Wallon 2006.

Methods	Cluster‐randomized trial. Units of randomization were prenatal clinics.
Participants	5023 seronegative pregnant women attending prenatal clinics in 7 counties in the area of Lyon (France). Completed both the pre‐test and the post‐test questionnaire 2790 women (1953 in the experimental arm and 837 in the control arm). Loss to follow‐up was 44.5%.
Interventions	Physicians in experimental cities were instructed to give their patients: ‐ a 20‐page brochure containing relevant information about different aspects of pregnancy, including 4 pages on toxoplasmosis prevention and transmission; ‐ an audiotape containing frequently asked questions during pregnancy, including questions on toxoplasmosis.
Outcomes	Knowledge and behavior change including: ‐ consumption of undercooked meat of any type; ‐ handwashing after contact with transmission factors (soil, raw meat, unwashed vegetables) and before meals. Seroconvertion for toxoplasmosis.
Notes	Knowledge and behavior was measured through a questionnaire at baseline (during the third and fourth months of pregnancy) and at follow‐up (at delivery). There was a large and differential loss‐to‐follow‐up: completed the follow‐up questionnaire and were therefore included in the analysis 60% of the women enrolled in the experimental group and 48% of the women enrolled in the control group. A multivariate analysis was conducted but it was not specified if the statistical plan of analysis considered to correct for cluster design.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Not described in the publication.
Allocation concealment (selection bias)	Unclear risk	Not described in the publication.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Blinding of the women reported, blinding of the personnel was not possible, but unlikely to influence the outcome as the outcome assessed by the personnel was seroconversion rate.
Blinding of outcome assessment (detection bias) All outcomes	High risk	Measure of knowledge and behavior were self‐reported.
Incomplete outcome data (attrition bias) All outcomes	High risk	Overall losses to follow‐up were 44.5%: women in the control group were more likely to be lost at follow‐up (52% were lost) than women in intervention group (40% were lost).
Selective reporting (reporting bias)	Low risk	All of the study’s pre‐specified outcomes and all expected outcomes of interest to the review (including seroconversion rate) have been reported.
Other bias	Unclear risk	Study method and statistical plan of analysis were not described in details, therefore it is not possible to exclude other bias.