Book a call
Book a call

Transforming Heart Failure Care

A case study

Executive Summary

Heart failure (HF) is a major public health concern, affecting nearly a million people in the UK and contributing to high rates of hospital admissions, morbidity, and healthcare costs. Despite well-established clinical guidelines, a large proportion of HF patients remain on suboptimal therapy—often due to system inefficiencies, siloed data, and clinical workforce pressures.

To tackle this, Red Star AI, in collaboration with the University of Dundee, developed a cutting-edge digital platform to proactively identify and optimise treatment for patients with heart failure with reduced ejection fraction (HFrEF). The solution integrates artificial intelligence, real-world data, and streamlined workflows to enhance care quality and outcomes at scale.

The Challenge

  • A significant proportion of HF patients—especially those in the community—are not on optimal treatment.
  • Traditional identification methods rely on manual case reviews, which are time-consuming and often occur only after patient deterioration.
  • Clinicians are burdened with data trawling and administrative work, limiting their ability to intervene early.

The Solution

Red Star’s AI-powered heart failure platform bridges the gap between diagnosis and treatment by:

  • Screening thousands of electronic health records (EHRs) and echocardiography reports using a machine learning algorithm (CNN) to identify patients with HFrEF.
  • Auditing treatment against guideline-based therapies to flag undertreated individuals.
  • Presenting prioritised lists to clinicians to guide treatment updates and streamline workflows.
  • Offering real-time dashboards for treatment planning, outcome monitoring, and quality improvement initiatives.

Pilot Study Highlights

Conducted in partnership with Dundee’s School of Medicine under the leadership of Dr Ify Mordi and Dr Mya Win, the pilot study assessed 2,000 echocardiograms and evaluated treatment gaps.

  • Accuracy of AI Detection:
    • 99% accuracy in identifying HFrEF patients
    • 97% accuracy in detecting RAAS inhibitor use
    • 92% accuracy in detecting beta-blocker use
  • Clinical Assessment:
    • 73 patients assessed
    • 27 opted for treatment optimisation
    • 77% of participants were male
    • Median NT-proBNP: 679 pg/ml
    • Mean KCCQ score: 70 (moderate HF symptom burden)
  • Outcomes at 12 Weeks:
    • 82% of patients showed reduced NT-proBNP levels (a biomarker of HF severity)
    • 67% reported improved quality of life (KCCQ score)

Impact

✅ Proactive Care: Patients are now identified before deterioration and hospital admission.

✅ Bureaucracy-Busting: Red Star eliminates the need for manual searches and enables real-time audits.

✅ Patient-Centred Outcomes: Clear improvements in symptom burden and biomarker levels indicate meaningful health gains.

✅ Scalable & Integrated: The platform is adaptable to national and local NHS workflows, with full integration into EHR systems under development.

Clinician Perspective

“Unfortunately, due to service constraints, sometimes the first opportunity to identify patients who might benefit from more intensive treatment is after their condition has deteriorated. If we could identify such patients at an earlier stage, we might be able to intervene before a deterioration.”
— Dr Ify Mordi, University of Dundee

Conclusion

Red Star AI’s heart failure platform demonstrates that digital transformation in cardiology is not only feasible but essential. By leveraging routinely collected data and AI-driven insights, Red Star is empowering NHS clinicians to deliver faster, more effective, and more personalised care—transforming how we manage heart failure across the UK.