Macours 2012.

Study characteristics
Methods	Study design: cRCT Study grouping: parallel How were missing data handled? N/A Randomisation ratio: about 1:1 Recruitment method: municipalities selected for their extreme levels of poverty and because they had been affected by a severe drought in previous year. Sample size justification and outcome used: NR Sampling method: in communities randomly selected to participate in the Atención a Crisis programme, the primary child carer (known as the 'titular'), mainly a woman, was invited to a registration assembly where the programme objectives and various components were explained. At the end of assembly, a lottery took place in each community. Participation in the assemblies and lotteries was close to 100%. Based on lottery, all eligible HHs within each community were assigned to 1 of 3 treatments. Study aim or objective: to analyse the impact of a cash transfer programme on early childhood cognitive development and the extent to which changes in child development could be explained solely by the cash component of the Atención a Crisis programme. Study period: Atención a Crisis pilot programme was implemented between November 2005 and December 2006, Baseline data for the evaluation collected in April–May 2005. A first follow‐up survey collected in July–August 2006, 9 months after the HHs had started receiving payments. A second follow‐up survey, covering the same HHs, was collected between August 2008 and May 2009. Unit of allocation or exposure: HHs, stratified by community
Participants	Baseline characteristics Intervention or exposure Age: HH members aged 0–5 years: 1.04; aged 5–14 years: 1.7; aged 15–24 years: 1.17; aged 24–64 years: 1.84; aged > 65 years: 0.13 Place of residence: number of rooms in the house: 1.57 Sex: female children. %: 50; male HH head, %: 85 Ethnicity and language: NR Occupation: NR Education: number of years of education: mother 4.05; father 3.81 SES: NR Social capital: NR Nutritional status: children aged 0–5 years: WAZ –1.06; HAZ –1.27; WHZ –0.18 Morbidities: NR Concomitant or previous care: NR Control Age: HH members aged 0–5 years: 1.06; aged 5–14 years: 1.69; aged 15–24 years: 1.21; aged 24–64 years: 1.88; aged > 65 years: 0.18 Place of residence: rural community in Nicaragua Sex: NR Ethnicity and language: NR Occupation: NR Education: number of years of education: mother 4.21; father 3.88 SES: NR Social capital: NR Nutritional status: children aged 0–5: WAZ –0.88; HAZ –1.08; WHZ –0.16 Morbidities: NR Concomitant or previous care: NR Overall: NR Inclusion criteria: baseline data on HH assets and HH composition were then used to define programme eligibility. The eligibility criteria were determined using the proxy means methodology developed for the RPS and based on the national HH data from 2001 (EMNV). Additional discussions with local leaders from each intervention community were conducted to identify possible exclusion or inclusion errors. Based on the discussions with leaders, 3.7% of all the HHs considered were re‐assigned from non‐eligible to eligible, and 3.7% from eligible to non‐eligible. To avoid any possible selection bias resulting from the re‐assignment by the leaders, the results they presented use eligibility by the proxy means as the ITT (without taking into account the reclassification by the community leaders). Exclusion criteria: NR Pretreatment: children aged 0–5 years in the intervention group at baseline were significantly more underweight and received fewer vitamins and deworming drugs in the previous 6 months when compared to those in the control group. Also, in the control compared to the intervention group, HHs had more members aged ≥ 65 years. Attrition per relevant group: attrition over the study period was minimal, < 1.3% in 2006 and 2.4% in 2008. Attrition is uncorrelated with treatment status, and does not differ across treatment packages. The low attrition rates were a result of repeat visits to recover temporary absence and extensive tracking of migrants. Migrant HHs and children were interviewed and tested in their new locations. Description of subgroups measured and reported: NR Total number completed and analysed per relevant group: overall: 3326 in 2006 and 4245 in 2008. Numbers of participants per group N/A. Total number enrolled per relevant group: NR Total number randomised per relevant group: NR
Interventions	Intervention characteristics Intervention or exposure Food access intervention category: increase buying power Intervention type: CCT Description: programme had 2 objectives. First, to serve as a short‐term safety net by providing cash transfers to reduce the need for adverse coping mechanisms, such as taking children out of school or reductions in food consumption. Second, to promote long‐term upward mobility and poverty reduction by enhancing HHs' asset base and income diversification capacity. There were 3 different intervention groups, all received the same cash transfers (the same amount paid to the child's primary carer every 2 months), but with differing conditionalities or other co‐interventions. Group 1: the conditionality of regular health check‐ups for children aged 0–5 years was not monitored, and thus HHs were not penalised if they did not comply. Group 2: 1 member per HH was offered a scholarship to choose out of a number of vocational training courses at the municipal headquarters. Group 3: HHs were offered a lump sum payment to start a small non‐agricultural activity; the lump sum was conditional on developing a business development plan. Duration of intervention period: Atención a Crisis pilot programme was implemented between November 2005 and December 2006. Frequency: bi‐monthly cash transfers Number of study contacts: 3; baseline, first follow‐up, second follow‐up Providers: Ministry of the Family and programme staff Delivery: cash transfers paid every 2 months. For Group 1 the educational condition was monitored in practice. For Group 3 the lump sum was paid at the end of May and September 2006. The repeated information and communication efforts were delivered by programme staff during enrolment and paydays. Regular meetings were delivered by local programme promoters. Co‐interventions: NR Resource requirements: NR Economic indicators: total transfer of USD 145 during the year of the programme. HHs with children aged 7–15 enrolled in and attending primary school received an additional USD 90 per HH, and an additional USD 25 per child (with all amounts referring to the total transfer received over the year), conditional on school enrolment and attendance. Control: no intervention
Outcomes	Proportion of HH expenditure on food: % of food in total expenditure; % staple/animal protein/fruit and vegetables in total food expenditure Anthropometry: WAZ; HAZ Anxiety and depression: depression score (CES‐D scale); Cognitive function and development: language test score (TVIP score) Morbidity: number of days ill in bed in past month
Identification	Sponsorship source: World Bank support including the Trust Fund for Environmentally and Socially Sustainable Development (TFESSD) made available by the governments of Finland and Norway, the Bank‐Netherlands Partnership Trust Fund Program (BNPP), as well as the Research Committee through a Research Support Budget (RSB) grant. Funding from BASIS was also received under the USAID Agreement No. EDH‐A‐ 00‐06‐0003‐00 awarded to the Assets and Market Access Collaborative Research Support Program. Country: Nicaragua Setting: poor HHs in 6 municipalities of rural Nicaragua Author's name: Karen Macours Email: karen.macours@parisschoolofeconomics.eu Declarations of interest: NR Study or programme name and acronym: Atención a Crisis Type of record: journal article
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (Selection bias)	Low risk	Quote: "These households were allocated one of three different packages through a participatory lottery."
Allocation concealment (Selection bias)	Low risk	Allocation at community level, and it performed on all units at start of study.
Baseline characteristics similar (Selection bias)	Low risk	Some small differences noted at baseline, but regressions were used to adjust for differences between treated and control groups (page 258).
Baseline outcome measurements similar (Selection bias)	Low risk	There were significant differences between the groups at baseline (the WHZ was higher in the control group) but these were controlled for in the analysis.
Blinding of participants and personnel (Performance bias)	Low risk	It is difficult to blind in these types of interventions, but lack of blinding is unlikely to influence participants or personnel behaviour during a trial of this nature.
Blinding of outcome assessment (Detection bias)	High risk	Unclear whether the data collectors were blinded towards the group allocations, which may have influenced the measurement of outcomes across groups as these were self‐reported.
Protection against contamination (Performance bias)	Low risk	Geographical communities were the unit of allocation. Therefore, it is unlikely that there was a spillover effect of the intervention.
Incomplete outcome data (Attrition bias)	Low risk	Quote: "Attrition over the study period was minimal, less than 1.3 percent in 2006 and 2.4 percent in 2008."
Selective outcome reporting (Reporting bias)	Unclear risk	Protocol or trial registration number NR. Specific outcomes were not specified in the article's Methods section.
Other bias	Unclear risk	Misclassification bias: unlikely. Measurement bias: unclear risk. Unclear how food intake data were collected. Incorrect analysis: low risk. SEs adjusted for clustering at the community level.