Can this patient be safely weaned from RRT?

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.

Weaning from Kidney Replacement Therapy in the Critically Ill Patient with Acute Kidney Injury

Around 10% of critically ill patients suffer acute kidney injury (AKI) requiring kidney replacement therapy (KRT), with a mortality rate approaching 50%. Although most survivors achieve sufficient renal recovery to be weaned from KRT, there are no recognized guidelines on the optimal period for weaning from KRT. A systematic review was conducted using a peer-reviewed strategy, combining themes of KRT (intermittent hemodialysis, CKRT: continuous veno-venous hemo/dialysis/filtration/diafiltration, sustained low-efficiency dialysis/filtration), factors predictive of successful weaning (defined as a prolonged period without new KRT) and patient outcomes. Our research resulted in studies, all observational, describing clinical and biological parameters predictive of successful weaning from KRT. Urine output prior to KRT cessation is the most studied variable and the most widely used in practice. Other predictive factors, such as urinary urea and creatinine and new urinary and serum renal biomarkers, including cystatin C and neutrophil gelatinase-associated lipocalin (NGAL), were also analyzed in the light of recent studies. This review presents the rationale for early weaning from KRT, the parameters that can guide it, and its practical modalities. Once the patient's clinical condition has stabilized and volume status optimized, a diuresis greater than 500 mL/day should prompt the intensivist to consider weaning. Urinary parameters could be useful in predicting weaning success but have yet to be validated.

When to discontinue renal replacement therapy. what do we know?

PURPOSE OF REVIEW

Acute kidney injury is common in intensive care patients. Supportive care involves the use of renal replacement therapies as organ support. Initiation of renal replacement therapy has been the subject of much interest over the last few years with several randomised controlled studies examining the optimal time to commence treatment. In contrast to this, little evidence has been generated regarding cessation of therapy. Given that this treatment is complex, not without risk and expensive it seems timely that efforts should be expended at examining this vexing issue.

RECENT FINDINGS

Although several studies have been reported examining the successful discontinuation of renal replacement therapies all studies reported to-date are observational in nature. Conventional biochemical criteria have been used as well as physiological parameters including urine output. More recently, more novel biomarkers of renal function have been studied. Although to-date no optimal variable nor threshold for discontinuation can be established.

SUMMARY

Several variables have been described which may have a role in determining which patients may be successfully weaned from renal replacement therapy. However, few have been exposed to vigorous examination and evidence is sparse in support of any potential approach although urine output currently is the most often described. More recently novel biomarkers have also been examined but again are limited by study design and heterogeneity. Further research is clearly needed focussing on proposed variables preferably in multivariate models to improve predictive ability and successful cessation of therapy.

Anticipation of recovery of native renal function and liberation from renal replacement therapy in critically ill patients with severe acute kidney injury

Background

Renal replacement therapy (RRT) is used to manage critically ill patients with severe acute kidney injury (AKI-D), and it is undoubtedly life-sustaining for most patients. However, the prolonged unnecessary use of these techniques may be harmful. At present, no consensus guidelines provide specific recommendations for clinicians on when (optimal timing of discontinuation) and how (liberation or weaning) to stop RRT in intensive care unit (ICU) patients with recovering native kidney function.

Methods and results

Numerous variables such as clinical parameters, classical surrogate markers for glomerular filtration rate, novel biomarkers of kidney function and damage, and new imaging techniques in AKI-D have been described to predict successful discontinuation of RRT. Most available studies are limited by study design, heterogeneity of variable assessment and thresholds of biomarkers, and lack of prospective validation. At present, the decision on discontinuation of RRT in ICU patients is based on three clinical scenarios: (a) intrinsic kidney function (defined as spontaneous urine output > 500 ml/24 h, timed creatinine clearance > 15 to 20 ml/min) has adequately improved to match the demands and continued RRT is no longer consistent with goals of care (transition to intermittent RRT); (b) the acute illness that prompted RRT has improved; (c) the clinical practice of switching haemodynamic stable patients with persistent AKI-D from continuous RRTs to intermittent RRTs is variable, but de-escalation of RRT (frequency, dose) may facilitate mobilization and discharge of ICU patients.

Conclusions

The predictive ability of novel kidney biomarkers, surrogate markers of kidney function, and direct measurements of kidney function should be evaluated in future studies.

Current Approach to Successful Liberation from Renal Replacement Therapy in Critically Ill Patients with Severe Acute Kidney Injury: The Quest for Biomarkers Continues

Recovery of sufficient kidney function to liberate patients with severe acute kidney injury (AKI-D) from renal replacement therapy (RRT) is recognized as a vital patient-centred outcome. However, no clinical consensus guideline provides specific recommendations on when and how to stop RRT in anticipation of renal recovery from AKI-D. Currently, wide variations in clinical practice regarding liberation from RRT result in early re-start of RRT to treat uraemia after premature liberation or in the unnecessary prolonged exposure of unwell patients after late liberation. Observational studies, predominantly retrospective in nature, have attempted to assess numerous surrogate markers of kidney function or of biomarkers of kidney damage to predict successful liberation from RRT. However, a substantial heterogeneity in the timing of measurement and cut-off values of most biomarkers across studies allows no pooling of data, and impedes the comparison of outcomes from such studies. The accuracy of most traditional and novel biomarkers cannot be assessed reliably. Currently, the decision to discontinue RRT in AKI-D patients relies on daily clinical assessments of the patient’s status supplemented by measurement of creatinine clearance (> 15 ml/min) and 24-h urine output (> 2000 ml/min with diuretics). Clinical trials objectively comparing the success of validated biomarkers for guiding optimal timed liberation from RRT in AKI-D will be required to provide high-quality evidence for guidelines.

Determining the optimal time for liberation from renal replacement therapy in critically ill patients: a systematic review and meta-analysis (DOnE RRT)

INTRODUCTION

Renal replacement therapy (RRT) is associated with high mortality and costs; however, no clinical guidelines currently provide specific recommendations for clinicians on when and how to stop RRT in recovering patients. Our objective was to systematically review the current evidence for clinical and biochemical parameters that can be used to predict successful discontinuation of RRT.

METHODS

A systematic review and meta-analysis were performed with a peer-reviewed search strategy combining the themes of renal replacement therapy (IHD, CRRT, SLED), predictors of successful discontinuation or weaning (defined as an extended period of time free from further RRT), and patient outcomes. Major databases were searched and citations were screened using predefined criteria. Studied parameters were reported and, where possible, data was analyzed in the pooled analysis.

RESULTS

Our search yielded 23 studies describing 16 variables for predicting the successful discontinuation of RRT. All studies were observational in nature. None were externally validated. Fourteen studies described conventional biochemical criteria used as surrogates of glomerular filtration rate (serum urea, serum creatinine, creatinine clearance, urine urea excretion, urine creatinine excretion). Thirteen studies described physiologic parameters such as urine output before and after cessation of RRT, and 13 studies reported on newer kidney biomarkers, such as serum cystatin C and serum neutrophil gelatinase-associated lipocalin (NGAL). Six studies reported sensitivity and specificity characteristics of multivariate models. Urine output prior to discontinuation of RRT was the most-studied variable, with nine studies reporting. Pooled analysis found a sensitivity of 66.2% (95% CI, 53.6-76.9%) and specificity of 73.6% (95% CI, 67.5-79.0%) for urine output to predict successful RRT discontinuation. Due to heterogeneity in the thresholds of urine output used across the studies, an optimal threshold value could not be determined.

CONCLUSIONS

Numerous variables have been described to predict successful discontinuation of RRT; however, available studies are limited by study design, variable heterogeneity, and lack of prospective validation. Urine output prior to discontinuation of RRT was the most commonly described and robust predictor. Further research should focus on the determination and validation of urine output thresholds, and the evaluation of additional clinical and biochemical parameters in multivariate models to enhance predictive accuracy.

Feasibility of a protocol to wean patients from continuous renal replacement therapy: A retrospective pilot observation

PURPOSE

To evaluate the feasibility of a protocol-based algorithm to wean acute kidney injury (AKI) patients from continuous renal replacement therapy (CRRT).

METHODS

The protocol was introduced on one of two similarly equipped ICUs, while on the other (reference) ICU, CRRT discontinuation was based on clinical judgement. Patients were allocated to either ICU and were subjected to physician- or protocol-directed weaning, respectively. According to the algorithm, periodical withdrawal trials (WTs) were mandatory. Interventions were recommended (administration of diuretics, fluid, vasopressors, inotropes, or human albumin) to achieve specific goals (sufficient urine output, balanced fluid status, adequate renal perfusion pressure, optimal oxygen delivery, normoalbuminemia). Clearly stated criteria defined when to abort a WT and to resume RRT for one cycle, followed by another WT.

RESULTS

Urine output and S_cvO₂ during WTs were higher with protocol-directed weaning, as well as the amount of administered fluids. WT abort ratio was 48% with a tendency to prolonged WT duration, compared to 64% in the reference patients. No relevant adverse side effects were observed.

CONCLUSION

Our data show the feasibility of a structured approach to wean AKI patients from RRT that bundles established interventions and brings the weaning into the physician's focus.

Determining the optimal time for liberation from renal replacement therapy in critically ill patients: protocol for a systematic review and meta-analysis (DOnE RRT)

INTRODUCTION

Renal replacement therapy (RRT) is a complex and expensive form of life-sustaining therapy, reserved for our most acutely ill patients. While a number of randomised trials have evaluated the optimal timing to start RRT among critically ill patients in the intensive care unit (ICU), there has been a paucity of trials providing guidance on when and under what circumstances to ideally liberate a patient from RRT. We are conducting a systematic review and meta-analysis to identify clinical and biochemical markers that predict kidney recovery and successful liberation from acute RRT among critically ill patients with acute kidney injury.

METHODS AND ANALYSIS

Our comprehensive search strategy was developed in consultation with a research librarian and independently peer-reviewed by a second librarian. We will search electronic databases: Ovid Medline, Ovid Embase and Wiley Cochrane Library. Selected grey literature sources will also be searched. Our search strategies will focus on concepts related to RRT (ie, intermittent haemodialysis, slow low-efficiency dialysis, continuous renal replacement therapy), intensive care (ie, involving any ICU setting) and discontinuation of therapy (ie, either clinical, physiological and biochemical parameters of weaning acute RRT) from 1990 to October 10, 2017. Citation screening, selection, quality assessment and data abstraction will be performed in duplicate. Studies will, where possible, be pooled in statistical meta-analysis. When deemed sufficiently clinically homogenous, and we have four or more studies reporting, sensitivities and specificities will be pooled simultaneously using a hierarchical summary receiver operator characteristic curve and bivariate analysis.

ETHICS AND DISSEMINATION

Our systematic review will synthesise the literature on clinical and biochemical markers that predict liberation from RRT. Research ethics approval is not required.

TRIAL REGISTRATION NUMBER

CRD42018074615.