Salem 2014.

Methods	Trial design: cluster‐randomised study design, where in 12 primary schools 2 classes from each grade were randomly assigned to 2 treatment arms Follow‐up: 24 months Study was started in 2009 The study was not considered in analyses of this review because results and data ‐ complex multi‐level model with teeth nested in a child nested in a school class nested in a school ‐ were not in useable form for this review (unit of analysis was chosen to be a tooth surface, but clustering of data was not taken into account in the analyses. Study author gave the following additional information on analyses of the trial: "During study period we found that the arrangement of pupils in each class is undergoing substantial changes every year. Hence there was a combination of both groups in each class at the second year and we decided to analyse the study population at surface level")
Participants	Location: university clinic, Iran Inclusion criteria: Children had to have ≥ 1 sound (by Nyvad criteria) and newly and completely erupted first permanent molar with deep occlusal fissure Age at baseline: 6 to 7 years Gender: girls 38%, boys 62% Baseline caries: sealant group: mean dmft index 4.41 (± 0.92); fluoride varnish group: 4.76 (± 2.75) (additional information was obtained from study authors) Number randomly assigned: 400 children (sealant 200, varnish 200) with 1579 occlusal surfaces (mean number of teeth 3.9 per child) Number evaluated: 352 at 2 years (173 in sealant group, 179 in fluoride varnish group)
Interventions	Comparison: resin‐based fissure sealant vs fluoride varnish 2 treatment arms Group 1: resin‐based sealant (Eco Seal), applied by dentist without assistant (additional information obtained from study authors). Partially and completely lost sealants were repaired/reapplied once if needed after 6 months Group 2: sodium fluoride varnish (Durafluor, NaF 5%), applied biannually by dentist without assistant (additional information obtained from study authors) Co‐interventions: All children participated in oral hygiene education sessions including restriction of sugary snacks, regular toothbrushing and toothbrush and 1450 ppm fluoride toothpaste given at every visit Fluoride concentration of water during the years 2009 to 2012 at study area was in the range of 0.45 to 0.8 ppm, on the basis of seasonal changes
Outcomes	Sound or carious occlusal surface of first molar: caries numbers for occlusal surfaces scored by Nyvad criteria. DMFT and DMFS scores reported for whole tooth surfaces One calibrated dentist carried out all examinations. Caries status was determined by 2 visual‐tactile measures: WHO criteria (DMF) and Nyvad criteria applied
Notes	Intra‐examiner reliability: Kappa co‐efficient for Nyvad and WHO criteria was 0.79 and 0.81, respectively Sealant retention: after 24 months: completely retained 43%, partial retention 46%, total loss 11% Funding source: granted by Institutional Review Board and Ethical Committee of Guilan University of Medical Sciences, Iran
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "In each school, one class from each grade was assigned to sealant and the other to varnish by coin tossing" Comment: Random sequence generation was adequate
Allocation concealment (selection bias)	Low risk	Quote from author correspondence: "The records were kept by dental nurses at dental clinic files. The nurses were the staff of university dental clinic and did not know the children or their dental history" Comment: Independent staff combined with cluster‐randomised study design gives the impression of adequate randomisation procedure
Blinding of outcome assessor (detection bias)	Low risk	Quote from author correspondence: "The examiner was not involved in the study design and had no access to records" Comment: Blinding of outcome assessor was adequate
Incomplete outcome data (attrition bias) All outcomes	Low risk	Missing data: 27/200 (13.5%) in sealant group, and 21/200 (10.5%) in fluoride varnish group Reason for drop‐outs in both groups: Children moved away from the area Comment: Missing data were less than 25%, and groups were balanced in numbers and reasons for missing data
Selective reporting (reporting bias)	Low risk	Outcomes reported: caries response, sealant retention Comment: Pre‐specified outcomes (in methods) were reported in pre‐specified way
Other bias	Low risk	Comparability of groups: Information on groups was available at 2 years (additional information was obtained from study authors): Baseline mean dmft in sealant group was 4.41 (± 0.92) and in fluoride varnish group 4.76 (± 2.75). Proportion of girls was 46% in sealant group and 31% in varnish group Quote: "The majority of children reported brushing their teeth once daily and sugary snacks 1‐2 times per day without significant difference between the groups" Comment: Detailed description of demographic characteristics and caries risk level was provided at baseline. Although the proportion of girls was greater in sealant group than in fluoride varnish group, we assessed that groups were balanced at baseline Co‐interventions: All children participated in oral hygiene education sessions including restriction of sugary snacks and regular tooth brushing, and received toothbrush and 1450 ppm fluoride toothpaste at every visit Comment: In both groups, the same co‐interventions were allowed