Developments in the United Kingdom

With the goal of significantly increasing public access to evidence-based psychological therapies for depression and anxiety, the UK's National Health Service (NHS) rolled out the Improving Access to Psychological Therapies (IAPT) initiative in 2008 (Reference O'Shea and ByrneO'Shea & Byrne, in press) By March 2011, 3660 new cognitive behavioural therapy (CBT) workers had been trained and 60% of the adult population had access to these services. Moreover, continued investment will ensure that by 2014, IAPT will provide interventions to 900 000 people with depression and anxiety, with recovery rates averaging 50% (Clark, Reference Clark2011; Centre for Economic Performance, 2012).

Although IAPT was informed by National Institute for Health and Clinical Excellence (NICE) best-practice guidelines (NICE, 2009; NICE, 2011) its successful rollout was strongly influenced by reports presented to the UK government by Lord Layard and his colleagues from the London School of Economics (Layard etal. Reference Layard, Bell, Clark, Knapp, Meacher, Priebe, Turnberg, Thornicroft and Wright2006; Layard etal. Reference Layard, Clark, Knapp and Mayraz2007). These (and follow-up) reports encompassed detailed economic analyses that demonstrated that IAPT more than pays for itself (i.e. it produces a net economic gain; see Table 1).

Table 1 How IAPT pays for itself (Layard etal. Reference Layard, Bell, Clark, Knapp, Meacher, Priebe, Turnberg, Thornicroft and Wright2006; Layard etal. Reference Layard, Clark, Knapp and Mayraz2007; IAPT, 2011)

IAPT, Improving Access to Psychological Therapies; GP, general practitioner; GDP, gross domestic product; NHS, National Health Service.

Mental health in the UK receives 13% of the NHS budget (Centre for Economic Performance, 2012). This is one of the highest health expenditure allocations in Europe (IAPT, 2011) but there are continued economically based arguments being put to government that make further increases in investment more likely. For example, a recent London School of Economics report indicated that increased expenditure on therapies for the most common mental disorders (e.g. through the IAPT initiative) would almost certainly cost the NHS nothing as it would lead to substantial reductions in the £10 billion per year spent on physical healthcare caused by mental ill-health (Centre for Economic Performance, 2012).

Developments in Ireland

The proportion of the health budgetallocated to mental health services in Ireland has steadily declined in recent years – from 13% in 1986 to 5.3% in 2010 (Mental Health Reform, 2010). To address the associated under-development of these services, A Vision for Change (VFC) (Department of Health & Children, 2006), proposed the provision of integrated, recovery-focused care that is delivered in the community, primarily by multi-disciplinary Community Mental Health Teams (CMHTs). However, despite recent developments such as the recruitment of over 400 health and social care professionals to professionally complete CMHTs and the rollout of counselling services in primary care (Ward, Reference Ward2012) on the whole the implementation of the recommendations from VFC has progressed at a slow pace [Health Service Executive (HSE), 2012].

Progress has undoubtedly been hindered by a recessionary economic climate in which the HSE has recently cut €53 million from its mental health and primary care budget to offset its current budget deficit (Wall, Reference Wall2012). However, figures also suggest that outdated approaches to mental health service provision remain prominent. For example, between 2007 and 2008, there was a 19% increase in the prescription of the anti-depressant Mirtazapine and a 10% increase for the benzodiazepine Alprazolam under the General Medical Services scheme [Mental Health Commission (MHC), 2011]. Furthermore, in both 2009 and 2010, over €100 million was spent by the HSE on prescriptions for mental health difficulties in primary care (Primary Care Reimbursement Service, 2009; Primary Care Reimbursement Service, 2010).

In contrast to the United Kingdom, there has been a profound neglect of economic evaluations of mental health services in Ireland (Gibbons etal. Reference Gibbons, Lee, Parkes and Meaney2012) and this may partly explain the lack of governmental investment in developing the area. However, one of the few evaluations that has been undertaken found that costs arising from mental health (i.e. direct care and decreased economic output) amounted to 2% (or €3 billion) of GNP in 2006 (O'Shea & Kennelly, Reference O'Shea and Kennelly2008). Due to this significant economic burden and the expressed willingness of surveyed members of the public (n = 435) to pay extra taxation to fund community-based care, the report stressed that investing in mental health services is essential from an economic perspective (O'Shea & Kennelly, Reference O'Shea and Kennelly2008).

Another welcome economic report undertaken by the Kildare/West Wicklow Adult Mental Health Service found that providing acute care in a community setting cost ∼27% less (per capita) than providing such care in traditional acute settings (Gibbons etal. Reference Gibbons, Lee, Parkes and Meaney2012). This ‘Comprehensive Community Model’ was also more efficient – it reduced waiting times and had higher attendance rates. However, far more reports of this nature are needed to move Ireland away from its long tradition of ignoring the economic aspects to mental health (O'Shea & Kennelly, Reference O'Shea and Kennelly2008). Moreover, if mental health services wish to secure funding for much-needed service development (as the NHS did for IAPT), it is essential that development proposals have a comprehensive economic rationale, especially in our recessionary economy. The next part of this paper highlights ways in which mental health services can incorporate an economic dimension into their service evaluations.

Cost-benefit analysis

In a cost-benefit analysis, monetary values are placed on both the inputs (costs) and outcomes (benefits) of a particular programme or service. This allows policy makers and stakeholders to determine whether or not an overall net gain (i.e. when total benefits exceed total costs) is offered economically (Robinson, Reference Robinson1993). Moreover, cost-benefit analyses allow policy makers to consider the relative efficiency of various potential investments and are thus a useful decision-making tool (Tudor-Edwards & Thalany, Reference Tudor-Edwards and Thalany2001).

In a cost-benefit analysis conducted on a mental health service, all of the benefits from the service's interventions (e.g. ‘recovery’ from a mood problem) are converted into monetary benefits so that they can be compared against the monetary costs of providing these interventions (Hoch & Smith, Reference Hoch and Smith2006). For example, in IAPT's cost-benefit analyses (Layard etal. Reference Layard, Bell, Clark, Knapp, Meacher, Priebe, Turnberg, Thornicroft and Wright2006; Layard etal. Reference Layard, Clark, Knapp and Mayraz2007; IAPT, 2011), estimated improvements in service users’ clinical outcomes were converted into arising monetary benefits such as increased workforce productivity (i.e. less sick days and reduced use of incapacity benefits) and reduced healthcare costs (also referred to as the ‘medical cost offset’). These benefits were projected over a 2-year period and compared against the cost of providing therapy (which took into account salary, equipment and other costs). As the benefits of IAPT easily exceeded its costs (as shown in Table 1), the analyses demonstrated that the initiative produced a net gain economically. Examples of costs and benefits that can be associated with mental health services are presented in Table 2.

Table 2 Costs and benefits that can be associated with mental health services

An essential element of a cost-benefit analysis is deciding which benefits to measure and how to measure them (Hoch & Smith, Reference Hoch and Smith2006). A structured method of measurement can be undertaken using the 26-item Trimbos and iMTA Questionnaire on Costs Associated with Psychiatric Illness (TiC-P) (Hakkaart-Van Roijen, Reference Hakkaart-Van Roijen2002). The TiC-P allows the systematic collection of data pertaining to both medical resource utilisation and costs attributable to production losses. The former is measured by asking service users how many contacts they had with different healthcare providers [e.g. general practitioner (GP), psychiatrist, medical specialist, physiotherapist, hospital] and their frequency of medication use during a set period of time (e.g. 3 months). The latter is measured by the reported number of days of absence from work, in terms of both short-term (<2 weeks) and long-term absence.

At 26 items, the TiC-P may be considered too lengthy by service providers and users. Hence, we present a newly formulated 12-item cost evaluation tool – the ‘EcoPsy 12’ (see Table 3). This tool's content and structure are informed by the TiC-P (Hakkaart-Van Roijen, Reference Hakkaart-Van Roijen2002), governmental reports (Layard etal. Reference Layard, Bell, Clark, Knapp, Meacher, Priebe, Turnberg, Thornicroft and Wright2006; O'Shea & Kennelly Reference O'Shea and Kennelly2008; Gibbons etal. Reference Gibbons, Lee, Parkes and Meaney2012) various rigorous cost-benefit analyses conducted within mental health services (Rollman etal. Reference Rollman, Belnap, Mazumdar, Houck, Zhu, Gardner, Reynolds, Schulberg and Shear2005; Soeteman etal. Reference Soeteman, Verheul, Delimon, Meerman, van, Rossum, Ziegler, Thunnissen, Busschbach and Kim2010; Gerhards, Reference Gerhards, de Graaf, Jacobs, Severens, Huibers, Arntz, Riper, Widdershoven, Metsemakers and Evers2010) and a 10-point checklist that can be used to assess the quality of economic evaluations (Drummond etal. Reference Drummond, O'Brien, Stoddard and Torrance1997). Whatever scale is used to collect output and health care usage data, these can be supplemented with measures that show general impairment (and thus probable reductions in productivity) such as the 5-item Work and Social Adjustment Scale (Mundt etal. Reference Mundt, Marks, Shear and Greist2002).

Table 3 The EcoPsy-12

Cost-utility analysis

A cost-utility analysis aims to reveal the ‘utility’ (or cost-effectiveness) associated with health gains, most commonly through a metric called ‘Quality Adjusted Life Years’ (QALYs) (Chisholm etal. Reference Chisholm, Healey and Knapp1997). QALYs measure the benefits of healthcare interventions using a single index that combines life-years and health-related quality of life during those years (Al-Janabi etal. Reference Al-Janabi, Flynn and Coast2011). QALYs are obtained by multiplying a weight representing quality of life in a health state [ranging from 0 (death) to 1.0 (perfect health)] by the length time spent in that health state (Hoch & Smith, Reference Hoch and Smith2006).

After the number of QALYs a service or intervention produces for service users is calculated, the cost per QALY is generated. The cost per QALY metric takes into account all the costs of service provision and shows the costs required to produce each QALY. In this way, from an investment perspective, comparisons can be made between those interventions that are relatively inexpensive (low cost per QALY) and those that are relatively expensive (high cost per QALY) (Phillips, Reference Phillips2009). Those interventions that have a low cost per QALY are considered to be more efficient than those that have a high cost per QALY (Drummond etal. Reference Drummond, Brixner, Gold, Kind, McGuire and Nord2009). The general consensus internationally is that up to around €30 000 per QALY represents the upper limit for VFM investment in a service or intervention, although figures such as these are open to debate (Hoch & Smith, Reference Hoch and Smith2006; Phillips, Reference Phillips2009).

For most clinical trials, the effects of an intervention on long-term life expectancy are difficult to predict, especially for relapsing mental health problems such as low mood. Thus, in these cases it is considered appropriate to exclude the (long-term) life years component of the QALY calculation (Edwards etal. Reference Edwards, Hounsome, Russell, Russell, Williams and Linck2004). Moreover, various clinically useful methods for calculating QALYs can be performed such as the ‘standard gamble’ and ‘the time trade off’ that use scaling techniques to incorporate service user preferences into QALYs (Mann etal. Reference Mann, Gilbody and Richards2009). However, as such techniques are often complex, expensive and time-consuming, many governmental institutions in the United Kingdom (e.g. NICE), the United States and Canada use brief, standardised instruments such as the (6-item) EuroQol-6 Dimensions (EuroQoL Group, 1990) and the (6-item) Short-Form 6-Dimensions (Brazier & Roberts, Reference Brazier and Roberts2004) for the calculation of the QALYs (Mann etal. Reference Mann, Gilbody and Richards2009; Petrou & Gray, Reference Petrou and Gray2011).

Although QALYs allow comparisons of a diverse range of treatments in a ‘common currency’ that shows treatment effectiveness and cost utility in a single index (Al-Janabi etal. Reference Al-Janabi, Flynn and Coast2011), unfortunately there exists no universally accepted way of measuring its quality of life weight and different methods of measuring QALYs tend to produce substantially different results (Hoch & Smith, Reference Hoch and Smith2006). If looking to calculate QALYs, mental health services should take into account the lack of consensus in the area that still exists, despite ongoing efforts internationally (Drummond etal. Reference Drummond, Brixner, Gold, Kind, McGuire and Nord2009). To facilitate understanding of QALYs and the cost per QALY metric, we provide a hypothetical example below (see Table 4).

Table 4 Hypothetical example of how QALYS and cost-per QALYs can be used

QALY, Quality Adjusted Life Years; VFM, value-for-money.

Cost-minimisation analysis

In a cost-minimisation analysis, only the costs of providing a service or intervention are focused on. This simple form of analysis is only appropriate when the benefits (i.e. improvements in service user outcomes) of two or more regimens have previously been shown to be equivalent (Hoch & Smith, Reference Hoch and Smith2006).

An example of a cost-minimisation analysis can be found in a recent observational study of IAPT that included 39 227 service users (Hammond etal. Reference Hammond, Croudace, Radhakrishnan, Lafortune, Watson, McMillan-Shields and Jones2012). Here it was initially found that face-to-face, and telephone-delivered, low-intensity CBT were equally effective for mild-to-moderate anxiety and depression. A subsequent cost-minimisation analysis concluded that as telephone-delivered CBT cost 36.2% less per session to provide, it could thus be considered to be the more efficient and accessible intervention option. Another such example can be found in comparisons between CBT and medication interventions for panic and depression. Here various studies have found that both interventions offer comparative effectiveness but that CBT interventions cost approximately one-third less (Hunsley, Reference Hunsley2003).