Skip to main content
. 2020 Jul 28;2020(7):CD011504. doi: 10.1002/14651858.CD011504.pub2

Lopez Arana 2016.

Study characteristics
Methods Study design: CBA
How were missing data handled? Children's study (secondary analysis): children with missing covariates at baseline as well as children LTFU were excluded from the analysis. Analysis of differential attrition was done for children LTFU (Lopez‐Arana 2016); Women's study (secondary analysis): ITT analysis was according to HH allocation, but women with missing covariates as well as those LTFU were excluded from the analysis. Analysis of differential attrition was done for women LTFU (Forde 2012).
Randomisation ratio: N/A
Recruitment method: NR
Sample size justification and outcome used: NR
Sampling method: stratified sample. 57 municipalities were randomly selected from 622 municipalities (with < 100,000 inhabitants) implementing FA (classified into 25 strata based on routine government data on region, health/education structure, population). Intervention municipalities were matched with 65 control municipalities from the same stratum. Of the eligible HHs within each municipality, 100 HHs were randomly sampled.
Study aim or objective: to evaluate the impact of the FA programme on under‐ and overnutrition of children as well as the BMI of women from poor HHs in Colombia.
Study period: June 2002–2006.
Unit of allocation or exposure: municipalities
Participants Baseline characteristics
Intervention or exposure

  • Age: HH head, mean, years: TCP HHs 45.44 (SE 13.13); TSP HHs: 44.15 (SE 12.95); children aged < 7 years, mean, n: TCP HHs 1.1 (SE 0.12); TSP HHs 1.25 (SE 1.19)

  • Place of residence: lived in a rural but sparsely populated part of the municipality, mean: TCP HHs 0.47 (SE 0.5); TSP HHs 0.41 (SE 0.49); lived in a rural but populous part of the municipality: TCP HHs 0.08 (SE 0.27); TSP HHs 0.14 (SE 0.34)

  • Sex: female adults, mean, n: TCP HHs 1.38 (SE 0.72); TSP HHs 1.36 (SE 0.7)

  • Ethnicity and language: NR

  • Occupation: NR

  • Education: HH head, TCP HHs, mean: incomplete primary schooling 0.48 (SE 0.50); complete primary schooling 0.15 (SE 0.35); secondary schooling 0.1 (SE 0.3); TSP HHs, mean: incomplete primary schooling 0.45 (SE 0.5); complete primary schooling 0.16 (SE 0.36); secondary schooling 0.09 (SE 0.28)

  • SES: HH members, mean, n: TCP HHs 5.86 (SE 2.35); TSP HHs 6.12 (SE 2.41); owns a house, mean: TCP HHs 0.97 (SE 0.17); TSP HHs 0.97 (SE 0.18); subsidised health insurance, mean: TCP HHs 0.63 (SE 0.48); TSP HHs 0.7 (SE 0.46)

  • Social capital: informally subsidised health insurance, mean: TCP HHs 0.22 (SE 0.42); TSP HHs 0.18 (SE 0.39)

  • Nutritional status: Attanasio 2006: number of different food types consumed during the previous week, mean: TCP HHs 8.6 (SE 0.92); TSP HHs 7.8 (SE 0.14); food consumption per month, mean: TCP HHs COP 317,339.1; TSP HHs COP 301,111.6; food consumption as proportion of HH consumption per month, mean: TCP HHs 0.715, TSP HHs 0.735; all intervention HHs (Lopez‐Arana 2016): HAZ, mean: –1.47 (SD 1.21); stunting, n (%): 391 (30.3); BMIZ, mean: 0.20 (SD 1.0); all intervention HHs (Forde 2012): BMI of women, mean: 25.17 (95% CI 25 to 25.34)

  • Morbidities: NR

  • Concomitant or previous care: children participating in Hogares Comunitarios, n (%): 521/1290 (40.4)


Control

  • Age: HH head, mean, years: 45.53 (SE 13.23); children aged < 7 years, mean, n: 1.12 (SE 1.15)

  • Place of residence: lived in a rural but sparsely populated part of the municipality, mean: 0.35 (SE 0.48); lived in a rural but populous part of the municipality: 0.07 (SE 0.26)

  • Sex: female adults, mean, n: 1.37 (SE 0.74)

  • Ethnicity and language: NR

  • Occupation: NR

  • Education: HH heads, mean: incomplete primary schooling 0.45 (SE 0.5); complete primary schooling 0.14 (SE 0.35); secondary schooling 0.09 (SE 0.29)

  • SES: HH members, mean, n: 6.07 (SE 2.47); owns a house, mean: 0.96 (SE 0.19); subsidised health insurance, mean: 0.7 (SE 0.46)

  • Social capital: informally subsidised health insurance, mean: 0.14 (SE 0.35)

  • Nutritional status: Attanasio 2006: number of different food types consumed during the previous week, mean: NR; food consumption per month, mean: COP 289,527.1; food consumption as proportion of HH consumption per month, mean: 0.71; control HHs (Lopez‐Arana 2016), HAZ, mean –1.42 (SD 1.13); stunting, n (%): 442 (SD 27.9); BMIZ: 0.25 (0.9); control HHs (Forde 2012): BMI of women, mean: 25.43 (95% CI 25.21 to 25.65)

  • Morbidities: NR

  • Concomitant or previous care: children participating in Hogares Comunitarios, n (%): 897/1584 (56.6)


Overall: NR
Inclusion criteria: families living in a municipality where the intervention (FA) was implemented were required to 1. hold a Colombian citizen card, 2. have children aged < 18 years and 3. be classified in the lowest level of the official socioeconomic classification in December 1999.
Exclusion criteria: none reported for IFS report summary (Attanasio 2005) or children's study (secondary analysis) (Lopez‐Arana 2016). Women's study (secondary analysis): women who were underweight (BMI 18.5 kg/m2) at baseline; women who were pregnant or breastfeeding at any point during the study (Forde 2012).
Pretreatment: in the children's study more children in the control group were participating in the Hogares Comunitarios programme. Control HHs were also less likely to have a mother with no education or be in an un urbanised (rural) area. In the women's study women in the treatment group were more likely to be slightly older and participate in community activities. Treatment HHs also had less persons per room, less piped water to the HH, and were less likely to be in an urban location; while control HHs had lower HH wealth. Treatment areas had larger populations, more intervention‐eligible families, slightly higher average HH wealth, and many more banks; while control areas a higher quality of life index, ratio of doctors to population, and proportion of HHs with piped water.
Attrition per relevant group: Attanasio 2006: attrition was 6% overall at the first follow‐up. Lopez‐Arana 2016: children's study (secondary analysis): attrition was 39.2% (833/2123) in the intervention group and 41.8% (1138/2722) in the control group, with no differential LTFU reported between the 2 groups. Forde 2012: women's study (secondary analysis): attrition was 38.8% (785/2023) in the intervention group and 38.0% (512/1347) in the control group with no differential LTFU reported between the 2 groups (Chi2 = 0.21, P = 0.64).
Description of subgroups measured and reported: women's study (secondary analysis): subgroup analysis excluding women in pre‐exposed areas (26 municipalities).
Total number completed and analysed per relevant group: Attanasio 2006: unclear. Forde 2012: women's study (secondary analysis): intervention group: 1238 women from 57 municipalities; control group: 835 women from 65 municipalities. Lopez‐Arana 2016: children's study (secondary analysis): intervention group: 1290 children (aged 7 years) from 31 municipalities; control group: 1584 children (aged 7 years) from 65 municipalities.
Total number enrolled per relevant group: Attanasio 2006: intervention group: 6293 HHs from 57 municipalities (2954 TSP HHs from 31 municipalities and 3339 TCP HHs from 26 municipalities); control group: 4424 HHs from 65 municipalities
Total number randomised per relevant group: N/A
Interventions Intervention characteristics
Intervention or exposure

  • Food access intervention category: increase buying power

  • Intervention type: CCT

  • Description: cash payments to mothers on condition that their children aged < 7 years regularly attended vaccination programmes as well as growth and development check‐ups (COP 40,000) and that their children aged 7–17 years attended school regularly – ≥ 80% of school lessons (additional payments of COP 14,000 per primary school and COP 28,000 per secondary school child).

  • Duration of intervention period: mid‐2002 to early 2006.

  • Frequency: money periodically transferred to the bank account of the beneficiaries

  • Number of study contacts: June–September 2002 (baseline survey); July–November 2003 (first follow‐up survey) and 2005–2006 (second follow‐up survey)

  • Providers: Colombian government through World Bank and Inter‐American Development Bank funding.

  • Delivery: transfer of cash into the HH bank account.

  • Co‐interventions: some children were also participating in the Hogares Comunitarios programme; a childcare supplementary nutrition and psychosocial stimulation programme.

  • Resource requirements: sufficient health and education infrastructure to service conditionalities without causing bottlenecks.

  • Economic indicators: programme paid USD 183,258,944 to mothers between April 2001 and November 2004.


Control: no intervention

  • Co‐interventions: some children were also participating in the Hogares Comunitarios programme; a childcare supplementary nutrition and psychosocial stimulation programme.
Outcomes Diet diversity: DDI
Anthropometry: HAZ, stunting, BMIZ, thinness
Adverse events: overweight, obesity
Identification Sponsorship source: IFS report summary: NR; children's study: (quote) "S. L.‐A. was supported by the European Union Erasmus Mundus Partnerships programme Erasmus‐Colombus (ERACOL) and Fundación para el Futuro de Colombia (COLFUTURO) at Erasmus MC in the Netherlands. M.A. was supported by the European Research Council (ERC) (grant no. 2636840), the National Institute on Ageing (award numbers R01AG040248 and R01AG037398), and the LIFEPATH project funded by the European Union's Horizon2020 research and innovation programme under grant agreement 633666. The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript."; women's study: (quote) "IF is funded by a Medical Research Council Fellowship which mandates open access publishing (grant code G0701333). In 2001, a partnership between the Institute of Fiscal Studies (London, a research institute), Econometria (Bogotá, a research institute) and SEI (Bogotá, a company specialising in the design and collection of social surveys) was commissioned by the Colombian Government to evaluate Familias, after open tendering."
Country: Colombia
Setting: very poor rural and urban HHs
Authors' names: Sandra Lopez‐Arana; Ian Forde; Orazia Attanasio
Email: o.attanasio@ucl.ac.uk; s.lopezarana@erasmusmc.nl; i.forde@ucl.ac.uk
Declarations of interest: Attanasio 2006: No. Forde 2012: Yes. "All authors have completed the Unified Competing Interest form and declare no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous 3 years; no other relationships or activities that could appear to have influenced the submitted work." Lopez‐Arana 2016: Yes. "The authors declare that no conflicts of interest exist."
Study or programme name and acronym: Familias en Acción (FA)
Type of record: Centre for the Evaluation of Development Policies: Institute for Fiscal Studies report summary; journal articles (Attanasio 2006; Forde 2012; Lopez‐Arana 2016)
Notes  
Risk of bias
Bias Authors' judgement Support for judgement
Random sequence generation (Selection bias) High risk Prospective controlled study. No randomisation carried out.
Allocation concealment (Selection bias) High risk Prospective controlled study. No randomisation carried out.
Baseline characteristics similar (Selection bias) Low risk Baseline non‐equivalence was detected for a number of pertinent characteristics: children's participation in the Hogares Comunitarios programme (P = 0.01) in the control group; while control children were significantly less likely to have mothers with no education (P = 0.006) or come from rural areas (P = 0.002). Women in the treatment group were older (P < 0.01), while control women came from significantly less wealthy (P < 0.001) HHs which were more likely to be in semi‐urban areas (P < 0.001). However, the study authors adjusted for covariates at the individual, HH and municipality level in the regression analyses.
Baseline outcome measurements similar (Selection bias) Unclear risk Study authors stated that the higher level of food consumption in TCP HHs was an early effect of the intervention (26 municipalities who received the intervention before the baseline survey was conducted). The true baseline comparability in terms of food consumption across the intervention municipalities was, therefore, unknown (Attanasio 2006).
Blinding of participants and personnel (Performance bias) Low risk No blinding as this was a CBA. Unclear if this lack of blinding and the awareness of follow‐up surveys would have resulted in a temporary performance bias in terms of the volume and quality of food purchased; which may have affected anthropometric outcomes.
Blinding of outcome assessment (Detection bias) Low risk Unclear whether the field workers were blinded but outcomes were objective and unlikely to have been influenced by lack of blinding.
Protection against contamination (Performance bias) Unclear risk Study authors provided no information on the geographical location of intervention vs control municipalities.
Incomplete outcome data (Attrition bias) High risk The secondary analysis by Forde 2012 reported high attrition among women (38.8% vs 38%). Women who were LTFU were older (P = 0.01), with lower formal educational attainment (P = 0.01) and greater parity (P < 0.001) compared to those with complete data. The secondary analysis by Lopez‐Arana 2016 also reported high attrition (39.2% vs 41.8%) in children. Children who were LTFU were older (P < 0.0001), less likely to be overweight (P = 0.02) and had lower BMIZ (P = 0.001).
Selective outcome reporting (Reporting bias) Unclear risk Study protocol N/A.
Other bias Unclear risk Misclassification bias: high. Information on receipt of cash transfer was self‐reported by HHs. Measurement bias: unlikely.