This White Paper presents an overview of several Daubert-related issues as they relate to the development and presenting of engineering and scientific evidence by an Expert Witness in a court of law for a Class Action Lawsuit. My commentary is based on actual court cases and other information.
Federal and state courts have a wide spectrum of rulings with respect to the conclusion that Daubert applies or does not apply to expert testimony at the certification stage of a Class Action Lawsuit. In some cases, the court may require a complete, conclusive Daubert inquiry, whereas in other cases the court may require only a limited, focused Daubert inquiry.
Importantly, Daubert challenges cut both ways, and may be used by both the Plaintiffs and Defendants to their benefit. Typically, however, Daubert inquiries tend to benefit the Defendants in Class Action Lawsuits.
II. Daubert Challenges in General
The general challenges that can be made by the Defendants in a Class Action Lawsuit with respect to determining the admissibility of engineering and scientific evidence include the following:
a. Whether the expert’s methodology can be proven wrong
b. Whether the expert’s methodology has undergone publication and peer review
c. Whether the expert’s methodology has a known or potential rate of error
d. Whether the expert’s methodology is generally accepted in the engineering or scientific community
e. Other challenges
These challenges generally fall under the category of “the reliability” of expert findings and opinions:
a. Failure to consider all relevant data
b. Failure to use or not use certain data
c. Failure to conduct sufficient relevant testing
d. Failure to evaluate data in an unbiased manner
e. Failure to show proximate cause of injury
f. Failure to use correct assumptions and facts
g. Failure to justify differences in technical assessment methods
h. Failure to be objective in data selection (i.e., cherry-picking data)
i. Failure to be objective by presenting findings and opinions
j. Other challenges
In sum, these challenges point to the necessity of “appropriately connecting the forensic dots” from the source of the contamination to the receptor vis-à-vis clear, defensible fate and transport mechanisms, and ultimately to probable injury to health or property.
III. Daubert Challenges in Specific
The specific challenges that can be made by the Defendants in a Class Action Lawsuit include the following:
a. Development, Documentation, and Use of a Definitive Overall Project Methodology
For many “simple” litigation efforts, the expert is frequently asked by the Legal Team to simply go out and collect a “few” samples and have them tested. The Legal Team will take it from there, and attempt to outmaneuver the opposition. While such data collection is not necessarily scientifically incorrect, the failure to prepare and document a formal Field Work Plan may present problems in terms of conducting a limited or biased field program which results in an incomplete, limited, or even erroneous data base. While the formalization of this element of work does not guarantee valid data, it does force the Legal Team and Scientific Team to think through the entire litigation process in terms of data collection and evaluation.
b. Development, Documentation, and Use of Definitive Field Investigation Methods
Again, for many “simple” litigation efforts, the expert is frequently asked by the Legal Team to simply go out and collect a “few” samples and have them tested. While the testing results may be correct, the process and documentation elements of this process are of equal importance as the data itself. The environmental matrix, field methods, field equipment, sampling timeframe, sampling locations, sampling depths, sample preservation, chain-of-custody, QA/QC, etc. are critical in the sample collection and evaluation process. Errors in the field must be avoided since they can bias or even invalidate the results received from the analytical chemistry laboratory. Again, the preparation of a Field Work Plan does not guarantee valid data, but it does force the Legal Team and Scientific Team to think through the entire litigation process in terms of data collection and evaluation.
c. Identification of Specific Regulatory Clean-up / Action Levels or Health Injury Levels
In many respects, the target clean-up levels, action levels, remediation levels, restoration levels, etc. are the linchpin to the Lawsuit. Given sufficient investigation, the presence of at least some level of a contaminant can likely be confirmed through sampling and analysis. But, as Paracelsus once said, “The dose makes the poison.” The mere presence of a toxic chemical at the receptor site does not equate to personal health injury or diminution of property value. Is the exposure and dose sufficient to cause injury? Does it support nuisance and trespass claims?
d. Assessment of the Site’s Compliance History
The compliance history of the site frequently becomes an issue of debate. Does compliance imply no injury or impact? Can injury occur even in the presence of compliance with regulations? Can someone driving under the speed limit still cause an accident?
e. Determination of the Nature of any Alleged Contamination
A primary task of the expert is to determine the presence and nature of any alleged contamination in soil, sediments, surface water, groundwater, the atmosphere, household dust, ecosystems, etc. This includes identifying the entire suite of toxic chemicals that are present at the receptor site, regardless of concentration or location.
f. Determination of the Extent of any Alleged Contamination
A primary task of the expert is to determine the extent of any alleged contamination in soil, sediments, surface water, groundwater, the atmosphere, household dust, ecosystems, etc. This includes identifying the entire suite of toxic chemicals that are present on the site, regardless of concentration, and demarcating the lateral extent and depth of impact (i.e., a “three dimensional analysis”). Without this information, it is virtually impossible to predict injury to human health and the environment.
g. Preparation of an Exposure Assessment
An exposure assessment correlates the nature and extent of contamination to the location of the impacted receptors, both humans and the environment, in general. Humans may be exposed to contaminants by dermal contact, inhalation, and ingestion. The routes of exposure of each toxic chemical identified must be evaluated to assess these three exposure vectors. Importantly, exposure can increase or decrease over time.
h. Preparation of a Health or Ecological Based Risk Assessment
Health injury can be evaluated by toxicologists and medical doctors. Ecological injury can be evaluated by scientists. In assessing health injury, the impacted population (Plaintiffs) should be examined by qualified medical experts to identify their health conditions, and to determine if such conditions are reasonably associated with the types and concentrations of the contaminants found at or near the point of exposure, and the three exposure vectors. And, disease latency periods must be considered.
i. Receptor Analysis
An analysis of the human receptor is critical in terms of exposure vectors, concentration of exposure, timing of exposure, duration of exposure, and associated injury.
j. Description of Environmental Background Conditions
Background conditions are defined as naturally-occurring levels of the contaminants of concern in the environment prior to any pollution by man-made facilities or actions. For example, toxic metals are inherently present in many surface soils and groundwaters, but the mere presence of these chemicals near the receptors does not necessarily mean that the chemicals came from the Defendant’s site or operations. The presence of Arsenic, a potent carcinogen, is an example – Arsenic is naturally occurring at high concentrations in many soils and groundwaters.
k. Identification of Upgradient or Nearby Sources
Frequently, the a priori determination is made that all contamination found at a given receptor location is associated with a certain source (e.g., the Defendant’s site). Some contaminants can migrate hundreds to thousands of feet from their actual source. Thus, it is important to differentiate the actual source from the site under investigation. This is especially true of contaminants found in the atmosphere, surface water, and groundwater which can be transported long distances, sometimes hundreds or thousands of feet, or even miles from the source.
l. Identification and Evaluation of Confounder Conditions
Confounder conditions and alternative explanations of the detected contamination should be fully evaluated as part of the overall scientific process, especially in the absence of definitive ambient information. For example, a receptor complains of asthma. Is the source the Defendant’s site, or area-wide pollen, or dust or cigarette smoke within the house?
IV. Examples of Daubert Challenges and Responses
The following are examples of Daubert challenge topics that have been raised to 3TM Consulting, along with responses:
a. The protocols used to collect the samples are not appropriate.
Response: 3TM uses equipment and protocols consistent with the equipment vendors; the EPA, DOD, DOE, USDA, USGS, ASTM, and other organizations; standard industry practice; good engineering practice; and common sense.
b. The use of household dust as a metric for human health risk assessment is not appropriate.
Response: EPA has sampled household dust at many sites and has required the remediation of houses contaminated with household dust. There is a significant amount of information in the scientific public domain on the topic.
c. The use of household attic dust as a metric for human health risk assessment is not appropriate.
Response: Many parties have sampled household attic dust at many sites. There is a significant amount of information in the scientific public domain on the topic.
d. The concentration of any given contaminant in surface soil, household dust, and attic dust are not the same.
Response: The concentration of any contaminant in soil can be less than, equal to, or greater than the corresponding concentration in either household dust or attic dust. Concentrations of contaminants in attic dust can be either lower or higher than corresponding concentrations in household dust. There is no set “rule per se.”
e. Contaminants in the environment degrade quickly and thus do not pose a health risk.
Response: Some do, some do not. There is significant evidence in the scientific public domain that indicates that most contaminants degrade slowly over time. This is particularly true of contaminants in household dust and attic dust which are not exposed to the same physical, chemical, and biological degradation processes as traditional environmental media such as soil, water, etc.
f. The sample was tested for Chemical X, and it was either absent or found at only a very low level.
Response: The concentration could in reality be low, or this could be a result of biased sampling locations. Or, it could be a residual from a much higher concentration from an impact years ago. Further, even when certain chemicals degrade, they degrade into degradation daughter chemicals which may be persistent, or which may have an equivalent, lower, or higher toxicity than the parent compounds.
g. Degradation daughter chemicals are of no concern since they are not regulated.
Response: This is not true. Degradation daughter chemicals can have a toxicity that can be lower, higher, or the same as the parent compound. Health injury or environmental impact can occur from degradation daughter compounds, even when the parent compound is absent.
h. The contaminant levels found at the site are low and do not pose a threat to human health.
Response: The latency period of many diseases can be years or decades. The “snapshot” sample that was collected and tested may not be indicative of the much higher levels that were present years ago which impacted the receptor.
i. The half-life of any chemical in the environment is dictated by studies conducted and memorialized in scientific literature.
Response: This may or may not be true. I have found many instances that contaminants with low half-lives persist at a site an order of magnitude longer than was expected. This is because the contaminated matrix was isolated from the physical, chemical, and biological processes that normally degrade chemicals.
j. The selection of sampling locations by the Plaintiff’s attorney invalidates the scientific work.
Response: While it can be argued that attorneys should or should not select sampling locations, the collection and testing of a sample does not invalidate its results or importance. Even random sampling events that are totally devoid of any pre-determined planning can yield important evidence, if the samples are collected and tested appropriately.
k. The information derived from the collection of only a few environmental samples can be extrapolated or interpolated to construct a large area of impact.
Response: Depending on the number and location, such extrapolation or interpolation may or may not have scientific merit. Generally, the meaningfulness of a data base is increased by a higher number of individual samples.
l. The Defendant went out and collected some samples, but they were all “non-detect.”
Response: If the sample collection process was not biased, then the area tested may not be contaminated. On the other hand, it is easy to collect samples with an a priori intent of not being representative of any contamination. This is the old Urban Cowboy method of “looking for contamination in all the wrong places…”
m. If one does not perform a classical statistical analysis of all the data, then that data is not valid.
Response: Someone once appropriately said, “Statistics can be used to prove anything, even truth.” The number of samples, the spatial distribution of samples, and other factors may influence the statistical significance of the field work as a whole. However, reverting to classroom statistical calculations to prove or disprove contamination may prove fruitless. A classical example: Three dogs were found dead in New York City the night of March 10. Statistically, that is not noteworthy. Three dogs were found dead in New York City the night of March 11, and each had a missing ear. Statistically, that may not be noteworthy, but it certainly begs for further investigation.
n. Standard statistical practice calls for the discarding of anomalously high measurements made in the environment.
Response: Simply because a high level of a contaminant was detected at a certain location does not mean that the value is in error or should not be used in any assessment. On the contrary, it could indicate major contamination, even in a localized area. This is akin to an offensive football team ignoring the 800 pound defensive lineman simply because he is anomalously big.
o. The highest level of contaminant observed is the highest level at the site which was sampled.
The probability of identifying the location with the absolute highest concentration of contaminant at the site sampled is statistically near zero. Move over one foot, and it could be higher.
p. High concentrations of chemicals at a site automatically implies a health hazard.
Response: High concentrations may or may not imply a health hazard. The routes of exposure (dermal contact, inhalation, and ingestion) must be evaluated. If there is not a completed pathway between the chemical and the receptor, then no health impact can occur. Notwithstanding, a potential for exposure may still exist.
q. The identified contaminants are ubiquitous in the world, and therefore not associated with the site.
Response: Some are. Some are not. Frequently, the argument is made that “Dioxins are ubiquitous in the world…” in an attempt to convince the Court that the Dioxins found at the site are of no consequence. However, the scientific issue is the concentration, fingerprint, and locations of where they were found, and their probable impact to humans.
r. Different laboratories were used for different phases of the work; therefore, the results are not valid.
Response: Different credible laboratories may be used throughout a lengthy project for a number reasons, including errors uncovered in early phases of the work, the inability of the lab to meet turnaround time deadlines, attempting to confirm findings, cost, schedule, logistics, etc.
s. Not all of the samples that were collected were tested; therefore, the party collecting the samples is trying to hide something.
Response: Frequently, a large number of samples are collected, and some are tested, and some are archived for whatever reason, including schedule delays, desire to minimize expenses, etc. This practice does not automatically imply that the samples that were tested lack validity.
t. Fingerprinting of the chemicals are required to support a legal position.
Response: Chemical fingerprinting to rule in or rule out possible sources of the contamination is frequently used in Lawsuits. In some Lawsuits, chemical fingerprinting may provide valuable evidence as to origin of the contamination. In other Lawsuits, chemical fingerprinting may not be necessary if the source is readily identifiable.
u. A formal confounder study was not conducted.
Response: The use of a formal confounder study may or may not be material to the Lawsuit. If multiple sources of the contamination are suspected, then such a study may prove useful. On the other hand, some sites are isolated and thus the obvious source of any contaminants that may be found off-site is associated with the operation of that site.
v. A formal medical injury control group study was not conducted.
Response: Control studies are expensive but useful in determining medical injury causation in many lawsuits. The absolute requirement for such a study is a function of the class and their alleged health injuries.
w. Occam’s Razor always prevails.
Response: This is always true except for the following circumstance: You walk through the woods and find a dead skunk. A few feet away is a stick. You conclude that the stick killed the skunk.
V. Using Daubert to Defend or Defeat Class Action Lawsuits
Plaintiffs seeking to establish Class Certification under Federal Rule 23 must prove:
a. An adequate class definition
g. Inadequacy of separate adjudication
These elements point to the necessity of a significant burden of proof by the Plaintiffs. This burden of proof may be straightforward for some lawsuits, such as a situation in which a large group of individuals were killed or injured by a single act, such as an explosion at a refinery. This burden of proof becomes very complex for other lawsuits, such as a situation in which a community is allegedly exposed to long-term historical releases of toxic chemicals from a site that has morphed over time.
a. The site may have operated for years or even decades.
b. The site may have housed a number of similar or distinct operations over a span of years or decades.
c. The site may have had different features, operations, and sources over a span of years or decades.
d. The site may have released different toxic chemicals over a span of years or decades.
e. The regulatory environment and requirements have evolved over the past several decades.
f. The compliance history of the site may have changed significantly over a span of years or decades.
g. The receptor population and associated locations may have changed over a span of years or decades.
h. The lifestyle of the receptor population may have changed over a span of years or decades.
i. Other nearby sources of pollution may have emerged or shut down over a span of years or decades.
All of these parameters in toto affect emissions and releases, fate and transport of contaminants, exposure scenarios, risk scenarios, and potential for health injury or diminution of property value.
In analyzing probable health injury, the first major tenet is that it will be virtually impossible to calculate exposure doses for every receptor over a long period of time. For a potential class of (say) 10,000 receptors that would equate to 10,000 separate studies, each of which is comprehensive.
The second tenet is that every individual will likely have a different exposure scenario, history, and dose. This dose, in theory, points to variable potential for and severity of health injury.
The third tenet is that even extensive sampling may yield significant anomalies. That is, the concentrations of a contaminant measured at one physical location may materially differ from concentrations measured at an adjacent locations. There are many technical reasons for this. The point is: one must carefully evaluate the available data in forming findings and opinions based on interpolation and extrapolation.
The fourth tenet is that individual lifestyles of the 10,000 plaintiffs confound health injury assessment. For example – Person A and Person B both live at the same address, but Person A follows a healthy lifestyle whereas Person B does not. If they are exposed to identical levels of the same contaminant, are their health injuries identical?
The fifth tenet is that medical experts sometimes have difficulty in distinguishing health injury from histrionic behavior, especially in a legal world. Some medical conditions such as skin dermatitis are easy to evaluate. Asthma may or may not be. This is essentially “differential diagnosis.”
My favorite war story about all of this is the following. During a class action lawsuit, I was asked to sample indoor household dust at a large number of plaintiff residences, and test the dust samples for a specific suite of contaminants. As luck would have it, the GC trace on some of the lab data showed the presence of cocaine in an indoor sample. The jury would have a tough time empathizing with the plaintiff in terms of exposure to toxic chemicals from a nearby plant. Further, in deposition, the plaintiff testified to drinking 12 beers per day and smoking 3 packs of cigarettes per day. So much for Daubert…