The maternal profile associated with the underlying root cause of neonatal mortality using machine learning and administrative health data

Abstract

AbstractInfant mortality is a reflection of an analysis of biological, socioeconomic, and assistance factors. An analytical analysis of this problem implies the processing of large sets of data from different areas. Data Science approaches have become increasingly widespread to deal with problems that require large datasets to perform deep analysis. Machine learning methods have become popular due to their efficiency and efficacy in discovering knowledge by identifying patterns in feature interactions of large datasets. This work proposes the use of a machine learning approach to evaluate the association between sociodemographic factors and preventable root causes of neonatal mortality. For this, demographic and epidemiological data from Brazilian public health birth and mortality (SINASC and SIM, respectively) information systems were used. Using an unsupervised approach, for instance, the K-Modes clustering algorithm, clusters were created, so we are able to evaluate the socio-demographic profile of each one of the clusters. In this way, it is possible to evaluate the differences between the profiles of each cluster. The profile consists of features such as maternal age, maternal years of schooling, race, number of consultations, type of delivery, public or private hospital, and date of first prenatal consultation. The analysis was performed using data from the period between 2012 and 2018, for the city of São Paulo, one of the richest regions of the country. The data quality for this region is considered to be very high, so there is no need to apply data correction methods. Besides that, the method adapted does not require data assumptions, and it’s suitable for categorical data, which is our case. Considering that this is a data-driven approach, preliminary results indicate that only a few assumptions can be made on profile using these features, although some associations between demographic variables and neonatal mortality by preventable root causes can be identified. We hope to encourage reflection on the newborn in the socioeconomic environment and contribute to public health policies.