Efficacy of Continuous Renal Replacement Therapy and Intermittent Hemodialysis in Patients with Renal Failure in Intensive Care Unit: A Systemic Review and Meta-analysis

Objective This study aimed to compare the clinical efficacy of continuous renal replacement therapy (CRRT) and intermittent hemodialysis (IHD) in patients with renal failure in intensive care unit (ICU). Methods Relevant studies were searched in the databases including EMBASE, Cochrane Library, and MEDLINE (PubMed) from inception to January 04, 2021. The inclusion of available studies and the collection of data were independently conducted by two authors after reviewing the full text. Pooled analyses of relative risk (RR) and weighted mean difference (WMD) were performed to compare the outcomes of renal recovery, short-term mortality, length of ICU stays, and length of in-hospital stays between the two different treatment groups. Publication bias was assessed by the funnel plot. Results A total of 11 RCT studies including 1740 patients with renal failure were eligible for final analysis. Among them, 894 patients (51.4%) underwent CRRT and 846 patients (48.6%) received IHD. Pooled analysis did not find significant differences in renal recovery and short-term mortality between the two groups. Interestingly, patients underwent CRRT showed significantly shorter length of ICU stay and in-hospital stay than those who underwent IHD (ICU stay: RR: −0.61, 95%CI: −1.10—−0.11, P < 0.05; I2 = 93.6%; in-hospital stay: RR: −0.56, 95%CI: −1.41–0.28, P < 0.05; I2 = 97.7%). No significant publication biases were observed on the funnel plots. Conclusion Compared with IHD, CRRT had similar effects on renal recovery and short-term mortality in patients with renal failure in ICU. As a promising technique in clinical practice, CRRT could significantly reduce the length of ICU stay and in-hospital stay of patients, which was of great significance for the reduction of medical costs and the long-term benefits of patients, thereby reducing the burden on society and individuals.


Introduction
Acute kidney injury (AKI) is a clinical syndrome characterized by an acute decrease in glomerular fltration rate [1][2][3]. Mild patients with AKI have no obvious symptoms and recover quickly, whereas severe acute patients have a higher mortality rate [4]. Due to the reduced glomerular fltration rate, metabolic wastes cannot be discharged in time, which will afect the balance of water and electrolytes in body. In severe cases, kidney injury can lead to multiple organ failure, endangering the life of the patient. Te incidence of AKI in general population was only 5%, and it was signifcantly higher in intensive care unit (ICU) patients, with an incidence of 36-70% and a mortality rate of 40-70% [5]. In China, the number of AKI patients reached 2.9 million in 2013, of which 700,000 died, and the treatment cost was 13 billion US dollars [6], causing a huge disease burden to the society.
At present, one of the efective clinical treatments for AKI is blood purifcation, among which intermittent hemodialysis (IHD) and continuous renal replacement therapy (CRRT) are more commonly used [7]. IHD adopts the principle of difuse convection to remove harmful substances such as infammatory mediators and endogenous and exogenous toxins in the blood of patients through dialysis equipment, which has a certain efect on improving the patient's condition [8]. However, due to the characteristics of signifcant changes in the composition of dialysate, blood gas parameters, and blood membrane reaction, patients are prone to hypotension after solute and water are removed in a short period of time. Meanwhile, the removal of small molecular substances such as norepinephrine and epinephrine will aggravate the hypotension, which has a serious impact on the recovery of renal function, resulting in decreased survival [9].
In recent years, with the development of medical technology, CRRT has been widely used in the treatment of patients with severe acute renal failure [10]. Similar to IHD, CRRT removes infammatory mediators mainly by convection and adsorption, and it continuously and slowly removes water and solutes in the body [11]. Te major advantage of CRRT is the stable hemodynamics. Because dehydration is slow, the incidence of hypotension and arrhythmia is low, and there is no renal ischemia, which is conducive to the recovery of renal function. Currently, there is still controversy about the selection of treatment strategy for the treatment of patients with AKI in ICU. Identifying the best treatment for renal impairment in patients with renal failure in ICU and clarifying the specifc advantages of various treatment options are important to efectively improve patient outcomes and reduce the economic burdens of social and families.
Herein, we conducted this systemic review and metaanalysis aimed to compare the clinical efcacy of CRRT and IHD in patients with renal failure in ICU and further provided evidence for helping clinicians to make a better treatment strategy.

Search of Studies.
We performed the systemic review and meta-analysis by strictly following the instructions and requirements in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [12]. Te key words "continuous renal replacement therapy" and "intensive care units" and "kidney failure, chronic" or "renal insufciency" or "acute kidney injury" and "randomized controlled trial" were used for searching the relevant studies in the databases including Cochrane Library, EMBASE, and MEDLINE (PubMed) from inception to January 04, 2021. Two authors independently conducted the literature search. First, the potentially relevant literature studies were retrieved from the abovementioned databases; second, these literature studies were screened by reviewing the tittle and abstract; third, the studies meeting the inclusion/exclusion criteria were eventually included after carefully reading the full-text. If there were discrepancies on the inclusion/exclusion of a certain study, a senior researcher was consulted and the study was determined by comprehensive discussion.

Inclusion and Exclusion of Studies.
Studies meeting all the following inclusion criteria were eligible: (1) patients included were those with renal failure in the ICU; (2) type of studies were randomized controlled study (RCT); (3) studies directly comparing CRRTmodalities to IHD modalities; and (4) studies with ethic statement and signed informed consent forms.
Te following studies were excluded: (1) included patients had undergone hemodialysis before admitting to the ICU; (2) included patients with acute drug poisoning; (3) studies with incomplete result data; (4) studies published in the form of reviews, case reports, systemic reviews, and conference abstract.

Extraction of Data.
After collecting all the eligible studies, the relevant data were extracted and recorded by the abovementioned two investigators. Any inconsistencies were resolved by consensus. Data for study identifer, the author, year of publication, country, study design, setting, sample size, mean age, proportion of males, other population characteristics, disease diagnosis, primary outcomes, secondary outcomes, risks, ethical approval, study limitations, and other important information were collected with a predesigned form.

Outcomes.
Te primary outcome was the recovery of kidney function. Te secondary outcomes included numbers of patients free of CRRT after discontinuing CRRT, days in ICU, days in hospital, mortality, serum creatinine (SCr), blood urea nitrogen (BUN), and hemodynamic index.

Evaluation of Bias and Study Quality.
All the eligible studies were RCTs, so the Revised Cochrane Risk of Bias Tool (RoB2.0) was used for assessing the risk of bias and quality of each study. Te publication bias was evaluated by drawing funnel plots and visually assessing the symmetry of the plots. Te quality of studies and evaluation of bias were shown in supplementary Figure 1 for details.

Statistical
Analyses. Data were analyzed by Stata 14.0, and the pooled analysis was performed by Review Manager 5.0. For studies with binary variables and continuous variables, we collected the relative risk (RR) and standard mean diference (SMD). Te between-study heterogeneity was evaluated by the Cochran Q test and I 2 statistic. A signifcant heterogeneity was indicated by a P < 0.01 in Q test or I 2 statistic >50%. In the meta-analysis, the random-efect model was adopted in the analysis of quantitative data. If I 2 statistic ≤50% or P ≥ 0.01 in Q test, it was indicated no signifcant heterogeneity, and the fxed-efect model was used for pooled analysis.

Inclusion of Studies.
After searching the literature studies in the abovementioned databases with the key words, a total of 811 articles were retrieved. Among them, 128 articles were duplicated literature and were excluded. By reviewing the title, 41 relevant articles were obtained. From these articles, 18 literature studies were excluded after carefully reading the abstract. Eventually, a total of 11 articles [13][14][15][16][17][18][19][20][21][22][23] were included since 12 articles did not meet the inclusion criteria (6 articles were cohort studies or reviews; 3 studies were lack of control group; 1 article with unavailable data; and 2 articles with the same study population or from the same institution).   A total of 11 RCT studies with 1740 patients with renal failure were eventually included. Among them, 894 patients (51.4%) underwent CRRT and 846 patients (48.6%) received IHD. As for the gender composition ratio, there were 1026 males (62.7%) and 610 females (37.3%). In terms of the CRRT in these 11 studies, continuous venous-venous hemodialyses (CVVH) were used in 9 studies, continuous arterial-venous hemodialysis was used in 1 study, and continuous hemodialysis fltration was used in 1 study. Te characteristics of included studies are shown in Table 1 [13,18,[20][21][22]. Metaanalysis showed that 170 and 176 patients obtained renal recovery after the CRRT and IHD, respectively. Pooled analysis did not fnd signifcant diferences in renal recovery between the two treatment strategies (RR: 1.00, 95% CI: 0.84-1.18, P � 0.112; I 2 � 46.6%, Figure 1). Te funnel plot did not show signifcant asymmetry (Supplementary Figure 2), indicating that there was no signifcant publication bias of these studies.

Length of ICU Stay.
In order to further compare the efcacy of CRRT and IHD in patients with renal failure in ICU, we conducted a pooled analysis of the length of ICU stay of these patients. Te results of 5 included studies showed that CRRT signifcantly shortened the length of ICU stay of patients compared with that of IHD (RR: −0.61, 95% CI: −1.10-−0.11, P < 0.05; I 2 � 93.6%, Figure 3). No signifcant publication bias was observed on the funnel plot (Supplementary Figure 4).

Length of In-Hospital Stay.
Finally, we compared the length of in-hospital stay of patients who underwent different treatment strategies. Te data were obtained from 5 studies. Te results of meta-analysis showed that the length of in-hospital stay was signifcantly shorter in the CRRT group than that of the IHD group (RR: −0.56, 95% CI: −1.41-0.28, P < 0.05; I 2 � 97.7%, Figure 4), which was similar to the results of length of ICU stay. We did not fnd signifcant publication bias of these studies according to the funnel plot (Supplementary Figure 5).

Discussion
In this study, we found that the CRRT had similar efcacy in renal recovery and short-term mortality compared with that of IHD in patients with renal failure in ICU. However, patients who underwent CRRT showed signifcant shorter length of ICU stay and in-hospital stay than that of patients received IHD. It highlighted the advantage of CRRT in ICU patients with renal failure. AKI is one of the complications of critically ill patients, with high morbidity and mortality [24]. Te main causes of AKI include sepsis, liver failure, cardiac surgery, hypovolemia, and antibiotic-related diseases. Clinically, it mainly manifests as elevated serum creatinine and decreased glomerular fltration rate, and the disease condition develops rapidly. Severe patients will cause anemia, nitrogen retention, dyspnea, if not promptly treated may endanger the patient's life [25]. Especially for patients with severe AKI, the prognosis is relatively poor due to the unstable hemodynamic status, high catabolism, and high-volume load [26]. In recent years, the incidence of AKI has been increasingly elevated, but there is no reliable clinical treatment. So, it is of great clinical signifcance to explore efective and appropriate treatment strategies for patients with renal failure in hospitals, especially in the ICU.
Currently, blood purifcation is usually used to remove toxins and infammatory substances in patients and to stabilize the cell structure, microenvironment, and function to achieve the purpose of treatment [27]. At present, currently commonly used blood purifcation techniques for the treatment of AKI include IHD and CRRT [28]. IHD removes toxic substances such as toxins and infammatory mediators from the patient's blood mainly by difusion and convection, thereby purifying the blood and keeping the internal environment in a stable state [29]. However, the disadvantages of IHD include volume overload, hypercatabolism, and hemodynamic stability. If the patient is not efectively treated, ascites, cerebral edema, pleural efusion, and hyponatremia are likely to occur [30]. Systemic diseases such as heart failure and gastrointestinal bleeding will seriously aggravate the patient's condition and even adversely afect the patient's life. For patients with severe acute renal failure, it is usually accompanied by organ dysfunction, coupled with changes in the patient's metabolic function, which is easy to produce overload conditions [31]. Terefore, it is important to explore an efective and appropriate way to treat acute renal failure in clinical practice, optimize the adverse symptoms of patients, and improve the treatment efect. In recent years, CRRT has been widely used in clinical practice due to its advantages of stable hemodynamics, high solute clearance, and good biocompatibility. CRRT can continuously and slowly remove solutes in the body through adsorption, difusion, and convection and can replace the fuid daily, of which the clearance rate is higher, helping to reduce the patient's systemic infammatory response and improve the patient's immune function [32]. It can also provide nutritional support, stabilize the internal environment of the patient's body, and improve the prognosis of the patient. Studies have shown that CRRT treatment can remove infammatory factors and reduce organ damage, thereby better maintaining the stability of vascular and intracellular and extracellular ion levels, acid-base balance, and osmotic pressure [33]. CRRT is therefore a promising technique to achieve better clinical outcomes. Our pooled results showed that CRRT showed similar efcacy in improving renal function and short-term mortality in patients with renal failure, further confrming the important role of CRRT in the treatment of renal impairment. ICU is a place where modern medical theory and high-tech modern equipment were applied to conduct specialized and centralized monitoring, treatment, and nursing for critically ill patients [34]. Most of the patients in the ICU are in serious disease conditions. If patients are not treated appropriately, the length of hospital stay is prolonged and the incidence of complications increases. CRRT is widely used in clinical practice, especially in the ICU [35]. It can maintain the stability of hemodynamic parameters and improve the clinical symptoms and organ function of patients, which are important for reducing complications. In this study, we found that CRRT shortened the length of ICU stay and in-hospital stay in patients with renal failure, which has social and individual signifcance for reducing the heavy economic burden of renal failure.

4.1.
Limitations. Tis study has several limitations. First, the number of included studies for analysis of renal recovery, length of ICU stay, and length of in-hospital stay is relatively small. More RCTs should be conducted in the future to further investigate the efcacy of these two strategies. Secondly, due to the lack of relevant data, we only investigated the health economic indicators including length of ICU stay and in-hospital stay. If sufcient data are available, other economic indicators such as the cost of illness should be analyzed. Tirdly, since the included studies for the renal recovery, length of ICU stay, and in-hospital stay are rare, it is difcult to precisely assess the publication bias of these studies. In-depth investigations with more studies are needed to further confrm the conclusions of this study.

Conclusion
Our comprehensive systematic review including 11 RCTs provides the available information on two common interventions to combat the renal failure in the ICU ward. Compared with IHD, CRRT had similar efects on renal recovery and short-term mortality in patients with renal failure in ICU. As a promising technique in clinical practice, CRRT could signifcantly reduce the length of ICU stay and in-hospital stay of patients, which was of great signifcance for the reduction of medical costs and the long-term benefts of patients, thereby reducing the burden on society and individuals.

Data Availability
Te datasets in this study are available from the corresponding author on reasonable request.

Conflicts of Interest
Te authors declare that they have no conficts of interest.

Acknowledgments
Tis study was supported by major scientifc and technological projects of Changzhou Municipal Health Commission (ZD201918).