Background: Large Language Models (LLMs) such as ChatGPT, Claude, and Gemini are being increasingly adopted in medicine; however, their reliability in cardiology remains underexplored. Purpose of the study: To compare the performance of three general-purpose LLMs in response to cardiology-related clinical queries. Study design: Seventy clinical prompts stratified by diagnostic phase (pre or post) and user profile (patient or physician) were submitted to ChatGPT, Claude, and Gemini. Three expert cardiologists, who were blinded to the model’s identity, rated each response on scientific accuracy, completeness, clarity, and coherence using a 5-point Likert scale. Statistical analysis included Kruskal–Wallis tests, Dunn’s post hoc comparisons, Kendall’s W, weighted kappa, and sensitivity analyses. Results: ChatGPT outperformed both Claude and Gemini across all criteria (mean scores: 3.7–4.2 vs. 3.4–4.0 and 2.9–3.7, respectively; p < 0.001). The inter-rater agreement was substantial (Kendall’s W: 0.61–0.71). Pre-diagnostic and patient-framed prompts received higher scores than post-diagnostic and physician-framed ones. Results remained robust across sensitivity analyses. Conclusions: Among the evaluated LLMs, ChatGPT demonstrated superior performance in generating clinically relevant cardiology responses. However, none of the models achieved maximal ratings, and the performance varied by context. These findings highlight the need for domain-specific fine-tuning and human oversight to ensure a safe clinical deployment.

Objective Despite minimally invasive techniques having gained wider application in cardiac surgery, current evidence on minithoracic aortic surgery is still limited. The aim of this study was to compare early and midterm outcomes of patients undergoing operations of the proximal thoracic aorta through ministernotomy (MS) versus full sternotomy (FS). Methods Data from 624 consecutive patients who underwent proximal aortic repair through MS (n = 214, 34.3%) and FS (n = 410, 65.7%) at two aortic centers were analyzed. Treatment selection bias was addressed using propensity score matching (MS vs. FS). After matching, two well-balanced groups of 202 patients each were created. Results Median cardiopulmonary bypass and cross-clamp times were 88 and 68 min, respectively, with no difference between groups. Overall, 30-day mortality was 1% (n = 2) in MS and 0.5% (n = 1) in FS (p = .6). No difference was found in the rates of stroke (MS n = 5, 2.5%; FS n = 5, 2.5%), dialysis (MS n = 1, 0.5%; FS n = 4, 2%), bleeding (MS n = 7, 3.5%; FS n = 7, 3.5%), and blood transfusions (MS n = 67, 33.3%; FS n = 57, 28.4%). Patients receiving MS showed a lower incidence of respiratory insufficiency compared with FS (0% vs. 2.5%, p = .04). Intensive care unit and hospital stays were similar between groups. Two-year survival rate was 97.2% in MS and 96.5% in FS (p = .9). Conclusions Mini proximal aortic operations can be performed successfully without compromising the proven efficacy and safety of conventional access. In selected patients, MS was associated with very low mortality and morbidity rates. Additionally, MS demonstrated superior clinical outcomes as regards respiratory adverse events, when compared with FS.

The clinical benefits of mitral valve repair over replacement in the setting of mitral infective endocarditis remain not clearly established. We aim to review the clinical experience in mitral valve surgery for infective endocarditis looking at mid–term survival. Prospectively collected data of patients who underwent cardiac surgery for infective endocarditis between 2001 and 2021 at two cardiac centres were reviewed. Among them, 282 patients underwent native mitral valve surgery and were included in the study. Nearest–neighbour propensity score matching, including twelve preoperative variables, was performed to account for differences in patients’ profile between the repair and replacement subgroups. Propensity matching provided 87 well–matched pairs. Mean age was 60±15 years, 76% of the patients were male. Twenty–two patients (12%) had a recent cerebral event. CABG and aortic valve replacement were associated in 10% and in 38% of the cases, respectively. Patients who underwent mitral valve replacement had a higher rate of annular, commissures and bileaflet lesions. Overall in–hospital mortality was 7% with no difference between the repair and the replacement cohorts (p=0.77). Survival probabilities at 1, 5– and 10–years were 89%, 70% and 67%, respectively after mitral repair, and 88%, 76% and 62%, respectively after mitral replacement; log–rank p=0.94. Freedom from mitral intervention by any cause at 10–years was 3.3% and 8.7%, respectively after mitral repair and mitral replacement; Gray’s p=0.37. Mitral valve repair was associated with acceptable in–hospital mortality and survival in a high–risk and comorbid cohort of patients with mitral valve infection. We found no difference in early, mid–term survival and mitral reoperation risk between patients who underwent repair or replacement of the mitral valve in the setting of infective endocarditis.