Delivery of cognitive behavioural therapy to workers: a systematic review

Transcription

1 Occupational Medicine 2016;66: Advance Access publication 25 September 2015 doi: /occmed/kqv141 Delivery of cognitive behavioural therapy to workers: a systematic review V. V. Naidu 1, E. Giblin 2, K. M. Burke 3 and I. Madan 2 1 GKT School of Medical Education, King s College London, London WC2R 2LS, UK, 2 Occupational Health Department, Guy s and St Thomas NHS Foundation Trust, London SE1 7EH, UK, 3 Medical School, St George s University of London, London SW17 0RE, UK. Correspondence to: I. Madan, Occupational Health Department, Guy s and St Thomas NHS Foundation Trust, Education Centre, 75 York Road, London SE1 7EH, UK. ira.madan@kcl.ac.uk Background Cognitive behavioural therapy (CBT) is a key intervention, enabling workers on sick leave with common mental health problems to return to work. It can be delivered by a variety of methods including face-to-face therapy and the Internet. It is not known which is the optimal method of delivery. Aims Methods Results To establish the optimum method of delivering CBT to workers with common mental health problems. We undertook a systematic search of the OvidMEDLINE and EMBASE biomedical databases from the start of electronic records to 31 July 2013 for randomized trials comparing one method of delivering CBT with another for treatment of mild-to-moderate depression, anxiety and adjustment disorders. We included publications that assessed at least one of four outcomes: clinical or costeffectiveness, accessibility and acceptability. A scoping search found no studies in the workplace. We therefore focussed on interventions in the year age group. We found six studies comparing methods of delivery of CBT for anxiety disorders but found no trials which compared methods of delivery for mild-to-moderate depression. All delivery methods led to an improvement in anxiety symptoms. Internet-delivered CBT with some input from a therapist was found to be as clinically effective as face-to-face CBT and more cost-effective. Conclusions Internet CBT should be made available in workplaces for workers with anxiety disorders as part of a stepped care plan. Key words Introduction Anxiety; cognitive behavioural therapy; common mental health disorders; occupational health; workers. It is estimated that one in six workers in the UK experience mild-to-moderate anxiety and depression at any one time [1]. These so-called common mental health disorders (CMHDs) are the leading cause of sickness absence in most high-income countries and account for ~35% of disability benefits [2]. Randomized controlled trials have shown that cognitive behavioural therapy (CBT) is effect ive in treating CMHDs and the UK National Institute for Health and Care Excellence (NICE) recommends CBT for the treatment of depression and anxiety disorders as a stand-alone treatment or in combination with medication, self-help, exercise and other talking therapies as part of a stepped care model [3]. However, access to CBT services is restricted by the high level of demand, limited availability of therapists and, in the workplace, by the cost to the business of the time the worker needs to spend away from their job. Internet- and computer-based delivery could improve access to CBT, as could bibliotherapy. Although computerized CBT has been shown to be effective in improving emotional distress in workers with stress-related absenteeism [4], it is not clear whether it is as effective as therapist-delivered CBT in reducing symptoms associated with CMHDs. The aim of this review was to establish the optimum method of delivering CBT for workers with mild-tomoderate depression or anxiety including panic, social phobia and adjustment disorders. The Author Published by Oxford University Press on behalf of the Society of Occupational Medicine. All rights reserved. For Permissions, please journals.permissions@oup.com

2 V. V. NAIDU ET AL.: DELIVERY OF COGNITIVE BEHAVIOURAL THERAPY TO WORKERS: A SYSTEMATIC REVIEW 113 Methods We used the population intervention control outcome approach to compile a search strategy [5]. Our initial population was workers with mild-to-moderate depression, anxiety (including panic and social anxiety disorders) or adjustment disorder. Post-traumatic stress disorder and obsessive compulsive disorder are considered stress disorders in the ICD-10, thus we excluded these from our search. CBT interventions were face to face (individual or group), computer, telephone and bibliotherapy. As we sought to compare delivery methods to find the optimum, we only included randomized control trials, which compared one method of CBT with another. Outcomes were clinical improvement, cost-effectiveness, acceptability and accessibility to the worker. We undertook a scoping search using >50 search terms (both medical subject headings and text words) relating to the above subjects but found no studies on workers or studies set in workplaces. We therefore amended our search criteria to include studies in any setting where the intervention was applied to the broad working age population of year olds. We performed a computerized literature search, restricting publication type to randomized controlled trials. We searched databases from inception to 31 July 2013, using the following search terms: OvidMEDLINE: *Depression/ [focus] or anxiety or panic or adjustment disorder and Cognitive therapy [subjects] PsychInfo: *Postpartum depression/ or *Endogenous depression/ or *Reactive depression/ or *Spreading depression/ or *Depression (emotion) or *Recurrent depression/ or *Beck depression inventory [focus] or anxiety or panic or adjustment disorder and Cognitive therapy [subjects]. EMBASE: *Depression/ [focus] or anxiety or panic or adjustment disorder and Cognitive therapy [subjects]. We hand searched key publications for additional references and consulted with leading experts in the field to identify any relevant missing publications. We limited the review to studies in English. Three reviewers (E.G., V.N. and K.B.) independently selected publications based on title and then abstract. They retrieved full texts of abstracts that met the inclusion criteria. Two reviewers (E.G. and V.N.) independently selected the full publications, which met the inclusion criteria and compared results. Where there was disagreement, a fourth reviewer (I.M.) was consulted to decide whether or not the publication should be included in the review. The critical appraisal tool for randomized controlled trials by the Critical Appraisal Skills Programme (CASP) [6] was used for the quality assessment of publications. The reviewers excluded publications if they did not meet the requirements of the CASP trial validity screening questions, i.e. the trial must have had a clearly focused question and the assignment of patients to treatment groups must have been random. I.M. extracted data from the remaining publications and assessed the internal and external validity of the research. Since it is recognized that the use of scales with summary scores to distinguish high- and low-quality studies is questionable [7], we assessed internal validity of the studies by considering possible biases including selection and attrition bias. In addition, we assessed blinding of participants, therapists and researchers and treatment allocation. We also assessed if the reported randomization was truly random. Results We identified 1447 references by the electronic literature search after de-duplication (Figure 1). Title screening resulted in 544 studies being selected for further scrutiny. After screening the abstracts, we retrieved 317 full text articles. We added no additional papers following hand-searching or consultation with experts in the field. Six papers met our inclusion criteria. Of these, all focused on patients with anxiety disorder including social anxiety, phobias and panic disorder (Table 1). Hedman et al. published two papers on one group of subjects, one comparing the clinical effectiveness of Internet-delivered CBT (icbt) versus face-to-face group therapy and the other comparing cost-effectiveness [8,9]. Two studies were set in Australia and the remainder in Sweden. The age range of the populations in the studies incorporated the working age of 18 65, although if studies did extend the upper age limit slightly, they were still included as the vast majority of participants were in the working age range. The trials in our review compared icbt with either individual or group face-to-face therapy and in all 1447 records identified by electronic search 544 selected by title 317 full-text articles retrieved following screening of abstracts Six studies included in qualitative analysis No additional articles identified by hand search Figure 1. Flow of information through the study selection process.

3 114 OCCUPATIONAL MEDICINE Table 1. Characteristics of the included trials Author and year of publication Study characteristics Participant characteristics Intervention Results Internal validity Andrews et al. (2011) [10] Hedman et al. (2011) [8,9] Bergstrom et al. (2010) [11] Aim: to compare the effectiveness of icbt with faceto-face group CBT in patients with social phobia. Recruitment: via psychiatry clinics. Setting: Australia Aim: to evaluate the clinical and cost-effectiveness of icbt with cognitive behavioural group therapy for the treatment of social anxiety disorder. Recruitment: via primary care physicians and psychiatrists. Setting: Sweden Aim: to compare the effectiveness of icbt and groupadministered CBT for panic disorder and to establish the costeffectiveness of these interventions. Recruitment: via psychiatrists, primary care physicians and self-referral. Screened by psychiatry nurse for presence of symptoms. Setting: Sweden N = 25 Mean age (SD): 31.9 (7.8) 41% N = 126 Mean age: icbt: 35.2 Face-to-face group: 35.6 icbt: 38% Face-to-face group: 34% icbt: shyness programme, including six online lessons, s, forums, text messaging and homework. Total treatment time: 8 weeks. Face-to-face group: weekly meetings under the guidance of the same clinician used in icbt group. Group therapy time = 4 h weekly for 7 weeks. Content followed a standardized programme. The average total clinician time was 240 min, 13 more than the icbt patients. icbt: 15 weekly text modules with a homework component. The patient had access to a therapist via an online secure messaging system who had to restrict time spent on each patient to <10 min per week. Face-to-face group: one initial individual session followed by 14 group sessions over 15 weeks. Groups were led by two therapists and had six to seven participants. Social anxiety levels were reduced in both groups. No significant difference between the two groups (P > 0.05) in outcome measures of social anxiety and associated disability measured at 8 weeks post-intervention. Social anxiety levels were markedly reduced in both groups. At post-treatment and 6 months follow-up, respectively, the 95% CI of the mean difference on the LSAS was and 2.5 to 15.69, favouring icbt. This was well within the noninferiority margin of 10 LSAS points for the lower bound. The gross total costs were significantly reduced at 6 months follow-up, compared with pre-treatment in both groups. As both treatments were equivalent in reducing social anxiety and gross total costs, icbt was more cost-effective due to lower intervention (therapist time) costs. N =104 Both treatments were 10 A majority of patients Mean age (SD): icbt: 33.8 (9.7) weeks long. icbt: 10 self-help responded to treatment (defined as 40% Face-to-face group: modules based on decrease in panic 34.6 (9.2) established CBT disorder severity scale). Treatment effects principles. Homework were maintained at icbt: 64% assignments with 6 months follow-up. Face-to-face group: 59% psychologist feedback. Non-mandatory online No significant discussion forums. differences were found between the two treatment groups in terms of reduction of symptoms. Small sample size. Follow-up time of 8 weeks, may not have given sufficient time for the intervention to be effective. Unclear whether the researchers were blinded to treatment allocation. Well-designed randomized controlled noninferiority trial. By design, there was no randomization to an active placebo condition, so the reduction in levels of social anxiety in the two groups may have been due to regression to the mean. By design, there was no randomization to an active placebo condition, so the reduction in levels of panic disorder in the two groups may have been due to regression to the mean.

4 V. V. NAIDU ET AL.: DELIVERY OF COGNITIVE BEHAVIOURAL THERAPY TO WORKERS: A SYSTEMATIC REVIEW 115 Table 1. Continued Author and year of publication Study characteristics Participant characteristics Intervention Results Internal validity Carlbring et al. (2005) [12] Kiropoulos et al. (2008) [13] Aim: to compare the effectiveness of individual CBT with self-help icbt for panic disorder. Recruitment: general population via advertisement in the media. Setting: Sweden Aim: to compare an icbt programme for panic disorder with individual face-to-face CBT. Recruitment: general population via advertisement in the media and mental healthy websites. Setting: Australia CI, confidence interval; LSAS, Liebowitz Social Anxiety Score. bar one [10], the icbt included some contact with a therapist. CBT led to a reduction in anxiety symptoms in the majority of the participants in all studies included in our review. There was no significant difference between the treatment groups in the trials except one [8], where icbt was found to be superior to group face-to-face therapy. Improvement was maintained for the periods measured in the trials (between the end of treatment and up to 6 months post-treatment). Only two trials compared the cost-effectiveness of delivery methods [9,11] and both found icbt cheaper due to less therapist time. Acceptability was directly measured in two trials: one compared icbt with face-to-face individual therapy [12] and the other compared icbt with face-toface group therapy [13]. Participants were satisfied with Face-to-face group: psychologist-led self-help programme during weekly 2 h sessions with the support of printed hand-outs. N = 49 icbt: 10 modules, 25 Mean age (SD): pages long via the icbt: 34.2 (6.0) web. Homework Face-to-face individual: included biblio 35.8 (9.3) therapy, essays and assignments with feedback from icbt: 68% therapists. Mean Face-to-face individual: total time spent 75% per participant was ~150 min. Face-to-face individual: participants received 10 weekly individual sessions, lasting min, between which they were expected to do homework. N = 86 Mean age: icbt: 72% Face-to-face individual: 73% icbt: panic online programme and psychologist interaction via . Participants were asked to read and practice one module per week. Face-to-face individual: a 12 week manualized CBT programme during which participants attended 1 h weekly sessions with a psychologist. Participants were designated weekly reading. their treatment but in one trial, participants reported that they found the pace of work too high especially in the icbt group, where only 28% of participants finished all modules in the intended 10 week time frame [12]. Kiropoulos found that participants in the face-to-face group reported a higher level of enjoyment in their communication with the therapist compared with the icbt subjects [13]. Accessibility of treatment was not reported in any of the studies. Discussion icbt was more costeffective than group treatment both at post-treatment and follow-up. Clinical effectiveness: the majority of participants reported a reduction in their symptoms posttreatment, but there was no significant difference in anxiety/ panic disorder measures in the two groups. P > Acceptability: most were satisfied with treatment, but felt the pace was too high, especially the icbt group. Clinical effectiveness: both treatment arms reported improvement in their symptoms, but there was no significant difference between treatment arms for panic attacks or depression/anxiety. P > 0.05 Acceptability: participants rated both treatments as satisfying but participants in the face-to-face group reported higher enjoyment with communicating with their therapist. Attrition rates in the two arms did not differ. Conclusions from cost-effectiveness analysis limited as calculations were solely based on therapist time. Study conducted by a research group closely affiliated with the Internet programme used. Therapists only had modest experience in working with patients with panic disorder. Small sample size. Selection bias due to self-selection of participants who would be willing to use an Internet-based treatment and have ready access to the Internet. All modes of delivery of CBT included in this review led to an improvement in symptoms of anxiety. Internetbased CBT was more cost-effective due to a reduction in therapist time compared with face-to-face therapy, either

5 116 OCCUPATIONAL MEDICINE in a group or individual setting, even when the icbt included some interaction with therapists by telephone or . We found no randomized controlled trials comparing face-to-face CBT with any other method of delivery apart from icbt. We found no trials comparing methods of delivery of CBT for mild-to-moderate depression in the working age group. This review is limited by the absence of studies set in workplaces. We cannot be sure that our findings can be extrapolated to workers with anxiety disorders, particularly if they are related to work where complex factors may lead to perpetuation of their symptoms. Furthermore, the employment status of the participants was not stated in any of the studies we reviewed. Therefore, although all the participants were broadly of working age, some may have been unemployed. Unfortunately, none of the studies we reviewed focussed on accessibility, which may be an important factor in anxiety disorders where individuals may find it difficult to attend face-to-face therapy sessions. Our finding that icbt is effective in reducing symptoms of anxiety disorders is supported by a meta-analysis which found that icbt was superior to the control group (who were usually on a waiting list and therefore not eligible for inclusion in our review). The number needed to treat was 2.15, but the analysis also included studies where CBT was used as a treatment for major depressive disorders [14]. This analysis also found that adherence to and satisfaction with icbt was good. Employers are usually interested in the effectiveness of a clinical intervention in reducing absenteeism from work or improving productivity in the workplace. Although none of the trials in our review were set in workplaces, a re-analysis of data from five Australian trials, which were included in the aforementioned meta-analysis, showed that icbt for generalized anxiety disorder (two studies), depression (two studies) and social phobia (one study) resulted in significant reductions in self-reported absenteeism compared with control groups (who were on a waiting list) [15]. All participants in the trials in our review reported that they were satisfied with their treatment. This concurs with the findings of a randomized controlled trial set in workplaces for employees with mild-to-moderate depression, which compared the clinical effectiveness of a freely available icbt package (MoodGYM) with directing workers to mental health information websites [16]. This trial found that workers were broadly positive about using icbt; moreover, some of the workers preferred the faceless approach to therapy. We recognize that anxiety may present as part of a mixed depressive/anxiety disorder, and since we found no trials comparing CBT delivery methods in working age people with mild-to-moderate depression, we can only draw conclusions on the optimal mode of delivery for working age people where anxiety is the prominent disabling symptom. The principal advantage of using icbt as part of stepped care of anxiety disorders in workers is that for many of these individuals, attending face-to-face sessions, especially group therapy, may increase symptoms of anxiety and therefore reluctance to attend. Moreover, icbt may be more acceptable to workers who may regard referral to face-to-face therapy as stigmatizing, due to their perception of the use of mental health services. icbt could be provided by the employer for use at home or at work in order to allow employees the flexibility to work through the modules at their own pace and, in some cases, without the need for the worker to be put on a waiting list for face-to-face therapy. However, it may not be suitable for all. In most industries, a digital divide of workers still exists so that some employees do not have the necessary confidence, skills or access to information technology to participate in an Internet-based treatment. Provided that caveat is heeded, our findings that icbt with some therapist input is as clinically effective and more cost-effective than faceto-face therapy for anxiety suggest that it should be made available to workers as part of a stepped care approach for those with anxiety disorders. Key points Internet-delivered cognitive behavioural therapy with some therapist input is as clinically and cost-effective as face-to-face cognitive behavioural therapy for working age people with anxiety disorders. We found no studies comparing delivery methods of cognitive behavioural therapy for people with mild-to-moderate depression. There is a need to establish which method of receiving cognitive behavioural therapy is most acceptable to workers with common mental health disorders. Conflicts of interest None declared. References 1. Lelliott P, Boardman J, Harvey S, Henderson M, Knapp M, Tulloch S. Mental Health and Work: A Report for the National Director for Work and Health. Working for Health. London: Royal College of Psychiatrists, Harvey SB, Henderson M, Lelliott P, Hotopf M. Mental health and employment: much work still to be done. Br J Psychiatry 2009;194: National Institute for Health and Care Excellence (NICE). Common Mental Health Disorders: Identification and Pathways to Care (Clinical Guideline CG123) nice. org.uk/cg123 (8 February 2015, date last accessed). 4. Grime PR. Computerized cognitive behavioural therapy at work: a randomized controlled trial in employees with

6 V. V. NAIDU ET AL.: DELIVERY OF COGNITIVE BEHAVIOURAL THERAPY TO WORKERS: A SYSTEMATIC REVIEW 117 recent stress-related absenteeism. Occup Med (Lond) 2004;54: Sackett DL, Richardson WS, Rosenberg W, Haynes RB. Evidence-Based Medicine: How to Practice and Teach EBM. New York: Churchill Livingston, Critical Appraisal Skills Programme (CASP). Critical Appraisal Skills Programme (CASP) casp-uk.net/ (8 February 2015, date last accessed). 7. Centre for Reviews and Dissemination. Systematic Reviews: CRD s Guidance for Undertaking Systematic Reviews in Health Care. New York: University of York, Hedman E, Andersson G, Ljótsson B et al. Internet-based cognitive behavior therapy vs. cognitive behavioral group therapy for social anxiety disorder: a randomized controlled non-inferiority trial. PLoS ONE 2011;6:e Hedman E, Andersson E, Ljótsson B, Andersson G, Rück C, Lindefors N. Cost-effectiveness of Internet-based cognitive behavior therapy vs. cognitive behavioral group therapy for social anxiety disorder: results from a randomized controlled trial. Behav Res Ther 2011;49: Andrews G, Davies M, Titov N. Effectiveness randomized controlled trial of face to face versus Internet cognitive behaviour therapy for social phobia. Aust N Z J Psychiatry 2011;45: Bergström J, Andersson G, Ljótsson B et al. Internet-versus group-administered cognitive behaviour therapy for panic disorder in a psychiatric setting: a randomised trial. BMC Psychiatry 2010;10: Carlbring P, Nilsson-Ihrfelt E, Waara J et al. Treatment of panic disorder: live therapy vs. self-help via the Internet. Behav Res Ther 2005;43: Kiropoulos LA, Klein B, Austin DW et al. Is Internetbased CBT for panic disorder and agoraphobia as effective as face-to-face CBT? J Anxiety Disord 2008;22: Andrews G, Cuijpers P, Craske MG, McEvoy P, Titov N. Computer therapy for the anxiety and depressive disorders is effective, acceptable and practical health care: a metaanalysis. PLoS ONE 2010;5:e Mackenzie A, Harvey S, Mewton L, Andrews G. Occupational impact of Internet-delivered cognitive behaviour therapy for depression and anxiety: reanalysis of data from five Australian randomised controlled trials. Med J Aust 2014;201: Schneider J, Sarrami Foroushani P, Grime P, Thornicroft G. Acceptability of online self-help to people with depression: users views of MoodGYM versus informational websites. J Med Internet Res 2014;16:e90.