Influence of timing of Levosimendan administration on outcomes in cardiac surgery.

Purpose: Though a subgroup analysis has shown improved survival for patients suffering severely reduced ventricular function undergoing coronary artery bypass grafting, RCTs were not able to demonstrate overall beneficial effects of perioperative Levosimendan in cardiac surgery. This might be due to Levosimendan's pharmacokinetics reaching a steady-state concentration only 4-8 h after administration. Thus, this study now analysed the influence of timing of Levosimendan administration on perioperative outcome in cardiac surgery patients preoperatively presenting with severely reduced ventricular function and therefore considered at high-risk for intra- or postoperative low cardiac output syndrome. We hypothesized that prolonged preoperative Levosimendan administration ("preconditioning") would reduce mortality. Methods: All adult patients undergoing cardiac surgery between 2006 and 2018 perioperatively receiving Levosimendan were included in this retrospective, observational cohort study (n = 498). Patients were stratified into 3 groups: Levosimendan started on the day prior to surgery ("preop"), Levosimendan started on the day of surgery ("intraop") or post ICU admission ("postop"). After propensity score matching (PSM) was performed, outcomes defined according to proposed standard definitions for perioperative outcome research were compared between groups. Results: After PSM, there were no significant differences in patients' characteristics, comorbidities and type/priority of surgery between groups. Compared to intraop or postop Levosimendan treatment, preop treated patients had significantly lower in-hospital-mortality (preop vs. intraop. vs. postop = 16,7% vs. 33,3% vs. 42,3%), duration of mechanical ventilation and rate of continuous renal replacement therapy. Conclusions: Prolonged preoperative treatment with Levosimendan of cardiac surgery patients preoperatively presenting with severely reduced left ventricular function might be beneficial in terms of postoperative outcome. Our results are in line with recent experts' recommendations concerning the prolonged perioperative use of Levosimendan. We strongly recommend that future randomized trials include this "preconditioning" treatment as an experimental arm.

Psychometric properties of the Reintegration to Normal Living Index for sepsis survivors.

PURPOSE: Return to a normal state of living is a key patient-relevant outcome for sepsis survivors. The Reintegration to Normal Living Index (RNLI) assesses self-perceived participation in patients with chronic disease, but its psychometric properties have been analyzed neither for patients after sepsis nor in a German patient cohort. This study aims to analyze the psychometric properties of the German version of the RNLI in sepsis survivors. METHODS: In a prospective multicenter survey study, 287 sepsis survivors were interviewed 6 and 12 months after hospital discharge. Multiple-group categorical confirmatory factor analyses with three competing models were used to explore the factor structure of the RNLI. Concurrent validity was evaluated in relation to the EQ-5D-3L and the Barthel Index of Activities of Daily Living (ADL). RESULTS: Regarding structural validity, all models showed an acceptable model fit. Because of high correlation between the latent variables in the two-factor models (up to r = 0.969) and for reason of parsimony, we opted for the common factor model to analyze the concurrent validity. Our analyses showed moderate positive correlations between RNLI score and ADL score (r ≥ 0.630), EQ-5D-3L visual analogue scale (r ≥ 0.656) and EQ-5D-3L utility score (r ≥ 0.548). The reliability assessed by McDonald's Omega was 0.94. CONCLUSION: We found convincing evidence for good reliability, structural and concurrent validity of the RNLI in German sepsis survivors. We propose to use the RNLI in addition to generic health-related quality of life measures to assess the reintegration to normal living after sepsis.

Pleural effusions are associated with adverse outcomes after cardiac surgery: a propensity-matched analysis.

BACKGROUND: Pleural effusions commonly occur in patients recovering from cardiac surgery; however, the impact on outcomes is not well characterized. The purpose of this study is to characterize the clinical outcomes of cardiac surgery patients with pleural effusion. METHODS: All patients undergoing cardiac surgery between 2006 and 2019 at a tertiary care university hospital were included in this observational, cross-sectional analysis using propensity matching. RESULTS: Of 11,037 patients that underwent cardiac surgery during the study period, 6461 (58.5%) had no pleural effusion (Group 0), 3322 (30.1%) had pleural effusion only (Group 1), and 1254 (11.4%) required at least one secondary drainage procedure after the index operation (Group 2). After propensity matching, the mortality of patients who underwent secondary drainage procedures was 6.1% higher than in Group 1 (p < 0.001). Intensive care unit (ICU) stay was longer for those with pleural effusions (18 [IQR 9-32] days in Group 2, 10 [IQR 6-17] days for Group 1, and 7 [IQR 4-11] days for Group 0, p < 0.001). Patients with pleural effusions had a higher incidence of hemodialysis (246 [20.0%] in Group 2, 137 [11.1%] in Group 1, 98 [7.98%] in Group 0), and a longer ventilation time in the ICU (57 [IQR 21.0-224.0] hours in Group 2, 25.0 [IQR 14.0-58.0] hours in Group 1, 16.0 [IQR 10.0-29.0] hours in Group 0). CONCLUSION: Pleural effusions, especially those that require a secondary drainage procedure during recovery, are associated with significantly worse outcomes including increased mortality, longer length of stay, and higher complication rates. These insights may be of great interest to scientists, clinicians, and industry leaders alike to foster research into innovative methods for preventing and treating pleural effusions with the aim of improving outcomes for patients recovering from cardiac surgery.

High Central Venous Pressure after Cardiac Surgery Might Depict Hemodynamic Deterioration Associated with Increased Morbidity and Mortality.

BACKGROUND: Cardiac surgery patients represent a high-risk cohort in intensive care units (ICUs). Central venous pressure (CVP) measurement seems to remain an integral part in hemodynamic monitoring, especially in cardio-surgical ICUs. However, its value as a prognostic marker for organ failure is still unclear. Therefore, we analyzed postoperative CVP values after adult cardiac surgery in a large cohort with regard to its prognostic value for morbidity and mortality. METHODS: All adult patients admitted to our ICUs between 2006 and 2019 after cardiac surgery were eligible for inclusion in the study (n = 11,198). We calculated the median initial CVP (miCVP) after admission to the ICU, which returned valid values for 9802 patients. An ROC curve analysis for optimal cut-off miCVP to predict ICU mortality was conducted with consecutive patient allocation into a (a) low miCVP (LCVP) group (≤11 mmHg) and (b) high miCVP (HCVP) group (>11 mmHg). We analyzed the impact of high miCVP on morbidity and mortality by propensity score matching (PSM) and logistic regression. RESULTS: ICU mortality was increased in HCVP patients. In addition, patients in the HCVP group required longer mechanical ventilation, had a higher incidence of acute kidney injury, were more frequently treated with renal replacement therapy, and showed a higher risk for postoperative liver dysfunction, parametrized by a postoperative rise of ≥ 10 in MELD Score. Multiple regression analysis confirmed HCVP has an effect on postoperative ICU-mortality and intrahospital mortality, which seems to be independent. CONCLUSIONS: A high initial CVP in the early postoperative ICU course after cardiac surgery is associated with worse patient outcome. Whether or not CVP, as a readily and constantly available hemodynamic parameter, should promote clinical efforts regarding diagnostics and/or treatment, warrants further investigations.

Machine learning identifies ICU outcome predictors in a multicenter COVID-19 cohort.

BACKGROUND: Intensive Care Resources are heavily utilized during the COVID-19 pandemic. However, risk stratification and prediction of SARS-CoV-2 patient clinical outcomes upon ICU admission remain inadequate. This study aimed to develop a machine learning model, based on retrospective & prospective clinical data, to stratify patient risk and predict ICU survival and outcomes. METHODS: A Germany-wide electronic registry was established to pseudonymously collect admission, therapeutic and discharge information of SARS-CoV-2 ICU patients retrospectively and prospectively. Machine learning approaches were evaluated for the accuracy and interpretability of predictions. The Explainable Boosting Machine approach was selected as the most suitable method. Individual, non-linear shape functions for predictive parameters and parameter interactions are reported. RESULTS: 1039 patients were included in the Explainable Boosting Machine model, 596 patients retrospectively collected, and 443 patients prospectively collected. The model for prediction of general ICU outcome was shown to be more reliable to predict "survival". Age, inflammatory and thrombotic activity, and severity of ARDS at ICU admission were shown to be predictive of ICU survival. Patients' age, pulmonary dysfunction and transfer from an external institution were predictors for ECMO therapy. The interaction of patient age with D-dimer levels on admission and creatinine levels with SOFA score without GCS were predictors for renal replacement therapy. CONCLUSIONS: Using Explainable Boosting Machine analysis, we confirmed and weighed previously reported and identified novel predictors for outcome in critically ill COVID-19 patients. Using this strategy, predictive modeling of COVID-19 ICU patient outcomes can be performed overcoming the limitations of linear regression models. Trial registration "ClinicalTrials" (clinicaltrials.gov) under NCT04455451.

Infectious Complications after Etomidate vs. Propofol for Induction of General Anesthesia in Cardiac Surgery-Results of a Retrospective, before-after Study.

BACKGROUND: Etomidate is typically used as an induction agent in cardiac surgery because it has little impact on hemodynamics. It is a known suppressor of adrenocortical function and may increase the risk for post-operative infections, sepsis, and mortality. The aim of this study was to evaluate whether etomidate increases the risk of postoperative sepsis (primary outcome) and infections (secondary outcome) compared to propofol. METHODS: This was a retrospective before-after trial (IRB EA1/143/20) performed at a tertiary medical center in Berlin, Germany, between 10/2012 and 01/2015. Patients undergoing cardiac surgery were investigated within two observation intervals, during which etomidate and propofol were the sole induction agents. RESULTS: One-thousand, four-hundred, and sixty-two patients, and 622 matched pairs, after caliper propensity-score matching, were included in the final analysis. Sepsis rates did not differ in the matched cohort (etomidate: 11.5% vs. propofol: 8.2%, p = 0.052). Patients in the etomidate interval were more likely to develop hospital-acquired pneumonia (etomidate: 18.6% vs. propofol: 14.0%, p = 0.031). CONCLUSION: Our study showed that a single-dose of etomidate is not statistically associated with higher postoperative sepsis rates after cardiac surgery, but is associated with a higher incidence of hospital-acquired pneumonia. However, there is a notable trend towards a higher sepsis rate.

Predicting lethal courses in critically ill COVID-19 patients using a machine learning model trained on patients with non-COVID-19 viral pneumonia.

In a pandemic with a novel disease, disease-specific prognosis models are available only with a delay. To bridge the critical early phase, models built for similar diseases might be applied. To test the accuracy of such a knowledge transfer, we investigated how precise lethal courses in critically ill COVID-19 patients can be predicted by a model trained on critically ill non-COVID-19 viral pneumonia patients. We trained gradient boosted decision tree models on 718 (245 deceased) non-COVID-19 viral pneumonia patients to predict individual ICU mortality and applied it to 1054 (369 deceased) COVID-19 patients. Our model showed a significantly better predictive performance (AUROC 0.86 [95% CI 0.86-0.87]) than the clinical scores APACHE2 (0.63 [95% CI 0.61-0.65]), SAPS2 (0.72 [95% CI 0.71-0.74]) and SOFA (0.76 [95% CI 0.75-0.77]), the COVID-19-specific mortality prediction models of Zhou (0.76 [95% CI 0.73-0.78]) and Wang (laboratory: 0.62 [95% CI 0.59-0.65]; clinical: 0.56 [95% CI 0.55-0.58]) and the 4C COVID-19 Mortality score (0.71 [95% CI 0.70-0.72]). We conclude that lethal courses in critically ill COVID-19 patients can be predicted by a machine learning model trained on non-COVID-19 patients. Our results suggest that in a pandemic with a novel disease, prognosis models built for similar diseases can be applied, even when the diseases differ in time courses and in rates of critical and lethal courses.

Improvements in Patient Monitoring in the Intensive Care Unit: Survey Study.

BACKGROUND: Due to demographic change and, more recently, coronavirus disease (COVID-19), the importance of modern intensive care units (ICU) is becoming apparent. One of the key components of an ICU is the continuous monitoring of patients' vital parameters. However, existing advances in informatics, signal processing, or engineering that could alleviate the burden on ICUs have not yet been applied. This could be due to the lack of user involvement in research and development. OBJECTIVE: This study focused on the satisfaction of ICU staff with current patient monitoring and their suggestions for future improvements. We aimed to identify aspects of monitoring that interrupt patient care, display devices for remote monitoring, use cases for artificial intelligence (AI), and whether ICU staff members are willing to improve their digital literacy or contribute to the improvement of patient monitoring. We further aimed to identify differences in the responses of different professional groups. METHODS: This survey study was performed with ICU staff from 4 ICUs of a German university hospital between November 2019 and January 2020. We developed a web-based 36-item survey questionnaire, by analyzing a preceding qualitative interview study with ICU staff, about the clinical requirements of future patient monitoring. Statistical analyses of questionnaire results included median values with their bootstrapped 95% confidence intervals, and chi-square tests to compare the distributions of item responses of the professional groups. RESULTS: In total, 86 of the 270 ICU physicians and nurses completed the survey questionnaire. The majority stated they felt confident using the patient monitoring equipment, but that high rates of false-positive alarms and the many sensor cables interrupted patient care. Regarding future improvements, respondents asked for wireless sensors, a reduction in the number of false-positive alarms, and hospital standard operating procedures for alarm management. Responses to the display devices proposed for remote patient monitoring were divided. Most respondents indicated it would be useful for earlier alerting or when they were responsible for multiple wards. AI for ICUs would be useful for early detection of complications and an increased risk of mortality; in addition, the AI could propose guidelines for therapy and diagnostics. Transparency, interoperability, usability, and staff training were essential to promote the use of AI. The majority wanted to learn more about new technologies for the ICU and required more time for learning. Physicians had fewer reservations than nurses about AI-based intelligent alarm management and using mobile phones for remote monitoring. CONCLUSIONS: This survey study of ICU staff revealed key improvements for patient monitoring in intensive care medicine. Hospital providers and medical device manufacturers should focus on reducing false alarms, implementing hospital alarm standard operating procedures, introducing wireless sensors, preparing for the use of AI, and enhancing the digital literacy of ICU staff. Our results may contribute to the user-centered transfer of digital technologies into practice to alleviate challenges in intensive care medicine. TRIAL REGISTRATION: ClinicalTrials.gov NCT03514173; https://clinicaltrials.gov/ct2/show/NCT03514173.

Indications for cellular migration from the central nervous system to its draining lymph nodes in CD11c-GFP+ bone-marrow chimeras following EAE.

The concept as to how the brain maintains its immune privilege has initially been based on observations that it is lacking classical lymph vessels and later, the absence of dendritic cells (DC). This view has been challenged by several groups demonstrating drainage/migration of injected tracers and cells into cervical lymph nodes (CLNs) and the presence of brain antigens in CLNs in the course of various brain pathologies. Using CD11c-diphtheria toxin receptor (DTR)-green fluorescent protein (GFP) transgenic (tg) mice, we have shown the existence of CD11c+ cells, a main DC marker, within the brain parenchyma. Since injecting tracers or cells may cause barrier artefacts, we have now transplanted wild type (wt)-bone marrow (BM) to lethally irradiated CD11c-DTR-GFP tg mice to restrict the CD11c-DTR-GFP+ population to the brain and induced experimental autoimmune encephalomyelitis (EAE), an animal model of multiple sclerosis (MS). We observed ramified GFP+ cells in the olfactory bulb, the cribriform plate, the nasal mucosa and superficial CLNs. We measured a significant increase of host gfp genomic DNA (gDNA) levels in lymph nodes (LNs) previously described as draining stations for the central nervous system (CNS). Using flow cytometry analysis, we observed an increase of the percentage of CD11c-GFP+ cells in brain parenchyma in the course of EAE which is most likely due to an up-regulation of CD11c of resident microglial cells since levels of gfp gDNA did not increase. Our data supports the hypothesis that brain-resident antigen presenting cells (APC) are capable of migrating to CNS-draining LNs to present myelin-associated epitopes.

CD11c-positive cells from brain, spleen, lung, and liver exhibit site-specific immune phenotypes and plastically adapt to new environments.

The brain's immune privilege has been also attributed to the lack of dendritic cells (DC) within its parenchyma and the adjacent meninges, an assumption, which implies maintenance of antigens rather than their presentation in lymphoid organs. Using mice transcribing the green fluorescent protein under the promoter of the DC marker CD11c (itgax), we identified a juxtavascular population of cells expressing this DC marker and demonstrated their origin from bone marrow and local microglia. We now phenotypically compared this population with CD11c/CD45 double-positive cells from lung, liver, and spleen in healthy mice using seven-color flow cytometry. We identified unique, site-specific expression patterns of F4/80, CD80, CD86, CX3CR1, CCR2, FLT3, CD103, and MHC-II. Furthermore, we observed the two known CD45-positive populations (CD45(high) and CD45(int) ) in the brain, whereas liver, lung, and spleen exhibited a homogeneous CD45(high) population. CD11c-positive microglia lacked MHC-II expression and CD45(high) /CD11c-positive cells from the brain have a lower percentage of MHC-II-positive cells. To test whether phenotypical differences are fixed by origin or specifically develop due to environmental factors, we transplanted brain and spleen mononuclear cells on organotypic slice cultures from brain (OHSC) and spleen (OSSC). We demonstrate that adaption and ramification of MHC-II-positive splenocytes is paralleled by down-regulation of MHC-II, whereas brain-derived mononuclear cells neither ramified nor up-regulated MHC-II in OSSCs. Thus, brain-derived mononuclear cells maintain their MHC-II-negative phenotype within the environment of an immune organ. Intraparenchymal CD11c-positive cells share immunophenotypical characteristics of DCs from other organs but remain unique for their low MHC-II expression.

CD11c-expressing cells reside in the juxtavascular parenchyma and extend processes into the glia limitans of the mouse nervous system.

Recent studies demonstrated that primary immune responses can be induced within the brain depending on vessel-associated cells expressing markers of dendritic cells (DC). Using mice transcribing the green fluorescent protein (GFP) under the promoter of the DC marker CD11c, we determined the distribution, phenotype, and source of CD11c+ cells in non-diseased brains. Predilection areas of multiple sclerosis (MS) lesions (periventricular area, adjacent fibre tracts, and optical nerve) were preferentially populated by CD11c+ cells. Most CD11c+ cells were located within the juxtavascular parenchyma rather than the perivascular spaces. Virtually all CD11c+ cells co-expressed ionized calcium-binding adaptor molecule 1 (IBA-1), CD11b, while detectable levels of major histocompatibility complex II (MHC-II) in non-diseased mice was restricted to CD11c+ cells of the choroid plexus. Cellular processes project into the glia limitans which may allow transport and/or presentation of intraparenchymal antigens to extravasated T cells in perivascular spaces. In chimeric mice bearing CD11c-GFP bone marrow, fluorescent cells appeared in the CNS between 8 and 12 weeks after transplantation. In organotypic slice cultures from CD11c-GFP mice, the number of fluorescent cells strongly increased within 72 h. Strikingly, using anti-CD209, an established marker for human DC, a similar population was detected in human brains. Thus, we show for the first time that CD11c+ cells can not only be recruited from the blood into the parenchyma, but also develop from an intraneural precursor in situ. Dysbalance in their recruitment/development may be an initial step in the pathogenesis of chronic (autoimmune) neuroinflammatory diseases such as MS.