A Critical Review of Multi Criteria Decision Analysis Method for Decision Making and Prediction in Big Data Healthcare Applications

Multi-criteria decision making (MCR) has a rich history, employed since the early 1950s to mathematically model decision problems and identify optimal solutions from a set of alternatives. Recently, MCR has found applications in biomedical engineering and healthcare big data. The healthcare domain, characterized by a multi-stakeholder perspective involving patients, doctors, medical device manufacturers, and insurers, presents complex decision-making challenges. With diverse criteria and extensive healthcare records, it becomes challenging to make fair decisions. This paper outlines the requirements for a fair decision-making algorithm, highlighting the limitations of MCR in meeting fairness criteria. An algorithmic framework incorporating fairness criteria is presented, employing a similarity-based approach for decision-making on Electronic Health Records (EHR) to predict prospective cohort groups for a target patient.