Decision Support for Simulated Emergency Triage by Nurses: A Randomized Crossover Study
Development, Validation, and Human-AI Collaboration Evaluation of a Taiwan Triage and Acuity Scale Decision Support System
This randomized, three-period crossover study will evaluate a Taiwan Triage and Acuity Scale (TTAS) decision support system in a controlled simulation. Thirty nurses with emergency triage training or experience will each assess three different sets of 50 standardized, de-identified emergency cases. The three study conditions are unassisted triage, case-specific TTAS decision support, and a sham assistance interface. The order of conditions will be randomized, with 7 to 14 days between study periods. The main outcomes are agreement between nurses' final TTAS levels and an expert reference standard, active time needed to assess 50 cases, and a prespecified safety-adjusted efficiency score. Additional outcomes include undertriage, overtriage, workload, usability, and responses to misleading advice. The system will not be used for actual patient triage or entered into clinical records.
Study Overview
Status
Conditions
Intervention / Treatment
Detailed Description
The study will develop a traceable TTAS rule knowledge base and evaluate a decision support prototype in simulated emergency triage. A language-processing component identifies candidate complaints, vital signs, and modifiers from standardized case narratives. A deterministic rules engine generates a suggested TTAS level, matched rules, prompts for additional information, and a draft note. Nurses retain final decision authority. Three balanced case sets of 50 de-identified cases will be prepared from retrospective emergency data. At least two experienced emergency physicians or triage nursing experts will independently establish the reference TTAS level; disagreements will be adjudicated by a third blinded expert. Participants will be stratified by clinical experience and site and assigned to one of six condition sequences. Each participant completes all three conditions with different case sets and a 7- to 14-day interval between periods. The sham interface resembles the active interface but provides suggestions that are not based on the individual case, allowing assessment of automation bias. Analysis will account for study condition, period, sequence, case set, participant experience, case acuity, and repeated observations within participants and cases. This is a simulation study; no decision support output is used in care of real patients.
Study Type
Enrollment (Estimated)
Phase
- Not Applicable
Contacts and Locations
Study Contact
- Name: Yen-Pin Chen, PhD
- Phone Number: +886 978732756
- Email: f06945029@g.ntuh.edu.tw
Participation Criteria
Eligibility Criteria
Ages Eligible for Study
Accepts Healthy Volunteers
Genders Eligible for Study
Description
Inclusion Criteria:
- Age 18 years or older.
- Able to read Chinese and use the study computer platform.
- Prior emergency triage training or practical emergency triage experience.
- Willing to provide written informed consent and complete the planned case assessments.
Exclusion Criteria:
- Unable to provide valid informed consent.
- Unable to perform basic operations on the study platform or to follow the study schedule.
- Involved in developing the study system or aware of details distinguishing active from sham assistance in a way that could compromise masking.
Study Plan
How is the study designed?
Design Details
- Primary Purpose: Health Services Research
- Allocation: Randomized
- Interventional Model: Crossover Assignment
- Masking: Single
Arms and Interventions
Participant Group / Arm |
Intervention / Treatment |
|---|---|
|
No Intervention: Sequence A[BC]
Period 1: unassisted triage (A); and then Period 2: active TTAS decision support (B); Period 3: sham assistance (C). Each period includes 50 different standardized cases. Period 2: sham assistance (C); Period 3: active TTAS decision support (B). Each period includes 50 different standardized cases. |
|
|
Active Comparator: Sequence B[AC]
Period 1: active TTAS decision support (B); and then Period 2: unassisted triage (A); Period 3: sham assistance (C). Each period includes 50 different standardized cases. Period 2: sham assistance (C); Period 3: unassisted triage (A). Each period includes 50 different standardized cases. |
During review of 50 standardized cases, a research-only interface displays candidate presenting complaints, matched TTAS rules, a suggested triage level, prompts for missing information, and a draft triage note. The nurse makes the final decision. |
|
Placebo Comparator: Sequence C[AB]
Period 1: sham assistance (C); and then Period 2: unassisted triage (A); Period 3: active TTAS decision support (B). Each period includes 50 different standardized cases. Period 2: active TTAS decision support (B); Period 3: unassisted triage (A). Each period includes 50 different standardized cases. |
During review of 50 standardized cases, an interface resembling the active system displays non-case-specific or distribution-matched suggestions. These suggestions are for simulation only and are not used in patient care. |
What is the study measuring?
Primary Outcome Measures
Outcome Measure |
Measure Description |
Time Frame |
|---|---|---|
|
Exact agreement with expert TTAS reference standard
Time Frame: During each of three approximately 60-minute study periods, separated by 7-14 days; up to approximately 4 weeks per participant.
|
Percentage of completed case assessments for which the nurse's final five-level TTAS assignment exactly matches the adjudicated expert reference level, evaluated separately under each condition. |
During each of three approximately 60-minute study periods, separated by 7-14 days; up to approximately 4 weeks per participant.
|
|
Active time to complete 50 case assessments
Time Frame: During each of three approximately 60-minute study periods, separated by 7-14 days; up to approximately 4 weeks per participant.
|
Total active time in minutes required to complete the 50-case set in each condition, excluding recorded breaks and system interruptions. The analysis of incomplete sets must be specified before enrollment. |
During each of three approximately 60-minute study periods, separated by 7-14 days; up to approximately 4 weeks per participant.
|
|
Safety-Adjusted Efficiency Score
Time Frame: During each of three approximately 60-minute study periods, separated by 7-14 days; up to approximately 4 weeks per participant.
|
Prespecified score combining TTAS accuracy and speed. The approved protocol assigns 1.0 for exact agreement, 0.8 or 0.6 for overtriage by one or two levels, 0.4 for undertriage by one level, and 0 for undertriage by two or more levels, with additional penalties for undertriage of Level 1-2 cases. The exact high-acuity penalty, time component, aggregation, score range, and handling of missing cases require final specification before submission. |
During each of three approximately 60-minute study periods, separated by 7-14 days; up to approximately 4 weeks per participant.
|
Secondary Outcome Measures
Outcome Measure |
Measure Description |
Time Frame |
|---|---|---|
|
Sensitivity for high-acuity cases (TTAS Levels 1-2)
Time Frame: During each of three approximately 60-minute study periods, separated by 7-14 days; up to approximately 4 weeks per participant.
|
Proportion of expert-reference Level 1-2 cases assigned Level 1 or 2 by the nurse under each condition. |
During each of three approximately 60-minute study periods, separated by 7-14 days; up to approximately 4 weeks per participant.
|
|
Undertriage rate
Time Frame: During each of three approximately 60-minute study periods, separated by 7-14 days; up to approximately 4 weeks per participant.
|
Percentage of completed case assessments in which the nurse's final TTAS level indicates lower urgency than the expert reference standard; severe undertriage will be reported separately according to the prespecified threshold. |
During each of three approximately 60-minute study periods, separated by 7-14 days; up to approximately 4 weeks per participant.
|
|
Overtriage rate
Time Frame: During each of three approximately 60-minute study periods, separated by 7-14 days; up to approximately 4 weeks per participant.
|
Percentage of completed case assessments in which the nurse's final TTAS level indicates higher urgency than the expert reference standard. |
During each of three approximately 60-minute study periods, separated by 7-14 days; up to approximately 4 weeks per participant.
|
|
Acceptance of incorrect suggestions in the sham condition
Time Frame: During each of three approximately 60-minute study periods, separated by 7-14 days; up to approximately 4 weeks per participant.
|
Percentage of sham-assisted cases in which an incorrect displayed recommendation is accepted in the nurse's final TTAS decision; denominator and incorrectness criteria to be prespecified. |
During each of three approximately 60-minute study periods, separated by 7-14 days; up to approximately 4 weeks per participant.
|
Collaborators and Investigators
Study record dates
Study Major Dates
Study Start (Estimated)
Primary Completion (Estimated)
Study Completion (Estimated)
Study Registration Dates
First Submitted
First Submitted That Met QC Criteria
First Posted (Actual)
Study Record Updates
Last Update Posted (Actual)
Last Update Submitted That Met QC Criteria
Last Verified
More Information
Terms related to this study
Keywords
Other Study ID Numbers
- 202608042RIND
Plan for Individual participant data (IPD)
Plan to Share Individual Participant Data (IPD)?
Drug and device information, study documents
Studies a U.S. FDA-regulated drug product
Studies a U.S. FDA-regulated device product
This information was retrieved directly from the website clinicaltrials.gov without any changes. If you have any requests to change, remove or update your study details, please contact register@clinicaltrials.gov. As soon as a change is implemented on clinicaltrials.gov, this will be updated automatically on our website as well.