Precision ICU Care: Evaluating an AI-Driven Nursing Decision Support System for Delirium Prevention

NCT07631494 | Not Yet Recruiting

• 1 other identifier: N202502041
• Study Type: interventional
• Target: 188
• Locations: 1 country
• Sites: 1

Brief Summary

Delirium affects up to 83% of mechanically ventilated patients in the Intensive Care Unit (ICU), often leading to longer hospital stays and long-term memory or cognitive problems. While standard care protocols (such as the ABCDEF bundle) exist, they are often difficult to implement fully due to their complexity and the heavy workload on nursing staff. To address these challenges, this study introduces a 'Precision Nursing' approach by integrating Artificial Intelligence (AI) and Virtual Reality (VR). We will implement an AI-driven system to assist nurses in making personalized care decisions more efficiently. Additionally, interactive VR technology will be used to stimulate patients' cognitive function and encourage early mobility. Our goal is to reduce the clinical burden on healthcare providers while significantly improving recovery outcomes for ICU patients.

Conditions

Delirium in the Intensive Care Unit; Sleep Quality

Keywords

Artificial Intelligence; delirium; intensive care unit; Virtual Reality; sleep; machine learning

Outcome Measures

Primary Outcomes (2)

• Incidence of ICU delirium

  The proportion of participants who develop delirium during their ICU stay. Delirium is assessed using the Intensive Care Delirium Screening Checklist (ICDSC). The checklist evaluates 8 clinical signs, with each item scored 0 or 1. A total score of 4 or higher is defined as a "positive" delirium diagnosis.

  From ICU admission to ICU discharge or maximum 28 days, assessed per shift

• Delirium-Free Days

  The number of days alive and free of delirium during ICU stay. Delirium status is assessed daily using the Intensive Care Delirium Screening Checklist (ICDSC). A "delirium-free day" is defined as a calendar day on which the patient is alive and has an ICDSC score of less than 4

  Through Day 28 after ICU admission.

Secondary Outcomes (11)

• Subjective Sleep Quality

  Daily from enrollment until ICU discharge or up to a maximum of 28 days.

• Objective Sleep Quality

  Continuous monitoring every night from enrollment until ICU discharge or up to a maximum of 28 days.

• ICU Mortality

  From ICU admission to ICU discharge (estimated up to 28 days)

• 90-day All-cause Mortality

  90 days after hospital discharge.

• Cumulative Dose of Sedatives and Analgesics

  Daily during the ICU stay, up to 28 days.

Study Arms (2)

Smart Computing Assisted Nursing Decision in Delirium Prevention(SCAN-D) group

EXPERIMENTAL

The algorithm will calculate variable weights and select features contributing to 80% of the risk. These variables will guide the implementation of personalized delirium prevention strategies, with making decision based on nurses' clinical expertise.

Device: MLP-based Clinical Decision Support System (MLP-CDSS); Behavioral: VR-based Cognitive Intervention; Behavioral: VR-assisted Early Mobility Exercise

routine care control group

NO INTERVENTION

The control group will receive routine care, including delirium assessments and sleep environment management.

Interventions

MLP-based Clinical Decision Support System (MLP-CDSS) DEVICE

An AI-integrated system utilizing a Multilayer Perceptron (MLP) model to predict delirium risk and guide targeted interventions. The system continuously analyzes real-time data from Electronic Health Records (EHR), including physiological parameters, medication history, and laboratory results.

Smart Computing Assisted Nursing Decision in Delirium Prevention(SCAN-D) group

VR-based Cognitive Intervention BEHAVIORAL

A virtual reality intervention designed to provide cognitive stimulation and reduce sensory deprivation. Patients engage in interactive games (e.g., traditional-themed tasks such as goldfish scooping or lantern festivals) that require attention, memory, and spatial orientation. These sessions are conducted twice daily for total 30-40 minutes, 3-4 times a week.

Smart Computing Assisted Nursing Decision in Delirium Prevention(SCAN-D) group

VR-assisted Early Mobility Exercise BEHAVIORAL

A VR-based physical activity program focused on promoting upper limb movement and range of motion. The virtual environment encourages patients to perform specific gestures or reaching tasks while in bed or a seated position, aiming to improve mobility and reduce the physical deconditioning common in ICU stays. These sessions are conducted twice daily for total 30-40 minutes, 3-4 times a week.

Smart Computing Assisted Nursing Decision in Delirium Prevention(SCAN-D) group

Eligibility Criteria

• Age: 18 Years+
• Sex: all
• Healthy Volunteers: Yes
• Age Groups: Adult (18-64), Older Adult (65+)

You may qualify if:

• Patients aged 18 years
• High risk of developing delirium are defined as PRE-DELIRIC (PREdiction of DELIRium in ICu patients) with a score of ≥40%
• Sedation assessment form (Richmond Agitation-Sedation). Scale, RASS) is assessed as between +1 and -1
• Patient is clearly aware and able to communicate with Chinese and Taiwanese customers.

You may not qualify if:

• Those who have developed delirium before joining the study (assessed by the Intensive Care Delirium Screening Checklist (ICDSC) \>4 points)
• Expected to stay in the ICU for less than 24 hours
• APACHE II score \> 25 (meaning \> 50% mortality)
• Delirium cannot be evaluated, such as those with severe acute brain injury, audio-visual impairment, speech problems, psychotic disorders, aphasia, or in the entire ICU During this period, patient remained comatose.

Sponsors & Collaborators

• Taipei Medical University: lead

Study Sites (1)

Taipei Medical University, Taipei, 110

Taipei, 110, Taiwan

Location

MeSH Terms

Conditions

Sleep Initiation and Maintenance Disorders; Delirium

Condition Hierarchy (Ancestors)

Sleep Disorders, Intrinsic; Dyssomnias; Sleep Wake Disorders; Nervous System Diseases; Mental Disorders; Confusion; Neurobehavioral Manifestations; Neurologic Manifestations; Signs and Symptoms; Pathological Conditions, Signs and Symptoms; Neurocognitive Disorders

Central Study Contacts

Hsiao-Yean Chiu, Professor

CONTACT

886-2-27361661; hychiu0315@tmu.edu.tw

Study Design

• Study Type: interventional
• Phase: not applicable
• Allocation: RANDOMIZED
• Masking: SINGLE
• Who Masked: OUTCOMES ASSESSOR
• Purpose: PREVENTION
• Intervention Model: PARALLEL
• Sponsor Type: OTHER
• Responsible Party: PRINCIPAL INVESTIGATOR
• PI Title: Principal Investigator

Study Record Dates

• First Submitted: May 6, 2026
• First Posted: June 8, 2026
• Study Start: June 1, 2026
• Primary Completion (Estimated): January 1, 2028
• Study Completion (Estimated): June 1, 2028
• Last Updated: June 8, 2026

Record last verified: 2026-06

Locations

TW (1)