Cradock 2011.

Methods	Design: CBA Timing: Retrospective Allocation to group: Quote: "This quasi‐experimental evaluation study design contrasted local trends in consumption of sugar‐sweetened beverages, measured before and after the policy change was implemented, with national trends in consumption among high‐school–aged adolescents" Number of clusters or sites: This was a repeat cross‐sectional analysis. In the IG students from 17 high schools were sampled at baseline, and students from 18 high schools were sampled at follow‐up. In the CG, data were not collected at the cluster level Number of individuals: This was a repeat cross‐sectional analysis. At baseline, 895 students were included in the IG, and 1196 in the CG. At follow‐up, 1138 students were included in IG, and 1233 in the CG (see tables 1 and 3 of the study's primary report). Length of intervention: 2 years (in the IG, baseline data collection took place from February to April 2004, and follow‐up data collection from February to April 2006. In the control group, baseline data were collected in 2003 ‐ 2004, and follow‐up data in 2005 ‐ 2006)
Participants	General description of participants: High‐school students in grades 9 through 12 in Boston, USA Age: Teenagers (no details on the age reported; IG participants were 9th to 12th grade students, and CG participants were aged 15 ‐ 19 years) Inclusion criteria: Quote: "Students [in the IG] with complete data on demographic covariates and consumption of sugar‐sweetened beverages were included in this analysis" Exclusion criteria: Quote: "We excluded from analysis [of participants in the CG] data from respondents who self‐identified as 'other' race/ethnicity because of the small sample size" Recruitment: The study is based on data from the Boston Youth Survey conducted by the City of Boston and the Harvard Youth Violence Prevention Center, as well as on nationally representative data from the National Health and Nutrition Examination Survey conducted by the National Center for Health Statistics in the USA. The study's primary report does not provide details on recruitment, but includes references to relevant publications Weight status at baseline: N/R SSB consumption at baseline: Baseline SSB consumption was 607 ml/day in the IG, and 618 ml/day in the CG Equity considerations: Quote: "The student body [in the IG] is diverse: 37% of students are black, 39% are Hispanic, 13% are white, and 9% are Asian, and approximately 74% of students are eligible for free or reduced‐price meals"
Interventions	Intervention: Reduced availability of SSB at schools. (Quote: "In Boston, a policy that restricts the sale of sugar‐sweetened beverages in vending and à la carte settings was approved by the Boston School Committee in June 2004 and initiated with the fall 2004 school year. The new Boston Public Schools Snack and Beverage Policy and subsequent detailed implementation guidelines required that beverages sold in schools or on school grounds adhere to the Massachusetts à la Carte Food and Beverage Standards to Promote a Healthier School Environment published by Massachusetts Action for Healthy Kids. The beverage guidelines specifically precluded the sale of soft drinks, fruit drinks (ie, non–100% vegetable or fruit juice beverages), and sports drinks anywhere in school buildings or on school campuses and had specifications that limited other beverage serving sizes") Behavioural co‐intervention: Nutrition education and awareness‐raising (Quote: "The passage of the Boston Public Schools beverage policy was the beginning of more widespread focus on promoting more healthful foods and beverages in Boston and Boston schools. For example, further initiatives included the implementation of nutrition‐related curricula in middle and primary schools and interdepartmental committees and collaborations charged with monitoring implementation and acceptance of related policy guidelines. Awareness‐raising activities in the Boston Public Schools system included a presentation of the new policy guidelines to principals before implementation, parent workshops on healthful snack choices, dissemination of pamphlets to teachers and school staff detailing alternatives for fundraising, and a brochure for school administrators and teachers entitled Healthy Beverages and Snack News. Boston city officials also negotiated new procurement contracts with vendors who would supply the new more healthful options to schools, and school vending machines were stocked with water and 100% juice instead of sugar‐sweetened beverages") Control: No, minimal or alternative intervention (the study compares trends in Boston (the IG) with nationally representative data from NHANES)
Outcomes	Measures of SSB intake: Intake of SSB in servings/day, assessed through a 7‐day dietary recall (IG) and a 24‐hour dietary recall (CG), at baseline and 2 years post‐intervention Measures of intake of alternatives to SSB: N/R Anthropometric measures: N/R Adverse outcomes: Substitution effects, assessed by noting that total SSB intake (including SSB intake outside school) decreased after access to SSB at school was restricted (Quote: "Boston’s results also suggest that youth may not compensate for in‐school restrictions on sugar‐sweetened beverages by increasing consumption outside of school") Other outcomes: None included in this review
Context and implementation	Setting: Schools Sector: Education Country: USA Year(s) when implemented: 2004 ‐ 2006 Mode of implementation: Mandatory government regulation Level of implementation: Policy‐level intervention
Declarations	COI: No information provided by study authors (the study's primary report does not contain a COI section) Funding: "This work was supported by cooperative agreement nos. U48/DP000064 and 1U48DP001946 (including the Nutrition and Obesity Policy Research and Evaluation Network) from the Centers for Disease Control and Prevention; a Steps to a Healthier US grant to the Boston Public Health Commission; and the Robert Wood Johnson Foundation (nos. 260639, 61468, and 66284)" Trial registration: N/R Protocol availability: Protocol mentioned in the report but not publicly available
Notes	None
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	This was a CBA, and allocation was not random and not concealed
Allocation concealment (selection bias)	High risk	This was a CBA, and allocation was not random and not concealed
Similarity of baseline outcome measurements (selection bias)	Low risk	Baseline outcome measurements were similar (baseline SSB consumption was 607 ml/day in the IG, and 618 ml/day in the CG)
Similarity of other baseline characteristics (selection and performance bias)	Unclear risk	Baseline differences were substantial. Differences in observed baseline characteristics were taken into account in the analysis, but unobserved baseline differences seem likely. Quote: "We used linear regression analysis to examine changes in mean servings per day of sugar‐sweetened beverages between 2004 and 2006, adjusting for potential differences in student composition. In regression models, we estimated change in consumption via an indicator variable identifying surveys completed in 2006 (postpolicy change; 2004 survey was reference), controlling for respondents’ sex, grade, race/ethnicity, and primary neighborhood of residence"
Incomplete outcome data (attrition bias) All outcomes	Low risk	This was a repeat‐cross‐sectional analysis, samples were chosen to ensure representativeness, and missing data seem unlikely to have substantially affected the results. Quote: "In 2004 and 2006, respectively, 3.5% and 2.8% of respondents were missing data on consumption of sugar‐sweetened beverages. (...) Regression analysis including covariates for age, sex, and race/ethnicity (non‐Hispanic white, non‐Hispanic black, and Mexican American) accounted for potential differences in the sample population demographics across the 2 survey periods"
Blinding (performance and detection bias) Subjective outcomes	High risk	Outcomes were self‐reported, and participants were not blinded
Contamination (performance bias)	High risk	The IG and CG were overlapping, and it is likely that parts of the CG received the intervention. This would have biased the observed effects towards null. Other cities and states in the USA, forming part of the CG, may have implemented policies aimed at reducing the consumption of SSB during the study period
Selective reporting (reporting bias)	Low risk	All results mentioned in the Methods section are reported. No outcome measures one would expect in a study of this kind are missing
Other bias	High risk	Data were not collected in parallel in the IG and CG, and data collection methods differed substantially. Quote: "Furthermore, NHANES and Boston Youth Survey estimates of consumption of sugar‐sweetened beverages are not directly comparable because of differences in wording and data collection methods"