Metrics for Identifying Food Security Status and the Population with Potential to Benefit from Nutrition Interventions in the Lives Saved Tool (LiST).

Background: The Lives Saved Tool (LiST) uses the poverty head-count ratio at $1.90/d as a proxy for food security to identify the percentage of the population with the potential to benefit from balanced energy supplementation and complementary feeding (CF) interventions, following the approach used for the Lancet's 2008 series on Maternal and Child Undernutrition. Because much work has been done in the development of food security indicators, a re-evaluation of the use of this indicator was warranted.Objective: The aim was to re-evaluate the use of the poverty head-count ratio at $1.90/d as the food security proxy indicator in LiST.Methods: We carried out a desk review to identify available indicators of food security. We identified 3 indicators and compared them by using scatterplots, Spearman's correlations, and Bland-Altman plot analysis. We generated LiST projections to compare the modeled impact results with the use of the different indicators.Results: There are many food security indicators available, but only 3 additional indicators were identified with the data availability requirements to be used as the food security indicator in LiST. As expected, analyzed food security indicators were significantly positively correlated (P < 0.001), but there was generally poor agreement between them. The disparity between the indicators also increases as the values of the indicators increase. Consequently, the choice of indicator can have a considerable effect on the impact of interventions modeled in LiST, especially in food-insecure contexts.Conclusions: There was no single indicator identified that is ideal for measuring the percentage of the population who is food insecure for LiST. Thus, LiST will use the food security indicators that were used in the meta-analyses that produced the effect estimates. These are the poverty head-count ratio at $1.90/d for CF interventions and the prevalence of a low body mass index in women of reproductive age for balanced energy supplementation interventions.