Health and economic consequences of low-value mental health care: the case of benzodiazepines

Mental health problems are a leading cause of disability and suicide worldwide and their incidence is increasing over time (James et al., 2018). This has many reasons, including the 24h society which has increased sleep problems and anxiety, which in turn lead to the onset of other mental and physical illnesses (e.g., Giuntella and Mazzonna, 2019). There is also evidence that the pandemic has led to a mental health crisis with life-style disruption, increasing anxiety and depression (Giuntella et al., 2021). Yet, the increasing awareness of the social and economic burden of mental illnesses has contributed to growing mental health care costs, and to an increasing consumption of psychotropic drugs (e.g., Jorm et al., 2017) and concerns about overdiagnosis and overtreatment (e.g., Hertzberg et al., 2021).

In this project we focus on the causes and consequences of benzodiazepine prescription and use. Benzodiazepines are part of the symptomatic management of mental health disorders. While these drugs are relatively cheap and provide short-term relief from insomnia and anxiety symptoms, they might be highly addictive, they have side-effects and there is no evidence on their long-term efficacy. As a result, the global Choosing Wisely initiative has included their use, especially among the elderly, in the top list of low-value care treatments. Despite the global effort to reduce their use, benzodiazepines are frequently prescribed and are often widely used without following current guidelines. Moreover, alternative treatments such as cognitive-behavioral therapy are scarcely used, even though they are known to be safer and more effective in the long-term (Cronin et al., 2020).

The overarching aim of this project is to understand the drivers of benzodiazepine prescriptions and use, and assess their impact on health and labor market outcomes. We aim to investigate the demand-side, supply-side and institutional factors that influence use, and to leverage these to identify the causal impact of (mis)use on a range of health and labor market patient outcomes. We will rely on various data sources including administrative and insurance claim data, from two countries with different institutional arrangements and regulations with regards to benzodiazepine prescriptions and practices: Switzerland and the Netherlands.

By combining prescription, diagnosis and other characteristics of patients, we aim to compare long-run outcomes of patients after their first treatment with benzodiazepines or with similar initial diagnoses, but that receive different treatments. Since the initial treatment might not be at random, we are going to leverage life event triggers and plausibly exogenous variation in doctors’ practice style. We also plan to exploit differences in insurance design and coverage in the Swiss setting (e.g., HMO and quality circles, where doctors commit to prescribe less benzodiazepines). Moreover, given the dramatic age gradient in benzodiazepine use, we will investigate the causes and consequences of the sharp increase of their use at nursing home admission.

Our results will contribute to allocate more efficiently resources in mental health care and help to improve patients’ health. More generally, they help to better understand how and at which level to intervene to reduce the use of low-value services and provide evidence on the broader economic consequences of low-value care. Both are key to address large inefficiencies in our health systems.