Schwarz 1998

Methods	Study design: controlled before‐and‐after study Conducted in: China Unit of randomisation: kindergarten Unit of analysis: children Setting: kindergarten Funded by: "This study was financially supported by the University of Hong Kong (CRCG). The study received continuous material support from Colgate‐Palmolive (HK) and Colgate (Guangzhou) Co. Ltd" Duration of the study: 3 years
Participants	Inclusion criteria: not reported Exclusion criteria: not reported Age at baseline: Of 289 children, 94% were 3 years old, 4% were not yet 3 years old and 2% were 4 years old N (controls baseline): 121 N (controls follow‐up):99 N (interventions baseline): 168 N (interventions follow‐up): 152 Recruitment: All children studying in grade 1 were recruited Gender: not reported
Interventions	Intervention: received supervised toothbrushing and oral health education sessions. Research team paid a visit every 4 months to ensure that activities were on track, observations of classroom activities were conducted and meetings with teachers were set up to discuss ways of standardising the approach Control: No dental health education or other information/activities were provided for control kindergarten teachers. However, they were aware of ongoing activities Duration of intervention: 3 years
Outcomes	Caries development (dmfs)
Implementation related factors	Theoretical basis: not reported Resources for implementation: teacher and parent, dentist/dental hygienist Who delivered the intervention: teachers in kindergartens PROGRESS categories assessed at baseline: not reported PROGRESS categories analysed at outcome: not reported Outcomes related to harms/unintended effects: not reported Intervention included strategies to address diversity or disadvantage: not reported Economic evaluation: not reported
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	All children studying in grade 1 of the 3 largest kindergartens in the township were recruited into the study. Children in the largest kindergarten constituted the test group (n = 168), and those in the other 2 kindergartens formed control groups (n = 121)
Allocation concealment (selection bias)	High risk	Teachers from the control group were aware of ongoing activities in the intervention group
Incomplete outcome data (attrition bias) All outcomes	Low risk	Overall loss of less than 20% of children after 3 years was considered low. Assessments compared children who remained in the study versus those who dropped out with regard to baseline dmfs, parent education level and household income, with no significant differences noted. Thus, the dropout level was considered to not seriously affect outcome evaluations
Selective reporting (reporting bias)	Low risk	Published report presents all expected outcomes of interest for the review
Other bias	Unclear risk	Unclear
Blinding of participants and personnel (performance bias) All outcomes	High risk	Teachers who were among programme implementers from the control group were aware of ongoing activities in the intervention group
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Unclear