Quite often, the media reports on a study that purports to link a drug with some adverse medical condition. Frequently, the media later reports on a different study that came to a seemingly contradictory conclusion. This highlights the unique issues with regard to proving causation in pharmaceutical cases: Studies are not always scientifically valid or, consequently, legally reliable.
Plaintiffs' attorneys often fall into a trap when they see case reports of a medical condition associated with a drug. All too often, they race to court and file a complaint alleging that their client suffered from that very same medical condition. However, they do not realize that they have a long road ahead and will face significant hurdles in proving that the drug is causally related to the medical condition, as opposed to merely being associated with it.
Spontaneous reports of medical events or anecdotal reports are misleading from a causal standpoint. Spontaneous reports are reports of the occurrence of medical conditions in individual patients taking particular drugs and are submitted to the Food and Drug Administration by the drug manufacturer or on a voluntary basis by doctors, pharmacists, patients, lawyers and others. Similarly, an anecdotal report is a letter to the editor or an article in a medical journal written by a physician reporting on an observation seen in a patient who has taken a medication and experienced a medical event.
But these types of reports, hereinafter referred to as "case reports," while medically significant for further evaluation and study, do not prove causation in the legal sense.
As all personal-injury litigators know, causation is an essential element of each theory of liability in any personal-injury action. See, e.g., Sindell v. Abbott Labs., 26 Cal. 3d 588 (1980). Furthermore, the law in California "is well settled that causation must be proven within a reasonable medical probability based on competent expert testimony. Mere possibility alone is insufficient to establish a prima facie case." Jones v. Ortho Pharm. Corp., 163 Cal. App. 3d 396 (1985).
In addition, plaintiffs in California must prove causation even in cases where it would be particularly difficult to do so. They must present reasonably reliable evidence proving causation in order to proceed with their case. Furthermore, California law establishes that expert opinions cannot be based on unreliable evidence (Evidence Code section 801), speculation or assumptions (Hyatt v. Sierra Boat Co., 79 Cal. App. 3d 325 (1978)) or insufficient data (Korsak v. Atlas Hotels Inc., 2 Cal. App. 4th 1516 (1992)).
Case reports are unreliable in part because they constitute classic hearsay, often containing triple or quadruple levels of out-of-court statements. Accordingly, experts may base their opinions on case reports only if such reports may be reasonably relied upon by other experts. See Evid. Code § 801(b).
However, with case reports, experts cannot vouch for the accuracy of the observations contained in the reports or the motivations of the reporters. Because of the inherent unreliability of such data, courts facing this issue in prescription-drug cases consistently have held that case reports are not the type of data reasonably relied upon by experts for causation opinions. See, e.g., Muzzey v. Kerr-McGee Chem. Corp., 921 F. Supp. 511 (N.D. Ill. 1996).
Besides courts, the scientific community has also expressed its opinion that case reports do not prove a causal connection. The FDA and the World Health Organization have publicly stated that case reports are not capable of generating incidence rates for an adverse medical condition among users of the particular drug or of determining the relative risk of the particular drug to cause the adverse medical condition. Therefore, they are not the type of data that establish a cause-and-effect relationship between a drug and a medical condition.
In fact, the FDA specifically warns that, for any case report, there is no certainty that the drug caused the medical condition. Indeed, the medical condition may have been related to the underlying disease for which the drug was given or to other drugs being taken, or the medical condition may have occurred by chance.
In addition, case reports cannot be used to establish causality for the following reasons:
- They do not take into account the incidence rates of the medical condition in the relevant population that is not taking the suspect drug.
- They often include the possibility that confounding factors, such as underlying illnesses or exposures to other medications, may have contributed to the patient's adverse experience.
- They fail to account for the number of patients who are prescribed and using a given drug.
- Their data are often incomplete and, in some cases, include unsubstantiated diagnoses.
For instance, if 1,000 patients on the same drug developed the same medical condition, one would assume that this would indicate that the drug is related to the condition. However, this figure must be put in perspective. If there were a total of 1 million patients on the drug, the incidence rate for that condition among patients on that drug would be very small: 0.1 percent.
Compounded further, if the incidence rate of that medical condition in the relevant population that has not been exposed to the particular drug is 0.1 percent or more, there is no higher incidence of the condition in those patients on the drug than in those not on the drug. Accordingly, the 1,000 case reports would be meaningless from a causal perspective.
To prove causation, a plaintiff clearly needs more than just case reports. Although California courts have not directly addressed this issue, most courts have routinely held that epidemiological evidence is necessary.
Epidemiology is a scientific discipline that, by its nature, employs the concept of cause through statistical probability, in which hypotheses are framed, tested and refined in a controlled setting and then used in identifying the connection between one event (such as exposure to a drug) and a subsequent event (such as the development of a disease).
Case reports are not controlled data and do nothing more than raise the possibility that the drug was related to the event. See Casey v. Ohio Med. Prods., 877 F. Supp. 1380 (N.D. Cal. 1995). In other words, case reports are simply observations of the coincidental occurrence of two events, i.e., an exposure and a condition.
These principles are illustrated in Jones v. United States, 933 F. Supp. 894 (N.D. Cal. 1996), aff'd, 127 F.3d 1154 (9th Cir. 1997), in which a plaintiff and her husband brought a wrongful-birth action against the United States and claimed that the government was responsible for her pregnancy because military doctors failed to warn her that penicillin could interfere with the effectiveness of her birth-control pills.
Plaintiffs' experts in Jones relied most heavily on 11 published journal articles, all of which were based on "anecdotal case reports, reviews of research done by other people, or studies lacking a control group." The court concluded that the journal articles could not support a causation opinion because they were not supported by "good science" and were "not derived through the scientific method."
Likewise, the court in Casey was faced with a situation in which an expert sought to base his opinion on an article that compiled 572 case reports of a certain alleged drug-induced medical reaction at issue in that case. The court emphasized the importance of epidemiology and held that reliance upon this journal article was improper.
The Casey court further observed that such case reports were not reliable scientific evidence of causation because they simply described reported phenomena without comparison to the rate at which the phenomena occur in the general population or in a defined control group; they did not isolate and exclude potentially alternative causes; and they did not investigate or explain the mechanism of causation.
A causal connection between a drug and an alleged injury can be hard to establish in a pharmaceutical product-liability case. Case reports show only a possible association between a drug and an adverse medical condition, not a causal connection. Evidence of an association may be sufficient for formulation of a hypothesis that can later be tested and confirmed, but it is not proof of causation in the courtroom or the scientific community. Accordingly, it is incumbent upon attorneys to be wary of the inherent limitations of case reports in proving causation before taking drug cases.