12 March 2026
In a large international collaboration, Human Technopole has contributed to the development of three complementary risk prediction models that estimate heart failure risk across the entire cardiovascular disease spectrum. Together, these tools enable transparent, data-driven risk assessment and support personalised treatment strategies and more sustainable healthcare delivery.
Heart failure is one of the most serious and rapidly growing cardiovascular conditions worldwide. More than 60 million people are currently living with the disease, and numbers continue to rise due to ageing populations, increasing rates of diabetes and obesity, and improved survival after heart attacks and stroke. Despite therapeutic advances, heart failure remains a leading cause of hospitalisations, reduced quality of life, and healthcare burden.
One key challenge is that heart failure risk develops silently and varies between individuals. It may arise in people without previous cardiovascular disease, in patients with established atherosclerotic cardiovascular disease (ASCVD), or in those already diagnosed with specific forms of heart failure such as heart failure with preserved ejection fraction (HFpEF). Yet until recently, clinicians lacked a coherent framework to quantify risk consistently across these stages.
Traditional risk prediction models rely on well-known factors such as age, blood pressure, diabetes, cholesterol levels, and prior cardiovascular disease. However, they do not always capture the full complexity of heart failure risk. As a result, high-risk individuals may remain undetected, while others may receive unnecessary monitoring or treatment. Improving risk prediction is therefore essential both for patients and for healthcare systems facing growing cardiovascular disease burdens.
In three major international collaborative studies published in the European Heart Journal and co-led by Professor Emanuele Di Angelantonio – Head of the Health Data Science Research Centre at Human Technopole, and Professor of Clinical Epidemiology at the University of Cambridge – researchers developed and validated three complementary risk prediction models. Together, these models form a continuous framework for assessing heart failure risk across the entire disease trajectory, from primary prevention through to advanced disease management.
The first study introduces SCORE2-HF, a model estimating 10-year and 30-year risk of heart failure in adults over 40 years without prior cardiovascular disease.
Developed using data from over 600,000 individuals across 14 European countries and validated in over 1.3 million people, the model uses routinely available clinical information, including age, smoking status, blood pressure, body mass index, kidney function, and Type 2 diabetes, to identify individuals at elevated risk of heart failure long before symptoms appear. SCORE2-HF is aligned with the previously established SCORE2 models for ASCVD risk estimation, enabling synergistic assessment of both ASCVD and heart failure risk across Europe, supporting integrated primary prevention strategies.
The second study presents SMART2-HF, focusing on patients with established ASCVD, such as coronary artery disease or stroke, who have not yet developed heart failure. While existing guideline-recommended risk models for these patients quantify the risk of heart attack and stroke, they do not estimate heart failure risk. Built using data from nearly 8,000 patients and externally validated in over 240,000 patients across six international cohorts, SMART2-HF predicts both 10-year and lifetime risk of heart failure using routinely collected clinical data, complementing existing cardiovascular risk tools.
The third study introduces LIFE-Preserved, addressing patients already living with heart failure with preserved ejection fraction (HFpEF), a complex and increasingly common form of heart failure characterised by substantial heterogeneity in prognosis. The model, based on data from over 20,000 patients and validated in trials and real-world registries including more than 28,000 additional patients, predicts both short-term and lifetime risk of heart failure hospitalisation or cardiovascular death. New therapies have recently become available for HFpEF, but not all patients benefit equally. LIFE-Preserved helps identify those at highest risk, supporting personalised treatment decisions and shared decision-making between patients and clinicians.
Rather than being isolated tools, these models provide an integrated strategy: identifying risk of heart failure earlier, refining prognosis more precisely, and supporting treatment decisions based on individual probability rather than averages.
More broadly, the studies exemplify Human Technopole’s mission to translate data science into improving prevention, guiding precision treatment, and contributing to measurable public health impact in ageing societies. As Professor Di Angelantonio explains, “building on this research, the Human Technopole Cardiometabolic Flagship Research Programme will advance a bold vision for cardiometabolic health by using large-scale data and translational science to predict disease earlier, personalise care, and deliver measurable impact for patients and healthcare systems worldwide”.
