Experiments on behaviour and decision making in health
Behrendt, Hannah Ariadne
Research in judgement and decision-making has identified numerous ways in which human decisions are likely to be biased, deviating systematically from the behaviour one would expect if humans were fully rational 'Econs', maximizing their utility with perfect self-control. Suboptimal decision-making as a result of these biases imposes large costs on individuals and society. In this thesis I investigate experimentally how decision-making can be improved, focusing on the health domain, where errors ultimately become a matter of life and death. Chapter 1: Providing feedback has been shown to be an effective way to change behaviour across several domains, including energy use and the workplace. However, there is little evidence on the effect of providing feedback in high stakes environments, where people make risky decisions with potentially catastrophic losses. In the first chapter I conduct a field experiment that provides feedback and information in such a context - emergency healthcare. Understanding the drivers of patients' decision-making relating to their demand for healthcare services and how they can be directed to the most appropriate services at any given time is a challenge for health systems across many countries. This trial aimed to reduce avoidable Emergency Department (A&E) attendances by sending a personal feedback letter to people who recently attended an Emergency Department but whose health concerns could have been dealt with elsewhere. Patients were randomly allocated to either receive a follow-up letter with information on alternatives to A&E or no letter (usual care). Overall, I do not find a statistically significant difference between the re-attendance rates of patients who did and did not receive the letters. However, the effect of the intervention interacted significantly with patient age, especially in men. I develop a conceptual framework that explores possible explanations for these heterogeneous effects. Chapter 2: Policy interventions drawing on insights from behavioural sciences are increasingly popular and have been successfully applied across a number of different policy areas. However, little attention has been paid to the extent to which the effects of repeated behavioural interventions are sustained over time. In the second chapter I study this question through a natural field experiment in the English National Health Service (NHS). The intervention consists of changing the salience of waiting time on the clinicians' e-Referral Service (ERS) interface. Waiting longer can negatively affect patients' health gains from receiving treatment, so referring patients to services with shorter waiting times can be of benefit. I find, through a stepped wedge trial, that putting a simple alert against services with high waiting times leads to a 35 percent reduction in the share of referrals to these services. The effect of the intervention is sustained over time and does not vary with prior referral habits. A small, low-cost intervention increasing the salience of waiting time to clinicians has a powerful and sustained effect on the choices of their patients. Chapter 3: Behavioural attitudes toward risk and time, as well as behavioural biases such as present bias, are thought to be important drivers of unhealthy lifestyle choices. While the first two chapters of this thesis take behavioural biases as given, the third chapter makes a first attempt at exploring the possibility of training the mind to alter these attitudes and biases, in particular relating to health-related behaviours, using a randomized controlled experiment. The intervention we consider is a well-known psychological technique called "mindfulness", which is believed to improve self-control and reduce stress. We conduct an experiment with 139 participants, around half of whom receive a four-week mindfulness training, while the other half are asked to watch a four-week series of historical documentaries. We find strong evidence that mindfulness training reduces perceived stress, but only weak evidence of its impact on behavioural traits and health-related behaviours. We do not see that engagement with mindfulness training is correlated with behavioural characteristics such as impulsiveness and impatience. Our findings have significant implications for a new domain of research on training the mind to alter behavioural traits and biases that play important roles in lifestyle.