| Study | Design and methods appraisal* | Setting and participants | Intervention | Implementation | Psychosocial outcomes (p<0.05)†,‡ | Health outcomes (p<0.05)‡ |
| Kawakami et al. (1997)42 | Prospective cohort study with comparison group 1- and 2-year follow-upFinal sample: n = 187 Methods appraisal: 1, 2, 3, 4, 5, 7, 8, 9, 10 | Factory, Japan Manual workers | Stress reduction “working committee” comprising worksite supervisors, personnel staff and corporate medical staff. More and smaller teams with sub-supervisors and more on-the-job training; and ergonomic improvements | Authors report that employers supported the intervention, although one aspect (on-the-job training) was not fully implemented. Prior support from employees is not reported | Work overload (D) ↓ Control (C) ↔ Problems with co-workers (S) ↔ Problems with supervisors (S) ↔ Chance to learn (O) ↔ | Mean depression (Zung SDS score) ↑ Absenteeism ↑ Systolic blood pressure ↔ Diastolic blood pressure ↔ |
| Evanoff et al. (1999)43 | Retrospective, repeat cross-sectional study (with “pre-” and “post-” routine data analysis on absenteeism) Baseline 1 month after intervention. 6- and 14-month follow-up Final sample: n = 87 Methods appraisal: 2, 4, 5, 7, 10 | Hospital, USA Hospital orderlies | Participatory ergonomics team consisting of three orderlies and one supervisor Ergonomic interventions include new procedures and training for heavy lifting and limited use of mechanical aids. Stated aim was to reduce injury rates | Few reported details on effectiveness of implementation or commitment of employers or employees | Psychological stressors (D)↑ Combined measures (C) ↑ Social support (S) ↑ Job satisfaction (O) | Self-reported musculoskeletal illness ↑ Absenteeism ↑ |
*Methods appraisal: 1 = prospective; 2 = representative sample; 3 = appropriate comparison group; 4 = baseline response >60%; 5 = follow-up >80% in cohort, >60% in cross-section; 6 = adjustment for non-response and drop-out; 7 = conclusions substantiated by data; 8 = adjustment for confounders; 9 = all intervention group exposed, non-contaminated comparison group; 10 = appropriate statistical tests. †D, demand; C, control; S, social support; O, other psychosocial outcome measures. ‡↑ = improvement; ↓ = worsening; ↔ = little change/inconclusive (with reference to the DCS hypothesis that reduced demands and increased control and support are ‘improvements’).