Evaluating the effectiveness of a clinical decision support system (AI-Antidelirium) to improve Nurses' adherence to delirium guidelines in the intensive care unit

Shan Zhang; Shu Ding; Wei Cui; Xiangyu Li; Jun Wei; Ying Wu

doi:10.1016/j.iccn.2024.103933

Evaluating the effectiveness of a clinical decision support system (AI-Antidelirium) to improve Nurses' adherence to delirium guidelines in the intensive care unit

Intensive Crit Care Nurs. 2025 Jan 8:87:103933. doi: 10.1016/j.iccn.2024.103933. Online ahead of print.

Authors

Shan Zhang¹, Shu Ding², Wei Cui³, Xiangyu Li⁴, Jun Wei⁵, Ying Wu⁶

Affiliations

¹ School of Nursing, Capital Medical University, 10 You-an-men Wai Xi-tou-tiao, Feng-tai District Beijing, China. Electronic address: shan1993ccmu@163.com.
² Beijing Chao-Yang Hospital, Capital Medical University, Beijing, CN 100020, China. Electronic address: dingshu@qq.com.
³ School of Nursing, Capital Medical University, 10 You-an-men Wai Xi-tou-tiao, Feng-tai District Beijing, China. Electronic address: cuiweihom@163.com.
⁴ School of Nursing, Capital Medical University, 10 You-an-men Wai Xi-tou-tiao, Feng-tai District Beijing, China. Electronic address: lixiangyu0729@163.com.
⁵ Respiratory Intensive Care Unit, Xuanwu Hospital, Capital Medical University, Chang Chun Street 45, Xi-Cheng District of Beijing, 100053, China.
⁶ School of Nursing, Capital Medical University, 10 You-an-men Wai Xi-tou-tiao, Feng-tai District Beijing, China. Electronic address: helenywu@vip.163.com.

PMID: 39787945
DOI: 10.1016/j.iccn.2024.103933

Abstract

Objectives: To evaluate the impact of Artificial Intelligence Assisted Prevention and Management for Delirium (AI-AntiDelirium) on improving adherence to delirium guidelines among nurses in the intensive care unit (ICU).

Research methodology/design: Between November 2022 and June 2023, A cluster randomized controlled trial was undertaken.

Setting: A total of 38 nurses were enrolled in the interventional arm, whereas 42 nurses were recruited for the control arm in six ICUs across two hospitals in Beijing, comparing nurses' adherence and cognitive load in units that use AI-AntiDelirium or the control group.

Main outcome measures: The AI-AntiDelirium tailored delirium preventive or treated interventions to address patients' specific risk factors. The adherence rate of delirium interventions was the primary endpoint. The other endpoints were adherence to risk factors assessment, ICU delirium assessment, and nurses' cognitive load. The repeated measures analysis of variance was utilized to explore the influence of time, group, and time × group interaction on the repeated measurement variable (e.g., adherence, cognitive load).

Results: A cumulative total of 1040 nurse days were analyzed for this study. The adherence to delirium intervention of nurses in AI-AntiDelirium groups was higher than control units (75 % vs. 58 %, P < 0.01). When compared to control groups, AI-AntiDelirium was found to be significantly effective in both decreasing extraneous cognitive load (P < 0.01) and improving germane cognitive load (P < 0.01).

Conclusions: This study supports the effectiveness of AI-AntiDelirium in enhancing nurses' adherence to evidence-based, individualized delirium intervention and also reducing extraneous cognitive load.

Implications for clinical practice: A nurse-led systemshould be applied by nursing administrators to improve compliance with nursing interventions among ICU nurses.

Keywords: Adherence; Artificial intelligence; Clinical Decision Supporting System; Delirium; Intensive care unit.