Slade 2011

Methods	Study design: cluster‐randomised controlled trial Conducted in: Australia Unit of randomisation: communities Unit of analysis: individuals Setting: Aboriginal communities of Australia’s Northern Territory Funded by: "Funding was provided by project grant from the Australian National Health and Medical Research Council. Additional support for conduct of the study was provided by the Cooperative Research Centre for Aboriginal Health and the Northern Territory Government Department of Health and Families" Declaration of competing interests: Colgate‐Palmolive Pty Limited of Australia provided free supplies of Duraphat varnish for the study and low‐cost toothbrushes and toothpaste to community stores. One of the authors, Robert Thomson, is Director of the Dental Practice Education Research Unit at the University of Adelaide, which receives funding from Colgate‐Palmolive Pty Limited. None of the authors or study personnel received or receive consulting payments nor any other form of personal benefit from Colgate‐Palmolive Pty Limited Duration of the study: 2 years
Participants	Inclusion criteria Communities: remote location (> 100 km from Darwin), classified as Aboriginal (i.e. management by an indigenous council of community members), sufficient population (≥ 5 births per annum), signed informed consent to participate in the study received from community council Children: within participating communities as follows: Aboriginal identity, as declared by parent or family member; permanent residency in the community, not an outstation, as defined by the council’s population list and updated after consultation with community leaders; age 18 months to less than 48 months; no reported history of asthma; signed informed consent of parent or family member Exclusion criteria: not reported Age at baseline Control: 33 months Intervention: 33.6 months N (controls baseline): 322 N (controls follow‐up): 262 N (interventions baseline): 344 N (interventions follow‐up): 281 Recruitment: Communities were assigned at random to intervention or control groups. A total of 30 consenting communities were randomised to 15 control and 15 intervention communities. All children within any given community were to undergo the same study procedures Gender Male control: 52% Male intervention: 50%
Interventions	Intervention: Three types of interventions were provided for all eligible children and communities in the intervention group Duraphat 3 fluoride varnish was applied to children’s teeth once every 6 months for 2 years with the aim to complete 5 applications per child. The first application took place after the baseline dental epidemiological examination, and the final application was administered after follow‐up examination Advice to parents and family groups about caries prevention was provided in 2 settings. The first was during varnish application when the clinician explained the causes of dental decay and methods to prevent it. This included advice about drinking water, limited sugar exposure, use of fluoride‐containing toothpaste and toothbrushing. After a demonstration of toothbrushing, each parent family member was given the toothbrush, a tube of low‐concentration fluoride toothpaste and a children’s sized, reuseable water bottle. The second setting consisted of children’s play groups and preschools, where the same information and products were provided to parents and family members Community health promotion engaged parents, store owners, community leaders and healthcare workers about oral health and prevention of dental decay in their community. This took place in settings ranging from ‘face painting days’ to formal presentations at community council meetings In addition to reinforcing information presented to parents and family groups, information was provided about community‐wide activities to promote oral health. Recognising that Aboriginal health workers are the principal healthcare providers who promote traditional health practices, we explained the process of tooth decay to them, placing emphasis on the potential caries‐preventive benefits of traditional health practices and ‘bush tucker’ (i.e. food gathered from the land). Reinforcement of the same health promotion messages was conveyed to primary healthcare workers at health centres. Health centre staff were trained in oral disease recognition and referral of children with dental decay to school dental services. Training was supported with chart books and DVD instruction Control: no type of intervention Duration of intervention: 2 years
Outcomes	Dental caries
Implementation related factors	Theoretical basis: socio‐ecological model Resources for implementation: training package, staff time, dental materials. Colgate‐Palmolive Pty Limited of Australia provided free supplies of Duraphat varnish and low‐cost toothbrushes and toothpaste Who delivered the intervention: clinical study personnel: dental therapists or dentists, health centre personnel trained in clinical procedures by the research team PROGRESS categories assessed at baseline: Gender, Residence, Race PROGRESS categories analysed at outcome: not reported Outcomes related to harms/unintended effects: not reported Intervention included strategies to address diversity or disadvantage: Programme targeted remote indigenous communities, families and children Economic evaluation: not reported
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Before randomisation, 6 strata were formed on the basis of 3 characteristics of study communities ‐ timing of community consent; population size; and geographical region. Within each stratum, communities were block‐allocated at random to achieve equal numbers of intervention and control communities within strata. A random allocation algorithm was created by a consultant statistician using Stata software
Allocation concealment (selection bias)	Low risk	Concealment was not possible. Community‐level health promotion activities were self evident, and no attempt was made to conceal community allocation from children, community groups or study personnel
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Unclear
Selective reporting (reporting bias)	Low risk	Published report presents all expected outcomes of interest to the review
Other bias	Unclear risk	Unclear
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Unclear
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Unclear