Economic costs of full-blown and subthreshold panic disorder

Abstract

Background

Data on the societal costs of mental disorders are necessary to inform health policies.

Methods

This study assessed the costs of panic disorder and subthreshold panic disorder, compared these with costs of other mental disorders, and assessed the effects of (psychiatric and somatic) comorbidity and agoraphobia on the costs of panic. Using a large, population-based study in The Netherlands (n = 5504), both medical and production costs were estimated from a societal perspective within a one-year timeframe.

Results

Annual per capita costs of panic disorder were €10,269, while subthreshold panic disorder generated €6384. These costs were higher than those of the other mental disorders studied. About one quarter of the costs could be attributed to comorbidity. Agoraphobia was associated with higher costs.

Limitations

Methodological choices influence cost estimates. In the present study most of these will result in conservative cost estimates.

Conclusions

Panic thus causes substantial societal costs. Given the availability of effective treatment, treatment may not only benefit individual patients, but also have economic returns for society.

Introduction

Mental health care suffers a lack of resources in most parts of the world (World Health Organization, 2001). In addition, mental health care is systematically underserved when the public health impact of disorders is compared to the investment made to provide adequate care (World Health Organization, 2001). To call attention to mental health care and to stimulate that allocation of resources to health care priorities are driven by the public health impact of disorders, empirical data regarding the economic ramifications associated with mental morbidity are needed. Cost-of-illness studies can provide these data as they translate morbidity into economic terms, thereby revealing the magnitude caused by a particular disorder in relation to the burden caused by other illnesses (Rice and Miller, 1998). Costs can be divided into direct medical costs, direct non-medical costs and indirect costs associated with morbidity. Direct medical costs consist of treatment costs incurred by a broad range of both formal and informal health services as well as the costs of prescription drugs. Direct non-medical costs consist of ‘out of pocket’ expenses that arise when people travel to health service providers. Indirect non-medical costs consist of costs due to production losses in both paid labour and in the domestic sphere that are caused by illness. Data regarding the cost-of-illness are especially relevant for disorders that both can be treated effectively but are often left untreated, and that afflict high costs on society. Treating these disorders more rigorously may well be cost-effective or even cost-cutting by reducing medical (non-psychiatric) costs and production losses (Magruder and Calderone, 2000).

From this perspective, the economic ramifications of panic disorder deserve attention. Panic disorder can be treated effectively (van Balkom et al., 1997, Bakker et al., 2002), but at least half the subjects with panic disorder do not receive help (Wang et al., 2005, Leon et al., 1997). Furthermore, panic disorder promotes several factors known to drive costs. Examples are inefficient help seeking behavior (Wang et al., 2005, Kessler et al., 1999, Wittchen et al., 1998), misdiagnosis and subsequent use of inappropriate services (Harvison et al., 2004), excess disability (Kessler et al., 2005, Alonso et al., 2004a, Kouzis and Eaton, 2000) an unfavourable natural course (Eaton et al., 1998, Robins et al., 1991, Wittchen, 1988), and is likely to be correlated with substantial production losses due to work loss and work cutback days (Alonso et al, 2004a). Furthermore, comorbidity of panic disorder with agoraphobia (Weissman et al., 1997) and other mental disorders (Alonso et al., 2004b) is common, which further raises costs. As panic disorder affects many people (Kessler et al., 2005), societal cost may be substantial.

Recently we were able to show that panic disorder indeed ranked among the most expensive common mental disorders, both expressed per sufferer (called per-capita costs) and on a societal level (Smit et al., 2006). This confirmation warranted further research on two aspects. First, it is quite possible that milder panic, called subthreshold panic disorder, generate substantial costs as well since it is highly prevalent (Kessler et al., 2006) and like panic disorder associated with cost rising factors such as service use (Leon et al., 1995), disabilities (Kessler et al., 2006, Batelaan et al., 2007), agoraphobia (Batelaan et al., 2007) and comorbidity with other disorders (Kessler et al., 2006, Batelaan et al., 2007), though to a lesser extent compared to panic disorder. Second, as both agoraphobia and other comorbid disorders are often present in panic, the effects on costs of panic disorder and subthreshold panic disorder need to be determined.

To our knowledge, this cost-of-illness study is the first to address both issues, while using data from a large population survey. Our first aim was to assess the total societal costs (i.e. direct medical costs, direct non medical costs and indirect costs) of subthreshold panic disorder and panic disorder and to compare these with total societal costs of other mental disorders. Our second aim was to assess the effects of (somatic and psychiatric) comorbidity and agoraphobia on the costs of panic.