Hidrobo 2014.

Study characteristics
Methods	Study design: cRCT Study grouping: parallel How were missing data handled? Overall attrition 10%. Authors excluded data from HHs LTFU and with missing data but they constructed worse‐case scenarios to assess effects of excluding missing data. Construct worst‐case scenarios by assuming those HHs who select into the sample because of treatment (marginal HHs) were at the very top or very bottom of the outcome distribution. No differences in estimates observed in these analyses. Randomisation ratio: neighbourhoods and clusters randomised into 4 treatment groups using percentages of 20/20 for the control and food groups, and 30/30 for the cash and food voucher groups. 80 neighbourhoods and 145 clusters were randomised. Recruitment method: each HH in the selected neighbourhood was visited, mapped and administered a 1‐page questionnaire with basic demographic and socioeconomic questions. Sample size justification and outcome used: NR Sampling method: Neighbourhoods within urban centres were chosen for the intervention by WFP in consultation with the United Nations High Commissioner for Refugees (UNHCR) as areas that had large numbers of Colombian refugees and relatively high levels of poverty. Each HH was visited, mapped and administered a 1‐page questionnaire that consisted of basic demographic and socioeconomic questions designed to develop a PMT to define programme eligibility. However, based on point scores by nationality, the decision was made to automatically enrol all Colombian and mixed‐nationality HHs. First, neighbourhoods were randomised to either treatment or control group; second, all treatment clusters (geographical units within neighbourhoods) were randomised to cash, food voucher or food transfer. 1 unexpected complication in study design was change in beneficiary criteria implemented during baseline survey data collection. In process of surveying HHs, it was concluded that the targeting for transfers was too broad, resulting in the inclusion of HHs who were relatively well off. This led to a retargeting process where HHs who were relatively well off were dropped from the programme. Since there were not enough HHs in existing neighbourhoods to replace those that had been excluded and still reach programme enrolment targets, the decision was made to expand coverage to additional neighbourhoods on the outer circle of urban areas. These areas were subsequently rerandomised into treatment groups according to the approximate percentage lost. Study aim or objective: to compare the impact and cost‐effectiveness of cash, food vouchers and food transfers on the quantity and quality of food consumed. Objectives were 3‐fold: 1. to improve food consumption by facilitating access to more nutritious foods, 2. to increase the role of women in HH decision‐making related to food consumption and 3. to reduce tensions between Colombian refugees and host Ecuadorian populations. Study period: May 2011 to October 2011 Unit of allocation or exposure: neighbourhoods and clusters (geographical units within neighbourhoods)
Participants	Baseline characteristics Control Age: mean, years: 41.71. Number of children aged 0–5 years, mean: 0.59. Number of children aged 6–15 years, mean: 1.02 Place of residence: 7 urban centres in Carchi and Sucumbíos Sex, %: female 0.26 Ethnicity and language: Colombian, %: 0.37 Occupation: NR Education: had secondary education or higher, %: 0.32 SES: HH size: 4.12. Floor type dirt, %: 0.06. See others in table 1. Social capital: NR Nutritional status: DDI: 17.02. HDDS: 9.11. FCS: 59.05 Morbidities: NR Concomitant or previous care: NR Food transfer Age: mean, years: 41.13. Number of children aged 0–5 years, mean: 0.66. Number of children aged 6–15 years, mean: 0.90 Place of residence: 7 urban centres in Carchi and Sucumbíos Sex, %: female 0.25 Ethnicity and language: Colombian, %: 0.28 Occupation: NR Education: had secondary education or higher, %: 0.35 SES: HH size: 3.91. Floor type dirt, %: 0.04. See others in table 1. Social capital: NR Nutritional status: DDI: 17.44. HDDS: 9.22. FCS: 60.93 Morbidities: NR Concomitant or previous care: NR Cash Age: mean, years: 41.42. Number of children aged 0–5 years, mean: 0.59. Number of children aged 6–15 years, mean: 0.89 Place of residence: 7 urban centres in Carchi and Sucumbíos Sex, %: female 0.28 Ethnicity and language: Colombian, %: 0.24 Occupation: NR Education: had secondary education or higher, %: 0.35 SES: HH size: 3.82. Floor type dirt, %: 0.03. See others in table 1. Social capital: NR Nutritional status: DDI: 17.41. HDDS: 9.23. FCS: 60.00 Morbidities: NR Concomitant or previous care: NR Food voucher Age: mean, years: 42.21. Number of children aged 0–5 years, mean: 0.62. Number of children aged 6–15 years, mean: 0.83 Place of residence: 7 urban centres in Carchi and Sucumbíos Sex, %: female 0.29 Ethnicity and language: Colombian, %: 0.26 Occupation: NR Education: had secondary education or higher, %: 0.38 SES: HH size: 3.75. Floor type dirt, %: 0.04. See others in table 1. Social capital: NR Nutritional status: DDI: 17.28. HDDS: 9.19. FCS: 59.75 Morbidities: NR Concomitant or previous care: NR Overall: NR Inclusion criteria: HHs residing in the selected neighbourhoods with low SES as measured by the PMT. All Colombian and mixed‐nationality HHs. Exclusion criteria: all HHs who reported receiving the government's social safety net transfer programme, the BDH. Pretreatment: higher proportion of Colombian heads in HHs in control vs cash group (P = 0.01). Larger HH size and number of children aged 6–15 years in HHs in control vs voucher group. Across 132 (22 × 6) difference‐in‐means tests between the treatment and control groups, only 4 were statistically different at the 5% level, which revealed that randomisation was, mostly, effective at balancing baseline characteristics. Attrition per relevant group: overall attrition 11.5% (235 HHs did not complete follow‐up survey and an additional 35 HHs did not have complete food consumption data and were excluded from analysis). Attrition rates: 11% in control group, 8% in food group, 9% in cash group and 11% in voucher group. Description of subgroups measured and reported: N/A Total number completed and analysed per relevant group: conducted analysis on the 2087 HHs that were in the baseline and follow‐up surveys and had complete data on food consumption. Number per group NR. Total number enrolled per relevant group: NR Total number randomised per relevant group: in total, 80 neighbourhoods and 145 clusters were randomised into the 4 intervention groups: control, cash, vouchers and food. Total number of HHs randomised = 2357. Number per group NR.
Interventions	Intervention characteristics Food transfer Food access intervention category: food prices Intervention type: conditional food transfer Description: valued according to regional market prices at USD 40 and included rice (24 kg), vegetable oil (4 L), lentils (8 kg) and canned sardines (8 cans of 0.425 kg). Although USD 40 was less than most HH's total monthly food consumption at baseline, the quantity of food received for each item was higher than what the median HH in the sample consumed at baseline, which suggests that for many HHs the items from the food transfer would be extra‐marginal. Nutrition sensitisation was a key component of the programme, aimed at influencing behaviour change and increasing knowledge of recipient HHs, especially in regard to dietary diversity. To ensure a consistent approach to knowledge transfer, a curriculum was developed by WFP to be covered during each monthly training session. Topics included: 1. programme sensitisation and information, 2. family nutrition, 3. food and nutrition for pregnant and lactating women, 4. nutrition for children aged 0–12 months and 5. nutrition for children aged 12–24 months. All participants regardless of transfer modality participated in training, and transfers were conditional on attendance. In addition to monthly meetings, posters and flyers on nutrition were developed and posted at distribution sites, including supermarkets, banks, food warehouses and community centres to further expose participants across all 3 modalities to messaging. Flyers covered topics such as recommended food groups, daily nutritional requirements, proper sanitation and food preparation processes. Duration of intervention period: May 2011 to October 2011 Frequency: monthly food transfer per HH Number of study contacts: 2: March–April 2011 (baseline) and October–November 2011 (follow‐up) Providers: WFP (NPO) Delivery: food transfer was valued according to regional market prices at USD 40 and included rice (24 kg), vegetable oil (4 L), lentils (8 kg) and canned sardines (8 cans of 0.425 kg). Although USD 40 was less than most HH's total monthly food consumption at baseline, the quantity of food received for each item was higher than what the median HH in the sample consumed at baseline, which suggested that for many HHs the items from the food transfer would be extramarginal. Transfers were conditional on attending nutrition sensitisation training. Across all modalities, beneficiaries reported extremely high rates of satisfaction with both the programme and programme transparency, believed that the programme was fair, and reported that programme employees treated them with respect. On average, 99% of beneficiaries reported receiving their transfers in totality and 97% reported that they received all information needed to understand how the programme worked. Across the 3 modalities, a minimum of 88% of beneficiaries stated that they received their scheduled payments on time and that they knew how many transfers they would receive. Knowledge gained from the nutrition sensitisation sessions, as measured by a set of questions at baseline and follow‐up, was also similar across modalities. < 1% of voucher and food beneficiaries reported selling their food or voucher. Food HHs reported that the remainder was saved for later use (29.4%) and shared with others outside the HH (6.8%). Co‐interventions: NR Resource requirements: food transfer was significantly more expensive due to the cost of transport to distribution sites and rental of storage facilities. Repackaging bulk items for distribution was also costly, accounting for approximately 30% of the cost of food distribution. Moreover, costs of food did not tend to decrease with economies of scale because much of the modality‐specific costs were physical resource costs such as transport and re‐packaging. In terms of opportunity costs from time spent travelling to the distribution point and waiting to receive their transfers, food beneficiaries spent on average 93 minute. Economic indicators: costs in per‐transfer terms: cost to provide a food transfer was USD 11.46 (Appendix Table B.8). It was considerably less expensive to provide cash (USD 42.99 per transfer) or vouchers (USD 43.27 per transfer) than food (USD 58.22 per transfer). Food recipients spend slightly more, USD 2.12, as many had to use taxis to carry home the heavy loads of food given at the distribution points. Cash transfer Food access intervention category: increase buying power Intervention type: CCT Description: USD 40 transferred monthly onto preprogrammed debit cards. Cash transfer HHs were able to retrieve the cash any time; however, it had to be taken out in bundles of USD 10. Nutrition sensitisation was a key component of the programme, aimed at influencing behaviour change and increasing knowledge of recipient HHs, especially in regard to dietary diversity. To ensure a consistent approach to knowledge transfer, a curriculum was developed by WFP to be covered during each monthly training session. Topics included: 1. programme sensitisation and information, 2. family nutrition, 3. food and nutrition for pregnant and lactating women, 4. nutrition for children aged 0–12 months and 5. nutrition for children aged 12–24 months. All participants regardless of transfer modality participated in training, and transfers were conditional on attendance. In addition to monthly meetings, posters and flyers on nutrition were developed and posted at distribution sites, including supermarkets, banks, food warehouses and community centres to further expose participants across all 3 modalities to messaging. Flyers covered topics such as recommended food groups, daily nutritional requirements, proper sanitation and food preparation processes. Duration of intervention period: May 2011 to October 2011 Frequency: monthly transfer to preprogrammed debit card. Recipients could withdraw money at any time but only in USD 10 bundles. Number of study contacts: 2: March–April 2011 (baseline) and October–November 2011 (follow‐up) Providers: WFP (NPO) Delivery: transfers were conditional on attending nutrition sensitisation training. Across all modalities, beneficiaries reported extremely high rates of satisfaction with both the programme and programme transparency, believed that the programme was fair and reported that programme employees treated them with respect. On average, 99% of beneficiaries reported receiving their transfers in totality and 97% reported that they received all information needed to understand how the programme worked. Across the 3 modalities, a minimum of 88% of beneficiaries stated that they received their scheduled payments on time and that they knew how many transfers they would receive. Knowledge gained from the nutrition sensitisation sessions, as measured by a set of questions at baseline and follow‐up, was also similar across modalities. Cash HHs reported that the remainder was spent on non‐food expenditures (6.3%), shared with others outside the HH (2.4%) and saved for later use (8.3%). Co‐interventions: NR Resource requirements: principal cost associated with the cash transfer was the production of debit cards. In terms of opportunity costs from time spent travelling to the distribution point and waiting to receive their transfers, cash recipients spent 45 minutes travelling and waiting. Economic indicators: costs in per‐transfer terms: the cost to provide cash transfer, USD 2.99 (Appendix Table B.8). It is considerably less expensive to provide cash (USD 42.99 per transfer) or vouchers (USD 43.27 per transfer) than food (USD 58.22 per transfer). Cash and recipients spend an average of USD 1.46 per month on transportation and other out‐of‐pocket expenses to retrieve transfers. Food voucher Food access intervention category: food prices Intervention type: conditional food voucher Description: food vouchers valued at USD 40 and given in denominations of USD 20, redeemable for a list of nutritionally approved foods at central supermarkets in each urban centre. List consisted of cereals, tubers, fruits, vegetables, legumes, meats, fish, milk products and eggs. Food vouchers could be used over a series of 2 visits per month and had to be redeemed within 30 days of receipt. Vouchers were serialised and printed centrally, and were non‐transferable. Nutrition sensitisation was a key component of the programme, aimed at influencing behaviour change and increasing knowledge of recipient HHs, especially in regard to dietary diversity. To ensure a consistent approach to knowledge transfer, a curriculum was developed by the WFP to be covered during each monthly training session. Topics included: 1. programme sensitisation and information, 2. family nutrition, 3. food and nutrition for pregnant and lactating women, 4. nutrition for children aged 0–12 months and 5. nutrition for children aged 12–24 months. All participants regardless of transfer modality participated in training, and transfers were conditional on attendance. In addition to monthly meetings, posters and flyers on nutrition were developed and posted at distribution sites, including supermarkets, banks, food warehouses and community centres to further expose participants across all 3 modalities to messaging. Flyers covered topics such as recommended food groups, daily nutritional requirements, proper sanitation and food preparation processes. Duration of intervention period: May 2011 to October 2011 Frequency: food vouchers provided monthly, which could be used over a series of 2 visits per month and had to be redeemed within 30 days of initial receipt of voucher Number of study contacts: 2: March–April 2011 (baseline) and October–November 2011 (follow‐up) Providers: WFP (NPO) Delivery: vouchers were serialised and printed centrally, and were non‐transferable. Transfers were conditional on attending nutrition sensitisation training. Beneficiaries were asked about how they used their most recent transfer. Voucher HHs reported using 98.8% on food consumption, compared to 83% for cash HHs and 63.2% for food HHs. < 1% of voucher and food beneficiaries reported selling their food or voucher. Across all modalities, beneficiaries reported extremely high rates of satisfaction with both the programme and programme transparency, believed that the programme was fair and reported that programme employees treated them with respect. On average, 99% of beneficiaries reported receiving their transfers in totality and 97% reported that they received all information needed to understand how the programme worked. Across the 3 modalities, a minimum of 88% of beneficiaries stated that they received their scheduled payments on time and that they knew how many transfers they would receive. Knowledge gained from the nutrition sensitisation sessions, as measured by a set of questions at baseline and follow‐up, was also similar across modalities. Co‐interventions: NR Resource requirements: significant staff costs were associated with supermarket selection and negotiation of contracts, and voucher reconciliation and payment. These staff costs accounted for nearly 90% of the cost of implementing the voucher component of the intervention. In terms of opportunity costs from time spent travelling to the distribution point and waiting to receive their transfers, voucher beneficiaries spent on average 92 minutes. Economic indicators: costs in per‐transfer terms: cost to provide a voucher, USD 3.27 (Appendix Table B.8). It was considerably less expensive to provide cash (USD 42.99 per transfer) or vouchers (USD 43.27 per transfer) than food (USD 58.22 per transfer) Voucher recipients spent an average of USD 1.65 per month on transportation and other out‐of‐pocket expenses to retrieve vouchers. Control: no intervention
Outcomes	Food expenditure per capita per month (log values) Dietary diversity: DDI; HDDS; FCS; proportion with poor food consumption; number of days a HH consumed foods from each individual food group Diet intake: log per capita caloric intake per person per day – total and per food group/item
Identification	Sponsorship source: Government of Spain received through the WFP and funding provided by the CGIAR's Policy, Institutions and Markets research programme to IFPRI. Country: Ecuador Setting: poor neighbourhoods within 7 urban centres in the provinces of Carchi and Sucumbíos Author's name: Melissa Hidrobo Email: m.hidrobo@cgiar.org Declarations of interest: NR 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	Authors stated that neighbourhoods and clusters were randomised to either intervention groups but there was no description of how the random sequence was generated.
Allocation concealment (Selection bias)	High risk	The unit of allocation was neighbourhoods. After the baseline survey there was (quote) "a retargeting process where households who were relatively well off were dropped from the programme. Since there were not enough households in existing barrios to replace those that had been excluded and still reach programme enrolment targets, the decision was made to expand coverage to additional barrios on the outer circle of urban areas. These areas were subsequently re‐randomised into treatment groups according to the approximate percentage lost." This rerandomisation of households was done after neighbourhoods had already been allocated to intervention groups, which could have introduced bias.
Baseline characteristics similar (Selection bias)	Low risk	Most baseline characteristics were comparable. Quote: "Across 132 (22 × 6) difference‐in‐means tests between the treatment and control groups, only four are statistically different at the 5% level, which reveals that randomisation was, for the most part, effective at balancing baseline characteristics." Although authors did not adjust for baseline characteristics in the analyses, they assessed the robustness of estimations in additional analyses, which provided similar effect measures.
Baseline outcome measurements similar (Selection bias)	Low risk	Quote: "… estimate the treatment effect using Analysis of Covariance (ANCOVA) which controls for the lagged outcome variable."
Blinding of participants and personnel (Performance bias)	Low risk	Blinding of participants and personnel was not done. Knowledge of intervention allocation was unlikely to have affected participants' experience of the intervention. Authors reported that most participants used the interventions as they were supposed to.
Blinding of outcome assessment (Detection bias)	High risk	No blinding was possible. Outcomes were based on self‐reports from recipients; if they were not satisfied with intervention received, this could have biased their reporting of food consumed.
Protection against contamination (Performance bias)	Low risk	Allocation was by neighbourhood and cluster and it was unlikely that interventions were implemented in the wrong group.
Incomplete outcome data (Attrition bias)	Low risk	Quote: "Table B.1 in the appendix reveals that across 126 difference in means test for those who attrited, only 3 are significant at the 5% level. Those who left the food and cash arm are significantly younger than those who left the control arm; and those who left the food arm are less likely than the voucher arm to have a dirt floor. However, baseline analysis across treatment and control groups for households that remained in the study (Table 1) reveals that differences in age and dirt floor are not significant; therefore, the bias due to the differential attrition of these variables is likely to be very small."
Selective outcome reporting (Reporting bias)	Unclear risk	All relevant outcomes seemed to be reported but no protocol available.
Other bias	Unclear risk	Misclassification bias: unlikely. Researchers knew who had been allocated to each group. Measurement bias: unclear. Authors reported different valid measures of food security and dietary diversity; however, this is based on information only measured once at baseline and once at follow‐up, which may not be sufficient for representative sample of food consumption. Incorrect analysis: unlikely.