Prevention and treatment of childhood malnutrition in rural Malawi: Lungwena nutrition studies

Evaluating the robustness of targeted maximum likelihood estimators via realistic simulations in nutrition intervention trials

Several recently developed methods have the potential to harness machine learning in the pursuit of target quantities inspired by causal inference, including inverse weighting, doubly robust estimating equations and substitution estimators like targeted maximum likelihood estimation. There are even more recent augmentations of these procedures that can increase robustness, by adding a layer of cross-validation (cross-validated targeted maximum likelihood estimation and double machine learning, as applied to substitution and estimating equation approaches, respectively). While these methods have been evaluated individually on simulated and experimental data sets, a comprehensive analysis of their performance across real data based simulations have yet to be conducted. In this work, we benchmark multiple widely used methods for estimation of the average treatment effect using ten different nutrition intervention studies data. A nonparametric regression method, undersmoothed highly adaptive lasso, is used to generate the simulated distribution which preserves important features from the observed data and reproduces a set of true target parameters. For each simulated data, we apply the methods above to estimate the average treatment effects as well as their standard errors and resulting confidence intervals. Based on the analytic results, a general recommendation is put forth for use of the cross-validated variants of both substitution and estimating equation estimators. We conclude that the additional layer of cross-validation helps in avoiding unintentional over-fitting of nuisance parameter functionals and leads to more robust inferences.

The prevalence and risk factors for acute respiratory infections in children aged 0-59 months in rural Malawi: A cross-sectional study

Background

Acute Respiratory Infections (ARI) are a leading cause of childhood mortality and morbidity. Malawi has high childhood mortality but limited data on the prevalence of disease in the community.

Methods

A cross‐sectional study of children aged 0‐59 months. Health passports were examined for ARI diagnoses in the preceding 12 months. Children were physically examined for malnutrition or current ARI.

Results

828 children participated. The annual prevalence of ARI was 32.6% (95% CI 29.3‐36.0%). Having a sibling with ARI (OR 1.44, P = .01), increasing household density (OR 2.17, P = .02) and acute malnutrition (OR 1.69, P = .01) were predictors of infection in the last year. The point prevalence of ARI was 8.3% (95% CI 6.8‐10.4%). Risk factors for current ARI were acute‐on‐chronic malnutrition (OR 3.06, P = .02), increasing household density (OR1.19, P = .05) and having a sibling with ARI (OR 2.30, P = .02).

Conclusion

This study provides novel data on the high prevalence of ARI in Malawi. This baseline data can be used in the monitoring and planning of future interventions in this population.

Severe malnutrition among children under the age of 5 years admitted to a rural district hospital in southern Mozambique

OBJECTIVE

To describe the burden, clinical characteristics and prognostic factors of severe malnutrition in children under the age of 5 years.

DESIGN

Retrospective study of hospital-based data systematically collected from January 2001 to December 2010.

SETTING

Rural Mozambican district hospital.

SUBJECTS

All children aged <5 years admitted with severe malnutrition.

RESULTS

During the 10-year long study surveillance, 274 813 children belonging to Manhiça’s Demographic Surveillance System were seen at out-patient clinics, almost half of whom (47 %) presented with some indication of malnutrition and 6% (17 188/274 813) with severe malnutrition. Of these, only 15% (2522/17 188) were eventually admitted. Case fatality rate of severe malnutrition was 7% (162/2274). Bacteraemia, hypoglycaemia, oral candidiasis, prostration, oedema, pallor and acute diarrhoea were independently associated with an increased risk of in-hospital mortality, while malaria parasitaemia and breast-feeding were independently associated with a lower risk of a poor outcome. Overall minimum communitybased incidence rate was 15 cases per 1000 child-years at risk and children aged 12–23 months had the highest incidence.

CONCLUSIONS

Severe malnutrition among admitted children in this Mozambican setting was common but frequently went undetected, despite being associated with a high risk of death. Measures to improve its recognition by clinicians responsible for the first evaluation of patients at the out-patient level are urgently needed so as to improve their likelihood of survival. Together with this, the rapid management of complications such as hypoglycaemia and concomitant co-infections such as bacteraemia, acute diarrhoea, oral candidiasis and HIV/AIDS may contribute to reverse the intolerable toll that malnutrition poses in the health of children in rural African settings.