Nida Shahid 1,2 (PhD student), Tim Rappon 1(MD/PhD student), and Whitney Berta 1 (PhD, MBA, BSc)
1 Institute of Health, Policy, Management and Evaluation, University of Toronto, Dalla Lana School of Public Health
2 Toronto Health Economics and Technology Assessment (THETA) Collaborative
Background
Health care organizations are leveraging machine-learning techniques, such as artificial neural networks (ANN), to improve delivery of care at a reduced cost. Applications of ANN to diagnosis are well-known; however, ANN are increasingly used to inform health care management decisions. We provide a seminal overview of the applications of ANN to health care organizational decision-making.
Methods
We used the Arksey & O’Malley (2005) framework for conducting a scoping review and screened 3,397 articles from six databases related to Health Administration, Computer Science and Business Administration. Using an iteratively built inclusion criteria, we extracted information related to study characteristics, aim, methodology and context (including level of analysis).
Findings
A total of 80 articles were used in the study. Articles were published from 1997-2018 and originated from 24 countries, with a majority of papers (N=26) published by authors from the United States. Types of ANN used included generic ANN (N=36), feed-forward networks (N=25), or hybrid models (N=23) and the reported accuracy varied from 50% to 100%. We found majority of ANN informed decision-making at the micro level (N=61), between patients and health care providers, in comparison to meso-level (intra-organizational, N=29) and macro- level (inter-organizational or policy, N=10) decision-making.
Conclusion
Our review identifies key characteristics and drivers for market uptake of ANN for health care organizational decision-making to guide further adoption of this technique.