Reflections on Healthcare Economics: The Costs of Defensive Medicine

Physicians frequently cite defensive medicine (i.e., ordering extra tests, imaging, studies to fend off potential lawsuits) as a contributor to soaring health costs. Though the jury is still out on the extent of the costs of all this extra, often unnecessary, work up, it is undeniable that a significant proportion of physicians frequently consider the legal implications of their medical decision-making. 

Brief background and statistics: Surveys of physicians have shown that significant proportions (80-90%) report practicing defensive medicine. Estimates of how much this practice contributes to healthcare costs vary widely, since the subjectivity of defensive medicine makes it very difficult to measure retrospectively. Most experts believe anywhere between 1-2% of total health spending is attributable to defensive medicine, but this is extremely challenging to calculate. If one ignores the costs of defensive medicine and focuses on the liability system, the costs amount to a puny 1.5% of healthcare spending. A recent prospective study attempting to quantify the incidence and cost of defensive medicine in Orthopedic Surgery revealed that defensive practices accounted for 19.1% of orders (x-rays, MRIs) and 34.7% of the cost. 

The controversy arises from the fact that many health policy experts do not believe that tort law reform will result in the dramatic cost reductions that physicians often predict. They point to states where capped damages have not necessarily resulted in reductions in cost. Furthermore, there is a concern that patients who are victims of gross negligence may suffer from inadequate compensation in states with capped damages. 

This line of thinking would suggest that tort reform may not be the most urgent priority for our legislators. We have yet to tackle increasing coverage, changing reimbursement, improving quality, etc. So why do physicians advocate for tort law reform?

Although the contribution to the total may seem minimal, 1-2% of billions of dollars is not insignificant. Furthermore, with the coming changes to the healthcare system that reward efficient provision of healthcare (i.e. not ordering unnecessary tests, imaging and procedures), it will be important to reduce the practice of defensive medicine. Consider a scenario: 

Imagine that you are an emergency medicine physician about to see a 50-year-old patient who experienced a witnessed fainting episode - his wife tells you he was standing up from sitting down on a hot day and fell backwards onto their sofa and was briefly unresponsive before regaining consciousness. He is otherwise healthy and has no other medical problems. He denies any other recent symptoms, medications, travel, or sick contacts. His vital signs and physical exam are totally within normal limits. Would you order a head CT scan? Or would you order a chest x-ray and cardiac injury markers routinely in young patients presenting with uncomplicated chest pain?

These scenarios often result in divergent answers from different clinicians. Some practitioners would want to be sure they are not missing something dangerous (like an intracranial hemorrhage or myocardial infarction, respectively). Others would be reassured by the benign presentation of these patients and hold off on ordering further testing. This spectrum exists across every clinical decision because of differing appetites for risk from clinician to clinician. This is a phenomenon influenced by many factors, such as level of training/experience, clinical context, the patient seen just before this one, and even the state of being in an active lawsuit! For example, my clinical experience may drive me to "want to be sure" and order the extra test, whereas my attending may feel comfortable without the extra testing. Miller et al show that clinicians with a recent malpractice suit are more likely to order tests for defensive reasons. Atul Gawande describes studies where the EKG diagnosis determined by a clinician is influenced by the diagnosis of the previous EKG. Experience can also be a relative handicap - if the first two patients you prescribe blood thinners to happen to end up with fatal gastrointestinal or intracranial hemorrhage, you may hesitate prior to prescribing the same medication to the third patient. This is the challenge of different risk appetites in different clinicians for different clinical situations.

Take this already complex situation and add the caveat of bundled payments that reward "efficient" practice. Will a doctor in the hypothetical clinical scenario described above forego the head CT scan if he or she knew that their paycheck may be affected? It depends. Personally, I would rather take the hit on my paycheck rather than risk malpractice claims that can destroy reputations and instill years of anxiety and distraction associated with litigation. 

To me, the strongest indication for tort law reform is to better enable clinicians to practice efficient medicine in the bundled payments model. Physicians will be more likely to only order the necessary testing/imaging if the persistent paranoia of litigation is attenuated, as articulated in this article in Journal of General Internal Medicine. 

How can we address this issue going forward?

There are many potential solutions:
1) Litigation for gross negligence: Some states (e.g., Georgia) have attempted to move towards medical malpractice only for gross negligence. This means that all reasonable, well-trained clinicians would be expected to make certain decisions, order certain tests, and prescribe certain medications in a given clinical situation. Physicians not adhering to this standard and making decisions that harm patients are viewed as practicing in a "grossly negligent" fashion. An example of gross negligence is the physician that fails to order cardiac enzymes and an EKG in a male patient with a cardiac history and diabetes who presented with "crushing" sub-sternal chest pain radiating down the left arm. Interestingly, in Georgia and Texas, two states with recent tort law reform, mortality rates have remained the same or declined. Although this is a crude measure of patient safety, there is no evidence that patient safety has been sacrificed. 
2) Capped damages: Some people have proposed that capping damages in medical malpractice claims can potentially reduce frivolous claims. However, in some places where this has been enacted, health care costs have not come down. Furthermore, there are concerns that patients with legitimate claims may not receive adequate compensation.
3) "Safe harbor": Similar to the gross negligence concept, the "Safe Harbor" model proposed by the Center for American Progress recommends the definition of national evidence-based guidelines and utilization of clinical-decision support systems. These would define the actions a trained and reasonable physician would be expected to take in a given clinical situation. Patients (and their lawyers) would have to establish that these clinical standards were not met when presenting a claim. This would reduce a problem faced currently: the fact that there are wide variations in practice from one health system to another. We have begun to address the issue - under the Choosing Widely initiative, 35 specialties released standard recommendations on common testing modalities and imaging. This effort is laudable and should be expanded to national standards for most common clinical situations - the challenge will be in nationwide implementation.  
4) Non-adversarial medical malpractice compensation: It is now widely accepted that individual blame is counter-productive when a medical error occurs and patient safety is compromised. Most experts recognize and advocate that medical errors occur due to systemic failures in the provision of healthcare (errors occurring in the "Swiss cheese model"). This concept is described in "To err is human" and has revolutionized the way hospital systems address and improve patient safety. While there is evidence that fear of lawsuits affects medical-decision making, there is no evidence for reduced rates of medical error with the current litigious atmosphere. For this reason, some have proposed different ways of addressing fair compensation when a medical error occurs. One option is a "workers' compensation" type of board (under the Patient Compensation System) where patients can take grievances, instead of having physicians going to court. Such a board would provide a means for compensation, physicians would still pay malpractice premiums, and there would be a venue for addressing negligence and suspension of licensure for egregiously gross negligence. 

Tort law reform will not be the panacea of health care cost reduction. However, it is clear that physicians' decisions are affected by fear of litigation. There is no evidence, however, that the threat of litigation has improved patient safety. The new paradigms focusing on the systemic causes of medical error provide more promising means of improving patient safety. Until the fear of litigation is attenuated, it may be difficult to achieve rapid gains in cost reduction under the bundled payments model. 

“No matter what measures are taken, doctors will sometimes falter, and it isn't reasonable to ask that we achieve perfection. What is reasonable is to ask that we never cease to aim for it.” 

― Atul Gawande, Complications: A Surgeon's Notes on an Imperfect Science