Andersen, 2014 [44]
Context	Denmark; multidisciplinary, coordinated and early return-to-work (RTW) intervention under the management of the municipal sickness benefit offices; program aimed at employed persons on sick leave due to complex health-related problems irrespective of medical diagnoses; participation obligatory if the municipal sickness benefit officer estimates that it would enhance the chances of RTW
Participants	17 workers (13 women); average age 44 years (range 23–61); on sick leave for approximately 8 weeks due to common mental problems; at the last interview 6 were still sick-listed, 3 were no longer sick-listed but unemployed, 8 had returned to work
Methods	Semi-structured interviews at 3 time points: (1) before intervention; (2) approximately 3 months after first interview; (3) 6–7 months after first interview; methodology in line with principles of the Interpretative Phenomenological Analysis
Areberg, 2013 [45]
Context	Sweden; individual placement and support to help people with severe mental illness to find regular employment; data collected 12 months after the first meeting between participants and employment specialist; ongoing support
Participants *	17 participants (7 female), 20–59 years old, with a diagnosis of schizophrenia, psychosis, affective disorder or other; 2 employed and working at the time of the interview, 2 in a position of repose, 4 in a practice place, 2 in education, 7 looking for a job
Methods	Semi-structured interviews 12 months after first contact with employment specialist; stratified sampling; content analysis
* Participants from the same RCT as Hasson [52] who published a qualitative study based on interviews performed in the first year of implementation.
Boycott, 2015 [46]
Context	UK; individual placement and support for unemployed persons with severe mental illness recruited from community mental health teams and early intervention in psychosis teams; service offered for at least 6 months
Participants	31 participants (9 female) of the program who received IPS for at least 6 months; mean age 30.8 years; 14 psychosis, 8 schizophrenia, 4 bipolar disorder, 4 depression, 1 other diagnosis; 12 participants had gained employment at interview
Methods	Semi-structured interviews; interview after at least six months receiving the intervention; purposive sampling; thematic analysis
Buus, 2015 [47]
Context	Denmark; two counselling sessions by an occupational physician encouraging low-back pain (LBP) patients who expressed concerns about their ability to maintain their current job to engage in their own treatment by changing work routines and by exercising; a workplace visit was carried out if required
Participants	25 LBP patients (14 female); mean age 46.8 years
Methods	Semi-structured interviews with patients 1–4 years after concluding the counselling intervention; purpose sample; theoretical framework: “interactionism”; analysis: interpretative thematic approach
Cameron, 2012 [48]
Context	UK, South of England; vocational guidance model that aims to help workers with mental health problems to retain their employment; project workers listen, clarify, give information, refer on to other services and collaborate to draw up client action plans; referrals to the project are done by general practitioners, other health workers, employers, or by clients themselves
Participants	14 participants (10 female), average age 43.5 years (range 29–54)
Methods	Semi-structured individual interviews with project users; critical realist methodology
De Blasi, 2014 [49]
Context	France; support groups for vocational rehabilitation integrated in a multidisciplinary hospital department for patients who wish to RTW or who have already returned to work and have difficulties or questions regarding to work; led by an occupational therapist and a psychologist; 4 monthly sessions of 90 min
Participants	20 cancer patients (17 female); mean age at diagnosis: 47 years (range 37–55); cancer site: 16 breast, 2 hematopoietic, 1 digestive, 1 ear-nose-throat; 19 of them employed at diagnosis; 12 of them on sick leave during support group, 8 already working
Methods	Open questions in the questionnaire sent to participants one month after the last session; thematic analysis
Germundsson, 2011 [50]
Context	Sweden; collaboration projects from a Swedish project group (NTG-New Routes to the Labour Market-Quality Assurance of Collaboration) within the Equal program; unemployed persons with mental health problems are part of the target population; duration: not found
Participants	(A) 8 service users (4 female), aged 19–52 years, unemployed and in vocational rehabilitation, sick-listed for 18–60 months due to mental health problems, six of them with little or no experience of paid employment; (B) 20 professionals (results for Group B not reported here)
Methods	Semi-structured interviews with service users on two occasions approximately six months apart; purposeful sample; thematic analysis guided by a theoretical framework
Glavare, 2012 [51]
Context	Sweden; multi-professional pain-rehabilitation followed by a coached work-training (CWT) program; inclusion criteria for CWT were: unemployment, motivation for return-to work and at least 50% work capacity; the CWT program lasted 3–6 months
Participants	11 participants (8 female); 22–58 years-old; with long-term musculoskeletal pain, who had participated in the CWT program, 8 employed at interview
Methods	Semi-structured interviews; interviews performed 4–21 (mean = 11) months after the program; Grounded theory approach
Hasson, 2011 [52]
Context	Sweden (Malmö, Southern Sweden); individual placement and support to help people with severe mental illness to find regular employment; ongoing support at work
Participants	15 clients of the service; characteristics not reported
Methods	Key informant interviews, non-participant observations and document analysis. Interviews with (A) 15 clients and (B) other actors (e.g., employment specialists, project head, mental health care staff, social insurance officers; results for Group B not considered in this review; content analysis
Haugli, 2011 [53]
Context	Norway; multidisciplinary occupational rehabilitation program for patients on long-term sick leave, considered to have a fair chance of return-to work; it included different physical activities and individual and group based counselling aiming to increase function and work related processes; duration: not found
Participants	20 individuals with musculoskeletal and/or psychological complaints; 10 had returned to work (3 men 46–58 years old, 7 women 41–56 years old) and 10 individuals were registered with a disability pension (3 men 41–53 years old, 7 women 41–56 years old)
Methods	Semi-structured telephone interviews with individuals who attended the rehabilitation program 3 years ago; purposive sampling; systematic text condensation inspired by Giorgi’s phenomenological analysis
Higgins, 2014 [54]
Context	UK, Northern Ireland; the role of the general practitioner when managing illness that may deserve long-term sick leave is investigated; duration: not applicable
Participants	(A) 9 employees who had experienced a long-term sickness absence and (B) other stakeholders (e.g., government-level policy makers, trust-level human resources managers, occupational health clinicians; results for Group B not presented here)
Methods	Comparative case study within three Health and Social Care Trusts; 61 semi-structured interviews and 3 focus group; purposive sampling for the Trusts, otherwise convenience sampling; theoretically framed by Talcott Parsons’s model of the medical contribution to the sick role; realistic approach
Hubertsson, 2011 [55]
Context	Sweden; two actors in the standard management of sickness absence are investigated: the Social Insurance Agency and Health Care; duration: not applicable
Participants	15 participants (11 female); aged 33–63 years; with experience of sickness absence due to musculoskeletal disorders, 5 partly working, 2 partly unemployed, 5 on sick leave, 3 on disability pension
Methods	Semi-structured interviews; purposive sampling; latent content analysis
Jansson, 2014 [56]
Context	Sweden, two counties; two strategies that aim to improve the work ability of persons with common mental disorders are investigated: (1) a pedagogical model of learning called problem-based method, lasting 12 weeks and (2) cognitive behavioral therapy, number of sessions not standardized
Participants	16 participants (14 female); aged 30 to 63 years; 8 people from each intervention and each county; 13 working, 3 in sick leave
Methods	Semi-structured interviews; purposive sampling; content analysis
Kalef, 2014 [57]
Context	Norway; job training program at Norway’s largest telecommunications company with financial support from Norway’s Labour and Welfare Organization (NAV), targeted at people with disabilities with a high school diploma and motivated to gain employment; three main components: computer training, personal development, and work-practice; duration: not found
Participants	(A) 15 completers of program, 14 current participants, and (B) other stakeholders (e.g., company leaders, program administrators, teachers, community members; results for Group B not presented here)
Methods	Personal observations, 29 semi-structured interviews and document review; purposive sampling; thematic analysis
Lewis, 2013 [58]
Context	UK; personalized tailored supported employment intervention to help people with disabilities to find employment; funded by the Department for Work and Pensions and managed by its agency, Jobcentre Plus; variable duration, depending on length of absence and follow-along supports
Participants	98 participants (40 female) from a sample of 11 providers; median age group 40–49 years, the primary disability of 48% of participants was a learning disability
Methods	Semi-structured interviews; purposive sampling; thematic analysis
Martin, 2012 [60]
Context	Denmark; multidisciplinary, coordinated and tailored RTW-intervention for employees with relatively mild mental health problems *; sick-listed for 4–12 weeks; collaboration between a Danish municipality and a private company specialized in RTW; duration of intervention: 12 weeks
Participants	(A) 10 persons who completed the intervention; (B) other stakeholders (i.e., municipal social insurance officers and the multidisciplinary team); results for Group B not considered here
Methods	Interviews with participants were audio-taped, transcribed verbatim and coded thematically with the software NVivo
* Same intervention as in Martin (2015) [59] but implemented in only one municipality.
Martin, 2015 [59]
Context	Denmark; multidisciplinary, coordinated and tailored return-to-work-intervention for employees with common mental health problems; sick-listed for 4–12 weeks; collaboration between three Danish municipalities and a private company specialized in RTW; duration of intervention: 12 weeks
Participants	(A) 10 persons who completed the intervention, from three different municipalities; and (B) other stakeholders (i.e., municipal social insurance officers and the multidisciplinary team); results for Group B not considered here
Methods	Interviews with participants were audio-taped, transcribed verbatim and coded thematically with the software NVivo.
* Same intervention as in Martin (2012) [60] but implemented in 3 further municipalities.
Mikkelsgard, 2014 [61]
Context	Norway; supported employment project mainly targeted at people under the age of 25 years who receive financial assistance from the NAV (Norwegian Labour and Welfare Administration) and have mental health problems; collaboration between NAV and mental health services; voluntary participation based on individual motivation; duration: not found
Participants	Seven out of 52 participants of the project (5 women, 2 men); 20–35 years of age
Methods	Semi-structured interviews, audio-recorded and transcribed verbatim; content analysis
Nieminen, 2012 [62]
Context	Eight European countries; lifelong learning strategy to facilitate social inclusion of unemployed persons with severe mental illness. Participants were offered training related to personal development and planning, employment and recovery, as well as opportunities for unpaid and paid activities *; duration: not found
Participants	23 service users (9 female, 11 male, 3 not known); mean age 42.8 years (range 24–57); with schizophrenia, schizoaffective or bipolar disorder
Methods	Interviews (2–5 key informants from each site), chosen at random and interviewed at baseline, and 10 and 20 months; thematic approach
* Same intervention as Ramon (2011) [63]; both publications seem to be based on the same interviews; only additional information to Ramon 2011 has been extracted.
Ramon, 2011 [63]
Context	(see Nieminen, 2012) *
Participants	Service users with schizophrenia, schizoaffective or bi-polar disorder
Methods	Semi-structured interviews (2–5 in each site) at baseline and 21 interviews 10 months later; 138 written self-reports at baseline and 99 ten months later
* Same intervention as Nieminen (2012) [62].
Reagon, 2011 [64]
Context	UK, Wales; condition management programs (CMPs) led by the National Health Service to help individuals to manage health conditions better in relation to a return-to-work; characterized by a biopsychosocial approach and cognitive behavioral therapy techniques; clients are voluntarily referred to the CMP by the Jobcentre; about 13 weeks intervention after the evaluation, depending on individual needs
Participants	(A) 15 participants (9 female) who had recently completed the program; 20–50 years-old; with musculoskeletal (n = 7) and mental health problems (n = 8) as primary health condition; (B) 12 staff members; results for Group B not considered in this review
Methods	Semi-structured interviews; tape-recorded and transcribed verbatim; purposive sample; thematic analysis
Secker, 2012 [65]
Context	UK; Condition Management Programmes aimed at supporting persons on incapacity benefits due to long-term conditions to RTW; joint venture between the UK Department of Health and the Department for Work and Pensions; duration: not found
Participants	39 participants of the program (22 female) with mental health problems (n = 20), musculoskeletal (n = 8), cardiovascular (n = 4), and a combination of conditions (n = 7); most of them (n = 31) in their middle years
Methods	Twelve focus groups conducted at the site of the corresponding program; the discussions were tape-recorded and transcribed verbatim; purposive sample; text analysis
Tiedtke, 2012 [66]
Context	Belgium; legislation affecting RTW of women with breast cancer; duration: not applicable
Participants	(A) 5 women who survived breast cancer and (B) 4 representatives of patient associations and further stakeholders (e.g., treating physicians, employers, social security physicians); results for Group B not considered in this review
Methods	Three multidisciplinary focus groups; discussions tape-recorded and transcribed verbatim; purposive sample; thematic analysis
Vandekinderen, 2012 [67]
Context	Belgium; labor-market training program for women with mental health problems; five linear modules: registration, observation, guidance, training for employment, and work placement; duration: not found
Participants	(A) 11 women with mental health problems and substance dependence and (B) other stakeholders (i.e., the chief manager, support workers, and project partners); results for Group B not considered in this review
Methods	Document analysis of all available project documents and 17 interviews with women with mental health problems; the interviews were audio-taped, transcribed and returned to the participants for review; interpretative research approach