Healthcare analytics: A techno-functional perspective

The use of data analytics, particularly Artificial Intelligence (AI), has significantly transformed decision-making in healthcare. However, academic scholarship has expressed concerns over the misalignment of research priorities and their impact on data usage and choice of analytics. On the one hand, an inadequate understanding of healthcare themes and algorithmic nuances has posed difficulty for healthcare professionals to evaluate AI solutions and understand their generalizability and potential biases. On the other hand, the abundance of analytical tools, their complexity, and a misalignment between data type and analysis methods have presented major challenges for analysts in selecting a suitable method. As a result, healthcare research has mostly developed in silos. We propose a techno-functional framework that combines the perspectives of analysts and healthcare professionals, identifies key research themes (i.e., healthcare delivery, patient engagement, data management, market design, and policy and governance), and presents critical research questions that can guide future studies. We also propose a schema for segmenting analysis techniques based on data type and analytical complexity. This schema may help analysts select a suitable analysis technique and data to solve a specific problem. Our study aims to improve the understanding of healthcare analytics among researchers, analysts, and healthcare professionals and present a roadmap for future research.