This JHSPH-ICDDR,B-Kumudini Hospital study is nested within a cluster randomized, controlled intervention trial of a package of preventive and curative maternal-neonatal health care in a population of 300,000 in rural Bangladesh (Etiology, Prevention and Treatment of Neonatal Infections, Mirzapur). The Verbal Autopsy Validation Study will validate an improved, standardized verbal autopsy instrument for diagnosing the common causes of neonatal mortality in a community-based setting that minimizes biases introduced by care-seeking behavior. The study asks the question: What is the sensitivity and specificity of a standardized verbal autopsy algorithm for diagnosis of two major causes of severe neonatal disease (sepsis and asphyxia) in comparison to gold standard diagnoses based on physician and laboratory assessment? The Clinical Algorithm study seeks to assess the validity of neonatal illness recognition and classification according to Integrated Management of Childhood Illness criteria by community health workers compared to a physician gold standard assessment. The study seeks to answer the following questions: - What is the ability of community health workers (CHWs) for recognition and classification of very severe neonatal illness according to standardized clinical algorithm compared to physician gold standard assessment?
- What is the positive and negative predictive value of symptoms and signs of the clinical algorithm both individually and in combination, to predict severe neonatal disease requiring referral for evaluation and treatment at the hospital as defined by physician diagnosis?
Investigators: | 
