Machine learning-based identification of risk-factor signatures for undiagnosed atrial fibrillation in primary prevention and post-stroke in clinical practice.

AIMS: Atrial fibrillation (AF) carries a substantial risk of ischemic stroke and other complications, and estimates suggest that over a third of cases remain undiagnosed. AF detection is particularly pressing in stroke survivors. To tailor AF screening efforts, we explored German health claims data for routinely available predictors of incident AF in primary care and post-stroke using machine learning methods. METHODS AND RESULTS: We combined AF predictors in patients over 45 years of age using claims data in the InGef database (n = 1 476 391) for (i) incident AF and (ii) AF post-stroke, using machine learning techniques. Between 2013-2016, new-onset AF was diagnosed in 98 958 patients (6.7%). Published risk factors for AF including male sex, hypertension, heart failure, valvular heart disease, and chronic kidney disease were confirmed. Component-wise gradient boosting identified additional predictors for AF from ICD-codes available in ambulatory care. The area under the curve (AUC) of the final, condensed model consisting of 13 predictors, was 0.829 (95% confidence interval (CI) 0.826-0.833) in the internal validation, and 0.755 (95% CI 0.603-0.890) in a prospective validation cohort (n = 661). The AUC for post-stroke AF was of 0.67 (95% CI 0.651-0.689) in the internal validation data set, and 0.766 (95% CI 0.731-0.800) in the prospective clinical cohort. CONCLUSION: ICD-coded clinical variables selected by machine learning can improve the identification of patients at risk of newly diagnosed AF. Using this readily available, automatically coded information can target AF screening efforts to identify high-risk populations in primary care and stroke survivors.

Blunted beta-adrenoceptor-mediated inotropy in valvular cardiomyopathy: another piece of the puzzle in human aortic valve disease.

OBJECTIVES: Heart failure induced by valvular cardiomyopathy occurs in a substantial proportion of patients undergoing heart valve surgery. We aimed (i) to quantify beta-adrenoceptor (beta-AR) function by measuring the inotropic effect of isoprenaline in left ventricular (LV) tissue and (ii) to correlate beta-AR-mediated inotropy with clinical markers of heart failure. METHODS: A total of 179 LV myocardial samples were obtained from 104 consecutive patients who underwent aortic valve (AV) surgery between 2017 and 2019. Beta-ARs were stimulated by increasing the concentrations of isoprenaline, followed by a single high concentration of forskolin and calcium. Beta-AR sensitivity was estimated as the concentration to achieve half maximum effects (EC50). Maximum effect size was calculated as the relative beta-AR-mediated inotropic response compared to the force in the presence of high calcium [FISO/Ca (%)]. In vitro data were correlated with the clinical indicators of LV disease. RESULTS: FISO/Ca was independent of age and sex and amounted to 79.6 ± 20.5%. In a multivariate regression model, we found a significant inverse association between FISO/Ca and preoperative left ventricular end-diastolic diameter increase per 10 mm (OR -9.24, 95% CI -16.66 to -1.82; P = 0.015). Furthermore, patients with end-stage heart failure showed a strong tendency towards more severe reduction of max beta-AR response, as indicated by reduced FISO/Ca in a multivariate model (OR -29.60, 95% CI -61.92 to 2.72; P = 0.055). CONCLUSIONS: Our study indicates that in vitro myocardial contractility testing can quantify beta-AR dysfunction in patients with AV disease. We found a significant association between reduced beta-AR sensitivity and increased LV diameter, which may indicate a role of beta-AR dysfunction in the development of heart failure in patients with AV disease.