TORTINI

For your delectation and delight, desultory dicta on the law of delicts.

Archives »

Federal Rule of Evidence 106 Applied to Statistical Evidence

June 18th, 2013

Watching your expert witness harassed by an unfair crossexamination, without protection from the trial judge, is one of the most frustrating, annoying, and enraging events in trial.  Not so long ago, I had the experience of having a lawyer put a chart together during his crossexamination of an expert witness. The crossexaminer took the upper bound of each confidence interval, and asked “Based upon this study, the true relative risk could be as high as X?”  When the witness, naturally and properly pointed out that the study data were also compatible with a relative risk as low as Y, the lower bound of the confidence interval, the crossexaminer whined that the witness was not being responsive, with the trial court’s chiming in to curb the expert’s reasonable desire to have a complete, correct answer.

Watching events such as this crossexamination has led me to believe that there really should be a rule of completeness for statistical evidence.  If sampling is random and unbiased, a point estimate may be the best estimate of the true value of a mean or a proportion, but that estimate may be rather crummy if the random error is large.  Presenting the sample statistic without some idea of the standard error seems wrong, but is there a remedy?

Statistical evidence did not play much of a role in the development of common law evidence, but the law is concerned with the sort of contextual accuracy that was being abused by the crossexamination of the expert witness, above.  The Federal Rules provide for completeness in at least two contexts.  Federal Rule of Evidence 106 provides:

“Remainder of or Related Writings or Recorded Statements

If a party introduces all or part of a writing or recorded statement, an adverse party may require the introduction, at that time, of any other part — or any other writing or recorded statement — that in fairness ought to be considered at the same time.”

A similar rule governs depositions presented in court.  Federal Rule of Civil Procedure 32(a)(4).

As the Advisory Committee notes to Rule 106 point out, Rule 106 is based upon two basic considerations:

“The first is the misleading impression created by taking matters out of context. The second is the inadequacy of repair work when delayed to a point later in the trial. …  The rule does not in any way circumscribe the right of the adversary to develop the matter on cross-examination or as part of his own case.”

Rule 106, Advisory Committee Notes (internal citations omitted).

When “in fairness” should an omitted portion be considered contemporaneously?  Here is how the Seventh Circuit put the matter in the more mundane context of a writing:

“To determine whether a disputed portion is necessary, the district court considers whether

(1) it explains the admitted evidence,

(2) places the admitted evidence in context,

(3) avoids misleading the jury, and

(4) insures fair and impartial understanding of the evidence.”

United States v. Velasco, 953 F.2d 1467, 1475 (7th Cir.1992).

Presentation of a point estimate should be accompanied by a measure of its variability; “[w]henever possible, an estimate should be accompanied by its standard error.” Reference Manual on Scientific Evidence at 117-18 (2d ed. 2000).  Similarly, trial courts should be vigilant against permitting a party to use the upper or the lower bound of a confidence interval to paint a misleading picture of “what the evidence shows.”

 

Posted in Scientific Evidence | Comments Off on Federal Rule of Evidence 106 Applied to Statistical Evidence

Power in the Reference Manual for Scientific Evidence

June 15th, 2013

The Third Edition of the Reference Manual on Scientific Evidence (2011) [RMSE3ed] treats statistical power in three of its chapters, those on statistics, epidemiology, and medical testimony.  Unfortunately, the treatment is not always consistent.

The chapter on statistics has been consistently among the best and most frequently ignored content of the three editions of the Reference Manual.  The most recent edition offers a good introduction to basic concepts of sampling, random variability, significance testing, and confidence intervals.  David H. Kaye & David A. Freedman, “Reference Guide on Statistics,” in RMSE3ed 209 (2011).  Kaye and Freedman provide an acceptable non-technical definition of statistical power:

“More precisely, power is the probability of rejecting the null hypothesis when the alternative hypothesis … is right. Typically, this probability will depend on the values of unknown parameters, as well as the preset significance level α. The power can be computed for any value of α and any choice of parameters satisfying the alternative hypothesis. … Frequentist hypothesis testing keeps the risk of a false positive to a specified level (such as α = 5%) and then tries to maximize power. Statisticians usually denote power by the Greek letter beta (β). However, some authors use β to denote the probability of accepting the null hypothesis when the alternative hypothesis is true; this usage is fairly standard in epidemiology. Accepting the null hypothesis when the alternative holds true is a false negative (also called a Type II error, a missed signal, or a false acceptance of the null hypothesis).”

Id. at 254 n.106

The definition is not, however, without problems.  First, it introduces a nomenclature issue likely to be confusing for judges and lawyers. Kaye and Freeman use β to denote statistical power, but they acknowledge that epidemiologists use β to denote the probability of a Type II error.  And indeed, both the chapters on epidemiology and medical testimony use β to reference Type II error rate, and thus denote power as the complement of β, or (1- β).  See Michael D. Green, D. Michal Freedman, and Leon Gordis, “Reference Guide on Epidemiology,” in RMSE3ed 549, 582, 626 ; John B. Wong, Lawrence O. Gostin, and Oscar A. Cabrera, Abogado, “Reference Guide on Medical Testimony,” in RMSE3ed 687, 724.  This confusion in nomenclature is regrettable, given the difficulty many lawyers and judges seem have in following discussions of statistical concepts.

Second, the reason for introducing the confusion about β is doubtful.  Kaye and Freeman suggest that statisticians usually denote power by β, but they offer no citations.  A quick review (not necessarily complete or even a random sample) suggests that many modern statistics texts denote power as (1- β).  See, e.g., Richard D. De Veaux, Paul F. Velleman, and David E. Bock, Intro Stats 545-48 (3d ed. 2012); Rand R. Wilcox, Fundamentals of Modern Statistical Methods 65 (2d ed. 2010).  At the end of the day, there really is no reason for the conflicting nomenclature and the likely confusion it engenders.  Indeed, the duplicative handling of statistical power, and other concepts, suggests that it is time to eliminate the repetitive discussions, in favor of one, clear, thorough discussion in the statistics chapter.

Third, Kaye and Freeman problematically refer to β as the probability of accepting the null hypothesis when elsewhere they more carefully instruct that a non-significant finding results in not rejecting the null hypothesis as opposed to accepting the null.  Id. at 253.  See also Daniel Rubinfeld, “Reference Guide on Multiple Regression,“ in RMSE3d 303, 321 (describing a p-value > 5% as leading to failing to reject the null hypothesis).

Fourth, Kaye and Freedman’s discussion of power, unlike most of their chapter, offers advice that is controversial and unclear:

“On the other hand, when studies have a good chance of detecting a meaningful association, failure to obtain significance can be persuasive evidence that there is nothing much to be found.”

RMSE3d at 254. Note that the authors leave open what a legal or clinically meaningful association is, and thus offer no real guidance to judges on how to evaluate power after data are collected and analyzed.  As Professor Sander Greenland has argued, in legal contexts, this reliance upon observed power (as opposed to power as a guide in determining appropriate sample size in the planning stages of a study) is arbitrary and “unsalvageable as an analytic tool.”  See Sander Greenland, “Nonsignificance Plus High Power Does Not Imply Support Over the Alternative,” 22 Ann. Epidemiol. 364, 364 (2012).

The chapter on epidemiology offers similar controversial advice on the use of power:

“When a study fails to find a statistically significant association, an important question is whether the result tends to exonerate the agent’s toxicity or is essentially inconclusive with regard to toxicity.93 The concept of power can be helpful in evaluating whether a study’s outcome is exonerative or inconclusive.94  The power of a study is the probability of finding a statistically significant association of a given magnitude (if it exists) in light of the sample sizes used in the study. The power of a study depends on several factors: the sample size; the level of alpha (or statistical significance) specified; the background incidence of disease; and the specified relative risk that the researcher would like to detect.95  Power curves can be constructed that show the likelihood of finding any given relative risk in light of these factors. Often, power curves are used in the design of a study to determine what size the study populations should be.96”

Michael D. Green, D. Michal Freedman, and Leon Gordis, “Reference Guide on Epidemiology,” RMSE3ed 549, 582.  Although the authors correctly emphasize the need to specify an alternative hypothesis, their discussion and advice are empty of how that alternative should be selected in legal contexts.  The suggestion that power curves can be constructed is, of course, true, but irrelevant unless courts know where on the power curve they should be looking.  The authors are correct that power is used to determine adequate sample size under specified conditions, but again, the use of power curves in this setting is today rather uncommon.  Investigators select a level of power corresponding to an acceptable Type II error rate, and an alternative hypothesis that would be clinically meaningful for their research, in order to determine their sample size. Translating clinical into legal meaningfulness is not always straightforward.

In a footnote, the authors of the epidemiology chapter note that Professor Rothman has been “one of the leaders in advocating the use of confidence intervals and rejecting strict significance testing.” RMSE3d at 579 n.88.  What the chapter fails, however, to mention is that Rothman has also been outspoken in rejecting post-hoc power calculations that the epidemiology chapter seems to invite:

“Standard statistical advice states that when the data indicate a lack of significance, it is important to consider the power of the study to detect as significant a specific alternative hypothesis. The power of a test, however, is only an indirect indicator of precision, and it requires an assumption about the magnitude of the effect. In planning a study, it is reasonable to make conjectures about the magnitude of an effect to compute study-size requirements or power. In analyzing data, however, it is always preferable to use the information in the data about the effect to estimate it directly, rather than to speculate about it with study-size or power calculations (Smith and Bates, 1992; Goodman and Berlin, 1994; Hoening and Heisey, 2001). Confidence limits and (even more so) P-value functions convey much more of the essential information by indicating the range of values that are reasonably compatible with the observations (albeit at a somewhat arbitrary alpha level), assuming the statistical model is correct. They can also show that the data do not contain the information necessary for reassurance about an absence of effect.”

Kenneth Rothman, Sander Greenland, and Timothy Lash, Modern Epidemiology 160 (3d ed. 2008).  See also Kenneth J. Rothman, “Significance Questing,” 105 Ann. Intern. Med. 445, 446 (1986) (“[Simon] rightly dismisses calculations of power as a weak substitute for confidence intervals, because power calculations address only the qualitative issue of statistical significance and do not take account of the results already in hand.”)

The selective, incomplete scholarship of the epidemiology chapter on the issue of statistical power is not only unfortunate, but it distorts the authors’ evaluation of the sparse case law on the issue of power.  For instance, they note:

“Even when a study or body of studies tends to exonerate an agent, that does not establish that the agent is absolutely safe. See Cooley v. Lincoln Elec. Co., 693 F. Supp. 2d 767 (N.D. Ohio 2010).  Epidemiology is not able to provide such evidence.”

RMSE3d at 582 n.93; id. at 582 n.94 (“Thus, in Smith v. Wyeth-Ayerst Labs. Co., 278 F.Supp. 2d 684, 693 (W.D.N.C. 2003), and Cooley v. Lincoln Electric Co., 693 F. Supp. 2d 767, 773 (N.D. Ohio 2010), the courts recognized that the power of a study was critical to assessing whether the failure of the study to find a statistically significant association was exonerative of the agent or inconclusive.”)

Here Green, Freedman, and Gordis shift the burden to the defendant and make the burden one of absolute certainty in the product’s safety.  This is not a legal standard. The cases they cite amplify the error. In Cooley, for instance, the defense expert would have opined that welding fume exposure did not cause parkinsonism or Parkinson’s disease.  Although the expert had not conducted a meta-analysis, he had reviewed the confidence intervals around the point estimates of the available studies.  Many of the point estimates were at or below 1.0, and in some cases, the upper bound of the confidence interval excluded 1.0.  The trial court expressed its concern that the expert witness had inferred “evidence of absence” from “absence of evidence.”  Cooley v. Lincoln Elec. Co., 693 F. Supp. 2d 767, 773 (N.D. Ohio 2010).  This concern, however, was misguided given that many studies had tested the claimed association, and that virtually every case-control and cohort study had found risk ratios at or below 1.0, or very close to 1.0.  What the court in Cooley, and the authors of the epidemiology chapter in the RSME3d have lost sight of, is that when the hypothesis is repeatedly tested, with failure to reject the null hypothesis, and with point estimates at or very close to 1.0, and with narrow confidence intervals, then the claimed association is probably incorrect.  See, e.g., Anthony J. Swerdlow, Maria Feychting, Adele C. Green, Leeka Kheifets, David A. Savitz, International Commission for Non-Ionizing Radiation Protection Standing Committee on Epidemiology, “Mobile Phones, Brain Tumors, and the Interphone Study: Where Are We Now?” 119 Envt’l Health Persp. 1534, 1534 (2011) (“Although there remains some uncertainty, the trend in the accumulating evidence is increasingly against the hypothesis that mobile phone use can cause brain tumors in adults.”).

The Cooley court’s comments have some validity when applied to a single study, but not to the impressive body of exculpatory epidemiologic evidence that pertains to welding fume and Parkinson’s disease.  Shortly after the Cooley case was decided, a published meta-analysis of welding fume or manganese exposure demonstrated a reduced level of risk for Parkinson’s disease among persons occupationally exposed to welding fumes or manganese.  James Mortimer, Amy Borenstein, and Lorene Nelson, “Associations of welding and manganese exposure with Parkinson disease: Review and meta-analysis,” 79 Neurology 1174 (2012).

Posted in Expert Witnesses, statistical evidence | Comments Off on Power in the Reference Manual for Scientific Evidence

Improper Claims That Studies Lack Power, Made Without Specifying An Alternative Hypothesis

June 14th, 2013

The Misuse of Power in the Courts

A claim that a study has low power is meaningless unless both the alternative hypothesis and the level of significance are included in the statement.  See Sander Greenland, “Nonsignificance Plus High Power Does Not Imply Support Over the Alternative,” 22 Ann. Epidemiol. 364 (2012); Sander Greenland & Charles, Poole, “Problems in common interpretations of statistics in scientific articles, expert reports, and testimony,” 51 Jurimetrics J. 113, 121-22 (2011).

Power can always be assessed as low by selecting an alternative hypothesis sufficiently close to the null. A study, using risk ratios, which has high power against an alternative hypothesis of 2.0, may have very low power against an alternative of 1.1. Because risk ratios greater than two are often used to attribute specific causation, measuring power of a study against an alternative hypothesis of a doubling of risk might well be a reasonable approach in some cases.  For instance, in Miller v. Pfizer, 196 F. Supp. 2d 1062, 1079 (D. Kan. 2002), aff’d, 356 F. 3d 1326 (10th Cir.), cert. denied, 543 U.S. 917 (2004), the trial court’s Rule 706 expert witness calculated the power of a study to exceed 90% probability to detect a doubling of risk in a pooled analysis of suicidality in clinical trial data of an anti-depressant.  Report of John Concato, M.D., 2001 WL 1793169, *9 (D.Kan. 2001). Unless a court was willing to specify the level at which it would find the risk ratio unhelpful or not probative, such as a relative risk greater than two, power analyses of completed studies are not particularly useful.

Plaintiffs’ counsel rightly complain when defendants claim that a study with a statistically “non-significant” risk ratio greater than 1.0 has no probative value. Although random error (or bias and confounding) may account for the increased risk, the risk may be real. If studies consistently show an increased risk, even though all the studies have reported p-values > 5%, meta-analytic approaches may very well help rule out chance as a likely explanation for the increased risk. The complaint that a study, however, is underpowered, without more, does not help plaintiff establish an association; nor does the complaint establish that the study provides no useful information.

The power of a study depends upon several variables, including the size of the alternative hypothesis, the sample size, the expected value and its variance, and the acceptable level of probability for false-positive findings, which level is reflected in the pre-specified p-value, α, at which level the study’s findings would be interpreted as not likely consistent with the null hypothesis. The lower α is set, the lower will be the power of a test or a study, all other things being equal.  Similarly, moving from a two-tailed to a one-tailed test of significance will increase power.  Courts have acknowledged that both Type I and Type II errors, and the corresponding α and β, are important, but they have overlooked that Type II errors are usually less relevant to the litigation process. See, e.g., DeLuca v. Merrell Dow Pharmaceuticals, Inc., 911 F.2d 941, 948 (3d Cir. 1990).  A single study that failed to show a statistically significant difference in the outcome of interest does not support a conclusion that the outcome is not causally related to the exposure under study.  In products liability litigation, the parties are typically not assigned a burden of proving the absence of causation.

In the Avandia litigation, plaintiffs’ key claim is that the medication, an oral anti-diabetic, causes heart attacks, even though none of the several dozen clinical trials found a statistically significant increased risk. Plaintiffs’ expert witnesses argued that all the clinical trials of Avandia were “underpowered,” and thus the failure to find an increased risk was a Type II (false-negative) error that resulted from the small size of the clinical trials. The Avandia MDL court, considering Rule 702 challenges to plaintiffs’ expert witness opinions, accepted this argument:

“If the sample size is too small to adequately assess whether the substance is associated with the outcome of interest, statisticians say that the study lacks the power necessary to test the hypothesis. Plaintiffs’ experts argue, among other points, that the RCTs [randomized controlled trials] upon which GSK relies are all underpowered to study cardiac risks.”

In re Avandia Mktg., Sales Practices & Prods. Liab. Litig., 2011 WL 13576, at *2 (E.D. Pa. 2011) (emphasis in original). The Avandia MDL court failed to realize that the power argument was empty without a specification of an alternative hypothesis. For instance, in one of the larger trials of Avandia, the risk ratio for heart attack was a statistically non-significant 1.14, with a 95% confidence interval that spanned 0.80 to 1.63. P.D. Home, et al., Rosiglitazone Evaluated for Cardiovascular Outcomes in Oral Agent Combination Therapy for Type 2 Diabetes (RECORD), 373 Lancet  2125 (2009). This trial, standing alone, thus had excellent power against an alternative hypothesis that Avandia doubled the risk of heart attacks; such an alternative hypothesis would clearly be rejected based upon the RECORD trial. On the other hand, an alternative hypothesis of 1.2 would not be. The confidence interval, by giving a quantification of random error, conveys results reasonably compatible with the study estimate; the claim of “low power” against an unspecified alternative hypothesis, conveys nothing.

Last year, in a hormone therapy breast cancer case, the Eight Circuit confused power with β, and succumbed to plaintiff’s expert witness’s argument that he was justified in ignoring several large, well-conducted clinical trials and observational studies because they were “underpowered,” without specifying the alternative hypothesis he was using to make his claim:

“Statistical power is ‘the probability of rejecting the null hypothesis in a statistical test when a particular alternative hypothesis happens to be true’. Merriam–Webster Collegiate Dictionary 973 (11th ed. 2003). In other words, it is the probability of observing false negatives. Power analysis can be used to calculate the likelihood of accurately measuring a risk that manifests itself at a given frequency in the general population based on the sample size used in a particular study. Such an analysis is distinguishable from determining which study among several is the most reliable for evaluating whether a correlative or even a causal relationship exists between two variables.”

Kuhn v. Wyeth, Inc., 686 F.3d 618, 622 n.5 (8th Cir. 2012), rev’g, In re Prempro Prods. Liab. Litig., 765 F. Supp. 2d 1113 (W.D. Ark. 2011). The Kuhn court’s formulation, “in other words,” is incorrect.  Power is not the probability of observing false negatives; it is the probability of correctly rejecting the null in favor of a specified alternative hypothesis, at a specified level of significance probability.  The court’s further discussion of “accurately measuring” mischievously confuses one aspect of statistical power concerned with random variability, with study validity.  The 8th Circuit’s opinion never discusses or discloses what alternative hypothesis the plaintiff’s expert witness had in mind when disavowing certain studies as underpowered.  I suspect that none was ever provided, and that the judges missed the significance of the omission.  The courts would seem better off using the confidence intervals around point estimates to assess the statistical imprecision in the observed data, rather than improper power analyses that fail to specify a legally significant alternative hypothesis.

Posted in statistical evidence | Comments Off on Improper Claims That Studies Lack Power, Made Without Specifying An Alternative Hypothesis

Sympathy for the Diablo — Peppermint Barry vs Spicy Seafood Pasta

June 12th, 2013

Barry S. Levy is a physician, author, performer, and political activist.   Dr. Levy is a past president of the American Public Health Association, which conferred upon him its oldest and most prestigious award, the Sedgwick Memorial Medal, in 2005, for his outstanding achievements in public health.

Levy has been received other, less favorable notice from trial and appellate judges.  For instance, one federal judge found Levy engaged in a dubious enterprise to manufacture silicosis claims in Mississippi.  In re Silica Products Liability Litigation, 398 F. Supp. 2d 563, 611-16, 622 & n.100 (S.D. Texas 2005) (expressing particular disappointment with Dr. Barry Levy, who although not the worst offender of a bad lot of physicians, betrayed his “sterling credentials” in a questionable enterprise to manufacture diagnoses of silicosis for litigation). See also Schachtman, Silica Litigation: Screening, Scheming & Suing; Washington Legal Foundation Critical Legal Issues Working Paper Series No. 135 (Dec. 2005) (exploring the ethical and legal implications of the entrepreneurial litigation in which Levy and others were involved).

Unfortunately, In re Silica was not an isolated case.  See, e.g., Lofgren v. Motorola, Inc., 1998 WL 299925, No. CV 93-05521 (Ariz. Super. Ct., Maricopa Cty. June 1, 1998); Harman v. Lipari, N.J. L. Div. GLO-L-1375-95, Order of Nov. 3, 2000 (Tomasello, J.) (barring the use of Barry Levy in class action for medical monitoring damages); Castellow v. Chevron USA, 97 F.Supp. 2d 780, 793-95 (S.D. Tex. 2000); Knight v. Kirby Inland Marine Inc., 482 F.3d 347 (5th Cir. 2007); Watts v. Radiator Specialty Co., 990 So. 2d 143 (Miss. 2008); Aurand v. Norfolk So. Ry., 802 F.Supp.2d 950 (2011).

Now, once again, right here in River City, Dr. Levy has stepped out of bounds, and has been stripped of his mantle of testifying expert witness.  Mallozzi v. Ecosmart Technologies, Inc., 2013 WL 2415677, No. 11-CV-2884 (SJF)(ARL) (E.D.N.Y. May 31, 2013).

Over the course of a few minutes, plaintiff Velio Mallozi applied a “couple of ounces” of EcoSmart Organic Home Pest Control product to the foundation of his home, and a few squirts inside.  The product used, EcoSmart, contains plant oils, including one percent peppermint oil.  Id. at *1.  Mallozi did not experience any acute ill effects from his organic pest control efforts, and later went to dinner at an [unidentified] Italian restaurant, where he enjoyed a meal of “spaghetti with seafood.”  Id. at *10.  Perhaps seafood diablo, but Judge Feuerstein doesn’t say.

After the seafood spaghetti meal, Mallozi experienced a serious bout of laryngopharyngeal reflux (LPR).  With sympathy for the diablo, Mallozi sued only EcoSmart, on the claim that his inhaling some spray, with its 1% peppermint oil, caused his LPR.

The Mallozis sought out B.S. Levy to support the claimed relationship.  Levy did not disappoint.  He rendered a report that asserted that the few minutes of inhaling minor amounts of peppermint oil causes relaxation of the lower esophogeal sphincter, and consequently LPR.  Id. at *5.  Levy relied upon four articles to support this claim, but Judge Feuerstein refused to accept the claim of support, at face value.  Her Honor reviewed each study, only to find that the exposures that were associated with relaxation of G.I. smooth muscle, for a short period of about 20 minutes, were trials of ingested peppermint oil, at significantly greater doses, over lengthy periods of observation.  Id. at *6.  The studies individually and collectively did not support Levy’s opinion.

One of the papers on Levy’s reliance list was a case report of a patient that suffered burns in her mouth from ingestion of pure peppermint oil. The court found that this case report, involving large doses of pure peppermint oil, with an outcome different from claimed by Mallozi, was irrelevant.  Id. Even if it were construed to offer some support, the anecdotal nature of the case report, lacking controls, renders the report an unreliable basis for a causal conclusion.  Id.

The trial court found that B.S. Levy’s analysis ignored the crucial roles of dose, duration, route of exposure, and the nature of plaintiff’s condition, LPR.   Id. at *8.  The court held that Levy’s opinion did not satisfy the requirements of Rule 702.  The plaintiffs failed to show that Levy’s opinion was supported sufficiently by facts or data, and they failed to show that his opinion was the product of applying reliable methods and principles.  Id.

Judge Feuerstein, having found that Levy’s opinion on general causation between EcoSmart and LPR was bereft of validity, could have stopped without addressing specific causation.  Such an approach would have had the virtue of judicial economy, but would have left out some delicious facts.  Levy purportedly used some sort of differential etiology to infer the cause of plaintiff’s LPR, but he omitted meaningful consideration of plaintiff’s having had a history of severe severe gastroesophageal reflux disease (GERD), which predated his bout of LPR.  Indeed, plaintiff’s GERD had been so severe that he had been previously hospitalized for GERD after ingesting coffee and donuts.   Judge Feuerstein found Levy’s iterative disjunctive syllogism invalid for having failed to address the prior history of GERD, and the intervening role of the pasta and seafood dinner, in bringing on the LPR.  Id. at *10-11.

The Material Safety Data Sheet Ploy

Judge Feuerstein also demonstrated a careful understanding of the meaning and role of the Material Data Safety Sheet (MSDS).  EcoSmart’s MSDS contained a warning that some sensitive persons may experience dizziness or irritation of their nasal passages, and that ingestion may cause GI irritation.  Id. at *13.  Dr. Levy had embraced this MSDS language as an “admission,” but he did not analyze the sources of information or their validity.  The MSDS warning, as required by law, addresses the potential hazard of the ingredient, peppermint oil, irrespective of dose, concentration, or route of administration.  Id., citing Moore v. Ashland Chem. Inc., 151 F.3d 269, 278 (5th Cir.1998); Turner v. Iowa Fire Equip. Co., 229 F.3d 1202, 1209 (8th Cir.2000); Ingram v. Solkatronic Chem., Inc., No. 04–CV–0287, 2005 WL 3544244, at *6 (N.D.Okla. Dec. 28, 2005)).

The Treating Physician Echo

Apparently one of Mr. Mallozzi’s treating physicians joined the fray with echoes of Levy’s opinion.  Judge Feuerstein recognized that treating physicians are subject to the requirements of Rule 702, and that a me-too opinion deserved the me-too result.  Id. at *13, citing Davids v. Novartis Pharm. Corp., 857 F.Supp. 2d 267, 280 (E.D.N.Y. 2012).

Pasta and Peppermints

Bad sense, innocence, cripplin’ my mind.
Old medical records I just can’t find.
Too much Cajun spice, and I forgot about GERD.
Incense and peppermints, haven’t you heard?
But who cares, it’s just litigation.
Lots to win, but nothing to lose.

Incense and peppermints, meaningless claims.
Turn on, drop in, toxins are all the same

Throw your pride to the side; it’s what you must do.

Daubert is politics; 702.
A yardstick for lunatics, your point of view.
The court cares what games you choose.
You have lots to win, but nothing to lose

(adapted from, and with apologies to, Strawberry Alarm Clock)

Posted in Causation, Expert Witnesses, Rule 702, Scientific Evidence | Comments Off on Sympathy for the Diablo — Peppermint Barry vs Spicy Seafood Pasta

State of the Art — Tasing in the Dark

June 11th, 2013

In American product liability litigation, there are two causation requirements:  warnings causation and factual causation.  The law imposes a duty to warn of risks, and the alleged failure to warn must have some connection with the plaintiff’s use of the product in such a way as to have caused him harm.  If an adequate warning, whatever that is, would have protected plaintiff from the harm, then there is the requisite nexus between the alleged failure to warn and plaintiff’s harm.

The product must also, however, be a factual cause of the harm.  For instance, a really scary warning that promised death and destruction might have kept the plaintiff from using the product, but still, the product itself must have caused the harm, for the plaintiff to satisfy the factual causation requirement.

The history of the last 50 years or so of American products liability law has been a narrative of unending assault on causation, both with respect to warnings causation, and to medical or factual causation.  Plaintiffs have sought to dilute, eliminate, create presumptions for, and shift the burden of proof of, both causation requirements.

The duty to warn arises, understandably, when the seller or manufacturer has knowledge that the product can cause harm, and the buyer lacks this knowledge.  For various reasons, sloppiness, ideology, or advocacy, some writers neglect the requirement that the buyer must lack the knowledge of the danger.  Quod vanum et inutile est, lex non requirit.  The law does not require what is vain and useless.  This principle of equity and law still informs the law of product liability, through doctrines and rules such as sophisticated user and sophisticated intermediary.

The duty to warn was traditionally couched in terms of a duty to impart knowledge, actual or constructive, or a connection between a foreseeable use of the product and a serious harm.  The plaintiffs’ bar’s war on causation has sought to lower the epistemic standard for both warning and factual causation.  There are too many cases to cite for the proposition that sellers must warn of potential risks as well as actually known or knowable risks.  This erosion would be harmless but for the similar erosion in standards of medical or factual causation, and the attempts to abandon meaningful standards of expert witness causation opinions.

It is thus refreshing to see the Ninth Circuit hold that California’s law of duty to warn applies to only to known or knowable risks.  Last year, the Court of Appeals held that a Taser manufacturer had no duty to warn of the risk that repeated shocks administered by its products could lead to death by inducing metabolic acidosis, given the state of the art at the time the product was manufactured. In Rosa v. Taser International, Inc., 684 F.3d 941 (9th Cir. 2012) , the Rosa’s son died after being “tased” by police officers.  The autopsy report attributed death to ventricular arrhythmia, caused by Michael Rosa’s having used methamphetamine, along with the police’s use of Tasers.  Expert witnesses postulated the mechanism of causation between tasing and death was metabolic acidosis, which consists of increased lactic acid levels, which in turn increased the risk of cardiac arrest.

The Rosas sued Taser International, Inc., the manufacturer of the “stun guns” on theories of strict liability and negligent failure to warn the purchasers, the police, of the dangers of the product.  (They did not sue the meth manufacturer or retailer.) The Rosas argued that Taser had a duty to warn of any risk, “no matter how unsubstantiated,” even if based upon speculative or conjectural statements, or inadequate studies.  Id. at 946.  The Court Appeals decisively rejected a duty to warn on this inconclusive evidentiary display.

The Ninth Circuit held that California law conditions the duty to warn on a high epistemic standard of knowledge, actual or constructive.  Id. at 946.  The manufacturer must warn of a “particular risk … known or knowable in light of the generally recognized and prevailing best scientific and medical knowledge available at the time of manufacture and distribution.” Id. (quoting from Conte v. Wyeth, Inc., 168 Cal.App.4th 89, 85 Cal.Rptr.3d 299, 310 (2008) (emphasis added by the Circuit Court).

Noting that the California courts have not articulated epistemological principles for knowledge of hazards that must find their way into warnings, the court, in Rosa, noted some accepted considerations.  First, ignorance is not a defense. ‘‘[A] manufacturer is held to the knowledge and skill of an expert in the field; it is obliged to keep abreast of any scientific discoveries and is presumed to know the results of all such advances.’’ Id. at 946 (citing Carlin v. Superior Court, 13 Cal.4th 1104, 56 Cal. Rptr.2d 162, 920 P.2d 1347, 1351 n. 3 (1996)).

Second, a manufacturer need not warn of “every report of a possible risk, no matter how speculative, conjectural, or tentative.” Id. (internal citations omitted).  A duty to warn of every possible risk would have the counter-productive result of diluting the effect of specific warnings of real risks, by “inundat[ing the public] indiscriminately with notice of any and every hint of danger.” Id.

The Rosas relied upon four peer-reviewed articles to argue that Taser had a duty to warn. Id. at 497. The Ninth Circuit looked past the fact of mere publication and peer review, to the actual nature and content of the relied-upon articles.  Some of the publications did not address stun guns, but two did, although in speculative ways.  See Raymond Fish & Leslie A. Geddes, “Effects of Stun Guns and Tasers,” 358 The Lancet 687 (2001).  The second publication was a Department of Defense report, which predated Michael Rosa’s death, but was not generally available until after his death.  John M. Kenny, et al., Human Effects Advisory Panel, Report of Findings Sticky Shocker Assessment (1999).  The court doubted whether such a report, unavailable to defendant, could be the predicate for defendant’s duty warn.  In any event, the content of the 1999 report was itself speculative, and merely concluded that  “deaths following Taser[] use may be due to acidosis.” Kenny at 31. The report did not state a causal conclusion, and it qualified its dataset as limited by small size.  Kenny at 32.

Taser added relevant warnings, six years after Michael Rosa’s death, and the Rosas sought to capitalize upon the warnings as admissions of some sort.  Id. at 948.  The Court of Appeals rebuffed this argument, not only on the basis of subsequent remedial measure, but also on the basis of “state of the art,” which requires that the evidence of the harm, which forms the basis for the duty to warn, be known or knowable at the time of marketing of the product.  Id.

In Rosa, the available evidence that might constitute “knowledge” of a harm shifted between the time of injury, and the time of litigation.  In latent disease litigation, the shift in evidence can be extreme, and hence the importance of “state of the art” as a limit to the duty to warn.

Can the evidence for a knowledge claim regress?  If what was in the medical literature at some earlier time turned out to be false later, then it would seem that it was never “knowledge” in the first place.  Still, plaintiffs would likely rely upon the older apparent knowledge to claim that it created a duty to warn.  For instance, in asbestos litigation, plaintiffs have claimed, since the late 1970s, that asbestos exposure causes gastrointestinal (GI) cancer.  The plaintiffs’ claims were supported  by Dr. Selikoff and his cadre of Mt. Sinai testifying physicians and scientists, and many textbooks and review articles stated, without qualification, that asbestos caused GI cancer.  In 2006, however, the Institutes of Medicine convened a committee to review non-pulmonary cancers and asbestos, and concluded that the studies, accumulated over 35 years since Dr. Selikoff’s ipse dixit, did not support a conclusion that asbestos causes colorectal cancer.  Institute of Medicine of the National Academies, Asbestos: Selected Health Effects (2006).  If the claim is unsupported in 2006, it was not established at times past, despite the assertions of plaintiffs and their partisan expert witnesses in litigation, and in publications.

Another example of the uneven path of knowledge is the claim that silica causes lung cancer.  Early writers suggested this hypothesis on the basis of flawed evidence, but the claim came to be rejected when astute authors realized that the lung cancer mortality experience of uranium miners was determined by their exposure to ionizing radiation, not to silica.  Starting in the 1980s, however, a group of advocate scientists started to press for recognition of silica itself as a lung carcinogen, and these scientists narrowly succeeded, by a deeply divided vote, in an IARC working group vote in 1996.  This IARC classification put crystalline silica into category I, but the rationale for this classification was soon thereafter shown to be lacking.  See Patrick A. Hessel, John F. Gamble, J. Bernard L. Gee, Graham Gibbs, Francis H.Y. Green, W. Keith C. Morgan, and Brooke T. Mossman, “Silica, Silicosis, and Lung Cancer: A Response to a Recent Working Group Report,” 42 J. Occup. Envt’l Med. 704 (2000).  A seller of crystalline silica should, under the Rosa case, be well within its rights not to warn of a risk of lung cancer.  The reality is that the litigation system is way too uncertain and even capricious, leading most companies to have warned of lung cancer “risk” from the mid-1980s, forward.

What is encouraging about Rosa is its implicit recognition that the Rule 702 factors for reliable scientific opinion are very much in play when evaluating historical scientific opinion, which is asserted to be the basis for a duty to warn.  The more typical “state of the art” decisions simply pass over validity and reliability of historical statements and conclusions, with a grand gesture that everything published serves to provide “notice” of a potential hazard.  This usual approach begs the question, however, whether there was a sound basis for a conclusion that the product was known to cause the harm alleged in the pending case.  The scientific nature of the historical determination may mean that historians, lacking in scientific credentials, will be inappropriate witnesses to support historical claims of knowledge of risks.  The Rosa court’s insistence upon historical study and conclusion validity and correctness requires a careful review of the historical epistemic claim in view of the full historical and contemporary record.

In the Rosa case itself, the claim of failure to warn failed as a result of a fair analysis of the extant evidence for the claim at the time of marketing of the Taser. The post-marketing shift in the epistemic basis for the failure-to-warn claim was irrelevant to the Rosas’ case.  Perhaps the warnings vel non on the methamphetamine used by Michael Rosa would have been a better predicate for the lawsuit?

Posted in Historian Testimony, Rule 702 | Comments Off on State of the Art — Tasing in the Dark

The Mt. Sinai Catechism

June 7th, 2013

There are mythologies in science as there in religion.  Back in the day, Dr. Irving Selikoff invented a catchy catechism about synergistic interaction between asbestos and smoking in producing lung cancer.  Selikoff’s study of insulators advanced lung cancer point estimates for asbestos alone, for smoking alone, and for asbestos and smoking together:  5-10-50, which became incorporated in innumerable textbooks, articles, expert witness reports, and court opinions.

E. Cuyler Hammond, Irving J. Selikoff,  and Herbert Seidman, “Asbestos Exposure, Cigarette Smoking and Death Rates,” 330 Ann. N.Y. Acad. Sci. 473, 487 (1979).  Selikoff tells us that each insulator, starting in 1961, answered “a detailed questionnaire,” and later, repeat questionnaires on their smoking histories.  Id. at 475.

“Of the 8220 men who answered the questionnaire, 891 (11%) said that they had never smoked regularly; 488 (6%) had a history of pipe or cigar smoking but never smoked cigarettes regularly and; 6841 (83%) were either current or ex-cigarette smokers. Some of the cigarette smokers also smoked pipes or cigars or had done so in the past.”

Id. at 481.

The 5 supposedly represented the relative risk of asbestos exposure alone in insulators who did not smoke; the 10 supposedly represented the relative risk of smoking alone in the non-insulator general population of smokers; and the 50 represented the relative risk of insulators who had smoked.  Selikoff claimed that his data supported a multiplicative model of synergistic interaction between smoking and asbestos exposure.

Like other mythologies, there was a good bit of exaggeration and fabrication in Selikoff’s story. First, the non-smoking relative risk was based upon the observed risk ratio among insulators who “never smoked regularly.”  Smoking data were collected by survey, and the actual questionnaires and definition have never been published or provided by Selikoff and his colleagues.  Even taken at face value, “never smoked regularly” is not the same as non-smoker.  Adding that the insulators union was widely engaged in personal injury litigation, and that the membership had a strong motive to downplay their smoking histories, this exposure information becomes suspect.

The non-asbestos smoking lung cancer point estimate (10) was equally dubious.  The ten-fold risk statistic represented an average smoking lung cancer risk in the first Cancer Prevention Survey (CPS-1).  The insulators who smoked were rarely average smokers, and the CPS-1 point estimate was superseded within a couple of years by the new, revised CPS-II overall smoking lung cancer point estimate that was greater than twenty.  The CPS-II results were published in the 1980s and early 1990s, but Selikoff and his protégés contained to testify to the 5-10-50. See, e.g., Shopland, “Smoking-Attributable Cancer Mortality in 1991: Is Lung Cancer Now the Leading Cause of Death Among Smokers in the United States?” 83 J. Nat’l Cancer Inst. 1142 (1991).

These problems and others led astute authors to note their dissent from the simplistic multiplicative model.  F. Douglas K. Liddell, “The Interaction of Asbestos and Smoking in Lung Cancer,” 45 Ann. Occup. Hyg. 341 (2001) (“Both cigarette smoke and inhaled asbestos fibres can cause lung cancer, but the assessment of how these agents act in combination is a matter of great difficulty.”)

Despite the serious problems with the multiplicative model, Selikoff’s wanton republication of the 5-10-50 data, and the popularity of these data with testifying expert witnesses and regulating agencies, cemented their use in litigation. Consider, for instance, how the orthodoxy infiltrated the ostensibly neutral Reference Manual on Scientific Evidence, which proclaimed the multiplicative synergy of smoking and asbestos in its second edition:

“Occupational asbestos exposure in nonsmokers increases the risk of lung cancer by a factor of about five, from about 11 per 100,000, for nonsmoking industrial workers not exposed to asbestos to about 58 per 100,000 for nonsmoking asbestos workers; a significant smoking history increases the rate of lung cancer by a factor of at least ten. See U.S. Surgeon Gen., U.S. Dep’t of Health & Human Servs., The Health Consequences of Smoking: Cancer and Chronic Lung Disease in the Workplace 216 (1985); see also Rodolfo Saracci, “The Interactions of Tobacco Smoking and Other Agents in Cancer Etiology,” 9 Epidemiologic Revs. 175, 176–80 (1987). Because the effects of smoking and asbestos are multiplicative for lung cancer, the population of smoking asbestos workers has a lung cancer incidence of 5 times 10, or 50 times the background rates, rather than the 15-fold increase predicted by adding the separate risks. See U.S. Surgeon Gen., U.S. Dep’t of Health & Human Servs., supra, at 216–17.”

Mary Sue Henifin, Howard M. Kipen, and Susan R. Poulter, “Reference Guide on Medical Testimony,” in Federal Judicial Center, Reference Manual on Scientific Evidence 439, 476 & n. 136 (2d ed. 2000).  Dr. Kipen was a protégé of Dr. Selikoff.  The reference to the multiplicative mantra was eliminated from the Third Edition of the Reference Manual.

Back in April of this year, some other Selikoff protégés were busy trying to rehabilitate the 5-10-50 mantra, with a new study. Science Daily earnestly reported news of the new study as though it were ground breaking.  See “Asbestos Exposure, Asbestosis, and Smoking Combined Greatly Increase Lung Cancer Risk” (Apr. 12, 2013)  (reporting on the in-press version of Steve Markowitz, Stephen Levin, Albert Miller, and Alfredo Morabia, “Asbestos, Asbestosis, Smoking and Lung Cancer: New Findings from the North American Insulator Cohort,” Am. J. Respir. & Critical Care Med. (2013)).  Interestingly, only Stephen Levin, who died over a year ago, is listed as affiliated with the Mt. Sinai School of Medicine.  Although some of the other authors had previous affiliations with Mt. Sinai, they are now listed at the Center for the Biology of Natural Systems, Queens College – CUNY.

Science Daily quotes lead author Markowitz as saying that “[t]he interactions between asbestos exposure, asbestosis and smoking, and their influence on lung cancer risk are incompletely understood.”  It seems unlikely that this new article will shed much light on the problem.   The authors claim to apply and extend Selikoff’s approach from previous publications.  This new publication thus incorporates the serious problems that had afflicted Selikoff’s papers, especially his use of CPS-I to provide a point estimate for smoking alone (the alleged 10 in the 5-10-50 theory).  Here is the abstract of the paper:

“Rationale

Asbestos, smoking, and asbestosis increase lung cancer risk in incompletely elucidated ways. Smoking cessation among asbestos-exposed cohorts has been little studied.

Objectives

To measure the contributions of asbestos exposure, asbestosis, smoking and their interactions to lung cancer risk in an asbestos-exposed cohort, and to describe their reduction in lung cancer risk when they stop smoking.

Methods

We examined lung cancer mortality obtained through the National Death Index for 1981-2008 for 2,377 male North American insulators for whom chest x-ray, spirometric, occupational and smoking data were collected in 1981- 1983 and for 54,243 non-asbestos exposed blue collar male workers from Cancer Prevention Study II for whom occupational and smoking data were collected in 1982.

Measurements and Main Results

Lung cancer caused 339 (19%) insulator deaths. Lung cancer mortality was increased by asbestos exposure among non-smokers [rate ratio = 3.6 (95% CI: 1.7-7.6)], by asbestosis among non-smokers [rate ratio = 7.40 (95% CI, 4.0-13.7], and by smoking without asbestos exposure [rate ratio = 10.3 (95% CI, 8.8-12.2)]. The joint effect of smoking and asbestos alone was additive [rate ratio = 14.4 (95% CI, 10.7-19.4)] and with asbestosis, supra-additive [rate ratio = 36.8 (95% CI, 30.1-46.0)]. Insulator lung cancer mortality halved within 10 years of smoking cessation and converged with that of never-smokers 30 years following smoking cessation.

Conclusions

Asbestos increases lung cancer mortality among non-smokers. Asbestosis further increases the lung cancer risk and, considered jointly with smoking, has a supra-additive effect. Insulators benefit greatly by quitting smoking.”

The big news that Steven Markowitz did not trumpet in the press releases was that in the absence of asbestosis, there was no multiplicative interaction.  The asbestos-alone (point estimate) is still suspect.  The paper characterizes this point estimate as applying to non-smokers without asbestosis, but given this group’s history of equivocating between “never smoked regularly” and “non-smoker,” and the difficulty in obtaining accurate smoking histories from men enmeshed in litigation over respiratory claims, the asbestos-alone point estimate remains doubtful.  Curiously, the smoking histories were collected in 1981-83, but there is no mention of any attempt to verify the histories against smoking histories provided by the study participants in earlier versions of Selikoff’s studies on the insulators.

Similarly, the smoking-alone point estimate is incredible.  The authors purport to draw the smoking-alone risk ratio from CPS-II; yet the risk ratio presented (10.3) is less than half that reported in multiple publications of the CPS-II study.  Markowitz and colleagues cite to two papers for the 10.3 point estimate; neither appears to support this number.  Steven Stellman & Lawrence, “Smoking habits and tar levels in a new American Cancer Society prospective study of 1.2 million men and women,” 76 J. Nat’l Cancer Inst. 1057 (1986); Michael Thun, Eugenia Calle, Carmen Rodriguez, and Phyllis Wingo, “Epidemiological research at the American Cancer Society. 9 Cancer Epidemiol. Biomarkers & Prevention 861 (2000).  I would welcome contradiction from anyone who can find the point estimate in the cited papers.

Consider one of the most recent publications, which drew upon CPS-II data, in the New England Journal of Medicine.  For men who were current smokers, compared with men who never smoked, the relative risk of lung cancer mortality was 24.97. Michael Thun, Brian Carter, Diane Feskanich, Neal D. Freedman, Ross Prentice, Alan D. Lopez, Patricia Hartge, and Susan M. Gapstur, “50-Year Trends in Smoking-Related Mortality in the United States,” 368 New Engl. J. Med. 351 (2013).

The Markowitz, Levin, Miller paper is still “in press.”  There is still time for the editors to take a closer look.  There is also time for the editors to insist upon a disclosure of the authors’ conflicts of interest. See, e.g., Wannall v. Honeywell Internat’l, Inc., Civ. Action No. 10-351 (BAH), 2013 U.S. Dist. LEXIS 68523 (D.D.C. May 14, 2013) (discussing Markowitz’s participation as an expert witness).

Posted in Causation, Conflicts of Interest, Expert Witnesses, Scientific Evidence | Comments Off on The Mt. Sinai Catechism

Irving Selikoff and the Right to Peaceful Dissembling

June 5th, 2013

Among concerned writers on corporate conflicts of interest, it is a commonplace that industrial sponsors of epidemiologic and other research selectively publish studies favorable to their positions in litigation and regulatory controversies.  In my experience, most companies are fairly scrupulous about publishing the studies they funded.  If there is a correlation in industry funding and outcome, it is largely the result of corporate funding being directed to areas in which weak or corrupt politically motivated, public-interested scientists have already published studies with dubious results.  Common sense suggests that a fair test of the their claims will result in exonerative results.

It is also a commonplace that academic and public-spirited researchers will not have similar motives to suppress unfavorable results.  Again, in my experience, the opposite is true.  Consider that paragon of public-interested, political scientist, the late Dr. Irving Selikoff. During the course of discovery in the Caterinnichio case, I obtained manuscripts of two studies that Selikoff and his colleague, Bill Nicholson, prepared, but never published.  One study examined the mortality, and especially the cancer mortality, of workers at a Johns-Manville asbestos product manufacturing plant in New Jersey.  William J. Nicholson& Irving J. Selikoff, “Mortality experience of asbestos factory workers; effect of differing intensities of asbestos exposure”(circa 1988).

Selikoff’s failure to publish this manuscript on the Manville plantworkers is curious given his tireless and repeated republication of data from his insulator cohort.  For those familiar with Selikoff’s agenda, the failure to publish this paper appears to have an obvious goal:  suppress the nature and extent of Johns Manville’s use of crocidolite asbestos in its products:

“[O]ther asbestos varieties (amosite, crocidolite, anthophyllite) were also used for some products. In general, chrysotile was used for textiles, roofing materials, asbestos cements, brake and friction products, fillers for plastics, etc.; chrysotile with or without amosite for insulation materials; chrysotile and crocidolite for a variety of asbestos cement products.”

Id.  The suppression of studies obviously takes place outside the world of commercial or industrial interests.  See “Selikoff and the Mystery of the Disappearing Amphiboles.”

There was yet another studied never published by Selikoff, his work, again with Bill Nicholson, on the mortality of shipyard workers at the Electric Boat Company, in Groton, Connecticut. Irving Selikoff & William Nicholson, “Mortality Experience of 1,918 Employees of the Electric Boat Company, Groton, Connecticut January 1, 1967 – June 30, 1978” (Jan. 27, 1984) [cited below as Electric Boat].

Many of the asbestos cases that worked their way through the legal system in the 1980s and 1990s were filed by shipyard workers.  Most of these shipyard workers were not insulators, but claimed asbestos bystander exposure from work near insulators.  Invariably, the expert witnesses for these shipyard worker plaintiffs relied upon risk data from the Selikoff of asbestos insulators, even though Selikoff himself cautioned against using the insulator data for non-insulators:

“These particular figures apply to the particular groups of asbestos workers in this study.  The net synergistic effect would not have been the same if their smoking habits had been different; and it probably would have been different if their lapsed time from first exposure to asbestos dust had been different or if the amount of asbestos dust they had inhaled had been different.”

Selikoff, et al., “Asbestos Exposure, Cigarette Smoking and Death Rates,” 330 Ann. N.Y. Acad. Sci. at 487 (1979).

Having access to Selikoff’s shipyard worker data would have been extremely useful to the fact-finding process, because these data failed to support the cancer projections used by testifying expert witnesses.  Selikoff and Nicholson pointed out that about 50% of the Electric Boat shipyard workers had X-ray abnormalities  Electric Boat at 2. (This finding must be interpreted in the darkness of Selikoff’s documented propensity to overread chest X-rays for asbestos findings.  Rossiter, “Initial repeatability trials of the UICC/ Cincinnati classification of the radiographic appearances of pneumoconioses.” 29 Brit. J. Indus. Med. 407 (1972) (reporting IJS’s readings as among the most extreme outliers in a panel of pulmonary and radiology physicians; showing that IJS films were read as showing abnormal profusion of small, irregular densities up to twice as often as the most reliable readers in the study.)).

Selikoff’s unpublished Electric Boat study cautioned that the mortality data reflected short duration and latency, and that the full extent of asbestos-related manifestations had not been reached.  Electric Boat at 3.  This assertion was not really borne out by the data.  Selikoff’s paper reported the following observed and expected data for lung cancer:

Years from onset of employment 10-14 15-19 20-24 25-29 30+ TOTAL
OBSERVED 4 23.3 15 3 4 35
EXPECTED 1.3 17.7 8.1 4.7 5.1 25.9

The study is primitive even by then contemporary standards.  There is no control for smoking; and no data on smoking habits.  There is no data on radiation exposure. (Electric Boat built nuclear submarines.) No p-values or confidence intervals are supplied; nor are any estimates of trends included.

Despite Selikoff’s assertion that the follow-up period was not sufficiently long to capture asbestos-related malignancies, the data tell a different story.  The lung cancer Obs./Exp. ratios are increased for 10-14 years, and for 15-19 years, and so these risk ratios reflect that the cohort likely had non-asbestos-related risks for lung cancer, which risks are at work before the cohort entered the lagged period in which they might have elevated asbestos-related risks.  Although the numbers are smaller for the time intervals that involve more than 20 years from first employment, the observed numbers and risk ratios of lung cancers hardly suggests very much in terms of an occupational asbestos risk.

These data were obtained only because Bill Nicholson often served as an expert witness for plaintiffs in personal injury actions.  When he did so in New Jersey, he was subject to fairly broad discovery obligations, and thus I was able to obtain his unpublished studies.  Otherwise, the public and the scientific community learned only what Selikoff selectively disclosed in media interviews.  See Samuel G. Freedman, “Worker’s suit over asbestos at Groton shipyard to open” New York Times (Jan. 19, 1982) (noting the 50% prevalence finding, but not the mortality data).

Posted in Data Sharing, Scientific Evidence, Scientific Misconduct, Underlying Data | Comments Off on Irving Selikoff and the Right to Peaceful Dissembling

The Doubling Dose

June 5th, 2013

Sander Greenland and others have raised various theoretical objections to the argument that relative risks should exceed two before attribution can be made in specific cases.  In large part, Greenland’s objections turn on conjecture that risks are not stochastically distributed in the population samples studied in epidemiologic studies.  These objections are potentially true, but the burden remains on the proponent.  When there is serious evidence of latency or bimodal distribution, then epidemiologic studies or clinical trials can be adapted to examine the risk in the relevant subgroup.  When this is done, the relative risk argument once again is unavoidable.  For many exposures and conditions of interest, epidemiologic studies have evolved to accommodate the relevant model of risk distribution.

Despite Greenland’s speculative objections, the concept of a doubling dose has been widely advocated by scientists. The following are a few of such endorsements:

Philip Enterline, “Attributability in the Face of Uncertainty,” 78 (Supp.) Chest 377, 377, 378 (1980);

Otto Wong, “Using Epidemiology to Determine Causation in Disease,” 3 Natural Resources & Env’t 20, 23 (1988); Ben Armstrong, Claude Tremblay, and Gilles Theriault, “Compensating Bladder Cancer Victims Employed in Aluminum Reduction Plants,” 30 J. Occup. Med. 771 (1988);

Joshua Muscat & Michael Huncharek, “Causation and disease: Biomedical science in toxic tort litigation,” 31 J. Occup. Med. 997 (1989);

Troyen A. Brennan, “Can Epidemiologists Give Us Some Specific Advice?” 1 Courts, Health Science & the Law 397, 398 (1991)(“This indeterminancy complicates any case in which epidemiological evidence forms the basis for causation, especially when attributable fractions are lower than 50%.  In such cases, it is more probable than not that the individual has her illness as a result of unknown causes, rather than as a result of exposure to hazardous substance.”);

Mark R. Cullen & Linda Rosenstock, “Principles and Practice of Occupational and Environmental Medicine,” chap. 1, in Linda Rosenstock & Marc Cullen, eds., Textbook of Clinical Occupational and Environmental Medicine 1, 13-14 (Phila. 1994);

David F. Goldsmith & Susan G. Rose, “Establishing Causation with Epidemiology,” in Tee L. Guidotti & Susan G. Rose, eds., Science on the Witness Stand:  Evaluating Scientific Evidence in Law, Adjudication, and Policy 57, 60 (OEM Press 2001)(“A relative risk greater than 2.0 produces an attributable risk (sometimes called attributable risk percent10) or an attributable fraction that exceeds 50%.  An attributable risk greater than 50% also means that ‘it is more likely than not’, or, in other words, there is a greater than 50% probability that the exposure to the risk factor is associated with disease.”)

Below, I have updated once again the case law on the issue of using relative and attributable risks to satisfy plaintiff’s burden of showing, more likely than not, that an exposure or condition caused a plaintiff’s disease or injury.

See , for the updated the case law on the issue of using relative and attributable risks to satisfy plaintiff’s burden of showing, more likely than not, that an exposure or condition caused a plaintiff’s disease or injury.

 

Posted in Causation, Expert Witnesses, Scientific Evidence | Comments Off on The Doubling Dose

Landrigan v. The Celotex Corporation, Revisited

June 4th, 2013

Old-fashioned torts presented few problems for attributing causation of the plaintiff’s harm.  Summers v. Tice, 33 Cal.2d 80, 199 P.2d 1 (1948), may have involved uncertainty about the shooter, but there was no doubt that a pellet from one of the two defendants’ guns hit the plaintiff and caused a legally recognized injury.

Specific causation has been, and remains, the soft underbelly of the toxic tort world, at least for those cases not involving so-called signature diseases.  A priori assessments of risk do not necessarily translate into post-exposure, post-diagnosis attributions of outcome to exposure.  Put simply, risk is not cause. Guinn v AstraZeneca Pharms. LP, 602 F.3d 1245, 1255 (11th Cir. 2010) (“An expert, however, cannot merely conclude that all risk factors for a disease are substantial contributing factors in its development.  The fact that exposure to a substance may be a risk factor for a disease does not make it an actual cause simply because the disease developed.”) Unless there is a “fingerprint of causation,” what scientists would call a completely specific biomarker, then specific causation opinions are mostly guesswork.

Tobacco companies and others exploited this fact, in face of large relative risks of lung cancer among smokers, to maintain that these epidemiologic assessment were not probative of specific causation.  Andrew See, “Use of Human Epidemiology Studies in Proving Causation,” 67 Def. Couns. J. 478, 478 (2000) (“Epidemiology studies are relevant only to the issue of general causation and cannot establish whether an exposure or factor caused disease or injury in a specific individual.”); Melissa Moore Thomson, Causal Inference in Epidemiology: Implications for Toxic Tort Litigation, 71 N.C. L. Rev. 247, 255 (1992) (“statistic-based epidemiological study results should not be applied directly to establish the likelihood of causation in an individual plaintiff”); Michael Dore, Commentary on the Use of Epidemiological Evidence in Demonstrating Cause-in-Fact, 7 Harv. Envt’l L. Rev. 429, 433 (1983) (“Epidemiological evidence, like other generalized evidence, deals with categories of occurrences rather than particular individual occurrences. . . . Such evidence may help demonstrate that a particular event occurred, but only when accompanied by more specific evidence.”).  See, e.g., In re Fibreboard Corp.,893 F.2d 706, 712 (5th Cir.1990) (“It is evident that these statistical estimates deal only with general causation, for population-based probability estimates do not speak to a probability of causation in any one case; the estimate of relative risk is a property of the studied population, not of an individual’s case.” (emphasis in original; internal quotation omitted)).

Indeed, some courts continue to uphold this extreme anti-probabilistic view, even when relative risks exceed 20, or more.  McTear v. Imperial Tobacco Ltd., [2005] CSOH 69, at ¶ 6.180 (Nimmo Smith, L.J.) (“epidemiological evidence cannot be used to make statements about individual causation… . Epidemiology cannot provide information on the likelihood that an exposure produced an individual’s condition. The population attributable risk is a measure for populations only and does not imply a likelihood of disease occurrence within an individual, contingent upon that individual’s exposure.

In past posts, I have addressed some misunderstandings and misrepresentations concerning the use of a priori risk to assessment of specific causation.  One of the more glaring examples of bad scholarship in this area comes in a text edited by Professor Joseph Gastwirth:

“The court in Landrigan v. Celotex Corp. (1992: 1087) arrived at a similar conclusion, finding that:

a relative risk of 2.0 is not so much a password to a finding of causation as one piece of evidence, among others, for the court to consider in determining whether the expert has employed a sound methodology in reaching his or her conclusion.”

Accordingly the court granted recovery for injuries alleged to have arisen as the result of exposure to asbestos, although the demonstrated relative risk was 1.5.

Sana Loue, “Epidemiological Causation in the Legal Context: Substance and Procedures,” in Joseph Gastwirth, ed., Statistical Science in the Courtroom 263, 277 (2000).

Now that is stunningly bad scholarship, from someone who is both a lawyer and a scientist. The New Jersey Supreme Court, in the cited case, reversed a directed verdict for the defendants, and remanded for reconsideration of the admissibility of the plaintiffs’ expert witnesses.  There was never even an opportunity for the Supreme Court to “grant recoveries.”  Indeed, Mrs. Landrigan never obtained a favorable verdict in her lawsuit.  After remand, she dismissed her action in the face of the daunting task faced by her expert witnesses.

The author of the chapter, Sana Loue, is a Professor and Director in the Department of Epidemiology and Biostatistics in the School of Medicine of Case Western Reserve University, Cleveland, Ohio. Dr. Loue holds doctoral degrees in epidemiology and medical anthropology, as well as a law degree.

Dr. Loue is not alone in misunderstanding the Landrigan case. Some of the confusion perhaps results from the New Jersey Supreme Court’s errant opinion.  Some language in the Supreme Court’s decision makes it seem that there was an objection to the admissibility of the plaintiff’s expert witnesses’ opinions.  There was none.  Unlike many gatekeeping decisions, the plaintiff had a full opportunity to be heard; the defendants moved for a directed verdict at the end of the plaintiff’s case.  In addressing the defendants’ motion, the trial court assumed, for the sake of argument, that asbestos can cause colorectal cancer.  General causation was, of course, contested, but the motion turned on whether there was evidence in the record that would support specific causation.  The trial court held that specific causation required expert witness opinion testimony, but that the testimony in the case failed to provide a basis on which a reasonable jury could conclude that Mr. Landrigan’s colorectal cancer was caused by his alleged occupational asbestos exposure.

The New Jersey Appellate Division affirmed in a published opinion.  579 A.2d 1268 (1990).  The Appellate Division’s decision is still worth reading, not only because it correctly decided the issues, but because it reports material facts that the Supreme Court chose to ignore.  First, the Appellate Division noted that the most that Mr. Landrigan had sustained in terms of respiratory effects from his occupational asbestos exposure was pleural thickening, which never caused him impairment in his lifetime.  Indeed, Mr. Landrigan never was aware of this radiographic change, which only an expert witness hired by plaintiff’s counsel could see.  Id. at 1269.  (Plaintiff’ pulmonary physician expert witness, Dr. Sokolowski, had failed his B-reader examination, but he was a favorite of the asbestos plaintiffs’ bar for his “liberal” readings of chest films.)

The Appellate Division also emphasized the record evidence that the cause of most cases of colon cancer was (and remains) unknown, and more important that Mr. Landrigan’s colorectal cancer was physically indistinguishable from almost all other cases of the disease.  Id. at 1270.  The plaintiff’s hired expert witnesses had only epidemiologic evidence of an increased risk of colorectal cancer among asbestos-exposed workers. Although most of the better conducted studies fail to support the claim of increased risk, Drs. Sokolowski and Wagoner, the plaintiff’s witnesses, relied upon Selikoff’s cohort study of insulation workers, and its mortality risk ratios.  Irving J. Selikoff, E. Cuyler Hammond, and Herbert Seidman, “Mortality Experience of Insulation Workers in the United States and Canada, 1943-1976,” 330 Ann. N.Y. Acad. Sci. 91, 103 (1979) (colorectal cancer risk ratio 1.55);  E. Cuyler Hammond, Irving J. Selikoff,  and Herbert Seidman, “Asbestos Exposure, Cigarette Smoking and Death Rates,” 330 Ann. N.Y. Acad. Sci. 473, 480 (1979) (colorectal cancer mortality ratios 1.59 to 1.81).

Mrs. Landrigan’s witnesses both relied upon evidence of an increased risk, while ignoring or dismissing studies that found no such risk, and upon what they claimed was an absence of risk factors, such as fatty diet, excessive alcohol consumption, family history, and prior bowel disease, in Mr. Landrigan.  The trial court, and the Appellate Division, realized that the reasoning that these witnesses advanced failed to support their conclusions, as a matter of science, logic, and law:

“Although not stated by Dr. Sokolowski in so many words, he seems to be saying that risk exposure equates with causation, a proposition which we find legally untenable.”

579 A.2d at 1270 (1990).  The hand waving about ruling out known risk factors left the most likely cause in plain view:  unknown:

“One cannot rule out the presence of other risk factors without knowing what those factors may be.”

Id. at 1271.

The New Jersey Supreme Court reversed and remanded the case for further inquiries into the reliability of the expert witnesses’ opinions.  Landrigan v. The Celotex Corp., 127 N.J. 404, 605 A.2d 1079 (1992).  Therese Keeley, the capable lawyer who tried the Landrigan case for the defense, had argued the appeal before the Appellate Division, but another lawyer, less familiar with the issues, argued for the defendant, in the Supreme Court.  The Supreme Court made much of the new lawyer’s concessions in oral argument:

“Defense counsel urges that the Appellate Division opinion may be read as requiring that an expert may not rely on an epidemiological study to support a finding of individual causation unless the relative risk is greater than 2.0. See 243 N.J.Super. at 457-59, 579 A.2d 1268. At oral argument before us, they agreed that such a requirement may be unnecessary. Counsel acknowledged that under certain circumstances a study with a relative risk of less than 2.0 could support a finding of specific causation. Those circumstances would include, for example, individual clinical data, such as asbestos in or near the tumor or a documented history of extensive asbestos exposure. So viewed, a relative risk of 2.0 is not so much a password to a finding of causation as one piece of evidence, among others, for the court to consider in determining whether the expert has employed a sound methodology in reaching his or her conclusion.”

Id. at 419.  Even so, these concessions, improvident as they might have been, would not permit the Supreme Court to resolve the case as it did.  There was nothing in the Landrigan case, however, which would count as a biomarker of individual causation, or as support for a claim that Mr. Landrigan’s exposure was so much heavier than average that his personal exposure put him on the dose-response curve at a point that corresponded to a relative risk greater than two.

Here is how the Supreme Court described Dr. Sokolowski’s attempted reasoning process:

“In the present case, Dr. Sokolowski began by reviewing the scientific literature to establish both the ability of asbestos to cause colon cancer and the magnitude of the risk that it would cause that result. Next, he assumed that decedent was exposed to asbestos and that his exposure, in both intensity and duration, was comparable to that of the study populations described in the literature. He then assumed that other known risk factors for colon cancer did not apply to decedent. After considering decedent’s exposure and the absence of those factors, Dr. Sokolowski concluded that decedent’s exposure more likely than not had been the cause of his colon cancer.”

Id. at 420-21, 1087-88.  The obvious fly in the ointment is simply that many people with no known risk factors for colon cancer develop the disease.  The assumption behind a cohort study is that all the risk factors are even balanced between the exposed and the unexposed cohorts, and so the relative risk reflects the role of the exposure in question.  Of course, this assumption is rarely true outside the context of a randomized clinical trial, and the Selikoff studies relied upon by plaintiff’s witnesses were particularly inept in controlling or accounting for confounding factors.  Assuming, however, that both the exposed and unexposed groups had the same proportion of men without “known” risk factors, then the most Sokolowski and Wagoner could say was that Mr. Landrigan’s risk of colorectal cancer had been increased by 55% or so, above that of the risk for men similarly situated but lacking occupational asbestos exposure.  This 55% increase was the basis for the Court’s observation that the attributable risk was about 35%.  What the Court left for another day was how, if at all, this evidentiary display could support a conclusion of specific causation.  The trial and intermediate appellate courts saw clearly that Sokolowski and Wagoner had utterly failed to support their specific causation opinions, but the Supreme Court was intent upon giving them another bite at the apple:

“Without limiting the trial court on remand, its assessment of Dr. Sokolowski’s testimony should include an evaluation of the validity both of the studies on which he relied and of his assumption that the decedent’s asbestos exposure was like that of the members of the study populations. The court should also verify Dr.  Sokolowski’s assumption concerning the absence of other risk factors. Finally, the court should ascertain if the relevant scientific community accepts the process by which Dr. Sokolowski reasoned to the conclusion that the decedent’s asbestos exposure had caused his cancer.”

Id. at 420, 1088.  The Court thus did not give plaintiff’s expert witnesses a free pass for trial number two.  When faced with the prospect of having to show that Sokolowski’s and Wagoner’s ipse dixit were reaccepted by the relevant scientific community, the plaintiff dismissed her case.

Posted in Causation, Expert Witnesses, Rule 702, Scientific Evidence | Comments Off on Landrigan v. The Celotex Corporation, Revisited

Styles of Judicial Opinion Writing

May 28th, 2013

Not Enough Original Words

You copy a little,
But not too much.
Some people say it’s degrading.

In Cojocaru v. British Columbia Women’s Hospital and Health Centre, the Supreme Court of Canada confronted an interesting question whether a litigant may challenge an adverse judgment on grounds that the trial judge plagiarized from the plaintiffs’ submissions.  Cojocaru (May 24, 2013).

In this complex medical malpractice case, the trial court issued its written judgment in 368 paragraphs, of which only 47 were discernibly original.  The overwhelming majority of the paragraphs, 321, were plagiarized from the plaintiffs’ submissions, without attribution.  The Supreme Court agreed that this degree of plagiarism raises a serious question whether the trial judge’s judgment should be set aside as not a truly independent evaluation of the facts and law.  Id. at para. 10-11.

The Court proceeded to note that trial court judgments are “entitled” to a presumption of integrity and impartiality.  The defendants argued that any such presumption was vaporized by the “the extent of the copying, the quality of the copying, the lack of attribution for the copying, the nature of the case and the failure to fulfill the basic functions of reasons for judgment.” Id. at para. 52.

Although acknowledging that the trial court had copied over 87% of its judgment from the plaintiffs’ briefs, the Supreme Court held that the defendants had failed to rebut the presumption of judicial integrity and impartiality.  Id. at  para. 73.  Judicial plagiarism is, according to the Supreme Court, “a longstanding and accepted practice.” Id. at para 30.  The Court, however, acknowledged that excessive plagiarism “may raise problems.”  Id.

The Supreme Court rejected the notion that judges should express themselves in their own language, and that their reasons should be their “original” work product.  Id. at 31.  Copying, according to the Court, does not mean that plagiarist judges have failed to think about and resolve the issues before them.  Id.

The Court conceded that it was not “best practice for judges to bulk up their judgments with great swaths of borrowed material.”  If the carried the copying to excess (presumably much greater than 87%), the copying “may raise problems.”

Rules and prohibitions against plagiarism do not apply to judges, say the Canadian Supreme Court judges.  Id. at para. 32.  According to the high court, lack of originality is a virtue in a system that depends upon precedent.  “Judges are not selected, and are only rarely valued, because of their gift for original expression.”  Id. (quoting, with attribution, thankfully, Simon Stern, “Copyright Originality and Judicial Originality” 63 Univ. Toronto L. J. 1 (2013)).

The Supreme Court cited precedential support of its judgment that judges are too busy to think and write in their own words.  See id. at para. 37, 38 (citing English v. Emery Reimbold & Strick Ltd., [2002] EWCA Civ 605, [2002] 3 All E.R. 385 (holding that copying does not invalidate a court’s decision); id. at para. 41 (citing United States v. El Paso Natural Gas Co., 376 U.S. 651, 656 (1964) (noting that when trial judge “adopt[s] verbatim” the findings of fact submitted by counsel, “[t]hose findings, though not the product of the workings of the … judge’s mind, are formally his; they are not to be rejected out-of-hand, and they will stand if supported by evidence”).

The majority of the intermediate appellate court, the Court of Appeal, had held that the extent and nature of the plagiarism was “cogent evidence” of lack of the trial judge’s judicial integrity.  Although the Supreme Court rejected the conclusion that the trial judge had failed to make an independent decision, the Supreme Court did display some original, independent thought in upholding the defendants’ assignments of error.  The Supreme Court reversed the judgment for plaintiffs on the merits.  Id. at para. 76.

Judicial Verbosity – Too Many Original Words

There is a tradition in New Jersey appellate cases to never issue a short, succinct opinion when a long one is at hand.  Plaintiff Alison Rosenberg took Merck’s drug, Fosamax, for osteoperosis. After undergoing a tooth extraction, Rosenberg developed bone complications of her jaw.  She claimed that she developed osteonecrosis of the jaw (ONJ), and that this condition was caused by her use of Fosamax, and by Merck’s negligent failure to warn that Fosamax causes ONJ. Rosenberg et al. v. Merck Sharp & Dohme Corp., No. A-5271-10T3, 2013 WL 1187916 (N.J. Super. Ct. App. Div. Mar. 25, 2013) (per curiam).

Merck defended and affirmatively sought to show that Rosenberg did not have ONJ.  Its expert witness, Brent Ward, D.D.S., M.D., an expert in ONJ and in maxillofacial surgery, testified that Rosenberg suffered from osteomyelitis, not from ONJ.

At the end of the case, the trial judge instructed the jury to make findings:

“(1) whether plaintiff proved it was more likely than not that plaintiff had ONJ;

(2) whether plaintiff’s development of ONJ was a result of taking Fosamax; and

(3) whether Merck negligently failed to adequately warn about the risks of ONJ from Fosamax.”

The jury resolved the first question by finding that the plaintiff had not suffered from ONJ, and pursuant to the trial court’s instructions, ceased deliberations and notified the court of their final verdict, upon which judgment was entered.  The jury thus never reached the questions about ONJ and Merck’s liability.

Rosenberg, et vir., appealed, contending that the trial judge erred in excluding evidence of Merck’s 2010 change in its warning label, which impeded her attempt to impeach Dr. Ward on the relationship between Fosamax and ONJ.  The problem is that the excluded evidence was relevant only to an issue that the jury never reached.  Somehow the New Jersey Appellate Division managed to draw out its per curiam opinion to eleven pages, in its own words.

Posted in Trial Procedure | Comments Off on Styles of Judicial Opinion Writing

« Older Entries
Newer Entries »

The opinions, statements, and asseverations expressed on Tortini are my own, or those of invited guests, and these writings do not necessarily represent the views of clients, friends, or family, even when supported by good and sufficient reason.