De Visschere 2011.

Methods	Cluster‐randomised controlled trial, 3 arms Study period: 2003 to 2008 Planned follow‐up period: 5 years
Participants	Country: Belgium (Flanders) Setting: 14 nursing homes (selected out of 36 nursing homes using stratified cluster sampling) Target audience of intervention: nursing staff (nurses, nursing assistants, and nurse aides) Number of randomised participants (nursing home residents): 1393 Number of analysed participants (nursing home residents): 70 to 214 depending on outcome and measurement point Age, mean, years (SD): 84.93 (6.97) in IG, 86.0 (7.36) in CG1, 83.18 (8.63) in CG2 Female: 73.9% in IG, 77.6% in CG1, 74.4% in CG2 Edentulous residents: 67.8% in IG, 67.9% in CG1, 68.5% in CG2 Residents with dental prosthesis: 77% (all groups) Inclusion criteria: all residents eligible Exclusion criteria: n/a
Interventions	Intervention group Implementation of an oral hygiene protocol including multiple components: Education/skills training: theoretical and practical training session (half day) for all oral health co‐ordinators, who had to educate the other nursing staff (train‐the‐trainer principle) Additional components: introduction session (1 hour) with the director of the nursing home appointment of registered nurses as oral health co‐ordinators (responsible for the implementation procedure on their ward) oral assessment of new residents using newly designed assessment forms preparation of an "individualized oral hygiene plan" for every resident by the oral health co‐ordinators, and integration into daily care by all caregivers involved Control group CG1: usual oral hygiene CG2: complex intervention was implemented in the nursing home, but residents did not receive intervention Co‐interventions n/a
Outcomes	Primary outcomes Oral hygiene measures were assessed at baseline, and every year after the start of the study for a period of 5 years. Dental plaque: Plaque Index (Silness 1964) Denture plaque: method of Augsburger 1982
Funding	GABA International (supported data collection)
Notes	Results/data for primary outcomes only available for baseline, 2 years and 5 years after the start of the study.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: "randomly selected" Insufficient information
Allocation concealment (selection bias)	Unclear risk	Insufficient information
Blinding of participants and personnel (performance bias) All outcomes	Low risk	No blinding of participants and personnel or incomplete blinding, but the review authors judge that the assessed outcomes are not likely to be influenced by lack of blinding.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "examiners were blind as to which residents were included in the intervention or not"
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Correspondence: "every year residents were randomly selected using stratified cluster sampling for outcome assessment (at least 20% residents per nursing home)" Comment: number of residents per assessment period (every year) unclear
Selective reporting (reporting bias)	High risk	No protocol found; results/data for primary outcomes only available for baseline, 2 years and 5 years after the start of the study.
Other bias	High risk	Possible contamination between groups, and high risk of recruitment bias (selection/recruitment of participants after randomisation)