image a shoulder? Undoubtedly. Would it pass when applied to the spine? That is less certain.
Another field that frequently is called to provide expert testimony is accident reconstruction - specifically, auto crash reconstruction (ACR). ACRs may offer testimony only on crash metrics (delta V or speed change, acceleration, etc.), or they may also opine as to the actual likelihood that a claimant would have been injured, or even as to the severity of these potential injuries. In many jurisdictions, these latter opinions may be disallowed by the trial judge because the ACR is not qualified to make a medical diagnosis, but this situation is often remedied by calling in a biomechanist, who then relies on the ACR's analysis to determine the likelihood of injury and other issues. Bear in mind, however, that this testimony is often objected to; again, because the testimony as to whether an injury occurred is really the logical equivalent of a medical diagnosis, which is also not within the purview of a biomechanist. The biomechanist's counterargument is that he or she is not making a diagnosis per se, but merely speculating as to the likelihood of an injury, given the known forces and loading conditions the plaintiff was subjected to. The outcomes of these objections have been quite variable from case to case, so from the standpoint of opposing counsel, they are always worth raising. Much case law regarding biomechanical testimony exists,1 but is beyond the ambit of this paper. Nevertheless, Kumho guarantees that Daubert will apply to both ACR and biomechanist testimony.
Another aspect of ACR concerns the basis for the reconstruction itself. Because ACR is based on simple vehicle dynamics, which has its basis in Newtonian physics, it has not been subjected to much scrutiny over the years. Some special applications of ACR have been validated using crash tests and other experiments, but the general field of ACR has largely enjoyed a relatively unfettered, albeit untested, acceptance until recent years. Bartlett, et al.,2 created something of a stir for ACRs with their report. They discovered that in some applications, the commonly used methods of ACR are fairly imprecise, providing an opening for the Daubert challenge.
Thus, the mechanics of the reconstruction are often the most vulnerable point of attack for the ACR. And, if a biomechanist bases his or her opinions on the ACR's numbers, those opinions will be equivalently diminished in credibility and reliability. Beyond this, there are a number of software applications for use in ACR. Almost all of them are based on the original CALSPAN CRASH algorithm, which is problematic because the program assumes a common velocity crash condition, and does not account for restitution in the collision. These concepts are described in detail elsewhere.3 Recent studies have compared ACRs performed with WinSmash - the software program used by the National Highway and Traffic Safety Administration (NHTSA) for development of the National Automotive Sampling System (NASS) database - and the car's on-board event data recorder (EDR), which measures the acceleration of a crash with fairly good accuracy. It was discovered that WinSmash underestimated the delta V in these crashes 77 percent of the time, with an average underestimation of 20 percent.4 These, of course, are not trivial variations and suggest that in many years of NASS data, which have served as the substrate for literally thousands of automotive safety research projects over the years, the reported delta V has been subject to a not-insignificant systematic underestimation. Here again is the point of attack for a Daubert challenge.
In summary, Daubert applies to any expert wishing to provide the court with expert testimony in these cases, whether the testimony relies upon the results of a medical test or other device or whether it is based on scientific theory. Experts should be prepared for the Daubert challenge. The four (or five) elements of Daubert will require that the expert be familiar with the state of the art of scientific, biomedical, or other relevant literature from whence his or her testimony comes. A similar knowledge is required in developing strategies for challenging opposing experts through motions in limine - a process in which experts can play an integral role.
References
- Tessier RM. To hire or not hire a biomechanical engineer. Advocate 2004:40-7.
- Bartlett W, Wright W, Masory O, et al. Evaluating the uncertainty in various measurement tasks common to accident reconstruction. In: Accident Reconstruction 2002. Warrendale, SAE Technical Paper Series 2002-01-0546, 2002:57-68.
- Croft A. Biomechanics. In: Foreman S, Croft A (eds). Whiplash: The Cervical Acceleration/Deceleration Syndrome. 3rd ed. Baltimore: Lippincott Willliams and Wilkins, 2002:1-129.
- Gabler HC, Hampton CE, Hinch J. Crash severity: a comparison of event data recorder measurements with accident reconstruction estimates. SAE Technical Paper Series 2004-01-1194, March 2004.
Arthur Croft, DC, MS, MPH, FACO
Director, Spine Research Institute of San Diego
San Diego, California
Click here for previous articles by Arthur Croft, DC, MS, MPH, FACO.