Mass Torts: State of the Art

As We Were Saying

David Oliver — Wed, 08 May 2013 22:02:27 -0600

It certainly appeared by the 1960s that humanity was leaving the so-called second epidemiologic tranistion. Ancient pathogens were being defeated at every turn as science went from triumph to triumph. And thus a whole generation of public health professionals committed themselves to combating man-made and degenerative diseases; by then believed to be the leading cause of human suffering.

It was sin that brought these afflictions. The sin of greed led owners to run factories demanding back-breaking work. The sin of gluttony led workers to pack on the pounds putting unnatural stresses on the back that wore it down too soon. Decades of litigation over sore backs followed and most were fought over the question of which sin was the greater. It wasn't until recently that some researchers began to wonder, and to investigate, whether or not something else was actually responsible for the epidemic of back pain that arose in the second half of the twentieth century.

That something else turns out to be an anaerobic bacteria and the proof of it would make Robert Koch proud. In more than 40% of all patients with low back pain who had signs on MRI of degenerative disc disease the bacteria was identified as being present. And when antibiotics targeted at that very bacteria were administered a big chunk of those cases resolved (versus placebo).

Not every case of lumbar pain was due to anaerobic bacteria and it also appears that sudden stress on the back may create an opportunity for the bug to exploit. Nevertheless, it's quite clear that antibiotics, rather than difficult surgery with its attendant permanent (though usually partial) disability, may be all it takes to restore sufferers of back pain to good health. The lesson here, as my colleague put it in a joint presentation we gave today, is that "we may forget to focus on germs but they never forget to focus on us."

A Duty To Mine Big Data

David Oliver — Wed, 01 May 2013 16:35:56 -0600

We had another jury trial and thus were offline for a few weeks but it didn't take long to dig up something I hope you'll find of interest. I ran across it in a recent opinion by the district court in the In Re Fosamax Product Liability Litigation. What is it? It's a duty to actively mine the FDA's data for a signal, a hint, that your approved pharmaceutical product might be associated with an adverse event.

As an initial matter I almost didn't get to the interesting part of In Re Fosamax Product Liability Litigation thanks to this eye-roller: "In applying the nine Bradford Hill factors, he [Dr. Cornell] reviewed Plaintiff's medical records from 1996 to present, the office notes and depositions of her treating physicians, and 'past and current medical literature on the topics of osteopenia, osteoporosis and their prevention and treatment with bisphosphonate drugs including alendronate'". Which was followed by: "The methodology Dr. Cornell used is sufficiently reliable because the Bradford Hill criteria are 'broadly accepted' in the scientific community 'for evaluating causation, and 'are so well established in epidemiological research". This business of giving expert witnesses a pass for doing nothing more than invoking the great epidemiologist's name and saying that their method consisted of peering at the evidence through the supposed lens of Hill's "criteria" is utterly appalling but it would take several paragraphs to explain and we've done it before so I'll leave it at that.

The portion worth pondering is found further down in the discussion of defendant's motion to exclude a different expert witness, Dr. Madigan. Madigan is a statistician from Columbia with an impressive resume. He was tasked with assessing "whether a signal of problematic oversuppression of bone turnover and associated [atypical femur fractures (AFF)] . . . existed for Fosamax, using industry standard pharmacovigilance techniques and data sources, and the adverse event terms selected by Merck to internally evaluate the same" and whether "the strength of that signal, if any, in comparison to the signal, if any, for such events in other products indicated for the prevention and treatment of osteoporosis". To identify and evaluate such a signal Madigan took several medical terms considered by the defendant to be possible indicators of "oversuppression of bone turnover" and/or AFF and, using a program called Qscan, ran them through FDA's Adverse Event Reporting System ("AERS") database looking for associations with Fosamax use.

The data revealed "the presence of a clear signal for oversuppression of bone turnover and associated atypical femur fracture events utilizing the terms selected by Merck for such analysis. By standard metrics of 'signal' detection, the signal is strong, consistent, and not ambiguous. Of perhaps greater concern, the signal was striking in comparison to that for other drugs indicated for the prevention and treatment of osteoporosis. As early as 2001-2002, the spontaneous report data for Fosamax provide signals for a number of indicators of suppression of bone turnover. For the comparator drugs, such signals either never appear or appear years later." As Qscan (or similar software) data mining is widely used (and often mandated in the pharmaceutical industry) and as peer-reviewed papers arising out of data generated by the software had been published the court concluded that "data mining in pharmacovigilance" is a reliable method. Fair enough.

The excitement comes in the next paragraph in which the court concludes that Madigan's testimony "fits", and so is relevant to, an issue in the case because it informs the question of whether the defendant should have warned of the later-perceived risk of AFF. The court thereby implicitly held, I think, that the defendant had a duty to mine the FDA's data as early as 2001-2002. In other words, the existence of powerful data mining tools capable of uncovering an early signal of a possible harm associated with defendant's product created a duty to use such a tool. Ultimately, that's a duty to discover any statistical association between your product and some harm in the FDA's (admittedly accessible) that might be causal and to thereafter warn about it; and it's a duty to mine not only your data but any data that might shed light on your product.

It's hard to know what to make of a duty to mine big data. Obviously it's a potential problem for defendants. "Should have known" is no longer what was reasonably knowable (by humans). Instead it's what was knowable given a duty to use "powerful data mining and signal detection capabilities" including "a powerful query-by-example module that allows users to mine and visualize data through inquiries utilizing multiple case data elements." We're already in the middle of something similar; a case in which the plaintiff is demanding to see all of the death certificates collected by our client's benefits department claiming that we had, or had assumed, a duty to look for mortality patterns among our workers that might suggest a work-related etiology - the idea being that the plaintiff's decedent had succumbed to one such workplace illness and we either knew or should have known from looking at the death certificates that danger had been lurking.

And of course there's the problem of experts coming in after the fact and running term after term after modified term in varying combinations until an association emerges. Working backwards they'll then construct a narrative about why the set of terms eventually founded to produce the association would have been the obvious choice at the time the data mining ought to have been done; and the failure to look for such an obvious potential association will thereafter be cast as willful ignorance.

Then there's the bane of defendants in all latent disease cases - the hindsight bias. It's the "knew it all along" fallacy that emerges when a jury is shown the picture on the box before they see the puzzle pieces inside. They can't thereafter imagine that the few anecdotes and case reports that constitute a handful of the pieces that make up the picture on the box could ever have suggested mere randomness. That means the ability to mine huge amounts of data will make signals easier to find while making it harder to mount a successful state-of-the-art defense any time Qscan can tease a signal from the data.

Data mining has led, and will lead to startling discoveries in the sciences. In the law it may well lead to startling liabilities - especially if defendants are made to pay for harms foreseeable only by the most powerful software available. Ponder that.

"FELA is a very favorable statute for railroad employees"

David Oliver — Fri, 05 Apr 2013 00:10:07 -0600

In fact, it is so favorable that the post hoc ergo propter hoc fallacy is enough to carry the day. Consider Clements v. Norfolk Southern.

Plaintiff, an employee of a railroad, either slipped off or fell off a tire. How he came to be on it is not made clear but the tire was four feet in diameter and off it he went. At trial he testified that prior to the accident he had been free of back pain but was afflicted by pain, and a herniated disk, thereafter. Though he had no expert willing to opine that his fall was the cause of his sore back the court held "[i]t is not beyond the province of an ordinary person that falling off a four foot high tire could cause a herniated disk."

While it has long been the case that "experts" can be found who are willing to testify that tying one's work boots will readily cause a herniated disk or that falling off the roof of a two story building can't possibly cause a herniated disk - and one side or the other of every scenario in between - courts ought not give up on the quest for sound science in such cases. That's because science, at long last, is starting to test various theories concerning the causes and appropriate treatment for herniated disks. So far, evidence-based assessments are giving the "province of an ordinary person" a thrashing.

Of Mice and Men and Their Breath

David Oliver — Wed, 03 Apr 2013 23:44:45 -0600

One of the longstanding disputes in toxic tort litigation revolves around whether or not animal models are good (i.e. predictive) models for assessing the likely outcome of exposures to the chemical in question. As we've written previously the 1,3 butadiene question hangs crucially upon the question of whether or not workers in SBR (styrene-butadiene rubber) plants are likely to metabolize butadiene the same way as B6C3F1mice. However, isn't there an assumption here that goes unexcavated?

Aren't we assuming that all people metabolize things like butadiene the same way? Undoubtedly. But guess what? Thanks to differences in the bacteria in their gut, their diets and likely other yet undiscovered factors, people metabolize things differently. Read all about it at: Human Breath Analysis May Support the Existence of Individual Metabolic Phenotypes

Is It Liability All The Way Down?

David Oliver — Wed, 27 Mar 2013 00:10:11 -0600

Back to the Burks problem from the last post.

Is there something wrong with shifting one element of plaintiff's burden of proof (defendant-specific "but for" causation in the case of alternative liability) to the defendants when nature may well have been, but probably wasn't, the cause of plaintiff's injury? The trial court in Burks didn't think so. After all, the rationale behind the decision in Summers v. Tice to engage in such burden shifting was that as between a defendant who committed a wrong, and an innocent and wronged plaintiff, the risk of loss ought to be borne by the wrongdoer. So why would it be necessary that all of those potentially responsible for the harm be present? After all, thanks to Summers the wrong (breach of a duty) has been decoupled from causation in the "wrong + 'but for' causation = legal causation" calculus.

Thus, if the defendants each breached a duty, and collectively the probability exceeded 50% that they, as opposed to nature, were responsible for the plaintiff's C. sakazakii infection, you have everything you need for legal causation and alternative causation ought to apply since plaintiff can't possibly prove where the bacteria came from, right? I don't think so.

Let's go back to David Hume and causation. Remember his quote: "We may define a cause to be an object, followed by another, and where all the objects, similar to the first, are followed by objects similar to the second. Or, in other words, where, if the first object had not been, the second had never existed." This is what we all today recognize as causation, or counterfactual causation; if "x" hadn't happened "y" wouldn't have followed. Easy enough.

Now think about causation in the case of Summers - in which the burden of proof was shifted from plaintiff to the two hunters, each of whom had failed to handle his firearm responsibly and had discharged his weapon in plaintiff's direction. What happens if the second hunter "had never existed"? It doesn't change anything. Causation falls on one defendant instead of two. The cause of plaintiff's harm would have been the act of the first hunter and liability would have been imposed on defendant by his own hand as he'd be the only cause left. But what happens in the Burks scenario if a defendant goes missing?

The Burks plaintiff sued two suppliers of a product potentially contaminated with C. sakazakii, a common environmental pathogen, and the probabilities are such that it's slightly more likely (51%) that plaintiff's infection was due to the contamination of one of defendants' products. Let's say the probabilities were 26% that it came from the first defendant, 25% from the second and 49% from Mother Nature who sprinkles the bacteria around many homes. The district court in Burks thought alternative liability would apply in such a case because the source of the pathogen was, collectively, more likely than not the defendants. But what if one of the defendants' breach of a duty "had never existed"? Suddenly the remaining defendant is not liable. What was determinative of liability?

If I've got this right then the rule in Burks is as follows: Whether or not you are liable for plaintiff's injury hangs on the conduct of third parties over whom you had no control. Right? And that can't be the law. Or is liability independent of likelihood such that it goes all the way down? More on that tomorrow (or the next day...)

Very Little Authority For Now, But Maybe Not For Long

David Oliver — Fri, 22 Mar 2013 10:44:45 -0600

Two yellow pads are full of notes, diagrams, thoughts and arguments and still the answer eludes me. So, rather than continue not to post anything about all the interesting stuff going on in mass torts while I try to figure it out I'll just throw out what I've got so far over a series of posts and move on to something less difficult (and likely more interesting). On then to the question that's been bugging me for a month.

Should alternative liability, in the form of burden of proof shifting, apply in the following case: (1) a neonate contracted a Cronobacter sakazakii infection and was seriously injured; (2) C. sakazakii is a ubiquitous pathogen and readily forms biofilms on stainless steel, inside household water pipes and upon other surfaces (3) C. sakazakii infections in neonates have also been repeatedly traced to PIFs; (4) plaintiff's experts opined that it's more likely than not that the source of C. sakazakii was the powdered infant formula (PIF) fed to the neonate; (5) PIFs are not sterile since the process of sterilization would destroy the nutritional value of the PIF; (6) there were two or more suppliers of PIF whose product was prepared for the neonate; (7) none of the remaining PIF fed to the newborn, nor any of the lots from which they were drawn, were found to contain C. sakazakii; and, (7) the wrong complained of was a failure to warn that full term neonates, like the plaintiff, were at risk of C. sakazakii infections.

Alternate liability analysis often begins with Summers v. Tice. In that case both hunters who had fired their weapons in plaintiff's direction had been sued and both could be shown to have breached their duty to safely handle those weapons. Because it was impossible for the plaintiff to show whose buckshot was responsible for his injury, and because it was deemed just that the consequence of that impossibility fall on the culpable defendants rather than the innocent and injured plaintiff, the court held that the defendants, rather than the plaintiff, should bear the burden of proving from which shotgun the pellets had originated. Fair enough.

Here, in Burks v. Abbott Laboratories, both of the manufacturers of PIF consumed by the plaintiff had also been sued. Furthermore, neither's product bore a warning about the hazard which plaintiff claimed they had a duty to disclose. However, not all the possible sources of C. sakazakii (Mother Nature being judgment-proof) were before the court. Nevertheless, the court held, though it "located very little authority on this specific question" that the determination of whether it was more likely than not that the source of plaintiff's C. sakazakii infection was the defendants' PIF should be applied to the defendants collectively.

Ponder the consequences. Let's say that 51% of all C. sakazakii infections in neonates are due to PIFs even when C. sakazakii can't be isolated from the product or the lot from which it was drawn, and that 49% of all C. sakazakii infections in neonates are due to C. sakazakii found in the household water used to reconstitute the PIF or on kitchen surfaces. Also, assume the warning could fairly be said to be lacking (which in this case is not at all a given and in fact raises as many questions as stacking defendants to get to the 51% threshold - but that'll be addressed in subsequent posts).

If there's only one supplier of PIF there's no need for alternative liability so assume there are at least two. But what happens when you start dividing up the 51% among the defendants? Unless one of them was responsible for ~99% of the product the result is that you've stacked two defendants who might, but probably didn't, have something to do with plaintiff's infection and handed them all of the liability for it. Is that fair? What if there were 100 suppliers? And what's the justification for stacking defendants? Is it because they're in the same business? If so, how similar must their businesses be to permit such stacking? If in order to get to a 51% likelihood as to the source of the infection you had to stack PIF manufacturers with stainless steel kitchen appliance manufacturers (because C. sakazakii happily lives on stainless steel and creates biofilms that makes it nearly impossible to remove with household cleaning products) could you justify handing them the burden of proof by saying they were all in the food business?

Or what about reducing defendants' liability by Mother Nature's share? Would that solve the "overdeterrence" problem? What follows from the fact that the potential for these sorts of infections, due to the nature of pathogens, is essentially binary (you get it or you don't) and not of the dose-response variety seen in typical mass tort cases? These are the sorts of questions that have led to a lot of head scratching but so far few answers.Over the next couple of days I'll type up more of these questions and the paths down which they lead (at least the ones I've thought of and followed).

How a fact pattern like the one in Burks gets resolved is I think a very big deal. That's because an awful lot of diseases laid over the last forty years at the feet of man-made substances and bad habits turn out to have been due to pathogens all along. It also appears we're entering an era in which old scourges reemerge thanks to having evolved antibiotic resistance and new ones arise thanks to globalization. The resulting morbidity and mortality will makesevery other mass tort pale in comparison and so far juries aren't having much trouble blaming defendants for the depredations of Mother Nature's tiniest critters. We think there's a wave of litigation coming in which plaintiffs will assert liability for facilitating the transmission of pathogenic agents. The answers to questions like those posed by Burks will be critical in determining how it all plays out.

The Texas Supreme Court Has Decided To Hear Bostic

David Oliver — Sun, 17 Feb 2013 02:03:00 -0600

Here's some good news.

Though at first they decided not to hear plaintiff's appeal, the Texas Supreme Court has reconsidered. Hopefully that means there'll be an argument about the Dallas Court of Appeals' interpretation of Borg-Warner v. Flores in its Bostic opinion wherein it held that every putative cause to which liability may legally attach must be a "but for" cause.

The problem with Bostic, for asbestos claimants or any other claimant asserting a cause of action against multiple defendants each responsible for an act which was potentially sufficient to have caused plaintiff's injury, is that it created a rule whereby a plaintiff could not recover from any defendant so long as more than one defendant put in motion a series of events that could have lead to plaintiff's injuries. As we've been saying for months, the idea that if three defendants each negligently and separately start a fire capable of burning down the plaintiff's house she can't recover from any of them for the destruction of her house because she can't prove which one was the "real" cause is not only unfair, it hasn't been the law in Texas for a very long time. Anyway, we'll report on the briefs, the arguments and the result as the case proceeds.

In the meantime, we've been speculating about why the Court decided to give Bostic a second look. My pet theory is that someone dug into Dow v. Abutahoun, an opinion just a week or so old (and also from the Dallas Court of Appeals) and decided that for the first time since Havner a philosophical discussion about causation is in order. In Abutahoun the appellate court held that our Chapter 95 (triggering a requirement that plaintiff prove control over the contractor employee's work and actual conscious knowledge of the risk created) applies to a case in which employees of the premises owner create, or contribute to, the hazard of which plaintiff complains; so long as it was the same improvement to real property on which both contractor's and premises owner's employees were working. The idea seems to be, again, that when a defendant creates a risk of harm indistinguishable once manifest from those created by others, then the plaintiff bears the burden of proving that the putative harm was the "real" harm. That hasn't been the law and we doubt that the legislature intended such a result when it enacted Chapter 95. We'll see.

Finally, lest you think we've gone in the tank for plaintiffs, we'd note for the record that in both cases we'd hold for the defendant if we got to wear the robes - but we'd do it on different grounds. While it's impossible to say which fiber or group of fibers caused any given case of mesothelioma, it's easy for all sides to calculate the risk posed by any given exposure. And nowadays the exposures we're fighting over are typically minuscule and the risk associated with them are literally about the same as having a large meteorite explode 25 miles above your head. Which is to say that they've fallen to the point of simply being among those "ordinary risks of life" that make this life so interesting and so terrifying.

Squeak Squeak

David Oliver — Tue, 12 Feb 2013 01:29:11 -0600

In the run up to the trial of a case in which we're arguing that the B6C3F1 mouse ain't a man and 1,3 butadiene ain't a human carcinogen just because it causes cancer in the B6C3F1 mouse, out comes "Mice Fall Short as Test Subjects for Humans’ Deadly Ills" by Gina Kolata of the NYTimes. And it's a bombshell. Kolata reports on the paper "Genomic responses in mouse models poorly mimic human inflammatory diseases" and its central finding that immune responses in the mouse, including those related to heart disease and cancer, are no more closely correlated with human responses to the same stimuli than the roll of a pair of dice. It's the long-sought explanation as to why e.g.

"every one of nearly 150 drugs tested at a huge expense in patients with sepsis has failed. The drug tests all were based on studies in mice. And mice, it turns out, can have something that looks like sepsis in humans, but is very different from the condition in humans."

Good stuff, though not all that surprising if you've been following the sad tale of the development of drugs that cure cancer in mice yet have no effect in humans. And it doesn't mean that all scientific studies done on mice are worthless. Far from it. The ability to produce for example so-called knockout mice, rodents lacking a particular gene required to make a particular protein, allows an otherwise forbidden glimpse into the workings of the tiny chemical factories that we call cells. Nevertheless, the study does shatter the assumption that those little factories in mice run just like their counterparts in humans.

However, that's not the end of the story. If you read the whole thing you'll find this:

The study’s investigators tried for more than a year to publish their paper, which showed that there was no relationship between the genetic responses of mice and those of humans. They submitted it to the publications Science and Nature, hoping to reach a wide audience. It was rejected from both.

... reviewers did not point out scientific errors. Instead, [one of the authors] said, “the most common response was, ‘It has to be wrong. I don’t know why it is wrong, but it has to be wrong’ ”

which leads to our final point. Daubert's peer review factor was intended to serve as an independent indicator of reliability. The Court assumed that disinterested scientists on the lookout for bad science served as gatekeepers of the journals through which "scientific knowledge" was disseminated. Perhaps when there were far fewer journals and far fewer academics desperate to be published peer reviewers served such a function. Nowadays they too often serve the status quo - barring from publication the sort of disruptive findings that would discomfit the guild they serve. Thus, if we're not careful, does Daubert risk being effectively transmuted, at least in part, into Frye - i.e. a test of general acceptance rather than a test of sound science.

Those Who Switched From Saturated Animal Fats To Polyunsaturated Vegetable Oils May Have Made A Fatal Mistake

David Oliver — Sun, 10 Feb 2013 19:49:01 -0600

The publication of "Use of dietary linoleic acid for secondary prevention of coronary heart disease and death: evaluation of recovered data from the Sydney Diet Heart Study and updated meta-analysis" in the British Medical Journal should be sobering news for twenty-first century public health advocates. The results of the biggest and most thorough study thus far to investigate what happens when people switch from animal fats to vegetable oils couldn't be more clear. Those who heeded the advice to switch from saturated fats to polyunsaturated vegetable oils dramatically reduced their odds of living to see 2013.

What happened? Fifty years ago the medical community did an about-face (it had previously thought the evidence equivocal at best and so cautioned against jumping to conclusions) and instead went all in on polyunsaturated fats. It reasoned that since (a) cholesterol is associated with cardiovascular disease and (b) polyunsaturated fats reduce serum cholesterol levels, it inescapably followed that (c) changing people's diet from saturated fats to polyunsaturated fats would save a lot of lives. In 1984 Uncle Sam got involved - Time magazine reported on it in "Hold the Eggs and Butter" - and he made a big push for citizens to swap out animal fat in their diet for the vegetable variety and a great experiment on the American people was begun.

Assuming the inferences drawn from the Sydney Diet Heart Study (~60% increase in risk of death by switching from animal fats to vegetable oils) apply to Americans and considering the population at risk of death from cardiovascular disease (~37% given recent NHANES data) we can't think of any mass tort, or combination of mass torts, that has produced as much harm as the advice to change to a plant oil-based diet. Maybe some clever lawyer will find a way to use the tort system to quell the fervor of those armed only with a desire to do good and evidence of an association between something they don't like and something else they don't like. We won't be holding our breath. But we will be hoping that those seeking to impose lean body styles and low salt diets on their fellow countrymen take a good long look at the polyunsaturated vegetable oil data before they order their riders to sally forth on the next crusade.

Discretizations

David Oliver — Fri, 08 Feb 2013 12:46:23 -0600

Getting the causation arrow pointed in the right direction: obesity leads to vitamin D deficiency and not the other way 'round.

Happy TRAIL to Pharma - homeruns are still possible with small molecule drugs.

Whew, no cases of rabies, plague, polio or hemorrhagic fever this week. Then again, only 1 person died in all of Miami last week?

Hormone replacement therapy for chronic illnesses gets a "D" from the USPSTF; and that included so-called "bioidenticals".

H. pylori infection more than doubles the risk of colon adenocarcinomas.

Be careful with unsubstantiated claims!

David Oliver — Tue, 05 Feb 2013 22:25:16 -0600

An interesting and thoughtful comment on one of our recent posts from Dr. Marc-Andre Gagnon:

I believe your interpretation of the study is simply wrong. The authors refer to a study by Chan et al., published in BMJ, which showed that such differences between protocols and publications were prevalent in a sample of 70 clinical trials, 56 of which received industry funding. The quoted paper never said that non-industry funded studies showed as much difference (or bias) as industry funded studies, it simply made the case that we needed access to full data for every clinical trial to be sure that there are no bias in the reporting of results.

On the link on reporting bias in research and industry funding, one can find a Cochrane systematic review by Lungh et al. published on the topic two months ago.

Marc-André Gagnon

School of Public Policy and Administration, Carleton University

All the best,

Marc-André

Which precipitated an exchange wherein we asked about discrepancies and biases in non-industry funded data and got this equally thoughtful reply:

I think the topic is a very complex one since you need to take in consideration publication bias: journals like crispy results and dull ones are often unpublished. I think this type of bias applies in the same way to industry and non-industry funded studies. But this is not the same thing as reporting bias when it comes to the important question of industry funding and research outcome. A recent Cochrane systematic review showed that the problem is not an imaginary one: http://www.pharmaceuticalpolicy.ca/research/industry-sponsorship-and-research-outcome

The article analyzes the existing literature about reporting bias.

One thing is likely not in dispute - we're entering (or have entered) another age of empiricism in which data, and the inferences statistically drawn from it, presumptively trump even the cleverest idea if the data underpining that idea is biased, not honestly gathered or missing (in whole or in part). And that's a good thing. See also: "'In God we trust, all others (must) bring data’".

Thanks for the comments, Dr. Gagnon

Discretizations

David Oliver — Mon, 04 Feb 2013 09:26:59 -0600

The hedgehog: actually it's lots of little things.

Some say being overweight protects against many diseases of aging, some those studies are biased.

Father of the bride, and future defendant in 47 A. butzleri cases?

"...it is now possible to determine with great precision the cellular origin of solid tumours in mice." Imagine what it would mean for toxic tort cases if the same could be done in people.

Older fathers have children who live longer.

Are Industry-Sponsored Clinical Trials Misleading Or Just As Sloppy As Those Done By Academics?

David Oliver — Fri, 01 Feb 2013 13:45:55 -0600

If you've read "Published clinical trials shown to be misleading", a story about a newly published article detailing discrepancies between a drug company's documents and the papers it published based on data drawn from those documents, and if you pondered this quote from one of the study's authors about their review of those internal company documents

[w]e could see all of the biases right in front of us all at once

then you probably think something really damning has been found. If you clicked through to the editorial published along with the article, "Getting More Generous with the Truth: Clinical Trial Reporting in 2013 and Beyond", you probably also agree that "[f]or many working in medical journal publishing, these results will sadly not be surprising" and would readily join the call to require greater transparency among drug companies conducting clinical trials of their products for safety and efficiency. Certainly that's been the common reaction seen across the various reports, blogposts and tweets about the article.

But if you actually read the article, "Differences in Reporting of Analyses in Internal Company Documents Versus Published Trial Reports: Comparisons in Industry-Sponsored Trials in Off-Label Uses of Gabapentin", you might reconsider. Take all of those biases right in front of the investigators. Presumably the nefarious drug company was at least clever enough to bias the results in the drug's favor, right? Consider this from the Results section:

...we did not assess whether disagreements in descriptions of types of analyses resulted in different participants being analyzed for efficacy and safety. We also did not examine the impact of the observed disagreements on the effectiveness of gabapentin for the indications specified in our study

Yet surely the discrepancies must be the unmistakable sign of the profit motive at work, distorting the truth, right? Well, there's also this

The discrepancies in description of types of analyses we observed in our study are not unique to trials sponsored by for-profit entities such as pharmaceutical companies. Using trial protocols obtained from institutional review boards and corresponding publications, previous research has shown that disagreements such as those we described in our study can be observed in trials with funding from both for-profit and not-for-profit entities.

So while many who read only the tweets and posts about the paper concluded that it's a familiar story of books being cooked to show a drug to be safe and effective when it's not, there was no such conclusion. It's sort of like a murder case with no body and no motive; and nobody missing. All that having been said, drug companies are held to a higher standard and when it comes to transparency they ought to assume that every mistake their employees make will eventually wind up on the front page of the NYTimes, or at least PLOS One.

Sam Hammared

David Oliver — Wed, 23 Jan 2013 15:04:12 -0600

Yet another court has sent packing Dr. Sam Hammar and his "every exposure", "every breath" method of causal attribution in asbestos/mesothelioma cases. This time it was a USDC in Utah (Central Division) that ordered such testimony be excluded; essentially by adopting the late Popperian view that "an untested and potentially untestable hypothesis" isn't the same thing as the "scientific knowledge" required by Rule 702. Ok, so why blog about it if it's not particularly surprising? Well, because it presents an opportunity to discuss (1) the perils of reasoning by analogy and (2) why fault is invariably and inextricably bound up with causation.

Hammar, no doubt invoking the linear no-threshold dose response model for carcinogens, countered the motion to exclude his causation opinion by elaborating that his attribution of causation to every exposure is proper because no exposure can be ruled out (since every fiber carries some risk). Then, either conflating risk with causation or using the "small glasses" approach (the order doesn't say which), Hammar concluded that if no exposure can be ruled "out" then every exposure must be ruled "in"; thereupon leading logically to his opinion that the plaintiff's mesothelioma "was caused by his total and cumulative exposure to asbestos, with all exposures and products playing a contributing role."

The court analogized Hammar's reasoning to that of a detective who, upon learning of a murder in a large family (and knowing that the killer is often found among family members), not only doesn't rule anyone out but proceeds to theorize that because none may be ruled out all must thereby be ruled "in". While agreeing that refusing to rule anyone out would be reasonable the court held that the same reasoning "would not allow the detective to attribute legal liability to every family member on the basis of such a theory."

Unfortunately there are a few problems with this analogy. First, unlike family members, within categories of fiber type asbestos is fungible and no technique exists that can differentiate by origin one fiber from another. Second, the scenario appears to assume a one-hit model of causation (i.e. that only one family member was actually responsible) - which is not (nowadays anyway) the causal model plaintiffs push in these cases. Finally, and most importantly, it ignores the fact that in a real case causation doesn't get plaintiff anywhere if he hasn't established breach of a duty (e.g. failure to warn).

Recasting the analogy to more closely mirror a typical case yields something like the following: the detective arrives at the scene of the crime and finds identical octuplets standing around the victim who has been crushed under the weight of several large rocks. Each sibling admits to having exposed the victim to one or more of the rocks. That I think is more analogous to what the Fifth Circuit grappled with in Borel v. Fiberboard and it illustrates why in such cases courts shift the burden of sorting it all out from the plaintiff to the defendants. It's because no matter which defendant or which combination of defendants is held liable they're necessarily blameworthy and no one will think an injustice has been done.

But now let's update the analogy from 1973 to 2013. The detective arrives on the scene and finds the identical octuplets, the recently crushed victim nearby and each sibling admitting to having exposed the victim to rocks. However this time three of the siblings are found to have been responsible for nothing more than a handful of sand whereas the other five, all of whom were responsible for large rocks, are judgment-proof (thanks to bankruptcy court). Just as asbestos plaintiffs can't find any evidence that brake work increases the risk of mesothelioma, the detective can find no report of a handful of sand ever crushing anyone. What's a clever problem solver intent on blaming each viable defendant to do?

Hatch a "straw that broke the camel's back" theory which makes even a single grain of sand a necessary cause. Couple it with the linear no-threshold dose response model for carcinogens such that each grain of sand carried with it some non-zero risk of death by crushing and you're golden - your medical expert now gets to testify not only to causation but to the imposition of a risk, the magnitude of which (death) is enormous, on the decedent and, worse yet, without warning.

A number of courts intuit that there's something very much wrong with the theory but often can't quite pin down what that something might be. Here the court settled on the fact that the theory is untested and untestable. Now while it's true that no group of people who took one and only one breath of asbestos dust can be found and compared to another (equally nonexistent) group who never inhaled so much as a fiber, and no group of individuals can ever be ethically exposed to a known carcinogen just to see what happens to them, it's not clear why this ought to be a fatal flaw in plaintiff's theory. First, the theory is theoretically, if not practically, falsifiable and so under Daubert would avoid the label of pseudoscience. Second, would anyone really claim that in a hypothetical case in which the plaintiff was exposed to levels of asbestos that were higher than have ever been studied before that he couldn't recover because his theory that such massive levels of asbestos causes mesothelioma is untested and untestable? You get the point.

So if it's not a lack of testability, what's wrong with Hammar's theory? The problem is that it fails to assess the risk imposed, which is to say the defendant's conduct at the time of the conduct, and so fails to establish legal causation.

You can't go through life without kicking up some dust and it may well be the case that sometimes it really is one last straw, or one last grain of sand, that breaks the camel's back. But the focus of courts when deciding whether liability may justly be imposed should be on the conduct sought to be condemned. In the case of Smith v. Ford plaintiff sought to blame the defendant for a risk that was either nonexistent (as no study has ever shown an appreciably increased risk of mesothelioma among brake workers) or so small that it would take a cohort of many more people than have ever worked as brake repairmen to detect it. How can the imposition of such a risk, even if it exists, be a wrong?

It can't, or at least it shouldn't.

Virginia Adopts the Davidson Rule (Hopefully)

David Oliver — Mon, 14 Jan 2013 11:54:48 -0600

Virginia's highest court is the latest to wrestle with causation in asbestos/mesothelioma cases. Ultimately reversing a judgment in favor of Plaintiff and remanding the case for a new trial, the court in Ford v. Boomer decided that the substantial factor (or here, substantial contribution) test is unreliable and in any event inconsistent with Virginia law. The court embraced the Restatement (Third) of Torts' view of substantial factor causation (that it's incoherent), writing: "If courts cannot be relied upon to consistently construe the language, we cannot expect lay jurors to accomplish the same task." Concerned that the trial court's instruction to the jury, specifically that plaintiff must prove that exposure to defendant's product "was a substantial contributing factor in causing plaintiff's injury" left the jury free to assign liability to any exposure that was more than de minimis (the term was not defined), the court held that defendant's objection to the charge should have been sustained.

Ok, so if substantial factor/contribution cause doesn't cut it in Virginia what does? And if some exposure beyond "de minimis" isn't enough, what is? Sufficient cause, i.e. sufficient exposure! In parsing the Restatement (Third) of Torts section 27 and its comments the court concluded that the logic of permitting a jury to assign liability to either or both defendants in a case in which each negligently started a fire sufficient on its own to have destroyed plaintiff's property would apply with equal force in an asbestos/mesothelioma case. Consequently, "[t]he exposure must have been 'a' sufficient cause..."

So how does a plaintiff show that a putative exposure was "a" sufficient cause? He'll need "medical testimony as to the requisite exposure necessary to cause mesothelioma, to determine whether the exposure attributable to each defendant was more likely than not sufficient to have caused the harm." And how does he show that the attributable exposure was more likely than not causative? That's the $64,000 question.

If all plaintiffs need is the ipse dixit of a high priced testifying expert then they'll just go from saying "any exposure above background was a substantial factor" to "any exposure above background was sufficient" and nothing will change. Presuming that Virginia is serious about requiring plaintiffs to prove that a given exposure (which is to say dose) could have been sufficient, plaintiffs will have to show that the exposure in question more than doubled the risk of the disease. Essentially it means drawing the line at the lowest dose credibly shown to double risk. That's nothing new. It's the rule Judge Davidson has been applying for years in the Texas MDL.

So good news for Virginia defendants, right? Well, in Texas we have a great shares defense and apportionment scheme. Virginia? This sentence from Boomer should answer the question: "Other sufficient causes, whether innocent or arising from negligence, do not provide a defense." The only defense suggested by the court for an exposure determined to be sufficient (other than a warning) would be to show that the exposure occurred after the plaintiff had already developed his cancer. That means Virginia's old school joint and several liability scheme prevents Boomer from being much of a boon to most defendants. Yet, because plaintiffs will have to prove something like a 0.1 f/cc-yr cumulative dose for each putative "sufficient cause", the most marginal of the remaining defendants should get some relief. At least until plaintiffs can show the background risk is say 1:2,000,000, at which time 0.05 f/cc-yr becomes "sufficient". Etc., etc. (Now do you see why we've been saying the Restatement (Third) of Torts is a Trojan horse for the Precautionary Principle?)

Discretizations

David Oliver — Wed, 09 Jan 2013 18:26:42 -0600

The Annual Report to the Nation on the Status of Cancer (covering 1979 - 2009) is out. The good news is that overall death rates for some of the biggest killers continues to decline. The bad news is that deaths from cancers caused by infectious agents like HPV and hepatitis-C are up; doubling in some cases. Get your daughters, and sons, vaccinated.

It took several decades but the egg has now been cleared on the charges of murder by heart attack and stroke.

It also took decades to clear caffeinated coffee on all charges of murder by cancer. Now drinking 4+ cups per day has been shown to cut your risk of some cancers by half.

Many common food ingredients have been scientifically demonstrated to double or halve, and often double and halve, your risk of cancer. Click here to find out why.

The human breast milk microbiome is far more diverse than previously imagined and it's composition is impacted by factors such as obesity and Cesarean delivery. Could it be that her microbes, rather than a mother's genes, are to blame for her obese children? Could it be that it's the absence of certain breast microbes in mothers who underwent Cesarean delivery that's responsible for the increased risk of asthma in their children?

In the past, proponents of the Precautionary Principle have tended to deny that hormesis exists. As the evidence in favor of hormesis has gotten harder to ignore their narrative has changed to (1) it hasn't been shown to be the sort of general phenomenon that would permit an inference of hormesis for all toxins and (2) too little is known about the mechanism to permit its incorporation in to risk models.

Mass Torts Made A Surprise Appearance In The MOOC I'm Taking

David Oliver — Sun, 06 Jan 2013 23:41:15 -0600

The MOOC (massive open online course) is PH207x: Health in Numbers: Quantitative Methods in Clinical and Public Health Research and it's fabulous. The lectures are concise (they're often called "chunks") and for the most part do a fine job, step by step, of explaining the statistical tools for testing hypotheses and making inferences from data collected in public health research. The couple of lectures that left me scratching my head were quickly resolved by a visit to Khan Academy for a more detailed discussion of the math.

In the sixth week, while discussing hypothesis testing with two or more samples, Dr. Marcello Pagano (a great lecturer btw) decided to use the Bendectin story to illustrate the harm done when science is ignored. The story related was that a famous plaintiff lawyer had planted the story of "hideous birth defects" caused by Bendectin in the National Enquirer and that the cost of defending the ensuing litigation caused the maker to remove the drug from the market - with dire consequences for pregnant women with severe morning sickness. So far so good. But then Dr. Pagano put up a slide stating: "Not one court case lost."

Cases were lost and money was paid (even though the science behind the claims was fraudulent). So finally to the point of this post.

Just a month ago an expert (and former academic) that we're using on a case said, upon being told of the plaintiff's claims, (I'm paraphrasing) "I had no idea this went on. A professor at the law school told us that juries and courts are really good about separating shabby science from sound science and so plaintiff lawyers will only file meritorious claims." Right. Just ask Dow Corning how it worked out with their silicone breast implants. Anyway, there's a group of people out there spreading the story that Daubert isn't needed because junk science in the courtroom was never really a problem. Sadly, some really smart people believe them.

Be Careful What You Wish For When You Wish For A Standardless Standard Like Lohrmann

David Oliver — Sat, 29 Dec 2012 01:00:40 -0600

As promised we're weighing in on Holcomb v. Georgia Pacific, et al - the most recent effort by a court, this time Nevada's supreme court, to paint a fig leaf over the judicial embarrassment that is modern asbestos litigation.

To recap, by 1989 (twenty years after Clarence Borel filed the complaint that launched the mother of all mass torts) the litigation appeared to be winding down. The personal injury (and school abatement) cases had bankrupted most of the companies that once had constituted the lion's share of the asbestos industry. A little over a decade later Pittsburgh-Corning, and soon thereafter Owens-Corning, sought bankruptcy protection. With that, market share-wise, the vast majority of the American "asbestos industry" had been put out of its misery. The remaining defendants (with the exception of Owens-Illinois which serendipitously exited the business back in 1958) were each responsible for a microscopic share of the asbestos used in the United States. Surely the end was near. Instead, because courts tended to create special causation rules in asbestos cases which conflated risk with causation and because those same courts assiduously avoided the question of whether some risks were so small that liability could not fairly be predicated on them, the litigation continued unabated.

Drowning defendants, desperate for anything that might help against the "every asbestos fiber poses risk, mesothelioma is the actualization of risk, therefore plaintiff's meso was caused by every fiber" sophistry that goes on down at the courthouse, often grab for Lohrmann v. Pittsburgh Corning . It's an 80's era asbestosis case holding, essentially, that two to three weeks of work cutting/applying Unibestos wasn't enough to impose liability on Pittsburgh Corning (though things went less well for the other defendants). There was no discussion of dose, nor of the relative potency of amosite nor of the risk posed by each of the defendants' products. Instead, Lohrmann accepted proof of frequent, regular and proximate exposure to a defendant's asbestos-containing product as a proxy for a quantitative assessment of exposure and thereby risk or whatever other consideration drove the court's approximation of the line between de minimis and non-de minimis exposures. But how frequent is frequent? How regular is regular? How proximate is proximate?

"It's better than nothing", one assumes the Holcomb defendants thought when they asked the Nevada Supreme Court to adopt the so-called Lohrmann standard. So how did it work out? Would plaintiff's testimony that exposures were "numerous" be sufficiently regular and frequent? Would working "around" a joint-compound user proximate enough?

According to the Nevada Supreme Court it's enough. And thus the problem with Lohrmann.

Today there's an extensive peer-reviewed literature demonstrating the typical distribution of exposures resulting from almost every conceivable use of asbestos, and a well refined risk model for asbestos-induced mesothelioma that lets anyone estimate the risk associated with the exposures described by a plaintiff and his co-workers. Coming up with a plausible range of exposure is neither too hard nor too expensive. Down here in Texas plaintiffs need someone to testify to a supportable dose range and a review of their experts' bills reveals a typical cost of about $2500 - a fraction of amount they spend on their experts who testify about the history of the use and recognition of the hazards of asbestos exposure.

So would it be too much of a burden to make plaintiff state a supportable range of his likely exposure? The Nevada Supreme Court thought so - though it didn't say how much it thought such an estimation would cost, nor how much is too much. It also bought the straw man argument that states like Texas which actually ask "about how much" are instead asking the impossible to answer "state precisely how much". From there the court went on to conclude that an inference of causation may reasonably be drawn from expert testimony that infrequent and low level exposures can cause mesothelioma plus evidence that the plaintiff sustained "numerous" instances of working "around" the asbestos-containing product. And sure enough, that's the gist of Lohrmann.

As for why defendants continue to go from court to court demanding the adoption of the Lohrmann standard, that remains a mystery. As for why plaintiffs hate to say "about how much", we already know the answer - compared to the risks posed by some of the products now caught up in the litigation, taking a shower is a death defying feat.

2013 will be the International Year of Statistics. Maybe it'll be the year more judges and lawyers come to appreciate how much better our decision-making can be when we have the courage to demand the data and to accept what it implies.

Bostic Asks For A Rehearing And "But For" Causation Takes Center Stage

David Oliver — Fri, 14 Dec 2012 15:59:05 -0600

Kudos (despite their being a longtime and bitter foe) to Baron & Budd for doing a masterful job of demonstrating the flawed thinking behind, and absurd results implied by, Georgia-Pacific v. Bostic. As you know, the Texas Supreme Court recently denied Bostic's Petition for Review. Now Bostic has filed a Motion for Rehearing and this time has narrowed her focus to the appellate court's interpretation of "but for" causation. Here's our take.

Five years ago our supreme court authored a real gem in Borg-Warner v. Flores. The opinion recognized that while a plaintiff must prove her asbestos-related illness would not have arisen "but for" the asbestos to which she was exposed, she need not prove which among multiple sources of that exposure was the cause. However, she still had to prove that each source of exposure on which she sought to impose liability was a substantial factor in causing her illness. Because some of the exposures for which liability is nowadays sought to be imposed are very low, the court further required plaintiff to demonstrate not precisely, but rather within a reasonable range, the amount of asbestos to which she was exposed. As the risk imposed is the measure of legal causation, or substantial factor causation, and as dose is the measure of risk, the requirement allowed courts to make informed decisions when deciding, as a matter of public policy, which exposures could fairly result in the imposition of liability and which were de minimis and so could not. Then came Bostic.

The appellate court in Bostic decided that the "but for" test applies not to the question of whether the plaintiff's illness was caused by asbestos but to the question of whether each subset of the causative dose was the cause of plaintiff's illness. The problem with the latter formulation becomes clear when you consider its application to any case where there are multiple sufficient causes of plaintiff's injury. When a "but for" causation test is applied in such a case defendant X says "plaintiff can't prove she wouldn't have developed her illness 'but for' my product because defendant Y would have caused it anyway" and defendant Y turns around and argues that it couldn't have been the "but for" cause of plaintiff's illness since defendant X would have caused it anyway. The result would be that none of the defendants could be found liable no matter how tortious their conduct.

As we've previously written, Texas' supreme court long ago recognized the absurd and unjust results that would follow from applying the "but for" test to each cause within the set of sufficient causes of a plaintiff's injury. It reaffirmed that view in Flores. Anyone interested in sound decision-making down at the courthouse should be hoping that the Texas Supreme Court reconsiders its decision not to look at Bostic. And if you're a big fan of Flores then you'll want Bostic's take on "but for" causation reversed. Maybe then other courts, like Nevada's Supreme Court which recently decided Flores leans too heavily in favor of defendants, will see it for what it is - not an algorithm that dictates outcomes but rather a way for courts to intelligently draw a line around the outer limit of liability. A line drawn according to each court's assessment of its impact; purely from a public policy perspective.

Next up: we'll put up a post on Holcomb v. Georgia Pacific, that Nevada Supreme Court opinion referenced above. Preview: as we've been saying for years defendants who run around asking for Lohrmann-esque standards need to think about what they're wishing for.

Gut science: truth is stranger than fiction

Joseph M. Schreiber — Fri, 14 Dec 2012 12:39:40 -0600

A new paper by UT Southwestern Medical Center scientists presented a study which found that bacteria in your gut are competing through bacteriophages – unique predatory bacteria harboring viruses – for dominance and greater access to nutrients. While we knew that bacteria in the gut outnumber the number of cells in the body and have a huge impact on our immune system, we were unaware that they were unleashing viruses on each other. Further studies are planned to look into how to harness this fight for the betterment of us, not just the bacteria. See a story here.