1,721,004 research outputs found
A cardiologist’s guide to machine learning in cardiovascular disease prognosis prediction
No full textAbstract A modern-day physician is faced with a vast abundance of clinical and scientific data, by far surpassing the capabilities of the human mind. Until the last decade, advances in data availability have not been accompanied by analytical approaches. The advent of machine learning (ML) algorithms might improve the interpretation of complex data and should help to translate the near endless amount of data into clinical decision-making. ML has become part of our everyday practice and might even further change modern-day medicine. It is important to acknowledge the role of ML in prognosis prediction of cardiovascular disease. The present review aims on preparing the modern physician and researcher for the challenges that ML might bring, explaining basic concepts but also caveats that might arise when using these methods. Further, a brief overview of current established classical and emerging concepts of ML disease prediction in the fields of omics, imaging and basic science is presented
Guideline-directed medical therapy assessment in heart failure patients undergoing percutaneous mitral valve repair.
Full text linkAIMS
Achieving optimized guideline-directed medical therapy (GDMT) is recommended prior to transcatheter mitral valve edge-to-edge repair (M-TEER) for secondary mitral regurgitation (SMR). We aimed to propose and validate an easy-to-use score for assessing the quality of GDMT in patients with heart failure with reduced ejection fraction (HFrEF) undergoing M-TEER.
METHODS AND RESULTS
Among the 1641 EuroSMR patients enrolled in the EuroSMR Registry who underwent M-TEER, a total of 1072 patients [median age 74, interquartile range (IQR) 67-79 years, 29% female] had complete data on GDMT and a left ventricular ejection fraction ≤ 40% and were included in the current study. We proposed a GDMT score that considers the dosage levels of three medication classes (angiotensin-converting enzyme inhibitors/angiotensin receptor blockers/angiotensin receptor-neprilysin inhibitors, beta-blockers, and mineralocorticoid receptor antagonists), with a maximum score of 12 points indicating optimal GDMT. The primary outcome was all-cause mortality. The median GDMT score was 4 points (IQR 3-6). All three domains of the scoring system were associated with all-cause mortality (P < 0.05 for all). The overall GDMT score was associated with all-cause mortality (hazard ratio 0.90, 95% confidence interval 0.86-0.95 for each 1-point increase in the GDMT score). This association remained significant after adjusting for renal function and co-morbidities.
CONCLUSIONS
This study demonstrates the utility of a simple GDMT scoring system for assessing the adequacy of GDMT in HFrEF patients with relevant SMR undergoing M-TEER. The GDMT score has potential applications in guiding the design of future clinical trials and aiding clinical decision-making processes
Temporal trends in management and outcome of pulmonary embolism: a single-centre experience
No full textBACKGROUND: Real-world data on the impact of advances in risk-adjusted management on the outcome of patients with pulmonary embolism (PE) are limited. METHODS: To investigate temporal trends in treatment, in-hospital adverse outcomes and 1-year mortality, we analysed data from 605 patients [median age, 70 years (IQR 56-77) years, 53% female] consecutively enrolled in a single-centre registry between 09/2008 and 08/2016. RESULTS: Over the 8-year period, more patients were classified to lower risk classes according to the European Society of Cardiology (ESC) 2014 guideline algorithm while the number of high-risk patients with out-of-hospital cardiac arrest (OHCA) increased. Although patients with OHCA had an exceptionally high in-hospital mortality rate of 59.3%, the rate of PE-related in-hospital adverse outcomes (12.2%) in the overall patient cohort remained stable over time. The rate of reperfusion treatment was 9.6% and tended to increase in high-risk patients. We observed a decrease in the median duration of in-hospital stay from 10 (IQR 6-14) to 7 (IQR 4-15) days, an increase of patients discharged early from 2.1 to 12.2% and an increase in the use of non-vitamin K-dependent oral anticoagulants (NOACs) from 12.6 to 57.2% in the last 2 years (09/2014-08/2016) compared to first 6 years (09/2008-08/2014). The 1-year mortality rate (16.9%) remained stable throughout the study period. CONCLUSION: In-hospital adverse outcomes and 1-year mortality remained stable despite more patients with OHCA, shorter in-hospital stays, more patients discharged early and a more frequent NOAC use
Prediction and prognostic importance of in-hospital major bleeding in a real-world cohort of patients with pulmonary embolism
No full textPerformance of artificial intelligence-generated vs human-authored abstracts in a real-world setting
No full textTreatment response to spironolactone in patients with heart failure with preserved ejection fraction: a machine learning-based analysis of two randomized controlled trials
No full textSummary: Background: Whether there is a subset of patients with heart failure with preserved ejection fraction (HFpEF) that benefit from spironolactone therapy is unclear. We applied a machine learning approach to identify responders and non-responders to spironolactone among patients with HFpEF in two large randomized clinical trials. Methods: Using a reiterative cluster allocating permutation approach, patients from the derivation cohort (Aldo-DHF) were identified according to their treatment response to spironolactone with respect to improvement in E/e’. Heterogenous features of response (‘responders’ and ‘non-responders’) were characterized by an extreme gradient boosting (XGBoost) algorithm. XGBoost was used to predict treatment response in the validation cohort (TOPCAT). The primary endpoint of the validation cohort was a combined endpoint of cardiovascular mortality, aborted cardiac arrest, or heart failure hospitalization. Patients with missing variables for the XGboost model were excluded from the validation analysis. Findings: Out of 422 patients from the derivation cohort, reiterative cluster allocating permutation identified 159 patients (38%) as spironolactone responders, in whom E/e’ significantly improved (p = 0.005). Within the validation cohort (n = 525) spironolactone treatment significantly reduced the occurrence of the primary outcome among responders (n = 185, p log rank = 0.008), but not among patients in the non-responder group (n = 340, p log rank = 0.52). Interpretation: Machine learning approaches might aid in identifying HFpEF patients who are likely to show a favorable therapeutic response to spironolactone. Funding: See Acknowledgements section at the end of the manuscript
Proteomics to improve phenotyping in obese patients with heart failure with preserved ejection fraction
No full textInfluences on the Invasive Estimation of Cardiac Output with the Thermodilution and Indirect Fick-method
No full textIntroduction: Invasive measurement of cardiac output (CO) is a key hemodynamic parameter. While thermodilution (TD) is considered the method of choice, the calculation of CO based on the Fick principle is often preferred due to reduced cost and time. However, the indirect Fick method (iFM) used in clinical practice relies on an assumed oxygen consumption (VO2). Formulas to predict VO2, such as LaFarge (LaF), Dehmer (De) and Bergstra (Be), have been proposed, all of which were revealed to have limited predictive value in pediatric and adult patients compared to TD. The correlation between the iFM and TD method in an aged, realworld cohort was tested. Furthermore, variables leading to a mismatch between estimated and measured CO were investigated.
Methods: A single, clinically-indicated right heart catheterization was performed on each patient with CO estimated by iFM and TD in 194 consecutive patients between April 2015 and August 2017. Six patients were excluded due to incomplete baseline data. The VO2 was assumed by applying the formulas of LaF, De and Be. Body fat estimation was performed with the formula proposed by Jackson and Pollock.
Results: We included 188 consecutive patients (70±13 years, 59% male) in the current analysis. Severe tricuspid and mitral regurgitation were present in 25 and 43 patients, respectively.
CO-TD exhibited an overall moderate correlation to CO-iFM as assessed by LaF, De and Be formulas with large limits of agreement (-1.22 to 1.62, -2.31 to 1.65, -2.80 to 1.17 l/m², respectively). The mean difference of the CO between methods was 0.40, -0.24 and -0.81 (all p<0.001 for difference to TD), respectively. A rate of error ≥20% occurred with the equations by LaF, De and Be in 32%, 29% and 51% of patients, respectively.
TD-method as compared to iFM with LaF formula underestimated CO in patients with severe tricuspid regurgitation (p=0.022) but not when iFM was calculated based on the De (p=0.229) or Be (p=0.418) formula. Body fat estimation (29%± 12%) was performed in a subgroup of patients (n=149). Mitral regurgitation, body fat as well as cardiac rhythm disturbances did not affect the correlation between TD- and iFM.
Conclusion: Although CO-eFM exhibits an overall reasonable correlation with CO-TD, the predictive value in a single patient is low. CO-eFM cannot replace CO-TD in elderly patients. Common variables leading to a mismatch between the estimated and the measured CO, such as morbid obesity did not lead to a significant difference in this cohort.Einführung: Die Bestimmung des Herzzeitvolumens (HZV) stellt eine zentrale hämodynamische Messgröße dar. Das HZV wird üblicherweise mittels ThermodilutionsMethode (TD) oder der Methode nach Fick gemessen. Die Fick-Methode benötigt zur Berechnung des HZV den Sauerstoffverbrauch (VO2). In der klinischen Praxis wird der VO2 häufig nicht gemessen, sondern anhand einer von drei empirischen Formeln nach LaFarge (Lf), Dehmer (De) oder Bergstra (Bg) geschätzt. Diese Formeln wurden jedoch vornehmlich an pädiatrischen Kohorten untersucht. Daher wird in der vorliegenden Arbeit die Korrelation dieser Methoden in einer gealterten kardiologischen Population untersucht und zusätzlich werden Variablen untersucht die potenziell zu einer Verschlechterung der Korrelation führen.
Methoden: Zwischen April 2015 und August 2017 wurden bei 194 Patienten mit der klinischen Indikation zur Rechtsherzkatheteruntersuchung Messungen des HZV mittels TD- und indirekter Fick-Methode vorgenommen. Sechs Patienten wurden aufgrund von fehlenden Daten aus der Analyse ausgeschlossen. Der VO2 wurde anhand der Formeln von Lf, De und Bg berechnet. Der Körperfettanteil wurde mittels der Formel nach Jackson und Pollock berechnet.
Ergebnisse: 188 Patienten (70 ± 13, 59 % männlich) wurden in die vorliegende Analyse eingeschlossen. Eine hochgradige Trikuspidal- bzw. Mitralklappeninsuffizienz lag bei 25 (13 %) bzw. 43 Patienten (23 %) vor. Es gab eine moderate Korrelation der TD-Methode und der indirekten Fick-Methode (iFM) berechnet nach den Formeln von Lf, De und Be mit einem großen Übereinstimmungsbereich in den Analysen der Bland-Altman-Graphiken (-1.22 bis 1.62 [Lf], -2.31 bis 1.65 [De] und -2.80 bis 1.17 l/m² [Be]). Der mittlere Unterschied zwischen dem HZV der TD-Methode und der iFM war 0.40 (Lf), -0.24 (De) und -0.81 l/min (Be) (alle p < 0.001). Ein Unterschied zwischen dem HZV nach TD-Methode und iFM von ≥20 % wurde bei 32 % (Lf), 29 % (De) und 51 % (Bg) der Patienten beobachtet.
Die TD-Methode zeigte niedrigere HZV-Werte im Vergleich zur iFM Lf (p = 0.022), nicht jedoch nach der Formel von De (p = 0.229) oder Be (p = 0.418), bei Patienten mit hochgradiger Trikuspidalklappeninsuffizienz. Körperfettmessungen (Körperfettanteil 29% ± 12 %) erfolgten in einer Subgruppe von 149 Patienten. Es konnte kein relevanter Einfluss von Mitraklappeninsuffizienzen, dem Körperfettanteil und dem Herzrhythmus auf die Diskrepanz zwischen TD- und iFM beobachtet werden.
Zusammenfassung: Die TD-Methode und iFM zeigen eine moderate Korrelation, jedoch mit großen individuellen Unterschieden. Die iFM kann die Messung des HZV in älteren kardiologischen Patienten nicht ersetzen. Übliche Variablen, die wie eine morbide Adipositas zu einer Fehleinschätzung des HZV führen sollten, zeigten keinen Effekt in der vorliegenden Arbeit
Guideline-Directed Medical Therapy and Survival After TEER for Secondary Mitral Regurgitation With Right Ventricular Impairment
Full text linkBackground: Right ventricular impairment is common among patients undergoing transcatheter edge-to-edge repair for secondary mitral regurgitation (SMR). Adherence to guideline-directed medical therapy (GDMT) for heart failure is poor in these patients. Objectives: The aim of this study was to evaluate the impact of GDMT on long-term survival in this patient cohort. Methods: Within the EuroSMR (European Registry of Transcatheter Repair for Secondary Mitral Regurgitation) international registry, we selected patients with SMR and right ventricular impairment (tricuspid annular plane systolic excursion ≤17 mm and/or echocardiographic right ventricular-to-pulmonary artery coupling <0.40 mm/mm Hg). Titrated guideline-directed medical therapy (GDMTtit) was defined as a coprescription of 3 drug classes with at least one-half of the target dose at the latest follow-up. The primary outcome was all-cause mortality at 6 years. Results: Among 1,213 patients with SMR and right ventricular impairment, 852 had complete data on medical therapy. The 123 patients who were on GDMTtit showed a significantly higher long-term survival vs the 729 patients not on GDMTtit (61.8% vs 36.0%; P < 0.00001). Propensity score-matched analysis confirmed a significant association between GDMTtit and higher survival (61.0% vs 43.1%; P = 0.018). GDMTtit was an independent predictor of all-cause mortality (HR: 0.61; 95% CI: 0.39-0.93; P = 0.02 for patients on GDMTtit vs those not on GDMTtit). Its association with better outcomes was confirmed among all subgroups analyzed. Conclusions: In patients with right ventricular impairment undergoing transcatheter edge-to-edge repair for SMR, titration of GDMT to at least one-half of the target dose is associated with a 40% lower risk of all-cause death up to 6 years and should be pursued independent of comorbidities
- …
