JMIR Publications

ABSTRACT

Background:

Artificial intelligence (AI) provides opportunities to identify the health risks of patients and thus influence patient safety outcomes.

Objective:

The purpose of this systematic literature review is to identify and analyze quantitative studies utilizing or integrating AI to address and report clinical level patient safety outcomes.

Methods:

We restricted our search to PubMed, PMC, and Web of Science databases to analyze research articles published in English between 2009 and August 2019. We focused on quantitative studies that reported positive, negative, or intermediate changes in patient safety outcomes using AI applications specifically machine learning algorithms and natural language processing. Quantitative research reporting only AI performance but not its influence on patient safety outcomes were excluded from further review.

Results:

We identified 53 eligible studies, which we summarized with respect to their patient safety sub-categories, the most frequently used AI, and reported performance metrics. Identified safety subcategories are clinical alarms (n=9; mostly decision tree models used), clinical report (n=21; support vector machine used), and drug safety (n=23; mostly decision tree models used).

Conclusions:

The systematic review indicated that AI-enabled decision support systems, when implemented properly, can aid in uplifting patient safety by improving error detection, patient stratification, and drug management. Future work is still needed for robust validation of these systems in prospective and real-world clinical environments to understand how well AI can predict safety outcomes in health care settings.