Global neonatal and perinatal mortality: a review and case study for the Loreto Province of Peru

Jamie B Warren,¹ William E Lambert,² Rongwei Fu,² JoDee M Anderson,¹ Alison B Edelman<sup>3

1</sup>Department of Pediatrics, ²Department of Public Health and Preventive Medicine, ³Department of Obstetrics and Gynecology, Oregon Health and Science University, Portland, OR, USA

Background: Millennium Development Goal 4 calls for the reduction of the under-five mortality rate by two-thirds between 1990 and 2015. To reach this goal, neonatal mortality must be decreased. The lack of information on global neonatal and perinatal mortality impedes appropriate implementation of interventions, as vital registration systems are not available for the majority of the world's neonatal deaths. Verbal autopsy (VA) is a tool that has been used to determine cause of death. Recent studies have attempted to standardize and validate the use of this tool in resource-limited areas. The World Health Organization (WHO) International Standard VA Questionnaire was used to conduct a needs assessment in nine rural Peruvian villages. The goal was to determine the neonatal mortality rate (NMR), perinatal mortality rate (PMR), and causes of, and risk factors for, death in these villages.
Methods: Eligible women were interviewed using the WHO International Standard VA Questionnaire or a set of questions based on the WHO VA Questionnaire. NMR and PMR were calculated using a generalized estimating equation model. Three neonatologists independently reviewed VA records to provide cause of death determination. Reviewer agreement was assessed using percent agreement. Fisher's exact test was used to determine risk factors associated with death.
Results: The NMR was 31.4 per 1000 live births and the PMR was 49.7 per 1000 pregnancies. The main contributor to neonatal death was infection (43%). Percent agreement among reviewers was 90.5% and 38.9% for cause of neonatal death and stillbirth, respectively. Risk factors for death were pregnancy with twins (P = 0.001), preterm delivery (P = 0.003), and cesarean section delivery (P = 0.049).
Conclusion: The WHO VA proved useful for NMR and PMR calculation, cause of death determination, and risk factor identification. Information gathered in this needs assessment will allow for the design and implementation of tailored interventions.

Keywords: neonatal mortality, perinatal mortality, verbal autopsy, needs assessment