Transforming mental healthcare delivery through implementation science and behavioral economics

Abstract

Psychiatric disorders account for more years lived with disability globally than any other disease category. Currently, effective mental health treatments are variably deployed in community practice and patient outcomes are less robust than clinical trials. A major frontier in psychiatry is to develop strategies that increase the use of evidence-based assessment, prevention, and intervention approaches, galvanizing the new field of implementation science. In this Viewpoint, we argue that we must bring concepts from behavioral economics to implementation science to accelerate the reach and impact of psychiatric treatments.

What is implementation science?

Implementation science is the scientific study of methods to increase the adoption, implementation, and sustainment of evidence-based practices into routine care. The objective of implementation science is to improve the quality of health and mental healthcare to improve lives. The emergence of the field stems from the realization that many scientific discoveries never make their way to the individuals who need them, a ubiquitous problem in health [1] and mental health [2]. To date, implementation science has generated frameworks, study designs that answer questions related to effectiveness and implementation, validated measures, and established implementation outcomes. A major contribution of the field is the understanding that the context in which one implements is paramount to maximizing successful implementation. Fundamentally, implementation science is focused on clinician behavior change within organizational constraints as a key target to improve care quality and patient outcomes. A range of approaches (e.g., organizational theory, psychology, human factors, and systems science) have been applied to understand how best to change clinician behavior change within organizations. A major limitation of the field is the assumption that clinicians behave as “rational actors,” or that clinician decision makers maximize utility when delivering clinical care. We argue that the rich literature of behavioral economics offers an alternative and complementary perspective to other leading approaches that implementation science borrows from in viewing clinician decision making as human decision-making, thus allowing us to leverage known decision making patterns to make it easier for clinicians to do the right thing.

How can BE take a fresh lens to implementation science?

Behavioral economics (BE) was galvanized by the work of psychologists Tversky and Kahneman, who observed unexpected patterns in human decision-making. BE were popularized by economists like the Nobel laureate Richard Thaler, who posited that these patterns of decision making offer opportunities to “nudge” individuals to make more effective decisions by changing their environment (i.e., choice architecture). A growing number of investigators and policy makers have applied these concepts to shape public health behavior, and more recently patient behavior. For example, a landmark trial demonstrated the effectiveness of a suite of behavioral economic interventions such as lottery financial incentives and deposit contracts for weight loss [3]. The rise of the electronic health record has also been leveraged to harness these psychological insights to shape clinician behavior. One study found that changing the default in the electronic health record dramatically changed clinicians’ prescribing behavior, spawning a highly applied research program and nudge units in health systems [4]. These studies inspired our group to apply these BE approaches to improving mental healthcare.

To date, implementation science and BE have rarely informed one another. The field has seldom acknowledged the “humanness” of clinician decision making or leveraged those patterns as targets of implementation strategies. Further, implementation strategies tend to be complex and time- and resource-intensive [5] whereas BE strategies tend to be low-cost and easy to scale [6]. We can infuse implementation efforts with BE strategies, which shape the environment to make the “right” (i.e., evidence-based) choice the “easy” choice. BE strategies include making desired choices the default, making them attractive (e.g., using graphics and simple approaches to share information), making them social (e.g., using social comparison), and making them timely (e.g., providing reminders and feedback immediately before and after the desired behavior) [7].

What have we learned in applying implementation science and BE in our work?

We have applied these concepts through our Penn NIMH ALACRITY Center that aims to increase implementation of evidence-based practice for individuals with psychiatric disorders across the lifespan. In one instance, we examined why clinicians do not use evidence-based psychosocial treatments in session, using a BE lens. For example, humans underestimate the influence of their emotional (e.g., anxiety) and physical (e.g., hunger) states and how these states might affect their behavior – termed the hot-cold empathy gap [8]. Learning about exposure therapy during training is different from applying exposure therapy in vivo. Drawing on the hot-cold empathy gap, in theory, clinician emotions and distress during the session may influence application of this evidence-based technique. Indeed, when we use BE taxonomies to delineate implementation barriers, we find evidence for the hot-cold empathy gap, such that clinicians underestimate the influence of their own anxiety on their behavior in session with clients [9]. In another example, we tested the effect of escalating and deescalating financial incentives to improve patient medication adherence in people with major depressive disorder, finding that escalating incentives for daily antidepressant adherence significantly improved adherence compared to a control group during the first 6 weeks of treatment, by tapping into the BE concept of loss aversion – or the idea that patients who initiate treatment face ever-greater lost opportunities if they discontinue medication use [10].

Challenges and Future Directions

Bringing a BE lens to implementation science can transform mental healthcare across the lifespan. However, BE is not a solution for every problem, nor is it a panacea for major infrastructural, social, and financial barriers often rife in mental health care settings. Identifying the right behaviors for which behavioral economics can boost implementation approaches is critical. BE approaches are best applied to behavioral barriers (i.e., cognitive or psychological processes that operate prior to or during decision-making, or behaviors that get in the way of achieving a target behavior) rather than structural barriers (e.g., policy) that include factors external to the person that may get in the way of engaging in a specific behavior. A related challenge is that implementing complex behaviors may be less amenable to BE approaches than implementing simple behaviors. For example, prescribing behavior for psychotropic medications may be ripe for BE approaches, whereas it may be more difficult to use BE approaches with complex psychosocial interventions because there are multiple contingent behaviors in play. Further, many of the most successful applications of BE concepts in healthcare have leveraged the electronic health record and other technologies that are not available in under-resourced settings such as community mental health. We need to better understand how to translate these concepts using low-tech solutions. Finally, ethical questions when applying these approaches, especially to patients, are important given the sensitivity of mental health interventions. Although researchers have conducted similar trials in other health conditions (e.g., smoking cessation), there may be other considerations for mental health interventions.

Implementation science offers the potential for realizing the discovery of our scientific discoveries taken to scale. We would do well to draw from the powerful insights and theories of behavioral economics to optimize the effectiveness of our implementation strategies to ensure that we design approaches that recognize how clinicians actually behave, rather than limiting our scope to how we hope clinicians will behave.