Seeing the unseen: targeting neonatal mortality in rural Vietnam.

Although there has been significant reduction in under-five mortality in developing countries over the last decade (1), rate of reduction in neonatal deaths has lagged behind reduction in post-neonatal mortality. It is currently estimated that almost 41% of all under-five deaths globally occur in the newborn period (2), the majority within the first few days of life. The importance of the relative resilience of neonatal deaths is underscored by recent findings from Southeast Asia, where despite impressive reductions in child mortality, newborn mortality rates remain high in many regions (3). With emerging data on the high burden of stillbirths globally, the importance of the link between neonatal mortality and intrapartum stillbirths in developing countries is being increasingly recognized (4).

The work by Målqvist (5) published in this issue of Global Health Action represents an important contribution to understanding why neonatal mortality poses enormous challenges to health systems. In a systematic analysis of available health system information and comparison with household level surveys in Quang Ninh province, Vietnam, they highlighted gross under-reporting of neonatal mortality by official government sources. Their findings, for the same level of education, socio-economic indicators, and facility births, ethnic minority women had a twofold higher risk of neonatal death, provide much food for thought. The relative invisibility of neonatal deaths in Vietnam, a country that has made remarkable progress in reducing child mortality underscores the importance of accurate information and health statistics. In the absence of a robust birth (and death) registration system, it is not surprising that many newborn deaths go unreported. The fact that ethnic minority women were at risk of adverse perinatal outcomes despite accessing care in the public health system is probably reflective of a range of issues including higher social risks, poor status of women, and subtle discrimination within the health system. Although traditional measures of inequity deploy socio-economic indicators, commonly used asset indices, and income quintiles (6), there is increasing recognition of the importance of other markers of inequity such as gender (7), religion (8), and race or ethnicity (9). The latter may yet be a major marker of marginalization and social exclusion and in some populations require interventions well beyond the health sector. In a recent survey of infant and young child feeding practices in Sri Lanka (10), it was evident that the worst indicators were seen among the tribal populations. The mechanisms for ethnic clustering of excess neonatal mortality may range across the entire spectrum of care and Målqvist (5) rely upon the recognized conceptual framework of delays by Thaddeus and Maine (11). It is important to recognize that the three delays model is probably an oversimplification and that at each level of care, there may be complex series of factors affecting care. To illustrate, the first delay at household level may be a consequence of inadequate awareness of problems and, hence, delay in recognition of problems in the mother or newborn or a decision on care seeking by the family based on gender, past experience within the health system, poverty and, frequently, a sense of fatalism. Future studies should focus on qualitative methods to understand some of these issues and relevant interventions.

What is the relevance of these findings to action? Clearly, the recognition of at-risk groups and inadequate data collection is the first step toward developing robust interventions and a policy response. The studies by Målqvist et al. (5) underscore the importance of linking socio-behavioral and anthropological research to traditional monitoring and evaluation of health indicators and developing a policy response. Current evidence supports the utilization of innovative strategies to reach the poorest sections of the population ranging from community-based women's groups (12) to deployment of trained community health workers (13), interventions that specifically target the marginalized poor.