TITLE:
Indicator Selection for Quality Measurement in Maternal Neonatal and Child Health Services: Application of Random Forest Classifier
AUTHORS:
Sarah Nyanjara, Dina Machuve, Pirkko Nykanen
KEYWORDS:
Indicator Selection, Machine Learning, Quality Measurement, Random Forest, Quality Indicators, Maternal Care Quality, Neonatal Care Quality
JOURNAL NAME:
Journal of Computer and Communications,
Vol.11 No.7,
July
28,
2023
ABSTRACT: Quality of Maternal, Neonatal and Child (MNCH) care is an important aspect in ensuring healthy outcomes and survival of mothers and children. To maintain quality in health services provided, organizations and other stakeholders in maternal and child health recommend regular quality measurement. Quality indicators are the key components in the quality measurement process. However, the literature shows neither an indicator selection process nor a set of quality indicators for quality measurement that is universally accepted. The lack of a universally accepted quality indicator selection process and set of quality indicators results in the establishment of a variety of quality indicator selection processes and several sets of quality indicators whenever the need for quality measurement arises. This adds extra processes that render quality measurement process. This study, therefore, aims to establish a set of quality indicators from a broad set of quality indicators recommended by the World Health Organization (WHO). The study deployed a machine learning technique, specifically a random forest classifier to select important indicators for quality measurement. Twenty-nine indicators were identified as important features and among those, eight indicators namely maternal mortality ratio, still-birth rate, delivery at a health facility, deliveries assisted by skilled attendants, proportional breach delivery, normal delivery rate, born before arrival rate and antenatal care visit coverage were identified to be the most important indicators for quality measurement.