Harrison 2010.

Study characteristics
Methods	Study design: cluster‐RCT Study recruitment: January 2005–October 2007; follow‐up 30 (SD 3) months Published protocol/trial: protocol published and trial registered; ISRCTN41467632
Participants	Description: women at 12–34 weeks of pregnancy and mothers of newborn, predentate infants from 9 communities (intervention n = 131 in 5 communities, control n = 141 in 4 communities) Exclusion criteria: any woman knowing of an impending, permanent move out of her community. Indigenous population: Quebec Cree Nation Eeyou Istchee Setting: Canada, remote Place of delivery: health clinic Principle health condition: early childhood caries Age of mother (years), mean: intervention 25.5 (SD 6.4), control 25.6 (SD 5.8) Gender of infant: none reported Number of children in family: has other children: intervention 83 (64.3%), control 92 (65.7%) Family unit: none reported Socioeconomic status: none reported Employment of primary carer: none reported Education of primary carer: none reported
Interventions	Intervention Intervention name: MI Intervention aim: to control caries in Indigenous children Theory used to develop intervention: follows the principles of MI, a client‐centred but directive counselling style. Consumer and community involvement: 2 years of community consultation was completed. This was mainly around study design and the use of RCT. The decision to use an RCT was decided based on no health promotion programme for oral health previously being completed. Community Health Representatives who delivered the intervention modified the intervention protocol and resources (menus) to fit better with the "way of being" of the Cree. The menus were customised for various stages of infant and toddler development and were printed on flip charts that included images of local children and families. Overall grouping: counselling Fees, reimbursement, or incentives: none reported Procedures: at the counselling sessions, mothers were given the resources to enable them to adopt their selected behaviours, e.g. infant toothbrushes and fluoride toothpaste, and xylitol gum for the mother. Menus were developed at specific age ranges to reflect changes over time in each child's feeding and snacking habits, and dental development. FV was provided as a choice of care after the child's first birthday. Mothers also received a 'Privilege Card' to expedite dental care. Materials: mothers received resources at each MI visit to enable them to implement selected behaviours, e.g. infant toothbrushes, toothpaste, and sippy cups. Mothers also received a pamphlet explaining children's dental care practices. Mode of delivery: well‐child visit provided individually to each family face‐to‐face with home visits as an option. When and how often was the intervention delivered? mothers were counselled once during pregnancy and up to 6 times postnatally to correspond with the 2‐, 4‐, 6‐, 12‐, 18‐, and 24‐month well‐child visits at local clinics. Who delivered the intervention? Community Health Representatives: existing health workers and, where not available, local women were employed. Was there any training provided to the people who delivered the intervention? a 2‐day training workshop for Community Health Representatives was held in the first year of the project in a community adjacent to Eeyou Istchee that was accessible by air or ground transport. An MI consultant delivered the training. The consultant presented theory and principles of MI. Explanatory notes were developed for the Community Health Representatives to guide their counselling sessions. The MI consultant followed up months later with an MI coaching conference call to problem‐solve MI with and support the Community Health Representatives. A second 2‐day workshop for all intervention Community Health Representatives was held in year 2 in an intervention community. Following this workshop, the Project Manager regularly visited each of the communities individually to problem‐solve issues with recruiting and delivery of the counselling interventions. The Programme Manager also maintained regular telephone contact with the project's staff in each community. Was the study modified or adapted? recruitment and delivery of the intervention changed between sites. In 2/5 intervention communities and 2/4 control communities, recruitment and, for the intervention communities, delivery of the intervention was eventually completed by the Project Manager. Local women who were not Community Health Representatives completed recruitment in the remaining 2 control communities. Was the fidelity assessed? no Comparison Pamphlet group This was an education programme. The mothers received a culturally appropriate educational pamphlet describing healthy dental care practices for young children. Pamphlets were mailed to mothers when their child was aged 6 and 18 months. FV was available to control children at local dental clinics. Fees, reimbursement, or incentives: none reported
Outcomes	Overall health and well‐being: tooth level caries Child psychological health and emotional behaviour: parent answered 'yes' to > 1 quality of life question Child physical health and development: tooth level caries Service access and utilisation: number of visits to the dentist for tooth pain Adverse events: whether there were adverse events Economic costs: cost‐effectiveness Time points: 30 (SD 3) months of age
Funding source and conflicts of interest	Funding: Canadian Institute of Health Research (grant #FRN 67817) Conflict of interest: none reported
Notes	Other outcomes recorded but not used within the review Child: data on dental caries including enamel caries, dentinal caries, pulpal caries, restorations, and absence due to caries Mother: dental health knowledge and home care behaviours
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Communities were alphabetically ordered to receive their intervention, which was drawn from lots. Quote: "Communities were randomized in each round by alphabetical ordering of the communities' names. For example, for each round, the first name on the alphabetical list of communities was announced, followed by the drawing of an envelope from the basket; the next name was announced, followed by another draw until all envelopes were allocated."
Allocation concealment (selection bias)	Unclear risk	There was little information provided about allocation concealment other than the drawing of envelopes. Quote: "Randomization was done over community radio with two "rounds" of a constrained randomization process. Two baskets contained envelopes marked "test" or "control": one basket for large communities (2 envelopes: 1 test, 1 control) and another for smaller communities (7 envelopes: 4 test, 3 control)."
Selective cluster recruitment	High risk	Communities knew of their allocation prior to enrolment. Individuals who recruited women also delivered the intervention. There was some baseline differences in recruitment including fewer intervention mothers had already delivered at time of enrolment, had visited a dentist for toothache, and had other children with a previous tooth extraction.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Participants and people delivering the intervention were not blinded to the intervention. Quote: "Mothers and interveners were aware of their community's allocation."
Blinding of outcome assessment (detection bias) All outcomes	Low risk	People collecting data were blinded to the intervention. Quote: "The examiners were blinded to allocation and were unfamiliar with the intervention."
Incomplete outcome data (attrition bias) All outcomes	High risk	Intention‐to‐treat analysis completed. Greater loss to follow‐up in control group compared to intervention group (83% vs 93%).
Selective reporting (reporting bias)	Unclear risk	Protocol available. Data provided as per definition; however, caries reported at tooth and child level and this was not distinguished in the protocol and was the primary outcome.
Other bias	Unclear risk	Deviation from intended intervention: intervention mothers were given "Privilege Cards" to allow for priority access to dental services. However, because of the turnover of dental staff, not all clinics honoured the cards. About one‐third recalled using the cards; however, it is unclear how many mothers were turned away. No other obvious sources of bias.