Abstract
Acute kidney injury (AKI) is a serious condition associated with high incidence, mortality and poor prognosis. The optimal time for discontinuing continuous renal replacement therapy (CRRT) in AKI remains controversial. This study aimed to explore the predictive value of the furosemide stress test (FST) for determining the optimal timing to discontinue CRRT in critically ill patients with AKI. This prospective, multicenter, randomized controlled study was conducted across eight centers. Patients with AKI who were admitted into the ICU were included and grouped into the FST or non-FST control groups. FST was performed on patients enrolled in the FST group by randomization. The probability of renal replacement therapy (RRT) re-initiation within 7 and 28 days post-FST was assessed to evaluate the capability of FST in determining CRRT discontinuation time in AKI. A total of 240 patients were included after screening. Ninety-five patients (81.2%) in the FST group were successfully weaned, compared to 79 patients (64.2%) in the non-FST group (p = 0.003). Within 7 days after weaning, 22 patients (18.8%) in the FST group restarted RRT, compared to 44 patients (35.8%) in the non-FST group (p = 0.003). Within 28 days after weaning, 35 patients (29.9%) in the FST group restarted RRT, compared to 45 (36.6%) in the non-FST group (p = 0.273). FST is validated as a reliable and safe method for predicting successful CRRT weaning in critically ill AKI patients. It significantly improves the CRRT weaning success rate.
Trial registration
ChiCTR2400091040
Keywords: Acute kidney injury, furosemide stress test, weaning, continuous renal replacement therapy
1. Introduction
Acute kidney injury (AKI) is a major global public health issue [1], with approximately 13.5% of affected patients requiring continuous renal replacement therapy (CRRT) and a mortality rate of up to 26.9% [2]. Patients with AKI are susceptible to both immediate and long-term health consequences [3]. Although CRRT is widely utilized as an extracorporeal renal replacement therapy (RRT) for patients with AKI, significant challenges and uncertainties persist, particularly the lack of clear criteria for treatment initiation timing, treatment goals, and treatment duration [4]. The Kidney Disease: Improving Global Outcomes (KDIGO) guidelines [5] recommend considering the weaning of CRRT once renal function has been restored to meet the body’s metabolic needs. However, existing clinical guidelines do not provide explicit recommendations on the timing for discontinuing CRRT in patients with AKI. Early discontinuation can lead to increased risks, including the need for early reinitiation of RRT, fluid overload, delayed weaning from mechanical ventilation, and complications related to electrolyte and acid-base imbalances [6]. Although the furosemide stress test (FST) has emerged as a valuable biomarker for predicting AKI progression and the initiation of RRT [7], its specificity and sensitivity in predicting weaning from CRRT have not been validated. Therefore, this study aimed to further validate FST as a predictor of successful CRRT weaning.
2. Methods
2.1. Patients
This prospective multicenter randomized controlled study recruited patients who received CRRT for AKI in the intensive care units (ICUs) from June 2022 to April 2024 at eight central hospitals in China: Zhejiang Provincial People’s Hospital, Ningbo First Hospital, The First Affiliated Hospital of Wenzhou Medical University, Huzhou Central Hospital, Quzhou People’s Hospital, Jinhua Central Hospital, Lishui People’s Hospital, and Yiwu Central Hospital. Patients’ clinical information was gathered to meet the established inclusion and exclusion criteria.
This clinical trial complied with the Declaration of Helsinki, was approved by the Ethics Committee of Zhejiang Provincial People’s Hospital (Ethics No. 2019KY339), and was registered on the website of the China Clinical Trial Center (www.chictr.org.cn) under the registration number ChiCTR2400091040. The investigators obtained written informed consent from patients or their authorized representatives.
2.2. Inclusion and exclusion criteria
Eligible participants were ICU-admitted adults (≥18 years) diagnosed with AKI. The exclusion criteria included: (1) Patients on the verge of death or with terminal tumors; (2) Patients with severe hepatic insufficiency (Child-Turcotte-Pugh class C); (3) Patients with severe cerebrovascular lesions with a Glasgow Coma Scale(GCS) score of 3, and patients with cranio-cerebral trauma; (4) Patients with end-stage kidney disease (ESKD), and patients on long-term maintenance dialysis; and (5) Patients who had been treated with RRT before admission to the ICU.
2.3. Study protocol
Participants were randomly assigned to either the experimental or control group using the Cheng Qian Yi Shu data platform (cqmdata.com). This study employed a single-blind design. Participants and their authorized designees were blinded to group assignment, whereas the treating healthcare personnel were not. In the FST group, senior physicians performed the FST prior to CRRT discontinuation following comprehensive clinical evaluations, including assessments of hemodynamic stability, electrolyte and acid-base balance, volume status improvement, and renal function parameters (creatinine clearance and urine output). These evaluations guided CRRT withdrawal decision-making for AKI patients. The control group followed standard CRRT discontinuation protocols without FST implementation.
The primary outcome was to compare the success rate of CRRT withdrawal between the FST group and the non-FST group. Additionally, the predictive ability of 2-h urine output following the FST for determining the optimal timing of CRRT discontinuation was assessed using the area under the receiver operating characteristic (ROC) curve; The secondary outcomes included comparing the rates of RRT re-initiation within 7 and 28 days between the FST group and the non-FST group. Furthermore, the 28-day and 90-day survival rates were compared between the two groups, as well as between the successful and unsuccessful CRRT withdrawal subgroups.
2.4. Definition and staging of AKI
According to the 2012 KDIGO guidelines [5], AKI is defined as any of the following: (1) An increase in serum creatinine(Scr) by ≥ 0.3 mg/dL (26.5 µmol/L) within 48 h; (2) A rise in Scr level to 1.5 times or more of the baseline value within 7 days; (3) Sustained urine output < 0.5 mL/kg/h for at least 6 h.
The severity of AKI is classified into three stages: stage 1 (6 h of oliguria (< 0.5 mL/kg/h) or≥ 0.3 mg/dl increase in Scr or 1.5–1.9 times increase above baseline Scr); stage 2 (12 h of oliguria (< 0.5 mL/kg/h) or 2.0–3.0 times increase above baseline Scr); stage 3 (24 h of oliguria (< 0.3 mL/kg/h) or anuria for ≥ 12 h or ≥ 3.0 times increase above baseline Scr or initiation of RRT).
2.5. Furosemide stress test
Chawla et al. [8] established a standardized FST protocol: For patients with AKI who had received furosemide within the past 7 days, a 1.5 mg/kg intravenous bolus of furosemide was administered. For furosemide-naïve patients, a 1.0 mg/kg intravenous bolus was administered. Urine output after FST was observed to predict AKI progression.
The definition of the FST in this study was as follows: patients in the FST group who had been treated with furosemide within 7 days before CRRT were given furosemide 1.5 mg/kg intravenously before CRRT and patients in the control group who had not received furosemide in the 7 days before CRRT were administered 1 mg/kg furosemide intravenously. Patients were classified as part of the positive FST group if the urine output was greater than 200 mL within 2 h post-FST, and as part of the negative FST group if the urine output was less than 200 mL. The rates of RRT reinitiation within 7 and 28 days after discontinuation were observed, and renal function recovery was assessed through indicators such as urine output and creatinine clearance.
2.6. Definition of weaning success
According to the literature [9], successful discontinuation is defined as the absence of dependence on RRT for a period ranging from 3 days to 2 weeks after CRRT discontinuation in patients with AKI. In this study, successful CRRT discontinuation was prospectively defined as the absence of reinitiation of RRT within 7 days after CRRT discontinuation, because functional improvement in most cases of reversible AKI, if recovery occurs, typically manifests within 7–10 days.
In the FST group, a positive FST result was obtained at the time of CRRT discontinuation. Patients with a positive FST who did not require RRT initiation within 7 days were classified as weaning success. If the FST result was negative, CRRT was reinitiated, and the case was classified as a discontinuation failure. These patients awaited another discontinuation attempt, during which a repeat FST was performed to reassess eligibility for successful discontinuation. If a patient in the non-FST group did not undergo re-initiation of RRT within 7 days of stopping CRRT, their CRRT withdrawal was considered successful.
2.7. Statistical analysis
Data processing and analysis were conducted using SPSS 25.0 software. For continuous data, mean, interquartile range (IQR) and standard deviation (SD) were reported. Categorical variables were presented as frequencies and percentages. Chi-square or Fisher’s exact tests were used to compare Categorical variables between treatment groups. The independent sample t-test was utilized to evaluate normally distributed data, whereas the Wilcoxon rank-sum test was used for non-normally distributed data. Logistic regression analysis was used for univariate and multivariate analyses. Receiver operating characteristic (ROC) curves and area under the curve (AUC) were computed using GraphPad Prism 9 software. The optimal cutoff value for the 2-h urine output following the FST was determined by maximizing the Youden index (Youden index = Sensitivity + Specificity–1) [10]. The sensitivity and specificity of this optimal cutoff value for the 2-h urine output were evaluated to determine its predictive value for successful CRRT withdrawal following the FST. Additionally, Kaplan–Meier (K–M) survival curves were generated to assess differences in the 28- and 90-day survival rates between the successful weaning group and failed weaning group, with a significance level set at p < 0.05.
2.8. Baseline data
Baseline data included patient demographics, primary ICU admission diagnosis, comorbidities, laboratory test results, AKI stage by KDIGO criteria, urine output measurements, concomitant therapies, CRRT intervention details, characteristics at CRRT initiation and discontinuation, Sequential Organ Failure Assessment (SOFA) score, Acute Physiology and Chronic Health Evaluation (APACHE II) score, and lengths of ICU and hospital stay.
2.9. Sample size determination
The sample size was estimated using PASS 15.0 software. A retrospective study [9] found that the withdrawal success rate for patients with AKI receiving CRRT was 59%, with a 40% withdrawal success rate when furosemide was used as a diuretic. Based on a significance level (α) of 0.025 and 80% power (β = 0.2), the study determined that 105 participants per group would be sufficient to detect a difference between the two groups. Accounting for an anticipated 10% loss to follow-up, the final sample size was set at a minimum of 115 patients per group.
3. Results
3.1. Baseline information
This multicenter study enrolled critically ill patients with AKI from the ICUs of eight participating centers. Through 1:1 randomization, 400 patients were evaluated for eligibility. Of these, 200 patients were assigned to the FST group and 200 to the non-FST group. In the FST group, 77 patients died without being weaned off CRRT, 5 patients had ESKD requiring long-term hemodialysis, 1 declined to participate, and ultimately, 117 patients were included in the FST group. A total of 123 patients were included in the non-FST group, excluding 2 patients who declined participation and 75 patients who were not weaned from CRRT due to death. The study flow chart is presented in Figure 1.
Figure 1.
Flow diagram for the study.
Abbreviations: AKI, Acute Kidney Injury; ICU, Intensive care unit; FST, Furosemide stress test; CRRT, Continuous renal replacement therapy; ESKD, End-stage kidney disease
The two groups exhibited statistically significant differences in baseline hemoglobin and hematocrit levels upon ICU admission (p < 0.05). However, no statistically significant differences were found (p > 0.05) in the following baseline characteristics: patient demographics, primary ICU admission diagnosis, comorbid conditions, laboratory test results, AKI staging according to KDIGO criteria, urine output measurements, concomitant therapies, CRRT implementation details (including initiation and discontinuation parameters), SOFA scores, APACHE II scores, as well as lengths of ICU stay and hospitalization (Tables 1 and 2).
Table 1.
Baseline information.
| Total (n = 240) | FST (n = 117) | Non- FST (n = 123) | p | |
|---|---|---|---|---|
| Gender (male) | 164 (68.3) | 79 (67.5) | 85 (69.1) | 0.792 |
| Age (years) (IQR) | 69 (54, 80) | 71 (57.5, 79) | 65.9 (52, 81) | 0.260 |
| MAP (mmHg) (IQR) | 84.67 (70, 98) | 84.33 (68.84, 99.34) | 86 (71.67, 97.33) | 0.657 |
| ICU diagnosis (medicine) (%) | 78 (32.5) | 34 (29.06) | 44 (35.77) | 0.267 |
| Concomitant therapies | ||||
| Respiratory support, n (%) | 168 (70) | 86 (73.5) | 82 (66.67) | 0.248 |
| IABP therapy, n (%) | 8 (3.33) | 4 (3.42) | 4 (3.25) | 0.943 |
| ECMO therapy, n (%) | 8 (3.33) | 4 (3.42) | 4 (3.25) | 0.943 |
| Antibiotic therapy, n (%) | 182 (75.8) | 89 (76.1) | 93 (75.6) | 0.934 |
| Comorbidities | ||||
| factors to AKI | ||||
| CKD, n (%) | 30 (12.5) | 18 (15.38) | 12 (9.76) | 0.188 |
| CHF, n (%) | 28 (11.67) | 13 (11.11) | 15 (12.2) | 0.794 |
| Coronary disease, n (%) | 31 (12.92) | 13 (11.11) | 18 (14.63) | 0.416 |
| Hypertension, n (%) | 129 (53.75) | 62 (52.99) | 67 (54.47) | 0.818 |
| Diabetes, n (%) | 70 (29.17) | 31 (26.5) | 39 (31.71) | 0.375 |
| Hepatic insufficiency, n (%) | 20 (8.33) | 11 (9.4) | 9 (7.32) | 0.559 |
| ARDS, n (%) | 41 (17.08) | 20 (17.09) | 21 (17.07) | 0.997 |
| Severe trauma, n (%) | 10 (4.17) | 3 (2.56) | 7 (5.69) | 0.226 |
| Surgical procedures, n (%) | 66 (27.5) | 35 (29.91) | 31 (25.2) | 0.414 |
| Sepsis | 59(24.6) | 32(27.4) | 32(22) | 0.332 |
| Exposure to nephrotoxic agents | 17(7.08) | 8(6.83) | 9(7.31) | 0.885 |
| Severe infection | 47(19.58) | 28(23.93) | 19(15.44) | 0.098 |
| Laboratory tests | ||||
| WBC (× 109/L) (IQR) | 11.75 (8.3, 16.01) | 11.51 (7.06, 15.09) | 12.13 (9.1, 17.39) | 0.770 |
| HGB (g/L) (IQR) | 90 (74.25, 112.25) | 84 (74.5, 101.5) | 98 (73, 117) | 0.009 |
| HCT (%)(IQR) | 27.95 (22.1, 32.975) | 25.7 (21.8, 31.05) | 28.77 (22.1, 36) | 0.020 |
| TnI (µg/L) (IQR) | 0.25 (0.04, 9.43) | 0.18 (0.04, 9.315) | 0.36 (0.05, 14.33) | 0.498 |
| BNP (pg/L) (IQR) | 1131.3 (321.7, 1951.36) | 1049 (310, 1951.36) | 1407.3 (342, 1951.36) | 0.974 |
| Scr (μmol/L) (IQR) | 249.35 (168.2, 377.25) | 246.2 (174.1, 387.15) | 250.5 (151.5, 367) | 0.591 |
| BUN (mmol/L) (IQR) | 17.3 (12.22, 27.86) | 17.91 (12.98, 32.26) | 16.25 (10.92, 25.21) | 0.058 |
| K+(mmol/L) (SD) | 4.55 ± 0.89 | 4.47 ± 0.79 | 4.62 ± 0.98 | 0.19 |
| PH (SD) | 7.34 ± 0.12 | 7.35 ± 0.11 | 7.32 ± 0.12 | 0.128 |
| Lac (IQR) | 2.3 (1.3, 4.2) | 2.1 (1.3, 3.8) | 2.4 (1.32,4.4) | 0.700 |
| CRP (mg/L) (IQR) | 97.07 (33.775, 141.9) | 95.1 (38.35, 137.4) | 98.6 (28, 142.4) | 0.699 |
| PCT (mg/L) (IQR) | 6.51 (1.095, 13.34) | 6.2 (0.89, 13.34) | 8.21 (1.2, 13.34) | 0.874 |
| AKI | 0.194 | |||
| stage 1, n (%) | 63 (26.3) | 26 (22.2) | 37 (26.3) | – |
| stage 2, n (%) | 97 (40.4) | 46 (39.3) | 51 (41.5) | – |
| stage 3, n (%) | 80 (33.3) | 45 (38.5) | 35 (28.5) | – |
| Others | ||||
| Duration of CRRT (hours/day) | 8.75 (4.17, 14.4875) | 8 (4.04, 13.16) | 9.46 (4.5, 14.98) | 0.174 |
| Sofa score (IQR) | 8 (6, 10) | 8 (6, 11) | 8 (6, 10) | 0.265 |
| APACHE II score (SD) | 23.33 ± 7.92 | 23.76 ± 8.08 | 22.91 ± 7.77 | 0.407 |
| ICU stay (IQR) | 17 (9, 29) | 18.5 (10.75, 29) | 15 (8,29) | 0.129 |
| Hospital stay (IQR) | 27 (18.5, 42.5) | 28.5 (19.75, 43.25) | 26 (17, 42) | 0.245 |
For continuous data, the mean, interquartile range (IQR) and standard deviation (SD) were reported; for categorical variables, percentage analysis was used.
Abbreviations: MAP, Mean arterial pressure; ICU, Intensive care unit; IABP therapy, Intra -aortic balloon pump therapy; ECMO therapy, Extracorporeal membrane oxygenation therapy; CKD, Chronic Kidney Disease; CHF, Chronic heart failure; ARDS, Acute Respiratory Distress Syndrome; WBC, White blood cell; HGB, Hemoglobin; HCT, Hematocrit; PT, Prothrombin time; APTT, Activated partial thromboplastin time; BNP, Brain natriuretic peptide; Tbil,Total bilirubin; Scr, Serum creatinine; BUN, Blood urea nitrogen; pH, Potential of hydrogen; Lac, Lactate; CRP, C-reactive protein; PCT, Procalcitonin; eGFR, estimated glomerular filtration rate; Sofa score: Sequential organ failure assessment score; APACHEII score, Acute Physiology and Chronic Health Evaluation II score.
Table 2.
Characteristics at CRRT initiation and discontinuation.
| Total (n = 240) | FST (n = 117) | Non- FST (n = 123) | p | |
|---|---|---|---|---|
| Urine output at the time of AKI diagnosis | ||||
| 6-h (mL) (IQR) | 177.5 (30, 300) | 110 (10, 325) | 200 (50, 300) | 0.121 |
| 12-h (mL) (IQR) | 350 (100, 700) | 320 (80, 700) | 400 (150, 750) | 0.267 |
| 24-h(mL) (IQR) | 700 (250, 1680) | 650 (200, 1600) | 750 (250, 1700) | 0.448 |
| Indication for initiating CRRT | ||||
| K+> 6.5 mmol/L, n (%) | 10 (4.17) | 4 (3.42) | 6 (4.88) | 0.572 |
| PH < 7.15, n (%) | 23 (9.58) | 11 (9.4) | 12 (9.76) | 0.926 |
| Acute pulmonary edema | 28 (11.67) | 14 (11.97) | 14 (11.38) | 0.888 |
| BUN ≥ 112 mg/dL (6.22 mmol/L), n (%) | 189 (78.75) | 92 (78.63) | 97 (78.86) | 0.965 |
| Oliguria ≥ 72 hor anuria ≥ 24 h, n (%) | 133 (55.42) | 67 (57.26) | 66 (53.66) | 0.574 |
| The first day of CRRT | ||||
| 24-h total fluid intake (mL) (IQR) | 2904.35 (1941.25,4232.75) | 2580 (1828, 3921.5) | 3258.5 (2020, 4697) | 0.069 |
| 24-h total fluid output (mL) (IQR) | 2748.5 (1750.5,4349) | 2633 (1654, 4537.5) | 2855 (1780, 4280) | 0.350 |
| 24-h ultrafiltration(mL) (IQR) | 1234.5 (600.25, 1529.25) | 1092 (515, 1440) | 1429.2 (650,1652) | 0.189 |
| 24-h urine output (mL) (IQR) | 510 (150, 1300) | 600 (166, 1390) | 480 (120,1300) | 0.724 |
| Scr (µmol/L) (IQR) | 250.1 (168.2, 380.85) | 249 (176.65, 389.85) | 250.5 (147.9,367) | 0.501 |
| BUN (mmol/L) (IQR) | 17.19 (12.06, 27.215) | 17.82 (12.915, 31.075) | 16.25 (10.7, 25.38) | 0.079 |
| eGFR (ml/min/1.73 m2) (IQR) | 24.85 (14.87, 38.46) | 27.08 (17.23, 38.35) | 23.45 (13.34, 41.86) | 0.483 |
| BNP (pg/L) (IQR) | 3250.8 (603.58, 3451.89) | 3451.89 (599.75, 3451.89) | 2773.7 (586.9, 3451.89) | 0.297 |
| Lac (mmol/L) (IQR) | 2.5 (1.3375, 7.85) | 2.4 (1.40, 5.71) | 2.5 (1.3, 13.5) | 0.360 |
| The day of CRRT weaning | ||||
| 24-h total fluid intake (mL) (IQR) | 2840.5 (2198.75, 3614.5) | 2920 (2275.5, 3650) | 2792 (2060, 3586) | 0.671 |
| 24-h total fluid output (mL) (IQR) | 3106 (2187.75, 4340.75) | 3248 (2231.5, 4584.5) | 3089 (2121, 4231) | 0.477 |
| 24-h ultrafiltration (mL) (IQR) | 870 (305, 1580.5) | 887 (334.5, 1535) | 816.5 (233, 1642.5) | 0.834 |
| 24-h urine output (mL) (IQR) | 1200 (385, 1900) | 1230 (485, 1895) | 1150 (250, 1900) | 0.192 |
| Scr (µmol/L) (IQR) | 140.8 (81.725, 191.225) | 130.1 (75.15, 165.56) | 157 (85.2, 204) | 0.139 |
| Bun (mmol/L) (IQR) | 9.46 (4.9675, 25.105) | 8.49 (4.225, 21.855) | 10.2 (5.32, 27.67) | 0.187 |
| eGFR (mL/min/1.73 m2) (IQR) | 109.82 (48.6975, 149.99) | 102.02 (55, 149.99) | 113.53 (38.38, 149.99) | 0.587 |
| BNP (pg/L) (IQR) | 650.5 (350.83, 2285.18) | 816.1 (349.55, 2745.65) | 616.35 (329.83, 1989.8) | 0.052 |
| Lac (mmol/L) (IQR) | 1.65 (1.18, 2.63) | 1.5 (1.05, 2.68) | 1.75 (1.18, 2.7) | 0.826 |
For continuous data, the mean, interquartile range (IQR) and standard deviation (SD) were reported; for categorical variables, percentage analysis was used.
Abbreviations: Scr, Serum creatinine; BUN, Blood urea nitrogen; Lac, Lactate; BNP, Brain natriuretic peptide.
3.2. Success rate of CRRT weaning and restart rate of RRT
The FST group included 117 critically ill patients with AKI, of whom 95 (81.2%) were successfully weaned from CRRT. The non-FST group included 123 patients, with a significantly lower success rate of 64.2% compared to the FST group (79/123 vs. 95/117, p = 0.003) (Table 3, Figure 2).
Table 3.
The success rate of CRRT weaning.
| Total (n = 240) | FST (n = 117) | Non- FST (n = 123) | p | |
|---|---|---|---|---|
| Weaning success, n (%) | 174 (72.5) | 95 (81.2) | 79 (64.2) | 0.003 |
| Weaning failure, n (%) | 66 (27.5) | 22 (18.8) | 44 (35.8) | – |
Abbreviations: FST, Furosemide stress test; CRRT, Continuous renal replacement therapy.
Figure 2.
The success rate of CRRT weaning.
Abbreviations: FST, Furosemide stress test; CRRT, Continuous renal replacement therapy
A total of 66 patients (27.5%) required RRT within 7 days post-discontinuation, comprising 22 patients from the FST group and 44 patients from the non-FST group (18.8 vs. 35.8%, p < 0.05). Among successfully weaned patients in the FST group, none required RRT within 7 days, contrasting with 22 cases of re-initiation in the weaning failure group (0/95 vs. 22/22, p < 0.05). Additionally, within the successfully weaned group, 13 patients restarted RRT within 28 days, compared to 22 cases in the weaning failure group (13/95 vs. 22/22, p < 0.05). However, no statistically significant differences were observed in 28-day outcomes between the FST and non-FST groups (p > 0.05) (Tables 4 and 5, Figure 3).
Table 4.
The 7-day and 28-day restart rates of RRT.
| Total (n = 240) | FST (n = 117) | Non- FST (n = 123) | p | |
|---|---|---|---|---|
| 7-day, n (%) | 66 (27.5) | 22 (18.8) | 44 (35.8) | 0.003 |
| 28-day, n (%) | 80 (33.3) | 35 (29.9) | 45 (36.6) | 0.273 |
Abbreviations: FST, Furosemide stress test; RRT, Renal replacement therapy.
Table 5.
The 7-day and 28-day restart rates of RRT in the FST group.
| Total | Weaning success group | Weaning failure group | p | |
|---|---|---|---|---|
| 7-day, n (%) | 22 (38) | 0 (0) | 22 (100) | 0.000 |
| 28-day, n (%) | 35 (29.9) | 13(13.7) | 22 (100) | 0.000 |
Abbreviation: RRT, Renal replacement therapy.
Figure 3.
Restart rate of RRT.
Abbreviations: FST, Furosemide stress test; CRRT, Continuous renal replacement therapy
3.3. ROC curves, univariate, and multivariate analysis of CRRT weaning
To better guide clinical practice, we evaluated the predictive value of 2-h post-FST urine output, 24-h urine output on the day of CRRT weaning, and SCr for predicting weaning success.
ROC curve analysis demonstrated significant predictive performance for CRRT weaning success. The 2-h post-FST urine output demonstrated greater predictive value with an AUC of 0.76 (95% CI: 0.655–0.865, p < 0.001), showing optimal sensitivity (74.7%) and specificity (72.7%) at the cutoff value of 202.5 mL, indicating acceptable discrimination or moderate accuracy. The 24-h urine output on the day of weaning showed an AUC of 0.748 (95% CI: 0.642–0.854), with optimal sensitivity (78.9%) and specificity (40.9%) at 525 mL (p < 0.05), while Scr showed an AUC of 0.641 (95% CI: 0.517–0.764), achieving maximal sensitivity (86.4%) and specificity (45.3%) at 104.4 μmol/L (p < 0.05) (Table 6, Figure 4). These results indicate that both 2-h post-FST and 24-h urine output parameters (AUC > 0.74) were significantly more predictive than Scr (AUC: 0.641) and served as independent indicators for successful CRRT discontinuation.
Table 6.
The ROC of 2-h urine output after FST.
| Variable | AUC | 95% CI | p | Sensitivity (%) | Specificity (%) | Cutoff |
|---|---|---|---|---|---|---|
| FST 2-h UO(mL) | 0.76 | (0.655–0.865) | 0.00 | 74.7 | 72.7 | 202.5 ml |
| FST 24-h UO(mL) | 0.748 | (0.642–0.854) | 0.00 | 78.9 | 40.9 | 525 ml |
| Scr (µmol/L) | 0.641 | (0.517–0.764) | 0.04 | 86.4 | 45.3 | 104.4mmol/l |
Abbreviations: UO, Urine output; Scr, Serum creatinine.
Figure 4.
The ROC of 2-h urine output post-FST.
Abbreviations: FST, furosemide stress test; CRRT, continuous renal replacement therapy; UO, urine output
According to univariate logistic analysis, both 2-h post-FST urine output (OR: 1.008, 95% CI: 1.002–1.014) and 24-h urine output (OR: 1.001, 95% CI: 1.000–1.002) were statistically significant predictors of successful CRRT withdrawal (p < 0.05). After controlling for potential confounders, multivariate logistic analysis was used to validate these findings. The results showed that both parameters had independent predictive value (2-h output: adjusted OR: 1.006, 95% CI: 1.000–1.011; 24-h output: adjusted OR: 1.001, 95% CI: 1.000–1.002, p < 0.05) (Table 7).
Table 7.
Multivariate analysis of CRRT weaning.
| Variable | Univariate analysis |
Multivariate analysis |
||||
|---|---|---|---|---|---|---|
| OR | 95% CI | p | OR | 95%CI | p | |
| FST 2-h UO (mL) | 1.008 | (1.002–1.014) | 0.005 | 1.006 | (1.000–1.011) | 0.040 |
| Scr(µmol/L) | 1.000 | (0.998–1.002) | 0.744 | |||
| Alb (g/L) | 1.059 | (0.982–1.141) | 0.137 | |||
| 24-h UO (mL) | 1.001 | (1.000–1.002) | 0.002 | 1.001 | (1.000–1.002) | 0.024 |
Abbreviations: UO, Urine output; Scr, Serum creatinine ;Alb, Albumin; CRRT, Continuous renal replacement therapy.
3.4. K–M survival analysis
In the survival analysis, 187 patients survived at 28 days and 163 patients at 90 days post-weaning. Although the non-FST group exhibited numerically higher survival counts than the FST group at both time points (28-day: 95 vs. 92; 90-day: 85 vs. 78), these differences were not statistically significant (p > 0.05) (Table 8).
Table 8.
Comparison of survival outcomes between FST and non-FST groups.
| Survival rate | Total (n = 240) | FST (n = 117) | Non- FST (n = 123) | p |
|---|---|---|---|---|
| 28-day | 187 (77.9) | 92 (78.6) | 95 (77.2) | 0.794 |
| 90-day | 163 (67.9) | 78 (66.7) | 85 (69.1) | 0.686 |
Abbreviation: FST, Furosemide stress test.
To further investigate long-term survival outcomes in patients with AKI receiving CRRT, we conducted a subgroup analysis by classifying all enrolled patients into successful weaning group and weaning failure group, followed by K–M survival analysis. The K–M survival analysis revealed significantly superior 28-day survival rates in the successful weaning group compared with the weaning failure group (HR: 0.5956, 95% CI: 0.344–1.031, p = 0.046), indicating a clinically meaningful survival advantage for patients who achieved CRRT discontinuation. However, no statistically significant difference in survival was observed between the groups at 90 days (HR: 0.8174, 95% CI: 0.5143–1.299, p = 0.37) (Figures 5 and 6).
Figure 5.
The 28-day survival curves of successful and unsuccessful withdrawal.
Abbreviations: HR, hazard ratio
Figure 6.
The 90-day survival curves of successful and unsuccessful withdrawal.
Abbreviations: HR, hazard ratio
4. Discussion
In this multicenter prospective randomized controlled trial, we utilized a 2-h urine output > 200 mL following FST as the criterion for discontinuing CRRT. Our results confirm that 2-h urine output post-FST has clinical significance in predicting the success of CRRT weaning in patients with AKI.
In recent years, numerous studies have employed the FST to predict the progression of AKI. Chawla et al. [8] reported that a low urine volume (< 200 mL) after FST predicts AKI progression, with an AUC of 0.87, sensitivity of 87.1%, and specificity of 84.1%. Building upon this prior work, our study adopted a urine volume < 200 mL at 2 h post-FST as the threshold for continuing CRRT treatment in the FST group, thereby providing theoretical support for our approach. However, the optimal timing of urine output assessment after FST has varied from study to study. Rewa et al. [11] demonstrated that AKI progressors produced considerably less urine over a 6-h period than non-progressors (624.13 vs. 1997.79 mL, p < 0.001). Importantly, their results upheld the same <200 mL cutoff threshold while showing superior predictive efficacy of 2-h measures (AUROC: 2h post-FST: 0.87 vs. 6h post-FST 0.86) for stage III AKI progression, which corroborating Chawla’s findings.
Despite multiple randomized controlled trials investigating the optimal timing for CRRT initiation in critically ill patients with AKI, no definitive consensus has been established. There has also been some research on the prediction of the timing of CRRT initiation by FST in patients with severe AKI. The trial by Lumlertgul et al. [12], which randomized FST-negative patients into an early initiation of RRT group and a standard RRT group by evaluating the patients’ responsiveness to diuretics, confirmed that the severity of AKI was less severe in FST-positive patients; whereas the accuracy of predicting the initiation of RRT was significantly higher in FST- negative patients compared to FST-positive patients (13.6 vs. 75%, p < 0.001). Meanwhile, Zhang et al. [13] further demonstrated that FST is an independent predictor of CRRT initiation, showing that patients who were FST-negative had a 2.379-fold higher likelihood of initiating CRRT than FST-positive patients (95% CI: 1.64–3.44, p < 0.001). Multivariable analysis revealed superior predictive performance of 2-h post-FST urine output (optimal cutoff: 156 mL; AUC: 0.966, 95% CI: 0.98–0.99) compared to Scr (optimal cutoff: 115.3 μmol/L; AUC: 0.723, 95% CI: 0.653–0.786) and SOFA score (optimal cutoff: 8; AUC: 0.846, 95% CI: 0.786–0.895), confirming FST clinical utility in guiding CRRT initiation decisions for patients with AKI. The AUC, sensitivity, and specificity of the creatinine threshold reported in that study far exceeded those observed in our research, where Scr showed optimal sensitivity and specificity of 86.4 and 45.3%, at a cutoff of 104.4 μmol/L, with an AUC of 0.641 (95% CI: 0.517–0.764). Notably, FST in combination with biomarkers for CRRT initiation is also increasingly being investigated. The combination of FST and urinary C-C motif chemokine ligand 14 (CCL14) predicted the initiation of CRRT with a significantly higher AUC than with FST or CCL14 alone (FST + CCL14 AUROC: 0.87, 95% CI: 0.82–0.92 vs. FST AUROC: 0.79, 95% CI: 0.74–0.85vs. CCL14 AUROC: 0.83, 95% CI:0.77–0.89, p < 0.05) [14].
Currently, the optimal timing for CRRT withdrawal remains controversial. Klouche et al. [15] suggest that the decision to discontinue CRRT should consider the resolution of the underlying condition causing renal function recovery, fluid overload, and other factors. Numerous studies have identified several potential biomarkers for successful CRRT weaning. Neutrophil gelatinase-associated lipocalin [16] demonstrated optimal predictive performance for CRRT discontinuation in non-septic patients with AKI, with a cutoff value of 403 ng/mL showing the highest specificity and sensitivity (AUROC:0.88, sensitivity of 81% and specificity of 89%). Serum cystatin C [17] showed an independent association with long-term RRT-free survival when levels were < 2.97 mg/L at discontinuation(p < 0.001). The 2-h creatinine clearance [18] proved to be a reliable predictor for avoiding RRT re-initiation within 7 days post-weaning, with sensitivity, specificity, and positive predictive value of 75.5%, 84.4% and 88.8% respectively, at a threshold of 23 mL/min. The 24-h urinary creatinine [19] emerged as an effective and independent marker when 24h-Ucr ≥ 5.2 mmol, the sensitivity, specificity, and positive predictive value are 75%, 88%, and 84%, respectively; Proenkephalin A 119–159 [20] exhibited optimal performance within 90–105 pmol/L, though its correlation with weaning success was weaker than urine output (sHR: 2.35, 95% CI: 1.45–3.81, p < 0.001 vs. sHR: 2.91, 95% CI:1.80–4.73, p < 0.001). A daily urea excretion rate >1.35 mmol/kg/day [21] predicted RRT weaning with higher accuracy than urine output (92.5% vs. 82.1%, p = 0.03) and a greater AUROC (0.96 vs. 0.86). Although these biomarkers may reflect renal recovery and outcomes, their clinical application remains limited due to infrequent measurement, susceptibility to confounding factors, and insufficient accuracy for practical assessment of renal function and prognosis in patients with AKI.
Among all biomarkers, urine output has been the most extensively studied predictor for successful CRRT discontinuation in clinical practice [22]. However, the threshold of urine output used to predict the success of CRRT withdrawal has also shown heterogeneity across studies, and the relationship between diuretic-induced urine output and successful discontinuation of RRT has been less frequently evaluated. A retrospective study by Jeon et al. [23] demonstrated that diuretics and high urine production on the day preceding CRRT weaning correlate with favorable outcomes. In patients undergoing CRRT weaning, a pre-weaning urine output > 191 mL/day demonstrated a sensitivity of 81.2%, a specificity of 71.6%, and an AUROC of 0.821 and even among oliguric patients receiving diuretics, a lower urine output cutoff of 125 mL/day maintained clinically relevant predictive value, with 72.1% sensitivity, 68.8% specificity, and an AUROC of 0.745. A prospective cohort study [24] found that higher urine output (1732.54 ± 1053.23 mL, p = 0.002), higher creatinine clearance (21.34 ± 12.15, p = 0.049), and low CRP values (9.39 ± 7.26 mg/L, p = 0.017) all contributed to successful withdrawal, however, the urine output curve did not show a good AUC, with a precision of 0.62, the cutoff value of 1,920 mL with diuretics in 24 h, and sensitivity and specificity lower than 75%.Furthermore, Uchino et al. [9] demonstrated that spontaneous urine output >400 mL/d at CRRT withdrawal in patients with AKI had the highest predictive value, with a sensitivity of 46.5% and a specificity of 80.9%, while the use of diuretics at CRRT withdrawal strengthened this association, with a higher sensitivity and specificity at 2,330 mL/d after the use of diuretics, with a positive predictive value of 87.9%.In our randomized controlled study of 240 critically ill patients with AKI (117 in the FST group), the FST group demonstrated significantly higher weaning success rates compared to the control group (81.2 vs. 64.2%, p = 0.003) at CRRT discontinuation. Evaluation of post-FST urine output revealed that a 2-h urine volume ≥ 202.5 mL showed superior predictive value for successful CRRT weaning, with higher sensitivity and specificity than Scr (FST 2hUrine Output (UO): OR: 1.008, 95% CI: 1.002–1.014 vs. Scr: OR: 1.000, 95% CI: 0.998–1.002). The AUROC for 2-h post-FST urine output was greater than 0.75, exceeding Scr and 24-h urine output (AUROC: 0.76 vs. 0.641 vs. 0.76), indicating strong clinical usefulness for guiding CRRT termination decisions. In a recent machine learning analysis by Liang et al. [25] involving 976 patients (including 349 successful RRT withdrawal cases), both a positive furosemide stress test (60.2 vs. 40.7%, p < 0.001) and greater urine output in the 3 days preceding RRT withdrawal (755 vs. 125 mL, p < 0.001) were significantly associated with successful discontinuation compared to failed withdrawal cases. Our findings corroborate these results, demonstrating that 2-h post-FST urine volume (OR: 1.008, 95% CI: 1.002–1.014) serves as a significant predictor of successful weaning. Notably, our study extends these observations by providing several novel insights: First, we established a clear association between FST results and subsequent renal recovery, showing that patients with successful CRRT withdrawal following FST had significantly better renal outcomes. This was evidenced by lower RRT re-initiation rates at 7 days (18.8 vs. 35.8%, p < 0.05) and 28 days (29.9 vs. 36.6%, p > 0.05) compared to unsuccessful weaning patients. Second, successful weaning patients were associated with markedly improved 28-day survival outcomes (HR: 0.5956, 95% CI: 0.344–1.031, p = 0.046). This study showed that FST results could effectively distinguish 28-day survival outcomes in AKI patients, confirming the value of early renal tubular function assessment in guiding acute-phase interventions. However, this difference disappeared at 90 days. First, as follow-up time extends, the impact of non-renal complications (such as cardiovascular events and infections) on mortality gradually increases, potentially diluting the predictive efficacy of FST. Second, the sample size was estimated based on short-term outcomes, which may result in insufficient statistical power for the 90-day analysis. This finding highlights a key dimension in the management of AKI patients: FST primarily reflects the kidney’s response capacity to acute stress, whereas long-term prognosis depends more on the progression of underlying diseases and overall recovery. These critical prognostic associations were not previously examined.
This study has some limitations. First, the lack of standardized criteria for initiating and discontinuing CRRT, combined with the unblinded nature of this study, meant that decisions regarding CRRT weaning were based on individual clinician judgment rather than standardized protocols. Attending physicians made weaning decisions by comprehensively assessing renal function recovery based on clinical indicators such as urine output, trends in creatinine clearance, and overall hemodynamic stability. Similarly, the decision to restart CRRT depended on the evolution of the patient’s clinical condition after weaning. The absence of a standardized protocol for CRRT weaning and re-initiation may have introduced variability in decision-making between clinicians, which could act as a confounding factor affecting the study results; however, due to the limitations of various research centers, it is not feasible to obtain identical test conditions in each of them. As a result, the current study did not compare with other related biomarkers or combine urine volume following FST with other related biomarkers. This is a notable limitation. Future studies should incorporate such comparisons to better define the role of the FST within the broader context of AKI biomarker research; Finally, although CRRT-related complications (e.g. thrombosis, arteriovenous fistula, and infections) are known to influence AKI prognosis, systematic monitoring of these adverse events was not performed, limiting comprehensive outcome assessment. Future studies should incorporate prospective collection of complication data to provide a more comprehensive assessment of the risks and benefits associated with FST-guided weaning strategies.
5. Conclusions
According to this prospective multicenter controlled study, a 2-h post-FST urine output of ≥202.5 mL is a better predictor of successful CRRT weaning in critically ill patients with AKI than both 24-h urine output and Scr levels at weaning (p < 0.05). Furthermore, the FST effectively predicts the need for RRT reinitiation within 7 days after CRRT discontinuation (p < 0.05). The clinical application of these findings may significantly improve CRRT weaning success rates in patients with AKI and help standardize the weaning protocol. This represents a paradigm shift from empirical weaning decisions to precision medicine approaches in CRRT management for patients with AKI.
Supplementary Material
Acknowledgments
We are grateful to all the patients, their families, and the participating sites and investigators at Zhejiang Provincial People’s Hospital, Ningbo First Hospital, The First Affiliated Hospital of Wenzhou Medical University, Huzhou Central Hospital, Quzhou People’s Hospital, Jinhua Central Hospital, Lishui People’s Hospital, and Yiwu Central Hospital for making this study possible. X.H. Yang proposed and designed the research protocol; B.B. Zhou, Y. Qu, B. Xu, and S.Y. Tang were involved in the recruitment of the participants and implementation of the protocol; B.B. Zhou, S.Y. Tang, Y.N. Wang and E.N. Ge contributed to the collection of experimental data; B.B. Zhou, Y. Qu analyzed the experimental data; and all authors drafted and substantially revised the work. All authors have read and approved the final version of the manuscript prior to publication.
Glossary
List of abbreviations
- AKI
Acute kidney injury
- AUC
Area under the curve
- APACHE
Acute Physiology and Chronic Health Evaluation
- BUN
Blood urea nitrogen
- CI
Confidence interval
- CRP
C-reactive protein
- CRRT
Continuous renal replacement therapy
- CysC
Cystatin C
- ECMO
Extracorporeal membrane oxygenation
- FST
Furosemide stress test
- Hb
Hemoglobin
- HCT
Hematocrit
- ICU
Intensive care unit
- IQR
Interquartile range
- KDIGO
Kidney Disease: Improving Global Outcomes
- K-M
Kaplan–Meier
- NGAL
Neutrophil gelatinase-associated lipocalin
- OR
Odds ratio
- PCT
Procalcitonin
- ROC
Receiver operating characteristic
- RRT
Renal replacement therapy
- SD
Standard deviation
- SOFA
Sequential organ failure assessment
- Scr
Serum creatinine
- WBC
White blood cell
- MAP
Mean arterial pressure
- IABP
Intra-aortic balloon pump
- CKD
Chronic kidney disease
- CHF
Chronic heart failure
- ARDS
Acute respiratory distress syndrome
- PT
Prothrombin time
- APTT
Activated partial thromboplastin time
- BNP
Brain natriuretic peptide
- pH
Potential of hydrogen
- Lac
Lactate
- eGFR
Estimated glomerular filtration rate
- UO
Urine output
- Alb
Albumin
- 2h-CrCl
2-hour creatinine clearance
- 24h-Ucr
24h urinary creatinine
- PenKid
Proenkephalin A 119–159
- CCL14
Urinary C–C motif chemokine ligand 14
- ESKD
End-stage kidney disease
- HR
Hazard ratio
Funding Statement
This work was supported by Zhejiang Province Key R&D Program Projects(2019C03024).
Disclosure statement
The authors report there are no competing interests to declare.
Data availability statement
The data sets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
References
- 1.Hoste EAJ, Kellum JA, Selby NM, et al. Global epidemiology and outcomes of acute kidney injury. Nat Rev Nephrol. 2018;14(10):607–625. doi: 10.1038/s41581-018-0052-0. [DOI] [PubMed] [Google Scholar]
- 2.Hoste EAJ, Bagshaw SM, Bellomo R, et al. Epidemiology of acute kidney injury in critically ill patients: the multinational AKI-EPI study. Intensive Care Med. 2015;41(8):1411–1423. doi: 10.1007/s00134-015-3934-7. [DOI] [PubMed] [Google Scholar]
- 3.Ostermann M, Basu RK, Mehta RL.. Acute kidney injury. Intensive Care Med. 2023;49(2):219–222. doi: 10.1007/s00134-022-06946-0. [DOI] [PubMed] [Google Scholar]
- 4.Mehta RL. Challenges and pitfalls when implementing renal replacement therapy in the ICU. Crit Care. 2015;19(Suppl 3):S9. doi: 10.1186/cc14727. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Khwaja A. KDIGO clinical practice guidelines for acute kidney injury. Nephron Clin Pract. 2012;120(4):c179–184. doi: 10.1159/000339789. [DOI] [PubMed] [Google Scholar]
- 6.Schiffl H, Lang SM.. Current approach to successful liberation from renal replacement therapy in critically ill patients with severe acute kidney injury: the quest for biomarkers continues. Mol Diagn Ther. 2021;25(1):1–8. doi: 10.1007/s40291-020-00498-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Pon AG, Vairakkani R, Mervin EF, et al. Clinical significance of frusemide stress test in predicting the severity of acute kidney injury. J Bras Nefrol. 2021;43(4):470–477. doi: 10.1590/2175-8239-JBN-2021-0003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Chawla LS, Davison DL, Brasha-Mitchell E, et al. Development and standardization of a furosemide stress test to predict the severity of acute kidney injury. Crit Care. 2013;17(5):R207. doi: 10.1186/cc13015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Uchino S, Bellomo R, Morimatsu H, et al. Discontinuation of continuous renal replacement therapy: a post hoc analysis of a prospective multicenter observational study. Crit Care Med. 2009;37(9):2576–2582. doi: 10.1097/CCM.0b013e3181a38241. [DOI] [PubMed] [Google Scholar]
- 10.Youden WJ. Index for rating diagnostic tests. Cancer. 1950;3(1):32–35. doi: 10.1002/1097-0142(1950)3:1<32::AID-CNCR2820030106>3.0.CO;2-3. [DOI] [PubMed] [Google Scholar]
- 11.Rewa OG, Bagshaw SM, Wang X, et al. The furosemide stress test for prediction of worsening acute kidney injury in critically ill patients: a multicenter, prospective, observational study. J Crit Care. 2019;52:109–114. doi: 10.1016/j.jcrc.2019.04.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Lumlertgul N, Peerapornratana S, Trakarnvanich T, et al. Early versus standard initiation of renal replacement therapy in furosemide stress test non-responsive acute kidney injury patients (the FST trial). Crit Care. 2018;22(1):101. doi: 10.1186/s13054-018-2021-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Zhang K, Zhang H, Zhao C, et al. The furosemide stress test predicts the timing of continuous renal replacement therapy initiation in critically ill patients with acute kidney injury: a double-blind prospective intervention cohort study. Eur J Med Res. 2023;28(1):149. doi: 10.1186/s40001-023-01092-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Meersch M, Weiss R, Gerss J, et al. Predicting the development of renal replacement therapy indications by combining the furosemide stress test and chemokine (CC motif) ligand 14 in a cohort of postsurgical patients. Crit Care Med. 2023;51(8):1033–1042. doi: 10.1097/CCM.0000000000005849. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Klouche K, Gibney RTN, Forni LG.. Can this patient be safely weaned from RRT?. Intensive Care Med. 2018;44(5):639–642. doi: 10.1007/s00134-017-4948-0. [DOI] [PubMed] [Google Scholar]
- 16.Chen X, Chen Z, Wei T, et al. The effect of serum neutrophil gelatinase-associated lipocalin on the discontinuation of continuous renal replacement therapy in critically ill patients with acute kidney injury. Blood Purif. 2019;48(1):10–17. doi: 10.1159/000499026. [DOI] [PubMed] [Google Scholar]
- 17.Yang T, Sun S, Lin L, et al. Predictive factors upon discontinuation of renal replacement therapy for long-term chronic dialysis and death in acute kidney injury patients. Artif Organs. 2017;41(12):1127–1134. doi: 10.1111/aor.12927. [DOI] [PubMed] [Google Scholar]
- 18.Fröhlich S, Donnelly A, Solymos O, et al. Use of 2-hour creatinine clearance to guide cessation of continuous renal replacement therapy. J Crit Care. 2012;27(6):744.e1–5–744.e5. doi: 10.1016/j.jcrc.2012.08.012. [DOI] [PubMed] [Google Scholar]
- 19.Viallet N, Brunot V, Kuster N, et al. Daily urinary creatinine predicts the weaning of renal replacement therapy in ICU acute kidney injury patients. Ann Intensive Care. 2016;6(1):71. doi: 10.1186/s13613-016-0176-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.von Groote T, Albert F, Meersch M, et al. Evaluation of proenkephalin A 119–159 for liberation from renal replacement therapy: an external, multicenter pilot study in critically ill patients with acute kidney injury. Crit Care. 2023;27(1):276. doi: 10.1186/s13054-023-04556-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Aniort J, Ait Hssain A, Pereira B, et al. Daily urinary urea excretion to guide intermittent hemodialysis weaning in critically ill patients. Crit Care. 2016;20(1):43. doi: 10.1186/s13054-016-1225-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Klouche K, Brunot V, Larcher R, et al. Weaning from kidney replacement therapy in the critically ill patient with acute kidney injury. J Clin Med. 2024;13(2):579. doi: 10.3390/jcm13020579. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Jeon J, Kim DH, Baeg SI, et al. Association between diuretics and successful discontinuation of continuous renal replacement therapy in critically ill patients with acute kidney injury. Crit Care. 2018;22(1):255. doi: 10.1186/s13054-018-2192-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Suarez MIR, Ohara CN, Kitawara KAH, et al. Successful liberation from acute kidney replacement therapy in critically ill patients: a prospective cohort study. Blood Purif. 2024;53(2):96–106. doi: 10.1159/000534103. [DOI] [PubMed] [Google Scholar]
- 25.Liang Q, Xu X, Ding S, et al. Prediction of successful weaning from renal replacement therapy in critically ill patients based on machine learning. Ren Fail. 2024;46(1):2319329. doi: 10.1080/0886022X.2024.2319329. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data sets used and/or analyzed during the current study are available from the corresponding author on reasonable request.