Subscribe Subscribe   Find us on Twitter Follow POL on Twitter  



How to lie with statistics, medical malpractice edition

Arizona State Professor Michael Saks has an op-ed in the Milwaukee Journal-Sentinel that is notable for its use of misleading statistics.

Saks starts off by arguing that doctors aren't affected by high malpractice insurance premiums, because they can pass the costs on to patients--and then, in the very next sentence, argues that unlimited non-economic damages are necessary to deter doctors from committing malpractice. The inherent incompatibility of these two arguments appears to escape him.

He then repeats the "About 5% of doctors are the focus of half the lawsuits" sham statistic; we've refuted this innumeracy elsewhere, as well as noted the economic implications that would result if this were really a useful statistic.

But the most jarring statistic is one we haven't seen used before, and it's worth noting why it's bogus: "How well does the existing system screen out "frivolous" claims? One telling finding is that a doctor is 29 times more likely to be sued by a patient who has been injured by malpractice than by one who has been injured non-negligently."

Though there's no source for it, it sure sounds impressive; if one reads it too quickly, it sounds like 97% of medical malpractice suits are justified. But it deserves a closer look. Even if the 29x figure is true, it is hardly an endorsement of the current system.

What is an injured patient? Many malpractice cases result from patients being unhappy with adverse results; the cerebral palsy cases where doctors are blamed for birth injuries are the most notable examples. Imagine a world where doctors commit malpractice on 1 in 200 injured patients. Already, this seems awfully high--if every American man, woman, and child saw a doctor and had an adverse result only once every three years, that would be a half-million cases of malpractice a year. Let's now say that the odds of a doctor who hasn't committed malpractice being sued is pi. By the Saks statistic, the odds of a doctor who has committed malpractice being sued is:

pm = 29 * pi

With these assumptions, what are the odds that a malpractice lawsuit is meritorious? I'll show the work. Recall, for every malpractice case, there are 199 patients who have been seen without malpractice.

pm ÷ (199 * pi + pm)
29 x pi ÷ (199 * pi + 29 * pi)
29 ÷ (199 + 29) = 29 ÷ 228 = 12.7%

In a world that generously assumes over a half-million cases of malpractice a year, and generously accepts as true Saks's assertion that malpractice victims are 29 times more likely to sue, seven out of eight malpractice lawsuits will be meritless. Even if one disputes the 1-in-200 number, doctors would have to be injuring one in thirty patients before plaintiffs' lawyers got to the 50% level. That impressive Saks statistic turns out to be one that condemns the current state of affairs. "How well does the existing system screen out [meritless] claims?" Not very well at all.

Saks also claims that "research finds that only 10 to 20 cents are paid on each dollar of economic loss," but gives no cite; a Google for that phrase turns up only the statistic that a small fraction of the money spent on medical malpractice litigation goes to compensate for economic losses, but that's a different denominator. Anyone know what he's referring to? Email me at tedfrank at overlawyered dot com.



Rafael Mangual
Project Manager,
Legal Policy

Manhattan Institute


Published by the Manhattan Institute

The Manhattan Insitute's Center for Legal Policy.