Assessing Clinical Impact of AI for Iron Deficiency

Evaluation of the Clinical Impact of Machine Learning-Based Risk Classification Using Blood Analysis on Iron Deficiency Detection

The goal of this clinical trial is to evaluate whether an AI-based risk notification system integrated into routine clinical care can improve the clinical detection of iron deficiency in adult patients attending Internal Medicine, Family Medicine, and Hematology/Oncology clinics at China Medical University Hospital in Taiwan.

The main questions this study aims to answer are:

1. Does displaying AI-generated iron deficiency risk classification to physicians increase the overall detection rate of iron deficiency at the population level?
2. Does the AI-based risk notification influence physicians' diagnostic behavior by increasing the rate at which ferritin testing is ordered specifically for suspected iron deficiency?
3. Among ferritin tests ordered for suspected iron deficiency, does the diagnostic yield (positivity rate) remain appropriate, reflecting efficient use of testing resources?
4. Are the effects of the AI-assisted intervention consistent among patients with anemia and without anemia?

Comparison Groups Researchers will compare clinical encounters in which physicians receive AI-generated iron deficiency risk information (the Prompt Group) with encounters in which physicians receive standard laboratory results without AI risk display (the Control Group). The comparison focuses on differences in iron deficiency detection, ferritin ordering behavior for suspected iron deficiency, and diagnostic yield.

What Participants Will Experience

1. No Additional Procedures:

   As this is a pragmatic study embedded in routine clinical care, participants will not undergo any additional blood draws, invasive procedures, or clinic visits beyond standard care.

2. Routine Care Only:

   Patients attend their scheduled outpatient visits and receive complete blood count (CBC) testing as ordered by their treating physician, independent of study participation.

3. Background Data Integration:

   The AI system operates within the hospital's information system, analyzing routinely collected CBC data after results become available. No additional data entry or action is required from patients.

4. Physician Autonomy Preserved:

The AI provides a non-mandatory risk classification as decision support. For patients identified as high risk, the system may display an informational prompt suggesting consideration of iron-related testing if no recent testing is found. All diagnostic and management decisions remain entirely at the discretion of the treating physician.

Study Overview

• Status: Recruiting
• Conditions: Iron Deficiency; Iron Deficiency Anemias
• Intervention / Treatment: Other: AI Risk Display

Detailed Description

Detailed Description: Machine Learning-Based Risk Notification for Iron Deficiency

1. Study Overview and Design This is a single-center, prospective, pragmatic randomized controlled trial (pRCT) conducted at China Medical University Hospital (CMUH) in Taiwan. The study aims to evaluate the clinical impact of a machine learning-based clinical decision support tool on iron deficiency detection and related diagnostic behavior in a real-world outpatient setting.

   Unlike traditional explanatory trials with restrictive protocols, this pragmatic design integrates the intervention directly into the existing Electronic Health Record (EHR) system.

2. Participant Population

   The study includes adult patients (aged 18 years or older) receiving outpatient care in the following departments at CMUH:

   1. Division of General Internal Medicine
   2. Department of Family Medicine
   3. Division of Hematology and Oncology

   All eligible outpatient encounters in which a routine complete blood count (CBC) test is performed as part of standard clinical care are included. No additional tests or procedures are required for study participation.

3. AI Intervention and Clinical Workflow

   The AI model used in this study estimates the risk of iron deficiency based on routinely available CBC parameters. The randomization and intervention occur automatically within the EHR system after CBC results become available.

   Randomization:

   Following completion of the CBC, each eligible encounter is automatically assigned, according to a predefined randomization mechanism, to either the Prompt Group or the Control Group. Randomization occurs prior to physician review of laboratory results.

   Prompt Group (Intervention):

   Physicians are shown an informational "AI Risk Hint," categorizing the patient as either high risk or low risk for iron deficiency, displayed on the laboratory results interface. For patients classified as high risk, the system automatically checks whether iron-related testing has been performed within the previous 30 days. If no recent testing is identified, an informational prompt suggests consideration of iron-related laboratory evaluation.

   Control Group:

   Physicians review laboratory results according to usual clinical practice, without display of AI-generated risk information or prompts.

   Clinical Independence:

   The AI system functions solely as a decision support tool. It does not generate orders, mandate testing, or recommend specific treatments. All clinical decisions, including whether to order follow-up iron studies (e.g., ferritin or other iron indices), remain entirely at the discretion of the treating physician.

4. Outcome Measures

   4.1 Primary Outcome:

   Overall iron deficiency detection rate, defined as the proportion of patients identified as having iron deficiency among all eligible encounters.

   4.2 Secondary Outcomes: 4.2.1 Iron deficiency detection rate among patients with anemia The proportion of patients with anemia who are subsequently confirmed to have iron deficiency among all anemic patients included in the study.

   4.2.2 Iron deficiency detection rate among patients without anemia The proportion of patients without anemia who are subsequently confirmed to have iron deficiency among all non-anemic patients included in the study.

   4.2.3 Rate of ferritin testing for suspected iron deficiency, defined as the proportion of encounters in which ferritin testing is ordered with the indication "to confirm iron deficiency" within 1 month after the CBC report becomes available.

   4.2.4 Marginal diagnostic yield of ferritin testing attributable to AI-assisted prompting This measure reflects the average additional number of iron deficiency diagnoses obtained per additional ferritin test attributable to AI-assisted decision support, defined as the difference in confirmed diagnoses between the AI-assisted group and the control group divided by the difference in ferritin tests ordered between the two groups.

   4.2.5 Incremental cost-effectiveness of AI-assisted iron deficiency detection This outcome represents the additional cost required to identify one additional case of iron deficiency attributable to AI-assisted decision support.

5. Risk Mitigation and Ethical Considerations This outcome represents the additional cost required to identify one additional case of iron deficiency attributable to AI-assisted decision support, calculated as the difference in ferritin testing costs between the AI-assisted group and the control group divided by the difference in confirmed iron deficiency diagnoses between the two groups.

Data Privacy:

All data processing and analysis are conducted within CMUH's secure internal systems in accordance with institutional policies and local regulatory requirements. No identifiable patient data are transmitted outside the hospital environment.

• Study Type: Interventional
• Enrollment (Estimated): 2196
• Phase: Not Applicable

Contacts and Locations

Study Locations

• Taiwan
  • Taichung, Taiwan, 404
    • Recruiting
    • China Medical University Hospital
    • Contact: Yu-Hsin Chang; Phone Number: 12587 +886 422052121; Email: 028702@tool.caaumed.org.tw

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Adult; Older Adult
• Accepts Healthy Volunteers: No

Description

Inclusion Criteria:

1. Adults aged 18 years or older.
2. Patients attending outpatient clinics of participating departments including 2.1. Internal Medicine, 2.2 Family Medicine 2.3. Hematology/Oncology
3. Completion of a routine complete blood count (CBC) as part of usual clinical care during the outpatient encounter.
4. Availability of the CBC report in the institutional laboratory information system, allowing sufficient data for analysis.

Exclusion Criteria:

1. Encounters with missing or incomplete key identifiers (e.g., patient identification number, ) that prevent determination of exposure status (AI information displayed vs not displayed) or assessment of outcomes within the defined follow-up period.
2. Encounters without a valid department code required by the information system to trigger the randomization mechanism and AI risk display.
3. Encounters with incomplete or missing laboratory data required to activate the AI system.
4. Repeated CBC encounters from the same patient during the study period, if applicable; only the first eligible encounter will be included to avoid duplication.

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Health Services Research
• Allocation: Randomized
• Interventional Model: Parallel Assignment
• Masking: Double

Number of Arms

2

Arms and Interventions

Participant Group / Arm	Intervention / Treatment
No Intervention: Control
Experimental: AI display	Other: AI Risk Display; Participants in this group receive AI-generated information showing their high or low risk of iron deficiency to assist clinical decision-making. For participants identified as high-risk, the system automatically checks for iron-related tests performed in the past 30 days and alerts the physician if no recent tests are found. The final decision to order any tests remains with the physician.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Detection Rate of Iron Deficiency	The primary outcome is the proportion of patients with laboratory-confirmed iron deficiency identified during routine clinical care. Iron deficiency is determined based on routine iron-related laboratory tests, such as ferritin, as ordered by the treating physician according to usual clinical practice. The detection rate is calculated as the number of patients diagnosed with iron deficiency divided by the total number of patients undergoing complete blood count testing during the study period.	Within 1 month after the complete blood count report is available

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Iron Deficiency Detection Rate in Patients With Anemia	The proportion of patients with anemia who are subsequently identified as having iron deficiency based on standard iron-related laboratory tests ordered during routine clinical care.	Within 1 month after the complete blood count report becomes available
Iron Deficiency Detection Rate in Patients Without Anemia	The proportion of patients without anemia who are identified as having iron deficiency based on standard iron-related laboratory tests ordered during routine clinical care.	Within 1 month after the complete blood count report becomes available
Rate of iron-related laboratory testing for suspected iron deficiency	The proportion of eligible encounters in which iron-related laboratory testing is ordered with the indication "to confirm iron deficiency" within 1 month after the CBC report becomes available.	Within 1 month after the complete blood count (CBC) report becomes available
Incremental cost-effectiveness	The incremental cost-effectiveness ratio was defined as the difference in total iron-related laboratory testing costs between the AI-assisted group and the control group, divided by the difference in the number of confirmed iron deficiency cases identified between the two groups.	Within 1 month after the complete blood count report is available
Incremental number of confirmed iron deficiency diagnoses per additional iron-related laboratory test attributable to AI-assisted decision support	The incremental diagnostic yield was defined as the difference in the number of confirmed iron deficiency cases between the AI-assisted group and the control group, divided by the difference in the number of iron-related laboratory tests ordered between the two groups. This measure reflects the average additional number of diagnoses obtained per additional test attributable to AI-assisted risk prompting.	Within 1 month after the complete blood count (CBC) report becomes available

Collaborators and Investigators

• Sponsor: China Medical University Hospital

Study record dates

Study Major Dates

• Study Start (Actual): April 1, 2026
• Primary Completion (Estimated): December 31, 2026
• Study Completion (Estimated): March 1, 2027

Study Registration Dates

• First Submitted: January 30, 2026
• First Submitted That Met QC Criteria: January 30, 2026
• First Posted (Actual): February 6, 2026

Study Record Updates

• Last Update Posted (Actual): May 26, 2026
• Last Update Submitted That Met QC Criteria: May 21, 2026
• Last Verified: May 1, 2026

More Information

Terms related to this study

• Keywords: machine learning; detection rate; iron deficiency; complete blood count; pragmatic randomized trial
• Additional Relevant MeSH Terms: Metabolic Diseases; Hematologic Diseases; Anemia; Iron Metabolism Disorders; Anemia, Hypochromic; Nutritional and Metabolic Diseases; Hemic and Lymphatic Diseases; Iron Deficiencies; Anemia, Iron-Deficiency
• Other Study ID Numbers: CMUH115-REC2-004

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No