EVENTS

Date:	Wednesday, April 5, 2006
Time:	10:00 AM -- 12:00 PM
Location:	Wohlstetter Conference Center, Twelfth Floor, AEI 1150 Seventeenth Street, N.W., Washington, D.C. 20036

April 2006

Medical Malpractice Liability and Physician Supply

 

Do doctors really flee certain states, avoid high-risk specialties such as neurosurgery and obstetrics, or even forgo practicing medicine as a result of ever-growing liability concerns? Can liability reforms prevent an exodus? Empirical answers to these questions are vital to informed policymaking but are sometimes lacking from the politically charged debate between doctors and lawyers. At an April 5 AEI event, two papers from leading health economists on the subject were presented. Professor David Dranove of Northwestern University’s Kellogg School of Management presented his new paper, “Has the Malpractice Crisis in Florida Really Affected Access to Care?” co-authored with Anne Gron and Andrew Sfekas. AEI adjunct scholar and Florida State University College of Law professor Jonathan Klick’s paper, “Medmal Reform and Physicians in High-Risk Specialties,” was also presented.
 
David Dranove
Kellogg School of Management, Northwestern University

Although high malpractice premiums have been a concern of policy analysts for some time, systematic research on the subject is quite lacking. This paper empirically examines the extent to which patients’ access to treatment is affected by physician exit caused by medical malpractice litigation. We do this by examining two fields that are frequently given special attention, neurosurgery and obstetrics, and examining physician activity levels over time in Florida, which is an American Medical Association (AMA) crisis state, and which provides unique hospital data at the patient level. We further examine high-risk procedures within those high-risk specialties, specifically craniotomies and high risk deliveries (HRDs). Taking the patient as the unit of observation, we then use a “difference in difference” regression to compare trends in travel times for high-risk versus low-risk procedures and for crisis years versus non-crisis years. We find that travel time for craniotomies increased more than lower-risk neurosurgery procedures at a statistically significant level. Notably, increases for rural patients were not driving our results, as the rural picture was decidedly mixed. As a second test, we examined physician exit from these procedures. For craniotomies and HRDs, we found that very low volume providers largely exited, a result that many people would praise as sensible. For craniotomies, however, high-volume providers also exited at much higher rates during the crisis years. In short, our findings lend scant support to the idea that medical malpractice liability has contributed to substantial physician exit, and suggest that even in the so-called “crisis” years, this was not a large enough phenomenon to be particularly troubling.

Jonathan Klick
AEI and Florida State University
[Professor Klick’s presentation was delivered on his behalf by Ted Frank due to unavoidable traveling difficulties.]

Existing attempts to measure the affect of medical malpractice liability on access to medical care have been problematic. The paper I am presenting here today attempts to present a better natural experiment based on the fact that not all specialties should be equally affected by medical malpractice reform. Indeed, some low-risk specialties should not really be affected at all. This study uses a triple differences model to compare the behavior of doctors in high-risk fields against the control group of those in low-risk specialties. The AMA defines various high- and low-risk specialties, and I used data from the American Tort Reform Association (ATRA) to determine when various reforms were in place. The results were statistically significant, with some reforms (non-economic damage caps, total damage caps, contingency fee caps, and collateral source reform) together producing an improvement in supply of 11 percent. Other reforms (joint and several liability reform, periodic payments, and victims’ funds) adversely affected supply of high-risk physicians, together producing a decrease of 8 percent. In short, doctors in some specialties do seem very sensitive to reforms, and medical malpractice reforms may affect access to care. This effect is likely to be concentrated in rural and minority areas, and targeted reforms seem more likely to be effective than broad medical malpractice reform. We must go on to ask whether the number of doctors actually matters to the quality of care, because we cannot assume a priori that fewer doctors in high-risk specialties will result in lower quality care. There are some studies (Klick and Stratman; Dubay, et al; Rubin and Shepherd; and Kessler and McClellan) suggesting this is the case, but more work needs to be done in this area.

Randy Bovbjerg
Urban Institute

These papers are a real improvement over much of the politically charged debate. Specifically, they ask the right question: how are the patients affected? They also seem to have the big picture right, in that they find an effect but not a huge one. One weakness of the papers is their failure to look at the problem as doctors and patients do. For example, Klick’s paper attempts to quantify the benefit of caps, but the truth is that the biggest benefit of caps is the boon they give doctors by eliminating the possibility of a verdict that will completely wipe them out, a basically irrational but not wholly baseless fear that cannot be insured against otherwise. The papers also fail to realize the extent of spillovers between specialties, both psychologically and in terms of premiums, all of which medical malpractice insurers set together rather than completely independent of one another. The data, which are used largely because they are available, also have real limitations. Trial lawyers will greet these papers by harping on those people who die in the ambulance and are therefore never included. Doctors will say that not enough time has gone by in the current malpractice crisis to properly measure the results. In the field of analysis and commentary, though, these papers are a real step forward. Another important avenue of research is the lawyers’ myth that the whole effect of the legal system on doctors is to make them more careful.

Ted Frank
AEI

While I have various concerns about some of the treatment of data in both of these studies, I think they both show that medical malpractice reforms can have an impact on patients’ access to care. In the Dranove study, I worry about the dulling effect of treating intra-zip code travel as taking just one minute, since zip codes vary so greatly in size. I also worry that the difference between emergency surgeries and scheduled surgeries is ignored, since there is a huge difference in the implications for access. Regarding obstetrics, I am concerned about the attempt to isolate HRDs as separate from delivering babies at all. When it comes to their insurance, doctors do not get a choice between doing all high-risk deliveries or no high-risk deliveries; the real choice is between delivering babies or not delivering at all. So when doctors exit from performing HRDs, they are likely exiting from all obstetrics. To fix this problem, I would suggest looking at other factors such as infant mortality. In the Klick study, I worry that the use of the ATRA data, which takes into account only the date of legislative passage, fails to account for the long lag time that a court system that often strikes down reforms imposes. Because insurance companies are often skeptical that courts will uphold reforms passed by legislatures, they may wait until the laws are battle-tested before there is ever any reduction in rates.

AEI research assistant Philip Wallach prepared this summary.