AI Tools Identified Five Heart Failure Types

A new study conducted by UCL researchers has discovered five kinds of heart failure that could possibly be used to predict future risk for individual patients. When the heart is unable to efficiently pump blood around the body, this is referred to as heart failure. Current classification systems for heart failure do not reliably anticipate how the condition will progress.

Researchers examined detailed anonymised patient data from nearly 300,000 adults aged 30 or older who were diagnosed with heart failure in the UK during a 20-year period for the study, which was published in The Lancet Digital Health.

They identified five subtypes using several machine learning methods: early onset, late onset, atrial fibrillation related (an irregular heart rhythm caused by atrial fibrillation), metabolic (linked to obesity but with a low rate of cardiovascular disease), and cardiometabolic (linked to obesity and cardiovascular disease).

The researchers discovered differences in the risk of death among the subgroups one year following diagnosis. At one year, the all-cause death risks were as follows: early onset (20%), late onset (46%), atrial fibrillation (61%), metabolic (11%), and cardiometabolic (37%).

The research team also created an app that clinicians might use to detect which subtype of heart failure a patient has, which could enhance projections of future risk and inform discussions with patients.

Lead author Professor Amitava Banerjee (UCL Institute of Health Informatics) said, “We sought to improve how we classify heart failure, with the aim of better understanding the likely course of disease and communicating this to patients. Currently, how the disease progresses is hard to predict for individual patients. Some people will be stable for many years, while others get worse quickly.”

Better distinctions between types of heart failure may also lead to more targeted treatments and may help us to think in a different way about potential therapies.

“In this new study, we identified five robust subtypes using multiple machine learning methods and multiple datasets.”

“The next step is to see if this way of classifying heart failure can make a practical difference to patients—whether it improves predictions of risk and the quality of information clinicians provide, and whether it changes patients’ treatment. We also need to know if it would be cost effective. The app we have designed needs to be evaluated in a clinical trial or further research, but could help in routine care.”

To eliminate bias from a single machine learning method, the researchers grouped cases of heart failure using four different ways. These methods were used to data from two large UK primary care datasets that were representative of the entire UK population and were also connected to hospital admissions and mortality records. (The datasets covered the years 1998 to 2018. They were Clinical Practice Research Datalink (CPRD) and The Health Improvement Network (THIN).)

The research team trained the machine learning techniques on data segments and then tested these categories using a separate dataset after selecting the most robust subtypes.

Age, symptoms, the presence of other diseases, the drugs the patient was taking, and the results of tests (e.g., blood pressure) and evaluations (e.g., renal function) were used to determine the subtypes (of a possible 635).

The researchers also examined genetic data from 9,573 people with heart failure from the UK Biobank project. They discovered a relationship between specific subtypes of heart failure and higher polygenic risk scores (all-gene risk scores) for diseases like hypertension and atrial fibrillation.

Source Link