Skip Navigation
Skip to contents

JPMPH : Journal of Preventive Medicine and Public Health



Page Path
HOME > J Prev Med Public Health > Volume 49(2); 2016 > Article
Special Article
Proving Causation With Epidemiological Evidence in Tobacco Lawsuits
Sun Goo Leeorcid
Journal of Preventive Medicine and Public Health 2016;49(2):80-96.
Published online: March 31, 2016
  • 144 Download
  • 1 Crossref
  • 5 Scopus

Gachon University School of Medicine, Seongnam, Korea

Corresponding author: Sun Goo Lee, SJD  1342 Seongnam-daero, Sujeong-gu, Seongnam 13120, Korea  Tel: +82-31-750-6925, Fax: +82-31-750-6925 E-mail:
• Received: January 9, 2016   • Accepted: March 15, 2016

Copyright © 2016 The Korean Society for Preventive Medicine

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License ( which permits unrestricted noncommercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

  • Recently, a series of lawsuits were filed in Korea claiming tort liability against tobacco companies. The Supreme Court has already issued decisions in some cases, while others are still pending. The primary issue in these cases is whether the epidemiological evidence submitted by the plaintiffs clearly proves the causal relationship between smoking and disease as required by civil law. Proving causation is difficult in tobacco lawsuits because factors other than smoking are involved in the development of a disease, and also because of the lapse of time between smoking and the manifestation of the disease. The Supreme Court (Supreme Court Decision, 2011Da22092, April 10, 2014) has imposed some limitations on using epidemiological evidence to prove causation in tobacco lawsuits filed by smokers and their family members, but these limitations should be reconsidered. First, the Court stated that a disease can be categorized as specific or non-specific, and for each disease type, causation can be proven by different types of evidence. However, the concept of specific diseases is not compatible with multifactor theory, which is generally accepted in the field of public health. Second, when the epidemiological association between the disease and the risk factor is proven to be significant, imposing additional burdens of proof on the plaintiff may considerably limit the plaintiff’s right to recovery, but the Court required the plaintiffs to provide additional information such as health condition and lifestyle. Third, the Supreme Court is not giving greater weight to the evidential value of epidemiological study results because the Court focuses on the fact that these studies were group-level, not individual-level. However, group-level studies could still offer valuable information about individual members of the group, e.g., probability of causation.
Recently, a number of lawsuits have been filed in Korea claiming tort liability for damages caused by allegedly unlawful acts of tobacco companies. The Supreme Court handed down decisions in lawsuits filed by plaintiffs who were smokers or descendants of deceased smokers (hereinafter “individual smoker lawsuit”) (Supreme Court Decision, 2011Da22092, April 10, 2014; Supreme Court Decision, 2011Da23422, April 10, 2014). On the other hand, the lawsuit filed by the National Health Insurance Service (Seoul Central District Court, 2014Gahap 525054, hereinafter “government agency tobacco lawsuit”) is still pending. The common issue in these cases is whether epidemiological evidence submitted by the plaintiffs can be used to prove the causal relationship between smoking and diseases as required by civil law.
The reason the plaintiffs attempted to prove causation with epidemiological evidence is because the types of evidence presented in other tort lawsuits were not useful in proving the causal relationship in tobacco lawsuits. Most compensatory damages in traditional tort lawsuits, e.g., an injury caused by a fistfight, result from a relatively recent event and have relatively simple causal links, usually explained by a basic understanding of physics [1]. However, proving the causal relationship in a tobacco case is much more complicated. Smokers can develop a disease 20 to 30 years after their first exposure to cigarette smoke. During this 20-year to 30-year period, various factors, including, among others, an unhealthy lifestyle, workplace stress, and environmental factors, may all play a part in the development of the disease.
In an effort to overcome these difficulties in proving causation, plaintiffs in tobacco lawsuits have been actively utilizing epidemiological research results. Epidemiology studies the distribution and determinants of diseases in a defined population group [2]. Hence, epidemiological evidence tends to have statistical and stochastic characteristics. The plaintiffs use epidemiological evidence to demonstrate a causal relationship between a risk factor and a disease by showing that there is a substantial probability of a certain disease developing in a certain group of people exposed to a risk factor, and that the plaintiffs themselves belong to that group and have developed the disease after being exposed to the risk factor. On the other hand, defendants in tobacco lawsuits argue that statistical and stochastic epidemiological evidence is not sufficient to prove causation in an individual lawsuit, and that the plaintiff in tort litigations must prove the specific causation between the defendant’s unlawful act and the plaintiff’s individual damages.
In this context, this study discusses the value of epidemiological evidence in the context of tobacco lawsuits. Part II explains the elements of tort liability under the Korean Civil Act and the definition of the burden of proof as well as providing some examples of court decisions where courts have applied relaxed standards to the burden of proof. Part III describes how the appellate court and the Supreme Court have dealt with the issue of causation in individual smoker lawsuits. Part IV reviews applicable legal rules concerning individual smoker lawsuits as well as the value of epidemiological evidence. Finally, Part V summarizes and explores ways to utilize epidemiological evidence in future tobacco lawsuits such as the government agency tobacco lawsuit.
Establishing Tort Liability Under the Korean Civil Act
Article 750 of the Korean Civil Act provides that “any person who causes losses to or inflicts injuries on another person by an unlawful act, intentionally or negligently, shall be bound to make compensation for damages arising therefrom.” Accordingly, the elements of tort liability are: (1) intention or negligence by the defendant; (2) an unlawful act; (3) plaintiff’s losses or injuries; (4) causal relationship between the unlawful act and the damages [3].
Regarding the definition of causal relationship, legal doctrine and precedents generally follow the theory of proximate cause [3]. Proximate cause is based on the standard of high probability barring contingency. In other words, proximate cause exists when the post-event, from an objective point of view, is likely to have been caused by the preceding event.
The Burden of Proof in Tort Lawsuits
Under the theory of Korean civil procedure law, the burden of proof implies risks or disadvantages a party bears when a certain fact necessary to establish the elements of legal rights has not been proven sufficiently [4]. In addition, “a court shall determine, by its free conviction, whether or not an allegation of facts is true, taking into account the whole purport of pleadings and the results of examination of evidence, on the basis of the ideology of social justice and equity and in accordance with the principles of logic and experience” (Article 202, Korean Civil Procedure Act).
The burden of proof in tort cases lies with plaintiff [4]. Accordingly, the plaintiff must produce evidence that can substantiate the existence of the elements of tort liability, including the existence of causation. Failure to do so leads to an adverse judgment that the right to recovery has not been established.
Relaxing the Burden of Proof of Causation
Proving the existence of causation could be complicated in many tort lawsuits. As mentioned in the introduction, losses or damages are often caused by a combination of multiple factors that take place over a long period of time. In such cases, imposing the burden of proof solely on the shoulders of the plaintiff could in effect infringe on the plaintiff’s right to receive any potential compensation. Therefore, efforts have been made to find ways to protect plaintiff’s rights to a greater extent, especially in cases of environmental and product liability lawsuits.

Environmental lawsuits

In environmental lawsuits where plaintiffs claim compensation for damages caused by air or water pollution, their exposure to a pollutant in question may have taken place in an indirect rather than direct way (e.g., Supreme Court Decision, 2000Da65666 [principal claim], 65673 [counterclaim], October 22, 2002). Unfortunately, modern science is not entirely capable of explaining the progression of environmental pollution, making it difficult or even impossible for plaintiffs to provide scientific evidence of the causal relationship between the unlawful act and the resulting damage.
Therefore, the so-called probability theory was devised as an alternative means to prove causation. The theory does not require that facts be proven based on rigid scientific methodology; it only requires evidence demonstrating a substantial degree of probability that the harm would not have taken place without the unlawful act of the defendant [5,6]. Proof of the existence of a substantial degree of probability of a causal relationship between the act by the defendant and the occurrence of losses is to be considered sufficient, and the burden of presenting counterevidence lies with the defendant. The two major approaches in this school of thought are factual presumption theory and the principle of superiority of evidence. The former claims the existence of a cause-and-effect relationship without direct or positive proof of the fact, but grounded in probabilistic evidence which entitles it to belief. The latter compares the probabilities of the claims from both sides to determine which claim has a relatively higher (51:49) chance of being the truth. In other words, the principle of superiority of evidence provides a rationale for possible judicial relief when there is evidence that can be reasonably treated as superior relative to other evidence [4]. Criticisms of the probability theory are as follows: causation is inferred as fact not based on legal grounds but rather based on reasonable belief only; the compensation for the plaintiff’s damages is often not realized because the probability theory also applies to the defendant’s counterevidence; and, since the probability theory does not provide a bright line rule, it can only give a caution to the court that tends to be conservative in accepting causation. [5,6].
Acknowledging such criticisms, some began to argue that the plaintiff’s burden of proof should be relieved in accordance with res ipsa loquitur or the so-called new probability theory. The Korean Supreme Court has since adopted a legal principle regarding the alleviation of the burden of proving causation based on res ipsa loquitur or the so-called new probability [5,6]. For instance, the Court made decisions based on these theories in a number of environmental law cases including the Supreme Court Decision, 2000Da65666 (principal claim), 65673 (counterclaim), October 22, 2002.
This case dealt with a claim by laver farm owners for compensation for damage allegedly caused by warm wastewater released from a nearby thermal power plant. The plaintiff (the defendant of the counterclaim, hereinafter referred to as “the plaintiff”) was operating a thermal power plant on the west coast of Korea, and the defendants (the plaintiffs of the counterclaim, hereinafter referred to as “the defendants”) were owners of laver farms located several kilometers away from the plant. Laver is a type of seaweed that lives in low-temperature waters, thus particularly vulnerable to the rise of water temperature. Warm wastewater had been leaking into the nearby ocean from the cooling system of the plaintiff’s facilities, with the amount gradually increasing every year. During the same period, the laver yield per unit production of the defendants’ laver farms dropped to an equivalent of just 21.4% of their competitors with similar natural conditions.
The appellate case (Daejeon High Court Decision, 96Na738 [principal claim], 96Na745 [counterclaim], October 25, 2000) confirmed the fact that the water released by the power plant had made its way to the nearby ocean where the plaintiffs’ laver farm is located and contributed to the rise in water temperature. The court found that the above fact was related to the significant decrease in the harvest of laver, which is highly vulnerable to high water temperature. The court stated that, since the causation between the release of wastewater by the plaintiff and the decreased laver production was proved with convincing evidence, the plaintiff could not avoid the liability unless counterevidence was provided.
The Supreme Court, affirming the appellate court decision, stated:
  • The burden of proof of causation between an unlawful act and damages incurred generally lies with the alleged victim in tort cases. However, source substances released by corporations often cause damages indirectly to victims, through air or water pollution, and even modern science fails to explain all the mechanisms that lead to pollution. Consequently, in most cases it is extremely difficult or even impossible to scientifically prove each and every link that constitutes the causal relationship between the act and the damage. Therefore, requiring victims to present rigid scientific evidence of the causal relationship in environmental lawsuits may in effect lead to a complete denial of judicial remedies. Meanwhile, corporations have better technological and financial resources to conduct inquiries, and they have a higher likelihood of attempting to cover-up their negligence. As a result, from the viewpoint of equity, it would be fair to conclude that, when certain toxic substances released by corporations reaches plaintiffs’ property and damage on the property occurs, a corporation shall not be relieved of its liabilities unless it presents evidence of the harmlessness of the substances released at its source and that such substances did not cause damage to the inflicted objects (emphasis added).

The above decision took account of the practical reality in environmental lawsuits that the causation between pollution and the inflicted damage is indirect and complex. The decision holds the company liable, unless the defendant proves harmlessness, if the plaintiff successfully proves the following: (1) the defendant released a certain toxic substance, (2) the substance reached the damaged objects, and (3) the damage did occur.

Product liability lawsuits

The Supreme Court established a precedent of relieving the burden of proving causation even in a pharmaceutical case which is a product liability lawsuit. For example, the Supreme Court Decision, 2008Da17776, September 29, 2011 involved a case where plaintiffs suffering from hemophilia sued a pharmaceutical company, claiming that they had been infected with the human immunodeficiency virus (HIV) through the transfusion of blood products manufactured and supplied by the defendant.
In order to prove a substantial probability that the defendant’s blood products had been contaminated by the virus, the plaintiffs submitted as evidence the results of epidemiological research. At the request of the Health and Welfare Committee of the National Assembly, the Committee for HIV Infection through Blood Products, consisting of 16 members including experts from the medical and pharmaceutical fields, consumer groups, and other relevant organizations, carried out research between September 1, 2002 and February 4, 2004. The results suggested that the blood product was the only domestic brand distributed in Korea during that period, and that the probability of HIV infection for hemophiliac patients injected with the product was statistically significant in comparison with those who did not receive the injection.
Considering the difficulties in proving causation between the defect of a product or the negligence of a pharmaceutical company and the damages incurred, the Supreme Court relieved the plaintiff’s burden of proof as follows:
  • In cases where a patient as victim files a lawsuit against a pharmaceutical company claiming liability for damages on the grounds that he/she was infected by the injection of a blood product contaminated with HIV virus, the burden of proof shall be relieved if the plaintiff successfully proves all of the following: ① The patient did not show any symptom of infection before receiving the blood product produced by the pharmaceutical company; ② The infection took place after the injection had been carried out; ③ There is a substantial probability that the blood product had been contaminated with the virus. Relieving the patient’s burden of proof in order to allow the patient to claim compensation for damages by presuming causation between the infection and the defective blood product manufactured by the pharmaceutical company or its negligence would contribute to meeting the goal of our compensation claim system, which has adopted fair and equitable compensation as its guiding principle. The substantial probability of virus infection, even though there is no clear scientific evidence, can be inferred from multiple factors such as the proximity of timing between the use of the blood product and the infection, statistical relevance, the manufacturing process of the product, the medical characteristics of the virus infection, and the accuracy of the diagnosis of the virus infection conducted on the source blood (emphasis added).

Based on this legal test, the Court found a substantial probability that the blood product manufactured by the defendant company had been contaminated with HIV, and decided that an assumption could be made that there was a causal relationship between the defective blood product or related negligence by the defendant and the HIV infection of the plaintiffs.
The plaintiffs filed lawsuits for compensatory damages against the Korean government and KT&G for manufacturing and selling cigarettes, arguing that either they or their decedents had contracted diseases such as small-cell carcinoma (a type of lung cancer), squamous cell carcinoma, non-small cell lung cancer, bronchioloalveolar carcinoma, etc. due to decades of smoking.
The Appellate Court Decision (Seoul High Court Decision, 2007Na18883, February 2, 2016)
The appellate court decided, in light of the similarity between tobacco litigation and environmental litigation, that: “there is a need to relax the burden of proving causation between smoking and contracting a disease such as lung cancer to some degree, although it may fall short of the probability test in environmental litigation.” Based on this rationale, the court concluded that it was remarkably difficult or nearly impossible to scientifically verify whether individual cases of lung cancer were caused by smoking due to the following reasons: 1) it is difficult to tell whether one inhalation of cigarette has any meaningful connection to cancer because the amount of carcinogens absorbed into the body from one inhalation is so small; 2) it is difficult to prove causation between smoking and contracting a disease based on scientific evidence alone, since there are many other risk factors that may cause lung cancer; 3) the chemical composition and characteristics of cigarettes and cigarette smoke have not yet been clearly identified; and 4) a biotest ascertaining the effects of such chemicals is not a realistic possibility.
The appellate court acknowledged that smoking had a considerable epidemiological connection to cases of small-cell carcinoma and squamous cell carcinoma, citing the results of the epidemiologic research that the plaintiffs submitted as evidence. Therefore, the court found that since “smoking is a key factor, or at least a significant factor, in causing lung cancer, a causal relationship between smoking and lung cancer can be presumed.” The court then shifted the burden of proof to the defendant, requiring them to prove that lung cancer among smokers developed solely, or essentially, from different factors other than smoking. Nevertheless, the plaintiff’s claim for compensation was dismissed because the appellate court found no defects in cigarettes manufactured by the defendant company.
Regarding non-small cell lung cancer and bronchioloaveolar carcinoma, the appellate court found that the causal relationship between smoking and them was not proven sufficiently. According to the appellate court, it was difficult to infer a causal relationship from the epidemiological evidence submitted by the plaintiffs due to the following reasons: 1) non-small cell lung cancer does not refer to a particular type of cancer, but refers to all types of cancer that are not small cell carcinoma, and it includes types of lung cancer that have no relation, or a considerably low relation, to smoking; 2) there are reports that bronchioloalveolar carcinoma could occur due to tuberculosis, pneumonia, viruses, etc.; 3) non-small cell lung cancer has a considerably lower relation to smoking compared to that of squamous cell carcinoma or small cell carcinoma; and 4) other factors, such as environmental pollutants, may be contributing factors to non-small cell lung cancer, since the rate of occurrence among non-smokers is high.
The Supreme Court’s Decision (Supreme Court Decision, 2011Da22092, April 10, 2014)
The plaintiffs argued before the Supreme Court that the appellate court’s decision, which denied the causal relationship between smoking and non-small cell lung cancer and bronchioloalveolar carcinoma, should be reconsidered. However, the Supreme Court affirmed the appellate court’s decision, stating that non-small cell lung cancer and bronchioloalveolar carcinoma were not specific diseases caused solely by smoking, but were non-specific diseases that result from complex interactions of external environmental factors (such as physical, biological, and chemical influences) and the body’s internal factors. And then, the Court decided it was difficult to see how smoking caused non-small cell lung cancer and bronchioloalveolar carcinoma based solely on the epidemiological evidence submitted by the plaintiffs, referring to the “Agent Orange Case” (Supreme Court Decision, 2006Da17539, July 12, 2013), in which Vietnam War veterans claimed compensation for their exposure to Agent Orange. The Court cited the case’s legal test that identified the causal relationship between risk factors and non-specific diseases:
  • “In order to prove that a risk factor has the probability to cause a non-specific disease, the following must be established: 1) the rate of contraction of the non-specific disease is significantly greater in a group exposed to the risk factor than in a group that is not exposed to said risk factor, and 2) the time and amount of exposure of an individual in the group [that was exposed to the risk factor], the time of the occurrence, the health conditions before exposure, life-style habits, the changing state of the disease, and family history (emphasis added).

The Court dismissed the plaintiffs’ claim for compensation for their diseases, because there was no defect in the product.
The Validity of Differentiating General Causation and Specific Causation
The Supreme Court made it clear that in individual smoker lawsuits, while the epidemiological evidence may prove the probability of contracting diseases within a group, it cannot prove the cause of contracting diseases for an individual. According to the Court, “Epidemiology focuses on investigating the causes of disease contraction as a group phenomenon; it does not examine the cause of a disease of an individual who belongs to that group” [7]. The Supreme Court’s distinction between general causation and specific causation makes sense because it is based on an exact understanding of epidemiology.
Problems in Making a Distinction Between Specific Diseases and Non-specific Diseases
The Supreme Court categorized illnesses into specific diseases and non-specific diseases following the standard suggested in the “Agent Orange Case.” In this case, the court applied different legal principles for proving the causal relationship between smoking and contracting a disease, depending on which category the disease belongs to. However, academics in the field of health and medicine are not familiar with the so-called “specific disease” concept [8,9]. The Supreme Court defined the concepts of specific and non-specific diseases as follows:
• Specific disease: A disease “that develops due to a specific cause, and in which the cause and effect clearly correspond to one another.”
• Non-specific disease: “A disease whose causes and mechanisms are complex, and which develops due to both innate factors (such as genetics and physical conditions) and acquired factors (such as drinking, smoking, age, eating habits, and professional and environmental influences).”
In other words, the Court cited the following two requisites for a specific disease and examined each one: 1) the disease should develop due to a single cause, and 2) the cause (a specific cause) and effect (the occurrence of the disease) should clearly correspond to one another.

The condition that a specific disease must develop by a single cause

Currently, experts in the field of medicine and public health do not believe that a disease develops from a single cause. According to the multifactor theory, any one factor is not the necessary and sufficient condition for disease development—a disease occurs due to the interaction of multiple factors [11]. Even if a necessary factor exists, a disease cannot develop without that factor interacting with other intermediary and environmental factors.
The appellate court’s decision gave cholera as an example of specific disease, but even in cases of such infectious diseases, not everyone who is exposed to the germ or virus becomes sick. Of course, there is a strong correlation between exposure to a virus and contracting a disease, but a disease develops due to a complex interaction of several factors, such as the individual’s immunity or everyday habits. Therefore, according to the Supreme Court’s division of specific diseases and non-specific diseases, all diseases are “diseases that occur due to a complex interaction of both innate factors (such as genetics) and acquired factors (such as drinking, smoking, age, eating habits, and professional and environmental influences).” In other words, all diseases belong to the non-specific diseases category.
In order to figure out why the Supreme Court adopted the concept of specific diseases, the appellate court’s decision in the “Agent Orange Case” (Seoul High Court Decision, 2002Na32662, January 26, 2006) needs to be reviewed, because it is the first case that adopted the notion of specific diseases. In the process of making a decision on the value of epidemiological evidence in proving legal causation, the appellate court cited the US National Academy of Sciences (NAS) report and recognized the fact that “because chloracne develops only among persons exposed to 2,3,7,8-tetrachlorodibenzodioxin (TCDD), it is used as a biomarker in epidemiologic studies to verify whether or not a person was exposed to TCDD.” The court judged that chloracne is “a specific illness that develops due to a single cause, TCDD (emphasis added).”
However, the NAS report never mentioned the concept of a specific disease (or, as the appellate court decision called it, “specific illness”) which the Supreme Court adopted [12]. The above report merely used the expression “specific” in the section that explained how TCDD exposure can trigger a “specific health outcome” [12]. The report went on to say, “although chloracne has been used in epidemiologic studies as a biomarker for TCDD exposure, the data indicate that it is neither a sensitive nor exclusive indicator. It is usually not long lasting, is difficult to diagnose, and is not at all sensitive to exposure to herbicides that are not contaminated with TCDD.” The report even clearly stated that there are other causes of chloracne besides TCDD: “Evidence is sufficient to conclude that there is a positive association between exposure to herbicides* (2,4-D; 2,4,5-T and its contaminant TCDD; cacodylic acid; and picloram) and chloracne.” Therefore, the appellate court seems to have misread the NAS report and misjudged that chloracne was a specific disease, and the Supreme Court made a decision based on the assumption that the appellate court’s fact-finding was true.

The condition that the cause (specific cause) and effect (the occurrence of the disease) clearly correspond to one another

The second condition examined by the Supreme Court, which states that the cause (specific cause) and effect (the occurrence of the disease) clearly correspond to one another, is similar to the “specificity” concept. Specificity is an indicator of how accurately one cause of an associated pair can predict the occurrence of the other. According to the US Surgeon General’s 1964 report [13], specificity “implies the precision with which one component of an associated pair can be utilized to predict the occurrence of the other.” On a similar note, Oxford University professor Bradford Hill explained that the concept of specificity has two variants: 1) a cause leads to a single effect, not multiple effects and 2) an effect has one cause, not multiple causes [9]. Hill said both variants must be considered when determining specificity [9].
Epidemiologists use different criteria to make causal inferences from an epidemiological association, and call for care when using specificity as a criterion [2,14]. For instance, the Surgeon General’s report pointed out the growing recognition among the scientific community that a disease may have multiple causes. The report explained that the causal inference must not be contradicted even if the predictive value of the causes might be small. In other words, the report said that while smoking and lung cancer have a low degree of specificity (since lung cancer is caused by many different factors other than smoking and smoking can cause other diseases besides lung cancer), an epidemiological causation can be inferred [13] because relative risk ratios average about nine to 10 for cigarette smokers compared to non-smokers. Dr. Hill also presented what is now widely known as the “Hill Criteria” for causal inference based on epidemiological association. He asserted that the specificity criterion must not be overly emphasized, and even if there is no specificity between two components, one should not be too quick to conclude that there is no causality to be inferred [15].
As such, scientists have warned against misusing the specificity concept [9], and as previously mentioned, the concept of specific diseases as a disease category is artificial and false. Therefore, it is questionable why the Supreme Court had to adopt the concept of specificity in the process of judging the validity of epidemiological evidence, a process around which issues will come up more frequently in the future. It is more reasonable to judge the causal relationship between risk factors and diseases by considering the degree of epidemiological association based on the premise that all diseases are non-specific. Even if the concept of specificity is applied, the cancers mentioned in this particular case can be said to have a very high degree of specificity.
The Relationship Between Epidemiological Association and Legal Causation
The Supreme Court ruled that 1) even though there is a high degree of epidemiological association in this particular case, 2) the decision on causal relationship in a lawsuit must be made by focusing on an individual, not on a group which that individual belongs to:
  • “After conducting a contrasting epidemiological investigation between a group that was exposed to a risk factor and a group that was not, one must prove the probability that the non-specific disease occurred due to the risk factor: 1) one must prove that the rate at which the non-specific disease is contracted in the group exposed to the risk factor is considerably greater than the rate for the group not exposed to that factor, and 2) an individual in the group must additionally prove the length of time and the amount of exposure, the time of the occurrence, the health conditions before exposure, living habits, the changing state of the disease, family history, etc. (emphasis and numbers added).

In other words, the Supreme Court required the plaintiff to prove both the epidemiological association and individual factors in order to qualify for liability compensation. However, according to Supreme Court decisions, there is a possibility that a plaintiff’s rights could be unduly restricted. In cases where the causal relationship (and the epidemiological association inferred from it) between a risk factor and the occurrence of disease in a certain group is very strong, the plaintiff should not be required to prove that his or her own situation had no influence on the occurrence; the defendant should instead be required to prove that the occurrence of the disease was due to a cause other than the risk factor. From the perspective that all diseases are caused by multiple factors (a common view among academics in the natural sciences), the defendant in toxic tort lawsuits is always in a position to argue that the cause of the plaintiff’s disease is due to a different factor, not the risk factor. If the plaintiff has to prove that all “different factor” is irrelevant to his or her disease, this can only become a “proving of non-existence” most of the time. In other words, the plaintiff has the burden of proving that his or her (unsuitable) living habits or family history, etc. were not the reasons for the disease. Besides, it is almost impossible for the plaintiff to prove the non-existence of these factors since they cannot be easily measured and because a disease develops over the course of time and through a complex process. When there are multiple factors that have causal relationships with the occurrence of a disease, each factor should be seen as having a causal relationship with the occurrence [16]. The fact that multiple factors contributed to occurrence of the end result could be considered in establishing the scope of liability by applying comparative negligence principles (Korean Civil Act Article 763, 396).
Even though epidemiologists carry out research targeting populations and draw conclusions based on group indicators such as relative risk and attributable risk, we cannot say that epidemiology provides information solely about a population and no meaningful information regarding individuals in the group. For instance, the probability of causation calculated by epidemiological study shows the probability of a randomly selected individual contracting a disease due to a risk factor. This indicator is useful in converting group probability indexes into individual ones. The burden of proof does not mean the proof has to be logical and scientific, with no room for any doubt or possibility of objection. It means there must be a high degree of probability that is generally thought to be true and socially accepted [4]. Therefore, if the epidemiological association is fully proven, we can infer that there is a probability of a causal relationship [17]. Then, the burden to reverse the decision lies with the defendant, by proving that an individual’s situation affected the occurrence of the disease.
The value of epidemiological evidence has become a major issue not only in tobacco litigation but in the growing number of toxic substance litigations. The value of epidemiological evidence was a contentious issue for the “Agent Orange Case” many years ago, in which Vietnam War veterans sought compensation from companies that manufactured Agent Orange (Supreme Court Decision, 2006Da17539, July 12, 2013; Supreme Court Decision, 2006Da17553, July 23, 2013). The same question came up in another case where the plaintiff argued he had contracted asthma from exposure to gas emissions and sued automobile companies for compensation (Supreme Court Decision, 2011Da7437, September 4, 2014).
In a government agency tobacco lawsuit that is currently pending, the reliability of epidemiological evidence is also a critical issue. The plaintiff in this case is demanding compensation for diseases that the appellate court of the individual smoker lawsuit decided as having legal causation with smoking; even though that court dismissed the plaintiff’s appeal because there were no defects in tobacco products, the court stated that there were causal relationships between squamous cell carcinoma and small cell carcinoma and smoking. In the government agency tobacco lawsuit, the plaintiff is excluding the types of cancers that had a relatively weak association with smoking and is only presenting types of cancers that show an 80 to 90 percent association with smoking.
By virtue of its research methodologies and concepts, an epidemiological study provides reliable information on not just probability within a group but probability for individuals in the group as well. Therefore, it is desirable that courts understand and appreciate the nature and the value of epidemiological evidence which might be the only convincing and powerful evidence of causation in many types of toxic tort lawsuits.
This study was conducted with financial support from Gachon University (no. GCU 2014-5107).


The author has no conflicts of interest associated with the material presented in this paper.

Supplemental material (Korean version) is available at
  • 1. Jee WL. Lecture on civil law. 12th ed. Seoul: Hongmoonsa; 2014. (Korean)
  • 2. Gordis L. Epidemiology. 4th ed. Philadelphia: Elsevier/Saunders; 2009
  • 3. Yang CS, Kwon YJ. Changes and protection of rights: civil law 2. Seoul: Pakyoungsa; 2012. (Korean)
  • 4. Min IY, Kim NH. Annotated civil procedure law 1. Seoul: Korean Association for Public Administration and Justice; 2012. (Korean)
  • 5. Jung DY, Yoo BH. Civil procedure law. Seoul: Pakyoungsa; 2014. (Korean)
  • 6. Lee SY. New civil procedure law. 9th ed. Seoul: Pakyoungsa; 2015. (Korean)
  • 7. Kim JH. 2014 An analysis of key judicial decisions by field (5) civil law 2. Law Times; 2015 Mar 13 [cited 2016 Jan 10]; Available from: (Korean)
  • 8. Porta MS, Greenland S, Hernán M, Santos Silva ID, Last LM; International Epidemiological Association. A dictionary of epidemiology. 6th ed. Oxford: Oxford University Press; 2014
  • 9. Rothman JK, Greenland S, Lash TL. Modern epidemiology. 3rd ed. Philadelphia: Lippincott Williams & Wilkin; 2008
  • 10. Lee SG. Using epidemiological evidence to prove causation in toxic torts: recent cases of the Supreme Court of Korea. Justice 2015;146(1):256-285. (Korean)
  • 11. Preventive Medicine Compilation Committee. Preventive medicine. 3rd ed. Seoul: Gyechuk Munhwasa; 2004. (Korean)
  • 12. Committee to Review the Health Effects in Vietnam Veterans of Exposure to Herbicides. Veterans and Agent Orange: health effects of herbicides used in Vietnam. Washington, DC: National Academy Press; 1994
  • 13. US Surgeon General’s Advisory Committee on Smoking and Health. Smoking and health; report of the advisory committee to the Surgeon General of the Public Health Service. Washington, DC: US Department of Health, Education, and Welfare, Public Health Service; 1964
  • 14. Rothman KJ, Greenland S. Causation and causal inference in epidemiology. Am J Public Health 2005;95 Suppl 1: S144-S150Article
  • 15. Hill AB. The environment and disease: association or causation? Proc R Soc Med 1965;58(5):295-300ArticlePubMedPMC
  • 16. Supreme Court Decision, 99Da43448∙99Da43455, Jan 14, 2000 (Korean)
  • 17. Yune JS. The main issues of the product liability. Yonsei Law Rev 2011;21(3):1-55. (Korean)

Figure & Data



    Citations to this article as recorded by  
    • Attribution of Cancer Origins to Endogenous, Exogenous, and Preventable Mutational Processes
      Vincent L. Cannataro, Jeffrey D. Mandell, Jeffrey P. Townsend, Melissa Wilson
      Molecular Biology and Evolution.2022;[Epub]     CrossRef

    JPMPH : Journal of Preventive Medicine and Public Health