An article in Nature Medicine on February 20, 2026, highlights the results of a large pragmatic trial of an AI stethoscope to detect cardiovascular disease. When used as intended, the AI stethoscope improved the detection of these diseases compared to traditional methods[1]. In the aforementioned American study, the AI stethoscope achieved a sensitivity of 92.3% in identifying moderate to severe valvular heart disease, while the traditional stethoscope only had a sensitivity of 46.2%[1]. This identified more than twice the number of cases during routine clinical examinations[1]. The AI stethoscope records high-quality heart sounds and uses machine learning algorithms to recognize acoustic patterns associated with valvular diseases[1]. Nevertheless, low adoption rates and workflow problems in a pragmatic trial significantly reduced the overall effectiveness of the technology. The authors of the trial highlight the risks of gaps in implementation, such as lower specificity leading to false positive results[1].