Machine Learning-Based Approaches for Measuring Quality of Maternal, Neonatal and Child Health Services in Developing Countries: A Literature Review

Abstract

High rates of maternal and neonatal mortality in developing countries are frequently linked to poor quality of healthcare provided to pregnant women and children. Quality measurement is one of the recommended step in improving the quality of maternal, neonatal and child health (MNCH) services. An effective quality measurement approach can enhance routine quality measurement and reporting hence improves MNCH quality. This study intended to review quality measurement approaches commonly used in developing countries and explore their effectiveness in quality measurement. Moreover the study highlights the capabilities of machine learning in facilitating effective quality measurement in MNCH. Review of academic journals, books and web pages related to MNCH was done. Knowledge related to quality measurement in MNCH, machine learning in MNCH and machine learning approaches for quality measurement was collected, organised and analysed. The review found that majority of existing quality measurement approaches are manual and paper based hence time consuming and resource inefficient. Furthermore, the use of machine learning in MNCH and machine learning based approaches for quality measurement in healthcare was observed. However, these approaches focus mainly on quality measurement in health services provided via the web pages and for the specific medical conditions. In this study the potential of machine learning in quality measurement has revealed and the proposal to use machine learning was presented. This would result into development of machine learning-based quality measurement approach suitable for resource-constrained countries such as Tanzania.