JMIR Publications

ABSTRACT

Background:

The digitization and automation of diagnostics and treatments promise to alter the quality of health care and improve patient outcomes, while undersupply of medical personnel, high workload on medical professionals, and medical case complexity increase. Clinical decision support systems (CDSS) have been proven to help medical professionals in their everyday work through their ability to process vast amounts of patient information, but still suffer from lack of adoption and standardized evaluation. With the rise of artificial intelligence (AI), CDSS became adaptive human-like technologies, able to learn and destined to change their characteristics over time. Yet, research has not reflected on the characteristics and factors essential for effective collaboration between human actors and AI-enabled CDSS.

Objective:

Our study seeks to summarize the factors influencing an effective collaboration between medical professionals and AI-enabled CDSS within a direct, individual interaction. These factors are essential for medical professionals, management, and technology designers to reflect on the adoption, implementation, and development of AI-enabled CDSS.

Methods:

Following the PRISMA guidelines, we conducted a literature review including three different meta-databases, screening over 1000 articles and including 101 of them for full-text assessment. In the end, seven met our inclusion criteria and were analyzed for our synthesis.

Results:

We identified the technological characteristics and human factors that appear to have an essential effect on the collaboration of medical professionals and AI-enabled CDSS in accordance with our research objective, namely training data quality, performance, explainability, adaptability, medical expertise, technological expertise, personality, cognitive biases, and trust. Comparing our results with those from research on non-AI CDSS, some characteristics/factors retain their importance, while others gain or lose relevance due to the uniqueness of human-AI interaction. However, only few studies mention theoretical foundations and patient outcomes related to AI-enabled CDSS.

Conclusions:

Our study provides a comprehensive overview of relevant characteristics and factors that influence the interaction and collaboration of medical professionals and AI-enabled CDSS. Rather limited theoretical foundations are currently hindering the possibility of creating adequate concepts and models to explain and predict interrelations between these characteristics and factors. For an appropriate evaluation of human-AI collaboration, patient outcomes and the role of patients within the decision-making process should be taken into consideration.