Merck’s Vioxx liability running estimate

Note: If you have not read the original post, please read that first below.

Update June 1, 2007

The latest count says that 27,250 lawsuits alleging 45,700 incidences of personal injury or loss due to Vioxx have been filed and that includes 266 filed as possible class actions. It appears that approximately 50% of the initially estimated (by Judge Fallon) victims have not acted for reasons we do not know.

Update March 13, 2007

After losing his first trial against Merck, Frederick “Mike” Humeston has now received damages of $47.5 million. This puts the average settlement amount at ~$29 million, but since Merck has won nine cases and lost five, it means that the probability of payment is 35%. Therefore, the estimated liability is 40,000X0.35X29 million or $406 billion. Obviously, this is way too high but at this point, I will keep this number and revise it as we get additional data.

Update October 6, 2006

Contrary to prediction by Judge Eldon Fallon, as of today, according to Bloomberg News, 40,000 claims have been filed against Merck. The situation is not completely clear if this is the end – there can be more in the future.

Update August 30, 2006

Judge Eldon Fallon has thrown out the $51 million damages but Merck is still liable. A retrial has been ordered. The effect is that we are back to the estimate of April 29, 2006, and that is, $90 billion.

Update August 17, 2006

In the Gerald Barnett case, the jury is directing Merck to pay $51 million. According to the latest estimates, the company now faces as many as 16,000 lawsuits. With the revised numbers, the average amount of compensation per lawsuit is now ~$25 million. So sticking to my formula (which seems to holding well for now; 50% probability that Merck win a lawsuit), total number adds up to $200 billion. It is clearly the highest number I am seeing so far.

Update April 29, 2006

In the Leonel Garza case, Merck will have to pay $7.75 million even though the actual award by the jury was much higher. Another update on the number of lawsuits: 11,500 lawsuits filed to date. So the average amount to be actually given to the plaintiffs who have won is now $15.8 million. The numbers have gotten slightly better now for Merck. Using the same logic as I used before (Merck wins half the cases and has to pay an average of $15.8 million to the rest), the total liability estimate right now is $90 billion.

Update April 11,2006

We do have some more data now though still not enough to make a statistically sound estimate. The number of lawsuits is currently at 10,000, though news reports indicate that more are likely to be filed. It seems that when I picked a 0.5 probability, it was pretty good – so far, Merck has lost and won two cases each. But the damages have been higher than the conservative estimate that I chose. Carol Ernst is likely to receive $26 million while the McDarby family is expected to receive $13.5 million. That puts the average at ~$20 million.

In other words, let us assume that the number of cases will stay at 10,000 and Merck will win half of them and pay an average of $20 million in damages in the ones that it loses. That makes the total Vioxx liabilities to be ~$100 billion. Ouch! No one has a clue how these numbers will actually turn out, but from the data that we have so far, it looks pretty scary.

Original post

After Merck lost the first trial in the Vioxx recall case, a topic of discussion has been an estimate of Merck’s legal liability. Since the day of the recall, a lot of estimates have been developed, ranging all the way from a few billion dollars to as much as $55 billion. Since then two estimates – $18 billion and $38 billion – have been cited most often, according to research conducted by iProceed.

Now that we know the actual amount of the damages, it is a good time to review previous numbers and to see if we can come up with better estimates. Below is a summary of research conducted by iProceed.

How many people have been hurt by Vioxx?

Since we only have good numbers for the United States, let us stick to them. Merck estimates that 105 million U.S. prescriptions were written for Vioxx between May 1999 and August 2004. The company says that approximately 20 million people in the U.S. used the drug (Source: Merck’s call with Wall Street analysts on December 14, 2004).

A Kaiser Permanente study showed that 25 out of every 10,000 people who took the drug had serious cardiovascular problems. That means some 50,000 people may have been injured in some way, using their numbers.

Dr. David Graham, an expert at the FDA, conducted a study of 1.4 million people in California and found that 69 out of every 10,000 people had either serious coronary disease or sudden cardiac death. That puts the number of those injured at 140,000 – an estimate that no one else has disputed and is supported by another expert, Dr. Eric Topol, the cardiologist at Cleveland Clinic. Dr. Graham presented his analysis to Congress and also published it in medical journal The Lancet. Another study notes that the case-fatality rate is 44% and that puts the toal number of deaths from Vioxx in the US at roughly 61,600, though in interviews Dr. Graham has used numbers ranging from 50,000 to 60,000.

What is the average settlement amount going to be?

Some of the earlier estimates looked at numbers ranging all the way from $100,000 to $650,000. However, Ms. Carol Ernst is likely to receive $26 million. And she is not the patient, but only the surviving widow, and her case was considered to be relatively weak. So let us make some very conservative assumptions here:

    Approximately 100,000 affected patients (or their family members) will sue Merck, according to U.S. District Court Judge Eldon Fallon.
    The average success rate is going to be 50% (though if Ms. Ernst can win, anybody can).
    The likely reward will be just 5% of what Ms. Ernst received, that is, $2.6 million (again highly unlikely that the amount will be this small but accounting for individual situations and local laws, let us assume this). Though it needs to be understood that a large number of cases will be tried in jurisdictions that do not put a limit on punitive damages, or do not cap them as severely as Texas. Note: I was almost tempted to use $26 million but the number was just too scary. You can still try it for the fun of it.

So what are Merck’s Vioxx liabilities in the United States alone?

100,000 X 0.50 X $1.3 million = $65 billion

Final remarks

I believe that a forecast is only as good as the assumptions behind it. While we do have good numbers on the number of injuries, likely number of lawsuits, the probability of success of a lawsuit (it has gone up after the Ernst vs. Merck verdict), we can only guess what the juries will do. While one verdict has given us a flavor of what we can expect, we will have better numbers sometime in 2006 when more cases may be decided.

Recommended article: Merck on the wrong track with its Vioxx strategy