Monday, August 29, 2005

Vioxx and Torts

A recent jury trial in Texas found Merek liable for a man's death. The award was large and knocked Merek's share price significantly lower.

Merek could be correct that the jury did not follow the law and that the Plaintiff had a weak legal case. At the same time, the outcome could still be correct for society in the long run.

Mark Donald reports it was Merek's position that "'Plaintiffs must establish to a reasonable degree of medical certainty that Vioxx probably -- not possibly -- caused Mr. Ernst's fatal cardiac arrhythmia. . . They cannot do so. This failure of proof is fatal to each and every claim.'"

Causation
To recover damages in a civil tort case, the Plaintiff must show causation: that the defendant's action caused the harm the Plaintiff suffered.

A hypothetical: a doctor performs surgery on the patient and negligently leaves a sponge inside the patient's body. The patent then dies. If the doctor can show that leaving the sponge inside the patient did not cause the death, but rather an unanticipated allergic reaction to a medicine, the doctor is not liable for the negligent act.

To prove causation a plaintiff must either show that the harmful outcome would not have occurred more likely than not, had the plaintiff not been negligent, or that the plaintiff's negligence was a "substantial factor" in causing the harmful outcome. The latter test is rather vague and allows for liability where there is more than one cause contributing to an event. Different jurisdiction have different formulations as to what proof of causation is sufficient.

In cases like the Vioxx trail, causation can be difficult to prove. For example, if taking Vioxx increased the chance of heart attack by 80% and a patient takes Vioxx and dies, more likely than not the patient would have died even if they had not taken Vioxx. Thus, Merek should not be liable for the death.

However, applying these causation rules can lead to bad results overall. Lets say there are 100,000 Vioxx taking patients, 50 die from hearth attack, and only 30 would have died if all the patients had not taken the medicine. Clearly, Vioxx caused 20 deaths, but we don't know which 20 out of the 50 dead. In each individual case, more likely than not, the dead patient probably would have had a heart attack and died even if they did not take Vioxx. Thus, none of the 50 dead patients would recover in court.

This same result occurs in many toxic tort cases, where a group of people are exposed to a dangerous chemical due to a company's negligence. Science can show that each person faces, say, a 50% greater risk of getting cancer, but each person who gets cancer cannot prove the exposure cause their cancer.

The Jury Result
The Texas jury might have ignored the rules of causation and focused on Merek's bad behavior, as indicated by the huge punitive damage award. Merek knew it had a potentially dangerous product and chose not to tell the public. The jury intuitively understood what I demonstrated above: traditional causation rules are not sufficient to deter negligent or malicious behavior in situations where the negligence causes a significant increase in the risk of death, but where causation in any one case cannot be proved more likely than not.

Class Actions
Class action suits might seem like a possible solution to the causation problem, but they are not. In a class action, each plaintiff's case must be proved separately. So, if the 50 patients discussed above brought a class action, they would lose because each individual could not demonstrate causation.

My Suggestion
I would propose the ability for plaintiffs to join together and sue as a group. Applied to the hypo above, Merek would be liable only for 20 deaths, not all 50, but payments would be divided between all 50 because it is impossible to tell which patients were killed by Vioxx and which would have died anyway.

This solution punishes a defendant for the harm they caused, and distributes the funds to those who were likely injured. It also protects a defendant from paying for all 50 deaths, when they only caused 20. It would make trial outcomes more predictable and, therefore, probably more likely to settle.

The result would give money to those who were not harmed by the plaintiff's actions. This does not trouble me. If there is a choice between letting the wrongful defendant profit from causing deaths, and giving a who has suffered and done nothing wrong some money. . .

More
econbrowser
liking to
a Reason article

In an October 5, 2004 Wall Street Journal, article entitled "Despite Vioxx Withdrawal, The Benefits of Medicines Can Outweigh the Risks," Andrea Petersen writes:

. . . In a study of 2,600 patients, there were 15 heart attacks or stroke per 1,000 patients among those who took Vioxx for longer than 18 months. Among those who took a placebo, there were 7.5 heart attacks or strokes per 1,000 patients. And those who took Vioxx for less than 18 months had no increased cardiovascular risk.



A doubling of risk is at the edge of proving causation. So, the particular facts of the patient become important. (And I wonder if anything further is known after Oct. 5, 2004.)

No comments: