AI-Assisted Versus Independent Heart Sound Screening for Congenital Heart Disease in Children

NCT07830251 | Not Yet Recruiting

• 1 other identifier: 2026QHFELL K14/W
• Study Type: interventional
• Target: 10,000
• Locations: 1 country
• Sites: 1

Brief Summary

This is a prospective, school-based, cluster-randomized controlled trial conducted in Qinghai Province, China. The study will evaluate whether artificial intelligence (AI) can help primary care physicians identify congenital heart disease (CHD) during large-scale screening of children and adolescents. Eligible schools, rather than individual participants, will be randomly assigned to one of two screening approaches: independent cardiac auscultation by primary care physicians or AI-assisted cardiac auscultation. In both groups, primary care physicians will perform cardiac auscultation and record heart sounds. In the AI-assisted group, the physicians will also receive AI analysis of the recorded heart sounds to support their final screening decision. Participants will undergo medical history collection, cardiac auscultation, heart-sound recording, and echocardiography. Echocardiography and expert review will serve as the reference standard. The primary outcome is the sensitivity of each screening approach for detecting CHD that may require clinical intervention or specialist management. The study will also assess detection of all CHD, referral completion, confirmed diagnoses after referral, and related screening outcomes.

Conditions

Congenital Heart Disease (CHD); Screening; Artifical Intelligence; Cardiac Auscultation; Primary Care Physician

Outcome Measures

Primary Outcomes (1)

• Sensitivity for Detecting Congenital Heart Disease Requiring Clinical Intervention

  From enrollment to the end of treatment at 6 months

Secondary Outcomes (5)

• Specificity for Detecting Congenital Heart Disease Requiring Clinical Intervention

  From enrollment to the end of treatment at 6 months

• Sensitivity for Detecting All Congenital Heart Disease

  From enrollment to the end of treatment at 6 months

• Positive Predictive Value for Detecting Congenital Heart Disease Requiring Clinical Intervention

  From enrollment to the end of treatment at 6 months

• Referral Completion Rate

  From the initial screening visit through completion of the protocol-specified referral follow-up at 9 months

• Confirmed Diagnosis Rate After Referral

  From the initial screening visit through completion of the protocol-specified referral follow-up at 9 months

Study Arms (2)

Primary care physicians

NO INTERVENTION

AI + Primary care physicians

EXPERIMENTAL

Other: AI-Assisted Cardiac Auscultation

Interventions

AI-Assisted Cardiac Auscultation OTHER

During the school-based screening visit, a trained primary care physician will perform standardized cardiac auscultation and record the participant's heart sounds. The recorded heart sounds will be analyzed by an artificial intelligence-assisted auscultation system, which will provide automated analysis and decision-support feedback to the physician. The physician will consider the AI output together with the clinical examination and make the final screening and referral decision. The AI output is not a definitive diagnosis and will not replace echocardiography or specialist evaluation. No medication, surgery, or other therapeutic treatment will be administered as part of this intervention.

AI + Primary care physicians

Eligibility Criteria

• Age: 0 Years - 18 Years
• Sex: all
• Healthy Volunteers: Yes
• Age Groups: Child (0-17), Adult (18-64)

You may qualify if:

• Children and adolescents aged 0-18 years who are enrolled in a participating school in Qinghai Province and are eligible for the school-based congenital heart disease screening program.
• Written informed consent provided by the parent or legal guardian, or ethics-approved electronic informed consent, with participant assent obtained when applicable.
• Ability to undergo the protocol-specified cardiac auscultation, heart-sound recording, and echocardiographic assessment.
• Participant and/or parent or legal guardian is willing to provide the required medical history and, if referral is recommended, follow-up information.

You may not qualify if:

• Inability to cooperate with cardiac auscultation, heart-sound recording, or echocardiography, when the required procedure cannot be completed after reasonable attempts.
• Known congenital heart disease previously treated with catheter-based intervention or cardiac surgery, when the primary endpoint is not applicable under the protocol.
• Any condition that prevents completion of the required study procedures or prevents obtaining an interpretable protocol-specified echocardiographic assessment.
• Informed consent or assent is not provided, or consent is withdrawn before completion of the study procedures.

Sponsors & Collaborators

• Kun Sun: lead
• Bill and Melinda Gates Foundation: collaborator
• Women and Children Hospital of Qinghai Province: collaborator

Study Sites (1)

Qinghai Provincial Women and Children's Hospital

Qinghai, Qinghai, China

Location

MeSH Terms

Conditions

Heart Defects, Congenital

Condition Hierarchy (Ancestors)

Cardiovascular Abnormalities; Cardiovascular Diseases; Heart Diseases; Congenital Abnormalities; Congenital, Hereditary, and Neonatal Diseases and Abnormalities

Study Officials

• Kun Sun

  Xinhua Hospital, Shanghai Jiao Tong University School of Medicine

  PRINCIPAL INVESTIGATOR

• Xiulian Jiang

  Women and Children Hospital of Qinghai Province

  PRINCIPAL INVESTIGATOR

Central Study Contacts

Liudan Zhao

CONTACT

0086-18368407858; sjdyszld@sjtu.edu.cn

Study Design

• Study Type: interventional
• Phase: not applicable
• Allocation: RANDOMIZED
• Masking: NONE
• Purpose: SCREENING
• Intervention Model: PARALLEL
• Sponsor Type: OTHER
• Responsible Party: SPONSOR INVESTIGATOR
• PI Title: Professor of Department of Pediatric Cardiology

Study Record Dates

• First Submitted: September 12, 2026
• First Posted: September 21, 2026
• Study Start: September 14, 2026
• Primary Completion (Estimated): December 31, 2027
• Study Completion (Estimated): December 31, 2027
• Last Updated: September 21, 2026

Record last verified: 2026-09

Data Sharing

• IPD Sharing: Will not share

Locations

CN (1)