Skip to main content
Download PDF

Anticoagulation options for continuous renal replacement therapy in critically ill patients: a systematic review and network meta-analysis of randomized controlled trials

Critical Care volume 27, Article number: 222 (2023) Cite this article

Abstract

Background

Continuous renal replacement therapy (CRRT) is a widely used standard therapy for critically ill patients with acute kidney injury (AKI). Despite its effectiveness, treatment is often interrupted due to clot formation in the extracorporeal circuits. Anticoagulation is a crucial strategy for preventing extracorporeal circuit clotting during CRRT. While various anticoagulation options are available, there were still no studies synthetically comparing the efficacy and safety of these anticoagulation options.

Methods

Electronic databases (PubMed, Embase, Web of Science, and the Cochrane database) were searched from inception to October 31, 2022. All randomized controlled trials (RCTs) that examined the following outcomes were included: filter lifespan, all-cause mortality, length of stay, duration of CRRT, recovery of kidney function, adverse events and costs.

Results

Thirty-seven RCTs from 38 articles, comprising 2648 participants with 14 comparisons, were included in this network meta-analysis (NMA). Unfractionated heparin (UFH) and regional citrate anticoagulation (RCA) are the most frequently used anticoagulants. Compared to UFH, RCA was found to be more effective in prolonging filter lifespan (MD 12.0, 95% CI 3.8 to 20.2) and reducing the risk of bleeding. Regional-UFH plus Prostaglandin I2 (Regional-UFH + PGI2) appeared to outperform RCA (MD 37.0, 95% CI 12.0 to 62.0), LMWH (MD 41.3, 95% CI 15.6 to 67.0), and other evaluated anticoagulation options in prolonging filter lifespan. However, only a single included RCT with 46 participants had evaluated Regional-UFH + PGI2. No statistically significant difference was observed in terms of length of ICU stay, all-cause mortality, duration of CRRT, recovery of kidney function, and adverse events among most evaluated anticoagulation options.

Conclusions

Compared to UFH, RCA is the preferred anticoagulant for critically ill patients requiring CRRT. The SUCRA analysis and forest plot of Regional-UFH + PGI2 are limited, as only a single study was included. Additional high-quality studies are necessary before any recommendation of Regional-UFH + PGI2. Further larger high-quality RCTs are desirable to strengthen the evidence on the best choice of anticoagulation options to reduce all-cause mortality and adverse events and promote the recovery of kidney function.

Trial registration The protocol of this network meta-analysis was registered on PROSPERO (CRD42022360263). Registered 26 September 2022.

Background

Continuous renal replacement therapy (CRRT) is a standard therapy for critically ill patients with acute kidney injury (AKI) in the intensive care unit (ICU) [1]. In contrast to traditional forms of renal replacement therapy (RRT), such as intermittent hemodialysis and peritoneal dialysis, CRRT can deliver solute clearance and acid–base regulation for patients with unstable hemodynamic status [2]. Although CRRT is prescribed as a continuous dialysis therapy, unexpected downtime can occur due to clot formation in the extracorporeal circuits [3]. Frequent clotting in the extracorporeal circuits limits the benefits of CRRT. Extracorporeal circuit clotting can cause potential blood loss that may affect hemodynamic stability and require frequent circuit changes [4, 5], increasing the workload of medical staff and treatment costs [6]. Therefore, effective strategies to prevent clotting in extracorporeal circuits during CRRT are crucial. In addition to catheter choice, blood flow rate, and therapy choice, anticoagulation is also an important and commonly used strategy to prevent clot formation during CRRT [7].

Anticoagulation is essential to prevent extracorporeal circuits clotting during CRRT. Currently, various anticoagulation options have been used in clinical settings, and unfractionated heparin (UFH) is the most commonly used anticoagulant [8] due to its low cost, easy monitoring, and simple reversal [9, 10]. However, the increased risk of bleeding and heparin-induced thrombocytopenia type II (HIT-II) might lead to life-threatening complications [8]. Regional citrate anticoagulation (RCA) is another common anticoagulation option for CRRT. RCA has also been recommended as a suitable form of CRRT anticoagulation for its longer circuit life, and safety even in patients with liver dysfunction [11]. However, potential disturbances, such as citrate toxicity, metabolic alkalosis, and hypocalcemia, might be observed in critically ill patients [12]. Other anticoagulation options such as low-molecular-weight heparin (LMWH), nafamostat mesilate (NM), Prostaglandin I2 (PGI2), and Regional-UFH are also available during CRRT [13,14,15,16]. Various anticoagulation options have been used during CRRT, but the efficacy and safety of these anticoagulants remain controversial. To date, no studies have synthetically compared the effects of all these anticoagulation options. Therefore, we conducted a network meta-analysis (NMA) to provide the most recent available evidence on the best choice of anticoagulant during CRRT.

Methods

We conducted a systematic review with NMA in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines [17], and the protocol of this NMA was registered on PROSPERO (CRD42022360263).

Search strategy

We systematically searched all relevant publications without language restriction in PubMed, Embase, Web of Science, and the Cochrane database from inception to October 31, 2022. The search terms were as follows: “renal replacement therapy,” “continuous venovenous hemofiltration,” “CVVH,” “continuous venovenous hemodialysis’’, “CVVHD’’, “continuous venovenous hemodiafiltration’’, “CVVHDF,” “slow continuous ultrafiltration,” or “SCUF”; “anticoagulation,” “citrate,” “heparin,” “UFH,” “LMWH,” “dalteparin,” “nadroparin,” “enoxaparin,” “bivalirudin,” “prostacyclin,” “nafamostat,” “hirudin,” “iloprost,” or “tirofiban.” In addition, proceedings from the references that were listed in all retrieved articles, and eligible studies from published meta-analyses were also searched to ensure a complete identification of all eligible studies. The detailed search strategy is presented in Additional File 1. The search process was performed and confirmed by two independent reviewers (ZFZ and CL).

Selection criteria

Participants: Adult critically ill patients receiving CRRT in the ICU.

Interventions: Interventions were regarded as pharmacological interventions for preventing clotting of extracorporeal circuits during CRRT with any type of anticoagulant.

Types of outcome measures: filter lifespan or occurrence of filter clotting, all-cause mortality, length of stay, duration of CRRT, recovery of kidney function, adverse events, and costs.

Type of studies: all randomized controlled trials (RCTs or cluster RCTs) and quasi-RCTs concerning pharmacological anticoagulation during CRRT.

The inclusion criteria were as follows: (1) RCTs concerning pharmacological anticoagulation during CRRT in critically ill patients; (2) studies reporting the following outcomes: filter lifespan or occurrence of filter clotting, all-cause mortality, length of hospital or ICU stay, duration of CRRT, recovery of kidney function, adverse events, and costs; and (3) sufficient data available to calculate risk ratios (RRs) and mean differences (MDs) with 95% confidence intervals (CIs).

Exclusion criteria were: (1) no relevant data; (2) conference abstracts without full-text manuscripts; (3) nonhuman studies or studies that enrolled pediatric patients; (4) studies that enrolled patients receiving dialysis treatment before admission to the ICU; and (5) studies that compared different doses of the same pharmacological anticoagulation.

Data extraction and quality assessment

Two investigators (ZFZ and YYY) independently screened the titles and abstracts of the articles, and extracted data from the included studies. Any disagreements were resolved by a third author (LZ). The following data from each study were extracted: (1) study characteristics: publication year, author names, study design, target population, number of participants, age, sex, and so on; (2) type of anticoagulation during CRRT, CRRT modalities, dilution, and blood flow rate; and (3) outcome characteristics: filter lifespan or occurrence of filter clotting, all-cause mortality, length of hospital or ICU stay, duration of CRRT, recovery of kidney function, adverse events, and costs. Quality assessment of the included RCTs was performed by two investigators (CL and FW) using the Cochrane Collaboration Risk of Bias Tool [18], and disagreements would be settled by discussion with a third investigator (PF). Selection bias (random sequence generation and allocation concealment), performance bias (blinding of participants and personnel), detection bias (blinding of outcome), attrition bias (incomplete outcome data), reporting bias (selective reporting), and other biases were categorized as “low,” “unclear,” or “high” risk of bias.

Data synthesis and analysis

RR with 95% CIs was calculated for dichotomous variables, and MD with 95% CIs was calculated for continuous variables. Conventional pairwise meta-analyses were initially conducted using Stata SE, version 14 (StataCorp, College Station, TX, USA), for direct comparisons between different pharmacological anticoagulants during CRRT. P < 0.05 was considered statistically significant. Homogeneity, transitivity, and consistency were assumed to underlie the validity of conclusions from NMA analyses [19]. Homogeneity assumption was satisfied when the magnitude of heterogeneity within direct pairwise comparisons was acceptable. Heterogeneity was evaluated using the I2 statistic. Specifically, I2 values of 0–24.9%, 25–49.9%, 50–74.9%, and 75–100% represented no, low, moderate, and significant heterogeneity, respectively [20, 21]. And the transitivity assumption was satisfied when studies included were sufficiently similar regarding methodological and clinical characteristics. Inconsistency between direct and indirect estimates in the entire network of each outcome was assessed locally with a loop-specific approach and globally with design-by-treatment interaction model [22, 23]. And the consistency assumption was rejected when the p value of the inconsistency test was less than 0.05. In addition, Begg’s funnel plots and Egger’s test were also performed to analyze potential publication bias when ten or more trials were included in each pairwise meta-analysis. We also identified the relative rankings of different anticoagulants based on the surface under the cumulative ranking curve (SUCRA) for each outcome, ranging from 0 to 100% [24]. And the higher the value, the more effective of the anticoagulation during CRRT.

The NMA was also performed for multiple comparisons using Stata SE, version 14 (StataCorp, College Station, TX, USA). And given the risk of excessively high or low blood flow rates on the robustness of the NMA results, we performed a sensitivity analysis of primary outcome by exclusion of studies with a blood flow rate > 200 ml/min or < 130 ml/min, as well as studies lacking detailed information on blood flow rate. In addition, in order to eliminate the potential influence of other major non-pharmacological interventions (modes of CRRT, and pre- or post-dilution) [25], two subgroup analyses were performed for populations with CVVH, CVVHD, or CVVHDF and pre-dilution or post-dilution, respectively.

Results

Search results and characteristics of the included studies

A flowchart depicting the selection process for this study is presented in Fig. 1. A total of 2854 articles were identified from our initial search, and after removing 868 duplicates, 1986 articles were assessed by screening the title and abstract of each article. Subsequently, 89 articles were read for full text for further evaluation, and 51 articles were excluded for the following reasons: irrelevant data (n = 38), overlapping data (n = 4), study protocol (n = 3), and conference abstracts (n = 6). Finally, 37 RCTs from 38 articles were included in our NMA. The eligible studies were conducted from 1993 to 2022 with a total number of 2648 adult patients. These trials were conducted in 14 different countries, with Australia contributing the most (8 trials, 21.6%). And among the included studies, 6 were conducted in multicenter ICUs, and the other 31 RCTs were conducted in a single-center ICU. The NMA evaluated 14 different anticoagulation options, including UFH, RCA, LMWH, NM, PGI2, and so on. Across the analyzed studies, the sample sizes ranged from 10 to 596, and the mean (or median) age of patients ranged from 45 to 78 years. Most studies reported two-arm comparisons, except for 5 studies [26,27,28,29,30] with three-arm comparisons. A range of outcomes were recorded in these studies, including filter lifespan or occurrence of filter clotting, all-cause mortality, length of hospital or ICU stay, duration of CRRT, recovery of kidney function, adverse events, and costs. The baseline characteristics of the included RCTs are presented in Table 1.

Fig. 1
figure 1

Flowchart of study selection

Full size image
Table 1 Characteristics of included randomized controlled trials in the network meta-analysis
Full size table

Risk of bias

For the evaluation of the quality of the included studies, we utilized the Kappa coefficient to test the consistency, and we showed a substantial coefficient with a Kappa value of 0.90. Discrepancies were resolved by a third author through consensus. The details for the risk of bias of individual RCTs are presented in Additional File 2. Overall, most RCTs had a high risk of bias in blinding of participants and personnel, and blinding of outcome due to not being blinded or being conducted as an open-label study. However, blinding of patients and clinicians was clinically impracticable in studies due to virtual practice issues. Specifically, the lowest risk was random sequence generation, with exceeding 70% of studies considered to be at low risk for bias. More than half of the trials were judged as having a low risk of bias in allocation concealment, attrition, and reporting. Additionally, 9 trials had a high risk of other bias due to the funding source or lack of conflict of interest.

Assessment of heterogeneity, transitivity, and inconsistency

The results of heterogeneity in direct pairwise comparisons are presented in Additional File 3. Overall, no statistically significant heterogeneity was observed in most direct pairwise comparisons, except for some direct comparison pairs for the reduction of creatinine (Cr) and blood urea nitrogen (BUN) (Additional File 3). Publication bias was not evaluated due to the lack of enough trials included in each pair. We have evaluated the inclusion and exclusion criteria of the eligible studies, particularly with respect to adult critically ill patients with AKI (Additional File 4). Moreover, the target anticoagulation levels were comparable among studies. UFH and RCA are the two most commonly used anticoagulants. In the UFH group, the heparin infusion was adjusted to maintain the activated partial thromboplastin time (APTT) values below 70 s, or the post-filter activated clotting time (ACT) between 180 and 210 s in most studies [13, 26, 31,32,33,34,35,36,37,38,39,40,41,42]. In the RCA group, the post-filter ionized calcium concentration in most studies was kept at 0.25–0.35 mmol/L [32, 37, 43,44,45]. And the targeted systemic ionized calcium levels in most studies were kept at 0.9–1.2 mmol/L or 1.0–1.35 mmol/L [41, 44,45,46,47]. Thus, we consider the assumption of transitivity to be valid. In addition, there was no evidence of significant inconsistency in the network, either at the global or local levels (Additional File 5).

Filter lifespan and filter clotting

Twenty-eight RCTs involving 2342 patients reported filter lifespan. A total of 12 anticoagulation options were included. The network geometry is shown in Fig. 2. No statistically significant heterogeneity was observed among the included RCTs within most direct pairwise comparisons except for three comparison pairs. Comparison pairs of LMWH versus UFH (I2 = 66.7%), RCA versus Regional-UFH plus protamine (I2 = 56.2%), and RCA versus no anticoagulation (I2 = 55.4%) showed moderate heterogeneity, but were not statistically significant (p > 0.10) (Additional File 3). The inconsistency test at the global and local levels also indicated no significant inconsistency (Additional File 5). Anticoagulation options ranking based on SUCRA values, which are shown in Fig. 3, from best to worst, were as follows: Regional-UFH + PGI2 98.7%, RCA + LMWH 85.5%, RCA 67.2%, NM 59.7%, Bivalirudin 58.8%, LMWH 51.9%, UFH + PGI2 44.4%, No 35.5%, PGI2 30.4%, UFH 23.1%, Regional-UFH 22.8%, and Hirudin 21.9%. The forest plot for direct comparison is shown in Fig. 4. Regional-UFH + PGI2 had the highest SCURA value, and Regional-UFH + PGI2 seemed to outperform RCA (MD 37.0, 95% CI 12.0 to 62.0), LMWH (MD 41.3, 95% CI 15.6 to 67.0), NM (MD 37.7, 95% CI 6.8 to 68.6), and other evaluated anticoagulation options in prolonging filter lifespan (Additional File 6: Fig. S1). However, only a single included RCT with 46 participants had evaluated Regional-UFH + PGI2. Therefore, additional high-quality studies are necessary before any recommendation. UFH and RCA are most frequently used to assess the filter lifespan, and there is strong evidence that RCA increases filter lifespan compared with UFH (MD 12.0, 95% CI 3.8 to 20.2) (Additional File 6: Fig. S1). Sensitivity analysis was also performed by excluding 6 trials with a blood flow rate > 200 ml/min or < 130 mL/min, or without details of blood flow rate. And compared to UFH (MD 18.0, 95% CI 0.0 to 36.0), RCA also showed superiority in prolonging filter lifespan (Additional File 6, Fig. S43).

Fig. 2
figure 2

Network geometry of all the included anticoagulation options for evaluating filter lifespan. The size of the nodes and the thickness of the lines were proportional to the sample size and number of direct comparisons, respectively. Lines do not connect nodes when there were no head-to-head trials between two treatments. Abbreviations: UFH: unfractionated heparin; RCA: regional citrate anticoagulation; LMWH: low-molecular-weight heparin; PGI2: prostaglandin I2; NM: nafamostat mesilate; and No: no anticoagulation

Full size image
Fig. 3
figure 3

Filter lifespan ranking among different anticoagulation options

Full size image
Fig. 4
figure 4

Forest plot in direct comparisons for evaluation of filter lifespan

Full size image

Filter clotting is the most common cause of terminating hemofiltration, and 12 anticoagulation options were evaluated to decrease the rate of filter clotting (Additional File 6: Fig. S2). No statistically significant heterogeneity (P > 0.1) was observed, although comparison pairs of RCA versus UFH (I2 = 64.1%) showed moderate heterogeneity (Additional File 3). The SUCRA analysis and the forest plot are also limited in this analysis of Regional-UFH + PGI2 as it was carried out on data from a single trial with a small sample size. RCA (64.3%) ranked higher than LMWH (60.5%), UFH (34.3%), Regional-UFH (62.9%), NM (58.4%), and PGI2 (47.6%) (Additional File 6: Fig. S3). However, the forest plot for all comparisons indicated that RCA only had a lower incidence of filter clotting than UFH (Additional File 6: Fig. S5).

All-cause mortality

A total of 24 RCTs with 1947 patients, which compared to 11 anticoagulation options for CRRT, were included in this NMA (Additional File 6: Fig. S6). There was no significant heterogeneity, or inconsistency among the included trials reporting all-cause mortality (Additional File 3; Additional File 5). RCA and UFH were the most frequently used anticoagulants to assess all-cause mortality; however, no significant difference was observed between the two anticoagulants in reducing all-cause mortality (RR 0.99, 95% CI 0.84 to 1.16) (Additional File 6: Fig. S9). Although based on SUCRA values, Bivalirudin as anticoagulation for CRRT had the highest SCURA value (76.0%) (Additional File 6: Fig. S7), its 95% CI was wide and contained the null effect when compared with the other 10 anticoagulation options (Additional File 6: Fig. S9). No significant difference was observed among the 11 anticoagulation options based on the forest plot for all comparisons. Therefore, the anticoagulation options ranking for all-cause mortality should be interpreted with caution.

Length of hospital or ICU stay

Only a small subset of trials provided information on the length of hospital stay, with 4 RCTs [43, 44, 48, 49] involving 3 anticoagulation options. Therefore, NMA for this outcome was not performed. A total of 10 trials with 1123 patients reported data on length of ICU stay, and 7 anticoagulation options were assessed (Additional File 6: Fig. S10). Bivalirudin ranked the best according to SUCRA statistic (95.5%), followed by UFH (58.6%), Hirudin (47.6%), Regional-UFH (40.3%), RCA (38.0%), UFH + PGI2 (35.0%), and no anticoagulation (35.0%) (Additional File 6: Fig. S11). However, only a single study with a small sample size had evaluated Bivalirudin, and potential overinterpretation of the statistical testing and SUCRA ranking probably exist. Further high-quality RCTs are desirable to strengthen the effectiveness of Bivalirudin during CRRT. Moreover, no significant difference was observed among anticoagulation with UFH, Regional-UFH, RCA, UFH + PGI2, and Hirudin (Additional File 6: Fig. S13).

Duration of CRRT

A total of 9 RCTs comparing 7 anticoagulation options reported data on duration of CRRT (Additional File 6: Fig. S14). The direct comparison between RCA and UFH was most frequently used to assess duration of CRRT. The anticoagulation options for CRRT ranking, based on SUCRA values, from largest to smallest, were as follows: Regional-UFH 73.0%, RCA 61.7%, LMWH 59.4%, Bivalirudin 48.8%, no anticoagulation 47.1%, UFH 37.9%, and NM 22.0% (Additional File 6: Fig. S15). However, according to the forest plot for all comparisons, Regional-UFH did not show an advantage in reducing the duration of CRRT compared to the other 6 anticoagulation options (Additional File 6: Fig. S17). Thus, the anticoagulation option ranking should be interpreted cautiously.

Recovery of kidney function

A total of 10 trials with 1248 patients, which were compared to 8 anticoagulation options, reported recovery of kidney function (Additional File 6: Fig. S18). No heterogeneity was found among the included studies, and the consistent assumption was deemed acceptable. There is weak evidence suggesting that NM is beneficial to the recovery of kidney function compared to the other 7 anticoagulation options during CRRT. Although NM had the highest SCURA value (86.9%), its 95% CI was wide and contained the null effect when compared to other anticoagulation options (Additional File 6: Fig. S21). The reduction of Cr and BUN was also analyzed to reflect the recovery status of kidney function. With no difference in baseline characteristics between the control and intervention groups, we collected data on the levels of Cr and BUN after treatment to reflect the changes. Ten RCTs involving 7 anticoagulation options reported data on the reduction of Cr (Additional File 6: Fig. S22). Substantial heterogeneity was observed across studies within the comparison of PGI2 and UFH (I2 = 88.2%, P < 0.01), or RCA and no anticoagulation (I2 = 94.7%, P < 0.01). And according to the forest plot for all comparisons, no significant difference was observed among these 7 anticoagulation options (Additional File 6: Fig. S25). As for the analysis of the reduction of BUN, a significant heterogeneity among the included trials was also observed (Additional File 3), and none of the anticoagulation option showed an advantage over each other (Additional File 6: Fig. S29). Therefore, further high-quality RCTs are necessary to strengthen the evidence.

Adverse events

A total of 13 RCTs comparing to 8 anticoagulation options contributed to the analysis of adverse events. However, two RCTs [50, 51] comparing NM and no anticoagulation were excluded from the analysis due to a lack of direct comparison with the other 6 anticoagulation options. Thus, only 11 RCTs comparing to 6 anticoagulation options were included in the NMA of adverse events (Additional File 6: Fig. S30). As for the direct comparison of NM and no anticoagulation, no statistically significant difference was observed (RR 1.07, 95% CI 0.75 to 1.51, P = 0.72). Unfractionated heparin plus Prostaglandin E1 (UFH + PGE1) ranked the best among the 6 anticoagulation options (Additional File 6: Fig. S31); however, there is weak evidence that it reduces the occurrence of adverse events (Additional File 6: Fig. S33). Compared to UFH, RCA had a higher SCURA value, and a lower occurrence of adverse events was observed according to the forest plot for all comparisons (Additional File 6: Fig. S33). We also analyzed bleeding events reported from 30 RCTs compared to 14 anticoagulation options (Additional File 6: Fig. S34). Although RCA + LMWH ranked highest for lower occurrence of bleeding events, its 95% CI was wide and contained the null effect when compared with other anticoagulation options. However, there is strong evidence that UFH (RR 3.09, 95% CI 2.08 to 4.60) and LMWH increase the risk of bleeding events compared with RCA (Additional File 6: Fig. S37). Metabolic disturbance was also analyzed in this NMA, with only 8 trials comparing 5 anticoagulation options (Additional File 6: Fig. S38). No statistically significant difference was observed between the 5 anticoagulation options (Additional File 6: Fig. S41). In summary, the most effective anticoagulation option for reducing the occurrence of adverse events should be interpreted cautiously. Anticoagulation with RCA during CRRT is superior to UFH in reducing adverse events, including bleeding events.

Costs

Cost-effectiveness is a critical consideration when selecting anticoagulants for CRRT in critically ill patients with AKI. However, only four RCTs [30, 41, 49, 52] have reported the data on the costs of RCA. Gao et al. [49] noted that the cost of CRRT per person was slightly lower in the RCA group than that in the no anticoagulation group. And Schilder et al. [41] indicated that the costs of the first 72 h of CRRT were lower in the RCA group, compared to the UFH group. Despite sodium citrate costing more per day than heparin, RCA was found to prolong filter lifespan, which resulted in an economic benefit. However, Fealy et al. [52] suggested that the magnitude of the gain in circuit life did not offset the additional cost of citrate sufficiently.

In addition, two studies have compared the total daily costs of LMWH and UFH. Joannidis et al. [36] observed a lower total daily cost of CRRT during anticoagulation with LMWH, which contradicts the finding in the study conducted by Reeves et al. [53].

Subgroup analyses

Two subgroup analyses were performed for populations with CVVH, CVVHD, or CVVHDF and pre-dilution or post-dilution, respectively. Overall, in the CVVH subgroup, no significant difference was observed between the evaluated anticoagulants, except for Regional-UFH + PGI2. However, only a single included RCT with 46 participants had evaluated Regional-UFH + PGI2, which should be interpreted with caution (Additional File 6: Fig. S44). In the CVVHDF subgroup, there was evidence that RCA is superior in prolonging filter lifespan compared with UFH. However, no significant difference was observed among other evaluated anticoagulants (Additional File 6: Fig. S45). Due to the lack of sufficient data, the network meta-model of CVVHD is not available. In the pre-dilution and post-dilution subgroups, compared to UFH, RCA and LMWH both showed superiority in prolonging filter lifespan (Additional File 6: Figs. S46, S47).

Discussion

Key findings

We performed an NMA of 37 RCTs comprising 2648 participants to evaluate the efficacy and safety of 14 different anticoagulation options during CRRT in adult critically ill patients. UFH and RCA are most frequently used to assess the outcomes. Our findings suggested that RCA has advantages not only in prolonging filter lifespan, but also in reducing bleeding events when compared to UFH. The SUCRA analysis and forest plot of Regional-UFH + PGI2 are limited as only a single study was included. Additional high-quality studies are necessary before any recommendation of Regional-UFH + PGI2. Moreover, there were similar effects in terms of length of ICU stay, all-cause mortality, duration of CRRT, recovery of kidney function, and adverse events among most evaluated anticoagulation options.

Comparison with previous studies

Several systematic reviews and meta-analyses have compared the effects of RCA and other anticoagulants such as UFH during CRRT among critically ill patients. Six pairwise meta-analyses have compared the efficacy and safety of RCA and UFH in CRRT for critically ill patients [54,55,56,57,58,59]. A systematic review including 3 RCTs and 17 observational studies assessed the efficacy and safety of RCA, UFH, and NM compared to anticoagulation-free, respectively [60]. Another systematic review evaluated the effects of RCA versus UFH, Regional-UFH, and LMWH among critically ill patients treated with CRRT [61]. Our study comprehensively evaluated the effects of various anticoagulation options altogether, considering filter lifespan, all-cause mortality, length of stay, duration of CRRT, recovery of kidney function, and adverse events as outcomes. Unlike previous meta-analysis focused on comparisons of two anticoagulation options, our study was able to construct various anticoagulation options into a network structure, comprehensively evaluating the effectiveness of anticoagulants from both direct and indirect comparison. Comparison of our study with previous relevant studies is presented in Table 2. Such a comprehensive NMA would be more likely to accurately evaluate the efficacy and safety of different anticoagulation options during CRRT.

Table 2 Comparison of our study with previous relevant studies
Full size table

Clinical implications and future studies

Although filter lifespan is influenced by various factors, such as the position and patency of the vascular access, catheter choice, and blood flow rate, the choice of anticoagulant is also an essential strategy to prolong filter lifespan during CRRT [62]. This NMA suggested that Regional-UFH + PGI2 seemed to outperform other evaluated anticoagulants in prolonging filter lifespan and decreasing the rate of filter clotting. However, only one RCT with 46 participants evaluated Regional-UFH + PGI2, and there was potential for attrition and reporting bias in this trial. Therefore, the SUCRA analysis and the forest plot of Regional-UFH + PGI2 have limitations. And additional high-quality studies are necessary before making any recommendation.

Our NMA revealed that RCA is more effective in prolonging filter lifespan than UFH during CRRT. Since the requirement for intravascular access and artificial circuits during CRRT may increase the risk of infection, the prolonged filter lifespan could be associated with an increased rate of infection. However, only a post hoc analysis of the RICH trial [63] had analyzed the relationship between prolonged filter lifespan and the risk of infection. Further investigation into the potential paradoxical and negative impact of prolonged filter lifespan and the risk of infection is needed to provide evidence for clinical practice. Another concern is the cost-effectiveness of RCA. Currently, there are still no firm conclusions about whether RCA has superior cost-effectiveness. Only four RCTs evaluated cost-effectiveness differences between RCA and other anticoagulation options. Future studies should therefore pay more attention to cost-effectiveness.

It should be noted that variations in RCA protocols do have an effect on filter outcomes. But unfortunately, none of the RCTs included in this NMA provided sufficient data to analyze the effects of variations in RCA protocols. Currently, a variety of RCA protocols involving mode of citrate delivery and different citrate-containing solutions have been described [64]. Our previous study indicated a simplified RCA-based CVVH protocol using calcium-containing replacement solution had a similar circuit lifespan compared to calcium-free replacement solution. However, we can continuously supplement calcium without the need for a separate intravenous catheter and the preparation of a large dose of intravenous calcium solution, making it more convenient to apply in RCA-CRRT practice [65]. Poh et al. [66] found that a low-dose RCA protocol (an initial citrate dose of 2.5 mmol/L instead of 3 mmol/L) had fewer citrate-related complications without loss of efficacy. However, there remain several reasons limiting the widespread use of RCA, including the scarcity of CRRT-specific citrate solutions and intravenous calcium, complexity of RCA protocols, and the increasing workload [64, 67]. Therefore, the development of better and simplified RCA protocols is crucial.

In our NMA, all-cause mortality was not significantly different among different anticoagulation options. Several meta-analyses had also indicated that anticoagulation with RCA or UFH during CRRT did not impact all-cause mortality [6, 57]. However, minimizing mortality is the ultimate goal of develop better adjuvant therapy. Therefore, further high-quality prospective controlled studies are urgently needed to investigate new anticoagulation options that may decrease mortality.

Our NMA did not find evidence supporting the superiority of any anticoagulant over another on recovery of kidney function. The reduction of Cr and BUN is a reflection of the recovery status of kidney function, and the recovery of kidney function ultimately benefits patient survival [68]. Thus, future studies should focus on the influence of recovery of kidney function. Adverse events including bleeding events and metabolic disturbance were reported in our study. Our NMA showed that the occurrence of adverse events was not significantly different among most anticoagulation options during CRRT. The most effective anticoagulation option for reducing the occurrence of adverse events remains unclear. Instead, our finding suggested the superiority of RCA in critically ill patients in reducing adverse events compared with UFH. Although critically ill patients with acute or chronic liver failure were excluded in this NMA, a meta-analysis [69] and several observational studies [70, 71] had indicated that RCA is also safe in patients with liver failure or at high risk of bleeding. Thus, further high-quality RCTs are desirable to investigate the most effective anticoagulation option for reducing adverse events when performing CRRT.

Strengths and limitations

To our knowledge, our study comprehensively evaluated the efficacy and safety of various anticoagulation options altogether for critically ill patients with AKI. Our NMA allows the comparison of multiple anticoagulation options simultaneously in a single analysis and improves the precision by combining direct and indirect estimates. The findings of our NMA would add evidence to the future choice of anticoagulant during CRRT for critically ill patients. Our search strategy was extensive and included all the relevant studies with no publication year restrictions. It included data from more than 2500 patients, 37 RCTs from 38 articles, and multiple countries, including Australia, Italy, Germany, France, China, and so on. To ensure homogeneity and improve transitivity, we set strict inclusion criteria that only adult critically ill patients with AKI could be included. And the target anticoagulation levels were comparable among included studies. A variety of outcomes such as filter lifespan, all-cause mortality, length of stay, during of CRRT, recovery of kidney function, and adverse events were analyzed in our NMA. In addition, we have performed sensitivity analysis to eliminate the potential influence of blood flow rate on the results. And in order to eliminate the potential influence of other major non-pharmacological interventions on filter lifespan, we conducted two subgroup analyses for populations with CVVH, CVVHD, or CVVHDF and pre-dilution or post-dilution, respectively. Furthermore, two independent investigators have thoroughly evaluated the methodological quality of this study.

Despite these strengths, this NMA still has some limitations. Firstly, only a single included RCT with a small sample size had evaluated Regional-UFH + PGI2, and there was potential for attrition and reporting bias in this trial. The SUCRA analysis and the forest plot of Regional-UFH + PGI2 will be limited. Secondly, due to the nature of the intervention and clinical problem, most included RCTs were not double-blinded or conducted as an open-label study. Although it might not influence the outcomes, there is still potential for bias. Thirdly, insufficient data were available to evaluate the efficacy and safety of different anticoagulation options on length of hospital stay, which is also an important outcome to strengthen the evidence on the best choice of anticoagulation options during CRRT. Fourthly, moderate or significant heterogeneity existed in some outcomes such as the reduction of Cr and BUN. This was a potential factor that might influenced the robustness of these results. The results of these outcomes should therefore be interpreted cautiously. Last, but not least, only published trials with selective databases were included in this NMA, and reporting bias could not be ruled out because not all trials reported filter lifespan or filter clotting, especially when filter lifespan was not the primary outcome. Regardless of these limitations, we minimized bias throughout the analysis by strict method identification, data selection, statistical analysis, sensitivity analysis, and subgroup analyses. These steps would strengthen the stability and accuracy of this NMA.

Conclusions

Between the RCA and UFH groups, RCA is the priority anticoagulant in prolonging filter lifespan and reducing the risk of bleeding. Regional-UFH + PGI2 and Bivalirudin were evaluated by a single study each. Thus, additional high-quality studies are necessary before any recommendation of Regional-UFH + PGI2 and Bivalirudin. No statistically significant difference was observed in all-cause mortality, duration of CRRT, recovery of kidney function, and adverse events among most evaluated anticoagulation options.

Availability of data and materials

All data generated or analyzed during this study are included in this article and its additional information files.

Abbreviations

CRRT:

Continuous renal replacement therapy

AKI:

Acute kidney injury

RCTs:

Randomized controlled trials

NMA:

Network meta-analysis

NM:

Nafamostat mesilate

ICU:

Intensive care unit

UFH:

Unfractionated heparin

RCA:

Regional citrate anticoagulation

LMWH:

Low-molecular-weight heparin

PGI2:

Prostaglandin I2

No:

No anticoagulation

RR:

Risk ratio

MD:

Mean difference

SUCRA:

Surface under the cumulative ranking curve

Cr:

Creatinine

BUN:

Blood urea nitrogen

References

  1. Legrand M, Tolwani A. Anticoagulation strategies in continuous renal replacement therapy. Semin Dial. 2021;34:416–22. https://doi.org/10.1111/sdi.12959.

    Article  PubMed  Google Scholar 

  2. See E, Ronco C, Bellomo R. The future of continuous renal replacement therapy. Semin Dial. 2021;34:576–85. https://doi.org/10.1111/sdi.12961.

    Article  PubMed  Google Scholar 

  3. Uchino S, Fealy N, Baldwin I, Morimatsu H, Bellomo R. Continuous is not continuous: the incidence and impact of circuit “down-time” on uraemic control during continuous veno-venous haemofiltration. Intensive Care Med. 2003;29:575–8. https://doi.org/10.1007/s00134-003-1672-8.

    Article  PubMed  Google Scholar 

  4. Joannidis M. Oudemans-van Straaten HM Clinical review: patency of the circuit in continuous renal replacement therapy. Crit Care. 2007;11:218. https://doi.org/10.1186/cc5937.

    Article  PubMed  PubMed Central  Google Scholar 

  5. Boyle M, Baldwin I. Understanding the continuous renal replacement therapy circuit for acute renal failure support: a quality issue in the intensive care unit. AACN Adv Crit Care. 2010;21:367–75. https://doi.org/10.1097/NCI.0b013e3181f95673.

    Article  PubMed  Google Scholar 

  6. Tsujimoto H, Tsujimoto Y, Nakata Y, Fujii T, Takahashi S, Akazawa M, et al. Pharmacological interventions for preventing clotting of extracorporeal circuits during continuous renal replacement therapy. Cochrane Database Syst Rev. 2020;12:Cd012467. https://doi.org/10.1002/14651858.CD012467.pub3.

    Article  PubMed  Google Scholar 

  7. Dirkes S, Wonnacott R. Continuous renal replacement therapy and anticoagulation: what are the options? Crit Care Nurse. 2016;36:34–41. https://doi.org/10.4037/ccn2016623.

    Article  PubMed  Google Scholar 

  8. Oudemans-van Straaten HM, Kellum JA, Bellomo R. Clinical review: anticoagulation for continuous renal replacement therapy–heparin or citrate? Crit Care. 2011;15:202. https://doi.org/10.1186/cc9358.

    Article  PubMed  PubMed Central  Google Scholar 

  9. Davies H, Leslie G. Anticoagulation in CRRT: agents and strategies in Australian ICUs. Aust Crit Care. 2007;20:15–26. https://doi.org/10.1016/j.aucc.2006.11.001.

    Article  PubMed  Google Scholar 

  10. Uchino S, Bellomo R, Morimatsu H, Morgera S, Schetz M, Tan I, et al. Continuous renal replacement therapy: a worldwide practice survey. The beginning and ending supportive therapy for the kidney (B.E.S.T. kidney) investigators. Intensive Care Med. 2007;33:1563–70. https://doi.org/10.1007/s00134-007-0754-4.

    Article  PubMed  Google Scholar 

  11. Lameire N, Kellum JA. Contrast-induced acute kidney injury and renal support for acute kidney injury: a KDIGO summary (Part 2). Crit Care. 2013;17:205. https://doi.org/10.1186/cc11455.

    Article  PubMed  PubMed Central  Google Scholar 

  12. Lanckohr C, Hahnenkamp K, Boschin M. Continuous renal replacement therapy with regional citrate anticoagulation: do we really know the details? Curr Opin Anaesthesiol. 2013;26:428–37. https://doi.org/10.1097/ACO.0b013e3283620224.

    Article  CAS  PubMed  Google Scholar 

  13. Garcés EO, Victorino JA, Thomé FS, Röhsig LM, Dornelles E, Louzada M, et al. Enoxaparin versus unfractioned heparin as anticoagulant for continuous venovenous hemodialysis: a randomized open-label trial. Ren Fail. 2010;32:320–7. https://doi.org/10.3109/08860221003606281.

    Article  CAS  PubMed  Google Scholar 

  14. Hwang SD, Hyun YK, Moon SJ, Lee SC, Yoon SY. Nafamostat mesilate for anticoagulation in continuous renal replacement therapy. Int J Artif Organs. 2013;36:208–16. https://doi.org/10.5301/ijao.5000191.

    Article  CAS  PubMed  Google Scholar 

  15. Gainza FJ, Quintanilla N, Pijoan JI, Delgado S, Urbizu JM, Lampreabe I. Role of prostacyclin (epoprostenol) as anticoagulant in continuous renal replacement therapies: efficacy, security and cost analysis. J Nephrol. 2006;19:648–55.

    CAS  PubMed  Google Scholar 

  16. Biancofiore G, Esposito M, Bindi L, Stefanini A, Bisà M, Boldrini A, et al. Regional filter heparinization for continuous veno-venous hemofiltration in liver transplant recipients. Minerva Anestesiol. 2003;69(527–534):534–528.

    Google Scholar 

  17. Hutton B, Salanti G, Caldwell DM, Chaimani A, Schmid CH, Cameron C, et al. The PRISMA extension statement for reporting of systematic reviews incorporating network meta-analyses of health care interventions: checklist and explanations. Ann Intern Med. 2015;162:777–84. https://doi.org/10.7326/m14-2385.

    Article  PubMed  Google Scholar 

  18. Higgins JP, Altman DG, Gøtzsche PC, Jüni P, Moher D, Oxman AD, et al. The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials. Bmj. 2011;343:d5928. https://doi.org/10.1136/bmj.d5928.

    Article  PubMed  PubMed Central  Google Scholar 

  19. Donegan S, Williamson P, D’Alessandro U, Tudur SC. Assessing key assumptions of network meta-analysis: a review of methods. Res Synth Methods. 2013;4:291–323. https://doi.org/10.1002/jrsm.1085.

    Article  PubMed  Google Scholar 

  20. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med. 2002;21:1539–58. https://doi.org/10.1002/sim.1186.

    Article  PubMed  Google Scholar 

  21. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. Bmj. 2003;327:557–60. https://doi.org/10.1136/bmj.327.7414.557.

    Article  PubMed  PubMed Central  Google Scholar 

  22. Veroniki AA, Vasiliadis HS, Higgins JP, Salanti G. Evaluation of inconsistency in networks of interventions. Int J Epidemiol. 2013;42:332–45. https://doi.org/10.1093/ije/dys222.

    Article  PubMed  Google Scholar 

  23. Higgins JP, Jackson D, Barrett JK, Lu G, Ades AE, White IR. Consistency and inconsistency in network meta-analysis: concepts and models for multi-arm studies. Res Synth Methods. 2012;3:98–110. https://doi.org/10.1002/jrsm.1044.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  24. Salanti G, Ades AE, Ioannidis JP. Graphical methods and numerical summaries for presenting results from multiple-treatment meta-analysis: an overview and tutorial. J Clin Epidemiol. 2011;64:163–71. https://doi.org/10.1016/j.jclinepi.2010.03.016.

    Article  PubMed  Google Scholar 

  25. Tsujimoto Y, Fujii T. How to prolong filter life during continuous renal replacement therapy? Crit Care. 2022;26:62. https://doi.org/10.1186/s13054-022-03910-8.

    Article  PubMed  PubMed Central  Google Scholar 

  26. Bellomo R, Teede H, Boyce N. Anticoagulant regimens in acute continuous hemodiafiltration: a comparative study. Intensive Care Med. 1993;19:329–32. https://doi.org/10.1007/bf01694706.

    Article  CAS  PubMed  Google Scholar 

  27. Kozek-Langenecker SA, Spiss CK, Gamsjäger T, Domenig C, Zimpfer M. Anticoagulation with prostaglandins and unfractionated heparin during continuous venovenous haemofiltration: a randomized controlled trial. Wien Klin Wochenschr. 2002;114:96–101.

    CAS  PubMed  Google Scholar 

  28. Langenecker SA, Felfernig M, Werba A, Mueller CM, Chiari A, Zimpfer M. Anticoagulation with prostacyclin and heparin during continuous venovenous hemofiltration. Crit Care Med. 1994;22:1774–81.

    Article  CAS  PubMed  Google Scholar 

  29. Schilder L, Nurmohamed SA, ter Wee PM, Paauw NJ, Girbes AR, Beishuizen A, et al. Citrate confers less filter-induced complement activation and neutrophil degranulation than heparin when used for anticoagulation during continuous venovenous haemofiltration in critically ill patients. BMC Nephrol. 2014;15:19. https://doi.org/10.1186/1471-2369-15-19.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  30. Wu B, Zhang K, Xu B, Ji D, Liu Z, Gong D. Randomized controlled trial to evaluate regional citrate anticoagulation plus low-dose of dalteparin in continuous veno-venous hemofiltration. Blood Purif. 2015;39:306–12. https://doi.org/10.1159/000381662.

    Article  CAS  PubMed  Google Scholar 

  31. Arcangeli A, Rocca B, Salvatori G, Ciancia M, De Cristofaro R, Antonelli M. Heparin versus prostacyclin in continuous hemodiafiltration for acute renal failure: effects on platelet function in the systemic circulation and across the filter. Thromb Res. 2010;126:24–31. https://doi.org/10.1016/j.thromres.2010.01.048.

    Article  CAS  PubMed  Google Scholar 

  32. Betjes MG, van Oosterom D, van Agteren M, van de Wetering J. Regional citrate versus heparin anticoagulation during venovenous hemofiltration in patients at low risk for bleeding: similar hemofilter survival but significantly less bleeding. J Nephrol. 2007;20:602–8.

    CAS  PubMed  Google Scholar 

  33. Birnbaum J, Spies CD, Klotz E, Hein OV, Morgera S, Schink T, et al. Iloprost for additional anticoagulation in continuous renal replacement therapy–a pilot study. Ren Fail. 2007;29:271–7. https://doi.org/10.1080/08860220601166222.

    Article  CAS  PubMed  Google Scholar 

  34. Fabbri LP, Nucera M, Al Malyan M, Becchi C. Regional anticoagulation and antiaggregation for CVVH in critically ill patients: a prospective, randomized, controlled pilot study. Acta Anaesthesiol Scand. 2010;54:92–7. https://doi.org/10.1111/j.1399-6576.2009.02031.x.

    Article  CAS  PubMed  Google Scholar 

  35. Hein OV, von Heymann C, Diehl T, Ziemer S, Ronco C, Morgera S, et al. Intermittent hirudin versus continuous heparin for anticoagulation in continuous renal replacement therapy. Ren Fail. 2004;26:297–303. https://doi.org/10.1081/jdi-120039529.

    Article  PubMed  Google Scholar 

  36. Joannidis M, Kountchev J, Rauchenzauner M, Schusterschitz N, Ulmer H, Mayr A, et al. Enoxaparin vs. unfractionated heparin for anticoagulation during continuous veno-venous hemofiltration: a randomized controlled crossover study. Intensive Care Med. 2007;33:1571–9. https://doi.org/10.1007/s00134-007-0719-7.

    Article  CAS  PubMed  Google Scholar 

  37. Kutsogiannis DJ, Gibney RT, Stollery D, Gao J. Regional citrate versus systemic heparin anticoagulation for continuous renal replacement in critically ill patients. Kidney Int. 2005;67:2361–7. https://doi.org/10.1111/j.1523-1755.2005.00342.

    Article  CAS  PubMed  Google Scholar 

  38. Kiser TH, MacLaren R, Fish DN, Hassell KL, Teitelbaum I. Bivalirudin versus unfractionated heparin for prevention of hemofilter occlusion during continuous renal replacement therapy. Pharmacotherapy. 2010;30:1117–26. https://doi.org/10.1592/phco.30.11.1117.

    Article  CAS  PubMed  Google Scholar 

  39. Margraf A, Liu C, Küllmar M, Meersch M, Rossaint J, Zarbock A. Analysis of leukocyte recruitment in continuous veno-venous hemofiltration with regional citrate vs. systemic heparin anticoagulation. Cells. 2022. https://doi.org/10.3390/cells11111815.

    Article  PubMed  PubMed Central  Google Scholar 

  40. Monchi M, Berghmans D, Ledoux D, Canivet JL, Dubois B, Damas P. Citrate vs. heparin for anticoagulation in continuous venovenous hemofiltration: a prospective randomized study. Intensive Care Med. 2004;30:260–5. https://doi.org/10.1007/s00134-003-2047-x.

    Article  PubMed  Google Scholar 

  41. Schilder L, Nurmohamed SA, Bosch FH, Purmer IM, den Boer SS, Kleppe CG, et al. Citrate anticoagulation versus systemic heparinisation in continuous venovenous hemofiltration in critically ill patients with acute kidney injury: a multi-center randomized clinical trial. Crit Care. 2014;18:472. https://doi.org/10.1186/s13054-014-0472-6.

    Article  PubMed  PubMed Central  Google Scholar 

  42. Vargas Hein O, von Heymann C, Lipps M, Ziemer S, Ronco C, Neumayer HH, et al. Hirudin versus heparin for anticoagulation in continuous renal replacement therapy. Intensive Care Med. 2001;27:673–9. https://doi.org/10.1007/s001340100907.

    Article  CAS  PubMed  Google Scholar 

  43. Stucker F, Ponte B, Tataw J, Martin PY, Wozniak H, Pugin J, et al. Efficacy and safety of citrate-based anticoagulation compared to heparin in patients with acute kidney injury requiring continuous renal replacement therapy: a randomized controlled trial. Crit Care. 2015;19:91. https://doi.org/10.1186/s13054-015-0822-z.

    Article  PubMed  PubMed Central  Google Scholar 

  44. Trakarnvanich T, Sirivongrangson P, Trongtrakul K, Srisawat N. The effect of citrate in cardiovascular system and clot circuit in critically ill patients requiring continuous renal replacement therapy. J Artif Organs. 2022. https://doi.org/10.1007/s10047-022-01329-0.

    Article  PubMed  PubMed Central  Google Scholar 

  45. Xun K, Qiu H, Jia M, Lin L, He M, Li D, et al. Treatment effect of regional sodium citrate anticoagulation in elderly patients with high-risk bleeding receiving continuous renal replacement therapy. Clin Appl Thromb Hemost. 2021;27:10760296211050640. https://doi.org/10.1177/10760296211050640.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  46. Tiranathanagul K, Jearnsujitwimol O, Susantitaphong P, Kijkriengkraikul N, Leelahavanichkul A, Srisawat N, et al. Regional citrate anticoagulation reduces polymorphonuclear cell degranulation in critically ill patients treated with continuous venovenous hemofiltration. Ther Apher Dial. 2011;15:556–64. https://doi.org/10.1111/j.1744-9987.2011.00996.x.

    Article  CAS  PubMed  Google Scholar 

  47. Gattas DJ, Rajbhandari D, Bradford C, Buhr H, Lo S, Bellomo R. A Randomized controlled trial of regional citrate versus regional heparin anticoagulation for continuous renal replacement therapy in critically ill adults. Crit Care Med. 2015;43:1622–9. https://doi.org/10.1097/ccm.0000000000001004.

    Article  CAS  PubMed  Google Scholar 

  48. Zarbock A, Küllmar M, Kindgen-Milles D, Wempe C, Gerss J, Brandenburger T, et al. Effect of regional citrate anticoagulation vs systemic heparin anticoagulation during continuous kidney replacement therapy on dialysis filter life span and mortality among critically Ill patients with acute kidney injury: a randomized clinical trial. JAMA. 2020;324:1629–39. https://doi.org/10.1001/jama.2020.18618.

    Article  CAS  PubMed  Google Scholar 

  49. Gao J, Wang F, Wang Y, Jin D, Tang L, Pan K. A mode of CVVH with regional citrate anticoagulation compared to no anticoagulation for acute kidney injury patients at high risk of bleeding. Sci Rep. 2019;9:6607. https://doi.org/10.1038/s41598-019-42916-1.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  50. Lee YK, Lee HW, Choi KH. Kim BS Ability of nafamostat mesilate to prolong filter patency during continuous renal replacement therapy in patients at high risk of bleeding: a randomized controlled study. PLoS One. 2014;9:e108737. https://doi.org/10.1371/journal.pone.0108737.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  51. Choi JY, Kang YJ, Jang HM, Jung HY, Cho JH, Park SH, et al. Nafamostat mesilate as an anticoagulant during continuous renal replacement therapy in patients with high bleeding risk: a randomized clinical trial. Medicine. 2015;94:e2392. https://doi.org/10.1097/md.0000000000002392.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  52. Fealy N, Baldwin I, Johnstone M, Egi M, Bellomo R. A pilot randomized controlled crossover study comparing regional heparinization to regional citrate anticoagulation for continuous venovenous hemofiltration. Int J Artif Organs. 2007;30:301–7. https://doi.org/10.1177/039139880703000404.

    Article  CAS  PubMed  Google Scholar 

  53. Reeves JH, Cumming AR, Gallagher L, O’Brien JL, Santamaria JD. A controlled trial of low-molecular-weight heparin (dalteparin) versus unfractionated heparin as anticoagulant during continuous venovenous hemodialysis with filtration. Crit Care Med. 1999;27:2224–8. https://doi.org/10.1097/00003246-199910000-00026.

    Article  CAS  PubMed  Google Scholar 

  54. Chang H, Gong Y, Li C, Ma Z. Clinical efficacy of regional citrate anticoagulation in continuous renal replacement therapy: systematic review and meta-analysis. Ann Palliat Med 2021;10:8939–8951. https://doi.org/10.21037/apm-21-1693.

  55. Wu MY, Hsu YH, Bai CH, Lin YF, Wu CH, Tam KW. Regional citrate versus heparin anticoagulation for continuous renal replacement therapy: a meta-analysis of randomized controlled trials. Am J Kidney Dis. 2012;59:810–8. https://doi.org/10.1053/j.ajkd.2011.11.030.

    Article  CAS  PubMed  Google Scholar 

  56. Bai M, Zhou M, He L, Ma F, Li Y, Yu Y, et al. Citrate versus heparin anticoagulation for continuous renal replacement therapy: an updated meta-analysis of RCTs. Intensive Care Med. 2015;41:2098–110. https://doi.org/10.1007/s00134-015-4099-0.

    Article  CAS  PubMed  Google Scholar 

  57. Liu C, Mao Z, Kang H, Hu J, Zhou F. Regional citrate versus heparin anticoagulation for continuous renal replacement therapy in critically ill patients: a meta-analysis with trial sequential analysis of randomized controlled trials. Crit Care. 2016;20:144. https://doi.org/10.1186/s13054-016-1299-0.

    Article  PubMed  PubMed Central  Google Scholar 

  58. Feng X, Deng L, Zhang Y, Chang L. Regional citrate versus heparin anticoagulation in continuous renal replacement therapy in critically ill patients: a meta-analysis. Zhonghua Wei Zhong Bing Ji Jiu Yi Xue. 2020;32:982–7. https://doi.org/10.3760/cma.j.cn121430-20200122-00068.

    Article  PubMed  Google Scholar 

  59. Li R, Gao X, Zhou T, Li Y, Wang J. Zhang P Regional citrate versus heparin anticoagulation for continuous renal replacement therapy in critically ill patients: a meta-analysis of randomized controlled trials. Ther Apher Dial. 2022;26:1086–97. https://doi.org/10.1111/1744-9987.13850.

    Article  CAS  PubMed  Google Scholar 

  60. Zhang W, Bai M, Yu Y, Chen X, Zhao L, Chen X. Continuous renal replacement therapy without anticoagulation in critically ill patients at high risk of bleeding: a systematic review and meta-analysis. Semin Dial. 2021;34:196–208. https://doi.org/10.1111/sdi.12946.

    Article  PubMed  Google Scholar 

  61. Zhang Z, Hongying N. Efficacy and safety of regional citrate anticoagulation in critically ill patients undergoing continuous renal replacement therapy. Intensive Care Med. 2012;38:20–8. https://doi.org/10.1007/s00134-011-2438-3.

    Article  CAS  PubMed  Google Scholar 

  62. Baldwin I. Factors affecting circuit patency and filter “life.” Contrib Nephrol. 2007;156:178–84. https://doi.org/10.1159/000102081.

    Article  PubMed  Google Scholar 

  63. Gerss J, Meersch M, Kindgen-Milles D, Brandenburger T, Willam C, Kellum JA, et al. The effect of filter lifespan during continuous renal replacement therapy in critically Ill patients with acute kidney injury on the rate of new-onset infection: analysis from the RICH randomized controlled trial. Am J Respir Crit Care Med. 2022;206:511–4. https://doi.org/10.1164/rccm.202201-0063LE.

    Article  PubMed  Google Scholar 

  64. Morabito S, Pistolesi V, Tritapepe L, Fiaccadori E. Regional citrate anticoagulation for RRTs in critically ill patients with AKI. Clin J Am Soc Nephrol. 2014;9:2173–88. https://doi.org/10.2215/cjn.01280214.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  65. Wei T, Tang X, Zhang L, Lin L, Li P, Wang F, et al. Calcium-containing versus calcium-free replacement solution in regional citrate anticoagulation for continuous renal replacement therapy: a randomized controlled trial. Chin Med J. 2022;135:2478–87. https://doi.org/10.1097/cm9.0000000000002369.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  66. Poh CB, Tan PC, Kam JW, Siau C, Lim NL, Yeon W, et al. Regional citrate anticoagulation for continuous renal replacement therapy—a safe and effective low-dose protocol. Nephrology. 2020;25:305–13. https://doi.org/10.1111/nep.13656.

    Article  CAS  PubMed  Google Scholar 

  67. Kirwan CJ, Hutchison R, Ghabina S, Schwarze S, Beane A, Ramsay S, et al. Implementation of a simplified regional citrate anticoagulation protocol for post-dilution continuous hemofiltration using a bicarbonate buffered. Calcium Contain Replace Solut Blood Purif. 2016;42:349–55. https://doi.org/10.1159/000452755.

    Article  CAS  Google Scholar 

  68. Schiffl H. Discontinuation of renal replacement therapy in critically ill patients with severe acute kidney injury: predictive factors of renal function recovery. Int Urol Nephrol. 2018;50:1845–51. https://doi.org/10.1007/s11255-018-1947-1.

    Article  PubMed  Google Scholar 

  69. Zhang W, Bai M, Yu Y, Li L, Zhao L, Sun S, et al. Safety and efficacy of regional citrate anticoagulation for continuous renal replacement therapy in liver failure patients: a systematic review and meta-analysis. Crit Care. 2019;23:22. https://doi.org/10.1186/s13054-019-2317-9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  70. Nurmohamed SA, Jallah BP, Vervloet MG, Yldirim G, ter Wee PM. Groeneveld AB Continuous venovenous haemofiltration with citrate-buffered replacement solution is safe and efficacious in patients with a bleeding tendency: a prospective observational study. BMC Nephrol. 2013;14:89. https://doi.org/10.1186/1471-2369-14-89.

    Article  PubMed  PubMed Central  Google Scholar 

  71. Slowinski T, Morgera S, Joannidis M, Henneberg T, Stocker R, Helset E, et al. Safety and efficacy of regional citrate anticoagulation in continuous venovenous hemodialysis in the presence of liver failure: the liver citrate anticoagulation threshold (L-CAT) observational study. Crit Care. 2015;19:349. https://doi.org/10.1186/s13054-015-1066-7.

    Article  PubMed  PubMed Central  Google Scholar 

  72. Brain MJ, Roodenburg OS, Adams N, McCracken P, Hockings L, Musgrave S, et al. Randomised trial of software algorithm driven regional citrate anticoagulation versus heparin in continuous renal replacement therapy: the filter life in renal replacement therapy pilot trial. Crit Care Resusc. 2014;16:131–7.

    PubMed  Google Scholar 

  73. Cui W, Deng X, Lv S, Hu Q, Zhang K. The observation on the comprehensive effect of extracorporeal citrate anticoagulation protocol in continuous blood purification. J Qiqihar Univ Med. 2011;32(12):1884.

    Google Scholar 

  74. Hetzel GR, Schmitz M, Wissing H, Ries W, Schott G, Heering PJ, et al. Regional citrate versus systemic heparin for anticoagulation in critically ill patients on continuous venovenous haemofiltration: a prospective randomized multicentre trial. Nephrol Dial Transplant. 2011;26:232–9. https://doi.org/10.1093/ndt/gfq575.

    Article  CAS  PubMed  Google Scholar 

  75. Link A, Girndt M, Selejan S, Rbah R, Böhm M. Tirofiban preserves platelet loss during continuous renal replacement therapy in a randomised prospective open-blinded pilot study. Crit Care. 2008;12:R111. https://doi.org/10.1186/cc6998.

    Article  PubMed  PubMed Central  Google Scholar 

  76. Oudemans-van Straaten HM, Bosman RJ, Koopmans M, van der Voort PH, Wester JP, van der Spoel JI, et al. Citrate anticoagulation for continuous venovenous hemofiltration. Crit Care Med. 2009;37:545–52. https://doi.org/10.1097/CCM.0b013e3181953c5e.

    Article  CAS  PubMed  Google Scholar 

  77. Reeves JH, Graan M. Randomised controlled trial: enoxaparin versus heparin in continuous renal replacement therapy. Blood Purif. 2003;21(2):203.

    Google Scholar 

  78. Schilder L, Nurmohamed SA, ter Wee PM, Paauw NJ, Girbes AR, Beishuizen A, et al. The plasma level and biomarker value of neutrophil gelatinase-associated lipocalin in critically ill patients with acute kidney injury are not affected by continuous venovenous hemofiltration and anticoagulation applied. Crit Care. 2014;18:R78. https://doi.org/10.1186/cc13838.

    Article  PubMed  PubMed Central  Google Scholar 

  79. van Doorn KJ, Hubloue I, Verbeelen D. Urea exchange ePicacy during unfractionated heparin anticoagulation versus low molecular weight heparin (dalteparin) anticoagulation in continuous veno-venous hemofiltration. Blood Purif. 2004;22(2):243.

    Google Scholar 

  80. van der Voort PH, Gerritsen RT, Kuiper MA, Egbers PH, Kingma WP. Boerma EC Filter run time in CVVH: pre-versus post-dilution and nadroparin versus regional heparin-protamine anticoagulation. Blood Purif. 2005;23:175–80. https://doi.org/10.1159/000083938.

    Article  CAS  PubMed  Google Scholar 

  81. Victorino JA, Otero E, Rohsig L, Dornelles E, Louzada M, Sti` J, et al. Low molecular weight heparin (LMWH) versus nonfractionated heparin (NFH) as anticoagulation system in continuous veno-veno hemodialysis (CVVH) in patients with acute renal (ARF) and respiratory failure (RF). In: American thoracic society international conference. 2007; pp. 18–23.

Download references

Acknowledgements

Not applicable.

Funding

This work was supported by the 1.3.5 Project for Disciplines of Excellence, Clinical Research Incubation Project, West China Hospital, Sichuan University (2018HXFH018), the Scientific and Technological Department of Sichuan Province, China (2020YFG0105), and the 1.3.5 Project for Disciplines of Excellence, West China Hospital, Sichuan University (ZYGD18027).

Author information

Authors and Affiliations

  1. Department of Nephrology, Kidney Research Institute, West China Hospital of Sichuan University, Chengdu, 610041, China

    Zhifeng Zhou, Chen Liu, Yingying Yang, Fang Wang, Ling Zhang & Ping Fu

Authors
  1. Zhifeng Zhou
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Chen Liu
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Yingying Yang
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Fang Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Ling Zhang
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Ping Fu
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

LZ, as the corresponding author of this paper, was mainly responsible for program design and modification. ZZ wrote the first draft. ZZ, CL, and YY assessed each study independently and recorded eligibility, quality, and outcomes. CL and FW provided arbitration in case of disagreement. FW and PF put their hands on searching for relevant studies. All of the authors have carefully examined this manuscript and agreed with the ideas presented in the article.

Corresponding author

Correspondence to Ling Zhang.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

All authors have agreed to the publication of this manuscript.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1

. File 1: Search strategy.

Additional file 2

. Details of the risk of bias of included studies.

Additional file 3

. Direct pairwise comparisons and heterogeneity of major outcomes.

Additional file 4

. Inclusion criteria and exclusion criteria of participants in the includedrandomized controlled trials.

Additional file 5

. Inconsistency assessment globally with design by treatmentinteraction model or locally with loop specific approach.

Additional file 6

. Additional Figure S1 to Figure S47.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Zhou, Z., Liu, C., Yang, Y. et al. Anticoagulation options for continuous renal replacement therapy in critically ill patients: a systematic review and network meta-analysis of randomized controlled trials. Crit Care 27, 222 (2023). https://doi.org/10.1186/s13054-023-04519-1

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s13054-023-04519-1

Keywords

Critical Care

ISSN: 1364-8535