Novel AI tool for predicting cardio-cerebrovascular disease risk among Chinese

Researchers at the University of Hong Kong (HKU) have developed an innovative artificial intelligence (AI) assessment model that has the potential to fundamentally change the prevention and management of cardio-cerebrovascular (CCV) diseases specifically in the Chinese population.

The Personalized Cardiovascular Disease Risk Assessment for Chinese (P-CARDIAC) tool, developed by an interdisciplinary team at HKU, was trained and validated on a dataset of >300,000 patient records spanning 15 years extracted from the Hospital Authority’s electronic health record (EHR) system. [Eur Heart J 2024;5:363-370]

P-CARDIAC takes into consideration 125 risk variables (of which eight classes of cardiovascular disease [CVD]–related drugs were interactive covariates) to predict patients’ 10-year CCV risk in the form of a risk score (a composite of coronary heart disease, ischaemic/haemorrhagic stroke, peripheral artery disease, and revascularization).

“P-CARDIAC was validated to have better risk prediction performance in the Chinese population than existing CVD risk scores, such as Thrombolysis in Myocardial Infarction [TIMI] Risk Score for Secondary Prevention [TRS-2^oP] and Secondary Manifestations of Arterial Disease [SMART2], which were developed on western populations,” said Assistant Professor Celine Chui of HKU's School of Nursing and School of Public Health.

A 2-year clinical study, HEARTWISE, scheduled to commence in the fourth quarter of 2024, will investigate P-CARDIAC’s effectiveness in secondary prevention of CCV disease through a three-pronged approach:

1)  Evaluating P-CARDIAC’s effectiveness in predicting CCV disease through mapping P-CARDIAC scores with EHR data and patient-reported symptoms, including medications and upcoming specialty-clinic follow-up appointments, enabling potential early warning signs to be identified;

2)   Determining P-CARDIAC risk thresholds for CCV disease by comparing P-CARDIAC scores with clinical judgments, which will provide a more accurate and convenient reference for healthcare professionals and enable personalized management plans based on patients’ risk profiles; and

3)   Comparing CCV disease outcomes between standard care and pharmacist-led services (including long-term medication management and follow-up provided by pharmacists who work collaboratively with cardiologists and nurses), to assess the potential benefits of pharmacist-led interventions.

HEARTWISE will recruit 3,000 patients from six public hospitals across Hong Kong: Queen Mary Hospital, Queen Elizabeth Hospital, Kwong Wah Hospital, Princess Margaret Hospital, Tuen Mun Hospital, and Pok Oi Hospital.

"By leveraging P-CARDIAC’s personalized risk assessment capabilities and a multidisciplinary approach, HEARTWISE aims to identify effective strategies for reducing CCV disease burden in Hong Kong and beyond,” commented Professor Kai-Hang Yiu of HKU’s Division of Cardiology. “The ultimate goal is to incorporate P-CARDIAC into routine clinical practice to improve CCV health outcomes.”

A pilot study involving P-CARDIAC has been launched in collaboration with the Hong Kong Society of Rehabilitation (HKSR) at Central & Western District Health Centre and a community pharmacy, to enhance self-management among patients with chronic conditions such as stroke, post-acute myocardial infarction, diabetes, and hypertension through medication education and support. Preliminary results indicate that >90 percent of the patients’ P-CARDIAC score improved by the second visit (2 months after the first). Of the 20 patients recruited, 15 had follow-up visits, and 80 percent reported an improved understanding of their CCV disease risk, which contributed to better health outcomes and enhanced medication utilization.