How Reasonable is a ‘Reasonable Degree of Medical Certainty?'
The Use of Treating Physicians as Causation Experts in Product
Russell T. Burke, JD1
James W. Lynch, Jr., MD, FACP2
“The expert physician is seductively drawn up to make statements, then
driven to retract or qualify them and then pressed to perjury, or so near it
that it will be difficult to draw the line. . .”
~JAMA editorial, 1892
"[T]he hermit clinician can usually testify to anything if he holds an M.D.
and is willing to mumble some magic words about 'reasonable medical
~Peter W. Huber, Galileo's Revenge: Junk Science in the Courtroom,177
“When I use a word,” Humpty Dumpty said in rather a scornful tone, “it
means just what I choose it to mean--neither more or less.”
“The question is,” said Alice, “whether you can make words mean different
“The question is,” said Humpty Dumpty, “which is to be master--that's all.”
~Lewis Carroll, Through the Looking Glass, ch. 6. (1872)
Physicians routinely diagnose and treat patients for illnesses, using a
differential diagnosis to identify the illness to develop a treatment plan. They
often then find themselves in the middle of a legal battle and are asked to render
an opinion not just on the identity of the illness, but also on the cause of the
illness. Their testimony is based on an evidentiary standard of ‘reasonable
medical certainty,’3 a standard that is not used in the medical world (and rarely
defined in the legal world). In such situations, the treating physician is essentially
being asked to go beyond his clinical expertise and make an educated guess on
Member, Nexsen Pruet LLC, Columbia, SC.
Professor and Fellowship Program Director and Chief, Hematology/Oncology Veteran’s
Administration Medical Center, University of Florida College of Medicine, Gainesville, FL .
The Supreme Court has recognized that “[w]ithin the medical discipline, the traditional standard
for ‘factfinding’ is a ‘reasonable medical certainty’ Addington v. Texas, 441 U.S. 418, 439 (1979).
Indeed, there are over 8,000 reported state and federal court decisions that recite testimony by
physicians and other experts to the evidentiary standard of ‘reasonable medical certainty’ (based
on a WESTLAW Allstates and Allfeds search for “reasonable degree of medical certainty’ or
“reasonable medical certainty”). This does not include the same term for other disciplines, e.g.,
“reasonable engineering certainty,” etc.
what caused the illness. He will usually resort to his clinical practice of utilizing
the methodology of a differential diagnosis – identifying and eliminating potential
That puts into place the perfect storm in the medico-legal world: treating
physicians testifying on behalf of their patients using a familiar methodology for
an unfamiliar purpose, to an evidentiary standard of ‘certainty’ that holds no
medical meaning. Does Daubert4 come to the rescue in these situations? Not
necessarily. Appellate courts have not consistently analyzed the parameters
what makes for a reliable ‘differential diagnosis’ to establish causation, and thus
the admissibility of opinion testimony is entirely up to district court judges under
the Joiner 5 abuse of discretion standard. Trial lawyers are thus left with the
situation in which the district court may be more of a doorman than a gatekeeper,
and the treating physician gets to decide the admissibility of his own opinions
based on the ‘reasonable certainty’ shibboleth.6
I. THE REASONABLE MEDICAL CERTAINTY STANDARD
The concept of ‘reasonable certainty’ cuts across expert testimony from
many disciplines. Many courts, if not most, require some form of these magic
words to admit opinion testimony on issues including the cause of injury
(including causation in complex toxic tort cases), extent and permanency of
injury, standard of care, and proof of future damages, among others.7 Generally
speaking, since the liberalization of evidence rules and the adoption of the
Federal Rules of Evidence in 1975, most jurisdictions do not require the exact
formulaic expression, but rather admit testimony that implies reasonable medical
certainty.8 Underlying such testimony, however, is the unspoken belief that
‘reasonable medical certainty’ actually has some meaning, and further that it has
Daubert v. Merrell Dow Pharmaceuticals, 509 U.S. 579 (1993).
General Electric v. Joiner, 522 U.S. 136 (1997).
See James E. Hullverson, Jr., Reasonable Degree of Medical Certainty: A Tort et a Travers, 31
St. Louis U. L. J. 577, 589 (1987).
See 1 Paul C. Giannelli & Edward J. Imwinkelried, Scientific Evidence 155 (2d ed. 1993) ("Some
jurisdictions require an expert to express an opinion in terms of reasonable scientific probability or
certainty."); Edward J. Imwinkelried, Evidentiary Foundations 255 (3d ed. 1995) ("Some
jurisdictions insist that the expert vouch that his or her opinion is 'reasonably certain.' Other
jurisdictions demand a 'reasonably probable' opinion. The witness must be willing to testify that
he or she has formed the opinion to a reasonable medical or scientific certainty or probability.");
Edward R. Stein, The Direct Examination of the Expert Witness, in Faust F. Rossi, Expert
Witnesses 193, 219 (1991) ("[The degree of certainty] can relate to the admissibility of the
opinion. Some jurisdictions, and some judges, require that all questions calling for an opinion
must call for it to a certain standard, usually 'a reasonable degree of medical certainty."'). See
generally Hullverson, supra note 6, at 578 (discussing the "semantic nuisance" of defining the
phrase in various jurisdictions); Michael M. Martin, The Uncertain Rule of Certainty: An Analysis
and Proposal for a Federal Evidence Rule, 20 Wayne L.Rev. 781, 782 (1974) (analyzing "the rule
of certainty ... in its various forms").
See Jeff L. Lewin, The Genesis and Evolution of Legal Uncertainty About "Reasonable Medical
Certainty," 57 Maryland L. Rev. 380, 397-98 (1998).
some gravitas – that the testifying expert is putting his professional reputation on
the line when he utters the phrase. Unfortunately, neither premise is true.
‘Reasonable medical certainty’ has no consistent meaning, although it
implies some objective and quantifiable standard.9 And because it has no
consistent meaning, it is an evidentiary hurdle only to those testifying physicians
who are either new to the legal world or who take their oath very seriously. 10 The
phrase is not only misunderstood by physicians, it is often not even understood
by the lawyers who ask for opinions based on this ‘standard,’ nor by judges or
appellate courts who insist upon it. And yet issues of causation and future
medical conditions in product liability cases – as well as medical malpractice
cases, workers’ compensation cases and garden-variety negligence cases - are
won or lost based on the willingness of a treating physician to utter an opinion
based on an evidentiary standard that sounds very important but is devoid
A. THE OBSCURE ETIOLOGY AND MEANING OF ‘REASONABLE
Since most trial lawyers instinctively phrase opinion questions in terms of
‘medical’ or ‘scientific’ certainty, one would expect that we would at least know
how the phrase originated. Like the causes of many illnesses, however, we do
not know from whence it arises. One commentator has maintained that the
origins are shrouded in the mists of time: “The origins of the phrase are obscure,
its evolution accidental . . .”11 Another commentator, however, has sifted through
the sands of early cases and believes that the phrase originated in Chicago
between 1915 and 1930.
Although its precise origins cannot be determined, the phrase
have been generated by the efforts of Illinois attorneys to
accommodate two inconsistent rules of evidence that were
adopted almost simultaneously by the Illinois Supreme Court in
the early years of this century: the "reasonable-certainty rule,"
which prohibited experts from expressing speculative opinions
about damages, and the "ultimate-issue rule," which prohibited
The Nebraska Supreme Court has aptly observed that the term "reasonable" is ambiguous and
that "an attempt to give specific meaning to the term 'reasonable' is [like] trying to count what is
not a number, and measure what is not space." Altshuler v. Coburn, 57 N.W. 836, 838 (Neb.
1894). The Court ultimately opted for a standard of "probability" rather than "reasonable medical
certainty." Lane v. State Farm Mut. Auto. Ins. Co., 308 N.W.2d 503, 512-13 (Neb. 1981).
There are movements afoot to reign in testifying physicians on the premise that doctors who
provide opinions under oath are engaged in practicing medicine and therefore subject to the
discipline of medical boards for their testimony. See, e.g., the Coalition and Center for Ethical
Medical Testimony website at http://www.ccemt.org/displayboard.cfm.
Hullverson, supra n. 6, at 578; cf. Martin, supra note 7 at 786-808.
experts from "invading the province of the jury" by expressing
definitive opinions on disputed issues.12
Once the phrase was in use in Illinois, it appeared in a popular trial
manual (Goldstein’s Trial Technique), and thereafter the phrase spread gradually
to other jurisdictions. By 1960 at least twenty-two other states had adopted the
evidentiary standard of ‘reasonable medical certainty.’ By 1970, the phrase had
made its way into the published decisions of forty-eight states.13 Thus, by the
start of the boom in products liability law, the phrase ‘reasonable medical
certainty’ had received not only general acceptance but also ‘talismanic
Given the breadth of use of the phrase, one would expect that its meaning
would have been long-ago established in the case law. But the early cases do
not define the term and with the rapid adoption of the phrase by the courts in the
1960’s, perhaps it was assumed that its meaning was self-evident. Given the
uncertainty in the origin of the phrase, we cannot resort to ‘original intent’ to
decipher its meaning. We are thus left with the situation that the phrase means
different things in different jurisdictions, and likely different things to different
people: “[c]ase law from around the country demonstrates that appellate courts
have not been able to come to a consensus on whether "reasonable medical
certainty" means "probably true," "more probable than not," "beyond a
reasonable doubt," or some combination of those concepts.”15 Further, the
phrase might mean something different depending on whether causation or proof
of future damages is the issue.16 Generally speaking, ‘reasonable medical
Lewin, supra note 8 at 407.
Lewin, at 435-457.
Lewin, at 457; several cases refer to ‘reasonable medical certainty’ as a talisman (see, e.g.,
Holbrook v. Lykes Bros. Steamship Co., 80 F.3d 777, 785 (3 Cir. 1996); Loura v. Adler, 664
N.E.2d 1002, 1005 (Ohio Ct. App. 1995) (per curiam); Hashimoto v. Marathon Pipe Line Co., 767
P.2d 158, 164 n. 9 (Wyo. 1989); see also Nancy Levit, Ethereal Torts, 61 Geo. Wash. L. Rev.
136, 168 (1992)).
Glenn E. Bradford, Dissecting Missouri’s Requirement of “Reasonable Medical Certainty,” 57
Missouri Bar J. No. 3 (May-June 2001) (see www.mobar.org/journal/2001/mayjun/index.htm); see
generally Lewin, supra note 8 at 385, 375-471 and n.17. One pre-Daubert commentator
suggests that it relates to substantive evidence rather than admissibility standards (thus
demonstrating that the confluence of issues that gave rise to the phrase in the first instance still
lives on): Testimony to a “reasonable degree of certainty” means that the opinion must be in
“conformity with a generally accepted explanatory theory used in the field to derive the opinions
expressed by the expert. By requiring that an expert’s opinion be derived according to a theory
substantially accepted in the particular field, the court ensures a degree of reliability that would
not be present if the opinion were to be held solely by that expert.” Miller, Cross Examination of
Expert Witnesses: Dispelling the Aura of Reliability, 42 U.Miami L.Rev. 1073, 1097-98 (1988).
“In sum, in the context of the ‘reasonable certainty’ rule with respect to future damages, the
courts employed quite divergent interpretations of the phrase ‘reasonable medical certainty.’
Some courts equated ‘reasonable certainty’ with ‘reasonable medical certainty,’ while other
viewed the phrases as distinct. Some courts held that the standard was satisfied by testimony
about probable future consequences, while others interpreted the ‘rule of certainty’ as requiring
more than a probability of future injury. Some courts approved of jury instructions employing the
phrase ‘reasonable medical certainty,’ while others did not.” Lewin, at 477.
certainty’ deals with probabilities rather than possibilities, but that depends on the
jurisdiction in which the case originates.
B. WHAT IS REASONABLY CERTAIN TO A PHYSICIAN?
Practicing physicians do not routinely think in terms of ‘reasonable medical
certainty.’17 It is not a useful medical phrase. For example, when considering a
possible treatment, if the degree of risk is slight, the cost minimal, and the
possible benefit significant, most physicians would readily concur that the
proposed treatment is indicated with a ‘reasonable degree of medical certainty.’
This type of ‘certainty’ is well below the legal standard of ‘more probable than
not.’ On the other hand, if the proposed procedure is risky and expensive, then
physicians would require a much higher degree of certainty that the benefits
outweigh the risks of the procedure. The level of certainty required in such an
instance is higher than the criminal standard of ‘beyond a reasonable doubt.’
Given this sliding scale of certainty in the medical world, most physicians,
especially those new to the game, have no idea what ‘medical certainty’ means
in the legal world.18 This creates a situation in which lawsuits are one or lost
based on the legal sophistication of physicians,19 as opposed to the facts of the
C. THE UNCERTAINTY OF REASONABLE MEDICAL CERTAINTY
If the concept of certainty is designed to root out merely speculative
testimony that lacks evidentiary foundation,20 the requirement of testimony to a
Bradford, supra note 15 (“It appears that practicing physicians do not routinely employ the
concept of reasonable medical certainty in everyday practice. In discussing this issue, one writer
has observed that physicians “are not accustomed to thinking in terms of certainty.” The
concepts of certainty and predictability seem to arise only when medical matters reach the
litigation stage, where decisions must be made on matters related to future events and
conditions” (quoting Earl F. Rose, A Pragmatic Approach to Medical Evidence and the Lawsuit, 5
U. Tol.L.Rev. 237, 252 (1974)).
Jonas R. Rappeport, Reasonable Medical Certainty, 13 Bull. Am. Acad. Psychiatry & L. 5, 8
(1985) (“Doctors testify to it daily, but they do not know what it means.” It is not a medical term at
all: “As men of law they are tying to force us aliens to speak their language.”); Mark D. Howard,
Proving Causation with Expert Opinion: How Much Certainty is Enough? 74 Ill. B.J. 580, 584
(1986) (“Most experts, other than professional witnesses, are unfamiliar with the ‘reasonable
certainty’ language used in court.”)
One court noted that "[p]hysicians differ in the degree of caution or lack of caution with which
they phrase their opinions, and one man's 'possibility' may be equivalent to another's
'probability.'" This creates an untenable situation: ". . . it is a shame that claims of equal merit
should stand or fall on just how the medical witnesses choose to portray their conclusion . . . ."
Inland Steel Co. v. Johnson, 439 S.W.2d 562, 564, n. 1 (Ky. 1969).
In Dallas v. Burlington Northern, Inc., 689 P.2d 273, 277 (Mont. 1984), the Montana Supreme
Court explained "reasonable medical certainty" as follows: “Although we still formally adhere to a
"reasonable medical certainty" standard, the term is not well understood by the medical
profession. Little, if anything, is "certain" in science. The term was adopted in law to assure that
testimony received by the fact finder was not merely conjectural but rather was sufficiently
probative to be reliable. We are striving for, what in fact, is a probability rather than a possibility.
‘reasonable medical certainty’ has not fulfilled its mission. As indicated, ‘medical
certainty’ can mean anything from ‘possibility’ to ‘more probable than not’ to
‘beyond a reasonable doubt.’ The fundamental problem is that certainty itself is
not an objective standard, and the legal evidentiary standard purports to be
Not only is ‘medical certainty’ uncertain, it is often tied to causation
opinions by treating physicians who have little training or knowledge to opine
about the etiology of a particular disorder. This is nowhere more true than the
use of ‘differential diagnosis’ by treating physicians who use the technique to
diagnose a disorder based on reported symptoms and clinical studies, but then
go on to opine about the external causes of the disorder. It is one thing for a
physician to diagnose a symptom, it is quite another to determine the actual
cause of that symptom. Most treating physicians do not use differential diagnosis
to determine the external causes of a disorder, because they generally do not
have to know what is causing a disorder in order to treat it. Nevertheless, many
courts admit such testimony, noting that the lack of support for such testimony
‘goes to the weight’ of the evidence, or that such testimony will be subject to
‘vigorous cross examination.’21 The result is that the courts allow physicians to
render opinions based on an undefined and misunderstood evidentiary standard
using methodologies that they themselves neither understand nor use in their
practice of medicine.
II. DIFFERENTIAL DIAGNOSIS IN MEDICINE AND IN THE LAW
As noted above, doctors routinely employ differential diagnosis in the
course of their everyday clinical practice. That is nothing new. Problems arise,
however, when doctors purport to utilize the differential diagnosis methodology in
areas beyond their expertise or for purposes in which the technique is not suited.
This is the difference, which is lost on many courts, between a clinical differential
diagnosis and a “causation assessment.”22 In such cases, a scientific
methodology that is perfectly suited for one purpose creates the veneer of
scientific validity – sufficient perhaps to satisfy Daubert standards – for a quite
different purpose. This is the result when treating physicians purport to utilize
differential diagnosis to ‘rule out’ potential causes of an illness without 1) an
Our evidentiary standards are satisfied if medical testimony is based upon an opinion that it is
"more likely than not."”
Daubert v. Merrell Dow Pharmaceuticals, 509 U.S. 579, 595-96(1993). “Respondent expresses
apprehension that abandonment of "general acceptance" as the exclusive requirement for
admission will result in a "free-for-all" in which befuddled juries are confounded by absurd and
irrational pseudoscientific assertions. In this regard, respondent seems to us to be overly
pessimistic about the capabilities of the jury and of the adversary system generally. Vigorous
cross-examination, presentation of contrary evidence, and careful instruction on the burden of
proof are the traditional and appropriate means of attacking shaky but admissible evidence.”
Ronald E. Gots, M.D., Ph.D., Differential Diagnosis versus Causation Assessment: Why they
are Separate Methodologies and how they Relate to Daubert, International Center for Toxicology
and Medicine. http://phys4.harvard.edu/~wilson/soundscience/differental%20diagnosis.html.
adequate factual foundation to do so, or 2) before science has been able to ‘rule
in’ all (or a substantial majority of) the purported causes in the first instance.
Despite the admonition in Daubert that “scientific validity for one purpose is not
necessarily scientific validity for other, unrelated purposes,”23 many courts ignore
that dicta when they allow treating physicians testify as to causation in toxic tort
A. CLINICAL USE OF DIFFERENTIAL DIAGNOSIS
The process for a clinical differential diagnosis is as follows. In a
differential diagnosis of a patient who is ill, the treating physician lists the
potential causes for a symptoms or set of symptoms. The list is then prioritized
based on judgment that includes epidemiology, a history and physical of the
patient, and diagnostic tests. The doctor then develops a "working diagnosis"
which is either confirmed by appropriate diagnostic testing, or is ruled out. If it is
ruled out, then a another diagnosis should be tested, and so on until an
explanation is found. This process works because it is in real time and assumes
(albeit erroneously) that 1) we understand all potential causes of a malady, 2) our
diagnostic tests are sufficiently accurate to adequately support decision-making
and 3) patients’ maladies are caused by a single disease. The reality is that
there are many as yet unidentified causes of illness. Further, diagnostic tests are
of variable sensitivity and specificity, and this needs to be accounted for in
diagnosing illness. Finally, many patients have multiple issues - a "conspiracy"
of medical problems - that create a confusing mix of symptoms. In other words,
particular symptoms may not be caused by medications, a stroke, low sodium in
the blood, or a urinary tract infection, but rather by the combination of all of them.
Consider, for example, a patient who presents at the ER for an evaluation
of chest pain. He describes pain in the middle of his chest, with tightness, that
wakes him up at night and is relieved by Tums. This history suggests acid reflux,
but the list of causes also includes ischemic heart disease. Both are common,
but the latter is clearly more life threatening, so the evaluation will seek to "rule
out" a cardiac cause. Blood tests are therefore performed to look for heart
damage, and risk factors for heart disease will be used to determine whether the
patient should have a further cardiac work-up or not. The patient may be started
on a medication to suppress acid secretion, and may even have an endoscopy to
look for an inflamed esophagus. This process proceeds as a real time evaluation
that adds clarity to the cause and effect question. However, even if the patient
has a cardiac catheterization and is found to have blocked arteries to the heart,
these clinical findings do not necessarily mean that the chest pain was caused by
heart blockage. Many people have both heart disease and acid reflux. A
definitive diagnosis would require a treadmill test to determine whether the
patient’s heart rate increase causes chest pain similar to the initial report.
Further, an echocardiogram may be performed to determine if the heart has
sufficient blood flow.
509 U.S. at 591.
This standard methodology of clinical differential diagnosis is not,
however, necessarily applicable to determine causation in toxic tort cases, such
as whether exposure to talcum powder at certain levels causes sinusitis or lung
disease. This is because diagnosing the direct physical cause of a symptom is
quite different than determining the precise external cause of disease. The initial
causation question from a medical standpoint is whether the state of knowledge
is sufficient to identify the major causes of a disease or disorder.24 This is
determined (in degree of importance) by randomized, double blind studies (which
would be highly unusual to find in most toxic tort cases), epidemiological studies
(population based studies which show an increased incidence of the disease
associated with exposure to a substance),25 animal toxicological studies based
on experimental models to establish a dose-response relationship, chemical
analogies (predicting effects of exposure based on the established effects of
exposure to similarly structured compounds), and clinical or case reports of
particular patients (anecdotal reports of symptoms after purported exposure to a
substance).26 This ‘general causation’ evidence is at best indirect and therefore
does a poor job of informing as to whether any particular person has disease X
because of exposure Y (“specific causation”). The notion of reasonable medical
certainty in such cases must be carefully analyzed by the quality of the evidence,
and by the possible confounding variables.
Consider the question of the causes of lung cancer. The evidence is that
85 to 90% of patients who develop lung cancer are smokers. The risk of lung
cancer is ten times higher in smokers than in nonsmokers. The experimental
As discussed infra, this is what the case law identifies as “general causation,” meaning that the
general causes of a disease must be known and ‘ruled in’ before the specific causes can be
‘ruled out’ by the general methodology of differential diagnosis.
Note that even when there is an epidemiological study of an association between exposure to a
substance and a disease does not mean that the exposure caused the disease. “There are three
reasons that a positive association may be observed in an epidemiological study: (1) bias; (2)
chance; and (3) real effect.” Joe G. Hollingsworth and Eric G. Lasker, The Case Against
Differential Diagnosis: Daubert, Medical Causation Testimony, and the Scientific Method, Journal
of Health Law 85, 90 (2004, Vol. 37, No. 1). See also Magistrini v. One Hour Martinizing Dry
Cleaning, 180 F. Supp. 2d 584, 591 (D.N.J. 2002); Michael D. Green et al., Reference Guide on
Epidemiology, in Reference Manual on Scientific Evidence 336 (2d. ed. 2000).
Actual case reports, just as the actual case that a treating physician may be opining on, are
most susceptible to post hoc fallacies. As one court noted, “Doctors in their day-to-day practices
stumble upon coincidental occurrences and random events and often follow human nature, which
is to confuse association and causation. They are programmed by human nature and the rigors
and necessities of their clinical practices to conclude that temporal association equals causation,
or at least that it provides an adequate proxy in the chaotic and sometimes inconclusive world of
medicine. This shortcut aids doctors in their clinical practices because their most important
objective day-to-day is to help their patients and “first, do no harm,” as their Hippocratic oath
requires. Consequently, they make a leap of faith . . [This type of] clinical impression is not the
sort of scientific methodology that Daubert demands.” Siharath v. Sandoz Pharm. Corp., 131 F.
Supp. 2d 1347, 1372 (N.D.Ga. 2001), aff’d sub nom. Rider v. Sandoz Pharm. Corp., 295 F.3d
1194, 1199 (11 Cir. 2002) (case reports “may rule out other potential causes of the effect, but
they do not rule out the possibility that the effect manifested in the reported patient’s case is
simply idiosyncratic or the result of unknown confounding factors.”)
evidence in animals is overwhelming that cigarette smoking leads to deposits of
carcinogens, that in turn leads to lung cancer. So, when Mr. Jones smokes three
packs a day for fifty years and develops lung cancer, there is a reasonable
degree of medical certainty that his lung cancer was caused by his cigarette
On the other hand, consider the meaning of medical “certainty” in a case
alleging that a plaintiff’s lung cancer was caused not by smoking but by radon
exposure in his home. Radon is found in high concentrations in uranium mines,
and there is epidemiological evidence that a statistically significant number of
uranium miners develop lung cancer. There appears to be a dose-response
effect, i.e., the risk of cancer increases the longer a miner works in a uranium
mine. Can we therefore conclude with ‘medical certainty’ that radon caused lung
cancer in a specific patient? That is a much harder question than diagnosing the
lung cancer in the first instance. As a further complicating factor, as a matter of
general causation, three recent epidemiological studies came to three different
conclusions on this same causation question.27
In the first study, researchers found “direct evidence of an association
between residential radon and lung cancer risk, a finding predicted using miner
data and consistent with results from animal and in vitro studies.”28 Presumably
these researchers could testify to a that radon can cause lung cancer in non-
smokers, and then a treating physician could ‘rule out’ other causes (e.g.,
smoking) and opine to a reasonable degree of medical certainty that radon did
cause his patient’s lung cancer.
A second recent study came to the opposite conclusion, finding that the
association between residential radon and lung cancer was not generally
statistically significant, and noting that “subjects with presumably lower
uncertainty in the exposure assessment showed a higher risk and also that
dietary antioxidants may act as an effect modifier.”29 In other words, this study
concluded that not only was there no statistical basis to link prolonged low-level
exposure to radon to lung cancer, in fact some people with lower exposure rates
actually had a higher incidence of lung cancer. (Ipso facto, these researchers
would become defense experts.)
Finally, a third study found that there is a “small” excess lung cancer risk
associated with indoor radon exposure after precise adjustment for smoking.30
This supports the contention that “[s]cience is the search for a consensus of rational opinion
among all competent researchers.” Alan Cromer, Uncommon Sense: The Heretical Nature of
D. Krewski, et al., Residential Radon and Risk of Lung Cancer: A Combined Analysis of 7
North American Case-Control Studies, Epidemiology 2005 Mar.16(2):137-145.
F. Bochicchio, et al., Residential Radon Exposure, Diet and Lung Cancer: A Case-control
Study in a Mediterranean region, Int. J. Cancer. 2005 Jan 1.
H. Baysson, et al., Indoor Radon and Lung Cancer in France, Epidemiology, 2004 Nov.
Hence, according to these researchers, there may be a link between household
radon and lung cancer, but it is at present only a tenuous link. Armed with these
studies, a medical history, and ‘differential diagnosis,’ could a treating physician
with ‘reasonable medical certainty’ base a causation opinion simply by ruling out
other known causes of lung cancer? Probably so, at least if the case arises in
the Second, Third, Fourth, Sixth, Eighth, Ninth, Tenth, and D.C. Circuit courts.31
But how could a treating physician ‘rule out’ other, unknown causes of lung
cancer when the epidemiological evidence linking radon to lung cancer is
inconclusive in the first instance? Daubert and Kumho Tire would seem to
require more scientific rigor to such opinions by treating physicians, but
unfortunately there are numerous cases that admit equally shaky opinions.
Treating physicians make judgment calls and treatment decisions every
day based on imperfect data, ruling out possible diagnoses based on experience,
available data, and probabilities, all in an effort to help their patients in real time.
That is what makes medicine an art. Once they are reasonably certain what an
illness is, they know what the treatment options are. Making such critical
judgments for treatment, based on such imperfect data, is the life work of a
treating physician. But the ability to treat an illness does not equate to scientific
knowledge of the external cause of the illness. Making treatment decisions
based on imperfect data should not become license to for physicians to render
causation opinions on even more imperfect data.
There are two natural biases inherent in the practice of medicine that
should be recognized in this regard. First, treating physicians may have a natural
bias to help their patients in litigation,32 sufficient to drive them to link the
exposure to some agent to his patient’s illness with ‘reasonable medical
certainty.’ Second, physicians, like the rest of us, are uncomfortable with
uncertainty. This drives them to come to draw conclusions in situations where
there is scant scientific basis to draw conclusions.33 But physicians must
acknowledge that there are many diseases for which there is either no known
cause, or for which science has not identified all the major causes. Even for
those diseases for which we presume to know the causes, it is likely that the
causes are complex, involving a combination of environmental, genetic, dietary,
and other factors. This uncertainty in knowing the pathogenesis of disease is
uncomfortable, and is reasonably certain to drive unscientific opinions from
treating physicians in the courtroom: “The origin of all science is the desire to
know causes, and the origin of all false science and imposture is the desire to
However, probably not in the Fifth and Eleventh circuits. See discussion of the use of
differential diagnosis in the federal circuits, infra at 23.
Hollingsworth et al., supra note 25, at 97 (2004, Vol. 37, No. 1) (“Differential diagnoses
conducted for tort-litigation purposes can raise unique issues of reliability, though, because they
are generally conducted post hoc and not in the context of medical treatment.”)
“The folly of mistaking a paradox for a discovery, a metaphor for a proof, a torrent of verbiage
for a spring of capital truths, and oneself for an oracle, is inborn in us.” Paul Valery, Introduction
to the Method of Leonardo da Vinci (1895).
accept false causes rather than none; or, which is the same thing, in the
unwillingness to acknowledge our own ignorance.”34
B. LEGAL USE OF DIFFERENTIAL DIAGNOSIS IN TOXIC TORT
For decades lawyers and courts accepted the use of differential diagnosis
as an acceptable methodology to determine causation in tort cases. With the
exponential increase in product liability litigation and the complexity of causation
in toxic tort cases, combined with the willingness of experts to make causative
links in the courtroom that have not been made in the laboratory, this acceptance
began to be questioned. Since Daubert and Kumho Tire require that all expert
testimony must both reliable and relevant, and that scientific expert testimony
must be based on “scientific knowledge,” there is ample room to question the
reliability of treating physicians to testify outside of their specific areas of
expertise to establish causation by means of ‘differential diagnosis.’ Thus, in the
last decade courts have begun to more closely examine the use of differential
diagnosis to establish causation.
As noted above, toxic tort cases discuss causation in terms of general
causation and specific causation. General causation refers to whether X can
cause Y, and specific causation refers to whether X did cause Y in a given
case.35 The process starts with the compilation of a comprehensive list of
competing causes that are generally capable of causing the patient’s disease or
illness. This is referred to as ‘ruling in’ all possible causes, or ‘general causation.’
General causation evidence in toxic tort cases may take many forms that have
different weight, but generally include epidemiology, animal studies, chemical
analogies, case reports, regulatory findings, MSDS information and other
secondary sources.36 Of these potential sources of evidence, epidemiology is
regarded as the most scientifically valid.37 Failure to rule in suspected causes in
Attributed to William Hazlitt (1778-1830). Although the origin of this oft-quoted phrase is not
reasonably certain, the following quote concerning the nature of lawyering is certain: “Mr. Tooke
had "the mind of a lawyer, a rigid and constant habit of attending to the exact import of every
word and clause in a sentence. ... Mr. Tooke, in fact, treated words, as the chemists do
substances; he separated those which are compounded of others from those which are not
decompoundable. He did not explain the obscure by the more obscure, but the difficult by the
plain, the complex by the simple." W. Hazlitt, Spirit Of The Age, “Mr. Horne Tooke.”
See, e.g., Kelley v. American Heyer-Schulte Corp., 957 F. Supp. 873, 875–76 (W.D. Tex. 1997)
(recognizing the different concepts of general causation and speciﬁc causation), appeal
dismissed, 139 F.3d 899 (5th Cir. 1998); Cavallo v. Star Enter., 892 F. Supp. 756, 771 n.34 (E.D.
Va. 1995), aff’d in part and rev’d in part, 100 F.3d 1150 (4th Cir. 1996), cert. denied, 522 U.S.
1044 (1998); Casey v. Ohio Med. Prods., 877 F. Supp. 1380, 1382 (N.D. Cal. 1995).
Hollingsworth, supra note 25 at 89; Michael D. Green, et al., Reference Guide on
Epidemiology, in Reference Manual on Scientific Evidence 335-36 and 381-382 (2d ed. 2000).
Hollander v. Sandoz Pharm. Corp., 95 F. Supp.2d 1230, 1235 n. 14 (W.D. Okla. 2000) (“In the
absence of an understanding of the biological and pathological mechanisms by which disease
develops, epidemiological evidence is the most valid type of scientific evidence of toxic
causation”), aff’d, 289 F.3d 1193 (10 Cir. 2002); Soldo v. Sandoz Pharm. Corp., 244 F. Supp.2d
434, 532 (W.D.Pa. 2003)(“epidemiology is ‘the primary generally accepted methodology for
the general causation analysis may make the overall opinion testimony
unreliable.38 One of the potential shortcomings in this process is how to account
for unknown causes as well as for the presence of possible confounding factors39
in the available studies.
The next step to determine causation is for the physician or other expert to
examine the available evidence concerning the specific patient, and engage in a
process of eliminating potential causes until the most likely cause is identified.
This is the ‘ruling out’ process of “specific causation.” The methodology, as
outlined by the courts, is a straightforward process of elimination (a methodology
made famous by Sherlock Holmes,40 and also used in other forensic areas such
as fire cause and origin investigations). A key question in the cases is whether
this ruling out process is done reliably: “The expert must provide reasons for
rejecting alternative hypotheses ‘using scientific methods and procedures’ and
the elimination of those hypotheses must be founded on more than ‘subjective
beliefs or unsupported speculation.’”41 However, in these cases, physicians often
rely on their experience as opposed to an objective factual analysis. 42 This is
where Daubert motions are often won or lost.
demonstrating a causal relation between a chemical compound and a set of symptoms or a
disease’”) (quoting Conde v. Velsicol Chem. Corp., 804 F. Supp. 972, 1025-26 (S.D. Ohio 1992),
aff’d, 24 F.3d 809 (6 Cir. 1994). Note, however, that epidemiological studies must sort out mere
association of two events from causation. See In re Diet Drugs, No. MDL 1203, 2001 WL 454586
(E.D.Pa. Feb.1, 2001); Kelley v. American Heyer-Schulte Corp., 957 F. Supp. 873
(W.D.Tex.1997) (also noting that observed association between exposure and condition may
reflect true cause-effect relationship or spurious finding, and to distinguish between these
alternatives, it is necessary first to consider confounding factors); Michael D. Green, et al.,
Reference Guide on Epidemiology, in Reference Manual on Scientific Evidence 336-37 (2d ed.
Clausen v. M/V NEW CARISSA, 339 F.3d 1049, 1058 (9th Cir. 2003); Hall v. Baxter Healthcare
Corp., 947 F. Supp. 1387, 1413 (D.Ore. 1996); Cavallo v. Star Enter., 892 F. Supp. 756, 774
(E.D.Va. 1995), aff’d in relevant part, 100 F.3d 1150 (4 Cir. 1996).
A confounding factor in a study is a variable that is related to one or more of the variables
defined in a study. A confounding factor may mask an actual association or falsely demonstrate
an apparent association between the study variables where no real association between them
exists. If confounding factors are not measured and considered, bias may result in the conclusion
of the study.
“How often have I said to you that when you have eliminated the impossible, whatever remains,
however improbable, must be the truth?“ Sir Arthur Conan Doyle, The Sign of Four (1890). Mr.
Holmes used other variants of this phrase: “Improbable as it is, all other explanations are more
improbable still.” Silver Blaze (1892); “We must fall back upon the old axiom that when all other
contingencies fail, whatever remains, however improbable, must be the truth.” The Adventure of
Bruce-Partington Plans (1908).
Clausen v. M/V NEW CARISSA, 339 F.3d 1049, 1058 (9th Cir. 2003) (quoting Claar v.
Burlington N.R.R. Co. 499, 502 (9 Cir. 1994). Differential diagnosis, at least according to some
courts, “does not by itself prove the cause, even for the particular patient. Nor can the technique
speak to the issue of general causation. Indeed, differential diagnosis assumes that general
causation has been proven for the list of possible causes it eliminates.” Hall; supra note 38, 947
F. Supp. at 1413.
Heller v. Shaw Indus., Inc., 167 F.3d 146, 154-55 (3d. Cir. 1999) (treating physician is not
required to rely on published studies, but instead can rely on “experience with hundreds of
patients, discussions with peers, [and] attendance at conferences and seminars” which are “tools
Some cases, however, meld general and specific causation together,
finding that a ‘properly’ conducted and explained differential diagnosis, by itself or
with little else, could sufficiently establish causation in a toxic tort case.43 This
result is accomplished at times with a little sleight of hand on the meaning of
‘differential diagnosis,’ or by weighing the strength of the evidence of the
temporal proximity of the exposure to the onset of the illness.
The Reference Manual on Scientific Evidence properly defines “differential
diagnosis” as “the process of determining which of two or more diseases with
similar symptoms and signs the patient is suffering from, by means of comparing
the various competing diagnostic hypotheses with the clinical findings.” 44 This is
the everyday job of the treating physician. This definition limits differential
diagnosis to determining the cause of symptoms, and does not reference
determining the potential causes of the disease that resulted in the symptoms.
Several courts, however, have adopted a broader definition of differential
diagnosis that would include the determination of the cause of the disease itself.
The courts have largely ignored the fact that this is outside the scope of what
most clinicians do: “The use of a differential diagnosis for the purpose of therapy
only ‘follow[s] the causal stream up to a point where intervention is possible’
because, typically, physicians ‘do not care about the disease’s etiology – the
theory of its original or cause – unless understanding causation would assist in
diagnosis and treatment.’”45 For example, the following, overly-broad, definition
of differential diagnosis has been stated by several courts: “[d]ifferential
diagnosis, or differential etiology, is a standard scientific technique of identifying
the cause of a medical problem by eliminating the likely causes until the most
probably one is isolated.“46
of the trade, and should suffice for the making of a differential diagnosis even in those cases in
which peer-review studies do not exist to confirm the diagnosis of the physician.”)
Turner v. Iowa Fire Equip. Co., 229 F.3d 1202, 1208-09 (8 Cir. Cir. 2000) (although holding
that the differential diagnosis in the particular case insufficient); Zuchowicz v. United States, 140
F.3d 381, 385, 389-90 (2d Cir. 1998) (differential diagnosis plus temporal relationship of overdose
and illness); McCullock v. H.B. Fuller Co., 61 F.3d 1038, 1043-44 (2d Cir. 1995) (rejecting rule
in/rule out requirement and specific dosage requirements); Hardyman v. Norfolk & W. Ry. Co.,
243 F.3d 255, 261-65 (6th Cir.2001) (outside toxic tort context); Westberry v. Gislaved Gummi,
178 F.3d 257, 262-265 (4 Cir. 1999); Kannankeril v. Terminix Int’l, Inc., 128 F.3d 802, 808-09
(3d Cir. 1997).
Mary Sue Henifin, et al., Reference Guide on Medical Testimony, in Reference Manual on
Scientific Evidence 481 (2d ed. 2000). One Court quoted Stedman’s Medical Dictionary at 428
(25 ed. 1990): “[d]ifferential diagnosis is defined for physicians as ‘the determination of which of
two or more diseases with similar symptoms is the one from which the patient is suffering, by a
systematic comparison and contrasting of the clinical findings.” Kannankeril v. Terminix Int’l, Inc.,
128 F.3d 802, 807 (3d Cir. 1997).
Gary Sloboda, Differential Diagnosis or Distortion?, 35 U.S.F.Law Rev. 301, 304 (2001)
(quoting Herbert A. Simon, Artificial-Intelligence Approaches to Problem Solving and Clinical
Diagnosis, in Logic of Discovery and Diagnosis in Medicine at 72, 87 (Kenneth F. Schaffner ed.,
Baker v. Dalkon Shield Claimants Trust, 156 F.3d 248, 252-253 (1 Cir. 1998) (emphasis
added); Westberry v. Gislaved Gummi AB, 178 F.3d 257 (4th Cir. 1999); Glastetter v. Novartis
By adding “etiology” to the definition – a term that is separately defined in
the Reference Manual on Scientific Evidence to relate to the determination of
external causes, not to the diagnosis of disease47 - the courts have expanded
differential diagnosis to include the determination of the external cause of the
disease itself. As one court noted, “[w]hereas most physicians use the term to
describe the process of determining which of several diseases is causing a
patient's symptoms, . . . courts have used the term in a more general sense to
describe the process by which causes of the patient's condition are identified.”48
By doing so, the courts have empowered clinicians to opine on matters often well
beyond their clinical practice. Further, by allowing treating physicians to testify
on causation, the courts are admitting testimony based on standards of
evidentiary reliability that are different than experts in the fields of epidemiology,
pathology, pharmacology, or toxicology, even though “determining the etiology of
a disease involves the same scientific exercise, whether the decision is made by
a clinician, an epidemiologist, or other scientist.”49
Pharmaceuticals Corp., 252 F.3d 986 (8th Cir. 2001); Hardyman v. Norfolk & Western Ry. Co.,
243 F.3d 255 (6th Cir.2001); Jennings v. Baxter Healthcare Corp., 14 P.3d 596 (Ore. 2000). See
also, McCullock v. H.B. Fuller Co., 61 F.3d 1038, 1044 (2d Cir. 1995) (describing differential
etiology as an analysis “which requires listing possible causes, then eliminating all causes but
one”); Glaser v. Thompson Med. Co, 32 F.3d 969, 978 (6th Cir. 1994) (differential diagnosis is “a
standard diagnostic tool used by medical professionals to diagnose the most likely cause or
causes of illness, injury and disease”). Following this overly broad definition, one court went on to
explain the proper role of differential diagnosis – diagnosing a disease – thus further confounding
the difference between diagnosis of a disease and causation: “A reliable differential diagnosis
typically, though not invariably, is performed after ‘physical examinations, the taking of medical
histories, and the review of clinical tests, including laboratory tests,’ and generally is
accomplished by determining the possible causes for the patient’s symptoms and then eliminating
each of these potential causes until reaching one that cannot be ruled out or determining which of
those that cannot be excluded is the most likely.” Westberry, 178 F.3d at 262 (emphasis added);
quoting Kannankeril v. Terminix Int’l, Inc., 128 F.3d 802, 807 (3d Cir. 1997).
Mary Sue Henifin, et al., supra n. 35: “differential etiology: A term used on occasion by expert
witnesses or courts to describe the investigation and reasoning that leads to a determination of
external causation, sometimes more speciﬁcally described by the witness or court as a process of
identifying external causes by a process of elimination” (emphasis added).
Clausen v. M/V NEW CARISSA, 339 F.3d 1049, 1057 n.4 (9th Cir. 2003) (citing Stedman's
Medical Dictionary 474 (26th ed. 1995) (emphasis in original). The court noted that there was not
universal agreement in the medical field for a definition of differential diagnosis, and that
environmental and occupational health physicians use the term in the broader sense. The court
concluded that since both experts in the case used the term in the broader sense, it did not want
to ‘muddy the waters’ with the distinction. See also Easum v. Miller, 92 P.3d 794, 802 (Wyo.
2004). Some courts have recognized that this distinction actually makes a difference. In In re
Breast Implant Litigation, 11 F. Supp.2d 1217, 1230 (D.Co.1998), the court held that a
"[d]ifferential diagnosis may be utilized by a clinician to determine what recognized disease or
symptom the patient has, but it is incapable of determining whether exposure to a substance [or
in this case a harm] caused disease in the legal sense." Simply put, "an untested hypothesis
cannot be a scientifically reliable basis for an opinion based on causation." Id.; see also Lennon
v. Norfolk and Western Ry. Co.123 F. Supp.2d 1143, 1154 (N.D.Ind. 2000) (“differential diagnosis
does not by itself prove the cause, even for the particular patient.”)
Moore v. Ashland Chemical, 151 F.3d 269, 275 n.6 (5th Cir. 1998) (en banc), cert. denied, 526
U.S. 1064 (1999); see also O’Conner v. Commonwealth Edison Co., 13 F.3d 1090, 1105 n. 14
(7 Cir. 1994) (“we do not distinguish the treating physician from other experts when the treating
physician is offering expert testimony regarding causation.”)
Causation in toxic tort cases involves overlapping areas of expertise, but
epidemiologists are not qualified to rule out potential causes of a disease for
specific causation, and treating physicians are not generally qualified to ‘rule in’
causes based on differential diagnosis. Further, although treating physicians
likely could reasonably ‘rule out’ certain potential causes of a disease in certain
cases, they are being asked to render post hoc opinions well after treatment
decisions are made, based, to one degree or another, on speculation. Whether a
treating physician’s opinion is appropriate under a Daubert analysis of course
depends on the particular facts of the case, but the courts would do well to
remember that "[t]he courtroom is not the place for scientific guesswork, even of
the inspired sort. Law lags science; it does not lead it."50
III. THE ETIOLOGY OF THE JUDICIAL MISUSE OF DIFFERENTIAL
DIAGNOSIS: WESTBERRY V. GISLAVED GUMMI, AB
One of the leading post-Daubert cases to allow a treating physician to
testify on causation based on a differential diagnosis is Westberry v. Gislaved
Gummi AB, 178 F.3d 257 (4th Cir. 1999). In Westberry, the plaintiff’s treating
physician testified to a ‘reasonable degree of medical certainty’ that exposure to
talc dust in an unknown quantity aggravated plaintiff’s sinus disease, despite his
admission that he lacked any scientific evidence to support that conclusion. The
Fourth Circuit found that the physician’s opinion was based on a ‘reliable’
differential diagnosis. Westberry now stands for the proposition that a treating
physician can testify as to causation in a toxic tort (or other) case based solely on
a differential diagnosis. This is not simply an isolated opinion, an ‘outlier’ in the
progression of Daubert causation cases. To date, Westberry has been cited in
55 state and federal cases51 related to the use of a reliable differential diagnosis
Rosen v. Ciba-Geigy Corp., 78 F.3d 316, 319 (7th Cir.1996) (Posner, J.); see also Moore v.
Ashland Chemical, 151 F.3d 269, 276 (5th Cir. 1998) (en banc), cert. denied, 526 U.S. 1064
(1999) (“In sum, the law cannot wait for future scientific investigation and research. We must
resolve cases in our courts on the basis of scientific knowledge that is currently available. The
inquiry authorized by Rule 702 is a flexible one; however, a scientific opinion, to have evidentiary
relevance and reliability, must be based on scientifically valid principles.”); contra Turner v. Iowa
Fire Equip. Co., 229 F.3d 1202, 1209 (8 Cir. 2000) (“The first several victims of a new toxic tort
should not be barred from having their day in court simply because the medical literature, which
will eventually show the connection between the victims’ condition and the toxic substance, has
not yet been completed. If a properly qualified medical expert performs a reliable differential
diagnosis through which, to a reasonable degree of medical certainty, all other possible causes of
the victim’s condition can be eliminated, leaving only the toxic substance as the cause, a
causation opinion based on that differential diagnosis should be admitted.”); cited with approval,
Hollander v. Sandoz Pharm. Corp., 289 F.3d 1193, 1211-12 (10 Cir. 2002).
Westberry was cited most recently this past week in Rink v. Cheminova, 2005 WL 428418 (11
Cir. 2005). Not all courts, however, follow the Westberry analysis. The decision is discussed
negatively in Coastal Tankships, U.S.A. v. Anderson, 87 S.W.3d 591, 605 ff. (Tex.App. 2002)
(noting split in federal cases whether a differential diagnosis, without more, can reliably show both
specific and general causation); Valentine v. PPG Industries, 821 N.E.2d 580, 598 ff. (Ohio App.
2004); Roche v. Lincoln Property Co., 278 F. Supp.2d 744, 751 ff. (E.D.Va. 2003); Praytor v. Ford
Motor Co., 97 S.W.3d 237, 244ff. (Tex.App. 2002); and Plourde v. Gladstone, 190 F. Supp.2d
708, 721ff. (D.Vt. 2002).
to prove causation, and 364 times generally in decisions, appellate briefs, and
secondary sources. What the Court’s opinion does not reveal, however, is that
the treating physician in the case did not perform his final differential diagnosis
until cross-examination in the courtroom, and clearly did not understand the
meaning and import of ‘reasonable medical certainty.’ Westberry has thus
helped create a legal standard wholly unsupported by the record in the case and
now followed by other state and federal courts.52
A. THE FACTS
Westberry worked for a window manufacturer. The manufacturer
purchased rubber products for installation into its windows from a Swedish
company, Gislaved Gummi AB (“GGAB”). A thin layer of talc was placed on the
gaskets as a lubricant in order to prevent the rubber from sticking at pinch points
in the plant’s machinery. Additionally, the presence of talc on the gaskets
prevented employees from suffering blisters and “rubber burns” from handling
and stapling the gaskets into place in the window frames.
Westberry began work at the plant in 1992. He worked on the production
line, adjacent to the gasket cutting area, for two years before switching to a job
as a gasket cutter in January, 1994. His job consisted of unpacking gaskets from
GGAB’s boxes, putting gasket rolls on a cutting machine, and cutting gaskets to
the various lengths needed for the production line. Westberry claimed that his
duties brought him into contact with high concentrations of airborne talc.53
Neither the manufacturer nor GGAB warned Westberry regarding any dangers
associated with airborne talc, and Westberry did not wear any mask or other
protective equipment during his employment.
Westberry had a long and complicated history of sinus problems dating
back to 1979. In 1980, his ENT doctor attributed his sinus disease to allergies
and a deviated septum. By 1988, four years before he began working with for
the window manufacturer, Westberry’s medical records revealed a history of
multiple nasal polyps, tinnitus, and ongoing problems with his sinuses. By 1992,
Westberry had undergone three separate surgeries for removal of nasal polyps
caused by sinusitis. In November 1992, he again developed nasal polyps, which
his doctor at the time decided not to remove.
In May 1994, three months after starting his job as a gasket cutter,
Westberry caught a cold and had a sore throat (a possible cause of sinusitis).
The cold lasted about two weeks. Westberry was hospitalized for a severe sinus
See, e.g., Hollander v. Sandoz Pharmaceuticals, 289 F.3d 1193 (10 Cir. 2002); Easum v.
Miller, 92 P.3d 794, 802 (Wyo. 2004); Shaver v. U.S.,319 F.Supp.2d 649, 665 (M.D.N.C.,2004);
CSX Transp., Inc. v. Miller, 858 A.2d 1025, 1072-1073 (Md.App. 2004) (citing Westberry as the
“definitive analysis of the use of differential diagnosis by a physician”); Praytor v. Ford Motor Co.
97 S.W.3d 237 (Tex.App. 2002); Keener v. Mid-Continent Cas.,817 So.2d 347 (La.App. 2002).
The opinion in Westberry does not mention that OSHA investigators tested the gasket cutting
area and found no measurable trace of respirable talc.
infection and chronic sinusitis on July 25, 1994, and was treated with antibiotics.
From that point on, Westberry was treated by Dr. David Isenhower, a board
certified otolaryngologist. Beginning in September 1994, Westberry underwent
several sinus surgeries, including a “sinus obliteration” in April 1995, in which his
frontal sinus was removed.
Westberry sued GGAB, alleging that the failure to warn of the dangers of
breathing airborne talc proximately caused the aggravation of his pre-existing
sinus condition. Dr. Isenhower was the sole witness for Westberry on causation.
The district court declined to consider GGAB’s Daubert motion, holding that
Daubert did not apply to ‘non-scientific’ opinion testimony.54 Despite his
admission that the etiology of sinusitis is not well known, that there were no
scientific studies linking talc exposure to sinusitis, that he did not know the levels
of exposure to talc at the plant, and that there were several other potential
causes of Westberry’s sinus troubles, and despite Westberry’s extensive history
of sinus ailments, Dr. Isenhower (to a reasonable degree of medical certainty)
attributed Westberry’s exacerbated condition to his inhalation of talc from
GGAB’s rubber gaskets. The jury returned a verdict for $400,000.00.
On appeal, the Fourth Circuit conducted a Daubert analysis of Dr.
Isenhower’s testimony. First, the Court noted that differential diagnosis is a
“standard scientific technique” that “has widespread acceptance in the medical
community, has been subject to peer review, and does not frequently lead to
incorrect results.”55 The Court noted that, with one exception, other circuit courts
have found that medical opinion testimony based upon a “reliable differential
diagnosis” is “sufficiently valid to satisfy the first prong of the Rule 702 inquiry.” 56
The case was tried before the Supreme Court’s decision in Kumho Tire Co. v. Carmichael, 526
U.S. 137 (1999), that definitively expanded the Daubert gatekeeping function to all expert opinion
testimony under the Rule 702 of the Federal Rules of Evidence. The Fourth Circuit ruled that the
failure of the lower court to conduct a Daubert review his was error, and therefore the opinion was
in fact the first Daubert review of the evidence. 178 F.3d at 262. Nevertheless, did not review the
decision de novo, but reviewed it under an abuse of discretion standard. 178 F.3d at 261; see
General Electric v. Joiner, 522 U.S. 136, 139 (1997).
178 F.3d at 262, citing In re Paoli R.R. Yard PCB Litig., 35 F.3d 717, 758 (3d Cir. 1994) and
Heller v. Shaw Indus., Inc., 167 F.3d 146, 154-55 (3d. Cir. 1999). This oft-repeated reference
designed to sweep in three of the four Daubert factors is inaccurate as it relates to the ‘known
error rate.’ The known error rate of differential diagnosis relates to the diagnosis of disease, not
the identification of external causes. Thus, the Fourth Circuit further confounded the definition of
178 F.3d at 263, citing Heller v. Shaw Indus., Inc., 167 F.3d 146, 154, 156-57 (3d. Cir. 1999);
Kennedy v. Collagen Corp., 161 F.3d 1226, 1228-1230 (9 Cir. 1998), cert. denied, 526 U.S.
1099 (1999); Baker v. Dalkon Shield Claimants Trust, 156 F3d 248, 252-53 (1 Cir. 1998);
Zuchowicz v. United States, 140 F.3d 381, 385-87 (2d. Cir. 1998); and Ambrosini v. Labarraque,
101 F.3d 129, 140-41 (D.C.Cir. 1996); contra Moore v. Ashland Chem., Inc., 151 F.3d 269, 277-
79 (5 Cir. 1998) (en banc), cert. denied, 526 U.S. 1064 (1999). See also Benedi v. McNeil-
P.P.C., Inc. 66 F.3d 1378, 1383-85 (4 Cir. 1995).
B. THE CHRONOLOGICAL DEVELOPMENT OF A REAL-TIME,
COURTROOM DIFFERENTIAL DIAGNOSIS IN WESTBERRY
Dr. Isenhower developed his causation opinion on demonstrably incorrect
information. While treating Westberry, he found no clinical evidence linking talc
to Westberry’s sinusitis, although he looked for clinical evidence in the form of
“granulomas” in the tissue samples that might have formed around talc
particles.57 When he met with Westberry’s counsel in 1995, before suit was filed,
he said that he could not render an opinion that the talc caused Westberry’s
sinus problems, because he could not make the necessary histological
connection without solid clinical data.58 He also knew, as he readily
acknowledged at trial, that there was there no medical evidence linking exposure
to talc to sinusitis:
Q: Now, you know of no medical journal article that specifically links the
exposure of talc to sinusitis?
A: That is correct.
Q: And you know of no medical journal article that specifically links the
exposure to talc to the development of polyps in the nose?
A: That is also correct.
Q: You know of no epidemiological studies that make those links, is that
A: That is also correct.
Q: You know of no animal studies that make those links, is that correct?
A: That is also correct.
Q: And you know of no medical textbooks that you’ve looked at that makes
A: Talc is not discussed.
The pathology reports found no talcum particles in the tissue samples. Joint Appendix, 313-
Joint Appendix, 307; 774. His medical notation at the time was as follows: “James Westberry
is here with his attorney’s (sic) to discuss the possibility of a law suit against a manufacturer of
the talcum powder that [the manufacturer] uses on its window rubber gaskets. I told the
attorney’s (sic) that since no granulomas have been identified in his tissue, then we could not
make a solid case for approximate (sic) cause from the talcum exposure. The time[,] course and
degree of illness remain very suspicious, but if we need a question of negligence raised or
approximate (sic) cause then we need at least that histologic[al] connection.”
Q: Right. In fact, any problems with talc are with high exposure and in the
lungs. That’s the only –
A: In the lungs and in the abdominal cavity.
Q: Okay, but we’re not talking about the sinus cavity?
A: Not the upper respiratory, no.59
Although he was initially reluctant to blame talc as the causative agent
without clinical evidence, Dr. Isenhower changed his mind when another one of
his patients, also a worker at the same plant, developed sinusitis. Based on this
fact, and solely on this fact, Dr. Isenhower rendered his opinion at deposition and
on direct examination that the aggravation of Westberry’s preexisting sinus
disease was caused by exposure to airborne talc:
Q: And it was that one other employee that made you be able to
testify with a reasonable degree of certainty on proximate cause, is that right?
A: I believe that was the only other factor, yes sir.60
According to Dr. Isenhower, he now had what he believed was the first
reportable case of airborne talc causing (or at least aggravating) sinus disease.
The problem with this analysis, however, and nowhere mentioned in the Fourth
Circuit’s opinion, was that Dr. Isenhower was confronted on the witness stand
with evidence that this other patient was never exposed to airborne talc at the
plant. It was only then – in the courtroom – that Dr. Isenhower performed what
the Fourth Circuit deemed a ‘reliable’ differential diagnosis. When confronted
with evidence that this other worker was in fact not exposed to talcum powder,
Dr. Isenhower nevertheless stuck to his opinion. The trial judge – in this case,
more of a doorman than a gatekeeper – admitted the testimony, and the Fourth
Circuit determined there was no abuse of discretion because the testimony of the
treating physician on causation was based on a ‘reliable’ differential diagnosis.
This despite the fact that the expert acknowledged that medical science does not
know the causes of sinusitis.
Further, in the course of formulating this ‘differential diagnosis’ on the
cause of Westberry’s sinusitis, Dr. Isenhower acknowledged that he had only a
temporal relationship between alleged cause and effect, that he did not know all
the causes of sinusitis, and did not understand the legal standards for ‘medical
certainty,’ even noting the difference between making a recommendation based
on clinical findings versus a definitive causation opinion. His testimony is
Joint Appendix, 302-303.
Q: Now, you would agree at the close of the day here, you have no scientific
evidence that links talc to this man’s problem?
A: That is correct.
Q: Is it still your opinion with a reasonable degree of medical certainty that
talc caused this man’s problem?
A: It’s the only exposure that changed that accelerated a pre-existing
condition that he had up to the multiple surgical problems that he had. He
had previous disease in younger years and it was a long term . . .
quittance (sic) of the disease. Again, this is the kind of situation where I
have a man who tells this story of exposure who has these problems
which he had known about in the past. He knows what they are because
he’s had volumes of, but similar problems in the past. He gets this
exposure and there’s an acceleration of his problems. It’s the kind of thing
where I say, well, he shouldn’t be there [at the plant]. Maybe there is a
problem. But as far as a rigorous duplicated scientific or laboratory proved
experiment, of course not. There is no such thing.
Q: You don’t know what caused his sinusitis and polyps in ‘79, ‘80, ‘81, ‘85,
‘92? . . .
A: No. . . .
Q: You don’t know the cause of all that?
A: No, we don’t know the true mechanism of the formation of polyps or
sinus disease, other than obstruction of drainage from the sinuses.
Q: And a common cold can cause polyps? A common cold can cause
A: Okay. You need to put that intercept in there. A common cold can cause
sinusitis, but again, most folks don’t get that. So there is something else
about your system –
Q: Well, he’s not most folks, right?
A: That’s true.
Q: So if he had a common cold, and we know he had a sore throat in April or
May of 1994, which could have been a common cold, correct?
Q: And then in July of 1994, the problems start, right?
A: That’s the date of my consultation with him, yes.
Q. All right. Now, despite the fact that you have absolutely no evidence as to
the amount of talc at the work place –
A. That’s right.
Q. You have no evidence that any other of your patients were exposed to talc
in the work place. You’ve got no evidence that any customer of Gislaved
Gummi has ever had a problem with talc. You have no evidence that any
other person at the Greenwood plant had problems with talc. You don’t
know what level of talc Mr. Westberry was exposed to. Is it still your
opinion with a reasonable degree of medical certainty – and there’s no
histological evidence that says there is talc in these samples, you are still
saying that it is your opinion with a reasonable degree of certainty that talc
A. I suppose I could be made to look very foolish by sticking to my guns, but -
Q. You don’t have to.
A. I will. Really, I have no other factor to account for the acceleration of the
abnormalities that I saw in this man.
Q. It is true, though, that back in March of 1995, you had no other factor?
A. That is true.
Q. And that other factor was supposedly [your other patient with sinusitis]?
A. Yes. . . .
Q. But you have no evidence that [the other patient] was exposed to talc and
that’s what you needed. That’s the link you needed in 1996 to draw this
connection. Now you don’t have that link, do you?
A. Let me think on that a second. I had suspected an environmental
exposure on Mr. Westbury (sic) before I met [my other patient]. I’d like to
— I was extremely hesitant to make comment that might be against one of
our local businesses because I need a good cause for that because I
considered them a responsible employer. But I had enough evidence to
affect clinical decisions on Mr. Westberry. And not having legal
training, it is hard to know where you go from enough evidence to
make a medical decision and take somebody to the operating room
and work on his head surgically to go to a medical opinion that’s
expressed to a jury.
Q. We’re talking about a reasonable degree of medical certainty, not legal
certainty. Your profession.
A. I have no other factor [i.e., causative agent], and I would love to learn of
another one with him, and to my understanding of the determination of
reasonable degree of medical certainty, that is a way for me to arrive at
that. I have no other change that would have accelerated his sinus
problems other than his exposure.
Q: And now you’re not so sure about that exposure to [the other patient], are
A: [He] had an exposure at the work place to a particul[ate] matter. And
again, with the records that I reviewed that you provided for me here
today, I did not document an exposure to talc. This is clearly a case that
has troubled me from the very beginning because we have
performance changes in the sinus, and acceleration of sinus
problems, and yet, not rigorous scientific proof, at least according to
what we’ve perceived the proof needed to be. We know that in other
areas of the body, the reaction is of a granuloma formation never seen in
the nose. We don’t know whether it does it there. And my statements
along the way were predicated upon believing that you need the same
immune reactivity in the nose as you would elsewhere. And I still think
that makes sense, but --
Q: Scientifically, you would want that, wouldn’t you?
A: Yes, you would.
Q: We don’t have that here, do we?
A: The only thing we have is the clinical change in the behavior of Mr.
Westberry’s sinus disease.61
The record is clear that Dr. Isenhower had no basis for his opinion other
than the temporal relationship between exposure and disease.62 Further, he had
no idea what a medically certain opinion meant –acknowledging even that his
Joint Appendix 330-331, 334-335.
Cf. Moore, 151 F.3d at 278 (noting that “[i]n the absence of established scientific connection
between exposure and illness, or compelling circumstances . . . the temporal connection between
exposure to chemicals and an onset of symptoms, standing alone, is entitled to little weight in
opinion could look “foolish.” The Fourth Circuit tried to provide a factual basis to
support the opinion, noting the MSDS63 (which was not relied on by Dr.
Isenhower), the purported differential diagnosis, and the temporal relationship.
As to the differential diagnosis, the court stated that Dr. Isenhower had reliably
ruled out other causes. That is both unsupported by the record,64 and
inaccurate: Dr. Isenhower never explained how he could rule out Westberry’s
cold as the cause of his sinusitis, did not reasonably exclude infection from
bacteria, and he admitted that he did not know the etiology of sinusitis. In other
words, he could not rule out unknown causes, and he did not rule out known
causes. Clearly, an epidemiologist or a toxicologist could not have opined on
these facts that there was causation.
B. CIRCUIT COURT ACCEPTANCE OF WESTBERRY AND
DIFFERENTIAL DIAGNOSIS BY TREATING PHYSICIANS
Although the Westberry opinion does not fully recite the record developed
in the case, it nevertheless is routinely cited as one of the leading cases to
support the proposition that a treating physician can opine on causation based on
a reliable differential diagnosis, without any epidemiological or toxicological
evidence. Most circuit courts now agree that a “reliable” or “scientifically valid”
differential diagnosis, either alone or with other minimal evidence such as a
temporal relationship between purported cause and its effect, is sufficient to
prove causation.65 In other words, most courts may not require an expert to
Further, the MSDS did not specify the amount of talc necessary to irritate the mucous
membranes (no dose-response information), and did not specify that sinusitis was a possible
outcome. Cf. Moore, 151 F.3d at 278 (noting that MSDS is of limited value to support causation
opinion by treating physician because it did not reveal level of exposure necessary to sustain
injuries about which the MSDS warned).
Dr. Isenhower ruled out exposure to bacteria from water-skiing (Westberry had water-skied
numerous times right before his hospitalization) because he had never seen such a cause and
effect – but then admitted he had also not seen a cause and effect relationship between talc and
Q: You could get an infection from [water-skiing], could you not?
A: That is correct.
Q: That is a possible cause, it is not, of this condition?
A: Yes, sir. However, I have never seen it happen from Lake Greenwood in my time there. If it’s
the cause, then Mr. Westberry would be the only case that I’ve seen.
Q: Well, this talc might be the only case you’ve seen, too, correct?
A: That is correct.
Joint Appendix, 325-327.
Zuchowicz v. United States, 140 F.3d 381, 385, 389-90 (2d Cir. 1998), Heller v. Shaw Indus.,
Inc., 167 F.3d 146, 154-55 (3d. Cir. 1999); Westberry, 178 F.3d 257, 262-266 (4 Cir. 1999);
Hardyman v. Norfolk & W. Ry. Co., 243 F3d 255, 260 (6 Cir. 2001) (citing Westberry); Turner v.
Iowa Fire Equip. Co., 229 F.3d 1202, 1208 (8 Cir. 2000) (citing Westberry), Glastetter v.
Novartis Pharms. Corp., 252 F.3d 986, 989 (8th Cir. 2001) (citing Westberry), Lauzon v. Senco
Products, Inc., 270 F.3d 681 (8 Cir. 2001) (citing Westberry), Mattis v. Carlon Electrical
Products, 295 F.3d 856, 861 (8 Cir. 2002), and Kudabeck v. The Kroger Co., 338 F.3d 856 (8th
Cir. 2003) (citing Westberry); Clausen v. M/V NEW CARISSA, 339 F.3d 1049, 1058 (9th Cir.
2003) (citing Westberry); Hollander v. Sandoz Pharms. Corp., 289 F3d. 11931209-1212 (10th Cir.
2002) (citing Westberry), Goebel v. Denver and Rio Grande Western R. Co., 346 F.3d 987 (10
always be able to ‘rule in’ causes before he rules them out.66 Only the Fifth and
Eleventh Circuits appear to have rejected the use of differential diagnosis,
standing alone, to establish causation.67
The legal battleground, at least for the time being, is therefore over the
reliability or validity of the treating physician’s (or other expert’s) process or
technique of ruling in potential causes (including unknown causes and
confounding factors), and ruling out specific causes. It is clear that when there is
competing evidence concerning the cause of a disease, then the experts must
engage in the “Holmesian” process of identifying potential causes and eliminating
unlikely causes. This is true not just in toxic tort cases, but also in most products
liability cases. “[N]o opinion we know of has concluded that differential etiology,
when properly performed, is inadmissible.”68 In terms of a Daubert analysis, it is
useful to analyze the determination of causation by dividing the question of
scientific validity into three categories: “(1) the validity of the underlying principle,
(2) the validity of the technique applying the principle, and (3) the proper
application of the technique on a particular occasion.”69 The “principle” in these
cases is simply the use of deductive reasoning, and the technique of “differential
diagnosis” applying deductive reasoning is clearly accepted by the courts
(although, when there is some judicial sleight of hand when courts intone that
“differential diagnosis” has been accepted by peer review and has a known error
rate, because that applies only to the diagnosis of symptoms, not to ‘differential
etiology’).70 It would seem from an examination of the cases that there is a
sliding scale of admissibility based on the availability of evidence on general
causation and on the strength of the evidence concerning the temporal order of
exposure and onset of symptoms (except in the Fifth and Eleventh Circuits).71
Until the courts grapple more closely with what constitutes a “proper” technique
of differential etiology, the battleground in the cases will be on the
reasonableness of each step of the expert’s evaluation, closely examining what
is not known and what was not done – and why it was not done.
If the expert in question is not the typical testifying expert, but rather a
treating physician unfamiliar with the courtroom, or with all potential causes of a
disease or illness, or with the legal certainty standard, he may be made to be
Cir. 2003) (citing Westberry), Bitler v. A.O. Smith Corp., 391 F.3d 1114, 1123-24 (10 Cir. 2004)
(citing Westberry, although not in medical context); Ambrosini v. Labarraque, 101 F.3d 129, 140-
41 (D.C.Cir. 1996).
Joseph Sanders and Julie Machal-Fulks, The Admissibility of Differential Diagnosis Testimony
to Prove Causation in Toxic Tort Cases: The Interplay of Adjective and Substantive Law, 64 Law
& Contemp. Probs. 107, 122-29 (Autumn 2001).
Moore v. Ashland Chemical, Inc., 151 F.3d 269, 275 (5th Cir. 1998) (en banc), cert. denied,
526 U.S. 1064 (1999); Rink v. Cheminova, 2005 WL 428418 at *6 (11th Cir. 2005) (distinguishing
Sanders et al, supra note 66 at 120.
Sanders, et al. at 120, quoting Professor Paul Gianelli, The Admissibility of Novel Scientific
Evidence: Frye v. United States, a Half-Century Later, 80 Colum.L.Rev. 1197, 1201 (1980).
See note 55, supra.
Sanders et al., supra note 66 at 129-34.
very uncomfortable in assessing ‘certainty’ to his opinions if there is little more
than a temporal relationship between purported cause and supposed effect. In
particular, most treating physicians do not have the luxury to examine the
veracity of statements made by their patients in the medical history. That can be
fertile ground for attacking the causation opinion of a treating physician. If the
physician relies on a medical history that omits important information relevant to
general causation, the opinion should be inadmissible under a Daubert analysis.
For example, in a pre-Daubert opinion, the Fifth Circuit excluded a physician’s
testimony on just those grounds:
We do not hold, of course, that admissibility of an expert opinion
depends upon the expert disproving or discrediting every
possible cause other than the one espoused by him. Here,
however, Dr. Johnson has admitted that Viterbo’s symptoms
could have numerous causes and, without support save
Viterbo’s oral history, simply picks the cause that is most
advantageous to Viterbo’s claim. Indeed, Dr. Johnson’s
testimony is no more than Viterbo’s testimony dressed up and
sanctified as the opinion of an expert. Without more than
credentials and a subjective opinion, an expert’s testimony that
“it is so” is not admissible.72
Even so, as in Westberry, it may not matter whether there is solid science or not
to validate an opinion: if a district court gatekeeper is inclined to ‘let the jury
decide,’ since the standard of review is abuse of discretion, then there is not
much hope of reversal on appeal.
Although most courts continue to confuse the two, there is a clear
difference between diagnosing a disease or illness and identifying the external
cause of that disease or illness. A treating physician may be ably equipped for
the former, but have no expertise for the latter. The concept of ‘reasonable
medical certainty,’ furthermore, seems to have little weight when treating
physicians are engaging in little more than guesswork when they ignore unknown
causes or attempt to ‘eliminate’ potential causes. Such testimony amounts to
little more than reciting what the plaintiff told them, as the treating physician has
no incentive to conduct any independent factual investigation to reliably eliminate
Viterbo v. Dow Chemical, 826 F.2d 420, 424 (5 Cir. 1987); see also In re Paoli R.R. Yard PCB
Litig., 35 F.3d 717, 762 (3d Cir. 1994) (otherwise allowing opinion testimony based on “standard
diagnostic techniques,” but affirming exclusion of two experts who based their causation opinion
“solely on plaintiff’s self-report of illness in preparation for litigation”); Diaz v. Johnson Mattehy,
Inc., 893 F. Supp. 358, 376 (D.N.J. 1995) (failure to rule out alternative causes); Pick v. Am. Med.
Sys., Inc., 958 F. Supp. 1151, 1168 (E.D.La. 1997) (same); Wooley v. Smith & Nephew Richards,
Inc., 67 F. Supp.2d 703, 709 (S.D.Tex. 1999) (failure to interview patient, examine patient, or
consider all patient’s medical records); Cooper v. Smith & Nephew, Inc., 259 F.3d 194, 201-03
(4 Cir. 2001) (same)
causes. Given that there is at present little hope, at least in the federal circuits,
that the appellate courts will reverse the admission of a treating physician’s
opinion, defense counsel must explore in detail the ‘certainty’ with which treating
physicians rule in specific causes, and rule out specific causes. Cross-
examination may reveal that the opinion is nothing more than an educated
Further, in cases where there is, relatively speaking, not much at stake, a
court may not be inclined to dig into the scientific and clinical issues involved to
ferret out what makes a “reliable” differential diagnosis. Defense counsel would
be well served to counter the testimony of treating physicians with clear and
concise testimony from their own experts explaining the difference between a
clinical diagnosis for treatment purposes, and the methodology necessary to
render a scientifically valid opinion on causation. Finally counsel should continue
to press the distinction between diagnosis and etiology at the lower court and
appellate levels (particularly pressing the issue of whether a treating physician’s
‘differential etiology’ is indeed a peer reviewed process with a known error rate),
until the courts develop a more closely reasoned framework for evaluating the
admissibility of causation opinions.