Short-term vital parameter forecasting in the intensive care unit: A benchmark study leveraging data from patients after cardiothoracic surgery.

Patients in an Intensive Care Unit (ICU) are closely and continuously monitored, and many machine learning (ML) solutions have been proposed to predict specific outcomes like death, bleeding, or organ failure. Forecasting of vital parameters is a more general approach to ML-based patient monitoring, but the literature on its feasibility and robust benchmarks of achievable accuracy are scarce. We implemented five univariate statistical models (the naïve model, the Theta method, exponential smoothing, the autoregressive integrated moving average model, and an autoregressive single-layer neural network), two univariate neural networks (N-BEATS and N-HiTS), and two multivariate neural networks designed for sequential data (a recurrent neural network with gated recurrent unit, GRU, and a Transformer network) to produce forecasts for six vital parameters recorded at five-minute intervals during intensive care monitoring. Vital parameters were the diastolic, systolic, and mean arterial blood pressure, central venous pressure, peripheral oxygen saturation (measured by non-invasive pulse oximetry) and heart rate, and forecasts were made for 5 through 120 minutes into the future. Patients used in this study recovered from cardiothoracic surgery in an ICU. The patient cohort used for model development (n = 22,348) and internal testing (n = 2,483) originated from a heart center in Germany, while a patient sub-set from the eICU collaborative research database, an American multicenter ICU cohort, was used for external testing (n = 7,477). The GRU was the predominant method in this study. Uni- and multivariate neural network models proved to be superior to univariate statistical models across vital parameters and forecast horizons, and their advantage steadily became more pronounced for increasing forecast horizons. With this study, we established an extensive set of benchmarks for forecast performance in the ICU. Our findings suggest that supplying physicians with short-term forecasts of vital parameters in the ICU is feasible, and that multivariate neural networks are most suited for the task due to their ability to learn patterns across thousands of patients.

LVAD as a Bridge to Transplantation-Current Status and Future Perspectives.

Heart failure (HF) is a common disease associated with high morbidity and mortality rates despite advanced pharmacological therapies. Heart transplantation remains the gold standard therapy for end-stage heart failure; however, its application is curtailed by the persistent shortage of donor organs. Over the past two decades, mechanical circulatory support, notably Left Ventricular Assist Devices (LVADs), have been established as an option for patients waiting for a donor organ. This comprehensive review focuses on elucidating the benefits and barriers associated with this application. We provide an overview of landmark clinical trials that have evaluated the use of LVADs as a bridge to transplantation therapy, with a particular focus on post-transplant outcomes. We discuss the benefits of stabilizing patients with these systems, weighing associated complications and limitations. Further technical advancements and research on optimal implantation timing are critical to ultimately improve outcomes and securing quality of life. In a world where the availability of donor organs remains constrained, LVADs are an increasingly important piece of patient care, bridging the critical gap to transplantation in advanced heart failure management.

Stroke outcomes following durable left ventricular assist device implant in patients bridged with micro-axial flow pump: Insights from a large registry.

BACKGROUND: Stroke after durable left ventricular assist device (d-LVAD) implantation portends high mortality. The incidence of ischemic and hemorrhagic stroke and the impact on stroke outcomes of temporary mechanical circulatory support (tMCS) management among patients requiring bridge to d-LVAD with micro-axial flow-pump (mAFP, Abiomed) is unsettled. METHODS: Consecutive patients, who underwent d-LVAD implantation after being bridged with mAFP at 19 institutions, were retrospectively included. The incidence of early ischemic and hemorrhagic stroke after d-LVAD implantation (<60 days) and association of pre-d-LVAD characteristics and peri-procedural management with a specific focus on tMCS strategies were studied. RESULTS: Among 341 patients, who underwent d-LVAD implantation after mAFP implantation (male gender 83.6%, age 58 [48-65] years, mAFP 5.0/5.5 72.4%), the early ischemic stroke incidence was 10.8% and early hemorrhagic stroke 2.9%. The tMCS characteristics (type of mAFP device and access, support duration, upgrade from intra-aortic balloon pump, ECMELLA, ECMELLA at d-LVAD implantation, hemolysis, and bleeding) were not associated with ischemic stroke after d-LVAD implant. Conversely, the device model (mAFP 2.5/CP vs. mAFP 5.0/5.5: HR 5.6, 95%CI 1.4-22.7, p = 0.015), hemolysis on mAFP support (HR 10.5, 95% CI 1.3-85.3, p = 0.028) and ECMELLA at d-LVAD implantation (HR 5.0, 95% CI 1.4-18.7, p = 0.016) were associated with increased risk of hemorrhagic stroke after d-LVAD implantation. Both early ischemic (HR 2.7, 95% CI 1.9-4.5, p < 0.001) and hemorrhagic (HR 3.43, 95% CI 1.49-7.88, p = 0.004) stroke were associated with increased 1-year mortality. CONCLUSIONS: Among patients undergoing d-LVAD implantation following mAFP support, tMCS characteristics do not impact ischemic stroke occurrence, while several factors are associated with hemorrhagic stroke suggesting a proactive treatment target to reduce this complication.

Temporary extracorporeal life support: single-centre experience with a new concept.

OBJECTIVES: The combination of veno-arterial extracorporeal membrane oxygenation with a micro-axial flow pump (ECMELLA) is increasingly used for cardiogenic shock (CS) therapy. We report our experience with a novel single-artery access ECMELLA setup with either femoral (2.0) or jugular venous cannulation (2.1), respectively. METHODS: Data from 67 consecutive CS patients treated with ECMELLA 2.0 (n = 56) and 2.1 (n = 11) from December 2020 and December 2022 in a tertiary cardiac center were retrospectively analyzed. RESULTS: The mean age was 60.7 ± 11 years, 56 patients (84%) were male. CS aetiology was acute on chronic heart failure (n = 35, 52%), myocardial infarction (n = 13, 19.5%), postcardiotomy syndrome (n = 16, 24%) and myocarditis (n = 3, 4.5%). Preoperatively 31 patients (46%) were resuscitated, 53 (79%) were on a ventilator and 60 (90%) were on inotropic support. The median vasoactive inotropic score was 32, and the mean arterial lactate was 8.1 mmol/l. In 39 patients (58%), veno-arterial extracorporeal membrane oxygenation was explanted after a median ECMELLA support of 4 days. Myocardial recovery was achieved in 18 patients (27%), transition to a durable left ventricular assist device in 16 (24%). Thirty-three patients (n = 33; 49%) died on support (25 on ECMELLA and 8 on Impella after de-escalation), 9 (13%) of whom were palliated. Axillary access site bleeding occurred in 9 patients (13.5%), upper limb ischaemia requiring surgical revision in 3 (4.5%). Axillary site infection occurred in 6 cases (9%), and perioperative stroke in 10 (15%; 6 hemorrhagic, 4 thromboembolic). CONCLUSIONS: ECMELLA 2.0/2.1 is a feasible and effective therapy for severe CS. The single-artery cannulation technique is associated with a relatively low rate of access-related complications.

Treatment of cardiac implantable electronic device endocarditis in pacemaker-dependent patients utilizing a percutaneous aspiration system.

With the increasing use of cardiac implantable electronic devices, the number of patients with cardiac implantable electronic device-related endocarditis is also rising. The treatment of this type of endocarditis is a challenging clinical task, in particular if device removal is required in patients who are pacemaker dependent. This video tutorial describes a treatment strategy for cardiac implantable electronic device-related endocarditis involving the tricuspid valve in pacemaker-dependent patients. The proposed treatment strategy consists of implanting an epicardial pacemaker via a minimally invasive subxiphoid approach, percutaneous aspiration of tricuspid valve vegetations and complete transvenous explantation of the infected cardiac implantable electronic device system using advanced lead extraction tools.

Temporary Mechanical Circulatory Support in Cardiogenic Shock Patients after Cardiac Procedures: Selection Algorithm and Weaning Strategies.

Mechanical circulatory support has proven effective in managing postcardiotomy cardiogenic shock by stabilizing patients' hemodynamics and ensuring adequate organ perfusion. Among the available device modalities, the combination of extracorporeal life support and a microaxial flow pump for left ventricular unloading has emerged as a valuable tool in the surgical armamentarium. In this publication, we provide recommendations for the application and weaning of temporary mechanical circulatory support in cardiogenic shock patients, derived from a consensus among leading cardiac centers in German-speaking countries.

Impact of preoperative mitral regurgitation on left ventricular assist device patients: propensity score-matched analysis of the EUROMACS dataset.

OBJECTIVES: Mitral regurgitation (MR) is frequently observed in patients undergoing left ventricular assist device implantation. We investigated the impact of preoperative MR on left ventricular assist device patients. METHODS: A retrospective propensity score-matched analysis of adult patients enrolled in the EUROMACS registry between 1 January 2011 and 30 November 2021 was performed. Patients were divided into 2 groups according to the grade of preoperative MR: none-to-mild (MR 0-II) or moderate-to-severe (MR III-IV). RESULTS: Following 1:1 propensity score matching, each group consisted of 914 patients. Incidence of postoperative temporary right ventricular support, reoperation for bleeding and dialysis was similar. MR III-IV demonstrated shorter median intensive care unit stay [14 days (6; 27.8) vs 10 days (5; 22), P = 0.004] and ventilation time [72 h (22, 320) vs 31 h (18, 150), P < 0.001]. Mortality was lower for MR III-IV patients [subdistribution hazard ratio: 0.66, 95% confidence interval (CI): 0.59-0.73, P < 0.001]. The 1-year survival was 68.1% (95% CI: 65.1-71.3%) in MR 0-II and 75% (95% CI: 72.1-78%) in MR III-IV. A lower incidence of total complications [odds ratio (OR): 0.93 (0.89-0.98), P = 0.003] and trend towards a lower risk of neurological dysfunction (subdistribution hazard ratio: 0.79; 95% CI: 0.61-1.01, P = 0.063) and sustained ventricular tachycardia [OR: 0.93 (0.54-1.03), P = 0.074] were demonstrated for MR III-IV. The risk of fatal stroke and pump thrombosis was similar. CONCLUSIONS: Moderate-to-severe MR in patients undergoing left ventricular assist device implantation is associated with better mid-term survival and lower incidence of total major adverse events and complications. The incidence of severe postoperative complications including fatal stroke and device thrombosis was similar.

HeartWare to HeartMate 3 left ventricular assist device exchange via a left lateral thoracotomy.

A continuous-flow left ventricular assist device implant is a well-established therapy for patients with end-stage heart failure. Currently, the HeartMate 3 device is the only commercially available durable left ventricular assist device. Therefore, patients on HeartWare HVAD support who require a pump exchange must have a HeartMate 3 implanted.

Improving Survival in Cardiogenic Shock-A Propensity Score-Matched Analysis of the Impact of an Institutional Allocation Protocol to Short-Term Mechanical Circulatory Support.

Temporary mechanical circulatory support (tMCS) is a life-saving treatment option for patients in cardiogenic shock (CS), but many aspects such as patient selection, initiation threshold and optimal modality selection remain unclear. This study describes a standard operating procedure (SOP) for tMCS allocation for CS patients and presents outcome data before and after implementation. Data from 421 patients treated for CS with tMCS between 2018 and 2021 were analyzed. In 2019, we implemented a new SOP for allocating CS patients to tMCS modalities. The association between the time of SOP implementation and the 30-day and 1-year survival as well as hospital discharge was evaluated. Of the 421 patients included, 189 were treated before (pre-SOP group) and 232 after implementation of the new SOP (SOP group). Causes of CS included acute myocardial infarction (n = 80, 19.0%), acute-on-chronic heart failure in patients with dilated or chronic ischemic heart failure (n = 139, 33.0%), valvular cardiomyopathy (n = 14, 3.3%) and myocarditis (n = 5, 1.2%); 102 patients suffered from postcardiotomy CS (24.2%). The SOP group was further divided into an SOP-adherent (SOP-A) and a non-SOP-adherent group (SOP-NA). The hospital discharge rate was higher in the SOP group (41.7% vs. 29.7%), and treating patients according to the SOP was associated with an improved 30-day survival (56.9% vs. 38.9%, OR 2.21, 95% CI 1.01-4.80, p = 0.044). Patient allocation according to the presented SOP significantly improved 30-day survival.

Interventional Procedures for Left Ventricular Assist Device-Associated Complications.

As patients on long-term left ventricular assist device (LVAD) face a substantial risk for open cardiac reoperation, interventional treatment approaches are becoming increasingly important in this population. We evaluated data of 871 patients who were on LVAD support between January 1, 2016 and December 1, 2020. Interventional treatments for LVAD-associated complications were performed in 76 patients. Seventeen patients underwent transcatheter aortic valve replacements (TAVR) and 61 patients underwent outflow graft interventions (OGI). TAVR improved symptoms in patients with severe symptomatic aortic regurgitation. Postinterventional complications included aggravation of preexisting right heart failure (RHF), third-degree atrioventricular block, and intrapump thrombosis (in 3 [16.7%], 2 [11.1%], and 1 [5.6%] patients, respectively). In outflow graft obstructions, OGI led to recovery of LVAD flow ( p < 0.001), unloading of the left ventricle ( p = 0.004), decrease of aortic valve opening time ( p = 0.010), and improvement of right heart function ( p < 0.001). Complications included bleeding, RHF, and others (in 9 [10.8%], 5 [6.0%], and 5 [6.0%] patients, respectively). Eight (9.6%) patients died within the hospital stay after OGI, including mortality secondary to prolonged cardiogenic shock. In conclusion, interventional procedures are a feasible and safe treatment modality for LVAD-associated complications.

Durable mechanical circulatory support in patients with heart failure with preserved ejection fraction.

Only few patients with heart failure with preserved ejection fraction undergo durable mechanical circulatory support implantation. We identified three patients with heart failure with preserved ejection fraction who needed durable mechanical circulatory support as bridge-to-transplant therapy. In two patients with hypertrophic cardiomyopathy, the hypertrophic papillary muscles and myocardium were resected to allow for subsequent left ventricular assist device implantation. In one patient, all visible parts of the mitral valve were additionally resected. The third patient with restrictive cardiomyopathy underwent Berlin Heart Excor BVAD implantation with left atrial cannulation.