(1). Introduction. Cases involving toxic substances often pose difficult problems of proof of factual causation. These problems can also arise in cases involving activities that may cause disease, such as continued repetitive motion. Sometimes it is difficult to prove which defendant was connected to the toxic agent, see Comment p, or whether an adequate warning would have prevented the plaintiff's harm, see Comment b. The special problem in these cases, however, is proving the connection between a substance and development of a specific disease. In all of these cases, the requirement to prove factual causation remains the same; the plaintiff must prove it by a preponderance of the evidence, and the standards for factual causation set forth in §§ 26- 27 continue to apply.

In most traumatic-injury cases, the plaintiff can prove the causal role of the defendant's tortious conduct by observation, based upon reasonable inferences drawn from everyday experience and a close temporal and spatial connection between that conduct and the harm. Often, no other potential causes of injury exist. When a passenger in an automobile collision suffers a broken limb, potential causal explanations other than the collision are easily ruled out; common experience reveals that the forces generated in a serious automobile collision are capable of causing a fracture. By contrast, the causes of some diseases, especially those with significant latency periods, are generally much less well understood. Even known causes for certain diseases may explain only a fraction of the incidence of such diseases, with the remainder due to unknown causes. Causal agents are often identified in group (epidemiologic) studies that reveal an increase in disease incidence among a group exposed to the agent as compared to a group not exposed. Biological mechanisms for disease development—i.e., a series of causally linked physiological changes from exposure to disease development—are frequently complicated and difficult to observe. Science continues to develop a better understanding of the biological steps in the development of diseases, but current knowledge in this respect is considerably more modest than for traumatic injury. As a consequence, courts in toxic-substances cases often must assess various alternative methods proffered with regard to factual causation.

Over the past several decades, courts have devoted a great deal of energy to the issue of causation in toxic-tort cases. Causation is a question of fact normally left to the jury, unless reasonable minds cannot differ. Appellate or trial-court review of jury findings affect the allocation of power between judges and juries. Until the early 1980s, a qualified expert witness's opinion that a toxic agent was a factual cause of the plaintiff's disease was treated as sufficient evidence. A few celebrated cases and case congregations, such as the Agent Orange and Bendectin litigations, led some courts to distrust juries' ability to resolve cases based on conflicting expert-opinion evidence. Courts began to scrutinize the scientific evidence employed and to examine carefully the bases for an expert's opinion on factual causation. Some courts then tried to develop bright-line rules based on science for adequate proof of factual causation. The high-water mark for this overreliance on scientific thresholds occurred in the Bendectin litigation when one court announced a blanket rule that a plaintiff could not make out a sufficient case without statistically significant epidemiologic evidence.

These courts may be relying on a view that “science” presents an “objective” method of establishing that, in all cases, reasonable minds cannot differ on the issue of factual causation. Such a view is incorrect. First, scientific standards for the sufficiency of evidence to establish a proposition may be inappropriate for the law, which itself must decide the minimum amount of evidence permitting a reasonable (and, therefore, permissible) inference, as opposed to speculation that is not permitted. See Comment b. Second, scientists report that an evaluation of data and scientific evidence to determine whether an inference of causation is appropriate requires judgment and interpretation. Scientists are subject to their own value judgments and preexisting biases that may affect their view of a body of evidence. There are instances in which although one scientist or group of scientists comes to one conclusion about factual causation, they recognize that another group that comes to a contrary conclusion might still be “reasonable.” These scientists' views reflect their scientific experience outside the courtroom. They may have different views about specific instances of conflicting scientific testimony in a courtroom. Scientists' judgments about causation outside the legal context may also be affected by the comparative costs of errors, as when caution counsels in favor of declaring an uncertain agent toxic because the potential harm it may cause if toxic is so much greater than the benefit forgone if it were not introduced. Courts, thus, should be cautious about adopting specific “scientific” principles, taken out of context, to formulate bright-line legal rules or conclude that reasonable minds cannot differ about factual causation.

This Comment is necessarily general. It addresses how methods of proof for traumatic injuries and for diseases may differ. Toxic-substance cases often involve statistical and group-based scientific studies that courts seldom confronted when the Restatement Second of Torts was published. Toxic agents and the diseases they cause differ, and methods of proof may vary accordingly. The law continues to evolve as courts are confronted with a variety of different circumstances related to different toxic substances, different diseases, and the varieties of available evidence.

Scientific methods may advance in the future to better facilitate causation determinations for individuals, thereby obviating the need for statistically based group studies. While such techniques are largely unavailable today, dramatic advances in microbiology, genetics, and related fields have been made. These developments may produce new forms of evidence to which courts will adapt legal treatment of proof of causation. This Comment is necessarily based on scientific methods available at the time it was published.

Proof of causation often involves the admissibility of expert-witness opinions. Admissibility is governed by the law of evidence, and nothing in this Comment addresses that law. However, admissibility cannot be determined without reference to the substantive law. Moreover, courts may be required to examine scientific evidence when it is offered to prove agent-disease causation. That examination may occur either in the admissibility determination or in the determination whether the evidence is sufficient to meet the burden of production. These usually are separate issues and are subject to different legal standards. Courts, however, sometimes conflate these issues in the process of determining whether there is an adequate basis for an expert's opinion. The requirement of causation, the elements of agent-disease causation that are sometimes required when group studies are employed as proof, and the sufficiency of the evidence to meet the burden of production on causation are matters of substantive tort law, and they are addressed in the Restatement.

Most causation issues are resolved under the “but-for” standard for factual cause. See § 26. The plaintiff must prove by a preponderance of the evidence that, but for the defendant's tortious conduct with respect to the toxic substance, the plaintiff would not have suffered harm. When group-based statistical evidence is proffered in a case, this means that the substance must be capable of causing the disease (“general causation”) and that the substance must have caused the plaintiff's disease (“specific causation”). In other cases, when group-based evidence is unavailable or inconclusive, and other forms of evidence are used, the general and specific causation issues may merge into a single inquiry. In any case, plaintiff's exposure to the toxic agent must be established.

Thus, courts often address “exposure,” “general causation,” and “specific causation.” Nevertheless, these items are not “elements” of a plaintiff's cause of action, and in some cases may not require separate proof. So long as the plaintiff introduces admissible and sufficient evidence of factual causation, the burden of production is satisfied. A court in a particular case may conclude that reasonable minds cannot differ about proof of factual causation under the general test because reasonable minds cannot differ on whether the plaintiff was exposed to the agent, whether the agent is generally capable of causing the disease, or whether the agent caused the plaintiff's disease in the specific case. These categories function as devices to organize a court's analysis, not as formal elements of the cause of action.

(2). Exposure to the agent. In evaluating factual causation, one issue that may arise is whether the plaintiff was exposed to the substance. Three primary means of exposure to toxic substances include inhalation, absorption, and ingestion, but others exist, such as injection or a fetus's transplacental exposure to agents in the mother's body. Often the method of exposure is critical to the type or extent of risk.

Exposure is frequently disputed in occupational-disease cases and hazardous-waste cases, while it is less often an issue in pharmaceutical cases. Proof of exposure may entail relatively straightforward historical facts, such as the presence of asbestos at the plaintiff's workplace or whether the plaintiff took a prescribed drug, or it may require complicated scientific evidence, such as dispersion modeling, to determine how and where the substance was transported. The latter form of evidence is often required in airborne- or groundwater-pollution cases. The intensity and duration of exposure (the “dose”) affects the magnitude of the risks posed and the likelihood of causation.

(3). General causation. “General causation” exists when a substance is capable of causing a given disease. The concept developed because a prominent form of scientific methodology investigates causation on a group basis and therefore addresses whether an agent causes an increased incidence of disease in the group being studied. These studies proceed by comparing the incidence of disease in a group that has been exposed to the agent with the incidence of disease in a group of unexposed persons. The latter group's disease, thus, is attributable to causes other than the agent being studied. Traumatic-injury cases, by contrast, do not require this form of evidence because other causes that might explain the injury are absent, and we have a reasonably good understanding of the causal mechanisms involved from trauma to injury.

Occasionally, biological-mechanism evidence is sufficiently developed to prove general causation. More frequently, however, the evidence consists of scientific studies comparing the incidence of disease in groups of individuals (epidemiologic evidence) or animals (toxicologic evidence) with different levels of exposure. When a study finds a difference in the incidence of disease in the exposed and unexposed groups, an “association” exists between exposure and disease. Another type of epidemiologic study compares the extent of exposure among those with and without the disease. These studies seek to identify toxic substances at the aggregate population level—by finding a higher incidence of a disease in a group exposed to the substance (an “association”).

Even when epidemiologic studies find an association between a substance and a disease, further analysis is necessary before a causal conclusion can be drawn. Scientists first systematically gather all of the studies that have been conducted and that are relevant to the causal question being investigated. When multiple studies exist, they are synthesized, either qualitatively in a review or quantitatively with a method known as meta-analysis. However, reasons may exist for disregarding or giving less weight to one or more of the available studies. If an association is found, epidemiologists use a number of factors (commonly known as the “Hill guidelines”) for evaluating whether that association is causal or spurious. A spurious association may be the result of study errors—such as biases (scientists use “bias” to mean a source of error rather than as a predisposition to testify or decide a matter in an improper way) and uncorrected confounding factors (alternative causes that are responsible for the association, rather than the agent under study)—or sampling error (the result of small numbers of subjects and random chance). Similarly, a study may incorrectly fail to find an association that exists, because of study errors, especially when the disease is rare and an insufficient number of subjects exist to reveal any relationship. Epidemiologists use statistical methods to estimate the range of error that sampling error could produce; assessing the existence and impact of biases and uncorrected confounding is usually qualitative.

Whether an inference of causation based on an association is appropriate is a matter of informed judgment, not scientific methodology, as is a judgment whether a study that finds no association is exonerative or inconclusive. No algorithm exists for applying the Hill guidelines to determine whether an association truly reflects a causal relationship or is spurious. Because the inferential process involves assessing multiple unranked factors, some of which may be more or less appropriate with regard to a specific causal assessment, judgment is required. For example, one of the Hill factors requires an assessment of other scientific evidence that bears on the causal relationship under consideration. In some cases, there may be a substantial body of other evidence, while in other cases there may be little. The saliency of other evidence of causation often entails considerable judgment. Thus, in some cases, reasonable scientists can come to differing conclusions on whether a body of epidemiologic data justifies an inference of causation. Similarly, reasonable scientists may, in some instances, disagree on whether the absence of an association is exonerative of the agent or is merely inconclusive.

Usually, other and unknown individual factors (causes) must concur with exposure to the agent for an individual to contract the disease. Group studies do not provide a basis for determining which individuals in a group suffer disease from exposure to the agent and which do not. More importantly, whenever other chemical, physical, or biological agents can produce the disease, group studies cannot distinguish which individual's disease was caused by exposure to a particular agent and which individual's disease was caused by another agent. So long as tort law adjudicates claims on an individual basis, specific causation requires attention even when general causation is established through the use of group studies.

Occasionally, courts have suggested or implied that a plaintiff cannot meet the burden of production on causation without epidemiologic evidence. Those cases often confronted a substantial body of epidemiologic evidence introduced by the defendant that tended to exonerate the agent as causal. Circumstances in individual cases, however, are sufficiently varied that almost all courts employ a more flexible approach to proof of causation—except in those cases with a substantial body of exonerative epidemiologic evidence. Epidemiologic studies are expensive and can take considerable time to design, conduct, and publish. For disease processes with long latency periods, valid studies cannot be performed until the disease has manifested itself. As a consequence, some plaintiffs may be forced to litigate long before epidemiologic research is available. Indeed, sometimes epidemiologic evidence is impossible to obtain, which may explain why neither the plaintiff nor the defendant is able to proffer supportive epidemiology. Thus, most courts have appropriately declined to impose a threshold requirement that a plaintiff always must prove causation with epidemiologic evidence, and, in some cases (as explained below), the evidence bearing on specific causation may be sufficient to pretermit the need to assess general causation.

(4). Specific causation. “Specific causation” exists when exposure to an agent caused a particular plaintiff's disease. Sometimes proof of specific causation is easy and collapses into proof of general causation, as when there are no alternative causal agents for a disease, and the disease is said to be a “signature” of the substance. In other cases, however, specific causation remains an issue even though general causation is established.

Scientists who conduct group studies do not examine specific causation in their research. No scientific methodology exists for assessing specific causation for an individual based on group studies. Nevertheless, courts have reasoned from the preponderance-of-the-evidence standard to determine the sufficiency of scientific evidence on specific causation when group-based studies are involved. Properly understood and applied, this analytical framework provides a reasonable basis for determining specific causation in the absence of more particularistic evidence about the cause of the plaintiff's disease.

Courts have reasoned that, when a group study finds that exposure to the agent causes an incidence in the exposed group that is more than twice the incidence in the unexposed group (i.e., a relative risk greater than two), the probability that exposure to an agent caused a similarly situated individual's disease is greater than 50 percent. Accordingly, when there is group-based evidence finding that exposure to an agent causes an incidence of disease in the exposed group that is more than twice the incidence in the unexposed group, the evidence is sufficient to satisfy the burden of production and permit submission of specific causation to a jury. In such a case, the factfinder may find that it is more likely than not that the substance caused the particular plaintiff's disease. The propriety of this “doubling” reasoning depends on group studies identifying a genuine causal relationship and a reasonably reliable measure of the increased risk. Expert witnesses may testify to specific causation based on the logic of the effect of a doubling of the risk and other considerations explained below that modify the probability of causation for a particular individual.

Additional considerations affect the propriety of determining the probability of specific causation based on the outcome of a group-based study. Depending on the state of the evidence about these additional matters, they may bear either on the sufficiency determination by the court or be relevant to the jury's determination. Thus, the extent to which the group-study outcome reflects the increased risk to the plaintiff depends on the plaintiff's similarity to those included in the group study. Relevant differences include whether: (a) the plaintiff was exposed to a comparable dose; (b) the plaintiff was not differentially exposed to other potential causes of the disease; and (c) the plaintiff has individual characteristics that might also bear on the risk of disease, such as age, gender, or general health, comparable to those in the study group.

The likelihood that an agent caused an individual's disease may be refined when there are independent, alternative known causes of the disease. The underlying premise is that each of these known causes is independently responsible for some proportion of the disease in a given population. Eliminating one or more of these as a possible cause for a specific plaintiff's disease increases the probability that the agent in question was responsible for that plaintiff's disease. Courts frequently refer to the elimination of other known causes for a plaintiff by employing the medical terminology of “differential diagnosis.” Assessing whether other causes can be ruled out (or in) as potential causes of a plaintiff's disease can provide probative evidence of specific causation. This technique is more accurately described as a “differential etiology.” It is most useful when the causes of a substantial proportion of the disease are known. Then, the presence (or absence) of these causes for the specific plaintiff affects the probability that the agent in question caused the plaintiff's illness. When the causes of a disease are largely unknown, however, differential etiology is of little assistance. Evidence about biological mechanisms may also alter the likelihood that exposure to the substance caused plaintiff's disease, either by ruling out other known causes or by explaining why the suspected agent is a more likely cause of the disease than others.

For all of these reasons, any judicial requirement that plaintiffs must show a threshold increase in risk or a doubling in incidence in a group study in order to satisfy the burden of proof of specific causation is usually inappropriate. So long as there is adequate evidence of general causation, courts should permit the parties to attempt to show, based on the sorts of evidence described above, whether the plaintiff's disease was more likely than not caused by the agent. Depending on the other factors detailed above, an increase of the incidence of disease less than a doubling may be sufficient to support a finding of causation, while in another case, even an increased incidence greater than two may not be sufficient. When the sufficiency of the evidence to meet the burden of production is at issue, courts should consider all of the evidence that bears on the matters discussed above and determine whether, in light of the general standard for sufficiency discussed in Comment b, the evidence would permit a reasonable jury to find that plaintiff's disease more probably than not was caused by exposure to the agent.

In most instances, differential etiology is not an appropriate technique for proving general causation. Nevertheless, in some limited circumstances courts have found that plaintiffs met their burden of proof of agent-disease causation without separate proof of general causation. Factors such as a good biological-mechanism explanation of how the agent could have caused the plaintiff's disease, a differential etiology ruling out other known causes, a reasonable explanation for the lack of general-causation evidence (and no contrary evidence of an absence of general causation), a short latency period and acute response, and the appropriate disease response to dechallenge (removal from exposure) and rechallenge (reexposure) to the agent, if combined and consistent, provide a persuasive basis for excusing the plaintiff from providing other proof of general causation.

(5). Multiple exposures and synergistic interactions. In some cases, a person may be exposed to two or more toxic agents, each of which is known to be capable of causing (general causation) the person's disease. The two agents may operate independently, in which case the incidence of disease in a group exposed to both will be additive—the excess incidence due to the first agent along with the excess incidence due to the second agent. Cases such as these present a relatively straightforward application of the principles set forth in Comment c(4). If the toxic agents are attributable to the tortious conduct of separate actors, courts then face the question whether to apply the rule developed for multiple exposures in asbestos cases. This rule permits finding each actor's asbestos products to which the person was exposed to be a factual cause of the person's disease. See § 27, Comment g. Alternatively, courts might employ the traditional rule, requiring proof of which of the multiple exposures was a cause of the harm. At least where the biological mechanism by which disease develops is unknown, the asbestos rule is quite analogous and attractive as a means for adapting proof requirements to the available scientific knowledge. Apportionment of liability among those actors held liable is based on the comparative—responsibility rules in Restatement Third, Torts: Apportionment of Liability §§ 1- 25. The alternative—the more traditional requirement of proof of which of the two toxic exposures was the cause of the disease—would require proof that does not exist, except on a probabilistic basis, as outlined in Comment c(4).

• Illustrations:
  • 1. Abby was exposed to two different solvents while working in a laboratory. Each solvent contained a toxic chemical; one contained brion, and the other contained choron. After developing a disease, myeplopia, several years later, she sues the manufacturers of each solvent, claiming that the manufacturers were negligent for including a toxic chemical in their solvents. Abby's evidence, presented by competent expert testimony based on valid scientific evidence, reveals that the increased risk of contracting myeplopia from the dose of brion to which she was exposed is insufficient to permit a finding of factual causation. Similarly, the increased risk of myeplopia from exposure to choron is insufficient to permit a finding of factual causation. However, Abby's evidence reveals that, while choron and brion operate independently (those exposed to both are only subject to an increased risk of the additive risks of each), the combined risk of contracting myeplopia due to exposure to both is sufficient to permit a finding of factual causation. Each of the manufacturers is subject to liability. See § 26, Comment c. Apportionment of liability between the manufacturers is governed by Restatement Third, Torts: Apportionment of Liability.
  • 2. Same facts as Illustration 1, except that competent evidence shows that choron exposure increases the risk of myeplopia by 10 times, as does brion exposure. Competent evidence also reveals that the mechanism by which myeplopia develops is different for choron exposure and for brion exposure and that exposure to one or the other, but not both, is the most likely explanation for Abby's myeplopia. Abby cannot prove, however, whether choron or brion caused her myeplopia. Pursuant to § 28(b), the burden of proof on agent-disease causation is shifted to the manufacturers of choron and brion.
  • 3. Same facts as Illustration 2, except that competent evidence reveals that choron and brion operate in precisely the same physiologic manner in the human body; they are interchangeable in their role in causing myeplopia. Exposure to each of choron and brion is a factual cause of Abby's myeplopia. See § 27, Comment g.

In some cases, as, for example, asbestos workers who smoke cigarettes, the two toxic agents together have a synergistic effect. This means that the excess incidence of disease among those exposed to both agents will be greater than the sum of the excess incidences found in those exposed to each separate agent. If the synergistic effect is sufficiently large, the excess incidence of disease due to the synergistic effect will be greater than the excess incidence due to each of the agents separately. In such circumstances, factfinders may infer that the combined exposure is a cause of the plaintiff's disease. This inferential process is similar to the one permitting a jury to find specific causation based on the increase in the incidence found from a general-causation study, such as those described in Comment c(4). Although the reasoning for synergistic agents differs from that for nonsynergistic agents, the outcome is similar if the synergistic effect of the interacting agents is sufficiently large.

However, identification of both of the synergistic agents as a cause of the disease does not end the inquiry. Many causes exist for a given harm. See § 26, Comment f. Only those causes attributable to tortious conduct are legally relevant in determining liability and apportioning liability for the plaintiff's harm. See Restatement Third, Torts: Apportionment of Liability § 26, Comment m. Thus, a natural condition, a genetic trait of the plaintiff, or a nonnegligent actor's conduct that are causes, in addition to a negligent actor's conduct, of the plaintiff's harm have no effect on the negligent actor's role as a cause of harm or on apportionment of liability. If more than one legally responsible agent is a cause of the plaintiff's harm, then apportionment of liability is based on comparative responsibility pursuant to Restatement Third, Torts: Apportionment of Liability §§ 1-25.

• Illustrations:

• • 4. Brett was occupationally exposed to asbestos for several decades. He also smoked cigarettes during approximately the same time period. Brett, who has developed lung cancer, sues Rossman, Inc., the manufacturer and supplier of the asbestos to which he was exposed, claiming that Rossman failed adequately to warn of the dangers of asbestos exposure. Brett provides competent expert testimony that, based on valid scientific studies, the dose of asbestos to which he was exposed increases the risk of contracting lung cancer by a factor of five (500%). The dose of cigarette smoke to which he was exposed increases the risk of lung cancer by a factor of 12 (1200%). However, the combined exposure to both asbestos and cigarette smoke increases the risk of lung cancer by a factor of 60 (6000%). Brett's evidence is sufficient to permit the factfinder to find that exposure to both asbestos and cigarette smoke were causes of his lung cancer. Because neither Brett nor Rossman claim that the smoking implicates tortious conduct, no apportionment of liability for Brett's lung cancer would occur, if the factfinder found in Brett's favor against Rossman.
  • 5. Same facts as Illustration 4, except that Rossman successfully persuades the factfinder that Brett's smoking constituted negligence on his part. Neither Brett nor Rossman alleges any tortious conduct by the cigarette manufacturers. Liability for Brett's lung cancer would be apportioned between Brett and Rossman based on comparative responsibility according to Restatement Third, Torts: Apportionment of Liability § 7.