Does Defensive Medicine Exist?

Despite the common thought that most physicians perform defensive medicine on a regular basis, the evidence bearing this out has been inconsistent at best.

Ninety-three percent of physicians report practicing defensive medicine; that is, performing tests or therapies that provide little or no value to the patient, or withdrawing care altogether, because of the threat of lawsuit.  However, studies looking into the practice habits of high-risk specialist physicians such as obstetricians before and after tort-reform often fail to show change.

One landmark study by Kessler and McClellan in 1996 does, however, suggest that tort reform does reduce the unnecessary testing that exists, defined in that study as reduced intensity of care with consistent measures of quality outcomes.  Any such trend in the setting of tort reform, they conclude, implies that the higher level of care formerly performed were extraneous and therefore defensive medicine.  Specifically, their study showed that in states that implement tort reform versus states that do not implement such reform, there is a five to nine percent decrease in the Medicare payments for hospital care of patients over 65 years old with ischemic heart disease. However, the rate of repeat admissions and mortality from heart disease stay constant.  Of note, this study analyzes the effect of two subcategories of reform: (1) tort that is aimed at directly decreasing the amount of payment awarded per claim, or direct reforms, and (2) tort reform that reduces the pressure tort has on care providers but only indirectly affects award amounts, or indirect reforms. Direct reforms most commonly include caps on damages and allowance of collateral source offset, while indirect reforms include joint-and-several liability reform and limits on plaintiff attorney contingency fees.  Both direct and indirect reforms were effective by Kessler and McClellan’s measures of reducing defensive medicine.

This current study by Sloan and Shadle used the same data set in patients over 65 years old from 1985 to 2000 to measure effects of tort reform, but attempted to make the analysis more generalizable.  In contrast to the 1996 study, they broadened their definition of medical care quality as measured by mortality from one disease state to four disease states (acute myocardial infarction, breast cancer, diabetes, and stroke) and measured all Medicare payments (not just hospital payments) in the one year period following diagnosis of the target illnesses.  Surprisingly, this analysis showed that while one-year mortality remained constant, there was no significant change in the amount of Medicare payments for any individual disease state for either direct or indirect tort reform.  When hospitalizations for all diseases were considered together, there was actually a slight increase in Medicare payments after indirect tort reforms with a concurrent increase in one-year mortality.  No such changes were seen for all hospitalizations affected by direct reforms.  The authors were unable to interpret the while attempting to single out effects of individual types of tort reform.

Commentary
Although almost all physicians report their clinical practice to be affected by the threat of malpractice suits, literature studying changes in malpractice insurance premiums, the frequency of medical malpractice claims, and various types of tort reform yield unreliable effects on physician practice and patient outcomes.  This current study, similar to the 1996 study, is measuring the wrong physician activity.  It is logical that in a highly litigious environment, physicians would more often tend to order unnecessary tests to diagnose a disease, or rather to not miss a disease, than tend to increase intensity of care once the diagnosis is already made, as was measured in these studies.  Moreover, unlike this study, analyses of practice habits and patient outcomes in response to tort reform will not be useful unless policymakers are able to derive which specific reform methods are effective and therefore would translate best to statutory change.

The Quarterly Journal of Economics. 1996; 111 (2): 353-390.

JAMA. 2005; 293 (21): 2609-2617.

Journal of Health Economics. 2009; 28: 481-491.

by
Lisa J. Maurer, MD

One Reply to “Does Defensive Medicine Exist?”

Comments are closed.