| Study | Design and methods appraisal∗ | Setting and participants | Intervention and implementation† | Psychosocial outcomes (p<0.05)‡§ | Health outcomes (p<0.05)§ | ||
| Kawakami et al (1997)41 | Prospective cohort with comparison group1- and 2-year follow-upsFinal sample n _ 187Methods appraisal: 1, 2, 3, 4, 5, 7, 8, 9, 10, | Factory floor, JapanManual workers | More and smaller teams with sub-supervisors; participatory committee; more on-the-job training; and ergonomic improvementsIntroduced to reduce stress as the intervention site had the highest levels within the company. 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)Co-worker problems (S)Supervisor problems (S) | ↓↔↔↔ | Mental health(Zung self-rating depression score)AbsenteeismCardiovascular proxies(systolic and diastolic blood pressure) | ↑↑↔ |
| Wahlstedt et al (2000, 2001)32 37 | Prospective cohort with comparison group1-month follow-upFinal sample (n _ 82)Methods appraisal: 1, 3, 4, 5, 7, 10 | Postal sorting office, SwedenManual workers and shop floor supervisors | Increased task variety, more teamwork, more personnel, more time to plan work, bonus schemeIntervention(s) introduced by employers to improve the psychosocial work environment and to reduce sickness absence and staff turnover. Employees volunteered to be involved in the intervention | Psychological work demands (D)Authority over decisions (C)Skill discretion (C)Social support (S)Contact with superior (S)Contact with team mates (S) | ↑↑↔↑↔↔ | Shoulder and thoracic symptoms(Nordic questionnaire on musculoskeletal complaints)Neck and low back symptoms (Nordic questionnaire on musculoskeletal complaints) | ↑↔ |
| Fredriksson et al (2001)40 | Prospective cohort. Some qualitative components9-month follow-upFinal sample (n _ 102)Methods appraisal: 1, 2, 4, 9, 10 | Factory floor, SwedenManual workers | Production line introduced, decreased task variety, teamworkingIntroduced to increase productivity, reduce the need for skilled labour and reduce sick leave. Few reported details on effectiveness of implementation or commitment of employees | High job demands (D)Possibilities to influence work (C)Poor social support at work (S) | ↔↓↔ | Musculoskeletal disordersVisits to health centreSickness absence | ↓↓↔ |
| Wahlstedt and Edling (1994, 1997)42 43¶ | Prospective cohort8- and 12-month follow-upsFinal sample (n _ 100)Methods appraisal: 1, 2, 4, 6, 7, 8, 9, 10 | Postal sorting office, SwedenManual workers and shop floor supervisors | More teamwork, more personnel, role clarification, production goals, fewer supervisors, partial change in shift system, increased feedback, new vending machine and microwave ovenIntervention(s) introduced by employers to improve the psychosocial work environment and to reduce sickness absence and staff turnover. Positive experience on one section of workers extended to others. Employees involved in planning the intervention | Psychological work demands (D)Skill discretion (C)Authority over decisions (C)Contact with team-mates (S)Contact with superiors (S) | ↔↑↑↔↔ | Sick leaveSleep and gastrointestinal complaints | ↑↑ |
| Korunka et al (2003)44 | Prospective cohort. Some qualitative components2-months and 1-year follow-upsFinal sample n _ 185Methods appraisal: 1, 2, 4, 5, 7, 9 | Local government office, AustriaManagerial, technical, administrative and customer service workers | More teamwork (through restructuring, team leaders with autonomous budgets), incentive systemIntervention implemented as part of a politically and economically motivated shift in the style of public sector management. Authors imply that implementation may have been of a “poor professional quality” due to “rigid bureaucratic structures” | Job control (C)Social support (S) | ↔↔ | Perceived stress(Weyer Questionnaire)Emotional strain(Weyer Questionnaire)Tiredness(Weyer Questionnaire) | ↔↔↔ |
| Sutherland and Cooper (1989, 1992)45 46 | Prospective repeat cross-section3-year follow-upFinal sample n _ 917Methods appraisal: 1, 2, 4, 5, 9, 10 | Health centres, UKGeneral practitioners (doctors) | More teamwork, new rolesIntervention(s) in the context of the introduction of a new employment contract “at the end of a period of considerable dispute and disagreement”. Few reported details on effectiveness of implementation | Time pressure (D)Pressure due to unrealistic expectations of others (D)Demands of job on family life (D)Demands of job on social life (D)Responsibility (D)Job varietyFreedom to choose method of working (C)Satisfaction with co-workers (S)Satisfaction with supervisors (S) | ↓↓↓↓↓↓↓↑↑ | Free floating anxiety(Crown–Crisp experimental index)Somatic anxiety(Crown–Crisp experimental index)Depression(Crown–Crisp experimental index) | ↓↓↓ |
| Appleton et al (1998)47†† | Retrospective cross-section7-year follow-upFinal sample n _ 285Methods appraisal: 2, 4, 7, ,9 10 | Health centres, UKGeneral practitioners (doctors) | More teamwork, new rolesIntervention(s) in the context of the introduction of a new employment contract. Few reported details on effectiveness of implementation or commitment of employees | Satisfaction with amount of responsibility given (C)Satisfaction with freedom to choose method of working (C)Satisfaction with opportunity to use abilities (C)Satisfaction with job varietySatisfaction with colleagues and fellow workers (S) | ↓↓↔↔↔ | Work-related physical healthPsychological symptoms | ↔↔ |
∗Methods appraisal (box 2): 1, prospective; 2, representative sample; 3, appropriate comparison group; 4, baseline response > 60%; , 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.
†See box 1.
‡D, demand; C, control; S, support.
§↑, Improvement; ↓, worsening; ↔, little change.
¶Study is set in the same workplace as Wahlstedt et al (2000, 2001).32 37
††Study uses baseline results from Sutherland and Cooper (1992) and Cooper et al (1989).45 46 Little change for sample as a whole but health outcomes worsened for lower-grade employees.