A systematic approach to identify health system resilience indicators using artificial neural network algorithm

Abstract

According to the World Health Organization (WHO), a health system comprises all organizations, people, and actions that aim to promote, restore, or maintain health [1]. Health systems perform multiple functions in society not only delivering healthcare services and other interventions aimed at maintaining or improving health. They serve multiple societal functions beyond merely delivering healthcare services, but also adapting to risks, disasters, and unpredictable events to maintain functionality. Health system resilience, therefore, refers to the system's capacity to adapt and return to previous performance levels as quickly as possible after such disruptions. The WHO has created a model called Six Building Blocks to measure the health system's resilience. This study aims to assign indicators obtained from a comprehensive literature review to the blocks established by the World Health. The study uses an Artificial Neural Network (ANN) algorithm to analyze data from 35 Organisation for Economic Co-Operation and Development (OECD) countries from 2020 to 2022. ANN is used to assess the varied relevance or weights of each building block's contribution to health system resilience, providing detailed knowledge of how different components impact overall health system performance during crises. As a consequence of this study, policymakers may identify which healthcare building block and accompanying indicator should be prioritized for investment in preparation for future pandemics. The findings suggest which characteristics are most important for increasing health system resilience during public health emergencies.