Abstract
Background
Regional citrate anticoagulation (RCA) is gradually adopted for intermittent kidney replacement therapy (IRRT) in critically ill patients to mitigate circuit clotting. However, evidence comparing its efficacy and safety remains limited. This study aimed to (1) validate the safety and efficacy of regional citrate anticoagulation (RCA) compared to conventional anticoagulation avoidance during intermittent renal replacement therapy (IRRT) in a critical care nephrology cohort, and (2) establish practical criteria for selecting RCA protocols based on individualized patient bleeding and clotting risk assessments.
Methods
This retrospective study analyzed 141 critically ill patients requiring IRRT without systemic anticoagulation: RCA (n = 48) vs. heparin-free (n = 93). Primary outcomes included IRRT completion rates and circuit clotting events. Secondary outcomes comprised filter lifespan, net ultrafiltration (UF), solute clearance (Kt/V, URR), and adverse events. Multivariate regression identified clotting predictors.
Results
Circuit clotting caused 93.9% of premature terminations. The RCA group demonstrated significantly higher IRRT completion rates (87.5% vs. 53.8%, p < 0.001). Net UF was superior with RCA (1.9 ± 1.0 kg vs. 1.4 ± 0.9 kg; P = 0.010), while Kt/V, URR and the occurrence of hypocalcemia and metabolic acidosis remained comparable. Platelet count, traditional clotting factors (such as fibrinogen, PT, and aPTT), and thromboelastograms-derived parameters (such as R time and maximum amplitude) were comparable between subgroups. Multivariate analysis confirmed RCA as an independent protective factor against clotting (OR 0.121; P < 0.001), particularly in patients with platelet counts > 130 × 109/L and hemoglobin > 90 g/L.
Conclusions
RCA with calcium-containing dialysate significantly improves IRRT completion rates, filter longevity, and ultrafiltration efficiency without increasing metabolic risks, in a specific group of patients with platelet counts > 130 × 109/L and hemoglobin > 90 g/L, positioning RCA as a safer and more effective anticoagulation strategy for critically ill populations. Prospective trials are needed to validate these findings and to optimize RCA protocols.
Clinical trial number
Not applicable.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12882-025-04406-7.
Keywords: Acute kidney injury, Critically ill, Circuit clotting, Heparin-free, Intermittent renal replacement therapy, Regional citrate anticoagulation
Background
Regional citrate anticoagulation (RCA) has emerged as the gold-standard strategy for balancing bleeding and coagulation risks in critically ill patients requiring continuous renal replacement therapy (CRRT) [1, 2]. Traditional heparin-based anticoagulation remains suboptimal in critically ill patients. Heparins (especially unfractionated heparin and low molecular weight heparins to varying degrees) are primarily cleared renally, leading to unpredictable drug accumulation, and the coexisting coagulation disorders in AKI add logistical burden and potential for suboptimal control. As an alternative to CRRT, sustained Low-Efficiency Dialysis (SLED) offers comparable hemodynamic stability through reduced fluid removal rates while utilizing conventional dialysis equipment, which can be further optimized with RCA protocol as reported in a prospective study by Di Mario et al. [3]. While recent studies have extended its application to intermittent hemodialysis (IHD) and SLED as an alternative approach for ICU patients [3–8] optimal implementation remains controversial. Protocol heterogeneity including variable use of calcium-free versus calcium-containing dialysate, predilution versus post-dilution hemodiafiltration, and institution-specific citrate dosing limits generalizability and complicates comparative efficacy assessments [4].
A critical determinant of RCA success is precise ionized calcium (iCa2+) regulation, typically targeting pre-hemofilter levels < 0.4 mmol/L through citrate infusion rates titrated to blood flow rate (BFR) [4–16]. However, direct translation of CRRT-derived protocols to IHD is problematic: while lower BFRs may enhance anticoagulation efficacy by reduced citrate loss achieving target iCa2+ levels, they risk compromising small solute clearance due to IHD’s shorter treatment duration. Dialysate calcium concentration further modulates this balance - calcium-free regimens sustain circuit hypocalcemia for robust anticoagulation but increase systemic hypocalcemia risk, whereas calcium-containing dialysate simplifies clinical management and improves safety profiles despite potential citrate underdosing. To mitigate citrate loss across the hemofilter, some protocols incorporate post-hemofilter citrate supplementation [5]. However, this approach necessitates advanced technical expertise for real-time iCa2+ monitoring, specialized dialysate preparation, and frequent blood gas analysis for iCa2+ monitoring, substantially increasing nursing workload and healthcare costs. These challenges are also compounded by the dynamic coagulation profiles of critically ill patients: Systemic inflammation may simultaneously drive platelet activation, fibrinogen production, and endothelial dysfunction, creating a paradoxical state of hypercoagulability despite thrombocytopenia [17]. Consequently, comprehensive coagulation assessment - integrating traditional parameters (e.g., platelet count, fibrinogen) and viscoelastic testing (e.g., thromboelastography) - is essential for protocol personalization.
Traditional heparin-free protocols demonstrate unacceptably high rates of circuit thrombosis in critically ill populations with dual bleeding and hypercoagulation risks [17], however it is logistically simpler and easy to operate, as well as applicable for patients with extremely increased bleeding risk such as severe thrombocytopenia.
Therefore, in this retrospective case-control study, we directly compared a modified RCA protocol using calcium-containing dialysate against standard heparin-free regimens in critically ill patients requiring IHD or hemodiafiltration in the real-world. Beyond evaluating circuit longevity, dialysis adequacy (Kt/V, urea reduction ratio), and ultrafiltration efficiency, we stratified patients by coagulation profiles (platelet count, hemoglobin, thromboelastography) and clinical risk factors to identify subgroups most likely to benefit from RCA versus heparin-free strategies. Our dual objectives were to (1) validate RCA’s safety and efficacy relative to conventional anticoagulation avoidance, and (2) establish practical criteria for protocol selection based on individualized bleeding/clotting risk assessments.
Methods
Patients and study design
A retrospective study was conducted on patients with kidney failure who underwent IRRT without systemic anticoagulation in the CCNU at Peking University First Hospital (PKUFH) between April and December 2018. The diagnosis of acute kidney injury (AKI) followed the KDIGO clinical practice guidelines [18]. The etiology of AKI was classified as prerenal, intrinsic renal, postrenal causes based on clinical assessment. The intrinsic renal causes were further classified based on the specific site of kidney damage (e.g. renal vascular diseases, glomerular diseases, acute tubular necrosis, acute interstitial nephritis, or other renal parenchymal disorders). The diagnosis of pre-existing CKD was based on prior clinical documentation within the patient’s medical history. Indication for KRT initiation in AKI-attended patients included refractory hyperkalemia, severe metabolic acidosis, refractory volume overload / pulmonary edema, or uremic complications. In CKD patients, KRT was performed as scheduled CKD management. The decision regarding the choice of anticoagulation approaches with RCA or heparin-free protocol, and the type of IRRT (hemodialysis or hemodiafiltration) was made by the discretion of clinicians. Specifically, the selection between hemodialysis (HD) and hemodiafiltration (HDF) was followed by a goal-directed, physiology-based protocol aligned with intensive care priorities, concerning volume management, toxin spectrum, and hemodynamic stability. The primary indication for selecting RCA or a heparin-free strategy was the presence of significant bleeding risk or contraindications to systemic heparinization. This decision was driven by patient-specific factors, including active bleeding (n = 53), recent surgery or invasive procedures (n = 55), and underlying coagulation disorders (n = 33). Patients with sepsis and those undergoing isolated ultrafiltration were excluded from this study, concerning of sepsis-induced coagulopathy as a confounding factor. Prior approval was obtained from the research ethics committee of PKUFH (MR-11-23-021020). The need for informed consent was waived by the Human Ethics Committee of the Peking University First Hospital due to the retrospective nature of the study and the use of anonymized patient data, in accordance with the regulations of PKUFH that equal or exceed the standards set by the Declaration of Helsinki.
Local RCA and heparin-free protocol
The RCA protocol was locally modified in accordance with the KDIGO recommendation for CRRT. For hemodialysis, the prescription involved using a calcium-containing dialysate with a concentration of either 1.25 or 1.5 mmol/L, with potassium concentration of 2.0, 2.5 or 3.0 mmol/L aiming to maintain normokalaemia. The bicarbonate load delivered to the patient was dynamically controlled with the base excess set to -2 to -4 mmol/L online. A 4% sodium citrate solution was infused immediately downstream from the connection between the dialysis catheter and arterial line of the circuit at a fixed rate of 1.2 times the BFR. The routine measurement of ionized calcium was not necessary, and the citrate infusion was not directly diverted into the venous line of the circuit. The dialysate fluid rate was primarily set at 500 ml/min (n = 131), or 300 ml/min (n = 10) if the patients presented with clinically significant, symptomatic intradialytic hypotension episode meeting the defined blood pressure (a decrease in systolic blood pressure ≥ 20 mm Hg or a decrease in mean arterial pressure ≥ 10 mm Hg) and symptom criteria (including muscle cramping, nausea/vomiting, dizziness, generalized weakness, chest pain or discomfort). For hemodiafiltration, predilution mode was selected, and the prescribed substitution fluid rate was set 1:3 − 1:2 ratio of BFR.
The heparin-free hemodialysis or hemodiafiltration protocol included a procedure of pre-heparinization of the dialysis circuit using a preflush of heparin saline at a concentration of 3000–5000 IU/L in a closed cycle for 30 min. The circuit was then rinsed thoroughly with saline for at least 10 min. During IRRT, the dialysis circuit was not flushed.
The dialyzers employed in the study included FX600 (Polyethersulfone membrane, 1.4 m², Fresenius Medical Care), FX60 (Polyethersulfone membrane, 1.3 m², Fresenius Medical Care), FX8 (Polyethersulfone membrane, 1.8 m², Fresenius Medical Care), 150G (Polymethylmethacrylate membrane, 1.5 m², Toray Medical), 15G (Polymethylmethacrylate membrane, 1.0 m², Toray Medical).
Coagulation-related parameters
Demographic and laboratory information was collected from the electronic medical recording system. Information regarding routine systemic anticoagulants and anti-platelet agents was also documented. Coagulation testing included levels of PT/INR, aPTT, thrombin time, D-dimer, fibrinogen, and fibrinogen degradation product. Thromboelastograms (TEG) assay was conducted and yielded parameters of reaction time (R), K time (K), α-angle, maximum amplitude (MA), and Ly30 [19].
IRRT parameters and outcomes
The following variables were documented: type of vascular access, BFR, citrate infusion rate, surface area of the dialysis or hemofilter membrane, dialysate calcium concentration, hourly ultrafiltration rate, net ultrafiltration, calculated urea clearance index (Kt/V), and urea reduction ratio (URR). Instances of circuit clotting caused by hemofilter clots or clots in the venous bubble trap were also recorded. The severity of hemofilter clots was categorized into three classes: class I (mild), II (moderate), and III (severe). Data on the number of premature terminations of IRRT related to clotting or other causes were collected. Premature termination of IRRT due to circuit clotting more than 30 min ahead of schedule, was defined as the poor IRRT outcome. Other adverse events including hypocalcemia (defined as serum total Ca2+<2.11mmol/L), severe alkalosis (defined as serum bicarbonate > 30mmol/L) and metabolic acidosis (defined as serum bicarbonate < 20mmol/L) during the whole course of IRRT in the current hospital stay were recorded. Patients with concurrent systemic symptoms were considered having life-threatening hypocalcemia or acid-base disturbance. Patients were considered to have life-threatening hypocalcemia if they exhibited signs of neuromuscular irritability (e.g., carpopedal spasm, laryngospasm, bronchospasm, seizures) and/or cardiovascular instability (e.g., severe hypotension, arrhythmias). Life-threatening acid-base disturbance was defined by the presence of profound neurological impairment (e.g., lethargy, coma), Kussmaul respirations (indicating severe metabolic acidosis), or significant respiratory distress/failure attributable to the acid-base imbalance.
Statistical analysis
Data processing was performed using SPSS version 25.0. Measurements with normal distribution were expressed as mean ± standard deviation. Median (quartile) was used for data with a non-normal distribution. T-tests and nonparametric tests were used for between-group comparisons based on the data distribution of each variable. Influencing factors for clotting-related premature termination of IRRT were analyzed by multivariate logistic regression, and the odds ratio (OR) was used to represent the strength of the association. Statistical significance was set at P < 0.05.
Results
Characteristics of patients
The study cohort comprised 141 patients with 48 undergoing IRRT with RCA. The rest of 93 patients received heparin-free IRRT. The baseline clinical characteristics of these patients are detailed in Table 1. The average age of patients was 54 years, with a slight male predominance of 53.9%. Past medical history included hypertension (116 patients, 82.3%), diabetes (15 patients, 10.6%), coronary heart disease (17 patients, 12.1%), and liver disease (8.5%). Among the 68 AKI-attended patients, 22 had pre-existing CKD according to prior clinical documentation within the patient’s medical history. Anemia was a common condition among the patients, as indicated by an average hemoglobin level of 90.7 ± 19.0 g/L. The platelet count ranged from 91.0 to 209.5 × 109/L, with a median of 140.0 × 109/L. The serum albumin concentration was measured at 32.0 ± 6.5 g/L.
Table 1.
Clinical features of patients receiving RCA and heparin-free protocol for IRRT
| Variables | Total (n = 141) | RCA (n = 48) | heparin-free (n = 93) | P value |
|---|---|---|---|---|
| Age(year), M(IQR) | 54 ± 19 | 54 ± 20 | 51 ± 19 | 0.375 |
| Male gender, n (%) | 76(54) | 33(69) | 43(46) | 0.011 |
| BMI(kg/m2) | 26(18,32) | 24(23,26) | 25 ± 6 | 0.587 |
| AKI | 68(48) | 26(54) | 42(45) | 0.311 |
| CKD | 73(52) | 22(46) | 51(55) | |
| Baseline diseases Hypertension, n (%) | 116(82) | 39(81) | 77(83) | 0.820 |
| Diabetes mellitus, n (%) | 15(11) | 5(10) | 10(11) | 0.934 |
| Liver disease, n (%) | 12(9) | 5(10) | 7(8) | 0.542 |
| Coronary heart Disease, n (%) | 17(12) | 6(13) | 11(12) | 0.908 |
| Medications Warfarin, n (%) | 17(12) | 5(10) | 12(13) | 0.667 |
| Anti-platelet drugs, n (%) | 21(15) | 4(8) | 17(18) | 0.116 |
| Mean arterial pressure at dialysis start (mmHg) | 99.9 ± 13.9 | 102.7 ± 14.7 | 98.4 ± 13.4 | 0.099 |
| Laboratory data Hemoglobin (g/L) | 90.7 ± 19.0 | 89.7 ± 16.7 | 91.3 ± 20.1 | 0.646 |
| Platelet count, (x 109) | 140.0(91.0,209.5) | 142.0(96.3,219.3) | 142.2 ± 74.1 | 0.295 |
| Serum albumin (g/L) | 32.0 ± 6.5 | 31.6 ± 5.4 | 32.1 ± 7.0 | 0.658 |
| Serum creatinine(µmol/L) | 554.0 ± 234.5 | 518.3(379.5,702.2) | 542.3 ± 228.1 | 0.702 |
| Magnesium(mmol/L) | 0.9(0.8,1.0) | 0.9(0.8,1.0) | 0.9 ± 0.2 | 0.219 |
| Phosphate(mmol/L) | 1.3(1.1,1.7) | 1.3(1.1,1.7) | 1.3(1.1,1.6) | 0.787 |
| Alanine aminotransferase(IU/L) | 12.0(7.0,17.0) | 12.0(6.8,18.0) | 11.5(7.8,17.0) | 0.958 |
| Aspartate aminotransferase(IU/L) | 16.0(13.0,22.0) | 16.0(12.8,21.0) | 17.0(14.0,22.0) | 0.331 |
| γ-Glutamyl transferase(IU/L) | 25.0(17.0,54.0) | 30.5(20.8,54.5) | 23.5(15.0,52.5) | 0.075 |
| Alkaline phosphatase(IU/L) | 68.0(53.0,87.0) | 68.0(55.8,84.3) | 68.5(52.8,87.0) | 0.986 |
| Total bilirubin(µmol/L) | 10.4(7.5,13.1) | 11.2(8.6,13.3) | 10.0(7.3,12.9) | 0.328 |
| Coagulation factors | ||||
| APTT (s) | 30.4(28.4,32.3) | 29.7(28.4,31.5) | 30.5(28.3,32.6) | 0.330 |
| PT (s) | 11.2(10.4,11.8) | 11.4(10.8,12.0) | 10.9(10.3,13.1) | 0.075 |
| INR | 1.0(0.9,1.0) | 1.0(0.9,1.0) | 1.0(0.9,1.0) | 0.103 |
| Fibrinogen (g/L) | 3.4 ± 1.2 | 3.5(2.6,4.4) | 3.2(2.5,3.9) | 0.126 |
| R (min) | 6.1(4.7,6.8) | 5.7(4.8,6.6) | 6.1(4.7,7.0) | 0.391 |
| K (min) | 1.4(1.2,1.8) | 1.4(1.1,1.7) | 1.4(1.2,2.0) | 0.286 |
| α-angle(deg) | 70.2(64.7,73.9) | 70.4(65.6,74.4) | 70.0(63.4,73.3) | 0.362 |
| MA (mm) | 61.6 ± 9.7 | 62.3(55.7,70.0) | 60.4(55.0,68.0) | 0.296 |
| Ly30(%) | 0.1(0.1,0.15) | 0.1(0.1,0.3) | 0.1(0.1,0.1) | 0.157 |
| Influencing factors to clotting-related premature termination | ||||
| Vascular access | 0.245 | |||
| Arteriovenous fistula | 24(17) | 5(10) | 19(20) | |
| Temporary central femoral catheter | 17(12) | 6(13) | 11(12) | |
| Temporary central jugular catheter | 53(38) | 17(35) | 36(39) | |
| Tunneled central venous access | 47(33) | 20(42) | 27(29) | |
| Dialyzers | < 0.001 | |||
| FX600 | 38 (27) | 2 (4) | 36 (39) | |
| FX60 | 56 (40) | 10 (21) | 46 (49) | |
| FX8 | 4 (3) | 1 (2) | 3 (3) | |
| 150G | 36 (26) | 30 (63) | 6 (6) | |
| 15G | 7 (5) | 5 (10) | 2 (2) | |
| Delivered blood flow rate (ml/min) | 220 (200,230) | 200 (180,200) | 220 (200,250) | < 0.001 |
| Citrate infusion rate | NA | 216.0 ± 20.6 | NA | |
| Dialysate with calcium concentration of 1.5mmol/L | 131(92.9) | 44(93.6) | 87(94.6) | 0.795 |
| Net ultrafiltration | 1.6 ± 1.0 | 1.9 ± 1.0 | 1.4 ± 0.9 | 0.010 |
| Kt/V* | 0.7 ± 0.4 | 0.7 ± 0.4 | 0.7 ± 0.4 | 0.993 |
| URR* | 0.4 ± 0.2 | 0.4 ± 0.2 | 0.4 ± 0.2 | 0.872 |
| Adverse events, n (%) | ||||
| Metabolic acidosis | 9 (6) | 3 (6) | 6 (6) | 1.000 |
| Severe alkalosis | 2 (1) | 1 (2) | 1 (1) | 1.000 |
| Hypocalcemia | 7 (5) | 1 (2) | 6 (6) | 0.423 |
Data are n (%) or median (IQR)
NA, not available
Anti-platelet drugs included aspirin, clopidogrel, and ticagrelor. AKI: Acute kidney injury; CKD: Chronic kidney disease; eGFR: Estimated glomerular filtration rate; Alb: Albumin; APTT: Plasma activated partial prothrombin time; PT: Prothrombin time; INR: International normalized radio; R: Reaction time; K: K-time; MA: Maximal amplitude; Kt/V: urea clearance index; URR: urea reduction ratio. P < 0.05 vs. CKD group. * 34 patients had data of blood urea nitrogen available for Kt/V and URR calculation (9 in RCA group and 25 in heparin-free group)
Coagulation and IRRT parameters
Table 2 and Supplementary Table 1 present the characteristics of coagulation and IRRT-related parameters. Patients with AKI exhibited significantly elevated levels of shortened PT/INR and aPTT. However, no significant differences were observed in the traditional coagulation factors PT/INR, aPTT, fibrinogen, and D-Dimer between subgroups of the RCA approach and heparin-free protocol. The TEG derived R, K, α-angle, MA, and Ly30 demonstrated comparable results among patients treated with the two types of anticoagulation approaches.
Table 2.
Logistic regression of factors associated with clotting-related premature termination of IRRT
| Index | Univariate | Multivariate | ||
|---|---|---|---|---|
| OR (95%CI) | P value | OR (95%CI) | P value | |
| Age (year) | 1.008(0.990–1.029) | 0.355 | 1.016(0.992–1.041) | 0.211 |
| Male Gender | 1.536(0.766–3.102) | 0.227 | 1.519(0.629–3.748) | 0.356 |
| AKI | 1.582(0.788–3.221) | 0.200 | 1.185(0.5052–2.798) | 0.697 |
| eGFR(ml/min/1.73m2) | 0.962(0.904–1.013) | 0.177 | 0.966(0.905–1.020) | 0.237 |
| Hemoglobin (g/dl) | 1.445(1.178–1.807) | < 0.001 | 1.482(1.180–1.914) | 0.001 |
| Hypertension | 1.865(0.725–5.446) | 0.218 | ||
| Diabetes mellitus | 1.287(0.408–3.808) | 0.652 | ||
| Liver disease | 0.933(0.239–3.135) | 0.914 | ||
| Warfarin use | 0.758(0.229–2.188) | 0.623 | ||
| Anti-platelet drugs | 1.500(0.569–3.842) | 0.400 | ||
| R(min) | 0.992(0.810–1.201) | 0.933 | ||
| MA(mm) | 1.007(0.972–1.046) | 0.693 | ||
| K(min) | 0.938(0.640–1.254) | 0.689 | ||
| α-angle(deg) | 1.002(0.963–1.044) | 0.940 | ||
| Platelet count (109/dl) | 1.045(1.000-1.094) | 0.056 | 1.072(1.014–1.137) | 0.015 |
| APTT(s) | 1.033(0.937–1.067) | 0.919 | ||
| INR | 1.427(0.591–5.068) | 0.435 | ||
| Fibrinogen (g/L) | 1.074(0.810–1.419) | 0.616 | ||
| Serum albumin (g/L) | 1.007(0.954–1.063) | 0.796 | ||
| Choice of anticoagulation | ||||
| Heparin-free (reference) | reference | reference | ||
| RCA | 0.166(0.059–0.403) | < 0.001 | 0.121(0.037–0.332) | < 0.001 |
| Delivered blood flow rate (ml/min) | ||||
| 210–220 | reference | |||
| 180–200 | 2.469 (0.914–7.018) | 0.080 | ||
| 230–260 | 0.635 (0.238–1.741) | 0.366 | ||
| Delivered dialysate flow rate (ml/min) | 0.450 (0.066–1.890) | 0.326 | ||
hb: hemoglobin; ALB: serum albumin; APTT: plasma activated partial prothrombin time; PT: prothrombin time; R: reaction time; MA: maximal amplitude; PLT: platelet count; OR: odds ratio; CI: confidence interval; IRRT: intermittent renal replacement therapy
Features of IRRT related parameters
Table 1 presents the efficacy and safety of RCA and heparin-free protocol for IRRT. Among all patients, 83.0% had a central venous catheter as their dialysis access, including temporary central femoral catheter (12.1%), temporary central jugular catheter (37.6%), and tunneled central venous access (33.3%). For patients with RCA approach, the citrate infusion rate was set at 216.0 ± 20.6 ml/min, and most of them (91.7%) received dialysate with a calcium concentration of 1.5 mmol/L. The median delivered blood flow rate (BFR) was 200 ml/min in the RCA group, whereas the heparin-free subgroup demonstrated a significantly higher median BFR of 220 ml/min (p < 0.001). The Kt/V and URR values of the RCA subgroup were comparable to those of the heparin-free group, and between subgroups stratified by platelet count and hemoglobin (Supplementary 2), while the net ultrafiltration was significantly higher in the RCA subgroup (1.9 ± 1.0 kg) compared to the heparin-free subgroup (1.4 ± 0.9 kg, P = 0.010). During the whole course of IRRT (from 1 day prior to dialysis through 1 day post-dialysis), metabolic acidosis was seldomly seen in RCA group (n = 3) and heparin-free group (n = 6), together with severe alkalosis (one patient in each group) and hypocalcemia (one patient from RCA group and six patients from heparin-free group). No cases of life-threatening hypocalcemia or acid-base disturbances were observed.
Clotted-related IRRT outcome and influencing factors
Out of the total of 49 patients who experienced premature terminated IRRT sessions, 93.9% were attributed to circuit clotting. The completion rate of the RCA approach was 87.5%, whereas it was only 53.8% in patients with the heparin-free protocol. The instance of circuit clotting caused by clots in the venous bubble trap was 37.0%. As demonstrated in Table 3, the on-schedule completion rates of the RCA and heparin-free protocols varied among patients categorized by hemoglobin level and platelet count. In patients with platelet count > 130 × 10⁹/L and hemoglobin > 90 g/L, the completion rate of IRRT reached 100% with RCA versus 18% with heparin-free protocol (p < 0.001). Multivariate logistic regression analysis of factors associated with clotting-related premature termination of IRRT revealed that hemoglobin (OR = 1.482, P = 0.001) and platelet count (OR 1.072, P = 0.015) were associated with an increased risk of premature circuit clotting related IRRT failure. However, receiving the RCA approach was inversely correlated with poor IRRT outcome (OR = 0.121, P < 0.001). No correlation was found with traditional and TEG derived coagulation parameters (Table 2).
Table 3.
Difference in the rate of on-schedule IRRT and net ultrafiltration between RCA and the heparin-free subgroups
| PLT < 130 × 109/L | PLT ≥ 130 × 109/L | |||||||
|---|---|---|---|---|---|---|---|---|
| HB < 90 g/L | HB ≥ 90 | HB < 90 g/L | HB ≥ 90 g/L | |||||
| N = 35 | N = 30 | N = 37 | N = 39 | |||||
| CR | NUF | CR | NUF | CR | NUF | CR | NUF | |
| RCA | 90% | 2.1 ± 0.9 | 64% | 1.9 ± 0.8 | 93% | 1.7 ± 1.2 | 100% | 1.9 ± 1.1 |
| Heparin-free | 72% | 1.6 ± 0.7 | 58% | 1.8 ± 0.9 | 73% | 1.5 ± 1.1 | 18% | 1.0 ± 0.8 |
| p value | 0.390 | 0.127 | 1.000 | 0.852 | 0.204 | 0.557 | < 0.001 | 0.008 |
RCA: Regional citrate anticoagulation; Hb: Hemoglobin; PLT: Platelet count; CR: Completion rate of kidney replacement therapy; NUF: net ultrafiltration(L); IRRT: intermittent renal replacement therapy
Discussion
Regional citrate anticoagulation (RCA) demonstrates unique advantages for critically ill patients requiring intermittent hemodialysis (IHD) or hemodiafiltration (HDF) in intensive care settings. Unlike protocols demanding frequent ionized calcium monitoring [20], our locally optimized RCA approach - using calcium-containing dialysate (1.5 mmol/L) and citrate infusion rates fixed at 1.2× blood flow - simplifies clinical workflows while maintaining safety [21]. This is particularly relevant given that only 5–10% of ICU patients undergo IRRT during their stay, often necessitating rapid, reliable anticoagulation to balance bleeding risks and clotting prevention. Our findings reveal that RCA achieves this equilibrium effectively: it prevented premature circuit clotting in 87.5% of sessions, outperformed heparin-free protocols in ultrafiltration efficiency (1.9 ± 1.0 kg vs. 1.4 ± 0.9 kg, P = 0.010), and avoided systemic heparinization without requiring real-time iCa2+ adjustments.
The need for effective dehydration in our cohort, as opposed to the lower net ultrafiltration observed in Leroy’s study, may account for the different outcomes of IRRT. For conditions fluid management supersedes solute clearance requiring strict volume control - such as cardiac dysfunction secondary to fluid overload - IRRT with RCA may offer distinct advantages over CRRT, enabling rapid, controlled ultrafiltration in fluid-sensitive populations. The clinical relevance of our findings lies in the potential of RCA to enhance patient outcomes in critical care by offering a safer and more effective dialysis option for those at high risk of both bleeding and clotting. This is particularly relevant for patients with conditions like active bleeding, recent surgery, or heparin-induced thrombocytopenia, where traditional anticoagulation strategies are not suitable.
The protocol’s robustness is further evident in high-risk subgroups. Patients with elevated level of platelets (> 130 × 109/L) or hemoglobin (> 90 g/L) achieved 100% circuit survival with RCA, contrasting starkly with heparin-free strategies (18% completion rate, P < 0.001). This aligns with the well-established role of activated platelets and erythrocytes in thrombogenesis: by chelating ionized calcium upstream, RCA disrupts both intrinsic and extrinsic coagulation cascades while mitigating cellular contributions to clot stability. Critically, our results challenge the paradigm that CRRT is mandatory for anticoagulation control in unstable patients. Instead, RCA-enabled IRRT emerges as a pragmatic alternative - delivering comparable urea clearance (Kt/V, URR) to heparin-free methods but with superior fluid management and operational simplicity, even in resource-constrained ICUs [22–26].
Clinical advantages of tailored RCA in critical care
Our study demonstrates that a modified RCA protocol using calcium-containing dialysate (1.5 mmol/L) achieves superior circuit longevity compared to both heparin-free strategies and prior RCA reports, particularly in critically ill patients requiring aggressive fluid removal. While Leroy et al. [4] observed a 37% clotting rate with similar dialysate calcium, our protocol’s success (87.5% completion rate) likely stems from two critical adjustments: (1) optimized citrate dosing (1.2×blood flow rate) to compensate for citrate loss at moderate blood flows of 196 mL/min (vs. 232 mL/min in Leroy’s cohort), and (2) prioritization of ultrafiltration goals over maximal solute clearance. This suggested that the severity of water overload was different from Leory’s study, that might result in relative reduced risk of circuit clotting. Despite that the URR was inferior, it is not controversary with the evidence that fluid balance - not urea kinetics - dictates outcomes in unstable patients [27]. Notably, RCA’s efficacy was magnified in subgroups of elevated platelet count or hemoglobin), where it achieved 100% circuit survival, contrasting starkly with heparin-free approaches (18% completion rate). These findings underscore RCA’s unique ability to address both iCa2+-dependent coagulation and cellular contributions to thrombosis (e.g., platelet activation, erythrocyte aggregation), making it indispensable for patients with inflammation-driven coagulopathy. Additionally, RCA with calcium-containing dialysate simplifies the procedure and has demonstrated long-term safety avoiding the occurrence of hypocalcemia [15, 22, 23].
Although our study focused on RCA for IKRT, the protocol’s safety profile—particularly the absence of major bleeding and citrate accumulation—supports its broader applicability to hybrid therapies like SLED. Recent evidence confirms that RCA-SLED is feasible even in high-risk patients (e.g., 51.8% with shock; Di Mario et al., 2025), reinforcing citrate’s role as a universal anticoagulant for hemodynamically fragile populations across KRT modalities.
Toward precision anticoagulation in IRRT
Building on our finding of a 37% venous bubble trap clotting rate—indicating incomplete regional citrate efficacy—we propose that sustaining ionized calcium suppression downstream warrants investigation. Subgroup variability in completion rates (64–100%) further suggests that patient-specific factors (e.g., platelet counts) may guide RCA personalization. Integrating viscoelastic testing (e.g., clot firmness, lysis time) into anticoagulation decisions may better stratify patients for RCA versus heparin-free strategies. Future studies should explore whether dual-site citrate infusion or biomarker-adjusted protocols mitigate circuit loss without compromising metabolic safety, particularly in high blood-flow settings.
This study has several limitations inherent to its retrospective design. First, the baseline characteristics of the two patient groups, including citrate metabolic capacity, and underlying comorbidities influencing the coagulation status, were not evaluated and uniformly matched. This inherent heterogeneity may introduce confounding effects when comparing outcomes between groups. Second, variations in dialysis protocols (e.g. dialyzer selection, and type of IRRT) were present across the cohorts. These differences reflect real-world clinical practice but limit our ability to standardize interventions and isolate the specific impact of citrate dosing optimization.
Conclusion
Our study suggests that a simplified RCA protocol using fixed calcium-containing dialysate (most 1.5 mmol/L) may reduce circuit clotting of IRRT, while maintaining safety in critically ill patients. The observed advantage of RCA appears particularly notable in patients with elevated platelet counts (≥ 130 × 10³/µL) and hemoglobin levels (≥ 9 g/dL), where heparin-free IRRT fails. The protocol’s standardized citrate dosing (1.2× blood flow) could potentially reduce iCa2+ monitoring needs and operational errors, enhancing practicality in critically ill patients. These findings should be interpreted in the context of retrospective data limitations, including potential unmeasured confounders. Future studies should prospectively verify these results while exploring venous-line citrate supplementation strategies to address residual clotting.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
We greatly thank Ying Tan, Chongyan Yu and Jinwei Wang for their help of data collection and analysis.
Author contributions
Q. J. and T. S. conceived and designed the study; Z.L. , J.Y. and T.Z. collected the data; T.Z., J.Y. and T. S. analyzed and interpreted the data; T. Z. drafted the report; J. Y. , Q. J. , and T. S. revised the draft. All authors have read and approved the final manuscript.
Funding
This study was supported by the National High Level Hospital Clinical Research Funding (Interdepartmental Clinical Research Project of Peking University First Hospital; grant number: 2022CR09), the National Science and Technology Major Projects for Major New Drugs Innovation and Development (grant number: 2017ZX09304028), and CAMS Innovation Fund for Medical Sciences (grant number: 2019-I2M-5-046).
Data availability
All data generated or analysed during this study are included in this article. Further enquiries can be directed to the corresponding authors.
Declarations
Ethics approval
This retrospective study complied with the Declaration of Helsinki. The study protocol was reviewed and approved by the Human Ethics Committee of the Peking University First Hospital (MR-11-23-021020). The need for informed consent was waived by the Human Ethics Committee of the Peking University First Hospital due to the retrospective nature of the study and the use of anonymized patient data.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Tao Zhao and Jie Yang contributed equally to the study.
Contributor Information
Qizhuang Jin, Email: jinqizhuang@bjmu.edu.cn.
Tao Su, Email: tao.su@bjmu.edu.cn.
References
- 1.Davenport A. What are the anticoagulation options for intermittent hemodialysis? Nat Rev Nephrol. 2011;7(9):499–508. PMID: 21727925. 10.1038/nrneph.2011.88 [DOI] [PubMed]
- 2.Jha V, Arici M, Collins AJ, et al. Understanding kidney care needs and implementation strategies in low- and middle-income countries: conclusions from a Kidney Disease: Improving Global Outcomes (KDIGO) controversies conference. Kidney Int. 2016;90(6):1164–74. 10.1016/j.kint.2016.09.009 [DOI] [PubMed] [Google Scholar]
- 3.Di Mario F, Regolisti G, Fani F, et al. Sustained Low-Efficiency Dialysis (SLED) with regional citrate anticoagulation (RCA) and new dialysis equipment: a prospective study with serum citrate measurements and electrolyte monitoring. J Nephrol. 2025;38(2):473–80. 10.1007/s40620-024-02201-5 [DOI] [PubMed] [Google Scholar]
- 4.Leroy C, Pereira B, Soum E, et al. Regional citrate anticoagulation versus heparin for intermittent Hemodialysis in ICU patients. Ann Intensive Care. 2021;11(1):13. 10.1186/s13613-021-00803-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Tang X, Chen D, Zhang L, et al. Regional citrate anticoagulation in high-risk bleeding patients during intermittent Hemodialysis. J Zhejiang Univ Sci B. 2022;23(11):931–42. 10.1631/jzus.B2200082. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Zhang Q, Ding D, Wang C, Lu J, Zhu Q, Ding F. Automated regional citrate anticoagulation system based on individualized dosing models in nonliver failure patients undergoing PIRRT therapy. Ren Fail. 2024;46(2):2387207. 10.1080/0886022X.2024.2387207. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Robert T, Bureau C, Lebourg L, et al. Regional citrate anticoagulation during intermittent Hemodialysis without calcium monitoring. Intensive Care Med. 2017;43(12):1927–8. 10.1007/s00134-017-4924-8. [DOI] [PubMed] [Google Scholar]
- 8.Fan Y, Wu F, Zou M, et al. An individualized regional citrate anticoagulation protocol for hemodialysis: a real-world retrospective study. Int Urol Nephrol. 2024;56(1):295–302. 10.1007/s11255-023-03677-z. [DOI] [PubMed] [Google Scholar]
- 9.Rhee H, Berenger B, Mehta RL, et al. Regional citrate anticoagulation for continuous kidney replacement therapy with calcium-containing solutions. Am J Kidney Dis. 2021;78(4):550–e5591. 10.1053/j.ajkd.2021.01.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Fiaccadori E, Regolisti G, Cademartiri C, et al. Citrate-based protocol for sustained low-efficiency Dialysis in AKI. Clin J Am Soc Nephrol. 2013;8(10):1670–8. 10.2215/CJN.00510113. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Evenepoel P, Maes B, Vanwalleghem J, et al. Regional citrate anticoagulation using a conventional calcium-containing dialysate. Am J Kidney Dis. 2002;39(2):315–23. 10.1053/ajkd.2002.30551. [DOI] [PubMed] [Google Scholar]
- 12.Buturovic-Ponikvar J, Cerne S, Gubensek J, et al. Calcium-free versus calcium-containing dialysate for regional citrate anticoagulation. Int J Artif Organs. 2008;31(5):418–24. 10.1177/039139880803100507. [DOI] [PubMed] [Google Scholar]
- 13.Apsner R, Buchmayer H, Gruber D, et al. Citrate for long-term hemodialysis: prospective study of 1009 consecutive high-flux treatments in 59 patients. Am J Kidney Dis. 2005;45(3):557–64. 10.1053/j.ajkd.2004.12.002. [DOI] [PubMed] [Google Scholar]
- 14.Faguer S, Saint-Cricq M, Nogier MB, et al. Heparin-free intermittent Hemodialysis using calcium-free citrate dialysate in critically ill patients. Crit Care Med. 2017;45(11):1887–92. 10.1097/CCM.0000000000002694. [DOI] [PubMed] [Google Scholar]
- 15.Morabito S, Pistolesi V, Tritapepe L, Fiaccadori E. Implementation of a simplified regional citrate anticoagulation protocol for post-dilution continuous hemofiltration using a bicarbonate-buffered, calcium-containing replacement solution [published correction appears in Blood Purif. 2017;43(1–3):144]. Blood Purif. 2016;42(4):349–355. 10.1159/000452755 [DOI] [PubMed]
- 16.Morabito S, Pistolesi V, Tritapepe L, et al. Regional citrate anticoagulation for RRTs in critically ill patients with AKI. Clin J Am Soc Nephrol. 2014;9(12):2173–88. 10.2215/CJN.01280214. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Liu J, Liu Z, Zhao T, Su T, Jin Q. Thromboelastography and traditional coagulation testing in Non-ICU-Admitted patients with acute kidney injury: an observational cohort study. Am J Nephrol. 2023;54(5–6):208–18. 10.1159/000530777. [DOI] [PubMed] [Google Scholar]
- 18.KDIGO Clinical Practice Guideline for Acute Kidney Injury. Kidney Int Suppl. 2012;2(1):19–36. 10.1038/kisup.2011.32. [Google Scholar]
- 19.Hemostasis and thrombosis. Curr Opin Hematol. 1994;1(5):B113–69. [PubMed] [Google Scholar]
- 20.Clark JA, Schulman G, Golper TA. Safety and efficacy of regional citrate anticoagulation. Clin J Am Soc Nephrol. 2008;3(3):736–42. 10.2215/CJN.03460807. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Oudemans-van Straaten HM, Kellum JA, Bellomo R. Anticoagulation for continuous renal replacement therapy. Crit Care. 2011;15(1):202. 10.1186/cc9358. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Meier K, Saenz DM, Torres GL, et al. Thrombelastography in renal dysfunction and intracerebral hemorrhage. J Stroke Cerebrovasc Dis. 2018;27(5):1350–6. 10.1016/j.jstrokecerebrovasdis.2017.12.026. [DOI] [PubMed] [Google Scholar]
- 23.Vaisman S, Kensey K, Cho YI. Hemodialysis and whole blood viscosity. Int J Artif Organs. 2009;32(6):329–35. 10.1177/039139880903200603. [DOI] [PubMed] [Google Scholar]
- 24.Wink J, Vaziri ND, Barker S, et al. Hemodialysis and blood viscosity. Int J Artif Organs. 1988;11(5):340–2. [PubMed] [Google Scholar]
- 25.Marchioli R, Finazzi G, Specchia G, et al. Cardiovascular events in polycythemia Vera. N Engl J Med. 2013;368(1):22–33. 10.1056/NEJMoa1208500. [DOI] [PubMed] [Google Scholar]
- 26.Ogasawara N, Oguro T, Sakabe T, et al. Hemoglobin and indoleamine 2,3-dioxygenase expression. J Cell Biochem. 2009;108(3):716–25. 10.1002/jcb.22308. [DOI] [PubMed] [Google Scholar]
- 27.Alvarez G, Chrusch C, Hulme T, et al. Renal replacement therapy: a practical update. Can J Anaesth. 2019;66(5):593–604. 10.1007/s12630-019-01306-x. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
All data generated or analysed during this study are included in this article. Further enquiries can be directed to the corresponding authors.