Estimating the burden of 'weighing less': A systematic review and meta-analysis of low birth-weight in India

Background. The National Family Health Survey-3 (NFHS-3; 2005-06) reports that the prevalence of low birth-weight babies is 22% in India. This old figure is probably an underestimate as this nationwide survey acquired information on birth-weight of only 34% of babies. We aimed to make a fresh estimate of the proportion of low birth-weight babies. Methods. A systematic search was done through PubMed, Google Scholar, Cochrane Library, Medline, IndMed, Embase, WHO and Biomed Central databases. Studies published from 2004 to 2014 were included. Study quality was assessed using the adapted Mirza and Jenkins checklist. An 'adjustment' of 24% was applied to the published estimates where data were collected through records or through a combination of records and anthropometry. The adjustment was done to account for the heaping of birth-weight data at 2500 g. Meta analysis using both random and fixed effects model was done to derive an estimate. Results. Nineteen studies with 44 133 subjects were included in the review. The pooled estimate for the prevalence of low birth-weight was 27% (95% CI 24%-30%) and the 'adjusted' pooled prevalence was 31% (95% CI 28%-33%). The prevalence in urban and rural areas was 30% (95% CI 23%-38%) and 26% (95% Cl 22%-30%), respectively. Region wise estimates revealed that the prevalence at 33% was comparatively higher in eastern regions (95% CI 29%-37%). Conclusion. The pooled prevalence of low birth-weight is higher than that reported by NFHS-3. Updated estimates should be used to guide future interventions and policies.