Collaborative for Healthcare research in behaviOral economIcs and deCision sciEnceS (CHOICES)
Decisions are integral to our lives—some decisions are simple, while others are more challenging. Multiple disciplines are interested in how we make decisions and change behavior. Psychologists study cognitive and behavioral aspects of decision making while consumer behavior is applied to optimize marketing and business campaigns. Economists are divided in their approach. Classical economic theory assumes that people make decisions in a rational way by seeking out their own best interest. This assumes people have the mental capacity to deal with huge amounts of information and can resistant attempts to bias preferences through manipulation. In a behaviorally-oriented perspective, behavioral economists acknowledge human limits on computational power, willpower, and self-interest may lead people to not act in an economically rational fashion. All disciplines agree that human beings have difficulty defining and making good choices. This is particularly true under time-limited, stressful, complex, or uncertain conditions.
Some of the more difficult decisions we make involve healthcare choices or interventions for ourselves or for others. Clinicians routinely make difficult decisions under conditions of uncertainty, time constraints, and in complex environments. These decisions can be informed by medical literature, clinical decision rules, or predictive models. In practice, however, clinicians employ a number of subjective “rules of thumb” to form their clinical judgments. These behavior patterns are usually based on personal experience and beliefs. Clinician behavior also appears to be significantly influenced by salience or recency. Indeed, many clinicians overemphasize personal experience by allowing recent incidents to bias decisions. Other evidence suggests it is difficult for them to interpret statistics and correctly apply novel clinical research. All of these subjective judgments result in an environment where it can take 12 to 17 years to translate research into practice. Therefore, to alter and improve clinician behavior, we need new approaches to understand the clinical decision making process, identify predictable decision making flaws, and act on them.
Starting in 2012, faculty from the Foster School of Business, the Natural Sciences Division in the College of Arts and Sciences (Department of Psychology), School of Public Health, School of Medicine (Departments of Global Health, Radiology and Surgery) and the School of Pharmacy started meeting monthly to discuss the application of behavioral economics and decision sciences to healthcare. We share information about different projects and ideas, and look for opportunities to collaborate across projects. Our main focus is to explore how the principles from behavioral economics and decision sciences are being used or can potentially be used to influence clinician behavior to improve the care of their patients and the efficiency of the healthcare system.
Acute appendicitis is a common surgical condition, and the traditional conceptual disease model identifies time as a critical determinant of perforation. From this perspective, most cases of acute appendicitis proceed toward perforation unless timely intervention occurs. In this model, behaviors play a prominent role in determining outcomes: how the individual behaves in seeking care and how the healthcare system behaves in diagnosing and treating the patient can both potentially affect outcomes. Studies that have tried to evaluate associations between perforation and socioeconomic factors have generally used demographic characteristics -- median income per ZIP code, race, insurance type, etc. -- to make inferences about behavior. But, to our knowledge, no studies have been performed in which patients are directly surveyed about their pre-hospital experience and decision-making.
We are conducting a multi-site survey to ascertain and quantify how decision-making patterns and patient reported socioeconomic characteristics impact time-to-presentation, and possibly, clinical outcomes such as perforation, when patients have acute appendicitis.
We are studying the impact of adverse events in risk-avoidance behavior. Specifically, we are assessing whether intraoperative colorectal leak-testing patterns change after an adverse event. Disruption of the colorectal anastomosis is a difficult complication that leads to significant morbidity and mortality at high cost to the patient, the health care system, and society. A standard method to test the potential for anastomotic leak is intraoperative anastomotic testing with isotonic sodium chloride solution, povidone-iodine, or air insufflation of the anastomosis. Using the Surgical Clinical Outcomes Assessment Program (SCOAP) database, we are quantifying any impact adverse events might have on risk-avoidance behavior. Specifically, we are assessing how leak-testing patterns change after adverse events that are either personally experienced or experienced by a colleague. This analysis will provide compelling data for additional funding to explore other behavior patterns in high-risk surgical situations that could help prevent and reduce postsurgical complications. For more information, contact Val Simianu.
We are exploring potential differences between the preferences patients and physicians can make facing an end of life scenario. It has been reported that during only the last 6 months of life a person in the US can expend up to 80% of their lifetime health expenditures and receive the most invasive and highest costs interventions without significantly improving their quality of life or their life expectancy. Also, research has found problems with end-of-life care, leading many analysts to conclude that existing patterns of care do not meet the needs and preferences of terminally ill patients. We are exploring the attributes and the value assigned to them by patients and physicians when faced with an end of life situation, specifically a terminal cancer. Using a method called Multiple Criteria Decision Analysis (MCDA) we are eliciting the criteria and the weights assigned by patient under an end of life scenario and comparing it to the results provided to a similar hypothetical scenario answered by physicians. If the results of the MCDA include the same criteria and similar weights, this model and the information provided can be very helpful in clinical practice to help physicians and patients to discuss alternatives at the end of life.
In healthcare, decisions made after diagnostic tests often lead to further testing and treatments for incidental findings. We are working to determine whether informing physicians of the frequency of abnormal findings unrelated to symptoms, the costs of the tests, and the efficacy of treatments for patients with low back pain reduces unnecessary additional testing and treatment, including surgery. Low back pain is the most common type of pain in adults with annual costs over $25 billion. Magnetic Resonance Imaging (MRI) is used for diagnosis and treatment decisions. Abnormalities in MRI of the spine can be very common in subjects without symptoms. In patients with low back pain these abnormal findings are frequently reported, causing unnecessary testing and potentially harmful or ineffective invasive interventions. Our experiment simulates real-life scenarios for 300 physicians treating patients with low back pain to determine the effect of providing additional information on MRI reports as previously described. We expect this additional information to change physician behavior regarding additional imaging and treatment of low back pain, including referral to surgery. For more information, contact Val Simianu.
An evolving project in our collaborative has attempted to quantify risk-thresholds for physicians, and determine whether surgeons and clinicians of different levels of training make different decisions than non-physicians in situations of high uncertainty. Using a model developed by our collaborators in the Department of Psychology, we are using revealed preferences in a model of salting roads in inclement weather (1) to describe a clinician’s inclination towards risk-avoiding choices and to see how they relate to choices made in clinical scenarios where there is uncertainty about the value of surgical intervention (2). For more information, contact Val Simianu.
The Spine Lumbar Fusion Outcomes Calculator can help facilitate discussion among physicians, patients, and their families about the likelihood of pain & function improvement 60 days to one year after lumbar fusion surgery.
Contact us to indicate your interest in learning more about CERTAIN CHOICES.