Determinants of low birth weight: a community based prospective cohort study.

The study aimed at identifying and quantifying determinants of low birth weight (LBW) by following a community based prospective cohort of pregnant women in 45 villages in Pune district. In the 1922 live births born to mothers without a chronic illness, in whom birth weight was available within 24 hours, the cumulative incidence of LBW (< 2500 g) was 29%. The unadjusted relative risks for LBW were significantly higher for lower socio-economic status (RR = 1.71), maternal age less than 20 years (RR = 1.27), primiparity (RR = 1.32), last pregnancy interval less than 6 months (RR = 1.48), non-pregnant weight less than 40 kg (RR = 1.3), height below 145 cm (RR = 1.51), hemoglobin less than 9 g/dl (RR = 1.53) and third trimester bleeding (RR = 1.87). Multivariate logistic regression analysis showed that the adjusted odds ratio for LBW decreased with increasing gestational duration, non-pregnant weight, parity and rising education level of the mother. Socio-economic status, non-pregnant weight, maternal height, and severe anemia in pregnancy had substantial attributable risk per cent for LBW (41.4%, 22.9%, 29.5% and 34.5%, respectively). The findings suggest that selectively targetted interventions such as improving maternal education and nutrition, specifically anemia, wider availability of contraception to delay the first pregnancy and to increase pregnancy intervals may help in identifying and ensuring adequate care for those women at greatest risk of LBW.

Foot tape measure for identification of low birth weight newborns.

The majority of births in rural India take place at home, Logistic constraints make early and reliable identification of low birth weight babies difficult. Using neonatal foot length as a proxy measure for birth weight, we devised a tri colored foot tape intended for use at home by the neonatal caretaker or birth attendant. The tape was field tested in a rural community in the Pune district. Results showed a sensitivity of 68.2% and a predictive value of 45.5% for identifying low birth weight. For very low birth weights (< 1500 g) the sensitivity was 100%, specificity 95.2% and the positive predictive value 60%. Interobserver reliability comparing a trained medical social worker and the household member was high (kappa score of 0.82). If implemented on a larger scale this simple, low cost technology has the potential to significantly enhance the yield of identification of low birth weight babies born at home.