Pathways to Infant Mortality in Urban Slums of Delhi, India: Implications for Improving the Quality of Community- and Hospital-Based Programmes
Bhandari, Nita, Bahl, Rajiv, Taneja, Sunita, Martines, Jose, Bhan, Maharaj K., Journal of Health Population and Nutrition
Mortality rates among children aged less than five years in developing countries have declined over the past years consequent to improving socioeconomic conditions and child-survival programmes (1). Infant mortality has also declined in most settings; the challenge now is how to reduce the current rates further. One of the ways could be to obtain greater insights into the processes underlying infant deaths, beginning from recognition of illness and its severity by the family, care-seeking practices, and the quality of care received when it is sought. This knowledge will help improve the impact of current health programmes.
We present results of an analysis of infant deaths in a birth cohort of 4,000 newborns enrolled in a multicentre randomized controlled trial that evaluated the safety and benefits of vitamin A administration to mothers within four weeks of delivery and to their infants at each polio and DPT immunization (2). In this trial, infants were randomly allocated to receive vitamin A or placebo at three weeks of age and followed up at monthly intervals until their first birthday. The point and period prevalence of common childhood illnesses were ascertained, and deaths occurring during infancy were identified. There was a modest benefit of the intervention on vitamin A status at six months but without any impact on morbidity or mortality. As the surveillance for morbidity was largely passive, the study offered a unique opportunity for indepth analysis of causes and underlying processes of deaths in depth.
MATERIALS AND METHODS
The study was conducted in two urban slums in Delhi, India, between February 1995 and August 1996. Residents here usually live in single-room hutments. About half of the families are nuclear. Four-fifths of women and two-fifths of men have never been to school. Malnutrition among children aged less than three years is widely prevalent. About 42% of them are stunted (height-for-age z-score ≤-2), and 17% are wasted (weight-for-height z-score ≤-2) at the age of 12 months (3,4). These children suffer from 6-8 episodes of diarrhoea and 1-2 episode(s) of acute lower respiratory infections per year (3,4).
This population is served by a pluralistic healthcare system that includes traditional and modern healthcare providers (5). The providers include physicians trained in the biomedical system of medicine (30%), in indigenous systems of medicine, primarily Ayurveda (20%), and registered medical practitioners or RMPs (30%). The government based on their work experience certifies the latter category, but they usually lack formal training. The remaining are either nurses or community health workers who have set up a practice. Within and around the study area, about 50 such healthcare providers attend to clients in small clinics or larger nursing homes; the latter also offer inpatient facilities. Outpatient services are provided by two clinics run by non-government organizations and one government clinic. Some families also visit faith healers who are usually based in places of worship. In this paper, we refer to public hospitals as source of care as those where medical care is governmentsponsored and is provided almost free of charge to patients. On the other hand, patients seeking care from private sources need to make out-of-pocket expenses for care.
Enrollment of birth cohort and identification of deaths
Pregnancies and new births were identified through a doorto- door survey covering the entire community of 125,000 inhabitants, and pregnant women were followed up weekly until delivery. Infants were enrolled in the intervention trial at three weeks of age. Whenever an identified newborn did not turn up at the clinic for enrollment, a home-visit was made to identify the live/dead status. After the initial nine months of the study and until the 18th month, i.e. the end of enrollment, an informant-based system was used for identifying newborns or infants aged less than three weeks. …