Article type
Abstract
"Background: Nowadays, clinical practice guidelines are used increasingly across many therapeutic areas so that evidence-based best practice becomes standard practice. Clinical practice guidelines can improve the consistency of medical staff's treatment behaviors by providing clear recommendations, thereby increasing the likelihood that patients in differently developed regions will have access to the same quality of care. Good guideline adherence is conducive to reducing patient mortality, improving their prognosis, and improving their quality of life. However, the difficulty and complexity of guideline implementation is much greater than guideline development, and improving guideline adherence faces many challenges. Artificial intelligence (AI) contributes to evaluate and improve adherence to clinical practice guidelines.
Objective: The aim of the study is to summarize the current application status of AI in clinical practice guideline compliance and analyze the effects of various AI tools.
Methods: Web of Science, PubMed, Embase, CNKI, and CBM were systematically searched without language restriction. All included studies were selected manually and coded by two reviewers independently, and any conflicts was consulted by a third reviewer.
Results: 3 597 studies were identified. From those studies we selected 73 for text review from which 19 articles fulfilled the inclusion criteria. The findings were as follows: (1) The guidelines were published in 1994~2022, and covered mainly genitourinary diseases (n=5, 26.3%) and respiratory diseases (n=4, 21.1%); (2) Two groups of people are included, healthcare providers (57.8%) and healthcare consumers (42.1%); (3) The most commonly used AI tool was clinical decision support system (n=5, 26.3%). Additionally, the two main purposes of AI tools are assessing the impact on adherence to clinical guidelines (73.7%) and predicting / evaluating factors influencing adherence to clinical guidelines (26.3%).
Conclusion: Artificial intelligence can provide assistance in assessing clinical guideline adherence, predicting adherence influencing factors, and more. In the future, it is recommended to develop AI tools specifically for clinical guideline adherence to improve clinical guideline implementation.
"
Objective: The aim of the study is to summarize the current application status of AI in clinical practice guideline compliance and analyze the effects of various AI tools.
Methods: Web of Science, PubMed, Embase, CNKI, and CBM were systematically searched without language restriction. All included studies were selected manually and coded by two reviewers independently, and any conflicts was consulted by a third reviewer.
Results: 3 597 studies were identified. From those studies we selected 73 for text review from which 19 articles fulfilled the inclusion criteria. The findings were as follows: (1) The guidelines were published in 1994~2022, and covered mainly genitourinary diseases (n=5, 26.3%) and respiratory diseases (n=4, 21.1%); (2) Two groups of people are included, healthcare providers (57.8%) and healthcare consumers (42.1%); (3) The most commonly used AI tool was clinical decision support system (n=5, 26.3%). Additionally, the two main purposes of AI tools are assessing the impact on adherence to clinical guidelines (73.7%) and predicting / evaluating factors influencing adherence to clinical guidelines (26.3%).
Conclusion: Artificial intelligence can provide assistance in assessing clinical guideline adherence, predicting adherence influencing factors, and more. In the future, it is recommended to develop AI tools specifically for clinical guideline adherence to improve clinical guideline implementation.
"