Table 1

Study characteristics of the seven included studies

Author, year, country	Population	Inclusion and exclusion criteria	Sample size, complete cases	Prognostic factors and confounders	Outcome, time point, case definition	Analysis	Results
Atlas et al, 2000, USA23	Working patients with diagnosed disc herniation, who received conservative or surgical (n=174/327; 53%) treatment, 68% male, mean age 40 years	Inclusion: diagnosed lumbar disc herniation Exclusion: previous spine surgery, cauda equina, spinal or other comorbidity, pregnancy	n=440, n=404 completed at least one follow-up, n=327 completed last follow-up	Receiving workers’ compensation, level of education, duration current episode, comorbidity, age, SF-36 general health, low back pain	Return to work; 4-year follow-up; case definition working on a job for pay	Stepwise multiple logistic regression	Receiving workers’ compensation OR 0.6 (0.3 to 1.2); age OR 0.7 (0.6 to 0.8); general health OR 1.1 (1.0 to 1.2); low back pain OR 0.8 (0.6 to 0.9), model adjusted for the other variables
Atlas et al, 2006, USA24	Working patients with sciatica, who received conservative or surgical treatment, 66% male, mean age 40 years	Inclusion: diagnosed lumbar disc herniation Exclusion: cauda equina, spinal or other comorbidity, pregnancy	n=394, who completed at least one follow-up between 5 and 10 years, n=352 competed last follow-up	Age, gender, initial treatment, physician expectation of surgery benefit, Quebec classification, category 4/6, low back frequency score, SF-36 physical function, SF-36 mental health	Return to work; 10 years; case definition employed at follow-up	Stepwise multiple logistic regression	Age OR 0.42 (0.3 to 0.58), male gender OR 0.33 (0.09 to 1.0), initial treatment, surgical OR 1.4 (0.46 to 4.6), Physician expected benefit OR 5.0 (1.65 to 17.7), physical function OR 1.4 (1.1 to 1.8), Quebec classification NS, Low Back Frequency Score NS, mental health NS, model adjusted for independent baseline prognostic factors
den Boer et al, 2006, Netherlands26	Working patients who underwent lumbar disc surgery, 59% male, mean age 41 years	Inclusion: >18 years, failure of conservative treatment, understand and read Dutch, having a paid job before the episode of complaints started. Exclusion: previous back surgery, physical comorbidity	n=200, n=182 complete cases	Fear of movement/ (re)injury, passive pain coping, physical work load, job satisfaction, duration sick leave Confounders: education, disability presurgery, neurological deficits presurgery, pain 3 days postsurgery	Work capacity, 6 months postsurgery, case definition: percentage work capacity (hours/week) compared with work capacity before the pain episode started=100%	Multiple logistic regression, only variables significant in univariable regression were entered into the model, and prespecified confounders	Fear of movement/ (re)injury OR 1.09 (SE 0.04 calculated 95% CI 1.01 to 1.18), passive pain coping OR 1.08 (SE 0.04, calculated 95% CI 1.0 to 1.17), physical work load OR 1.19 (SE 0.06, calculated 95% CI 1.06 to 1.34), job satisfaction OR 0.98 NS, duration sick leave OR 1.26 NS
Grøvle et al, 2013, Norway25	Working patients with sciatica and disc herniation, who underwent conservative or surgical (30%) treatment, 69% male, mean age 44 years	Inclusion:≥18 years, radiating pain below the knee and/or paresis, lumbar disc herniation Exclusion: pregnancy, spinal fracture, tumour, infection, previous surgery same disc, not able to read Norwegian	n=297, n=237 included complete cases (n=9 who were student, retired or homemaker at follow-up, were excluded)	Age, gender, marital status, current smoker, duration current sciatica episode > 3 months, had sciatica before, duration back problems>1 year, subjective health complaints, sciatica bothersomeness, disability, fear avoidance beliefs (work), fear of movement/reinjury, general health, emotional distress back pain, leg pain, positive SLR, motor weakness, reflexes depressed, sensory decrease	Return to work, 2 years; case definition: return to full-time work by self-report	Multiple logistic regression, only variables significant in univariable regression were entered into the model (p values < 0.2)	Age OR 0.97 (0.93 to 1.00), female OR 0.61 (0.31 to 1.22), bothersomeness OR 0.89 (0.82 to 0.97), fear avoidance beliefs OR 0.93 (0.90 to 0.97), general health OR 1.03 (1.01 to 1.05), positive SLR OR 0.44 (0.20 to 0.95)
O’Donnell et al, 2017, USA27	Workers’ compensation patients who received reoperation discectomy with or without fusion, 77.2% male, mean age 39.4 years	Inclusion: lumbar disc herniation after workplace injury, receiving lost-work compensation, injuries between 2005 and 2012, same level revision surgery exclusion: spondylolisthesis, spinal deformity, vertebral fractures, epidural haematomas and abscesses, spinal tumours, smoking history or using smoking deterrents	n=298, n=196 with fusion, retrospective cohort, therefore only complete cases	Revision surgery: fusion or no fusion, age, sex, marital status, labor-intensive occupation, permanent disability benefits, legal representation, psychiatric comorbidities, physical therapy and chiropractic care, opioid analgesic use, household income, permanent disability, time from primary surgery to reoperation surgery	Return to work, 3 years; case definition: ability to return within 2 years and work for at least 6 months within 3 years	Multiple logistic regression	Revision surgery: fusion OR 0.56 (0.33 to 0.97), psychiatric comorbidity before revision surgery OR 0.19 (0.05 to 0.68), opioids use within 1 month of revision surgery OR 0.44 (0.26 to 0.75)
Schade et al , 1999, Switzerland22	Patients who underwent lumbar disc surgery, 74% male, mean age 35 years (demographic data from Boos et al)38	Inclusion: a scheduled discectomy, age 20–50 years, continued employment at the time of surgery, no previous back surgery, failed conservative treatment, availability for additional clinical+MRI examination Exclusion: no Swiss residency, rapid progressive severe motor deficit or cauda equina syndrome	n=46, n=42 complete cases	Anxiety, depression, self-control, well-being, vitality, general health, occupational mental stress, job satisfaction, job-related resignation, social support confounders: pain and/or disability presurgery	Return to work, 2 years; case definition: return to ‘any’ work (time in months)	Univariable regression and stepwise multiple regression (medical data, general psychological factors and psychosocial aspects of work)	Depression beta ln 0.43 (estimated OR 1.54), occupational mental stress beta ln 0.28 (estimated OR 1.32), pain and/or disability presurgery beta ln 0.35 (estimated OR 1.42)
Than et al, 2016, USA28	Patients who underwent lumbar discectomy, 51% male, mean age 45 years	Inclusion: 18–80 years, symptomatic lumbar disc herniation recalcitrant to non-invasive therapies for at least 6 weeks exclusion: history of previous lumbar spinal surgery at the level of disc herniation; significant motor weakness (such as foot drop) or cauda equina syndrome; cancer, infection or fracture involving any portion of the spine; pregnancy	n=127, n=123 complete cases at 1-year follow-up	Physical function/general health (SF-36 scale), physical function (Oswestry Disability Index), BMI, back pain (VAS), age, sex, insurance type, work status, smoking status, baseline health status measures, self-reported work/disability status	Return to work, 3 months, case definition: employed at 3 months following the lumbar discectomy	Stepwise logistic regression	Age OR 0.92 (0.85 to 0.99), male sex OR 0.22 (0.04 to 1.09), BMI OR 0.90 (0.78 to 1.04), general health OR 1.03 (0.98 to 1.08), physical function OR 1.06 (0.997 to 1.13), smoking status OR 4.37 (0.82 to 23.27)

OR with 95% CI in brackets.