The Generalizability Gap in Aortic Stenosis: Analysis of 87 Clinical Trials vs. Real-World Practice
This study reveals a significant generalizability gap in aortic stenosis trials, finding that while TAVR trial participants closely mirror real-world demographics, SAVR trials enroll notably younger populations, potentially limiting the applicability of surgical outcomes to older, more vulnerable patients.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Heart valves are the one-way doors that keep blood flowing in the right direction through the heart. When the aortic valve, which sits between the heart and the main artery, becomes stiff and narrow, a condition called aortic stenosis develops. This blockage forces the heart to work harder to push blood through, a problem that becomes increasingly common as people grow older. Deciding how to fix this valve is a major decision for doctors and patients. They rely on large medical studies, known as clinical trials, to understand which treatments work best. These trials test new procedures on groups of people and then use those results to guide care for everyone else. But for these results to be useful, the people in the studies must look like the people doctors see in their offices every day. If the study participants are very different from the real-world patients, the advice based on those studies might not fit.
A team of researchers at the Royal Free London NHS Foundation Trust set out to check if this mismatch exists for aortic stenosis treatments. They focused on two main ways to replace the valve: a surgical approach where the chest is opened to swap the valve, and a less invasive method where a new valve is threaded through a blood vessel to the heart. The researchers gathered data from eighty-seven different clinical trials registered in the United States, covering more than twenty-six thousand patients. They compared the ages and genders of the people in these trials against a massive database of real patients who had undergone these surgeries across the country. This real-world database served as a mirror to see if the trials were reflecting the actual population of people with the disease.
The researchers found a clear split in how well the trials matched reality, depending on which treatment was being tested. For the less invasive procedure, the people in the studies looked very much like the people getting the treatment in real life. The average age of participants in these trials was about eighty-one years, which is nearly identical to the average age of eighty-one and a half years seen in the national database. The proportion of women in these trials also matched the real world closely. However, the picture was quite different for the surgical procedure. The trials for surgery included patients who were significantly younger than those typically treated in hospitals. The average age of people in the surgical trials was about seventy-five and a half years, which is roughly six years younger than the average patient in the real-world database. Furthermore, while the real-world patients included nearly half women, the surgical trials included fewer women, and the studies for the less invasive procedure outnumbered the surgical studies by more than four to one.
This difference matters because age and gender often influence how well a patient recovers from a procedure and how risky it might be. The researchers suggest that the surgical trials may have historically focused on selecting patients who were expected to do well, perhaps because they were younger or had fewer other health problems. In contrast, the less invasive procedure was initially tested on older, sicker patients who could not survive open surgery, so the trials naturally included people who looked more like the typical elderly patient seen today. The authors note that this gap means the evidence supporting surgery comes from a population that is, on average, six years younger and more often male than the people who actually need the surgery. This does not mean the surgical trials were done poorly, but it does mean that doctors must be careful when applying the results from those studies to older, frailer patients, particularly elderly women, who are underrepresented in the research. The findings highlight that while the evidence for the less invasive method fits the current reality well, the evidence for surgery may need to be interpreted with extra caution when treating the oldest and most vulnerable members of the population.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.