Researchers at the Icahn School of Medicine at Mount Sinai (New York, USA) have received a US$3.6 million, four-year grant from the National Heart, Lung, and Blood Institute (NHLBI), part of the National Institutes of Health (NIH), to develop an approach to extend evidence from clinical trials to a broader population of patients treated in everyday medical practice.
The project, called EVOLVE-MR (Extending valve-surgery outcomes from landmark trials to varied everyday patients with mitral regurgitation), will link data from major randomised clinical trials with long-term Medicare records. Researchers will use advanced statistical methods and machine learning to determine whether benefits seen in clinical trials extend to a broader, more diverse population and persist over a longer period.
The study is led by Mayte Suárez-Fariñas, co-director of Icahn’s Division of Biostatistics and Data Science in the Department of Population Health Science and Policy; and Emilia Bagiella, co-director of the Division of Biostatistics and Data Science and Director of the Center for Biostatistics in the Department of Population Health Science and Policy.
The project builds on more than a decade of research conducted through the Cardiothoracic Surgical Trials Network (CTSN), an NHLBI-funded network that has conducted landmark randomised clinical trials to determine the best treatment approaches for patients with heart valve disease.
“Landmark trials tell us what works on average for the patients who enrolled, and those are not always the ones we see in everyday practice. Surgeons and patients need evidence that reaches beyond the trial population, down to the individual in front of them, and over a lifetime,” said Suárez-Fariñas. “EVOLVE-MR links trial data to Medicare records so we can follow these patients and use causal machine learning to identify who benefits most from each surgical strategy.”
A major goal of the study is to move beyond asking whether one surgical approach is better overall and instead determine which treatment is most likely to benefit each patient.
The researchers will use modern causal inference methods and machine learning to identify differences in treatment benefit based on patients’ clinical characteristics, demographic factors, and other available information.
The team will also analyse echocardiographic imaging features to explore whether characteristics visible on heart imaging can help predict which surgical strategy may provide the greatest benefit for an individual patient.
“This project combines the power and rigour of clinical trials with the richness of real-world data, allowing the trial results to be generalised to broader patient populations and longer time horizons,” said Bagiella.
