A minimal neonatal dataset (mND) for low- and middle-income countries as a tool to record, analyse, prevent and follow-up neonatal morbidity and mortality

Abstract

# Background Neonatal mortality accounts for the most significant and today increasing proportion of under-5 mortality, especially in sub-Saharan Africa. The neonatal population is a sharp target for intervention for these 2.5 million annual deaths. The limited availability of quality data on morbidities leading up to this mortality hampers the development and follow-up of effective interventions. For leverage, undoubtedly more detailed and standardized data adapted to low and middle-income countries (LMICs) is urgently needed. # Methods Drawing on existing databases such as the Swiss Neonatal Network and Vermont Oxford Network, 267 clinical, administrative, and structural variables of neonatal health and healthcare services were selected and submitted for ranking to 42 experts through two Delphi rounds. An empirically limited number of variables with the highest ranking for availability and relevance in low and middle-income countries were field-tested in three centres in Burkina Faso during one year for improvement and practicality. # Results We report the database development process according to the Standards for Quality Improvement Reporting Excellence (SQUIRE 2.0) recommendations. The final dataset is composed of 73 clinical and 6 administrative patient variables, and 21 structural healthcare center variables. Two-thirds of clinical variables maintain matching definitions with high-income countries. # Conclusions The developed minimal neonatal dataset is standardized and field-tested for relevance and availability in LMICs allowing south-south and some south-north cross-comparison.