Truace
Subscribe
70/100

Moderate evidence

Cite
More
Export CSVExport JSONEmailLinkedInX

In a Ghana study of 3,800 adults, nurses with an AI stethoscope flagged who needs a heart specialist

Study title: Identifying Patients for Cardiovascular Evaluation Using Artificial Intelligence and Team-Based Care in Ghana: The H-TRIAGE Study

Evidence type
Controlled or validated comparison
Sample
n = 3,800
Sources
1 sourcePeer-reviewed
Published
2 min read
Truvace record
TRV-2026-2465cite or check it

Study title: Identifying Patients for Cardiovascular Evaluation Using Artificial Intelligence and Team-Based Care in Ghana: The H-TRIAGE Study

A prospective study in Ghana tested H-TRIAGE, an index that combines symptoms, vital signs and an AI-enabled digital stethoscope to decide which adults should get cardiovascular evaluation first. Trained frontline nurses and community health workers did the assessments on 3,800 adults between February and July 2026.

Read the study abstract

Background: Cardiovascular diagnostic capacity is limited in many low-resource settings. We evaluated the HeartLINK Triage for Risk-Informed Allocation and Guided Evaluation (H-TRIAGE) index, a prespecified strategy combining clinical information and AI-enabled digital auscultation to prioritize cardiovascular evaluation. Methods: In this prospective study in Ghana (February-July 2026), consecutively enrolled adults underwent a standardized assessment by trained frontline nurses and community health workers. H-T…

The gain

In a prospective study of 3,800 adults in Ghana, a triage score built from symptoms, vital signs and an AI stethoscope, applied by nurses and community health workers, flagged who needed cardiologist-adjudicated evaluation and could defer roughly half of those assessed.

In brief

Among participants, 1,707 (44.9%) were judged by cardiologists to need further testing, specialist evaluation or both. The share needing it rose sharply by risk group: 2.5% in the Low group, 87.9% in Intermediate and 100% in High.

The authors conclude that the approach could focus scarce specialist and diagnostic capacity on patients most likely to need it, deferring roughly half of those assessed. Sensitivity stayed high at a facility in a different district, though the work was done in one country over about six months.

Main points

  1. The study enrolled 3,800 adults: 3,123 in the primary cohort across 5 facilities in 3 districts of the Ashanti Region and 677 in external validation.
  2. Median age was 52 years and 77.7% of participants were women.
  3. Four cardiologists blinded to H-TRIAGE reviewed every participant, and a fifth adjudicated disagreements.
  4. In total, 1,707 participants (44.9%) had an adjudicated indication for further testing, cardiologist-led evaluation or both.
  5. The index assigns Low (0-3), Intermediate (4-7) or High (8 or more) risk using predefined symptom, vital-sign and AI-stethoscope weights.
  6. The proportion with an indication was 2.5% in Low, 87.9% in Intermediate and 100% in High risk.

The full story

Cardiovascular diagnostic capacity is limited in many low-resource settings, which is the problem H-TRIAGE targets. The strategy uses predefined additive weights rather than a trained model fitted to the study data, and was specified before the analysis.

Nurses and community health workers performed standardized assessments, while the reference standard came from four blinded cardiologists, with a fifth resolving disagreements. The reference was a recommendation for more diagnostic testing, cardiologist-led evaluation, or both.

In a modeled allocation analysis, Intermediate- and High-Risk participants counted as prioritized, and the team calculated sensitivity, specificity, positive predictive value and negative predictive value. External validation used a held-out facility in the Eastern Region, and sensitivity remained high there.

Talk it overFour questions for a class, a team or the dinner table

  1. Who gets this benefit, and who was left out of the study or story?

    The gain In a prospective study of 3,800 adults in Ghana, a triage score built from symptoms, vital signs and an AI stethoscope, applied by nurses and community health workers, flagged who needed cardiologist-adjudicated evaluation and could defer roughly half of those assessed.

  2. How do we know? What can this kind of evidence show, and what can’t it?

    The evidence This rests on a controlled comparison with a sample of 3,800, from Circulation: Population Health and Outcomes.

  3. What does it leave open? What else would you want to know before acting on it?

    The record’s own caveat The allocation analysis was modeled rather than a real-world deployment, and external validation covered one held-out facility with 677 participants.

  4. Have you seen AI used this way where you study, work or live? Did it go the same way?

The questions are the same on every record. The details under them are taken from this record as written.

Sources

1 on file
  1. Circulation: Population Health and Outcomesdoi:10.1161/circoutcomes.126.014561
    Peer-reviewed2026-11-06

The debate

More from Health