Skip to main content
. 2015 Apr 8;2015(4):CD010306. doi: 10.1002/14651858.CD010306.pub2

Wu 2006.

Methods Cluster‐randomized trial
Participants 961 teachers in 8 middle schools. 4 intervention (n = 459) and 4 control (n = 502) schools in Sichuan Province, China. After adjusting for clustering, the number of participants used in the analysis were 338 and 370 respectively
Interventions 1‐ Intervention strategies involving both individual (stress management) and organisational approaches (e.g. redesigning the task, establishing flexible work schedules and redesigning the work environment)
2‐ No‐intervention control
Outcomes Occupational Stress Inventory‐Revised Edition, Work Ability Index
Surveys conducted at baseline and 12 months post intervention
Notes Results of baseline measures were fed back to teachers in the intervention schools in order to develop the organisational component of the intervention. There is no further information on the final content of the organisational component
Date of the intervention: Not reported
Funding: National Natural Science Foundation of China (Grant 39970623)
Conflict of interest: Unknown
Risk of bias
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) Unclear risk Schools were randomly selected into study group and control group, but no mention of method of sequence generation
Allocation concealment (selection bias) Unclear risk Not reported
Blinding of participants and personnel (performance bias) 
 All outcomes Unclear risk No evidence, not applicable
Blinding of outcome assessment (detection bias) 
 All outcomes Unclear risk Not reported
Incomplete outcome data (attrition bias) 
 All outcomes High risk No mention of attrition in data and no mention if ITT analysis. Mean difference calculation in Review Manager 5 revealed a significant effect in the Role Insufficiency subscale, while the study does not report it to be significant. This may be due to the authors calculating the mean difference with the actual numbers after attrition, while we had to assume numbers pre‐ and post‐intervention to be the same
Selective reporting (reporting bias) Low risk All stated outcome data reported
Recruitment bias Unclear risk Not reported
Baseline imbalance Low risk No relevant baseline differences in age, gender, educational level and marital status (P > 0.05)
Incorrect analysis High risk There is no mention of accounting for clustering in the analysis, and so we assume that clustering has not been taken into account. We corrected for clustering by employing the intracluster correlation coefficient for the burnout scale (0.003) provided by Brown 2010
Other bias Unclear risk Not reported