A simple machine learning model for the prediction of acute kidney injury following noncardiac surgery in geriatric patients: a prospective cohort study.

BACKGROUND: Surgery in geriatric patients often poses risk of major postoperative complications. Acute kidney injury (AKI) is a common complication following noncardiac surgery and is associated with increased mortality. Early identification of geriatric patients at high risk of AKI could facilitate preventive measures and improve patient prognosis. This study used machine learning methods to identify important features and predict AKI following noncardiac surgery in geriatric patients. METHODS: The data for this study were obtained from a prospective cohort. Patients aged ≥ 65 years who received noncardiac surgery from June 2019 to December 2021 were enrolled. Data were split into training set (from June 2019 to March 2021) and internal validation set (from April 2021 to December 2021) by time. The least absolute shrinkage and selection operator (LASSO) regularization algorithm and the random forest recursive feature elimination algorithm (RF-RFE) were used to screen important predictors. Models were trained through extreme gradient boosting (XGBoost), random forest, and LASSO. The SHapley Additive exPlanations (SHAP) package was used to interpret the machine learning model. RESULTS: The training set included 6753 geriatric patients. Of these, 250 (3.70%) patients developed AKI. The XGBoost model with RF-RFE selected features outperformed other models with an area under the precision-recall curve (AUPRC) of 0.505 (95% confidence interval [CI]: 0.369-0.626) and an area under the receiver operating characteristic curve (AUROC) of 0.806 (95%CI: 0.733-0.875). The model incorporated ten predictors, including operation site and hypertension. The internal validation set included 3808 geriatric patients, and 96 (2.52%) patients developed AKI. The model maintained good predictive performance with an AUPRC of 0.431 (95%CI: 0.331-0.524) and an AUROC of 0.845 (95%CI: 0.796-0.888) in the internal validation. CONCLUSIONS: This study developed a simple machine learning model and a web calculator for predicting AKI following noncardiac surgery in geriatric patients. This model may be a valuable tool for guiding preventive measures and improving patient prognosis. TRIAL REGISTRATION: The protocol of this study was approved by the Committee of Ethics from West China Hospital of Sichuan University (2019-473) with a waiver of informed consent and registered at www.chictr.org.cn (ChiCTR1900025160, 15/08/2019).

Preoperative inflammatory biomarkers reveal renal involvement in postsurgical mortality in hip fracture patients: an exploratory study.

Background: Hip fractures in frail patients result in excess mortality not accounted for by age or comorbidities. The mechanisms behind the high risk of mortality remain undetermined but are hypothesized to be related to the inflammatory status of frail patients. Methods: In a prospective observational exploratory cohort study of hospitalized frail hip fracture patients, 92 inflammatory markers were tested in pre-operative serum samples and markers were tested against 6-month survival post-hip fracture surgery and incidence of acute kidney injury (AKI). After correcting for multiple testing, adjustments for comorbidities and demographics were performed on the statistically significant markers. Results: Of the 92 markers tested, circulating levels of fibroblast growth factor 23 (FGF-23) and interleukin-15 receptor alpha (IL15RA), both involved in renal disease, were significantly correlated with 6-month mortality (27.5% overall) after correcting for multiple testing. The incidence of postoperative AKI (25.4%) was strongly associated with 6-month mortality, odds ratio = 10.57; 95% CI [2.76-40.51], and with both markers plus estimated glomerular filtration rate (eGFR)- cystatin C (CYSC) but not eGFR-CRE. The effect of these markers on mortality was significantly mediated by their effect on postoperative AKI. Conclusion: High postoperative mortality in frail hip fracture patients is highly correlated with preoperative biomarkers of renal function in this pilot study. The effect of preoperative circulating levels of FGF-23, IL15RA, and eGFR-CYSC on 6-month mortality is in part mediated by their effect on postoperative AKI. Creatinine-derived preoperative renal function measures were very poorly correlated with postoperative outcomes in this group.

Elevated Serum KIM-1 in Sepsis Correlates with Kidney Dysfunction and the Severity of Multi-Organ Critical Illness.

The kidney injury molecule (KIM)-1 is shed from proximal tubular cells in acute kidney injury (AKI), relaying tubular epithelial proliferation. Additionally, KIM-1 portends complex immunoregulation and is elevated after exposure to lipopolysaccharides. It thus may represent a biomarker in critical illness, sepsis, and sepsis-associated AKI (SA-AKI). To characterise and compare KIM-1 in these settings, we analysed KIM-1 serum concentrations in 192 critically ill patients admitted to the intensive care unit. Irrespective of kidney dysfunction, KIM-1 serum levels were significantly higher in patients with sepsis compared with other critical illnesses (191.6 vs. 132.2 pg/mL, p = 0.019) and were highest in patients with urogenital sepsis, followed by liver failure. Furthermore, KIM-1 levels were significantly elevated in critically ill patients who developed AKI within 48 h (273.3 vs. 125.8 pg/mL, p = 0.026) or later received renal replacement therapy (RRT) (299.7 vs. 146.3 pg/mL, p < 0.001). KIM-1 correlated with markers of renal function, inflammatory parameters, hematopoietic function, and cholangiocellular injury. Among subcomponents of the SOFA score, KIM-1 was elevated in patients with hyperbilirubinaemia (>2 mg/dL, p < 0.001) and thrombocytopenia (<150/nL, p = 0.018). In univariate and multivariate regression analyses, KIM-1 predicted sepsis, the need for RRT, and multi-organ dysfunction (MOD, SOFA > 12 and APACHE II ≥ 20) on the day of admission, adjusting for relevant comorbidities, bilirubin, and platelet count. Additionally, KIM-1 in multivariate regression was able to predict sepsis in patients without prior (CKD) or present (AKI) kidney injury. Our study suggests that next to its established role as a biomarker in kidney dysfunction, KIM-1 is associated with sepsis, biliary injury, and critical illness severity. It thus may offer aid for risk stratification in these patients.

Deciphering AKI in Burn Patients: Correlations between Clinical Clusters and Biomarkers.

Acute kidney injury (AKI) is a significant complication in burn patients, impacting outcomes substantially. This study explores the heterogeneity of AKI in burn patients by analyzing creatinine time-series data to identify distinct AKI clusters and evaluating routine biomarkers' predictive values. A retrospective cohort analysis was performed on 2608 adult burn patients admitted to Hangang Sacred Heart Hospital's Burn Intensive Care Unit (BICU) from July 2010 to December 2022. Patients were divided into four clusters based on creatinine trajectories, ranging from high-risk, severe cases to lower-risk, short-term care cases. Cluster A, characterized by high-risk, severe cases, showed the highest mortality and severity, with significant predictors being PT and TB. Cluster B, representing intermediate recovery cases, highlighted PT and albumin as useful predictors. Cluster C, a low-risk, high-resilience group, demonstrated predictive values for cystatin C and eGFR cys. Cluster D, comprising lower-risk, short-term care patients, indicated the importance of PT and lactate. Key biomarkers, including albumin, prothrombin time (PT), cystatin C, eGFR cys, and total bilirubin (TB), were identified as significant predictors of AKI development, varying across clusters. Diagnostic accuracy was assessed using area under the curve (AUC) metrics, reclassification metrics (NRI and IDI), and decision curve analysis. Cystatin C and eGFR cys consistently provided significant predictive value over creatinine, with AUC values significantly higher (p < 0.05) in each cluster. This study highlights the need for a tailored, biomarker-driven approach to AKI management in burn patients, advocating for the integration of diverse biomarkers in clinical practice to facilitate personalized treatment strategies. Future research should validate these biomarkers prospectively to confirm their clinical utility.

Development and external validation of a machine learning model for the prediction of persistent acute kidney injury stage 3 in multi-centric, multi-national intensive care cohorts.

BACKGROUND: The aim of this retrospective cohort study was to develop and validate on multiple international datasets a real-time machine learning model able to accurately predict persistent acute kidney injury (AKI) in the intensive care unit (ICU). METHODS: We selected adult patients admitted to ICU classified as AKI stage 2 or 3 as defined by the "Kidney Disease: Improving Global Outcomes" criteria. The primary endpoint was the ability to predict AKI stage 3 lasting for at least 72 h while in the ICU. An explainable tree regressor was trained and calibrated on two tertiary, urban, academic, single-center databases and externally validated on two multi-centers databases. RESULTS: A total of 7759 ICU patients were enrolled for analysis. The incidence of persistent stage 3 AKI varied from 11 to 6% in the development and internal validation cohorts, respectively and 19% in external validation cohorts. The model achieved area under the receiver operating characteristic curve of 0.94 (95% CI 0.92-0.95) in the US external validation cohort and 0.85 (95% CI 0.83-0.88) in the Italian external validation cohort. CONCLUSIONS: A machine learning approach fed with the proper data pipeline can accurately predict onset of Persistent AKI Stage 3 during ICU patient stay in retrospective, multi-centric and international datasets. This model has the potential to improve management of AKI episodes in ICU if implemented in clinical practice.

The elevated lactate dehydrogenase to albumin ratio is a risk factor for developing sepsis-associated acute kidney injury: a single-center retrospective study.

BACKGROUND: There is no evidence to determine the association between the lactate dehydrogenase to albumin ratio (LAR) and the development of sepsis-associated acute kidney injury (SAKI). We aimed to investigate the predictive impact of LAR for SAKI in patients with sepsis. METHODS: A total of 4,087 patients with sepsis from the Medical Information Mart for Intensive Care IV (MIMIC IV) database were included. Logistic regression analysis was used to identify the association between LAR and the risk of developing SAKI, and the relationship was visualized using restricted cubic spline (RCS). The clinical predictive value of LAR was evaluated by ROC curve analysis. Subgroup analysis was used to search for interactive factors. RESULTS: The LAR level was markedly increased in the SAKI group (p < 0.001). There was a positive linear association between LAR and the risk of developing SAKI (p for nonlinearity = 0.867). Logistic regression analysis showed an independent predictive value of LAR for developing SAKI. The LAR had moderate clinical value, with an AUC of 0.644. Chronic kidney disease (CKD) was identified as an independent interactive factor. The predictive value of LAR for the development of SAKI disappeared in those with a history of CKD but remained in those without CKD. CONCLUSIONS: Elevated LAR 12 h before and after the diagnosis of sepsis is an independent risk factor for the development of SAKI in patients with sepsis. Chronic comorbidities, especially the history of CKD, should be taken into account when using LAR to predict the development of AKI in patients with sepsis.

Proximal tubular dysfunction as a predictor of AKI in Hospitalized COVID-19 patients.

BACKGROUND: High concentration of Angiotensin converting enzyme receptors in the proximal tubules make kidneys an early target in COVID-19. Proximal tubular dysfunction (PTD) may act as an early predictor of acute kidney injury (AKI) and more severe disease. METHODS: This prospective observational study was conducted in the COVID unit, Bangabandhu Sheikh Mujib Medical University. 87 COVID-19 patients without known kidney disease were screened for 6 markers of PTD on admission-hyperuricosuria, normoglycemic glycosuria, proteinuria, renal phosphate leak, sodium leak and potassium leak. Positivity of 2 of the first 4 markers was considered as PTD. 35 patients with PTD and 35 without PTD were followed up throughout their hospital stay. RESULTS: 52.9% had PTD on admission. The most prevalent markers were renal sodium leak (67%), followed by proteinuria (66.7%), hyperuricosuria (42.5%), potassium leak (32.2%), phosphate leak (28.7%) and normoglycemic glycosuria (20.7%). Mean age was 55.7 years. 32.9% patients developed AKI. PTD group had higher odds of developing AKI (odds ratio 17.5 for stage 1, 24.8 for stage 2 and 25.5 for stage 3; p<0.0001). The mean duration of hospital stay was 9 days higher in the PTD group (p<0.001). PTD group also had higher odds of transferring to ICU (OR = 9.4, p = 0.002), need for mechanical ventilation (OR = 10.1, p = 0.002) and death (OR = 10.3, p = 0.001). 32.6% had complete PTD recovery during follow-up. CONCLUSION: Proximal tubular dysfunction is highly prevalent in COVID-19 patients very early in the disease and may act as a predictor of AKI, ICU transfer, need for mechanical ventilation and death.

Predictive approach for liberation from acute dialysis in ICU patients using interpretable machine learning.

Renal recovery following dialysis-requiring acute kidney injury (AKI-D) is a vital clinical outcome in critical care, yet it remains an understudied area. This retrospective cohort study, conducted in a medical center in Taiwan from 2015 to 2020, enrolled patients with AKI-D during intensive care unit stays. We aimed to develop and temporally test models for predicting dialysis liberation before hospital discharge using machine learning algorithms and explore early predictors. The dataset comprised 90 routinely collected variables within the first three days of dialysis initiation. Out of 1,381 patients who received acute dialysis, 27.3% experienced renal recovery. The cohort was divided into the training group (N = 1135) and temporal testing group (N = 251). The models demonstrated good performance, with an area under the receiver operating characteristic curve of 0.85 (95% CI, 0.81-0.88) and an area under the precision-recall curve of 0.69 (95% CI, 0.62-0.76) for the XGBoost model. Key predictors included urine volume, Charlson comorbidity index, vital sign derivatives (trend of respiratory rate and SpO2), and lactate levels. We successfully developed early prediction models for renal recovery by integrating early changes in vital signs and inputs/outputs, which have the potential to aid clinical decision-making in the ICU.

[Development and validation of a nomogram for predicting 3-month mortality risk in patients with sepsis-associated acute kidney injury].

OBJECTIVE: To develop and evaluate a nomogram prediction model for the 3-month mortality risk of patients with sepsis-associated acute kidney injury (S-AKI). METHODS: Based on the American Medical Information Mart for Intensive Care- IV (MIMIC- IV), clinical data of S-AKI patients from 2008 to 2021 were collected. Initially, 58 relevant predictive factors were included, with all-cause mortality within 3 months as the outcome event. The data were divided into training and testing sets at a 7 : 3 ratio. In the training set, univariate Logistic regression analysis was used for preliminary variable screening. Multicollinearity analysis, Lasso regression, and random forest algorithm were employed for variable selection, combined with the clinical application value of variables, to establish a multivariable Logistic regression model, visualized using a nomogram. In the testing set, the predictive value of the model was evaluated through internal validation. The receiver operator characteristic curve (ROC curve) was drawn, and the area under the curve (AUC) was calculated to evaluate the discrimination of nomogram model and Oxford acute severity of illness score (OASIS), sequential organ failure assessment (SOFA), and systemic inflammatory response syndrome score (SIRS). The calibration curve was used to evaluate the calibration, and decision curve analysis (DCA) was performed to assess the net benefit at different probability thresholds. RESULTS: Based on the survival status at 3 months after diagnosis, patients were divided into 7 768 (68.54%) survivors and 3 566 (31.46%) death. In the training set, after multiple screenings, 7 variables were finally included in the nomogram model: Logistic organ dysfunction system (LODS), Charlson comorbidity index, urine output, international normalized ratio (INR), respiratory support mode, blood urea nitrogen, and age. Internal validation in the testing set showed that the AUC of nomogram model was 0.81 [95% confidence interval (95%CI) was 0.80-0.82], higher than the OASIS score's 0.70 (95%CI was 0.69-0.71) and significantly higher than the SOFA score's 0.57 (95%CI was 0.56-0.58) and SIRS score's 0.56 (95%CI was 0.55-0.57), indicating good discrimination. The calibration curve demonstrated that the nomogram model's calibration was better than the OASIS, SOFA, and SIRS scores. The DCA curve suggested that the nomogram model's clinical net benefit was better than the OASIS, SOFA, and SIRS scores at different probability thresholds. CONCLUSIONS: A nomogram prediction model for the 3-month mortality risk of S-AKI patients, based on clinical big data from MIMIC- IV and including seven variables, demonstrates good discriminative ability and calibration, providing an effective new tool for assessing the prognosis of S-AKI patients.

[Predictive value of albumin-to-fibrinogen ratio for acute kidney injury in infants undergoing ventricular septal defect repair with cardiopulmonary bypass].

OBJECTIVE: To investigate the predictive value of albumin-to-fibrinogen ratio (AFR) for postoperative acute kidney injury (AKI) in infants with ventricular septal defect repair under cardiopulmonary bypass (CPB). METHODS: A retrospective analysis was conducted on infants diagnosed with ventricular septal defect in Anhui Children's Hospital from January 2019 to July 2023. The infants were divided into AKI group and non-AKI group according to whether AKI occurred in hospital after operation. Demographic data, preoperative data, intraoperative data, postoperative data and laboratory results during CPB were collected. Multivariate Logistic regression analysis was used to find the factors of AKI after ventricular septal defect repair with CPB. Receiver operator characteristic curve (ROC curve) was drawn to analyze the predictive value of AFR for postoperative AKI after ventricular septal defect repair with CPB. RESULTS: A total of 215 children were collected, including 28 in AKI group and 187 in non-AKI group. There were no significant differences in age, gender, body weight, height, history of pneumonia and history of chronic heart failure between the two groups, but the left ventricular ejection fraction (LVEF) in the AKI group was significantly lower than that in the non-AKI group (0.526±0.028 vs. 0.538±0.030, P = 0.048). The duration of CPB (minutes: 74.1±12.1 vs. 65.8±11.3, P < 0.001), aortic cross-clamping (minutes: 41.7±9.7 vs. 37.2±9.4, P = 0.021) and hypothermic circulation arrest (21.4% vs. 8.6%, P = 0.047) in AKI group were significantly higher than those in non-AKI group, but there were no significant differences in the proportion of ultrafiltration and urine volume between the two groups. The length of intensive care unit (ICU) stay in AKI group was significantly longer than that in non-AKI group (days: 5.3±2.0 vs. 4.0±1.7, P < 0.001), but there were no significant differences in duration of mechanical ventilation and the proportion of postoperative hypotension between the two groups. During CPB, the levels of blood glucose (mmol/L: 9.4±1.3 vs. 8.8±0.8, P < 0.001), blood lactic acid (mmol/L: 2.2±0.3 vs. 2.0±0.3, P = 0.015) and serum creatinine (µmol/L: 79.7±11.5 vs. 74.4±10.9, P = 0.018) in AKI group were significantly higher than those in non-AKI group, while the AFR was significantly lower than that in non-AKI group (8.5±1.3 vs. 10.2±1.6, P < 0.001), but there were no significant differences in the levels of hemoglobin, blood urea nitrogen, alanine aminotransferase and aspartate aminotransferase between the two groups during CPB. Multivariate Logistic regression showed that AFR was a protective factor for AKI after ventricular septal defect repair with CPB [odds ratio (OR) = 0.439, 95% confidence interval (95%CI) was 0.288-0.669, P < 0.001]. Blood glucose (OR = 2.133, 95%CI was 1.239-3.672, P = 0.006) and blood lactic acid (OR = 5.568, 95%CI was 1.102-28.149, P = 0.038) were risk factors for AKI after ventricular septal defect repair with CPB. ROC curve analysis showed that the area under the curve (AUC) of AFR in predicting AKI after ventricular septal defect repair with CPB was 0.804 (95%CI was 0.712-0.897, P < 0.001). When the optimal cut-off value was less than 9.05, the corresponding sensitivity was 75.0% and the specificity was 72.7%. CONCLUSIONS: Low AFR (≤9.05) during CPB is an independent risk factor for AKI after ventricular septal defect repair with CPB. AFR during CPB has a high predictive value for postoperative AKI after ventricular septal defect repair with CPB.

Monocytes to lymphocytes multiplying platelets ratio as an early indicator of acute kidney injury in cardiac surgery with cardiopulmonary bypass: a retrospective analysis.

OBJECTIVE: The monocyte-to-lymphocyte multiplying platelets ratio (MLPR) is a novel systemic inflammatory marker, deriving from the monocyte-to-lymphocyte ratio (MLR). However, the link between MLPR and acute kidney injury following cardiac surgery (CSA-AKI) with cardiopulmonary bypass (CPB) has not been investigated yet. We comprehensively explored the potential linear and nonlinear relationship between MLPR or MLR and CSA-AKI. METHODS: Data of patients who underwent cardiac surgery with CPB between December 2018 and April 2021 were retrospectively collected at Fuwai Hospital, Beijing, China. MLPR was defined as monocyte count (×109/L) × 1000/(lymphocyte count (×109/L) × platelets (×109/L)). MLR was defined as monocyte count (×109/L)/lymphocyte count (×109/L). Logistic regression and restricted cubic spline (RCS) were used for linear and nonlinear analysis. The primary outcome was postoperative AKI within 48 h of after cardiac surgery. RESULTS: Of the 2420 patients screened, 2387 eligible patients were enrolled in the final analysis; the mean age was 54.7 years, and 1501 [62.9%] were men. The incidence of AKI was 25.8%. Logistic regression showed that MLPR (odds ratio [OR] = 1.31, 95% confidence interval [CI]: 1.16-1.48, p < .001) and MLR (OR = 3.06, 95% CI: 1.29-7.29, p = .012) were independent risk factors for AKI. Moreover, in the RCS model with adjustment for age (median: 56), female sex, and history of diabetes, a significant statistical difference was detected between preoperative MLPR, MLR, and AKI (p for non-linearity <.001). The subgroup analyses revealed similar results. CONCLUSIONS: The study revealed a nonlinear relationship between MLPR and MLR with AKI. MLPR exhibited a J-shaped curve, and MLR showed a favorable S-shaped curve in relation to AKI. Particularly, MLPR emerges as a promising clinical composite index for early CSA-AKI prediction. These findings emphasize the significance of MLPR as a valuable tool in clinical practice for timely identification and management of CSA-AKI.

Clinical impact of Wnt5a expression on persistence of acute kidney injury in patients with urosepsis.

Abnormal Wnt5a expression is associated with dysregulated inflammation and organ dysfunction. However, the effect of Wnt5a activation on the duration of organ dysfunction remains unclear. This prospective study investigated the association between Wnt5a levels and persistent acute kidney injury (AKI) in patients with urosepsis. Serum creatinine and Wnt5a levels were measured on days 1 and 5 and at discharge in 87 patients diagnosed with urosepsis. Patients with urosepsis were classified into an improving acute kidney injury (AKI) group and a persistent or worsening AKI group according to the AKI stage on days 1 and 5. AKI recovery was defined as a discharge-to-baseline serum creatinine ratio of <1.5. Twenty-eight patients with urosepsis (32.2%) had persistent or worsening AKI, and their Wnt5a levels were higher on days 1 and 5 and at discharge than those with improving AKI. The association between Wnt5a levels and persistent or worsening AKI was maintained after adjusting for age, sex, baseline serum creatinine levels, and disease severity. Moreover, elevated Wnt5a levels were associated with an increased risk of major adverse kidney events. High Wnt5a levels at discharge were associated with unrecovered AKI and participants with AKI recovery had a steeper Wnt5a slope over time than those without recovery, irrespective of age, sex, baseline serum creatinine level, or disease severity. Assessment of Wnt5a expression was helpful in predicting AKI persistence and adverse outcomes in patients with urosepsis. Therefore, Wnt5a may serve as a valuable bio-marker for identifying the risk of persistence of AKI.

Doppler-Derived Renal Functional Reserve in the Prediction of Postoperative Acute Kidney Injury in Patients Undergoing Robotic Surgery.

BACKGROUND: Postoperative acute kidney injury (PO-AKI) is a frequent complication after surgery. Various tools have been proposed to identify patients at high risk for AKI, including preoperative serum creatinine or estimated glomerular filtration rate (eGFR), urinary cell cycle arrest, and tubular damage biomarkers; however, none of these can appropriately assess AKI risk before surgery. Renal functional reserve (RFR) screened by the Doppler-derived intraparenchymal renal resistive index variation (IRRIV) test has been proposed to identify patients at risk for AKI before a kidney insult. IRRIV test has been developed in healthy individuals and previously investigated in cardiac surgery patients. This study aims to evaluate the value of the IRRIV test in identifying PO-AKI among patients undergoing robotic abdominal surgery in the Trendelenburg position for pelvic oncological disease. METHODS: We performed a prospective, double-blinded, observational study. Preoperative baseline renal function and RFR were assessed in 53 patients with baseline eGFR >60 mL/min/1.73 m2, undergoing robotic surgery in the Trendelenburg position for pelvic oncological disease. The capability of Doppler-derived RFR in predicting PO-AKI was investigated with the area under the receiver operating characteristic curve (ROC-AUC). RESULTS: Approximately 15.1% of patients developed AKI within the first 3 postoperative days. Thirty-one (58.5%) patients had a physiologic delta-RRI (ie, ≥0.05), while 22 (41.5%) patients did not. The ROC-AUC for PO-AKI was 0.85 (95% confidence interval [CI], 0.74-0.97; P = .007) for serum creatinine, 0.84 (95% CI, 0.71-0.96; P = .006) for eGFR, and 0.84 (95% CI, 0.78-0.91; P = .017) for delta-RRI. When combined with eGFR, the ROC-AUC for delta-RRI was 0.95 (95% CI, 0.9-1). CONCLUSIONS: Our findings show that the preoperative assessment of Doppler-derived RFR combined with baseline renal function improves the capability of identifying patients at high risk for PO-AKI with eGFR >60 mL/min/1.73 m2 after robotic abdominal surgery in Trendelenburg position for pelvic oncological disease.

AKI-Pro score for predicting progression to severe acute kidney injury or death in patients with early acute kidney injury after cardiac surgery.

BACKGROUND: No reliable clinical tools exist to predict acute kidney injury (AKI) progression. We aim to explore a scoring system for predicting the composite outcome of progression to severe AKI or death within seven days among early AKI patients after cardiac surgery. METHODS: In this study, we used two independent cohorts, and patients who experienced mild/moderate AKI within 48 h after cardiac surgery were enrolled. Eventually, 3188 patients from the MIMIC-IV database were used as the derivation cohort, while 499 patients from the Zhongshan cohort were used as external validation. The primary outcome was defined by the composite outcome of progression to severe AKI or death within seven days after enrollment. The variables identified by LASSO regression analysis were entered into logistic regression models and were used to construct the risk score. RESULTS: The composite outcome accounted for 3.7% (n = 119) and 7.6% (n = 38) of the derivation and validation cohorts, respectively. Six predictors were assembled into a risk score (AKI-Pro score), including female, baseline eGFR, aortic surgery, modified furosemide responsiveness index (mFRI), SOFA, and AKI stage. And we stratified the risk score into four groups: low, moderate, high, and very high risk. The risk score displayed satisfied predictive discrimination and calibration in the derivation and validation cohort. The AKI-Pro score discriminated the composite outcome better than CRATE score, Cleveland score, AKICS score, Simplified renal index, and SRI risk score (all P < 0.05). CONCLUSIONS: The AKI-Pro score is a new clinical tool that could assist clinicians to identify early AKI patients at high risk for AKI progression or death.

Cardiac Surgery-Associated Acute Kidney Injury.

AKI is a common and serious complication of cardiac surgery that has a significant impact on patient morbidity and mortality. The Kidney Disease Improving Global Outcomes definition of AKI is widely used to classify and identify AKI associated with cardiac surgery (cardiac surgery-associated AKI [CSA-AKI]) on the basis of changes in serum creatinine and/or urine output. There are various preoperative, intraoperative, and postoperative risk factors for the development of CSA-AKI which should be recognized and addressed as early as possible to expedite its diagnosis, reduce its occurrence, and prevent or ameliorate its devastating complications. Crucial issues are the inaccuracy of serum creatinine as a surrogate parameter of kidney function in the perioperative setting of cardiothoracic surgery and the necessity to discover more representative markers of the pathophysiology of AKI. However, except for the tissue inhibitor of metalloproteinase-2 and insulin-like growth factor binding protein 7 ratio, other diagnostic biomarkers with an acceptable sensitivity and specificity are still lacking. This article provides a comprehensive review of various aspects of CSA-AKI, including pathogenesis, risk factors, diagnosis, biomarkers, classification, prevention, and treatment management.

Perioperative Acute Renal Injury: Revisiting Pathophysiology.

BACKGROUND: Acute renal dysfunction and subsequent acute renal failure after cardiac surgery are associated with high mortality and morbidity. Early therapeutic or preventive intervention is hampered by the lack of an early biomarker for acute renal injury. Recent studies showed that urinary neutrophil gelatinase-associated lipocalin (NGAL or lipocalin 2) is upregulated early (within 1 to 3 h) after murine renal injury and in pediatric acute renal dysfunction after cardiac surgery. The authors hypothesized that postoperative urinary NGAL concentrations are increased in adult patients developing acute renal dysfunction after cardiac surgery compared with patients without acute renal dysfunction. METHODS: After institutional review board approval, 81 cardiac surgical patients were prospectively studied. Urine samples were collected immediately before incision and at various time intervals after surgery for NGAL analysis by quantitative immunoblotting. Acute renal dysfunction was defined as peak postoperative serum creatinine increase by 50% or greater compared with preoperative serum creatinine. RESULTS: Sixteen of 81 patients (20%) developed postoperative acute renal dysfunction, and the mean urinary NGAL concentrations in patients who developed acute renal dysfunction were significantly higher early after surgery (after 1 h, mean ± SD, 4,195 ± 6,520 vs. 1,068 ± 2,129 ng/ml; P < 0.01) compared with patients who did not develop acute renal dysfunction. Mean urinary NGAL concentrations continued to increase and remained significantly higher at 3 and 18 h after cardiac surgery in patients with acute renal dysfunction. In contrast, urinary NGAL in patients without acute renal dysfunction decreased rapidly after cardiac surgery. CONCLUSIONS: Patients developing postoperative acute renal dysfunction had significantly higher urinary NGAL concentrations early after cardiac surgery. Urinary NGAL may therefore be a useful early biomarker of acute renal dysfunction after cardiac surgery. These findings may facilitate the early detection of acute renal injury and potentially prevent progression to acute renal failure.

Interpretable machine learning model for predicting acute kidney injury in critically ill patients.

BACKGROUND: This study aimed to create a method for promptly predicting acute kidney injury (AKI) in intensive care patients by applying interpretable, explainable artificial intelligence techniques. METHODS: Population data regarding intensive care patients were derived from the Medical Information Mart for Intensive Care IV database from 2008 to 2019. Machine learning (ML) techniques with six methods were created to construct the predicted models for AKI. The performance of each ML model was evaluated by comparing the areas under the curve (AUC). Local Interpretable Model-Agnostic Explanations (LIME) method and Shapley Additive exPlanation values were used to decipher the best model. RESULTS: According to inclusion and exclusion criteria, 53,150 severely sick individuals were included in the present study, of which 42,520 (80%) were assigned to the training group, and 10,630 (20%) were allocated to the validation group. Compared to the other five ML models, the eXtreme Gradient Boosting (XGBoost) model greatly predicted AKI following ICU admission, with an AUC of 0.816. The top four contributing variables of the XGBoost model were SOFA score, weight, mechanical ventilation, and the Simplified Acute Physiology Score II. An AKI and Non-AKI cases were predicted separately using the LIME algorithm. CONCLUSION: Overall, the constructed clinical feature-based ML models are excellent in predicting AKI in intensive care patients. It would be constructive for physicians to provide early support and timely intervention measures to intensive care patients at risk of AKI.