Project Details

Emollient Therapy for Improved Survival and Growth of Very Low Birth Weight Infants in Zimbabwe

Gary Darmstadt, MD, MS

Summary

Medical problem as it relates to children: Up to 30% of neonatal deaths occur in very low birthweight [VLBW, <1500 grams (g)] infants who have a poorly developed and dysfunctional skin barrier, which puts them at risk for loss of water and heat and penetration of pathogens into the bloodstream through the skin, leading to poor growth and mortality. Incidence/prevalence of problem in children: Mortality among VLBW infants is 50% at Zimbabwe’s top public neonatal care unit at Sally Mugabe Central Hospital (SMCH), which is representative of sub-Saharan Africa. Background that will lead to the research gap: Oil massage of newborns is a widespread practice throughout Africa, Asia and the Mediterranean region, but studies of locally available products and oils routinely applied to newborn infants in high-mortality regions show harmful effects. In LMICs, emollient therapy with sunflower seed oil (SSO) in VLBW infants has been shown to improve skin barrier function, reduce the risk for bloodstream proven infections, and enhance growth during the neonatal period. Primary research gap this study will address. Data on the impact of emollient therapy on neonatal mortality is limited – especially from Africa – but suggestive of a 25% reduction. Additional data is needed to inform the development of global guidelines by the World Health Organization (WHO) on emollient therapy for hospitalized VLBW infants. Hypotheses: Hypothesis 1: Topical emollient therapy with SSO will reduce mortality by 25% compared to standard of care among VLBW infants 700-1500 g at SMCH, Harare, Zimbabwe. Hypothesis 2. Topical emollient therapy will improve skin condition compared to standard of care among VLBW infants at SMCH. Hypothesis 3: Topical emollient therapy will increase growth compared to standard of care among VLBW infants at SMCH. Hypothesis 4: There will be no strata-specific emollient treatment effect compared to the control for infants who are 700-1000 g, or 1000-1500 g. Hypothesis 5. Emollient therapy will improve skin and gut microbiome diversity compared to standard of care among VLBW infants at SMCH. Hypothesis 6: Improved weight gain during emollient therapy vs. standard of care will be linked to improved survival. Hypothesis 7. Mothers of infants who participate in emollient applications to their infants will have reduced rates of depression and anxiety and increased self-efficacy in care of their infants compared to mothers of infants in the control group. Specific Aims: Aim 1: Compare mortality rates for VLBW infants treated with emollient vs. control while hospitalized for up to 28 completed days. Aim 2: Compare growth rates for VLBW infants treated with emollient vs. control while hospitalized for up to 28 completed days. Aim 3. Estimate a strata specific treatment effect, comparing mortality rates in infants 700-1000 g or 1000-1500 g treated with emollient vs. control. Aim 4. Compare microbiome diversity on the skin and in the gut for VLBW infants treated with emollient vs. control while hospitalized for up to 28 completed days. Aim 5. Model the causal connection of neonatal weight gain with mortality. Study Design: We will individually randomize infants who are <72 hours of age, weigh 700-1500 at birth, and are hospitalized at SMCH to either treatment with emollient or to the control group which consists of standard of care (without application of products to the skin). Maternal demographic data and health history will be recorded at enrollment. Infants in the intervention arm will receive gentle applications of 3 grams of SSO per kg body weight twice daily until 28 completed days after birth or until discharge from the hospital. The impact of emollient treatment on the primary outcome of rates of death while hospitalized up to 28 completed days will be evaluated according to group assignment using a logistic model with fixed effects for study arm and strata random effects. If the randomization is unable to achieve good balance for a specific covariate (i.e., absolute standardized mean difference of < 0.20) then we will add an exploratory analysis which adjusts for the baseline covariates (e.g., sex, birth weight, gravidity, maternal age, HIV, multiple births) of the infants which were imbalanced. Infants will be weighed five times per week (Monday-Friday) and secondary analysis on differences in growth will use a linear model with fixed effects for the study arm and strata. Exploratory analysis will estimate a strata-specific treatment effect using a logistic model with fixed effects for study arm and strata, and an interaction term between study arm and strata. Skin swabs will be obtained from two skin sites (e.g., mid-abdomen, volar forearm) and from stool at 3-6 time points (e.g., baseline, day 7, and weekly until day 28 or discharge) and will be examined using 16S rRNA gene amplicon sequence data and ecological statistics.