Continuous renal replacement therapy (CRRT) is a form of dialysis prescribed to severely ill patients who cannot tolerate regular hemodialysis. However, as the patients are typically very ill to begin with, there is always uncertainty whether they will survive during or after CRRT treatment. Because of outcome uncertainty, a large percentage of patients treated with CRRT do not survive, utilizing scarce resources and raising false hope in patients and their families. To address these issues, we present a machine learning-based algorithm to predict short-term survival in patients being initiated on CRRT. We use information extracted from electronic health records from patients who were placed on CRRT at multiple institutions to train a model that predicts CRRT survival outcome; on a held-out test set, the model achieves an area under the receiver operating curve of 0.848 (CI = 0.822-0.870). Feature importance, error, and subgroup analyses provide insight into bias and relevant features for model prediction. Overall, we demonstrate the potential for predictive machine learning models to assist clinicians in alleviating the uncertainty of CRRT patient survival outcomes, with opportunities for future improvement through further data collection and advanced modeling.

BACKGROUND: The aim of this experimental study was to elucidate whether different distances between central venous catheter tips can affect drug clearance during continuous renal replacement therapy (CRRT). Central venous catheters (CVCs) are widely used in intensive care patients for drug infusion. If a patient receives CRRT, a second central dialysis catheter (CDC) is required. Where to insert CVCs is directed by guidelines, but recommendations regarding how to place multiple catheters are scarce. There are indications that a drug infused in a CVC with the tip close to the tip of the CDC, could be directly aspirated into the dialysis machine, with a risk of increased clearance. However, studies on whether clearance is affected by different CVC and CDC tip positions, when the two catheters are in the same vessel, are few.
METHODS: In this model with 18 piglets, gentamicin (GM) and vancomycin (VM) were infused through a CVC during CRRT. The CVC tip was placed in different positions in relation to the CDC tip from caudal, i.e., proximal to the heart, to cranial, i.e., distal to the heart. Serum and dialysate concentrations were sampled after approximately 30 min of CRRT at four different positions: when the CVC tip was 2 cm caudally (+ 2), at the same level (0), and at 2 (- 2) and 4 (- 4) cm cranially of the tip of the CDC. Clearance was calculated. A mixed linear model was performed, and level of significance was set to p < 0.05.
RESULTS: Clearance of GM had median values at + 2 cm, 0 cm, - 2 cm and - 4 cm of 17.3 (5.2), 18.6 (7.4), 20.0 (16.2) and 26.2 (12.2) ml/min, respectively (p = 0.04). Clearance of VM had median values at + 2 cm, 0 cm, - 2 cm and - 4 cm of 16.2 (4.5), 14.7 (4.9), 19.0 (10.2) and 21.2 (11.4) ml/min, respectively (p = 0.02).
CONCLUSIONS: The distance between CVC and CDC tips can affect drug clearance during CRRT. A cranial versus a caudal tip position of the CVC in relation to the tip of the CDC led to the highest clearance.

Optimal strategy for volume control and the clinical implication of achieved volume control are unknown in patients with sepsis-associated acute kidney injury (AKI) receiving continuous renal replacement therapy (CRRT). This randomized controlled trial aimed to compare the survival according to conventional or bioelectrical impedance analysis (BIA)-guided volume control strategy in patients with sepsis-associated AKI receiving CRRT. We also compared patient survival according to achieved volume accumulation rate ([cumulative fluid balance during 3 days × 100]/fluid overload measured by BIA at enrollment) as a post-hoc analysis. We randomly assigned patients to conventional volume control strategy (n = 39) or to BIA-guided volume control strategy (n = 34). There were no differences in 28-day mortality (HR, 1.19; 95% CI, 0.63-2.23) or 90-day mortality (HR, 0.99; 95% CI 0.57-1.75) between conventional and BIA-guided volume control group. In the secondary analysis, achieved volume accumulation rate was significantly associated with patient survival. Compared with the achieved volume accumulation rate of ≤  - 50%, the HRs (95% CIs) for the risk of 90-day mortality were 1.21 (0.29-5.01), 0.55 (0.12-2.48), and 7.18 (1.58-32.51) in that of  - 50-0%, 1-50%, and > 50%, respectively. Hence, BIA-guided volume control in patients with sepsis-associated AKI receiving CRRT did not improve patient outcomes. In the secondary analysis, achieved volume accumulation rate was associated with patient survival.

There remains no optimal anticoagulation protocol for continuous renal replacement therapy (CRRT) with regional citrate anticoagulation (RCA) in pediatric patients with elevated D-dimer levels. We aimed to assess the effects of different anticoagulation strategies on the risk of CRRT filter clotting in these patients. Pediatric patients undergoing CRRT were retrospectively grouped based on pre-CRRT D-dimer levels and anticoagulant: D-RCA group (normal D-dimer, RCA only, n = 22), D+ RCA group (elevated D-dimer, RCA only, n = 50), and D+ RCA+ systemic heparin anticoagulation (SHA) group (elevated D-dimer, RCA combined with SHA, n = 55). The risk of filter clotting and incidence of bleeding were compared among the groups. Among the groups, the D+ RCA+ SHA group had the longest filter lifespan; further, the incidence of bleeding was not increased by concurrent use of low-dose heparin for anticoagulation. Moreover, concurrent heparin anticoagulation was associated with a decreased risk of filter clotting. Contrastingly, high pre-CRRT hemoglobin and D-dimer levels and post-filter ionized calcium level > 0.4 mmol/L were associated with an increased risk of filter clotting. RCA combined with low-dose heparin anticoagulation could reduce the risk of filter clotting and prolong filter lifespan without increasing the risk of bleeding in patients with elevated D-dimer levels undergoing CRRT.

Current continuous kidney replacement therapy (CKRT) protocols ignore physiological renal compensation for hypercapnia. This study aimed to explore feasibility, safety, and clinical benefits of pCO2-adapted CKRT for hypercapnic acute respiratory distress syndrome (ARDS) patients with indication for CKRT.
We enrolled mechanically ventilated hypercapnic ARDS patients (pCO2 > 7.33 kPa) receiving regional citrate anticoagulation (RCA) based CKRT in a prospective, randomized-controlled pilot-study across five intensive care units at the Charité-Universitätsmedizin Berlin, Germany. Patients were randomly assigned 1:1 to the control group with bicarbonate targeted to 24 mmol/l or pCO2-adapted-CKRT with target bicarbonate corresponding to physiological renal compensation. Study duration was six days. Primary outcome was bicarbonate after 72 h. Secondary endpoints included safety and clinical endpoints. Endpoints were assessed in all patients receiving treatment.
From September 2021 to May 2023 40 patients (80% male) were enrolled. 19 patients were randomized to the control group, 21 patients were randomized to pCO2-adapted-CKRT. Five patients were excluded before receiving treatment: three in the control group (consent withdrawal, lack of inclusion criteria fulfillment (n = 2)) and two in the intervention group (lack of inclusion criteria fulfillment, sudden unexpected death) and were therefore not included in the analysis. Median plasma bicarbonate 72 h after randomization was significantly higher in the intervention group (30.70 mmol/l (IQR 29.48; 31.93)) than in the control group (26.40 mmol/l (IQR 25.63; 26.88); p < 0.0001). More patients in the intervention group received lung protective ventilation defined as tidal volume < 8 ml/kg predicted body weight. Thirty-day mortality was 10/16 (63%) in the control group vs. 8/19 (42%) in the intervention group (p = 0.26).
Tailoring CKRT to physiological renal compensation of respiratory acidosis appears feasible and safe with the potential to improve patient care in hypercapnic ARDS.
The trial was registered in the German Clinical Trials Register (DRKS00026177) on September 9, 2021 and is now closed.

BACKGROUND: Acute kidney injury (AKI) is a common and serious condition, particularly among elderly patients. It is associated with high morbidity and mortality rates, further compounded by the need for continuous renal replacement therapy in severe cases. To improve clinical decision-making and patient management, there is a need for accurate prediction models that can identify patients at a high risk of mortality.
METHODS: Data were extracted from the Dryad Digital Repository. Multivariate analysis was performed using least absolute shrinkage and selection operator (LASSO) logistic regression analysis to identify independent risk factors and construct a predictive nomogram for mortality within 28 days after continuous renal replacement therapy in elderly patients with acute kidney injury. The discrimination of the model was evaluated in the validation cohort using the area under the receiver operating characteristic curve (AUC), and calibration was evaluated using a calibration curve. The clinical utility of the model was assessed using decision curve analysis (DCA).
RESULTS: A total of 606 participants were enrolled and randomly divided into two groups: a training cohort (n = 424) and a validation cohort (n = 182) in a 7:3 proportion. A risk prediction model was developed to identify independent predictors of 28-day mortality in elderly patients with AKI. The predictors included age, systolic blood pressure, creatinine, albumin, phosphorus, age-adjusted Charlson Comorbidity Index (CCI), Acute Physiology and Chronic Health Evaluation II (APACHE II) score, and sequential organ failure assessment (SOFA) score. These predictors were incorporated into a logistic model and presented in a user-friendly nomogram. In the validation cohort, the model demonstrated good predictive performance with an AUC of 0.799. The calibration curve showed that the model was well calibrated. Additionally, DCA revealed significant net benefits of the nomogram for clinical application.
CONCLUSIONS: The development of a nomogram for predicting 28-day mortality in elderly patients with AKI receiving continuous renal replacement therapy has the potential to improve prognostic accuracy and assist in clinical decision-making.

Background and objective Continuous renal replacement therapy (CRRT) is a blood purification therapy modality for the treatment of renal failure in critically ill hospitalized patients with multiorgan dysfunction, effectively preventing uremia and multiple organ failure while improving renal function. However, the perfusion of patient blood through extracorporeal circulation often results in unexpected early occlusion of the CRRT circuit or hemofilter, leading to frequent interruptions in CRRT and wastage of medical resources. Moreover, clinical research on such circuit occlusions is limited. In Japan, CRRT circuits require long-term perfusion, often lasting 24 hours or more, indicating the need for a model capable of inducing occlusion at any arbitrary time; this model can evaluate various aspects, including causes and underlying mechanisms, and contribute to the development of an occlusion prediction method. Hence, we hypothesized the need for a model for inducing occlusion at arbitrary time points. Consequently, we strove to develop an ex vivo circuit occlusion model involving the injection of calcium into circulating citrated animal blood to evaluate the relationship between the amount of calcium chloride injected, circuit occlusion time, and changes in circuit pressure over time. Methods We developed a circuit occlusion model using a commercially available CRRT circuit, polysulfone membrane hemofilter, heating extension tube, and thermostatic water bath, along with commercially available citrated bovine whole blood. The circuit was filled with blood over a 10-min duration using a roller pump and was occluded after a specific period by varying the flow rate of calcium injected into bovine whole blood. Additionally, continuous injection of 1 mEq/mL calcium chloride into the circuit was maintained while bovine whole blood circulated. Measurements were performed at each calcium injection flow rate (2, 3, and 4 mL/h), with each measurement performed five times. The group that did not receive calcium injection was used as the control (0 mL/h: Con), and the experiment was performed three times. Groups were defined as \"0, 2, 3, and 4\" for each calcium injection flow rate. The relationship among the amount of calcium chloride injected, circuit occlusion time, and changes in circuit pressure over time was evaluated. Furthermore, blood tests and blood viscoelastic tests were performed at arbitrary times. Results The circuit occlusion time varied with each calcium injection flow rate, and a significant difference was observed between each group (p<0.05). Circuit pressure gradually changed at four min before occlusion when calcium was injected at 2, 3, and 4 mL/h, with a more rapid change at one min before occlusion. We measured circuit pressure at four and one min before occlusion (-4 min, and -1 min, respectively), and at the time of circuit occlusion (0 min) in the Con and 4 mL/h groups. Significant differences were observed in AP between -4 min and 0 min and -1 min and 0 min at a calcium flow rate of 4 mL/h. Additionally, significant differences were seen in prefilter and return pressures between -4 min and 0 min, -4 min and -1 min, and -1 min and 0 min at a calcium flow rate of 4 mL/h (p<0.05). Conclusions Our proposed model accurately estimated the occlusion time based on changes in circuit pressure. This model can be used to create various experimental systems depending on the desired occlusion time.

UNASSIGNED: A patent dual-lumen dialysis catheter is one of the basic requirements for efficient extracorporeal (EC) therapy.
UNASSIGNED: The objective of this study was to measure the resistance to blood flow offered by straight and curved-extension dual-lumen dialysis catheters used for continuous renal replacement therapy (CRRT).
UNASSIGNED: Twenty dogs suffering from acute kidney injury (AKI) were subjected to CRRT. The dogs were allocated randomly to Group-I (curved extension catheter, n=12) or Group II (straight extension catheter, n=8), based on the type of dual-lumen catheter used in CRRT. The catheter outflow and inflow pressures were recorded at blood pump speeds of 50 ml/min and 99-100 ml/min. Data were tested for normality, and differences in mean inflow and outflow catheter resistances were evaluated for statistical significance using independent samples t-tests.
UNASSIGNED: Straight extension catheters offered lower inflow resistance than curved extension catheters at both 50 ml/min (41.50 ± 5.84 mm Hg vs. 63.75 ± 6.88 mm Hg, P=0.03) and 99-100 ml/min (63.00 ± 8.11 mm Hg vs. 86.92 ± 7.02 mm Hg, P=0.04) blood flow rates. Straight extension catheters also offered lower outflow resistance than curved catheters at 99-100 ml/min blood flow rate (-94.12 ± 7.91 mm Hg vs. -128.25 ± 7.56 mm Hg, P=0.01; the negative signs only indicate the direction of blood flow).
UNASSIGNED: These findings suggest that straight-extension dual-lumen dialysis catheters perform better than the curved model in extracorporeal renal replacement therapy by considering their lower resistance to blood flow.

This study aimed to characterize the impact of extracorporeal membrane oxygenation (ECMO) on the pharmacokinetics (PK) of meropenem in neonates and children and to provide recommendations for meropenem dosing in this specific population of patients. Therapeutic drug monitoring (152 meropenem plasma concentrations) data from 45 patients (38 received ECMO) with a body weight (BW) of 7.88 (3.62-11.97) kg (median (interquartile range)) and postnatal age of 3 (0-465) days were collected. The population PK analysis was performed using NONMEM V7.3.0. Monte Carlo simulations were performed to assess the probability of target achievement (PTA) for 40% of time the free drug remained above the minimum inhibitory concentration (fT > MIC) and 100% fT > MIC. BW was found to be a significant covariate for the volume of distribution (Vd) and clearance (CL). Additionally, continuous renal replacement therapy (CRRT) was associated with a two-fold increase in Vd. In the final model, the CL and Vd for a typical patient with a median BW of 7.88 kg that was off CRRT were 1.09 L/h (RSE = 8%) and 3.98 L (14%), respectively. ECMO did not affect meropenem PK, while superimposed CRRT significantly increased Vd. We concluded that current dosing regimens provide acceptably high PTA for MIC ≤ 4 mg/L for 40% fT > MIC, but individual dose adjustments are needed for 100% fT > MIC.

Acute kidney injury (AKI) requiring continuous renal replacement therapy (CRRT), secondary to cardiovascular disease and sepsis, is associated with high in-hospital mortality. Although studies have examined cardiovascular disease and sepsis in AKI, the association between AKI and hepatic functional impairment remains unclear. We hypothesized that hepatic function markers would predict mortality in patients undergoing CRRT. We included 1,899 CRRT patients from a multi-centre database. In Phase 1, participants were classified according to the total bilirubin (T-Bil) levels on the day of, and 3 days after, CRRT initiation: T-Bil < 1.2, 1.2 ≤ T-Bil < 2, and T-Bil ≥ 2 mg/dL. In Phase 2, propensity score matching (PSM) was performed to examine the effect of a T-Bil cutoff of 1.2 mg/dL (supported by the Sequential Organ Failure Assessment score); creating two groups based on a T-Bil cutoff of 1.2 mg/dL 3 days after CRRT initiation. The primary endpoint was total mortality 90 days after CRRT initiation, which was 34.7% (n = 571). In Phase 1, the T-Bil, aspartate transaminase (AST), alanine transaminase (ALT), and AST/ALT (De Ritis ratio) levels at CRRT initiation were not associated with the prognosis, while T-Bil, AST, and the De Ritis ratio 3 days after CRRT initiation were independent factors. In Phase 2, T-Bil ≥1.2 mg/dL on day 3 was a significant independent prognostic factor, even after PSM [hazard ratio: 2.41 (95% CI; 1.84-3.17), p < 0.001]. T-Bil ≥1.2 mg/dL 3 days after CRRT initiation predicted 90-day mortality. Changes in hepatic function markers in acute renal failure may enable stratification of high-risk patients.