The rate of neonatal mortality in Latifabad, Pakistan, is about 47 deaths per 1,000 births, even though most pregnant women receive antenatal care and most births occur in hospitals with trained personnel, according to a population-based study conducted in 2003– 2005.1 The rate of perinatal mortality (stillbirths plus all neonatal deaths within 28 days of delivery) is 83 per 1,000 births, equivalent to one death per 12 births. The odds of such deaths are elevated among infants born before 37 weeks' gestation (odds ratio, 5.0) or delivered by cesarean section (2.3); most deaths are related to immaturity (26%), asphyxia or hypoxia (lack of oxygen; 26%) or infection (23%).
Pakistan ranks third in the world in number of neonatal deaths—about 300,000 annually. The country's neonatal mortality rate is generally estimated to be 45–50 per 1,000 births, but reliable population-based data on the number and causes of such deaths are lacking. The authors of the current study hypothesized that the rate would be lower than those levels in Latifabad (a town in Hyderabad), because women in the area have good access to obstetric care. With the help of government-employed lady health workers, who keep logs of pregnancies and provide basic maternal and child care through home visits, the researchers recruited pregnant women who were 16 or older, had no serious medical conditions, planned to deliver locally and were at 20–26 weeks' gestation. Eighty-three percent of eligible women enrolled in the study; participants provided demographic and health data, and underwent routine antenatal testing. Lady health workers tracked participants until delivery; a physician and nurse visited each woman within 48 hours of delivery to collect information on the delivery and the birth outcome, and made an additional visit about 28 days after delivery to assess longer-term outcomes. Overall, 28-day outcomes were obtained for 1,121 deliveries between September 2003 and August 2005. For all neonatal deaths and stillbirths, a physician and nurse interviewed the mother regarding the circumstances of the delivery, and a neonatologist and the lead investigator reviewed the maternal reports as well as hospital records to determine a likely cause of death.
The researchers examined univariate relationships between maternal characteristics (including age, education level, weight and number of antenatal care visits) and neonatal death, and between clinical and delivery variables (such as gestational age, birth weight, place of delivery, type of birth attendant, type of delivery and amniotic fluid characteristics) and neonatal mortality. Variables associated with neonatal death in univariate analyses were included in a multivariate analysis.
Of the deliveries with complete outcome data, 80% took place in hospitals or other health care facilities, and 19% were done by cesarean section. Most were attended by a doctor (60%) or a nurse or midwife (24%). Nonetheless, 53 resulted in neonatal death within 28 days of delivery—nearly half of them (45%) within 48 hours of delivery and three-quarters (74%) within the first week. In addition, 43 stillbirths occurred. These findings translate to rates of 47 neonatal deaths per 1,000 live births, 34 stillbirths per 1,000 births, and 83 cases of perinatal mortality per 1,000 births—one perinatal death per 12 births.
The vast majority (88%) of the 53 deceased infants had received treatment before death, and 75% died in a hospital. The researchers attributed 26% of the deaths to immaturity-related factors, 26% to lack of oxygen (asphyxia or hypoxia), 23% to infection, 8% to congenital abnormalities and the remaining 17% to other causes. Low birth weight was not considered an independent cause, but about half of neonatal deaths occurred among infants who weighed less than 2,500 g at birth.
No maternal characteristics were associated with neonatal mortality. Six clinical and delivery variables showed associations in univariate analyses, but in the multivariate analysis the only factors linked to mortality were gestational age of less than 37 weeks (relative risk, 5.0) and cesarean delivery (2.3).
The researchers call the 28-day neonatal mortality rate "striking," given that the deaths occurred "in an urban cohort in which a high proportion of births took place in a health facility assisted by skilled attendants, and a high proportion of sick neonates were cared for in the formal health-care system." Indeed, the neonatal mortality rate (47 per 1,000 live births) was similar to most estimates for Pakistan as a whole, leading the researchers to suspect that the country's neonatal mortality rate is actually higher than generally thought. In addition, the high proportion of neonatal deaths that occurred—despite the availability of antenatal, perinatal and postnatal care—suggests that "the quality [of care] may have been suboptimal." Although interventions designed to reduce neonatal mortality often focus on increasing access to care, the researchers emphasize that "without improved quality, increased health-care coverage is unlikely to substantially improve perinatal and neonatal outcomes."—P. Doskoch
1. Jehan I et al., Neonatal mortality, risk factors and causes: a prospective population-based cohort study in urban Pakistan, Bulletin of the World Health Organization, 2009, 87(2):130–138.