The economic cost of depression and anxiety disorders

SCMH estimated that the economic cost of mental health problems to society in England in 2002/2003 amounted to approximately £77 billion (SCMH, 2003). Output losses, such as non-employment, unpaid work, sickness absence and premature mortality, equated to over £23 billion of these costs. Furthermore, when these costs were adjusted to include only depression and chronic anxiety they totalled £12 billion a year – 1% of the total national income (CEP, 2006). The direct cost to UK employers of mental health problems (the majority of which constituted depression and anxiety disorders) has been estimated at approximately £26 billion (SCMH, 2007). This amount includes not only absenteeism but also ‘presenteeism’ which accounts for staff turnover and reduced productivity at work. In addition, the UK psychiatric morbidity survey (Office for National Statistics, 2000) showed that 42% of working-age adults suffering from depression and 40% suffering from an anxiety disorder were on incapacity benefit or income support compared to only 8% with no mental health disorder.

Data from other countries confirm the significant negative impact of depression and anxiety on the economy. For example, a North American study put the work-loss costs for those in employment with mood disorders (the vast majority of which were depressive disorders) at $11.5 billion (Druss et al. Reference Druss, Marcus, Olfson and Tanielian2001). Another USA study found that people with a pure affective disorder (i.e. excluding comorbid problems) had 10 times more work-loss days and work cutback days compared to those with no disorder (Kessler & Frank, Reference Kessler and Frank1997). An Australian mental health survey (Andrews et al. Reference Andrews, Henderson and Hall2001) found that people suffering from affective or anxiety disorder were between 2.1 and 2.8 times more likely to be unemployed compared to the general population.

There is a plethora of research documenting the psychological sequelae of unemployment. For example, Warr et al. (Reference Warr, Jackson and Banks1988) found that unemployed people experience higher levels of depression and anxiety together with lower self-esteem and wellbeing. Graetz (Reference Graetz1993) found that employed people report significantly higher levels of psychological health and wellbeing than the unemployed and students although, as Graetz recognizes, the benefits of employment are confined to those who have or find a satisfying job. This is an important point in the RtW agenda. Although the relationship between work and health can be complex, a recent review of over 400 pieces of scientific evidence concluded that work is beneficial to health and wellbeing and ‘outweighs the risks associated with it and by the negative consequences of unemployment and sickness absence’ (Waddell & Burton, Reference Waddell and Burton2006, p. 24). Waddell & Burton's recommendations included that ‘those with common health problems, including anxiety and depression, should be encouraged and supported to remain in, enter or re-enter work as soon as possible because it is therapeutic, helps to promote recovery, leads to better health outcomes, minimizes the harmful effects of long-term absence and incapacity, and promotes inclusion’ (Waddell & Burton, Reference Waddell and Burton2006, p. viii).

When these statistics and findings are considered in conjunction with the substantial evidence base for CBT for depression and anxiety (as evidenced by NICE) it is perhaps unsurprising that the Layard report (Reference Layard2005) highlighted the potential economic benefits that may follow from increasing the provision of CBT for people with depression and anxiety disorders in the UK. This was reflected in the IAPT health and wellbeing framework which explicitly incorporated ‘inclusion (including employment)’ as one of its four main outcome measures (DoH, 2007, p. 15); helping long-term incapacity benefit claimants back to work and helping people who have mental health problems remain at work. It is important then to review the evidence that CBT can assist individuals in either remaining or returning to work.

CBT and employment

Typically, many healthcare professionals do not consider employment to be a key objective for people with mental health problems (SEU, 2004). Indeed, as Black comments in a recent review, historically there has been poor coordination between the UK National Health Service (NHS) and employment and skills programmes leading to care that is symptom-focused with little regard to RtW or workplace adjustments (Black, Reference Black2008). Unsurprisingly, therefore, there is limited UK (and international) research to date examining the occupational outcomes of mental health treatment. A review of the NICE guidelines for depression and anxiety disorders did not reveal a single RtW recommendation (Hashtroudi & Paterson, Reference Hashtroudi and Paterson2009); while NICE (2009) reported a lack of evidence for the effectiveness of healthcare interventions that help people RtW.

There is some evidence that CBT can help non-clinical groups who are on long-term unemployment RtW. For example, Proudfoot et al. (Reference Proudfoot, Guest, Carson, Dunn and Gray1997) found that a CBT group-training programme led to statistically significant improvements in mental health and employment rates compared to a non-CBT-based control programme. The CBT group involved many aspects seen in standard CBT including goal-setting, challenging of automatic thoughts, relapse prevention as well as specific application of techniques for work situations. However, any mental health gains during training had been lost at 3 months for those who had not found employment which suggests the need for ‘booster’ CBT sessions. Della-Posta & Drummond (Reference Della-Posta and Drummond2006) compared the effects of offering 4 hours of CBT and job search assistance training to standard training alone for worker's seeking employment following a work-related injury. The CBT group showed statistically significant improvements in mental health scores and secured employment more rapidly than participants in the standard job search group. Creed et al. (Reference Creed, Machin and Hicks1999) reported improvements in mental health and coping behaviours in a group of unemployed people following a CBT-based programme but there was no difference in employment rates between the CBT and control group, although the numbers at follow-up were small. Similarly, Harris et al. (Reference Harris, Lum, Rose, Morrow, Comino and Harris2002) found no differences in job-seeking behaviours following a group CBT training programme compared with a non-CBT-based skills programme. However, as the authors acknowledge, their population had very substantial barriers to returning to work (e.g. limited education or training, very long periods of unemployment) while their CBT programme also de-emphasized behavioural components such as activity scheduling and experiments which may be critical ingredients in successful job-seeking.

CBT may also have a role in helping people suffering from comorbid physical and mental health problems RtW. Programmes based on cognitive behavioural techniques have been shown to be effective in treating such conditions as chronic low back pain (Guzman et al. Reference Guzman, Esmail, Karjalainen, Malmivaara, Irvin and Bombardier2001) and arthritis (Keefe et al. Reference Keefe, Smith, Buffington, Gibson, Studts and Caldwell2002) by addressing unhelpful beliefs around pain and recovery. However, even though pain and depression have been shown to be the most significant variables for predicting RtW (Corbiere et al. Reference Corbiere, Sullivan, Stanish and Adams2007) there has been limited research into vocational outcomes for patients with chronic pain following CBT programmes. A randomized control trial by Li et al. (Reference Li, Li-Tsang, Lam, Hui and Chan2006) found that a ‘work readiness’ programme, which incorporated CBT techniques, improved injured workers’ motivation and employment readiness and reduced their anxiety levels; however, Li et al. did not measure long-term occupational outcomes. A small study by White et al. (Reference White, Beecham and Kirkwood2008) found that group CBT for chronic pain led to improvements in psychological and physical health and occupational functioning. Other studies have also shown that CBT interventions for pain-related work disability can improve RtW rates (e.g. Kendall & Thompson, Reference Kendall and Thompson1998; Ektor-Anderson et al. Reference Ektor-Anderson, Ingvarsson, Kullendorff and Orbaek2008; Sullivan et al. Reference Sullivan, Adams, Thibault, Corbiere and Stanish2006).

There has been some interesting recent research on workplace CBT interventions. For example, Seymour & Grove (Reference Seymour and Grove2005) report that brief individual therapeutic interventions, including CBT, are effective for employees with job-related distress. However, the positive effects of CBT were mainly found in occupations where there was a high degree of control over the work environment which seems to suggest that CBT might not be beneficial if the work environment is both low in control and high in demand (Grove, Reference Grove2006). Mino et al. (Reference Mino, Babazono, Tsuda and Yasuda2006) found that a stress-management programme based on a CBT approach reduced symptoms of depression in workers at a highly stressful workplace but unfortunately the study did not investigate the impact of absenteeism or ‘presenteeism’. Wang et al. (Reference Wang, Simon and Avorn2007) compared telephone outreach and a care management programme, which included CBT, to treatment as usual (TAU) for workers with depression. The enhanced care included a comprehensive assessment of needs, facilitated entry to treatment [both CBT (either in-person or telephone), and antidepressant medication] and supported treatment compliance. At 6- and 12-month follow-up the intervention group showed improved clinical and workplace outcomes including higher job retention rates and an annual effect of 2 weeks more work than the usual care workers. However, there was little detail as to how much CBT was utilized in the treatment group and what this involved.

In contrast to these two studies, de Vente et al. (Reference De Vente, Kamphius, Emmelkamp and Blonk2008) found no differences in symptoms or absenteeism between CBT-based stress management training (SMT), whether group or individually delivered, for work-related stress complaints compared to TAU. However, as the authors acknowledge, individuals in the SMT group were not urged to resume work as soon as possible, instead only encouraged to at least partly resume work if their symptoms reduced to acceptable levels based on their own judgement. De Vente et al. (Reference De Vente, Kamphius, Emmelkamp and Blonk2008) suggest that a more explicit RtW agenda may have led to enhanced outcomes for the intervention group. An interesting study by Blonk et al. (Reference Blonk, Brenninkmeijer, Lagerveld and Houtman2006) examined interventions for self-employed people on sick leave owing to work-related psychological complaints. They compared CBT delivered by psychotherapists with a brief CBT informed intervention which included workplace and individual interventions delivered by labour experts. They found the combined individual and workplace intervention achieved significantly higher levels of both partial and full RtW over the standard CBT group. The results lead the authors to recommend that RtW should be addressed earlier on in CBT. Finally, findings have recently been released on the occupational outcomes achieved by the two IAPT demonstration sites (Clark et al. Reference Clark, Layard and Smithies2008). When combined the sites showed that CBT correlated to a net increase in employment of 5% which is in line with the forecasts made by Layard et al. (Reference Layard, Clark, Knapp and Mayraz2007). However, the re-employment rates varied considerably across the two sites with one site achieving a 10% increase in employment and the other a 4% increase. It should be noted that although disturbance in occupational functioning was captured using psychometric tools there was no mention of whether or how RtW was addressed in therapy.

In summary, even though research is somewhat limited and the relationship between mental health and RtW is a complex phenomenon, there is evidence to suggest that CBT can be effective at reducing absenteeism and in helping people RtW. CBT is a generic term for a multitude of approaches, therefore it is difficult to identify which particular aspects may lead to improved RtW rates. However, as Blonk et al. (Reference Blonk, Brenninkmeijer, Lagerveld and Houtman2006) suggest, incorporating workplace interventions and making RtW more explicit in treatment may improve outcomes. Therefore, it is important to consider how CBT can maximize both positive clinical and occupational outcomes, while maintaining the core components of CBT such as the collaborative relationship. Key questions which we will now explore are: What are the likely challenges of incorporating RtW explicitly into treatment? Will it cause problems in the therapeutic relationship? Can RtW be formulated and treated as a form of avoidance to be identified, challenged and overcome?

Collaboration and the therapeutic relationship

‘Collaborative empiricism’ is seen as a cornerstone of effective cognitive therapy (Beck et al. Reference Beck, Rush, Shaw and Emery1979). Although faced with common symptoms and diagnoses, the cognitive therapist should view each client's problems as idiosyncratic and should actively work ‘shoulder-to-shoulder’ with the client to develop a shared understanding of the client's problems which facilitates learning in therapy (Durham et al. Reference Durham, Swan and Fisher2000). There has been an increasing interest in the role of the therapeutic alliance in CBT over the last two decades and more recently its relationship to therapy outcomes. For example, Trepka et al. (Reference Trepka, Rees, Shapiro, Hardy and Barkham2004) found a clear positive association between alliance and outcome in cognitive therapy for depression and this association was stronger than therapist competence. A collaborative approach is regarded as an important factor in developing a strong therapeutic alliance (Ledley et al. Reference Ledley, Marx and Heimberg2005). Martin et al.'s (Reference Martin, Garske and Davis2000) review of the theoretical definitions of the therapeutic relationship found three main themes of the alliance; the collaborative nature of the relationship, the affective bond between the patient and therapist, and the ability to agree on treatment goals and tasks.

It seems natural that the cognitive therapist might shy away from bringing the RtW issue into treatment as it clearly introduces a therapist-/service-driven goal and therefore seems to go against the collaborative nature of CBT which could be a clear bias towards the therapist taking the lead. It could be perceived as taking an authoritative or autocratic stance which is against the principles of a good cognitive behaviour therapist (Curwin et al. Reference Curwin, Palmer and Ruddell2000). It could also be a potential barrier in the development of a good therapeutic alliance, which in turn could lead to poorer therapy outcomes, or it could make the therapist appear overly controlling leading to clients dropping-out of treatment. However, from our own experience this has not necessarily been the case. Avoidance of the RtW issue by the therapist may have more to do with their assumptions and beliefs that it will automatically damage the therapy. Certainly, keeping the RtW issue as a hidden agenda item would go against the explicit collaborative nature of CBT (Sanders & Wills, Reference Sanders and Wills2005), especially where an objective of the service is to help clients return to, or stay in work.

One of the potential problems with bringing the RtW agenda into therapy is that it might be seen by the client as a preconceived idea of the therapist or that the therapist is prioritizing the goals of the service over the clients, which could then create a significant barrier to the therapy becoming a collaborative venture. Therefore, a RtW agenda that is delivered as a therapist-driven goal may not only effect collaboration but the whole therapeutic relationship. Further consideration is needed on how to balance the goals of the therapist and service, and respect and respond to the preferences of individual patients. Imposing or subtly coercing a RtW agenda on a client who does not wish to establish RtW as a treatment goal raises significant ethical issues. However, in our experience, at an in-house occupational mental health service, we have mostly found that a RtW agenda facilitates both therapy and the therapeutic relationship and that once prompted the client is often keen to have it incorporated within their therapy goals. Potential ways of balancing these issues can be illustrated through two short examples from our own clinical practice.

Clinical example 1

Mike was a 45-year-old senior communications manager for a large organization. He had suffered from chronic fatigue syndrome for several years and had been signed-off work for the last 6 months. He had been referred by a consultant psychiatrist after his condition had not improved through self-management strategies such as ‘pacing’ and medication. Initially the therapeutic relationship was somewhat tense; Mike felt that being referred for CBT meant that ‘all my symptoms are in my head’. Early on in treatment a problem list was developed which in itself can be viewed as the beginnings of the formulation (Persons, Reference Persons1989). Mike placed being signed-off sick near the top of his problem list as for him it highlighted his lack of progress in his condition and left him without an important aspect of his social identity. Setting goals can help to increase hope and a sense of control for the client (Charlesworth & Greenfield, Reference Charlesworth and Greenfield2004), and this was particularly important in engaging Mike to the therapy.

For Mike returning to work was a key feature of his medium- and long-term goals, and demonstrated that RtW was made explicit early on in therapy, which accords with the recommendations of Blonk et al. (Reference Blonk, Brenninkmeijer, Lagerveld and Houtman2006). Setting goals and generating a problem list is a collaborative process (Josefowitz & Myran, Reference Josefowitz and Myran2005) and helped Mike overcome some of his initial scepticism as well as instilling some hope in the therapy and the future. It also helped forge the beginnings of a therapeutic relationship. By session 8 he had been able to start on a graduated RtW programme.

Clinical example 2

Simon described a 2-year history of panic attacks that were initially brought on during large work meetings but his difficulties had now generalized to most places where large crowds were present and he had recently been signed-off work by his GP (see Fig. 1). With regard to his problem list, Simon was particularly concerned that the problems might have a long-term negative impact on his new career as an engineer. This together with some of his work-related therapy goals, proved strong motivating factors for Simon in his recovery. Simon used his motivation to good effect in designing challenging experiments between sessions. These included behavioural experiments around his attendance at work meetings. There was time spent during therapy to overcome specific cognitions he had around the potential embarrassment of fainting in front of work colleagues. Role-play was used in sessions to practice discussions with his manager around his fears.

Fig. 1. Maintenance cycle for clinical example 2.

Clinical example 3

Mary was a 25-year-old single healthcare worker who was suffering from depression. Mary reported a great sense of relief when she was originally signed-off work by her GP and during the initial goal-setting did not detail any specific work-related goals. Due to her clear relief at not being at work the therapist did not question whether RtW should be addressed at this stage for fear that it would cause a rupture in the therapeutic alliance. However, as the cross-sectional formulation was developed it became clear to both Mary and her therapist that Mary's absence from work was reinforcing the negative view of herself that she could not cope with life (see Fig. 2).

Fig. 2. Maintenance cycle for clinical example 3.

The formulation helped to illustrate that without her attendance at work her everyday life now lacked structure or goal-directed activity, limiting her ability to experience a sense of mastery from activities which lowered her mood further. The use of Socratic questions at this stage also helped highlight this issue (e.g. ‘So Mary, if it now appears to you that being isolated at home and off work is making the problem worse, how might you start to break that pattern?’). Following these observations strategies were developed to address some of her avoidance of activities to both increase her sense of mastery and pleasure along with testing out some of her negative predictions. Behavioural experiments were designed to test her belief that her colleagues would judge her negatively when she contacted them. The RtW plan was informed by the formulation and incorporated testing out key cognition and behaviours. In many ways it followed the principles and benefits of activity scheduling (Beck et al. Reference Beck, Rush, Shaw and Emery1979). Openly discussing some of her anxieties and objections to aspects of the RtW plan revealed some cognitive biases that could be either worked through in session or tested out over the following weeks. It also provided opportunities during therapy to learn how she could deal with and overcome potential workplace and other setbacks.

Clinical example 4

Stuart, a 28-year-old van driver, was suffering from anxiety and depressive symptoms for the last 6 months. This had been triggered when his wife left him for another man taking their son with her. Stuart was off work which meant he could avoid colleagues asking questions about how he was or his home situation which triggered thoughts such as ‘I've lost everything’ and strong feelings of emptiness. However, through the formulation process it became apparent to Stuart and the therapist that his absence from work was playing a part, among others, in maintaining his symptoms (see Fig. 3) and therefore a RtW plan might help him overcome this.

Fig. 3. Maintenance cycle for clinical example 4.

Formulating Stuart's sickness absence helped bring into his awareness the effect it was having on maintaining his anxiety and low mood. Although still anxious about going back to work, this awareness led Stuart to become much more motivated in working towards a return. The RtW plan involved making some contact with his colleagues and managers which were set up in the form of behavioural experiments to test out his negative predictions of what would happen. Stuart, his manager, GP and therapist were all involved in devising a graduated RtW plan in an attempt to make it as likely to succeed as possible. Stuart was prompted to discuss his anxieties about returning to work openly in the session. This helped highlight some of his thinking errors around this issue and allowed opportunities to work with these. There was discussion and problem-solving around how he could deal with the potential reality of negative reactions from colleagues. Carrying out the RtW plan during the therapy allowed Stuart to bring issues from his return to the sessions and for them to be worked through.

Future research

There is a pressing need for further research into the efficacy and effectiveness of CBT in helping people with depression and anxiety disorders RtW. IAPT is one service model which may achieve this but it is important not to become blinkered by this model but to continue to explore other ways in which mental health treatment might take responsibility for promoting vocational outcomes or for integrating itself with employment issues (SEU, 2004; Black, Reference Black2008). This may then highlight whether or not CBT can have a specific role as an ‘early intervention to either replace or complement sickness absence’ for those suffering from depression or anxiety disorders (Wilday & Dovey, Reference Wilday and Dovey2005).

Further investigation is also needed into how the therapist can best incorporate RtW into the collaborative relationship which is at the heart of effective CBT. This may then give cognitive therapists more confidence that a RtW agenda can be explicitly addressed ‘shoulder-to-shoulder’ with the client instead of being directed at them. Work forms a fundamental part of most people's lives and of their self-concepts. Those people not in employment due to depression or anxiety disorders not only suffer from their symptoms but also suffer from their exclusion from society. Cognitive therapists can have an important role in helping those excluded to become engaged again (Richards & Suckling, Reference Richards and Suckling2008).