To drive health innovation that meets the needs of all and democratize healthcare, there is a need to assess the generalization performance of deep learning (DL) algorithms across various distribution shifts to ensure that these algorithms are robust. This retrospective study is, to the best of our knowledge, an original attempt to develop and assess the generalization performance of a DL model for AF events detection from long term beat-to-beat intervals across geography, ages and sexes. The new recurrent DL model, denoted ArNet2, is developed on a large retrospective dataset of 2,147 patients totaling 51,386 h obtained from continuous electrocardiogram (ECG). The model’s generalization is evaluated on manually annotated test sets from four centers (USA, Israel, Japan and China) totaling 402 patients. The model is further validated on a retrospective dataset of 1,825 consecutives Holter recordings from Israel. The model outperforms benchmark state-of-the-art models and generalized well across geography, ages and sexes. For the task of event detection ArNet2 performance was higher for female than male, higher for young adults (less than 61 years old) than other age groups and across geography. Finally, ArNet2 shows better performance for the test sets from the USA and China. The main finding explaining these variations is an impairment in performance in groups with a higher prevalence of atrial flutter (AFL). Our findings on the relative performance of ArNet2 across groups may have clinical implications on the choice of the preferred AF examination method to use relative to the group of interest.
Background: Diagnosis of AF-induced cardiomyopathy can be challenging and relies on ruling out other causes of cardiomyopathy and, after restoration of sinus rhythm, recovery of left ventricular (LV) function. The aim of this study was to identify clinical and echocardiographic predictors for developing cardiomyopathy with systolic dysfunction in patients with atrial tachyarrhythmia. Methods: This retrospective study was conducted in a large tertiary care centre and compared patients who experienced deterioration of LV ejection fraction (EF) during paroxysmal AF, demonstrated by precardioversion transoesophageal echocardiography with patients with preserved LV function during AF. All patients had documented preserved LVEF at baseline (EF >50%) while in sinus rhythm. Results: Of 482 patients included in the final analysis, 80 (17%) had reduced and 402 (83%) had preserved LV function during the precardioversion transoesophageal echocardiography. Patients with reduced LVEF were more likely to be men and to have a more rapid ventricular response during AF or atrial flutter (AFL). A history of prosthetic valves was also identified as a risk factor for reduced LVEF. Patients with reduced LVEF also had higher incidence of tricuspid regurgitation and right ventricular dysfunction. Conclusion: In ‘real-world’ experience, male patients with rapid ventricular response during paroxysmal AF or AFL are more prone to LVEF reduction. Patients with prosthetic valves are also at risk for LVEF reduction during AF/AFL. Finally, tricuspid regurgitation and right ventricular dysfunction may indicate relatively long-standing AF with an associated reduction in LVEF.
1 August 2022 in Journal of Cardiovascular Pharmacology
Patients with heart failure (HF) with iron deficiency (ID) have worse New York Heart Association class and are at a higher risk of recurrent hospitalizations. Intravenous (IV) iron has been shown to improve exercise ability and reduce hospitalizations. IV sodium ferric gluconate complex (SFGC) has been found to be safe and affordable but has not been studied in this population in a randomized trial. This was a prospective, single-blind, investigator-initiated, randomized controlled trial. Patients admitted for acute heart failure with ID were randomly assigned 1:1 to receive IV SFGC on top of optimal medical treatment. The primary outcome was the change in the 6-minute walk test (6MWT) from baseline to 3 and 6 months. Between September 2019 and May 2021, 34 patients were randomized. 19 patients (55%) were randomized to the treatment arm receiving 125 mg of IV SFGC per day for 3–5 days. COVID-19 was a major barrier to the implementation of the study follow-up protocol, which caused the study to end early. Both groups of patients had similar clinical characteristics, comorbidities, median left ventricular ejection fraction, and rate of death and readmissions due to HF. A higher level of NT-proBNP was observed in patients treated with IV iron (7902 pg/mL vs. 3158, P = 0.04). There was no difference in 6MWT change between groups at 3 months (improvement of 21.6 vs. 24.1 meters) or 6 months (−5 meters vs. 46 meters). In conclusion, IV SFGC–treated patients had a comparable 6-minute walk at 3 and 6 months despite suffering from more severe HF with higher baseline NT-proBNP (NCT04063033).