Unconditional cash transfers for assistance in humanitarian disasters: effect on use of health services and health outcomes in low- and middle-income countries.

Abstract

Unconditional cash transfers (UCTs) are a common social protection intervention that increases income, a key social determinant of health, in disaster contexts in low- and middle-income countries (LMICs). To assess the effects of UCTs in improving health services use, health outcomes, social determinants of health, health care expenditure, and local markets and infrastructure in LMICs. We also compared the relative effectiveness of UCTs delivered in-hand with in-kind transfers, conditional cash transfers, and UCTs paid through other mechanisms. We searched 17 academic databases, including the Cochrane Public Health Group Specialised Register, the Cochrane Database of Systematic Reviews (The Cochrane Library 2014, Issue 7), MEDLINE, and EMBASE between May and July 2014 for any records published up until 4 May 2014. We also searched grey literature databases, organisational websites, reference lists of included records, and academic journals, as well as seeking expert advice. We included randomised and quasi-randomised controlled trials (RCTs), as well as cohort, interrupted time series, and controlled before-and-after studies (CBAs) on UCTs in LMICs. Primary outcomes were the use of health services and health outcomes. Two authors independently screened all potentially relevant records for inclusion criteria, extracted the data, and assessed the included studies' risk of bias. We requested missing information from the study authors. Three studies (one cluster-RCT and two CBAs) comprising a total of 13,885 participants (9640 children and 4245 adults) as well as 1200 households in two LMICs (Nicaragua and Niger) met the inclusion criteria. They examined five UCTs between USD 145 and USD 250 (or more, depending on household characteristics) that were provided by governmental, non-governmental or research organisations during experiments or pilot programmes in response to droughts. Two studies examined the effectiveness of UCTs, and one study examined the relative effectiveness of in-hand UCTs compared with in-kind transfers and UCTs paid via mobile phone. Due to the methodologic limitations of the retrieved records, which carried a high risk of bias and very serious indirectness, we considered the body of evidence to be of very low overall quality and thus very uncertain across all outcomes.Depending on the specific health services use and health outcomes examined, the included studies either reported no evidence that UCTs had impacted the outcome, or they reported that UCTs improved the outcome. No single outcome was reported by more than one study. There was a very small increase in the proportion of children who received vitamin or iron supplements (mean difference (MD) 0.10 standard deviations (SDs), 95% confidence interval (CI) 0.06 to 0.14) and on the child's home environment, as well as clinically meaningful, very large reductions in the chance of child death (hazard ratio (HR) 0.26, 95% CI 0.10 to 0.66) and the incidence of severe acute malnutrition (HR 0.44, 95% CI 0.24 to 0.80). There was also a moderate reduction in the number of days children spent sick in bed (MD - 0.36 SDs, 95% CI - 0.62 to - 0.10). There was no evidence for any effect on the proportion of children receiving deworming drugs, height for age among children, adults' level of depression, or the quality of parenting behaviour. No adverse effects were identified. The included comparisons did not examine several important outcomes, including food security and equity impacts.With regard to the relative effectiveness of UCTs compared with a food transfer providing a relatively high total caloric value, there was no evidence that a UCT had any effect on the chance of child death (HR 2.27, 95% CI 0.69 to 7.44) or severe acute malnutrition (HR 1.15, 95% CI 0.67 to 1.99). A UCT paid in-hand led to a clinically meaningful, moderate increase in the household dietary diversity score, compared with the same UCT paid via mobile phone (difference-in-differences estimator 0.43 scores, 95% CI 0.06 to 0.80), but there was no evidence for an effect on social determinants of health, health service expenditure, or local markets and infrastructure. Additional high-quality evidence (especially RCTs of humanitarian disaster contexts other than droughts) is required to reach clear conclusions regarding the effectiveness and relative effectiveness of UCTs for improving health services use and health outcomes in humanitarian disasters in LMICs.