Ponce 2017.

Study characteristics
Methods	Study design: cRCT Study grouping: parallel group How were missing data handled? NR Randomisation ratio: 1:2:1 Recruitment method: NR Sample size justification and outcome used: power estimates as well as sample size were computed using the Optimal Design software, working with a power of 80%, at 5% significance and with a minimum detectable effect of 0.25 (SD). However, it was NR which outcome was used in this calculation. Sampling method: 3 groups of approximately 200 HHs per group were randomly selected in 3 provinces. Study aim or objective: to differentiate the effects of food vouchers and training in health and nutrition on consumption and dietary diversity in Ecuador by using an experimental design. Study period: baseline data collected between September and November 2013, and the intervention lasted 1 year thereafter. Unit of allocation or exposure: HHs
Participants	Baseline characteristics Intervention T1: food vouchers Age: HH head, years: 32.6 (no variance provided) Place of residence: NR Sex: female head of HH, %: 17 Ethnicity and language: Head of HH Mestizo race, %: 76.6 Occupation: NR Education: head of HH schooling, mean, years: 7.731 (no variance reported) SES: NR Social capital: NR Nutritional status: HH energy intake, mean, kCal/day: 7529.6 (no variance reported). HH per capita energy intake, mean, kCal/day: 1635,6 (no variance reported). DDI, mean: 5.96 (no variance reported) Morbidities: NR Concomitant or previous care: NR Intervention T2: food vouchers + training on health and nutrition Age: HH head, mean, years: 34.0 (no variance reported) Place of residence: NR Sex: female head of HH, %: 17.5 Ethnicity and language: head of HH Mestizo race, %: 81 Occupation: NR Education: head of HH schooling, mean, years: 7.638 (no variance reported) SES: NR Social capital: NR Nutritional status: HH energy intake, mean, kCal/day: 7415.6 (no variance reported). HH per capita energy intake, mean, kCal/day: 1609.6 (no variance reported). DDI, mean: 5.83 (no variance reported) Morbidities: NR Concomitant or previous care: NR Control: received no intervention Age: HH head, mean, years: 33.8 (no variance reported) Place of residence: NR Sex: female head of HH, %: 17 Ethnicity and language: head of HH Mestizo race, %: 79.5 Occupation: NR Education: head of HH schooling, mean, years: 7.535 (no variance reported) SES: NR Social capital: NR Nutritional status: HH energy intake, mean, kCal/day: 7638.6 (no variance reported). HH per capita energy intake, mean, kCal/day: 1660.8 (no variance reported). DDI, mean: 5.89 (no variance reported). Morbidities: NR Concomitant or previous care: NR Overall: NR Inclusion criteria: NR Exclusion criteria: NR Pretreatment: except for the first intervention group (T1, food voucher only) that had significantly fewer (P = 0.008) HH members aged 45–64 years when compared to the second intervention group (T2, food voucher + training), there were no significant differences, indicating that the randomisation worked properly. Attrition per relevant group: unclear if 'cases' referred to HHs or individuals (conflicting statements), and, therefore, it is unclear how many HHs/individuals were enrolled in the study and how many completed the study or what data were included in the analyses. Description of subgroups measured and reported: NR Total number completed and analysed per relevant group: T1, food voucher only: individuals analysed 336; T2, food voucher + training: individuals analysed 512; control: unclear (from Table 3) Total number enrolled per relevant group: T1, food voucher only: HHs at baseline 171; individuals at baseline 920; T2, food voucher + training: HHs at baseline 401; individuals at baseline 2426; control: HHs at baseline 201; individuals at baseline 997 Total number randomised per relevant group: T1 group (food voucher only): 171 HHs; T2 group (food voucher + training): 401 HHs; control: 201 HHs.
Interventions	Intervention characteristics Intervention T1: food vouchers Food access intervention category: food prices Intervention type: food vouchers Description: HHs received a food voucher of USD 40 monthly Duration of intervention period: unclear, but seemed to have been for 1 year because follow‐up assessment was performed at 1 year after baseline Frequency: monthly Number of study contacts: NR Providers: NR Delivery: NR Co‐interventions: NR Resource requirements: NR Economic indicators: NR Intervention T2: food vouchers + training on health and nutrition Food access intervention category: food prices Intervention type: food vouchers + training on health and nutrition Description: HHs received a food voucher of USD 40 monthly + monthly training sessions on topics that included malnutrition, food preparation, children's health, mother's health, women's rights and women's empowerment. Unclear whether only the head of the HH received this training or whether more (e.g. all adults) or all members of the HH received this training. Duration of intervention period: unclear, but seemed to have been for 1 year because follow‐up assessment was performed at 1 year after baseline Frequency: monthly Number of study contacts: NR Providers: NR Delivery: NR Co‐interventions: NR Resource requirements: NR Economic indicators: NR Control: no intervention
Outcomes	Dietary diversity: FCS Dietary intake: per capita energy consumption at HH level in the previous 7 days
Identification	Sponsorship source: WFP in Ecuador and the Spanish Ministry of Science and Innovation (grant number HAR2013‐47182‐C2‐1‐P). Country: Ecuador Setting: 3 provinces in Ecuador, 2 from the Sierra region (Carchi and Chimborazo) and 1 from the Coastal region (Santa Elena)." Comments: no protocol or trial registry number reported Author's name: Jesus Ramos‐Martin Institution: N/A Email: jramos@flacso.edu.ec Address: N/A Declarations of interest: none Study or programme name and acronym: N/A Type of record: journal article
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (Selection bias)	Unclear risk	Quote: "In our study, we randomly assigned households to one of the following groups …" Comment: however, the method of generating the random sequence was NR.
Allocation concealment (Selection bias)	Unclear risk	NR in the paper.
Baseline characteristics similar (Selection bias)	Unclear risk	Not many baseline characteristics were reported, thus it is not possible to make a fair judgement. e.g. HH income and farming activities are important characteristics for this study's research question but these were NR.
Baseline outcome measurements similar (Selection bias)	Low risk	HH energy intake, per capita energy intake and dietary diversity were measured at baseline and there were no meaningful differences between groups.
Blinding of participants and personnel (Performance bias)	Low risk	Blinding was not done, but it was unlikely that the lack of blinding influenced the participants' performance.
Blinding of outcome assessment (Detection bias)	High risk	Blinding of study participants was not possible. Outcomes were self‐reported and could have been influenced by lack of blinding.
Protection against contamination (Performance bias)	Unclear risk	Unclear how close HHs that participated in the study, and who were allocated to different intervention/control groups, lived from each other. It is possible that HHs from the intervention groups could have sold, shared or exchanged the extra food they got through the vouchers with other HHs; however, unclear from the manuscript whether this was likely or not.
Incomplete outcome data (Attrition bias)	High risk	Unclear how many individuals were analysed at endpoint in the control group, but there appeared to be high attrition in the 2 groups.
Selective outcome reporting (Reporting bias)	Unclear risk	Study authors did not mention whether they had an a priori protocol or if they registered their trial prospectively in a trial registry. All relevant outcomes in the methods section were reported in the results section of the manuscript.
Other bias	Unclear risk	Risk of misclassification bias: N/A as this was a trial. Risk of measurement bias: unclear, because the method used to collect 'Weekly HH values of food consumption' was NR. Risk of incorrect analysis: low, because outcomes were analysed at the HH level. Risk of recruitment bias: unclear because it is not clear whether recruitment took place before or after randomisation of HHs.