In hospitals in sub-Saharan Africa,heart failure often affects patients in their fifties, with a worse prognosis than in Europe, although there is a lack of means to detect it early. An international team followed nearly 6,000 Kenyan adults in eight care centers between June and December 2024. Their objective: to check if a simple classic electrocardiogram, reread by an algorithm, could track a silent stage, a precursor to heart failure.
Heart failure: difficult prevention in Southern countries
Heart failure is a chronic condition where the heart cannot pump enough blood to meet the body’s needs. This problem, on the rise globally, is particularly alarming in sub-Saharan Africa, where medical resources are limited. Patients there develop this disease at a younger age and with a more unfavorable prognosis, even if comorbidities are less frequent there than elsewhere.
Before the onset of heart failure, many patients experience warning signs, such as left ventricular systolic dysfunction (LVSD), which manifests as the inability of this ventricle to pump blood effectively. Cardiac echocardiography, which uses ultrasound to visualize the heart, is the standard test for diagnosing LVSD and other precursors to heart failure. However, these tests are very expensive and developing countries often lack the equipment and skills to perform them.
To overcome this disparity, the team of Dr. Ambarish Pandey, cardiologist at UT Southwestern Medical Center, explored a more economical technical solution.
How AI-assisted EKG screens for heart failure in Kenya
Using an Artificial Intelligence-assisted electrocardiogram (AI-ECG) involves enhancing a traditional ECG – a test of the heart’s electrical function – with an AI algorithm that looks for signs of left ventricular systolic dysfunction (LVSD) and other precursors to heart failure. ECG-IA has shown promising results in trials conducted in developed countries. However, its evaluation in developing countries remains rare. Researchers at UT Southwestern Medical Center and the Kenya Cardiac Society tested it in a subset of 1,444 patients who also had echocardiography.
Result: he identified a left ventricular systolic dysfunction in 14.1% of people, with a sensitivity of 95.6% and a negative predictive value of 99.1%. A positive test was frequently associated with left ventricular hypertrophy or diastolic dysfunction.
A low-cost test in the face of the lack of ultrasound for heart failure
For Dr. Pandey, “These results support AI-ECG as a practical and scalable screening tool that can effectively identify individuals at risk for heart failure in resource-limited settings where access to echocardiography is constrained, filling a critical gap in global cardiovascular care“.
An AI-assisted ECG costs barely more than a standard tracing, and much less than a cardiac ultrasound. Thanks to a negative predictive value of 99.1%, the tool reassures the vast majority of patients without heavy imaging and reserves ultrasound slots for the 18.3% reported as at risk. For Bernard Samia, president of the Kenya Cardiac Society, this approach paves the way for mass screening of heart failure in fragile health systems.
Researchers now want to test IA‑ECG in other African countries and follow patients to see if this screening translates into more treatments and fewer hospitalizations. This “software” could also be of interest to rural hospitals or even general medicine practices in France, where ECG is common but little used to prevent heart failure.