Minimal dataset.

Introduction
Newborn deaths are still a concern to global health systems. Increased rates of documented hospital-based births are incongruent to newborn survival chances worldwide. Factors for that rather paradoxical observation are largely unknown. We aimed to assess predictors of early newborn deaths at representative metropolitan referral health facilities in Africa.
Materials & methods
We designed a prospective, analytical, hospital-based study in neonatal units at Dar es Salaam public regional referral hospitals, Tanzania. Neonates who died within the first 7 days of life were the target population. A pre-designed Case Report Form was the main tool for data collection. Multivariable binary logistic regression model was fitted to account for predictors of early newborn deaths after appropriate linear model validation. Proportion of early newborn deaths was the outcome variable.
Results
We recruited and analysed 2212 neonate-days of follow-up. Prevalence of early newborn deaths was 28.1%. Birth asphyxia (χ2 = 20.4, df = 1), preterm delivery (χ2 = 5.36, df = 1) and respiratory distress syndrome (χ2 = 30.94, df = 1) were associated with early neonatal outcomes. Predictors of early newborn deaths were neonatal respiratory rate (Tachypnoea – A.O.R.: 2.28 (95% CI.: 1.44–5.79); Bradypnoea – A.O.R.: 1.9 (95% C.I.: 1.02–12.3) and gestational age (Preterm delivery – A.O.R.: 1.48, 95% CI.: 1.11–2.09 and Post-term delivery – A.O.R.: 5.05, 95% C.I.: 4.49–32.0).
Conclusions
Early newborn deaths rate was relatively high in this study population. Newborns’ respiratory rates and gestational age at delivery were significant clinical factors associated with early newborn deaths in this study.