Continuous renal replacement therapy (CRRT) is now widely used to treat acute renal failure (ARF) in critically ill patients. 1 One of the concerns with the use of CRRT is that it typically requires some form of anticoagulation to prevent circuit clotting. Although heparin is the most common choice of anticoagulant, 1 it is thought to be associated with bleeding complications, especially in high risk patients. 2 Furthermore, its anticoagulation effect might be limited in patients with antithrombin III deficiency, 3 and heparin can cause thrombocytopenia. 4 Therefore, several drugs have been used during CRRT as an alternative anticoagulant: low molecular weight heparin, 5,6 citrate, 7,8 prostacyclin, 9,10 hirudin, 11,12 and nafamostat mesilate. 13,14 Most of these drugs, however, are much more expensive than heparin. They also have their own shortcomings, such as a long half life, inability to reverse anticoagulation, 6,11 difficulty in monitoring, 6,11 hypocalcemia, alkalosis, 8,15 hypotension, 10 and unavailability (except in Japan). 13,14
Given the above concerns about the risks and costs of anticoagulation, the simplest technique in patients deemed at risk of bleeding might be that of CRRT with “no anticoagulation.” This technique has been previously reported in the literature. 2,16–18 However, information on this approach is limited because of its retrospective nature 2 or small sample size. 16–18 Therefore, we have conducted a large prospective observational study of circuit life during continuous venovenous hemofiltration (CVVH) where no anticoagulation was used in patients who fulfilled criteria for a risk of bleeding and now report our findings.
Materials and Methods
All patients who had ARF treated with CVVH in the intensive care unit (ICU), Austin Hospital, Melbourne, Australia from February 2001 to July 2002 and who did not die within 24 hours were prospectively observed as part of our ongoing quality assurance program. Prospective collection of data on circuit life has been and is an ongoing quality assurance activity in our unit. Thus the collection of data in relation to this manuscript was classified as a quality assurance activity. In our institution, the Ethics Committee waives the need for informed consent under such circumstances.
Continuous Venovenous Hemofiltration Technique and Anticoagulation Protocol
CVVH was conducted with the AK-10 machine (Gambro, Lund, Sweden), the BM-10/BM-11 machine (Edwards Life Sciences, Sydney, Australia), or the Hygieia machine (Kimal, Sydney, Australia). Venovenous access was secured by inserting a 13.5 Fr double lumen catheter (Niagara, Vascath, Ontario, Canada) into one of the large veins (most commonly femoral veins). Either AN69 (Hospal, Lyon, France) or APS650 (Asahi Medical, Tokyo, Japan) hollow fiber hemofilters were used. Blood flow was kept at 200 ml/min. Commercially prepared lactate buffered hemofiltration replacement fluid (Hemofiltration Solution, Edwards Life Science, Sydney, Australia) was infused pre- or postcircuit at 2 L/hour. In the February 2001 to December 2001 period, replacement fluid was infused precircuit. From January 2002 to July 2002, because of a change in our unit protocol, all patients were treated with postdilution CVVH (replacement fluid was infused postcircuit). When lactic acidosis (lactate > 5 mmol/L) occurred, by protocol, the replacement fluid was switched to a lactate free, bicarbonate buffered fluid (Hemosol, Gambro, Lund, Sweden).
Anticoagulation was conducted according to our protocol. Five IU/kg/hour of heparin were given as standard circuit anticoagulation unless one of the following conditions applied. If the prothrombin time-international normalized ratio (PT-INR) was less than 1.5, the activated partial thromboplastin time (APTT) was less than 40 seconds, and platelet counts more than 150 × 103/μl, then 10 IU/kg/hour of heparin were administered. If a patient had more than one of the following criteria, no anticoagulation was used: PT-INR more than 2.5, APTT more than 60 seconds, platelet counts less than 60 × 103/μl, active bleeding, or in the 24 hour postsurgery period. When circuit life was unacceptably short with no anticoagulation (e.g., < 6 hours for two consecutive circuits), regional anticoagulation (typically, heparin 1,000 IU/hour precircuit plus protamine 10 mg/hour postcircuit) was applied.
Data Collection and Statistics
Age, gender, admission diagnosis, SAPS-II, 19 and APACHE-II 20 were obtained at ICU admission, as is routine for our unit database. For each circuit, the following data were also obtained: starting date and time, ending date and time, heparin usage, heparin dose, and protamine usage. Blood tests were performed at 5 A.M. daily and results (platelet count, PT-INR, and APTT) were collected, as is routine in our institution. No additional anticoagulation tests were performed to monitor circuit anticoagulation. Hospital mortality was also obtained. For each circuit, platelet count, PT-INR, and APTT were defined as the values measured in the morning of the same day as the circuit was used. If a circuit lasted more than 2 days, averaged values were used.
For analysis, circuits were divided into three groups: heparin group (circuits anticoagulated with heparin), no anticoagulation group (circuits that were not anticoagulated), and protamine group (circuits regionally anticoagulated with heparin/protamine). Data are expressed as means (standard deviations). Chi-squared analysis and the Kruskal-Wallis test were used for nominal values and numerical variables, respectively, to compare the three groups (StatView, Abacus Concepts, Berkeley, CA). When significant, post hoc analysis was conducted with the Mann-Whitney test. Circuit life in the three groups was presented graphically using Kaplan-Meier product limit survival plots. The log-rank (Mantel-Cox) test was used to compare circuit life among the three groups. A p value of < 0.05 was considered statistically significant.
Forty-eight patients were included in this study. The demographics of these patients are shown in Table 1. Among these, 35 patients were treated with at least one no anticoagulation CVVH for at least one filter, and five patients were treated with regional anticoagulation CVVH for at least one filter. Three hundred circuits were observed. Among these, 143 circuits (47.6%) received no anticoagulation, and 31 (10.3%) received regional anticoagulation. The reasons why low dose heparin anticoagulation was not chosen for these circuits were (several circuits had more than one reason) low platelet count in 98 circuits, postsurgery period in 36, high APTT in 31, high PT-INR in 20, and active bleeding before CVVH in 10 (7 cerebral hemorrhage and 3 gastrointestinal bleed). No patient experienced any significant complication (e.g., major bleeding, thrombocytopenia, or hemodynamic instability) secondary to heparin or protamine during CVVH.
Clotting variables and circuit life are shown in Table 2. There was no significant difference in circuit life among the three groups (heparin, 20.9 hours; no anticoagulation, 19.3 hours; protamine, 21.2 hours; p = 0.38).
Kaplan-Meier survival plots of circuit life are shown in Figure 1A and 1B. Because the filter life was significantly shorter in the postdilution period (18.0 vs. 13.0 hours, p = 0.021), the three groups were divided into pre- and postdilution periods. However, there was no significant difference in circuit survival among the three groups, both in the pre- and postdilution periods.
Our anticoagulation policy during CVVH is very simple: 1) use low dose of heparin (5–10 IU/kg/hr), 2) if a patient is at risk of bleeding according to predefined criteria, use no anticoagulation, and 3) if circuit life is too short without anticoagulation, use regional anticoagulation with heparin/protamine. Using this protocol in our ICU, approximately half of our circuits received no anticoagulation. However, these circuits achieved an acceptable lifespan with no risk to the patient. Furthermore, circuit life in these patients was not different from that of patients receiving low dose heparin or regional heparin-protamine anticoagulation, and there were no bleeding complications secondary to CVVH. Our findings have practical implications and require detailed discussion.
Continuous Venovenous Hemofiltration With Low Dose Heparin Anticoagulation
The relatively high incidence of bleeding complications secondary to full heparin anticoagulation during CRRT has been reported and is quoted as a major risk and shortcoming of CRRT. 2,6,21–23 However, with low dose heparin (5–10 IU/kg/hour), we experienced no bleeding complications during 2,600 hours of CVVH (126 circuits). Gretz et al.24 also used low dose heparin (5.6 IU/kg/hr) in 216 circuits during CVVH and experienced no bleeding. These findings suggest that, at least in patients with relatively low risk of bleeding (PT-INR < 2.5, APTT < 60 seconds, platelet count > 60 × 103/μl, not immediately postsurgery), low dose heparin anticoagulation can be safely performed during CRRT. One might argue that a mean circuit life of 20.9 hours is too short to be accepted in daily practice. However, if heparin dose is increased to attempt to prolong circuit life, bleeding complications will become inevitable. Indeed, van de Wetering et al.25 found that, for every 10 second increase in APTT, the incidence of circuit coagulation was decreased by 25%, whereas the risk of hemorrhage increased by 50%. Considering the risk of full anticoagulation (bleeding in a critically ill patient) and its possible benefits (reduced cost and workload), we think it preferable to conduct low dose heparin anticoagulation during CVVH in patients not deemed to be at a particular risk of bleeding.
Continuous Venovenous Hemofiltration Without Anticoagulation
In a pilot study, we previously reported a mean circuit life of 32 hours during CVVH without anticoagulation in high risk patients. 16 Using different techniques of CRRT, others have reported a mean circuit life of between 22.1 and 35 hours. 2,17,18 Our current study markedly extends the experience with this approach to CVVH in high risk patients and confirms its feasibility in this population. We also found that there was no difference in circuit life between the low dose heparin group and the no anticoagulation group. Martin et al.2 conducted a retrospective study comparing no anticoagulation with low dose heparinization (100–700 IU/hr) and found a similar mean circuit life between the two groups (22.1 vs. 24.7 hours). These investigators also found that platelet count was different between the two groups (42 vs. 142 × 103/μl), consistent with our finding. Platelets are an important factor for clot formation in the circuit. 26 Therefore, CVVH with no anticoagulation appears particularly appropriate in patients with marked thrombocytopenia.
Continuous Venovenous Hemofiltration With Heparin/Protamine Regional Anticoagulation
We observed 31 circuits with regional anticoagulation and found that a mean circuit life was 21.2 hours with no complication (bleeding or prolonged APTT). We previously conducted a randomized crossover study comparing low dose heparin and regional anticoagulation with heparin/protamine during continuous venovenous hemodiafiltration 17 and found a trend in favor of regional anticoagulation. We also observed no bleeding complication nor prolonged APTT during regional anticoagulation. There are some concerns about regional anticoagulation with heparin/protamine (potential side effects, meticulous dose adjustment, and risk of bleeding). However, given the evidence available and the findings of our study, we consider this technique to be a useful method of rescue anticoagulation in a patient with a high risk of bleeding and a short circuit life in the absence of any anticoagulation.
Alternative Approaches to Anticoagulation During Continuous Renal Replacement Therapy
Several other drugs have been used during CRRT to prolong circuit life: low molecular weight heparin (LMWH), citrate, prostacyclin, hirudin, and nafamostat mesilate. Among these drugs, LMWH, prostacyclin, and hirudin have been compared with heparin in randomized controlled studies, none of which showed a significant increase in circuit life. 6,10,11 Furthermore, for an 80 kg patient in Australia, the cost of one day of LMWH therapy is $9.00 (U.S.) (dalteparin, 10 IU/kg/hr) and that of prostacyclin is $105.00 (4 ng/kg/min). Hirudin is not available in Australia, but the cost of this drug is $8.00/day in Europe (lipirudin, 4 μg/kg/hr). Heparin anticoagulation costs only $1.30/day (10 IU/kg/hr).
Citrate and nafamostat mesilate have not been compared with heparin in randomized controlled trials. Citrate causes anticoagulation by chelation of ionized calcium. Its effect is reversed by the postcircuit infusion of calcium so that it does not cause systemic anticoagulation. However, patients with liver dysfunction cannot metabolize infused citrate at a sufficient rate and may develop severe hypocalcemia. 15 Moreover, this technique requires a special dialysate/replacement fluid (containing no calcium), which is not commercially available. It costs approximately $520.00 to produce enough of this fluid to perform CVVH at 2 L/hr of ultrafiltration with citrate anticoagulation, 27 compared with $80.00 for the commercially available lactate replacement fluid used in our patients.
Nafamostat mesilate is a synthetic serine protease inhibitor, which causes an anticoagulation effect by blocking factors Xa, XIIa, and thrombin. Because of its short half life (approximately 8 minutes), it doesn’t affect systemic coagulation. 14 However, nafamostat mesilate is available only in Japan, and it is very expensive (approximately $520.00/day with 0.5 mg/kg/hr for a 60 kg patient).
Mean circuit life for CVVH with citrate anticoagulation has been reported by Palsson and Niles 8 to be 29.5 hours. Mean filter life for CVVHDF with nafamostat mesilate anticoagulation has been reported by Kitamura et al.28 to be 29.2 hours. Therefore, to prolong circuit life from 20 hours to 30 hours (saving 0.4 circuits per day), one needs to spend approximately $500/day beyond standard costs, if citrate or nafamostat mesilate is chosen rather than low dose heparin or no anticoagulation. Moreover, these techniques have never been shown to reduce bleeding complication compared with low dose heparin in a randomized study. It is unlikely that, from the bleeding risk point of view, they would be safer than no anticoagulation. Thus, although these techniques of anticoagulation are available, their superiority in terms of costs or safety during CVVH remains doubtful.
There are several system modifications with which filter life might be prolonged, such as high blood flow and reduction of air-blood mixing. 29,30 With such techniques, our simple anticoagulation protocol can be conducted safely and at low costs.
Our investigation is not a randomized controlled study, and thus our findings might have been affected by patient characteristics. This is a major limitation of our investigation. However, our study is prospective in design and the observational period was relatively short (totally 19 months), attenuating any possible time-related confounding variables. We observed 300 circuits, which is more than any other published report of circuit life, to our knowledge. Furthermore, the protocol of circuit anticoagulation management was simple and is easily reproducible. There is no definition of what an adequate circuit life should be. However, with a conservative and low cost approach to anticoagulation, we achieved a mean circuit life of approximately 20 hours. This value is similar to those previously reported with CVVH. 2,9,10,11
In summary, in a group of patients deemed to be at risk of bleeding, we found that CVVH without anticoagulation is feasible, safe, and able to achieve circuit life similar to low dose heparin anticoagulation and regional anticoagulation with heparin/protamine, as well as similar to previously reported values for this technique. Furthermore, using a simple protocol for anticoagulation, CVVH was conducted safely, with acceptable circuit life, and at low cost.
This study was supported by the Austin & Repatriation Medical Center Anaesthesia and Intensive Care Trust Fund.
1. Silvester W, Bellomo R, Cole L: Epidemiology, management, and outcome of severe acute renal failure of critical illness in Australia. Crit Care Med 29: 1910–1915, 2001.
2. Martin PY, Chevrolet JC, Suter P, et al: Anticoagulation in patients treated by continuous venovenous hemofiltration: a retrospective study. Am J Kidney Dis 24: 806–812, 1994.
3. Langley PG, Keays R, Hughes RD, et al: Antithrombin III supplementation reduces heparin requirement and platelet loss during hemodialysis of patients with fulminant hepatic failure. Hepatology 14: 251–256, 1991.
4. Warkentin TE, Kelton JG: Temporal aspects of heparin-induced thrombocytopenia. New Engl J Med 344: 1286–1292, 2001.
5. De Pont AC, Oudemans-van Straaten HM, Roozendaal KJ, et al: Nadroparin versus dalteparin anticoagulation in high-volume, continuous venovenous hemofiltration: A double-blind, randomized, crossover study. Crit Care Med 28: 421–425, 2000.
6. Reeves JH, Cumming AR, Gallagher L, et al: A controlled trial of low-molecular-weight heparin (dalteparin) versus unfractionated heparin as anticoagulant during continuous venovenous hemodialysis with filtration. Crit Care Med 27: 2224–2228, 1999.
7. Tolwani AJ, Campbell RC, Schenk MB, et al: Simplified citrate anticoagulation for continuous renal replacement therapy. Kidney Int 60: 370–374, 2001.
8. Palsson R, Niles JL: Regional citrate anticoagulation in continuous venovenous hemofiltration in critically ill patients with a high risk of bleeding. Kidney Int 55: 1991–1997, 1999.
9. Fiaccadori E, Maggiore U, Rotelli C, et al: Continuous haemofiltration in acute renal failure with prostacyclin as the sole anti-haemostatic agent. Int Care Med 28: 586–593, 2002.
10. Langenecker SA, Felfernig M, Werba A, et al: Anticoagulation with prostacyclin and heparin during continuous venovenous hemofiltration. Crit Care Med 22: 1774–1781, 1994.
11. Vargas Hein O, von Heymann C, Lipps M, et al: Hirudin versus heparin for anticoagulation in continuous renal replacement therapy. Int Care Med 27: 673–679, 2001.
12. Fischer KG, van de Loo A, Bohler J: Recombinant hirudin (lepirudin) as anticoagulant in intensive care patients treated with continuous hemodialysis. Kidney Int 72 (Suppl): S46–50, 1999.
13. Hu ZJ, Iwama H, Suzuki R, et al: Time course of activated coagulation time at various sites during continuous haemodiafiltration using nafamostat mesilate. Int Care Med 25: 524–527, 1999.
14. Ohtake Y, Hirasawa H, Sugai T, et al: Nafamostat mesylate as anticoagulant in continuous hemofiltration and continuous hemodiafiltration. Contrib Nephrol 93: 215–217, 1991.
15. Meier-Kriesche HU, Gitomer J, Finkel K, et al: Increased total to ionized calcium ratio during continuous venovenous hemodialysis with regional citrate anticoagulation. Crit Care Med 29: 748–752, 2001.
16. Tan HK, Baldwin I, Bellomo R: Continuous veno-venous hemofiltration without anticoagulation in high-risk patients. Int Care Med 26: 1652–1657, 2000.
17. Bellomo R, Teede H, Boyce N: Anticoagulant regimens in acute continuous hemodiafiltration: a comparative study. Int Care Med 19: 329–332, 1993.
18. Holt AW, Bierer P, Glover P, Plummer JL, et al: Conventional coagulation and thromboelastograph parameters and longevity of continuous renal replacement circuits. Int Care Med 28: 1649–1655, 2002.
19. Knaus WA, Draper EA, Wagner DP, Zimmerman JE: APACHE II: a severity of disease classification system. Crit Care Med 13: 818–829, 1985.
20. Le Gall J-R, Lemeshow S, Saulnier F: A new simplified acute physiology score (SAPS II) based on an European/North American multicenter study. JAMA 270: 2957–2963, 1993.
21. Gabutti L, Marone C, Colucci G, et al: Citrate anticoagulation in continuous venovenous hemodiafiltration: a metabolic challenge. Int Care Med 28: 1419–1425, 2002.
22. Ward DM, Mehta RL: Extracorporeal management of acute renal failure patients at high risk of bleeding. Kidney Int 41 (Suppl): S237–244, 1993.
23. Kaplan AA, Longnecker RE, Folkert VW: Continuous arteriovenous hemofiltration. A report of six months’ experience. Ann Intern Med 100: 358–367, 1984.
24. Gretz N, Quintel M, Ragaller M, et al: Low-dose heparinization for anticoagulation in intensive care patients on continuous hemofiltration. Contrib Nephrol 116: 130–135, 1995.
25. van de Wetering J, Westendorp RG, van der Hoeven JG, et al: Heparin use in continuous renal replacement procedures: the struggle between circuit coagulation and patient hemorrhage. J Am Soc Nephrol 7: 145–150, 1996.
26. Davenport A: The coagulation system in the critically ill patients with acute renal failure and the effect of an extracorporeal circuit. Am J Kidney Dis 30 (Suppl 4): 20–27, 1997.
27. Jahansouz F, Lamott J, Shephanson LG, et al: Pharmacy cost analysis of citrate anticoagulation (CIT) for continuous renal replacement therapy (CRRT) (Abstract). J Am Soc Nephrol 12: 171A, 2001.
28. Kitamura N, Hirasawa H, Sugai T, et al: Experience of using APF-06W (PAN-DX) during continuous hemodiafiltration. ICU and CCU 18: 197–202, 1994.
29. Schetz M: Anticoagulation in continuous renal replacement therapy. Contrib Nephrol 132: 283–303, 2001.
30. Gretz N, Quintel M, Ragaller M, et al: Low-dose heparinization for anticoagulation in intensive care patients on continuous hemofiltration. Contrib Nephrol 116: 130–135, 1995.