Exploring updates to the Lives Saved Tool for maternal and child nutrition outcomes
This is a summary of the following paper: Tong H, Piwoz E, Ruel M et al (2022) Maternal and child nutrition in the Lives Saved Tool: Results of a recent update. Journal of Global Health, 12. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9801341/
The Lives Saved Tool (LiST) is a mathematical modelling tool that calculates changes in cause-specific mortality based on intervention coverage change, intervention effectiveness for that cause, and the percentage of cause-specific mortality sensitive to that intervention. LiST assumes that changes in coverage for health and nutrition interventions drive health outcomes. The model relies on determining base year coverage levels of an evidenced nutrition and health intervention, along with risk factors, health status and other pertinent factors for women and children in low- and middle-income countries. The countries are defined by the World Bank and data are compiled from a comprehensive list of high-quality data sources, with most evidence taken from systematic reviews.
LiST is useful as intervention effectiveness estimates focus on low-income countries, so findings are not extrapolated from high-income settings to those with more complex development profiles – a common issue in global health. The model is also reproducible and transparent, as inputs are collaboratively reviewed and modified as needed. LiST can also be applied across a broad array of settings, as activities and many inputs are user-defined.
The approach is limited in humanitarian settings, where data collection and routine monitoring, which LiST relies upon, are often limited. Protracted crises are also difficult to model, as such stable crises may not align with LiST assumptions. It is important to update the model by adding novel interventions when new evidence emerges on efficacy, as well as updating existing estimates for proven interventions. For nutrition modelling, the ‘affected fraction’ – the population benefiting from an intervention (defined as those with a nutrient deficiency or, as a proxy indicator, those residing in an area known to have poor dietary diversity or food insecurity)1 – also requires regular attention to ensure model accuracy.
This study reviewed evidence from systematic reviews on 53 nutrition-related intervention-outcome (I-O) pairs for women and children under the age of five. An example of an intervention-outcome pairing is zinc supplementation in children under the age of five, which is paired with the outcomes of stunting, diarrhoea and pneumonia incidence. An external advisory group decided whether there was sufficient evidence of benefit for particular I-O pairs and how these could best be incorporated into LiST. Of the 53 pairs, 34 were incorporated into the updated model (an increase from 25 prior to review) and included 14 interventions (six for women of reproductive age and pregnant women, and eight for infants and children) and 16 nutrition, disease incidence and cause-specific mortality outcomes. The new set includes nine new I-O pairs, 13 existing links with updated efficacy and/or affected fractions, and 12 existing links with no changes to efficacy or affected fractions. The authors reaffirm the need for continuous updates to LiST in order for it to remain a useful tool for global health application.
1 As many nutrition intervention trials do not screen for deficiency, it is assumed that most of the population (‘affected fraction’) will benefit from a given nutrition-sensitive intervention in areas with high poverty, poor dietary diversity or food insecurity. These socioeconomic measures are therefore often used in place of biomarkers, as they are more practical.