Donald Rumsfeld once justified attacking Iraq without firm evidence that his target possessed the ultimately apocryphal “weapons of mass destruction” by citing fear of the unknown:
Reports that say that something hasn’t happened are always interesting to me, because as we know, there are known knowns; there are things we know we know. We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns − the ones we don’t know we don’t know. And if one looks throughout the history of our country and other free countries, it is the latter category that tend to be the difficult ones.
That worked really well in Iraq. But it works even worse with differential etiology, often referred to, not particularly correctly, as “differential diagnosis.” That’s because the “differential” part of this testimonial technique is really just a form of circumstantial evidence – a “process of elimination.” “Differential diagnosis 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.” Guinn v. AstraZeneca Pharmaceuticals LP, 602 F.3d 1245, 1253 (11th Cir. 2010). It is “a process of identifying external causes by a process of elimination.” McClain v. Metabolife International, Inc., 401 F.3d 1233, 1252 (11th Cir. 2005).
A differential etiology is a process-of-elimination approach to determining a subject’s cause of injury. Under this method, an expert considers all relevant potential causes of the symptoms and then eliminates alternative causes. Although the parties and the district court below refer to this method as a “differential diagnosis,” that term is really a misnomer. A “diagnosis” is concerned only about naming the condition or ailment, not establishing its cause.
C.W. v. Textron, Inc., 807 F.3d 827, 832 n.4 (7th Cir. 2015) (citation omitted).
“Unknowns” are a recurring problem with this “differential” process. If one doesn’t know what the possible causes are, then it will be impossible to eliminate them from the equation. A significant level of idiosyncratic causation is fatal to the valid use of differential etiology.
Although differential etiologies are a sound methodology in principle, this approach is only valid if general causation exists and a substantial proportion of competing causes are known. Thus, for diseases for which the causes are largely unknown, such as most birth defects, a differential etiology is of little benefit.
Michael D. Green, et al., “Reference Guide on Epidemiology,” in Federal Judicial Center, Reference Manual on Scientific Evidence 549, 618 (3d ed. 2011).
This technique is more accurately described as a “differential etiology.” It is most useful when the causes of a substantial proportion of the disease are known. . . . When the causes of a disease are largely unknown . . ., differential etiology is of little assistance.
Restatement (Third) of Torts: Physical & Emotional Harm § 28, comment c(4) (2010).
The validity of a differential diagnosis depends on a substantial proportion of the independent causes of the disease being known. . . . A differential diagnosis is of limited utility when a substantial portion of the incidence of a disease is of unknown etiology.
Id. at Reporter’s Notes to comment c(4).
The Third Restatement relies on commentary that has long understood that “unknowns” preclude the proper use of differential etiology:
[It] becomes particularly important when . . . one cannot assume that most causes of any given ailment are known. Otherwise, . . . the elimination of other risk factors would not significantly increase the likelihood that the exposure was the cause of the plaintiff’s injury. In such situations, one cannot make a Sherlock Holmes-like deduction that simply because all other known causes have been eliminated, the only known cause left, no matter how improbable, must be the actual cause.
Joseph Sanders & Julie Machal-Fulks, “The Admissibility of Differential Diagnosis Testimony to Prove Causation in Toxic Tort Cases: The Interplay of Adjective & Substantive Law,” Law & Contemp. Probs., 107, 133-34 (Autumn 2001). Differential etiology “can only be valid, however, when risk factors that account for most cases of the plaintiff’s injury and their interactions are understood.” Susan R. Poulter, “Science & Toxic Torts: Is There A Rational Solution to the Problem of Causation?, 7 High Tech. L.J. 189, 235 (1992).
Plentiful precedent likewise recognizes the incompetency of differential etiology where the condition at issue is often idiopathic in nature. “Because idiopathy accounts for more than half of the cases of [the disease at issue], a differential diagnosis could be considered inherently unreliable here.” Hall v. Conoco Inc., 886 F.3d 1308, 1315 (10th Cir. 2018) (affirming exclusion).
[T]hat decision is particularly critical here given the extensive number of [disease] cases that are idiopathic. Under such circumstances, eliminating a number of potential causes − without properly and explicitly “ruling in” a cause − is simply “of little assistance.” . . . [Plaintiff’s expert] needed some other method to “rule out” an idiopathic diagnosis. She did not provide one. As such, the district court acted within its discretion to conclude that the extraordinary number of idiopathic  cases, coupled with the lack of a reliable means to rule out an idiopathic diagnosis here, muted [plaintiff’s expert’s] ability to reliably apply this methodology.
Milward v. Rust-Oleum Corp., 820 F.3d 469, 476 (1st Cir. 2016) (quoting Restatement Third, §28, comment c(4), discussed above). Similarly, in Kilpatrick v. Breg, Inc., 613 F.3d 1329 (11th Cir. 2010), the plaintiff’s expert “clearly testified that he could not explain why potentially unknown, or idiopathic alternative causes were not ruled out. . . . Thus, the key foundation for applying differential diagnosis was missing.” Id. at 1343.
[Plaintiff’s expert] ignored such background risks. While recognizing the existence of idiopathic (or unknown) causes of [the condition], he dismissed them by merely stating that the risk of idiopathic [disease] is essentially zero. The failure to take into account the potential for idiopathically occurring [disease] − particularly when [this condition] is a relatively new phenomenon in need of further study − placed the reliability of [the expert’s] conclusions in further doubt.
Id. at 1342 (citations to McClain, supra, omitted). Accord Chapman v. Procter & Gamble Distributing, LLC, 766 F.3d 1296, 1311 (11th Cir. 2014) (following Kilpatrick; one reason being that plaintiff’s expert “omitted consideration of idiopathic causes for [plaintiff’s condition], additionally rendering his differential diagnosis unreliable”).
An early post-Daubert case, Black v. Food Lion, Inc., 171 F.3d 308 (5th Cir. 1999), similarly affirmed exclusion of an opinion since:
neither [the expert] nor medical science knows the exact process that results in [the condition] or the factors that trigger the process. Absent these critical scientific predicates, . . . no scientifically reliable conclusion on causation can be drawn.
Id. at 314. In Bland v. Verizon Wireless, (VAW) LLC, 538 F.3d 893, 897 (8th Cir. 2008), held that an expert’s “attempt to use a differential diagnosis . . . fails because . . . the cause of [the condition] in the majority of cases is unknown.” Id. at 897. Likewise, in Tamraz v. Lincoln Electric Co., 620 F.3d 665, 675 (6th Cir. 2010), “the other possibility − unknown (idiopathic) causation − currently accounts for the vast majority of [the disease’s] cases, making it impossible to ignore and difficult to rule out . . . via a differential-diagnosis.” Id. at 675.
State, as well as federal, appellate courts have recognized the fundamental incompatibility between the circumstantial method of proof represented by differential etiology and the presence of significant unknown causal factors. In Blackwell v. Wyeth, 971 A.2d 235 (Md. 2009), for example, the court affirmed exclusion of an expert whose purported differential diagnosis ignored “unknown genetics,” holding that:
[The trial court] did not err in finding that a gene or series of interacting genes that have not yet been identified is the “most prevalent alleged cause of [the condition], based upon our review of the record. We agree that [plaintiff’s expert] did not sufficiently consider genetics in his differential diagnosis equation.
Id. at 260. This failure to consider medically unknown causes meant that the expert’s “theory is no more than hypothesis and conjecture.” Id. at 261.
Likewise, Valentine v. PPG Industries, Inc., 821 N.E.2d 580 (Ohio App. 2004), recognized that “[t]o state that nothing else caused the [plaintiff’s injury] is contrary to the medical and scientific fact that the cause of [that injury] is unknown.” Id. at 599.
At this point, medical science does not enable physicians and other scientists to pinpoint a cause. . . . Thus, under the circumstances of this case . . . differential diagnosis is not a reliable technique for identifying causation. . . . [T]he present state of scientific knowledge on the cause of [the disease] precludes reliability in this context.
Id. at 599-600. The Ohio Supreme Court affirmed, holding that “[a]lthough differential diagnosis is a standard scientific method for determining causation, its use is appropriate only when considering potential causes that are scientifically known.” Valentine v. Conrad, 850 N.E.2d 683, 688 (Ohio 2006) (citation omitted). See Blanchard v. Goodyear Tire & Rubber, 30 A.3d 1271, 1277 (Vt. 2011) (“a large percentage of cases of plaintiff’s type . . . are of unknown origin, so any attempt to establish causation by ruling out other causes must fail”); Garcia v. City of New Orleans, 115 So.3d 515, 519 (La. App. 2013) (by admitting “that it is possible that the [condition] was idiopathic,” plaintiff’s expert “did not perform a differential diagnosis”); Kerns v. Hobart Brothers Co., 2008 WL 1991909, at *8 (Ohio App. May 9, 2008) (following Valentine; “It is axiomatic that one cannot rule out causes that are unknown or unable to be properly discerned and isolated.”); Coastal Tankships, U.S.A., Inc. v. Anderson, 87 S.W.3d 591, 614 (Tex. App. 2002) (“in this case there are both unknown and ubiquitous causes, both of which it is impossible to ‘rule out’”). Cf. Johnson & Johnson Talcum Powder Cases, 249 Cal. Rptr.3d 642, 676 (Cal. App. 2019) (recognizing that “a differential diagnosis alone may be insufficient as the sole basis for an opinion on the etiology of a largely idiopathic disease”; holding “that is not the situation before us”).
Lots of trial-level cases reach the same result. In Soldo v. Sandoz Pharmaceuticals Corp., unknown causes were fatal to the plaintiff’s attempted differential etiology:
[G]iven plaintiff’s experts’ admissions that many [incidents] occur for which a particular cause cannot be ascertained even after extensive investigation, consistent application of their own methodology requires them to rule out such idiopathic [incidents] before reliably concluding that [the drug] caused [this incident]. It is impossible to reasonably rule out a cause that cannot even be specifically identified.
Id. at 517. See G v. Fay School, Inc., 282 F. Supp.3d 381, 391 (D. Mass. 2017) (plaintiff’s expert’s “use of the differential diagnosis method failed to reasonably survey other potential causes” because she “neglects to account for the possibility of an idiopathic etiology”), aff’d, 931 F.3d 1 (1st Cir. 2019); Jones v. Novartis Pharmaceuticals Corp., 235 F. Supp.3d 1244, 1280 (N.D. Ala. 2017) (“However, an unreliable application of a background risk methodology leads to the same result as a failure to consider the background risk at all: the expert’s opinion will be excluded.”), aff’d, 720 F. Appx. 1006 (11th Cir. 2018); K.E. v. GlaxoSmithKline LLC, 2017 WL 440242, at *16 (D. Conn. Feb. 1, 2017) (“Since both parties agree that [plaintiff’s condition] can be caused by a variety of factors and that the majority . . . have an unknown cause, these omissions further reveal the inadequacy of the differential diagnosis performed here.”); McCarty v. Arch Wood Protection, Inc., 2016 WL 1306067, at *7 (E.D. Ky. March 31, 2016) (“Plaintiffs’ experts did not perform a reliable differential diagnosis because they did not rule out an idiopathic origin”); Hendrian v. Safety-Kleen Systems, Inc., 2014 WL 1464462, at *7 (E.D. Mich. April 15, 2014) (“Plaintiff makes no attempt at explaining how, if at all, [her expert] ‘ruled out’ all potential alternative causes, including, most importantly . . ., idiopathic origin”); In re Denture Cream Products Liability Litigation, 795 F. Supp.2d 1345, 1367 (S.D. Fla. 2011) (plaintiffs’ expert “did not perform a reliable differential diagnosis” because he “did not consider the possibility of an idiopathic cause”); Nozinich v. Johnson & Johnson, Inc., 2011 WL 13124085, at *10 (W.D. Tenn. July 6, 2011) (due to “idiopathic causes,” the plaintiff’s expert’s “differential diagnosis was flawed”); Pritchard v. Dow Agro Sciences, 705 F. Supp.2d 471, 492 (W.D. Pa. 2010) (excluding expert who “did not attempt to rule out that [plaintiff’s condition] is the result of an idiopathic or unknown cause despite the fact that the general cause of [that condition] is unknown”), aff’d, 430 F. Appx. 102 (3d Cir. 2011); Henricksen v. Conoco Phillips Co., 605 F. Supp.2d 1142, 1162 (E.D. Wash. 2009) (“the presence of a known risk factor is not a sufficient basis for ruling out idiopathic origin in a particular case, particularly where most cases of the disease have no known cause. [I]n [these] cases, analysis beyond a differential diagnosis is required.”); Perry v. Novartis Pharmaceuticals Corp., 564 F. Supp.2d 452, 470 (E.D. Pa. 2008) (“Standing alone, the presence of a known risk factor is not a sufficient basis for ruling out idiopathic origin in a particular case, particularly where most cases of the disease have no known cause.”); Redfoot v. B.F. Ascher & Co., 2007 WL 1593239, at *11 (N.D. Cal. June 1, 2007) (“differential diagnosis is faulty because [the expert] failed to consider . . . that the cause of [the disease] is not known today”); Doe 2 v. OrthoClinical Diagnostics, Inc., 440 F. Supp.2d 465, 477-78 (M.D.N.C. 2006) (“Although [the expert] apparently has considered a number of specific genetic disorders in performing his differential diagnosis, the Court finds that his failure to take into account the existence of such a strong likelihood of a currently unknown genetic cause . . . serves to negate [his] use of the differential diagnosis technique as being proper in this instance.”); Cano v. Everest Minerals Corp., 362 F. Supp.2d 814, 846 (W.D. Tex. 2005) (“given that [plaintiff’s expert] does not appear to consider . . . unknown origins, it is possible that none of the possible causes he lists was an actual cause”; opinion excluded) (footnote omitted); Nelson v. American Home Products Corp., 92 F. Supp.2d 954, 971 (W.D. Mo. 2000) (“With no other basis for ruling out idiopathic [disease], [plaintiff’s expert’s] differential diagnosis does not constitute reliable scientific knowledge.”); Whiting v. Boston Edison Co., 891 F. Supp. 12, 21 n.41 (D. Mass. 1995) (“If 90 percent of the causes of a disease are unknown, it is impossible to eliminate an unknown disease as the efficient cause of a patient’s illness.”); Zandi v. Wyeth, 2007 WL 3224242 (Minn. Dist. 2007) (plaintiff’s condition “does not lend itself to differential diagnosis because the scientific community has not accepted that [it]has a limited number of discrete and recognized possible causes such that ruling out one cause would implicate another”).
Finally, since we’ve discussed Wendell v. GlaxoSmithKline LLC, 858 F.3d 1227, 1237 (9th Cir. 2017), before, and included it in our bottom ten decisions of the year, we would be remiss not to point out that the Ninth Circuit has gone its unique, science-ignoring way on this issue:
Finally, the district court erred when it excluded Plaintiffs’ experts’ opinion testimony because of the high rate of idiopathic [occurrence] and the alleged inability of the experts to rule out an idiopathic origin. . . . We do not require experts to eliminate all other possible causes of a condition for the expert’s testimony to be reliable. It is enough that the proposed cause “be a substantial causative factor.” This is true in patients with multiple risk factors, and analogously, in cases where there is a high rate of idiopathy.
Id. at 1237 (citations omitted). That’s the sum total of Wendell’s reasoning – ignoring unknown causes is OK “by analogy” to cases with multiple known risk factors. That’s pretty thin gruel to reverse a discretionary Daubert ruling, we think, but Wendell is out there, so anyone arguing this issue needs to be aware of Wendell and deal with its flabby rationale.
But on the whole, “[a] rose by another name may smell as sweet − but simply calling an analysis a differential diagnosis doesn’t make it so.” In re Lipitor (Atorvastatin Calcium) Marketing, Sales Practices & Products Liability Litigation (No II) MDL 2502, 892 F.3d 624, 643 (4th Cir. 2018). One way that commonly happens is experts ignoring an elephant in the room – a significant possibility that the disease or condition at issue arose from unknown causes. Since differential etiology is simply a form of circumstantial evidence – ruling out all other significant possible causes – any significant likelihood of idiopathic causation blows a fatal hole in this methodology.