Kajita 2010.

Methods	Study design: C‐RCT Unit of allocation: municipal health centres Type of comparison: PEM only vs. nothing group A: no intervention group B: Mmiled information packet
Participants	Nurses, public health nurses, and allied health professionals in the field of community health Clinical speciality: community health Level of training: fully trained Setting/country: community‐based (e.g. community health centre, public health department)/Japan
Interventions	The intervention was the distribution of an evidence‐based guideline. The guideline was entitled "Evidence‐based guideline for the prevention of osteoporosis and osteoporotic fractures in community health", a purely evidence‐based practice guideline written in Japanese for the prevention of osteoporosis published in October 2004. This guideline was developed and formatted in accordance with recommendations for evidence‐based guidelines, according to formal assessment procedures specified in the Japanese version of the AGREE instrument
Outcomes	46 process outcomes, including implementation rate of evidence‐based health education items for osteoporosis prevention (see Table 5 for a complete list)
Notes	‐
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote, pg. 2: "after the pre‐intervention assessment, the 100 centers were randomly allocated in a 1:1 ratio to the intervention and control group by a minimization method that defined region and city/town as stratification factors"
Allocation concealment (selection bias)	Low risk	Quote, pg. 2: "the allocation was performed by the controller of the trial (M. I.), who was not involved in the assessment as an evaluator"
Baseline characteristics similar (selection bias)	Unclear risk	Quote, pg. 4: "there were no significant differences between the intervention and control groups in municipality type, population, population aging rate, number of permanent health center staff, or the qualifications of the staff (physicians, public health nurses, nurses, dieticians, physical therapists, and clerks). There was no significant difference between the intervention and control groups in the implementation rate for osteoporosis screening or any type of health education or counseling before the intervention" COMMENT: numerical data to support this was not provided
Baseline outcome measurements similar (selection bias)	Low risk	Quote, pg. 4: "there was no significant difference in the overall score for the implementation status of evidence‐based health education items, as recommended by the guideline, between the intervention (median, 10; first and third quartiles: 3, 17) and control (median, 9; first and third quartiles: 1.5, 18.5) groups in the pre‐intervention assessment. The Table shows the implementation status of each health education item in these groups"
Incomplete outcome data (attrition bias) All outcomes	Low risk	Quote, pg. 4: "all 100 municipal health centers completed the preintervention assessment. Of these, 3 centers declined to participate in the trial and 1 center was absorbed into another municipality (Figure 1). We performed the post‐intervention assessments for the remaining 96 centers (48 in the intervention group and 48 in the control group; 96% follow‐up rate)"
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote, pg. 3: "the post‐intervention assessment was performed 1 year after the distribution of the guideline under blinded conditions in which the evaluators were unaware of the allocation"
Contamination protection (contamination bias)	Low risk	COMMENT: the unit of allocation was by institution (health centre)
Selective reporting (reporting bias)	Low risk	All relevant outcomes in the methods section were reported in the results section
Other bias	High risk	Quote, pg. 9: "the study did not use a double‐blind design because it was not possible to use a placebo guideline. Instead, we offered to reimburse the control centers for the cost for materials needed to revise their health education programs. Although only 3 centers claimed reimbursement, our offer may have increased the use of information other than the guideline in the control group and may have improved the evidence‐based status of the programs of the control centers, thereby decreasing the magnitude of differences in the outcome measures between the groups"