Believing six improbable things: medical malpractice and "legal fear".

• Author: Hyman, David A.

"Alice laughed." "There's no use trying," she said; "one can't believe impossible things."

"I daresay you haven't had much practice," said the Queen. "When I was your age, I always did it for half an hour a day. Why, sometimes I've believed as many as six impossible things before breakfast." (1)

1. INTRODUCTION

   Philip Howard believes. Howard believes that "[d]octors, teachers, ministers, even [L]ittle [L]eague coaches, find their daily decisions hampered by legal fear." (2) "Legal fear" exists when "[f]ear of litigation ... undermine[s] our freedom to make sensible decisions." (3) Howard has advanced this claim in a best-selling book, in numerous op-eds and speeches, and through his disarmingly-named public-interest organization, Our Common Good. (4) Howard has many allies. The list of co-sponsors and supporters of Our Common Good is a "Who's Who" of business and politics, public intellectuals, organized medicine, and health policy scholars. (5)

   Belief is one thing; proof is quite another. Testing Howard's beliefs requires framing them as empirically falsifiable propositions and then assessing them in light of the data. Howard's recent Washington Post op-ed on medical malpractice separates his global claim ("[f]ear of litigation [is] undermin[ing] our freedom to make sensible decisions" (6)) into more manageable and concrete sub-claims--and data is available to test each sub-claim. (7) Howard's op-ed indicates he believes at least six things about medical malpractice:

   1. The tort system causes physicians and other health care providers to hide their mistakes.

   2. Physicians and other health care providers once dealt with mistakes more openly.

   3. Liability encourages substantial defensive medicine.

   4. Liability undermines access to needed medical services.

   5. Liability creates an "extortion lottery."

   6. A specialized medical court would be an improvement.

   These individual sub-claims are at best unproven and at worst flatly wrong. This is not to suggest that the medical malpractice system is operating optimally, or that alternative institutional arrangements might not outperform it. As we have detailed elsewhere, the medical malpractice system (like the health care delivery system) has a whole series of pathologies. (8) Nonetheless, it does not follow that reform (whether that advocated by Howard or by anyone else) will necessarily make things better. Institutional imperfection is an unfortunate reality of public policy. (9) It is foolish to choose Policy A over Policy B by listing the deficiencies of Policy B and the virtues of Policy A; one must also consider the virtues of Policy B and the deficiencies of Policy A if the decision is to be informed by anything other than faith in one's own belief system.

2. BELIEF AND REALITY

   Medical malpractice is the best-studied area of the tort system. (10) Over thirty years of empirical studies and extensive historical inquiry have made it possible to assess the performance of the system based on empirical data, not anecdotes. This Part goes through each of Howard's sub-claims regarding medical malpractice and considers how well the sub-claims correspond to what is actually known about the performance of the medical malpractice system.

   Assertion 1: The tort system causes physicians and other health care providers to hide their mistakes.

   Howard claims that "[t]ragic human errors occur, for example, in prescription dosage, because people fearful of legal consequences are reluctant to speak up." (11) In other words, fear of liability is encouraging a culture of silence: the higher the risk of liability, the lower the probability of voluntary error reporting. Howard is not alone in making this claim; the Institute of Medicine asserted in a famous 1999 report that "[p]atient safety is [] hindered through the liability system and the threat of malpractice, which discourages the disclosure of errors." (12)

   There is one minor difficulty with this claim: no systematic empirical evidence indicates that it is true, and a fair amount of evidence shows that this claim is either dramatically overstated or simply wrong. As one prominent patient safety advocate (and firm believer in error reporting) noted in a recent article in the New England Journal of Medicine, "[n]o link between [error] reporting and litigation has ever been demonstrated." (13) No empirical study finds that the frequency of error reporting falls as the likelihood of being sued increases. Instead, the empirical literature indicates that there is massive underreporting of errors throughout the health care system, regardless of the level of liability risk that providers face. (14)

   Consider the incentive to report errors for physicians and nurses who work in the Veterans Administration hospital system. By statute, such health care providers are immune from liability for negligence. (15) If Howard's first sub-claim were correct, one would expect VA hospitals to be hotbeds of voluntary error reporting. Freed from the fear of individual liability, physicians and nurses in the VA should "speak up" about any and all deficiencies in care--and care in the VA should have long been dramatically better than in the rest of the health care system. Instead, studies of VA hospitals show that there is significant underreporting of errors. (16) VA hospitals have also had serious quality problems--and improvements in quality have resulted from external (Congressional) pressure and scandals, not voluntary action from within. (17)

   Next, consider no-harm errors and near misses by providers that are subject to liability. It is difficult to see those kinds of errors as sources of malpractice exposure; actual damages are a necessary component of a malpractice suit. No-harm errors and near misses are important sources of learning, however; one can learn a lot about the reliability of a delivery system by studying its harmless failures. Yet no-harm errors and near misses do not appear to be reported at a higher rate than errors for which liability is a possibility. (18)

   Similarly, even though different specialties face different levels of liability risk, there is no evidence of differences in error reporting rates across specialties. Nor have we found any evidence that error reporting rates vary within or across states, even though malpractice insurance premiums and liability risks clearly vary substantially within and across states. (19) Shifting to an international perspective, error tracking and reporting systems are much more highly developed in the United States than in the United Kingdom, where physicians face much lower malpractice exposure risks--an observation that suggests that liability actually encourages openness about errors. (20)

   Error reporting fails to happen for a variety of reasons, but tort liability is unlikely to be a major contributor to that problem. Few people like to report their own errors, whether in a medical context or otherwise. Few people like to "rat" on their friends. There are ways to address these problems, but scrapping the tort system is not one of them.

   To summarize, the available evidence is inconsistent with Howard's first sub-claim. Although "legal fear" may discourage some voluntary error reporting at the margins, it is not responsible for the rampant underreporting of errors that exists...