Incidence, risk factors and prediction of post-operative acute kidney injury following cardiac surgery for active infective endocarditis: an observational study

Effects of dexmedetomidine on renal function after cardiac surgery for infective endocarditis: An interim analysis of a randomized controlled trial

BACKGROUND

Patients undergoing cardiac surgery for infective endocarditis (IE) are at a high risk of postoperative acute kidney injury (AKI) owing to heightened systemic inflammation. Therefore, we aimed to investigate the effect of dexmedetomidine on postoperative AKI in patients who underwent cardiac surgery for IE.

METHODS

A total of 63 patients who underwent cardiac surgery for IE were randomly assigned to receive either intravenous dexmedetomidine infusion of 0.4 μg kg-1 h-1 (DEX group) or normal saline infusion (control group) for 24 h after induction of anesthesia. The occurrence of AKI within seven days postoperation, epinephrine, norepinephrine, and interleukin-6 levels, as well as postoperative morbidities, were assessed. An intertrim analysis was conducted using Pocock's alpha spending function at α = 0.05 and β = 0.2.

RESULTS

This trial was early terminated according to the results of interim analysis performed when 60 % of the pre-set number of patients have been collected. The incidence of AKI was significantly lower in the DEX group than in the control group (32.3 % vs. 9.4 %, p = 0.025). Patients in the DEX group had significantly lower epinephrine levels than those in the control group, whereas norepinephrine and interleukin-6 levels were similar. Perioperative mean arterial pressure or heart rate did not differ between the groups.

CONCLUSIONS

Dexmedetomidine administration for 24 h starting from induction of anesthesia significantly reduced the incidence of postoperative AKI after cardiac surgery for IE (by 29 % vs. control) without hemodynamic side effects. This was accompanied by a significant attenuation of postoperative increase in serum epinephrine levels.

Infective Endocarditis by Lactobacillus Species-A Narrative Review

Bacteria of the genus Lactobacillus are microaerophilic or aerotolerant anaerobic Gram-positive non-spore-forming rods. They are considered essential members of the human gut microbiome; however, recent studies have revealed that these microorganisms are less predominant in the gut microbiome than initially thought. Lactobacillus spp. is mainly known for its use as a probiotic in foods and supplements to prevent and treat specific issues such as infectious diseases, irritable bowel syndrome, and diabetes mellitus. However, Lactobacillus spp. may occasionally cause infections such as bacteremia or infective endocarditis (IE). The present study aimed to review all cases of IE by Lactobacillus spp. and describe the epidemiology, microbiology, clinical characteristics, treatment, and outcomes of this infection by collecting relevant data from studies existing in Pubmed and Scopus until 28 September 2023. A total of 77 studies containing data for 82 patients were included. The median age was 56 years, and 69.6% were male. A prosthetic valve was present in 16% of patients, and 17.3% had previously been on probiotics. The aortic valve was the most commonly involved intracardiac site, followed by the mitral valve. Fever, embolic phenomena, sepsis, and heart failure were the most common clinical presentations. Aminoglycosides and penicillin were the most commonly used antimicrobials for definitive treatment. Surgery was performed in 53.7% of patients. Overall mortality was 17.1%. IE in prosthetic valves and presentation with shock were independently associated with overall mortality.

Vancomycin associated acute kidney injury in patients with infectious endocarditis: a large retrospective cohort study

Background: Vancomycin remains the cornerstone antibiotic for the treatment of infective endocarditis (IE). Vancomycin has been associated with significant nephrotoxicity. However, vancomycin associated acute kidney injury (AKI) has not been evaluated in patients with IE. We conducted this large retrospective cohort study to reveal the incidence, risk factors, and prognosis of vancomycin-associated acute kidney injury (VA-AKI) in patients with IE. Methods: Adult patients diagnosed with IE and receiving vancomycin were included. The primary outcome was VA-AKI. Results: In total, 435 of the 600 patients were enrolled. Of these, 73.6% were male, and the median age was 52 years. The incidence of VA-AKI was 17.01% (74). Only 37.2% (162) of the patients received therapeutic monitoring of vancomycin, and 30 (18.5%) patients had reached the target vancomycin trough concentration. Multiple logistic regression analysis revealed that body mass index [odds ratio (OR) 1.088, 95% CI 1.004, 1.179], duration of vancomycin therapy (OR 1.030, 95% CI 1.003, 1.058), preexisting chronic kidney disease (OR 2.291, 95% CI 1.018, 5.516), admission to the intensive care unit (OR 2.291, 95% CI 1.289, 3.963) and concomitant radiocontrast agents (OR 2.085, 95% CI 1.093, 3.978) were independent risk factors for VA-AKI. Vancomycin variety (Lai Kexin vs. Wen Kexin, OR 0.498, 95% CI 0.281, 0.885) were determined to be an independent protective factor for VI-AKI. Receiver operator characteristic curve analysis revealed that duration of therapy longer than 10.75 days was associated with a significantly increased risk of VA-AKI (HR 1.927). Kidney function was fully or partially recovered in 73.0% (54) of patients with VA-AKI. Conclusion: The incidence of VA-AKI in patients with IE was slightly higher than in general adult patients. Concomitant contrast agents were the most alarmingly nephrotoxic in patients with IE, adding a 2-fold risk of VA-AKI. In patients with IE, a course of vancomycin therapy longer than 10.75 days was associated with a significantly increased risk of AKI. Thus, closer monitoring of kidney function and vancomycin trough concentrations was recommended in patients with concurrent contrast or courses of vancomycin longer than 10.75 days.

Impact of Intermediate Susceptibility to Penicillin on Antimicrobial Treatment and Outcomes of Endocarditis Caused by Viridans and Gallolyticus Group Streptococci

BACKGROUND

Evidence supporting combination treatment with a beta-lactam plus an aminoglycoside (C-BA) for endocarditis caused by viridans and gallolyticus group streptococci (VGS-GGS) with intermediate susceptibility to penicillin (PENI-I) is lacking. We assessed the clinical characteristics and outcomes of PEN-I VGS-GGS endocarditis and compared the effectiveness and safety of C-BA with third-generation cephalosporin monotherapy.

METHODS

Retrospective analysis of prospectively collected data of a cohort of definite endocarditis caused by penicillin-susceptible and PENI-I VGS-GGS (penicillin minimum inhibitory concentration ranging from 0.25 to 2 mg/L) between 2008 and 2018 in 40 Spanish hospitals. We compared cases treated with monotherapy or with C-BA and performed multivariable analyses of risk factors for in-hospital and 1-year mortality.

RESULTS

A total of 914 consecutive cases of definite endocarditis caused by VGS-GGS with complete or intermediate susceptibility to penicillin were included. A total of 688 (75.3%) were susceptible to penicillin and 226 (24.7%) were PENI-I. Monotherapy was used in 415 (45.4%) cases (cephalosporin in 331 cases) and 499 (54.6%) cases received C-BA. In-hospital mortality was 11.9%, and 190 (20.9%) patients developed acute kidney injury. Heart failure (odds ratio [OR]: 6.06; 95% confidence interval [CI]: 1.37-26.87; P = .018), central nervous system emboli (OR: 9.83; 95% CI: 2.17-44.49; P = .003) and intracardiac abscess (OR: 13.47; 95% CI: 2.24-81.08; P = .004) were independently associated with in-hospital mortality among PEN-I VGS-GGS cases, while monotherapy was not (OR: 1.01; 95% CI: .26-3.96; P = .982).

CONCLUSIONS

Our findings support the use of cephalosporin monotherapy in PEN-I VGS-GGS endocarditis in order to avoid nephrotoxicity without adversely affecting patient outcomes.

Application of targeted maximum likelihood estimation in public health and epidemiological studies: a systematic review

PURPOSE

The targeted maximum likelihood estimation (TMLE) statistical data analysis framework integrates machine learning, statistical theory, and statistical inference to provide a least biased, efficient, and robust strategy for estimation and inference of a variety of statistical and causal parameters. We describe and evaluate the epidemiological applications that have benefited from recent methodological developments.

METHODS

We conducted a systematic literature review in PubMed for articles that applied any form of TMLE in observational studies. We summarized the epidemiological discipline, geographical location, expertize of the authors, and TMLE methods over time. We used the Roadmap of Targeted Learning and Causal Inference to extract key methodological aspects of the publications. We showcase the contributions to the literature of these TMLE results.

RESULTS

Of the 89 publications included, 33% originated from the University of California at Berkeley, where the framework was first developed by Professor Mark van der Laan. By 2022, 59% of the publications originated from outside the United States and explored up to seven different epidemiological disciplines in 2021-2022. Double-robustness, bias reduction, and model misspecification were the main motivations that drew researchers toward the TMLE framework. Through time, a wide variety of methodological, tutorial, and software-specific articles were cited, owing to the constant growth of methodological developments around TMLE.

CONCLUSIONS

There is a clear dissemination trend of the TMLE framework to various epidemiological disciplines and to increasing numbers of geographical areas. The availability of R packages, publication of tutorial papers, and involvement of methodological experts in applied publications have contributed to an exponential increase in the number of studies that understood the benefits and adoption of TMLE.

Perioperative Complications in Infective Endocarditis

Infective endocarditis is a challenging condition to manage, requiring collaboration among various medical professionals. Interdisciplinary teamwork within endocarditis teams is essential. About half of the patients diagnosed with the disease will ultimately have to undergo cardiac surgery. As a result, it is vital for all healthcare providers involved in the perioperative period to have a comprehensive understanding of the unique features of infective endocarditis, including clinical presentation, echocardiographic signs, coagulopathy, bleeding control, and treatment of possible organ dysfunction. This narrative review provides a summary of the current knowledge on the incidence of complications and their management in the perioperative period in patients with infective endocarditis.

Characterization of Risk Prediction Models for Acute Kidney Injury: A Systematic Review and Meta-analysis

Importance

Despite the expansion of published prediction models for acute kidney injury (AKI), there is little evidence of uptake of these models beyond their local derivation nor data on their association with patient outcomes.

Objective

To systematically review published AKI prediction models across all clinical subsettings.

Data Sources

MEDLINE via PubMed (January 1946 to April 2021) and Embase (January 1947 to April 2021) were searched using medical subject headings and text words related to AKI and prediction models.

Study Selection

All studies that developed a prediction model for AKI, defined as a statistical model with at least 2 predictive variables to estimate future occurrence of AKI, were eligible for inclusion. There was no limitation on study populations or methodological designs.

Data Extraction and Synthesis

Two authors independently searched the literature, screened the studies, and extracted and analyzed the data following the Preferred Reporting Items for Systematic Review and Meta-analyses guideline. The data were pooled using a random-effects model, with subgroups defined by 4 clinical settings. Between-study heterogeneity was explored using multiple methods, and funnel plot analysis was used to identify publication bias.

Main Outcomes and Measures

C statistic was used to measure the discrimination of prediction models.

Results

Of the 6955 studies initially identified through literature searching, 150 studies, with 14.4 million participants, met the inclusion criteria. The study characteristics differed widely in design, population, AKI definition, and model performance assessments. The overall pooled C statistic was 0.80 (95% CI, 0.79-0.81), with pooled C statistics in different clinical subsettings ranging from 0.78 (95% CI, 0.75-0.80) to 0.82 (95% CI, 0.78-0.86). Between-study heterogeneity was high overall and in the different clinical settings (eg, contrast medium-associated AKI: I2 = 99.9%; P < .001), and multiple methods did not identify any clear sources. A high proportion of models had a high risk of bias (126 [84.4%]) according to the Prediction Model Risk Of Bias Assessment Tool.

Conclusions and Relevance

In this study, the discrimination of the published AKI prediction models was good, reflected by high C statistics; however, the wide variation in the clinical settings, populations, and predictive variables likely drives the highly heterogenous findings that limit clinical utility. Standardized procedures for development and validation of prediction models are urgently needed.

Effect of intraoperative haemoadsorption therapy on cardiac surgery for active infective endocarditis with confirmed Staphylococcus aureus bacteraemia

OBJECTIVES

Sepsis caused by infective endocarditis (IE), due to Staphylococcus aureus, is associated with significant morbidity and mortality. Blood purification using haemoadsorption (HA) may attenuate the inflammatory response. We investigated the effect of intraoperative HA on postoperative outcomes in S. aureus IE.

METHODS

Patients with confirmed S. aureus IE undergoing cardiac surgery were included in a dual-centre study between January 2015 and March 2022. Patients treated with intraoperative HA (HA group) were compared to patients not treated with HA (control group). The primary outcome was vasoactive-inotropic score within the first 72 h postoperatively and secondary outcomes were sepsis-related mortality (SEPSIS-3 definition) and overall mortality at 30 and 90 days.

RESULTS

No differences in baseline characteristics were observed between groups (haemoadsorption group, n = 75, control group, n = 55). Significantly decreased vasoactive-inotropic score was observed in the haemoadsorption group at all time points [6 h: 6.0 (0-17) vs 17 (3-47), P = 0.0014; 12 h: 2 (0-8.3) vs 5.9 (0-37), P = 0.0138; 24 h: 0 (0-5) vs 4.9 (0-23), P = 0.0064; 48 h: 0 (0-2.1) vs 0.1 (0-13), P = 0.0192; 72 h: 0 (0) vs 0 (0-5), P = 0.0014]. Importantly, sepsis-related mortality (8.0% vs 22.8%, P = 0.02) and 30-day (17.3% vs 32.7%, P = 0.03) and 90-day overall mortality (21.3% vs 40%, P = 0.03) were also significantly lower with haemoadsorption.

CONCLUSIONS

Intraoperative HA during cardiac surgery for S. aureus IE was associated with significantly lower postoperative vasopressor and inotropic requirements and resulted in lower sepsis-related and overall 30- and 90-day mortality. In this high-risk population, improved postoperative haemodynamic stabilization by intraoperative HA appears to improve survival and should be further tested in future randomized trials.

High Incidence of Acute Kidney Injury in Patients Treated with High-Dose Amoxicillin and Cloxacillin Combination Therapy

High-dose amoxicillin and cloxacillin combination therapy is recommended for the empiric treatment of selected patients with infective endocarditis despite a low level of evidence. The main objective of this study was to evaluate the renal tolerance of high-dose intravenous amoxicillin and cloxacillin combination. We studied 27 patients treated with amoxicillin and cloxacillin (≥100 mg/kg daily) for at least 48 h. The primary endpoint was the occurrence of acute kidney injury (AKI). The median patient age was 68 ± 8 years, and 16 (59%) were male. The indication for this combination therapy was suspected or confirmed endocarditis with no bacterial identification in 22 (81%) patients. The primary endpoint occurred in 16 (59%) patients after initiating this combination therapy within an average of 4.4 ± 3.6 days. Among them, seven (26%) patients developed severe AKI, including four (15%) patients who required hemodialysis. Other risk factors for AKI were identified in all patients, including injection of iodinated contrast media in 21 (78%), acute heart failure in 18 (67%), cardiac surgery in 11 (41%), and aminoglycoside use in 9 (33%) patients. This study reports an incidence of 59% of AKI after initiating amoxicillin and cloxacillin combination therapy in a population at high renal risk.

Prediction of acute kidney injury risk after cardiac surgery: using a hybrid machine learning algorithm

Background

Acute kidney injury (AKI) is a serious complication after cardiac surgery. We derived and internally validated a Machine Learning preoperative model to predict cardiac surgery-associated AKI of any severity and compared its performance with parametric statistical models.

Methods

We conducted a retrospective study of adult patients who underwent major cardiac surgery requiring cardiopulmonary bypass between November 1st, 2009 and March 31st, 2015. AKI was defined according to the KDIGO criteria as stage 1 or greater, within 7 days of surgery. We randomly split the cohort into derivation and validation datasets. We developed three AKI risk models: (1) a hybrid machine learning (ML) algorithm, using Random Forests for variable selection, followed by high performance logistic regression; (2) a traditional logistic regression model and (3) an enhanced logistic regression model with 500 bootstraps, with backward variable selection. For each model, we assigned risk scores to each of the retained covariate and assessed model discrimination (C statistic) and calibration (Hosmer–Lemeshow goodness-of-fit test) in the validation datasets.

Results

Of 6522 included patients, 1760 (27.0%) developed AKI. The best performance was achieved by the hybrid ML algorithm to predict AKI of any severity. The ML and enhanced statistical models remained robust after internal validation (C statistic = 0.75; Hosmer–Lemeshow p = 0.804, and AUC = 0.74, Hosmer–Lemeshow p = 0.347, respectively).

Conclusions

We demonstrated that a hybrid ML model provides higher accuracy without sacrificing parsimony, computational efficiency, or interpretability, when compared with parametric statistical models. This score-based model can easily be used at the bedside to identify high-risk patients who may benefit from intensive perioperative monitoring and personalized management strategies.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12911-022-01859-w.

Anesthesia Considerations in Infective Endocarditis

The management of infective endocarditis is complex and inherently requires multidisciplinary cooperation. About half of all patients diagnosed with infective endocarditis will meet the criteria to undergo cardiac surgery, which regularly takes place in urgent or emergency settings. The pathophysiology and clinical presentation of infective endocarditis make it a unique disorder within cardiac surgery that warrants a thorough understanding of specific characteristics in the perioperative period. This includes, among others, echocardiography, coagulation, bleeding management, or treatment of organ dysfunction. In this narrative review article, the authors summarize the current knowledge on infective endocarditis relevant for the clinical anesthesiologist in perioperative management of respective patients. Furthermore, the authors advocate for the anesthesiologist to become a structural member of the endocarditis team.

Outcomes of surgically treated infective endocarditis in a Western Australian population

Background

Infective endocarditis is a disease that carries high morbidity and mortality. The primary endpoint of this study is to assess factors associated with in-hospital mortality in patients undergoing valvular surgery for infective endocarditis. The secondary endpoint of this study is to assess the incidence of post-operative stroke, renal failure, complete heart block and recurrence.

Methods

Between the years of 2015 to 2019, a total of 89 patients underwent surgery for infective endocarditis at Fiona Stanley Hospital, Western Australia. Data was collected from the Australia and New Zealand Cardiac Surgery Database from 2015 to 2019 as well as patients electronic medical record. A number of preoperative and perioperative factors were assessed in relation to patient mortality and morbidity. Univariate and multivariate logistical regression analysis was done to assess for the association between factors and in-hospital morbidity and mortality.

Results

A total of 89 patients underwent surgery for infective endocarditis from 2015 to 2019, affecting a total of 101 valves. The mean age of patients was 53.7 ± 16.5. A total of 79 patients had a positive blood culture pre-operatively, with Staphylococcus Aureus being the most frequently cultured organism (39%). Fourteen patients (16%) were deemed emergent and underwent surgery within 24 h of review. A total of five patients died within their hospital stay postoperatively. Variables significantly associated with mortality on univariate analysis were intravenous drug use, emergent surgery, perioperative dialysis, perioperative inotropes, cardiopulmonary bypass time and cross clamp time. Only CBP time was significantly associated with mortality on multivariate analysis. A total of 19 patients (21%) required hemodialysis after surgery, 10 patients sustained a postoperative stroke (11%), 11 patients developed a complete heart block post operatively (12%) and endocarditis recurred in 10 patients (11%).

Conclusion

Prolonged cardiopulmonary bypass times were significantly associated with mortality. This study is novel to report a lower mortality rate than previously quoted in the literature. We also report our findings of organisms, preoperative embolic phenomena and surgery in a Western Australian population. We recommend that all patients with endocarditis are discussed in multidisciplinary forum.

Outcome of Dialysis-Requiring Acute Kidney Injury in Patients With Infective Endocarditis: A Nationwide Study

BACKGROUND

Infective endocarditis (IE) may be complicated by acute kidney injury, yet data on the use of dialysis and subsequent reversibility are sparse.

METHODS

Using Danish nationwide registries, we identified patients with first-time IE from 2000 to 2017. Dialysis-naïve patients were grouped into: those with and those without dialysis during admission with IE. Continuation of dialysis was followed 1 year postdischarge. Multivariable adjusted Cox proportional hazard analysis was used to examine 1-year mortality for patients surviving IE according to use of dialysis.

RESULTS

We included 7307 patients with IE; 416 patients (5.7%) initiated dialysis treatment during admission with IE and these were younger, had more comorbidities and more often underwent cardiac valve surgery compared with nondialysis patients (47.4% vs 20.9%). In patients with both cardiac valve surgery and dialysis treatment (n = 197), 153 (77.7%) initiated dialysis on or after the date of surgery. The in-hospital mortality was 40.4% and 19.0% for patients with and without dialysis, respectively (P < .0001). Of those who started dialysis and survived hospitalization, 21.6% continued dialysis treatment within 1 year after discharge. In multivariable adjusted analysis, dialysis during admission with IE was associated with an increased 1-year mortality from IE discharge, hazard ratio = 1.64 (95% confidence interval, 1.21-2.23).

CONCLUSION

In dialysis-naïve patients with IE, approximately 1 in 20 patients initiated dialysis treatment during admission with IE. Dialysis identified a high-risk group with an in-hospital mortality of 40% and an approximate 20% risk of continued dialysis. Those with dialysis during admission with IE showed worse long-term outcomes than those without.

Machine Learning for the Prediction of Red Blood Cell Transfusion in Patients During or After Liver Transplantation Surgery

Aim: This study aimed to use machine learning algorithms to identify critical preoperative variables and predict the red blood cell (RBC) transfusion during or after liver transplantation surgery. Study Design and Methods: A total of 1,193 patients undergoing liver transplantation in three large tertiary hospitals in China were examined. Twenty-four preoperative variables were collected, including essential population characteristics, diagnosis, symptoms, and laboratory parameters. The cohort was randomly split into a train set (70%) and a validation set (30%). The Recursive Feature Elimination and eXtreme Gradient Boosting algorithms (XGBOOST) were used to select variables and build machine learning prediction models, respectively. Besides, seven other machine learning models and logistic regression were developed. The area under the receiver operating characteristic (AUROC) was used to compare the prediction performance of different models. The SHapley Additive exPlanations package was applied to interpret the XGBOOST model. Data from 31 patients at one of the hospitals were prospectively collected for model validation. Results: In this study, 72.1% of patients in the training set and 73.2% in the validation set underwent RBC transfusion during or after the surgery. Nine vital preoperative variables were finally selected, including the presence of portal hypertension, age, hemoglobin, diagnosis, direct bilirubin, activated partial thromboplastin time, globulin, aspartate aminotransferase, and alanine aminotransferase. The XGBOOST model presented significantly better predictive performance (AUROC: 0.813) than other models and also performed well in the prospective dataset (accuracy: 76.9%). Discussion: A model for predicting RBC transfusion during or after liver transplantation was successfully developed using a machine learning algorithm based on nine preoperative variables, which could guide high-risk patients to take appropriate preventive measures.

Using interpretability approaches to update "black-box" clinical prediction models: an external validation study in nephrology

Despite advances in machine learning-based clinical prediction models, only few of such models are actually deployed in clinical contexts. Among other reasons, this is due to a lack of validation studies. In this paper, we present and discuss the validation results of a machine learning model for the prediction of acute kidney injury in cardiac surgery patients initially developed on the MIMIC-III dataset when applied to an external cohort of an American research hospital. To help account for the performance differences observed, we utilized interpretability methods based on feature importance, which allowed experts to scrutinize model behavior both at the global and local level, making it possible to gain further insights into why it did not behave as expected on the validation cohort. The knowledge gleaned upon derivation can be potentially useful to assist model update during validation for more generalizable and simpler models. We argue that interpretability methods should be considered by practitioners as a further tool to help explain performance differences and inform model update in validation studies.

Risk-benefit Assessment of Systematic Thoracoabdominal-pelvic Computed Tomography in Infective Endocarditis

BACKGROUND

In the management of infective endocarditis (IE), the presence of extracardiac complications has an influence on both diagnosis and treatment. Current guidelines suggest that systematic thoracoabdominal-pelvic computed tomography (TAP-CT) may be helpful. Our objective was to describe how systematic TAP-CT affects the diagnosis and the management of IE.

METHODS

In this multicenter cohort study, between January 2013 and July 2016 we included consecutive patients who had definite or possible IE according to the Duke modified criteria, validated by endocarditis teams. We analyzed whether the Duke classification and therapeutic management were modified regarding the presence or the absence of IE-related lesion on CT and investigated the tolerance of this examination.

RESULTS

Of the 522 patients included in this study, 217 (41.6%) had 1 or more IE-related lesions. On the basis of CT results in asymptomatic patients, diagnostic classification was upgraded from possible endocarditis to definite endocarditis for only 4 cases (0.8%). The presence of IE-related lesions on CT did not modify the duration of antibiotic treatment (P = .55), nor the decision of surgical treatment (P = .39). Specific treatment of the lesion was necessary in 42 patients (8.0%), but only 9 of these lesions (1.9%) were asymptomatic and diagnosed only on the TAP-CT. Acute kidney injury (AKI) within 5 days of CT was observed in 78 patients (14.9%).

CONCLUSIONS

The TAP-CT findings slightly affected diagnosis and treatment of IE in a very small proportion of asymptomatic patients. Furthermore, contrast media should be used with caution because of the high risk of AKI.

Defining an Intraoperative Hypotension Threshold in Association with De Novo Renal Replacement Therapy after Cardiac Surgery

BACKGROUND

Acute kidney injury (AKI) is a frequent and deadly complication after cardiac surgery. In the absence of effective therapies, a focus on risk factor identification and modification has been the mainstay of management. The authors sought to determine the impact of intraoperative hypotension on de novo postoperative renal replacement therapy in patients undergoing cardiac surgery, hypothesizing that prolonged periods of hypotension during and after cardiopulmonary bypass (CPB) were associated with an increased risk of renal replacement therapy.

METHODS

Included in this single-center retrospective cohort study were adult patients who underwent cardiac surgery requiring CPB between November 2009 and April 2015. Excluded were patients who were dialysis dependent, underwent thoracic aorta or off-pump procedures, or died before receiving renal replacement therapy. Degrees of hypotension were defined by mean arterial pressure (MAP) as less than 55, 55 to 64, and 65 to 74 mmHg before, during, and after CPB. The primary outcome was de novo renal replacement therapy.

RESULTS

Of 6,523 patient records, 336 (5.2%) required new postoperative renal replacement therapy. Each 10-min epoch of MAP less than 55 mmHg post-CPB was associated with an adjusted odds ratio of 1.13 (95% CI, 1.05 to 1.23; P = 0.002), and each 10-min epoch of MAP between 55 and 64 mmHg post-CPB was associated with an adjusted odds ratio of 1.12 (95% CI, 1.06 to 1.18; P = 0.0001) for renal replacement therapy. The authors did not observe an association between hypotension before and during CPB with renal replacement therapy.

CONCLUSIONS

MAP less than 65 mmHg for 10 min or more post-CPB is associated with an increased risk of de novo postoperative renal replacement therapy. The association between intraoperative hypotension and AKI was weaker in comparison to factors such as renal insufficiency, heart failure, obesity, anemia, complex or emergent surgery, and new-onset postoperative atrial fibrillation. Nonetheless, post-CPB hypotension is a potentially easier modifiable risk factor that warrants further investigation.

Risk factors and outcomes of early acute kidney injury in infective endocarditis: A retrospective cohort study

OBJECTIVES

The incidence of acute kidney injury (AKI) in infective endocarditis (IE), its risk factors and consequences on patient and renal survival remain debated.

METHODS

Patients hospitalized for a first episode of IE (possible or definite according to modified Duke criteria) between 2013 and 2016 were included. The primary endpoint was to determine risk factors for early AKI (E-AKI) during the first week of management of IE.

RESULTS

A total of 276 patients were included: 220 (79.7%) had definite IE and 56 (20.3%) had possible IE. E-AKI occurred in 150 patients (53%). IE due to Staphylococcus aureus (OR 3.41; 95% CI 1.83-6.39; p<0.01), history of diabetes (OR 2.34; 95% CI 1.25-4.37; p<0.01), peripheral arterial disease (OR 2.59; 95% CI 1.07-6.23; p<0.05), immunological manifestations (OR 3.11; 95% CI 1.31-7.39; p=0.01), and use of norepinephrine (OR 3.44; 95% CI 1.72-7.02; p<0.01) were associated with E-AKI. In subgroup analysis, infectious disease consultation was associated with a lower risk of AKI at day 7 (OR 0.41; 95% CI 0.16-0.88; p=0.04). E-AKI was associated with 1-year mortality (OR 1.65; 95% CI 1.03-2.64; p=0.04) and chronic kidney disease progression (OR 2.23; 95% CI 1.30-3.82; p<0.01).

CONCLUSIONS

E-AKI is common in IE and often associated with non-modifiable variables. Multidisciplinary management should be mandatory, and awareness of AKI diagnosis and etiological explorations should be raised.

Surgical Results for Infective Endocarditis Complicated With Cardiogenic Shock

BACKGROUND

Infective endocarditis remains associated with substantial mortality and morbidity rates, and the presence of acute heart failure (AHF) compromises clinical results after valve surgery; however, little is known in cardiogenic shock (CGS) patients. This study evaluated the clinical results and risk of mortality in CGS patients after valve surgery.Methods and Results:This study enrolled 585 patients who underwent valve surgery for active endocarditis at 14 institutions between 2009 and 2017. Of these patients, 69 (12%) were in CGS, which was defined as systolic blood pressure <80 mmHg and severe pulmonary congestion, requiring mechanical ventilation and/or mechanical circulatory support, preoperatively. The predictors of CGS were analyzed, and clinical results of patients with non-CGS AHF (n=215) were evaluated and compared.Staphylococcus aureusinfection (odds ratio [OR] 2.19; P=0.044), double valve involvement (OR 3.37; P=0.003), and larger vegetation (OR 1.05; P=0.036) were risk factors for CGS. Hospital mortality occurred in 27 (13%) non-CGS AHF patients and in 15 (22%) CGS patients (P=0.079). Overall survival at 1 and 5 years in CGS patients was 76% and 69%, respectively, and there were no significant differences in overall survival compared with non-CGS AHF patients (P=1.000).

CONCLUSIONS

Clinical results after valve surgery in CGS patients remain challenging; however, mid-term results were equivalent to those of non-CGS AHF patients.

Factors associated with major adverse kidney events in patients who underwent veno-arterial extracorporeal membrane oxygenation

OBJECTIVE

To describe acute kidney injury (AKI) natural history and to identify predictors of major adverse kidney events (MAKE) within 1 year in patients supported by veno-arterial extracorporeal membrane oxygenation (VA-ECMO).

DESIGN

Retrospective observational study.

SETTING

Medical French intensive care unit between January 2014 and December 2016.

PATIENTS

Consecutive patients implanted with VA-ECMO ≥ 16 years, VA-ECMO for at least ≥ 48 h, and without end-stage chronic kidney disease (CKD).

INTERVENTION

None.

MEASUREMENTS

Multivariate logistic regression of factors associated with MAKE at 1 year defined as one of the following criteria within day 360: death and receipt of renal replacement therapy (RRT) or persistent renal dysfunction, i.e., CKD ≥ stage 3 corresponding to an estimated glomerular filtration rate (eGFR) ≤ 60 ml/min/1.73 m2 and MAKE at day 30 and day 90 defined as one of the following criteria within day 30 or day 90: death, receipt of renal replacement therapy and serum creatinine ≥ threefold increase.

MAIN RESULTS

158 consecutive patients were included (male sex: 75.9%; median and interquartile range: age: 59 [47-66], Simplified Acute Physiology Score II: 55 [39-66], Sepsis-related Organ Failure Assessment Score: 9 [7-12], time on VA-ECMO: 7.5 [4-12] days). Among them 145 (91.8%) developed an AKI during the intensive care unit (ICU) stay and 85 (53.8%) needed renal replacement therapy (RRT). 59.9% (91/152), 60.5% (89/147) and 85.1% (120/141) evaluable patients had a MAKE-30, MAKE-90 and MAKE-360, respectively. Factors significantly associated with MAKE-360 were eGFR at baseline (odds ratio (OR) 0.98, confidence interval 95% (CI) [0.97;1.00], p 0.02), Kidney Disease Improving Global Outcome (KDIGO) stage at cannulation (p = 0.03), e.g., stage 3 vs. reference stage 0 OR 10.20 [1.77-58.87], and number of red blood cell (RBC) packs received while under ECMO (OR 1.14, CI 95% [1.01;1.28], p = 0.03). At 1 year among the 51 survivors, almost half of the alive patients (n = 20/51) had a decline of estimated glomerular filtration (eGFR) > 30% mL/min/1.73 m2. Their median eGFR decline was - 26.3% [- 46.6;- 10.7].

CONCLUSION

Patients undergoing VA-ECMO had a high risk of AKI during the ICU stay. Factors associated with MAKE 360 were mainly eGFR at baseline, KDIGO stage at cannulation and, number of RBC packs received while under ECMO. Among survivors at 1 year, almost half of the alive patients (n = 20/51) had a decline eGFR > 30%.

Preoperative platelet count, preoperative hemoglobin concentration and deep hypothermic circulatory arrest duration are risk factors for acute kidney injury after pulmonary endarterectomy: a retrospective cohort study

Background

Acute kidney injury (AKI) is a major postoperative morbidity of patients undergoing cardiac surgery and has a negative effect on prognosis. The kidney outcomes after pulmonary endarterectomy (PEA) have not yet been reported; However, several perioperative characteristics of PEA may induce postoperative AKI. The objective of our study was to identify the incidence and risk factors for postoperative AKI and its association with short-term outcomes.

Methods

This was a single-center, retrospective, observational, cohort study. Assessments of AKI diagnosis was executed based on the Kidney Disease Improving Global Outcomes (KDIGO) criteria.

Results

A total of 123 consecutive patients who underwent PEA between 2014 and 2018 were included.

The incidence of postoperative AKI was 45% in the study population. Stage 3 AKI was associated with worse short-term outcomes and 90-day mortality (p < 0.001, p = 0.002, respectively). The independent predictors of postoperative AKI were the preoperative platelet count (OR 0.992; 95%CI 0.984–0.999; P = 0.022), preoperative hemoglobin concentration (OR 0.969; 95%CI 0.946–0.993; P = 0.01) and deep hypothermic circulatory arrest (DHCA) time (OR 1.197; 95%CI 1.052–1.362; P = 0.006) in the multivariate analysis.

Conclusion

The incidence of postoperative AKI was relatively high after PEA compared with other types of cardiothoracic surgeries. The preoperative platelet count, preoperative hemoglobin concentration and DHCA duration were modifiable predictors of AKI, and patients may benefit from some low-risk, low-cost perioperative measures.

Aminoglycosides for infective endocarditis: time to say goodbye?

BACKGROUND

Based on experimental studies showing synergism with β-lactams and glycopeptides, aminoglycosides have long been considered essential in the treatment of infective endocarditis (IE). However, their use is associated with a high risk of renal failure, especially in elderly patients.

AIMS

The aim of this narrative review was to summarize the evidence to support reducing or even avoiding the use of aminoglycosides for the treatment of IE. We also analysed data supporting the use of aminoglycosides in specific subgroup of IE patients.

SOURCES

PubMed database was searched up to July 2019 to identify relevant studies.

CONTENTS

Recent European Guidelines reduced the use of aminoglycosides in IE, no longer recommended in Staphylococcus aureus native-valve IE, and shortened to 2 weeks for IE related to Enterococcus faecalis and streptococci with penicillin MIC >0.125 μg/mL. In addition, an alternative regimen without aminoglycosides (ampicillin or amoxicillin plus ceftriaxone) is proposed for E. faecalis. Observational studies suggested that gentamicin would not be necessary in the case of staphylococcal prosthetic valve IE as long as rifampicin is maintained. Recent clinical studies showed that for streptococcal IE, gentamicin could be restricted to isolates with penicillin MIC >0.5 μg/mL. For the empirical and definitive treatment of E. faecalis IE, amoxicillin or ampicillin plus ceftriaxone may be considered, irrespective of high-level of aminoglycoside resistance.

IMPLICATIONS

In a scenario of progressive increase in the age and frailty of IE patients, the use of aminoglycosides can be reduced or avoided in ~90% cases. This should result in reduced incidence of renal failure, an important prognostic factor in IE.

Evidence-based Clinical Decision Support Systems for the prediction and detection of three disease states in critical care: A systematic literature review

Background: Clinical decision support (CDS) systems have emerged as tools providing intelligent decision making to address challenges of critical care. CDS systems can be based on existing guidelines or best practices; and can also utilize machine learning to provide a diagnosis, recommendation, or therapy course. Methods: This research aimed to identify evidence-based study designs and outcome measures to determine the clinical effectiveness of clinical decision support systems in the detection and prediction of hemodynamic instability, respiratory distress, and infection within critical care settings. PubMed, ClinicalTrials.gov and Cochrane Database of Systematic Reviews were systematically searched to identify primary research published in English between 2013 and 2018. Studies conducted in the USA, Canada, UK, Germany and France with more than 10 participants per arm were included. Results: In studies on hemodynamic instability, the prediction and management of septic shock were the most researched topics followed by the early prediction of heart failure. For respiratory distress, the most popular topics were pneumonia detection and prediction followed by pulmonary embolisms. Given the importance of imaging and clinical notes, this area combined Machine Learning with image analysis and natural language processing. In studies on infection, the most researched areas were the detection, prediction, and management of sepsis, surgical site infections, as well as acute kidney injury. Overall, a variety of Machine Learning algorithms were utilized frequently, particularly support vector machines, boosting techniques, random forest classifiers and neural networks. Sensitivity, specificity, and ROC AUC were the most frequently reported performance measures. Conclusion: This review showed an increasing use of Machine Learning for CDS in all three areas. Large datasets are required for training these algorithms; making it imperative to appropriately address, challenges such as class imbalance, correct labelling of data and missing data. Recommendations are formulated for the development and successful adoption of CDS systems.

Evidence-based Clinical Decision Support Systems for the prediction and detection of three disease states in critical care: A systematic literature review

Background: Clinical decision support (CDS) systems have emerged as tools providing intelligent decision making to address challenges of critical care. CDS systems can be based on existing guidelines or best practices; and can also utilize machine learning to provide a diagnosis, recommendation, or therapy course. Methods: This research aimed to identify evidence-based study designs and outcome measures to determine the clinical effectiveness of clinical decision support systems in the detection and prediction of hemodynamic instability, respiratory distress, and infection within critical care settings. PubMed, ClinicalTrials.gov and Cochrane Database of Systematic Reviews were systematically searched to identify primary research published in English between 2013 and 2018. Studies conducted in the USA, Canada, UK, Germany and France with more than 10 participants per arm were included. Results: In studies on hemodynamic instability, the prediction and management of septic shock were the most researched topics followed by the early prediction of heart failure. For respiratory distress, the most popular topics were pneumonia detection and prediction followed by pulmonary embolisms. Given the importance of imaging and clinical notes, this area combined Machine Learning with image analysis and natural language processing. In studies on infection, the most researched areas were the detection, prediction, and management of sepsis, surgical site infections, as well as acute kidney injury. Overall, a variety of Machine Learning algorithms were utilized frequently, particularly support vector machines, boosting techniques, random forest classifiers and neural networks. Sensitivity, specificity, and ROC AUC were the most frequently reported performance measures. Conclusion: This review showed an increasing use of Machine Learning for CDS in all three areas. Large datasets are required for training these algorithms; making it imperative to appropriately address, challenges such as class imbalance, correct labelling of data and missing data. Recommendations are formulated for the development and successful adoption of CDS systems.

Surgery-first treatment improves clinical results in infective endocarditis complicated with disseminated intravascular coagulation†

OBJECTIVES

Infective endocarditis (IE) is a critical infection with a high mortality rate, and it usually causes sepsis. Though disseminated intravascular coagulation (DIC) sometimes occurs in IE patients, no definitive treatment strategy for IE patients with DIC as a complication exists. Therefore, we evaluated the prevalence, surgical results and treatment strategy for IE complicated with DIC.

METHODS

Between 2009 and 2017, a total of 585 patients undergoing valve surgery for active IE were enrolled at 14 institutions, of whom 116 (20%) had DIC as a complication. For further evaluation, we divided DIC patients into medical treatment-first (n = 45, group M) and valve surgery-first (n = 51, group S) groups after excluding 20 patients with intracranial haemorrhage.

RESULTS

The overall survival rates at 1 and 5 years were 91% and 85% in the non-DIC group and 65% and 55% in the DIC group, respectively (P < 0.001). Recurrence-free survival rates at 1 and 5 years were 99% and 95% in the non-DIC group and 94% and 74% in the DIC group, respectively (P < 0.001). The overall survival rates at 1 and 5 years were 77% and 64% in group S and 51% and 46% in group M, respectively (P = 0.032). Multivariable analysis revealed that 'medical treatment first' was an exclusive independent risk factor [hazards ratio 2.26 (1.13-4.75), P = 0.024] for overall mortality.

CONCLUSIONS

Mortality and IE recurrence were statistically significantly higher in DIC patients. Valve surgery should not be delayed because most patients proceeding with medical treatment eventually require emergency surgery and their clinical outcomes are worse than those of patients undergoing early surgery.

[Clinical features and prognosis of infective endocarditis patients with acute kidney injury]

OBJECTIVE

To investigate the clinical features and treatment of infective endocarditis (IE) patients with acute kidney injury (AKI), and to compare the adverse complications and outcome with IE patients without AKI.

METHODS

Clinical data of 100 IE cases in Peking University First Hospital from January 2002 to June 2018 were retrospectively reviewed. The patients were divided into AKI group (n=21) and non-AKI group (n=79) based on the AKI network (AKIN) definition. The clinical data and prognosis were compared between the two groups.

RESULTS

The incidence of AKI was 21%. The average age was (43.7±15.7) years, and the ratio of male to female was 3 ∶1. There was no significant difference in age and gender between the two groups. Compared with non-AKI group, the AKI group had more rash and lower limbs edema (P=0.017 and P=0.001), higher urine blood and protein positive rate (both P<0.001). Lower hemoglobin and serum albumin level (both P<0.001), worse clinical cardiac function (NYHA III-IV, P=0.033) were found in AKI group compared with non-AKI group. There was no significant difference in microbiologic positive rate and pathogenic bacteria sorts between the two groups. Nine patients refused surgery, and the other 91 cases underwent cardiac surgery with cardiopulmonary bypass under general anesthesia, including 19 cases of AKI group and 72 cases of non-AKI group. The ventilation time and intensive care unit (ICU) stay time were longer in AKI group than in non-AKI group (P=0.028 and P=0.003). AKI group needed more red blood cell transfusion (P=0.010). Using the last serum creatinine before surgery as basic level, there was more new-onset AKI cases in AKI group than in non-AKI group. During the median follow-up time 42 months, there was no significant difference in perioperative and follow-up mortality between the two groups (P=0.463 and P=0.581).

CONCLUSION

More perioperation complications occurred in IE patients with AKI, but no significant difference in in-hospital and follow-up mortality between the AKI and non-AKI groups was observed.

Acute kidney injury in hospitalized patients who underwent percutaneous kidney biopsy for histological diagnosis of their renal disease

BACKGROUND

Performing a kidney biopsy is necessary to accurately diagnose diseases such as glomerulonephritis and tubulointerstitial nephritis, among other such conditions. These conditions predispose patients to chronic kidney disease, as well as acute kidney injury (AKI). Notably, most epidemiological studies describing AKI have not investigated this patient population.

METHODS

Included patients admitted to the nephrology ward of a tertiary hospital who underwent percutaneous kidney biopsy. AKI was diagnosed based on the Kidney Disease: Improving Global Outcomes criteria.

RESULTS

Of the 223 patients investigated, 140 (62.8%) showed AKI. Of these, 91 (65%), 19 (13.6%), and 30 (21.4%) presented with AKI classified as stages 1, 2, and 3, respectively. The primary indication for performing biopsy was nephrotic syndrome or nephrotic proteinuria (73 [52.1%] in the AKI vs. 51 [61.4%] in the non-AKI group, p = 0.048). Focal segmental glomerulosclerosis was the most prevalent primary disease (24 [17.1%] in the AKI vs. 15 [18.0%] in the non-AKI group, p = 0.150). Multivariate analysis of risk factors associated with AKI showed hemoglobin levels (odds ratio [OR] 0.805, 95% confidence interval [CI] 0.681-0.951, p = 0.011), serum high-density lipoprotein cholesterol levels (HDL-c, OR 0.970, 95% CI 0.949-0.992, p = 0.008), and baseline serum creatinine levels (OR 2.703, 95% CI 1.471-4.968, p = 0.001) were significantly associated with AKI.

CONCLUSIONS

We observed a high prevalence of AKI in hospitalized patients who underwent kidney biopsy to investigate their renal disease, particularly glomerulonephritis. Higher levels of hemoglobin and serum HDL-c were associated with a lower risk of AKI.

Acute kidney injury in infective endocarditis: A retrospective analysis

BACKGROUND

Acute kidney injury (AKI) is associated with high case fatality in infective endocarditis (IE), but epidemiological data on the frequency of AKI during IE is scarce. We aimed to describe the frequency and risk factors for AKI during the course of IE using Kidney Disease: Improving Global Outcomes consensual criteria.

METHODS

Using the French hospital discharge database (French acronym PMSI), we retrospectively reviewed the charts of 112 patients presenting with a first episode of probable or definite IE between January 2010 and May 2015.

RESULTS

Seventy-seven patients (68.8%) developed AKI. In univariate analysis, risk factors for AKI were cardiac surgery for IE (n=29, 37.7% vs. n=4, 1.4%, P<0.0005), cardiac failure (n=29, 36.7% vs. n=1, 2.9%, P<0.0005), diabetes mellitus (n=14, 18.2% vs. n=1, 0.9%, P=0.034), and prosthetic valve IEs (n=24, 31.2% vs. n=4, 11.4%). No differences were observed for gentamicin exposure (n=57, 64% vs. n=32, 86.5%, P=0.286). Prosthetic valve IE, cardiac failure, and vancomycin exposure were independently associated with AKI with respective odds ratio of 5.49 (95% CI 1.92-17.9), 4.37 (95% CI 4.37-465.7), and 1.084 (1.084-16.2). Mean length of hospital stay was significantly longer in patients presenting with AKI than in controls (respectively 52.4±22.1 days vs. 39.6±12.6, P<0.005).

CONCLUSION

AKI is very frequent during IE, particularly in patients with prosthetic valve IE, cardiac failure, and those receiving vancomycin.

Impact of an Acinetobacter baumannii outbreak on kidney events in a burn unit: A targeted machine learning analysis

BACKGROUND

Multidrug-resistant (MDR) bacteria outbreaks represent a major threat in intensive care units. Patients may then be exposed to drug-related direct toxicity during such outbreaks. The objective of this study was to explore the impact of an outbreak of imipenem-resistant Acinetobacter baumannii (IR-AB) on renal outcomes.

METHODS

We performed a before-and-after observational study in a French burn intensive care unit during an IR-AB outbreak: a 13-month period before (period A, October 2013-October 2014) and a 13-month period after outbreak control (period B, December 2014-December 2015). A total of 409 patients were included, 195 during period A and 214 during period B. The main endpoint was major adverse kidney events at day 90 (MAKE 90). Secondary endpoints were acute kidney injury (AKI) and persistent renal dysfunction.

RESULTS

Incidence of MAKE 90 was 15.9% during period A versus 11.2% during period B (P = .166) and AKI 28.2% versus 18.7% (P = .023). The use of colistin was associated with renal outcomes in univariate analysis. After adjustment of potential confounding factors using a targeted Machine Learning Analysis (ie, IR-AB-related infection, septic shock, severity scores, other nephrotoxics, chronic kidney disease, serum creatinine at admission, Staphylococcus aureus), colistin remained associated with the risk of MAKE and AKI (relative risk = 2.909, 95% confidence interval [CI] [1.364, 6.204], P = .006 for MAKE 90, and relative risk = 2.14, 95% CI [1.52, 3.02], P<.0001 for AKI).

CONCLUSIONS

The episode of IR-AB outbreak was associated with an increased risk of kidney events, which appears to be driven by the use of colistin.

Predicting acute kidney injury at hospital re-entry using high-dimensional electronic health record data

Acute Kidney Injury (AKI), a sudden decline in kidney function, is associated with increased mortality, morbidity, length of stay, and hospital cost. Since AKI is sometimes preventable, there is great interest in prediction. Most existing studies consider all patients and therefore restrict to features available in the first hours of hospitalization. Here, the focus is instead on rehospitalized patients, a cohort in which rich longitudinal features from prior hospitalizations can be analyzed. Our objective is to provide a risk score directly at hospital re-entry. Gradient boosting, penalized logistic regression (with and without stability selection), and a recurrent neural network are trained on two years of adult inpatient EHR data (3,387 attributes for 34,505 patients who generated 90,013 training samples with 5,618 cases and 84,395 controls). Predictions are internally evaluated with 50 iterations of 5-fold grouped cross-validation with special emphasis on calibration, an analysis of which is performed at the patient as well as hospitalization level. Error is assessed with respect to diagnosis, race, age, gender, AKI identification method, and hospital utilization. In an additional experiment, the regularization penalty is severely increased to induce parsimony and interpretability. Predictors identified for rehospitalized patients are also reported with a special analysis of medications that might be modifiable risk factors. Insights from this study might be used to construct a predictive tool for AKI in rehospitalized patients. An accurate estimate of AKI risk at hospital entry might serve as a prior for an admitting provider or another predictive algorithm.

Risk factors and short-term prognosis of preoperative renal insufficiency in infective endocarditis

Background

The incidence of postoperative complications and the in-hospital mortality rate of infective endocarditis (IE) complicated with renal insufficiency are relatively high. This study aimed to analyze the clinical features, etiological characteristics, diagnosis and treatment, and prognosis of IE with renal insufficiency and to explore the risk factors for renal damage.

Methods

IE patients undergoing valvular surgery between 2008 and 2017 in two cardiac centers were retrospectively analyzed. They were divided into renal insufficiency (RI) [endogenous creatinine clearance rate (Ccr) <60 mL/min/1.73 m²] and normal renal function (NRF) (Ccr ≥60 mL/min/1.73 m²) groups. The disease conditions at admission, etiology, treatment, and prognosis were compared between the two groups. Multivariate regression analysis was performed for the related factors.

Results

A total of 8,055 cases of valvular surgery was performed during the study period. We analyzed 401 IE patients [average age 43.9±15 years; RI, n=56 (14%); NRF, n=345 (86%)], after the exclusion of 2 patients with primary glomerulonephritis. RI patients showed higher perioperative mortality (14.3% vs. 4.5%, P=0.042) and streptococcal infection (71.4% vs. 43.8%, P=0.001) rates. The RI group was also older and had worse heart function, greater decreases in hemoglobin and platelet levels, a higher rate of prosthetic valve involvement, more cases of postoperative dialysis, and worse prognosis (all P<0.05). Binary logistic multivariate regression analysis showed that the incidence of streptococcal infection [odds ratio (OR) =4.271, 95% confidence interval (CI), 1.846-9.884; P=0.001], age ≥51 years (OR =5.138, 95% CI, 2.258-11.694; P<0.001), and New York Heart Association (NYHA) functional class III-IV (OR =10.768, 95% CI, 2.417-47.972; P=0.002) were independent risk factors for preoperative renal insufficiency.

Conclusions

IE patients with preoperative renal insufficiency had a high mortality rate and poor prognosis, with streptococcal infection predisposing to a higher risk of renal insufficiency. Moreover, older the age and worse heart function in IE resulted in a greater risk for renal insufficiency.

Plasma gelsolin level predicts acute kidney injury after cardiopulmonary bypass in infants and young children

BACKGROUND

Acute kidney injury (AKI) after cardiopulmonary bypass (CPB) is a common complication especially in pediatric population. Plasma gelsolin (pGSN) is an anti-inflammatory factor through binding with actin and pro-inflammatory cytokines in circulation. Decrease in pGSN has been reported in some pathologic conditions. The purpose of the study was to determine the alterations of pGSN level in infants and young children after CPB and the role of pGSN as a predictor for the morbidity and severity of post-CPB AKI.

METHODS

Sixty-seven infants and young children at age ≤ 3 years old undergoing CPB were prospectively enrolled. PGSN levels were measured during peri-operative period with enzyme-linked immuno-sorbent assay and normalized with plasma total protein concentration. Other clinical characteristics of the patients were also recorded.

RESULTS

In patients developing AKI, the normalized pGSN (pGSN_N) levels significantly decreased at 6 h post-operation and remained low for 24 h post-operation as compared to the patients with non-AKI. PGSN_N at 6 h post-operation combining with CPB time presents an excellent predictive value for AKI.

CONCLUSIONS

Decreased pGSN_N identifies post-CPB AKI in the patients ≤ 3 years old, and is associated with adverse clinical outcomes. The findings suggest that circulating GSN in post-CPB patients may have beneficial effects on diminishing inflammatory responses.

Calibration drift in regression and machine learning models for acute kidney injury

Objective

Predictive analytics create opportunities to incorporate personalized risk estimates into clinical decision support. Models must be well calibrated to support decision-making, yet calibration deteriorates over time. This study explored the influence of modeling methods on performance drift and connected observed drift with data shifts in the patient population.

Materials and Methods

Using 2003 admissions to Department of Veterans Affairs hospitals nationwide, we developed 7 parallel models for hospital-acquired acute kidney injury using common regression and machine learning methods, validating each over 9 subsequent years.

Results

Discrimination was maintained for all models. Calibration declined as all models increasingly overpredicted risk. However, the random forest and neural network models maintained calibration across ranges of probability, capturing more admissions than did the regression models. The magnitude of overprediction increased over time for the regression models while remaining stable and small for the machine learning models. Changes in the rate of acute kidney injury were strongly linked to increasing overprediction, while changes in predictor-outcome associations corresponded with diverging patterns of calibration drift across methods.

Conclusions

Efficient and effective updating protocols will be essential for maintaining accuracy of, user confidence in, and safety of personalized risk predictions to support decision-making. Model updating protocols should be tailored to account for variations in calibration drift across methods and respond to periods of rapid performance drift rather than be limited to regularly scheduled annual or biannual intervals.

Risk factors for acute kidney injury associated with the treatment of bacterial endocarditis at a tertiary academic medical center

BACKGROUND

Acute kidney injury (AKI) is a common complication of endocarditis.

OBJECTIVE

To determine risk factors for the development of AKI in patients treated for endocarditis.

METHODS

This single centre, retrospective univariate and multivariate analysis to determine risk factors for the development of AKI included patients diagnosed with endocarditis between January 2009 and October 2013.

RESULTS

Of 211 included patients, a total of 84 (39.8%) patients developed AKI. We identified multiple independent variables associated with the development of AKI, including: age ≥ 65 years, presence of hardware, chronic kidney disease, AKI on admission, infection with Staphylococcus spp, receipt of nafcillin or oxacillin or aminoglycoside and nafcillin or oxacillin or aminoglycoside and vancomycin, vancomycin trough level ≥ 20.0 mcg/ml, aminoglycoside total daily dose reduction, duration of vancomycin exceeding three days, receipt of loop diuretic or more than three concomitant nephrotoxins and duration of loop diuretic or non-steroidal anti-inflammatory drug therapy exceeding seven days.

CONCLUSIONS

In patients treated for endocarditis, multiple risk factors for AKI were identified. Prospective studies are needed to evaluate these variables for causation of AKI in patients treated for endocarditis.

Effect of perioperative sodium bicarbonate administration on renal function following cardiac surgery for infective endocarditis: a randomized, placebo-controlled trial

Background

Patients with infective endocarditis (IE) have an elevated risk of renal dysfunction because of extensive systemic inflammation and use of nephrotoxic antibiotics. In this randomized, placebo-controlled trial, we investigated whether perioperative sodium bicarbonate administration could attenuate postoperative renal dysfunction in patients with IE undergoing cardiac surgery.

Methods

Seventy patients randomly received sodium chloride (n = 35) or sodium bicarbonate (n = 35). Sodium bicarbonate was administered as a 0.5 mmol/kg loading dose for 1 h commencing with anesthetic induction, followed by a 0.15 mmol/kg/h infusion for 23 h. The primary outcome was peak serum creatinine (SCr) level during the first 48 h postoperatively. The incidence of acute kidney injury, SCr level, estimated glomerular filtration rate, and major morbidity endpoints were assessed postoperatively.

Results

The peak SCr during the first 48 h postoperatively (bicarbonate vs. control: 1.01 (0.74, 1.37) mg/dl vs. 0.88 (0.76, 1.27) mg/dl, P = 0.474) and the incidence of acute kidney injury (bicarbonate vs. control: 29% vs. 23%, P = 0.584) were similar in both groups. The postoperative increase in SCr above baseline was greater in the bicarbonate group than in the control group on postoperative day 2 (0.21 (0.07, 0.33) mg/dl vs. 0.06 (0.00, 0.23) mg/dl, P = 0.028) and postoperative day 5 (0.23 (0.08, 0.36) mg/dl vs. 0.06 (0.00, 0.23) mg/dl, P = 0.017).

Conclusions

Perioperative sodium bicarbonate administration had no favorable impact on postoperative renal function and outcomes in patients with IE undergoing cardiac surgery. Instead, it was associated with possibly harmful renal effects, illustrated by a greater increase in SCr postoperatively, compared to control.

Trial registration

ClinicalTrials.gov, NCT01920126. Registered on 31 July 2013.

Electronic supplementary material

The online version of this article (doi:10.1186/s13054-016-1591-z) contains supplementary material, which is available to authorized users.

Effect of acute kidney injury on mortality and hospital stay in patient with severe acute pancreatitis

AIM

Severe acute pancreatitis (SAP) is believed to be a major risk factor leading to acute kidney injury (AKI) among critically ill patients, but little is known about SAP-induced AKI. We study the incidence of AKI defined by the Acute Kidney Injury Network (AKIN) criteria and the risk factors associated with outcomes among SAP-induced AKI patients.

METHOD

We conducted a multicenter retrospective study of critically ill SAP-induced AKI patients during the period August 2009 to June 2013. Data on enrolled patients were retrieved from electronic records. Univariate and multiple regression analyses were performed.

RESULTS

Among a total of 414 SAP patients admitted to intensive care units(ICU), 287 (69.3%) developed AKI during their ICU stay, with 16.7%, 18.4%, and 34.3% classified as AKI stage I,II, and III, respectively. SAP-induced AKI patients experienced a significantly higher ICU mortality than those without AKI. The risk factors associated with ICU mortality among SAP-induced AKI patients included ACS (odds ratio (OR) 10.58), RRT (OR 3.31), sepsis (OR 2.46), CTSI (OR 3.01), APACHE II score (OR 1.82), AKI III (OR 1.38), ICU-length-of-stay (OR 1.04), and multi-organ failure.

CONCLUSIONS

The paper represents the first attempt to investigate the etiology and epidemiology of AKI following SAP under the AKIN criteria among critically ill patients. Several independent risk factors were found to be associated with ICU mortality for AKI patients. The findings may pinpoint crucial therapeutic measures for preventing AKI among a vulnerable population and for more effective management of SAP-induced AKI to improve the quality of intensive care.

Minor Postoperative Increases of Creatinine Are Associated with Higher Mortality and Longer Hospital Length of Stay in Surgical Patients

BACKGROUND

Surgical patients frequently experience postoperative increases in creatinine levels. The authors hypothesized that even small increases in postoperative creatinine levels are associated with adverse outcomes.

METHODS

The authors examined the association of postoperative changes from preoperative baseline creatinine with all-cause in-hospital mortality and hospital length of stay (HLOS) in a retrospective analysis of surgical patients at a single tertiary care center between January 2006 and June 2012.

RESULTS

The data of 39,369 surgical patients (noncardiac surgery n = 37,345; cardiac surgery n = 2,024) were analyzed. Acute kidney injury (AKI)-by definition of the Kidney Disease: Improving Global Outcome group-was associated with a five-fold higher mortality (odds ratio [OR], 4.8; 95% CI, 4.1 to 5.7; P < 0.001) and a longer HLOS of 5 days (P < 0.001) after adjusting for age, sex, comorbidities, congestive heart failure, preoperative hemoglobin, preoperative creatinine, exposure to radiocontrast agent, type of surgery, and surgical AKI risk factors. Importantly, even minor creatinine increases (Δcreatinine 25 to 49% above baseline but < 0.3 mg/dl) not meeting AKI criteria were associated with a two-fold increased risk of death (OR, 1.7; 95% CI, 1.3 to 2.4; P < 0.001) and 2 days longer HLOS (P < 0.001). This was more pronounced in noncardiac surgery patients. Patients with minor creatinine increases had a five-fold risk of death (OR, 5.4; 95% CI, 1.5 to 20.3; P < 0.05) and a 3-day longer HLOS (P < 0.01) when undergoing noncardiac surgery.

CONCLUSIONS

Even minor postoperative increases in creatinine levels are associated with adverse outcomes. These results emphasize the importance to find effective therapeutic approaches to prevent or treat even mild forms of postoperative kidney dysfunction to improve surgical outcomes.

Utilizing electronic health records to predict acute kidney injury risk and outcomes: workgroup statements from the 15(th) ADQI Consensus Conference

The data contained within the electronic health record (EHR) is "big" from the standpoint of volume, velocity, and variety. These circumstances and the pervasive trend towards EHR adoption have sparked interest in applying big data predictive analytic techniques to EHR data. Acute kidney injury (AKI) is a condition well suited to prediction and risk forecasting; not only does the consensus definition for AKI allow temporal anchoring of events, but no treatments exist once AKI develops, underscoring the importance of early identification and prevention. The Acute Dialysis Quality Initiative (ADQI) convened a group of key opinion leaders and stakeholders to consider how best to approach AKI research and care in the "Big Data" era. This manuscript addresses the core elements of AKI risk prediction and outlines potential pathways and processes. We describe AKI prediction targets, feature selection, model development, and data display.

My paper 10 years later: infective endocarditis in the intensive care unit

INTRODUCTION

Although the recent literature contains plenty of studies concerning all aspects of infective endocarditis (IE), very few focus on severe IE requiring admission to the ICU.

RESULTS

In 2004, we published a report on the clinical spectrum and prognostic factors in 228 consecutive critically ill patients with IE. Septic shock, neurological complications and immunocompromised state were independently associated with in-hospital mortality. Cardiac surgery during the acute phase of EI was associated with better survival. A lot of information has been accumulated during the past 10 years on management of IE. Although three sets of blood cultures allow the identification of about 90% of cases, culture-negative IE still remains a diagnostic challenge. Blood-polymerase chain reaction in valve tissue may yield a microbiologic diagnosis. New imaging techniques such as positron emission tomography computed tomography (PET-CT) have shown additive value in patients with an intracardiac device or valvular prosthesis. Systematic cerebral magnetic resonance imaging can lead to modification of therapeutic plans. The decision to operate and the timing of cardiac surgery should take into account the presence of congestive heart failure, neurological complications, renal failure and multiorgan dysfunction syndrome. In 2011 and 2013, we published the results of a multicentre prospective observational study of 198 ICU patients with left-sided IE and confirmed that cardiac surgery was associated with better outcome. The strongest independent predictor of post-operative mortality was the pre-operative multiorgan failure score. Neurological failure also represented a major determinant of mortality, regardless of the mechanism of neurological complication.

CONCLUSION

In the present paper, we propose algorithms to optimize the medico-surgical approach.