Understanding newborn care in Uganda : Towards future interventions

Abstract: Background: The highest rates of newborn deaths are in Africa. Existing evidence-based interventions could reduce up to 72% of the 3.8 million newborn deaths which occur every year worldwide, but are yet to be operationalised at scale in sub-Saharan health systems. Aim: To explore community perceptions, determine uptake of evidence-based newborn care practices, and identity delays leading to newborn deaths in Uganda. Methods: Studies were conducted from 2007 to 2009 in Iganga and Mayuge districts in eastern Uganda, and in an embedded Health Demographic Surveillance Site (HDSS) as follows: Qualitative methods with focus group discussions and in-depth interviews (I and IV); a population based cross-sectional study (II) and a case series approach of newborn deaths in the HDSS (III); and a health facility survey (III and IV). A wealth index was generated using principal component analysis of household assets, and was used as a proxy for socio-economic status (II and III). Verbal and social autopsy and a modified maternal mortality delay model were used to code causes and care-seeking delays of newborn deaths (III). Standard descriptive analysis (III) and content analysis were done (I and IV). Newborn care practices were coded as binary composite outcomes (optimal thermal care, good cord care, and good neonatal feeding) and multiple logistic regression analysis was done (II). Results: Most of the evidence-based newborn care practices were acceptable to community members but not promoted by health providers (I and IV). There was poor uptake of newborn care practices among both the poorest and least poor (II). Some practices like putting nothing on the umbilical cord and delaying bathing were less acceptable to caregivers (I). Only 42%, 38%, and 57% of newborns were judged to have had optimal thermal care, good cord care, and good neonatal feeding, respectively (II). Some mothers were putting powder on the cord; using a bottle to feed the baby and mixing/replacing breast milk with various substitutes (I, II and IV). Multiparous mothers were less likely to have safe cord practices (OR 0.5, CI 0.3 0.9), and so were mothers whose labour began at night (OR 0.6, CI 0.4 0.9) (II). 33% of 64 newborn babies had died in a hospital/health centre, 13% in private clinics and 54% died elsewhere (III). The median time to seeking care was 3 days from illness onset (IQR 1-6) (III). Major delays related to deaths of newborn babies were Delay 1 (delay in problem recognition and deciding to seek outside care) (50%) and Delay 3 (delay in receiving treatment at a health facility) (30%) (III). Health facilities lacked equipment, drugs, supplies and protocols for newborn care, and health workers had correct knowledge on only 31% of the survey questions related to newborn care (III). Care practices for preterm babies at home and at health facilities were of poor quality and potentially harmful (IV). Discussion: Implementation of evidence-based newborn care practices needs to be tailored to the local context. In order to reduce newborn deaths, a universal strategy targeting the entire population is needed and should utilise the many missed opportunities in current programmes. Capacity to manage newborns should be built at health facilities, including private clinics and those at the lower level. Community health workers in health facilitylinked preventive and curative newborn programmes may assist in underserved areas.