Regional hemodialysis anticoagulation: hypertonic tri-sodium citrate or anticoagulant citrate dextrose-A

Regional Hemodialysis Anticoagulation: Hypertonic Tri-Sodium Citrate or Anticoagulant Citrate Dextrose-A Michael J. Flanigan, MD, FACP, Lisa Pillsbury, MD, George Sadewasser, and Victoria S. Lim, MD, FACP

BSN,

0 Regional citrate anticoagulation should be a simple process of substituting hypertonic (1.6 mol/L) citrate for heparin and adjusting the infusion to obtain an arterial activated clotting time of 150 to 266 seconds. Serious, documented complications of citrate anticoagulation involve citrate intoxication during isolated ultrafiltration; hyperaluminemia, hyperammonemia, and hypernatremia during sorbent dialysis; and profound alkalosis, paresthesias, arrhythmia, and cardiac arrest during bicarbonate dialysis. We suspected that some of these complications could be avoided by using anticoagulant citrate dextrose-A (ACD) rather than hypertonic tri-sodium citrate (TSC) as the anticoagulant. In a cross-over study with random assignment order eight adults underwent mid-week dialyses with ACD (0.113 mol/L citrate) and TSC (1.6 mol/L citrate) regional citrate anticoagulation. Predialysis to postdialysis changes in Na (mEq/L), Ca (mg/dL), ionized Ca (mg/dL), pH, and HCO, (mEq/L) are listed below.

A TSC ACD Data

A

Na

6.8 * 2.1 4.8 k 2.1 are

given

as mean

2.4 2.1 values

A

Ca + 0.4 c 1.0

-0.9 -1.5

iCa

A

A PH

t 0.2 t 0.2

0.13 0.09

? 0.01 f 0.02

HCOB

10.4 + 1.7 6.6 2 0.9

? 1 sem.

Using continuous blood flow and avoiding isolated ultrafiltration and sorbent dialysis should prevent the delivery system complications of regional citrate anticoagulation. During this evaluation isotonic and hypertonic citrate resulted in similar serum sodium changes, and standard dialysate effectively reversed the citrate/calcium interaction of both hypertonic and isotonic citrate infusions to restore homeostasis without a separate calcium infusion. The combination of TSC and bicarbonate dialysate does produce a profound metabolic alkalosis, which is lessened by using ACD. In general, regional citrate anticoagulation is simplified by using standard dialysate with a hypertonic rather than an isotonic citrate infusion, and dangerous complications are further evaded by adjusting the dialysate bicarbonate to 25 to 36 mmol/L or substituting a mixture of citric acid and TSC (ACD) for TSC. 0 1996 by the National Kidney Foundation, inc. INDEX

WORDS:

Hemodialysis;

anticoagulation;

anticoagulant

T

AND

Journal

of Kidney

METHODS

Diseases,

Vol 27, No 4 (April),

tri-sodium

citrate.

From the Department of Medicine, University of Iowa Hospitals and Clinics, and the Department of Medicine, University of Iowa College of Medicine, Iowa City, IA. Received August 31, 1995; accepted in revised form December 12, 1995. Address reprint requests to Michael J. Flanigan, MD, FACP, Department of Medicine, W-346.GH, University of Iowa Hospitals and Clinics, Iowa City, IA 52240-6040. 0 1996 by the National Kidney Foundation, Inc. 0272~6386/96/2704-0008$3.00/O

The study was a pretest-posttest design with balanced assignment order. All studies took place in the University of Iowa Hospitals and Clinics Outpatient Hemodialysis Center, American

dextrose;

a hospital-based nonprofit facility in a state-owned teaching hospital. Study subjects were consenting adults who had received regular hemodialysis for more than 6 months. Each subject was assigned to undergo sequential midweek dialyses using either TX followed by ACD or vice versa based on the last digit of their hospital identification number. Tri-sodium citrate consisted of 47.6% or 1.6 mol/L TSC dihydrate in water (triCitraso1; Cytosol Laboratories, Braintree, MA). Anticoagulant citrate dextrose-A was an isotonic mixture of dextrose, citric acid, and TSC (0.038 mol/L anhydrous citric acid, 0.075 mol/L TSC dihydrate; ACD-A, Cytosol Laboratories) used by blood banks for cell storage, plasma, and platelet pheresis. As shown in Table 1, the starting doses of anticoagulant

HERE are multiple regional citrate anticoagulation methodslm4; some are particularly cumbersome,2-4 but all are capable of instigating profound metabolic derangements.5-9 Since trisodium citrate (TSC) anticoagulates by chelating calcium, it is reasonable to expect the therapeutic effect to be proportional to citrate activity rather than sodium content. Blood banks use anticoagulant citrate dextrose-A (ACD), a mixture of dextrose, citric acid, and TSC, to anticoagulate and store blood. We believed that this readily available compound might substitute for hypertonic TSC and reduce the postdialysis alkalosis and hypernatremia of TSC anticoagulation. This presentation describes ACD regional hemodialysis anticoagulation and compares its effects with those of hypertonic TSC. MATERIALS

citrate

1996:

pp 519-524

519

520

FLANIGAN Table

1. Estimated

(mUmin)

TSC (mUmin)

TSC (mL/hr)

100 200 300 400

0.25 0.50 0.75 1 .oo

15.0 30.0 45.0 60.0

Qe

Anticoagulant

Infusion

TSC (mudialysis)

NOTE. The initial ACD and TSC infusion rates chosen concentration of approximately 4 mmol/L. This concentration dialysis patient’s Lee-White clotting time.13

were dependent on the blood flow rate and calculated to achieve a whole blood citrate of 4 mmol/L, a level consistent with clinical anticoagulation.’ Infusion rates were then adjusted up or down by 5 to 10 mL/hr for TSC and 7.5 to 150 mL/hr for ACD to achieve a predialyzer activated clotting time 25% to 75% greater than the predialysis baseline. Dialysis followed the patient’s standard orders and dialysate bicarbonate was not modified for these isolated dialyses. Patients used F-80 polysulphone dialyzers (Fresenius USA, Walnut Hills, CA) and received enough dialysis to provide aKt/V urra2 1.2. The final dialysate contained 140 i 2 mmol/ L sodium, 1.625 ir 0.15 mmol/L calcium, 0.6 ? 0.1 mmol/ L magnesium, 38 !c 2 mmol/L bicarbonate, and 3 mmol/L acetate (DiaLyte; Dial Medical Supply, Cobe-Gambro-Hospal, Denver, CO). Dialysis and dialyzer reprocessing records were reviewed for each dialysis session to determine the amount of anticoagulant used and whether complications (paresthesias, hypotension, cramps, or arrhythmia) occurred, and to identify dialyses with clotting problems (a ~5% decrease in fiber bundle volume). The following laboratory tests were obtained to monitor the electrolyte changes of regional citrate anticoagulation dialysis: predialysis: sodium, calcium (total and ionized) and arterial blood gases (ABGs); postdialysis: calcium (total and ionized), and ABGs; and 15 minutes postdialysis: sodium, calcium (total and ionized), and ABGs. Laboratory samples were obtained anaerobically from the arterial blood line at a location proximal to the citrate infusion. Analyses were run within 30 minutes and used standard techniques, including ion-specific electrode measurement of hydrogen, sodium, and calcium activities. Descriptive data are presented as mean values + 1 SEM. Comparative analyses used a paired Student’s r-test and two-tailed P values 5 0.05 were considered significant.

Rates

ACD (mUmin)

60 120 180 240

ACD (mUhr)

3.5 7.0 10.5 14.0 for

regional produces

ET AL

ACD (ml/dialysis)

210 420 630 840

anticoagulation achieve near infinite prolongation

840 1680 2520 3360 a whole blood of the average

citrate hemo-

and one set of ACD data was lost due to premature dialysis termination. Twelve of 13 (seven of seven TSC and five of six ACD) citrate anticoagulations were successfully completed. The incidence of paresthesias was 21%; there was one episode of hypotension (TSC group) and no episodes of tetany or cardiac arrest. There were four instances of clotting noted in the nursing records, but only one of these was associated with dialyzer volume loss 2 5%; this occurred in the ACD group with a predialysis fiber bundle volume of 109 mL decreasing to 84 mL postdialysis. In total the dialyzer fiber bundle volume changed from 120 + 3 mL predialysis to 114 + 6 mL postdialysis. The volume of citrate solution delivered ranged from 130 mL in the TSC group to 3,500 mL in the ACD group. Absolute citrate consumption, 3 18 and 305 mm01 for the TSC and ACD groups, respectively, was similar in both groups (Table 2). The consequences of these hypertonic and isotonic infusions are summarized in Table 3. Generally there were increases in serum sodium, bicarbonate, and pH. Although total calcium increased during and after dialysis, ionized calcium

Table

2. Citrate

Infusion

During

Hemodialysis

mL Infused

mm01

RESULTS

Eight hemodialysis patients (four men and four women) were recruited from the University of Iowa Hospitals and Clinics Outpatient Dialysis Unit. The subjects were 33 to 76 years of age (5 1 + 6.1 years) and weighted 83 f 11 kg (range, 54 to 153 kg). One individual was unable to complete both dialyses because of hospitalization,

TSC ACD

200 2,700

k 20 i 275

320 305

t 32 rf- 31

NOTE. The average hemodialysis session lasted 3.8 hours and achieved a KW Urea of 1.3. Anticoagulation was adjusted to maintain the activated clotting time at 125% to 175% of its baseline level (generally 130 to 200 seconds). The actual quantity of citrate required was independent of the ACD or TSC formulation.

t

CITRATE

ANTICOAGULATION Table

521

3. Arterial Blood Gasses Citrate Anticoagulation

Predialysis Serum

During

Postdialysis

15 Minutes Postdialysis

sodium

WW TSC ACD Arterial pH TSC ACD Arterial Pco?

137.3 138.6

* 1.7 k 1.7

144.0 rt 1.1 144.4 2 1.7

7.42 2 0.14 7.43 2 0.02

7.55 + 0.02 7.52 + 0.01

7.52 k 0.01 7.51 2 0.01

41 + 1.6 41 i 1.6

43 t 2.2 40 k 1.7

47 t 2.7 43 2 1.8

WdL) TSC ACD Serum calcium

25.5 t 0.9 24.7 k 0.4

35.9 2 2.1 31.4 + 0.9

35.9 5 1.6 33.0 k 0.7

O-WW TSC ACD Ionized calcium

9.7 k 0.4 9.8 k 0.4

12.1 2 0.7 12.1 i: 0.7

11.8 ? 0.7 11.7 + 0.7

4.7 2 0.2 5.0 k 0.3

3.8 + 0.1 3.6 2 0.2

4.2 f 0.2 4.2 ? 0.2

(mm W TSC ACD Serum bicarbonate

OWW TSC ACD

NOTE. Serum electrolyte and arterial blood gas changes induced by regional citrate anticoagulation and bicarbonate dialysis are little changed by using ACD rather than hypertonic TSC. While ACD delivers less sodium and base, the clinically apparent difference is that of reduced metabolic alkalosis following ACD dialysis.

universally decreased during citrate anticoagulation. Additionally, some patients incurred a remarkable metabolic alkalosis when bicarbonate dialysate and regional citrate anticoagulation were combined (Fig 1). The 15minute postdialysis values of total calcium, ionized calcium, and pH suggest that ionized calcium remains depressed in response to the postdialysis alkalosis since citrate should have cleared the circulation by 15 minutes postdialysis (Table 3, Fig 2). DISCUSSION

Hemodialyzers convert a milliliter of blood into a 0.10~mm film contacting 150 cm’ of nonendothelial surface. This inevitably catalyzes contact-activated responses, including cellular adherence; first-phase platelet aggregation, macrophage activation, and neutrophil degranulation; protein absorption with fibronection coating; and compliment activation. “-” As a result of these surface interactions, fibrinolysins, kinins, leukotrienes, peroxidases, proteases, and other inflammatory components, including the humoral and cellular immune systems, are initiated.‘“‘7 Activating the clotting cascade is unavoidable and occurs dependably dhg

hem~~ysis.‘0~“~18~20

Early dialysis systems used immense quanti-

=65 $60

G65

g55 N 50 &45

g55 N 50 845

2 60

f40 g 35 ;30

i 35 ;30 F 6 25 = 20

g 25 = 20 012345678

tiii345678 Subject Fig 1. The metabolic alkalosis anticoagulation can be profound 7.6. Total venous carbon dioxide at 15 minutes postdialysis (A). metabolic alkalosis.

Subject

following bicarbonate dialysis (dialysate bicarbonate, 66 mmol/L) and citrate and depends on compensatory hypoventilation to keep the arterial pH below or HC03 predialysis (0) increased immediately postdialysis (0) and equilibrated Postdialysis Pcop values (U) often exceed 46 mm Hg to compensate for the

FLANIGAN

522

bii3;156?8

ET AL

01234567 Subject

Subject

Fig 2. Serum total and ionized calcium are altered in opposite directions by citrate anticoagulation. The persistent reduction of ionized calcium even 15 minutes postdialysis is likely the result of metabolic alkalosis. Predialysis total and ionized calcium values are represented by (0); the immediate postdialysis total and ionized ionized calcium are represented by (0). Equilibrated or 15minute postdialysis total serum calcium (A) and ionized calcium (V) reveal a stable total serum calcium and an increase in ionized calcium as residual citrate is metabolized.

ties of heparin to prevent extracorporeal thrombosis and the literature was replete with reports of atraumatic, gastrointestinal, and subdural hemorrhage.2’ Technical advances in systems design and use have reduced, but not eliminated, the need for anticoagulants and the incidence of spontaneous hemorrhages. Dialysis anticoagulation continues to have adverse effects in “high” and “very high risk” settings’.21.22;consequently, physicians have used peritoneal dialysis; heparin-free and “low-dose” heparin dialysis; prostaglandin, nafamostat mesilate, and low molecular weight heparin infusions; and regional citrate anticoagulation to reduce the incidence of bleeding diatheses.1~2~5,2’-25 While “heparin-free” hemodialysis sounds appealing, dialysis without anticoagulants is only successful 70% to 90% of the time and inevitably leads to a “subclinical” disseminated intravascular coagulopathy. ’ ‘,22,26-28 Only “low-dose” heparin and regional citrate anticoagulation have demonstrated efficacy in reducing hemodialysis-associated bleeding, and a randomized prospective trial comparing these protocols confirmed the superiority of regional citrate anticoagulation in “high”and “very high’ ‘-risk patients. ’ Several citrate anticoagulation protocols have

been devised and all appear equiefficacious. The choice of which protocol to follow should therefore be determined by availability, ease of performance, and safety. Isotonic citrate solutions have been used to prevent postdialysis hypernatremia.2 In this study isotonic ACD increased ultrafiltration requirements by 1,300% or 700 mL/hr (Table 1). While large-volume ultrafiltration is technically feasible, it is not desirable and adds complexity to the citrate anticoagulation procedure. In our dialysis unit conventional heparin anticoagulation results in a mean postdialysis serum sodium of 140 + 0.5 mmol/L. Following citrate anticoagulation the postdialysis serum sodium is 144 + 1.4 mmol/L and is not different between isotonic (0.113 mol/L) ACD and hypertonic (1.6 mol/L) TSC. Since sodium delivery is equal whether one uses hypertonic or isotonic citrate and both solutions produce equal increases in postdialysis sodium activity, use of the lessdemanding hypertonic protocol seems preferable. Similarly, early citrate anticoagulation proposals used calcium-free dialysate and intravenous calcium infusions to prevent dialyzer clotting and restore hemostasis. Later reports successfully used standard dialysate and no calcium infusion. Using standard dialysate reduces procedural

CITRATE ANTICOAGULATION

523

complexity and reduces the risks of hypocalcemia and hypercalcemia involved in the use of a calcium-free dialysate. When citrate is infused to achieve a blood citrate level of 4 to 5 mmoV L, standard dialysate containing 1.25 to 1.75 mmol/L calcium maintains calcium balance and restores hemostasis during isolated and repetitive use. IA5 This greatly simplifies regional citrate anticoagulation without incurring excess risk. Tri-sodium citrate functions as an anticoagulant because citrate anions chelate free calcium ions to lower serum calcium activity and prevent activation of calcium-dependent procoagulants. Blood normally contains approximately 0.05 mmol/L citrate, but at levels of 4 to 6 mmol/L activated clotting times are infinitely prolonged and at concentrations of 12 to 15 mmol/L blood can be stored for transfusion therapy. The parent compound of citrate, citric acid, has a plasma half-life of approximately 5 minutes and is readily metabolized to CO, and Hz0 by liver, kidney, and muscle cells. Tri-sodium citrate is converted to citric acid prior to its metabolism, and in that process each millimole of TSC NaCrtrate

+ 3H2C03 ++ Citric acid + 3NaHCO? 3H2C07 + H20 + 3C0, ++ 4H20 + 6C02

yields 3 mmol/L of sodium bicarbonate (NaHC03). Citric acid (C6H80,) is a tri-carboxylic acid with acid dissociation constants (pKas) of 3.14, 4.77, and 6.39. At the physiologic pH of 7.4 each millimole of citric acid combines with 3 mmol/L of sodium bicarbonate and is converted to its trisodium salt (TSC or NaQtrate). Anticoagulant citrate dextrose-A is 67% TSC and 33% citric acid; thus, while equimolar infusions of ACD and TSC yield equal numbers of citrate anions, the final metabolism of ACD produces 203 mm01 of bicarbonate per dialysis rather than the 320 mm01 produced by TSC. While compensatory postdialysis hypoventilation keeps the postdialysis pH the same in both dialysis groups, the 117 mm01 reduction in base delivery during ACD anticoagulation is reflected in a lower postdialysis total venous carbon dioxide content (bicarbonate). Repetitive citrate anticoagulation, espe-

cially when combined with bicarbonate dialysis, exacerbates dialysis-induced metabolic alkalosis and results in morbidity.6-9 Avoiding profound alkalosis involves reducing the dialysate HC03 to 530 mmol/L or, alternatively, compounding a mixture rich in citric acid to avoid citrate-induced alkalosis and sodium loading. In conclusion, because standard dialysate can restore calcium balance and hemostasis without a separate calcium infusion and because isotonic citrate solutions do not favorably influence dialysis sodium balance, it seems likely that the combination of a hypertonic citrate infusion and calcium-containing dialysate will minimize the complexity and risks of regional citrate anticoagulation. Regional citrate anticoagulation should be a simple process of substituting hypertonic TSC (1.6 mol/L TSC) for heparin and adjusting the infusion to maintain an arterial activated clotting time of approximately 150 to 200 seconds during continuous-flow hemodialysis. The serious, documented complications of citrate anticoagulation involve (1) continued infusion during isolated ultrafiltration with resultant “citrate intoxication,” (2) hyperaluminemia, hyperammonemia, and hypernatremia during sorbent dialysis, and (3) profound alkalosis with paresthesias, arrhythmia, and cardiac arrest.5-9’29 These complications are readily preventable if dialysis is continued even when ultrafiltration is the sole indication for therapy, if citrate use is limited to nonregenerating dialysis systems, and dialysate bicarbonate concentrations are reduced to 25 to 30 mmol/L. Citrate is relatively nontoxic and asymptomatic levels exceeding 40 mmol/L have been documented during transfusion therapy. Citrate anions, however, are ultimately converted to bicarbonate and can produce a profound metabolic alkalosis with resultant paresthesias, hypotension, and arrhythmia.4-9 This needs to be avoided, and preparation of a concentrated citrate solution with equal parts (molar) citric acid and TSC could minimize this risk. ACKNOWLEDGMENT The authors gratefully by Cytosol Laboratories, the Dialysis Unit Quality

acknowledge donation of ACD-A Braintree, MA, and the support of Assurance Committee.

REFERENCES 1. Flanigan MJ, Von Reducing the hemorrhagic

Brecht J, Freeman RM, Lim complications of hemodialysis:

VS: A

524

FLANIGAN

controlled comparison of low-dose heparin and citrate anticoagulation. Am J Kidney Dis 9:147-153, 1987 2. Pinnick RV, Wiegmann TB, Diederich DA: Regional citrate anticoagulation for hemodialysis in the patient at high risk for bleeding. N Engl J Med 308258261, 1983 3. Collart F, Wens R, Dratwa M: Anticoagulation regionale au citrate de sodium: Utilisation chronique chez le patient hemodialyse. Apropos de trois cas. Nephrologie 14: 151154, 1993 4. Janssen MJ, Huijgens PC, Bouman AA, Oe PL, van der Meulen J: Citrate anticoagulation and divalent cations in hemodialysis. Blood Purif 12308-316, 1994 5. Von Brecht J, Flanigan MJ, Lim VS, Freeman RM: Regional hemodialysis anticoagulation: Hemodialysis with hypertonic t&sodium citrate. Am J Kidney Dis 8:196-201, 1986 6. Silverstein FJ, Oster JR, Perez GO, Materson BJ, Lopez RA, Al-Reshaid K: Metabolic alkalosis induced by regional citrate hemodialysis. ASAIO Tram 35:22-25, 1989 7. Kelleher SP, Schulman G: Severe metabolic alkalosis complicating regional citrate hemodialysis. Am J Kidney Dis 9:235-236, 1987 8. Charney DI,

Salmond R: Cardiac arrest after hypertonic citrate anticoagulation for chronic hemodialysis. ASAIO Tram 36:M217-M219, 1990 9. van der Meulen J, Janssen MJ, Langendijk PN, Bouman AA, Oe PL: Citrate anticoagulation and dialysate with reduced buffer content in chronic hemodialysis. Clin Nephrol 37:36-41, 1992 10. Colton CK, Ward RA, Shaldon S: Scientific basis for assessment of biocompatibility in extracorporeal blood treatment. Nephrol Dial Transplant 9: 11-17, 1994 (suppl 2) 11. Lane DA, Bowry SK: The scientific basis for selection of measures of thrombogenicity. Nephrol Dial Transplant 9:18-28, 1994 (suppl 2) 12. Kjellstrand P, Okmark P, Odselius R, Thysell H, Riede G, Holmquist B: Adherence of blood cells to dialyzer membranes as a measure of biocompatibility. Int J Artif Organs 14:698-702, 1991 13. Gawaz MP, Mujais SK, Schmidt B, Gurland HJ: Platelet-leukocyte aggregation during hemodialysis. Kidney Int 46:489-495,

1994

14. Rosenkranz AR, Templ E, Traindl 0, Heinz1 H, Zlabinger GJ: Reactive oxygen product formation by human neutrophils as an early marker for biocompatibility of dialysis membranes. Clin Exp Immunol 98:300-305, 1994 15. David S, Tetta C, Camussi G, Canino F, Ahnici C, Rizzoli V, Cambi V: Adherence of human monocytes to haemodialysis membranes. Nephrol Dial Transplant 8: 12231227, 1993

El

AL

16. Pereira BJ, King AJ, Pouts&a DD, Strom JA, Dinarello CA: Comparison of first use and reuse of cuprophan membranes on interleukin- 1 receptor antagonist and interleukin-l beta production by blood mononuclear cells. Am J Kidney Dis 22:288-295, 1993 17. Gutierrez A, Alvestrand A, Bergstrom I, Beving H, Lantz B, Henderson LW: Biocompatibility of hemodialysis membranes: A study in healthy subjects. Blood Purif 12:95105, 1994 18. Tetta C, David S, Biancone L, Canino F, Cambi V, Camussi G: Role of platelet activating factor in hemodialysis. Kidney Int 39:S154-S157, 1993 (suppl) 19. Martin-Malo A, Velasco F, Rojas R, Castillo D, Rodriguez M, Torres A, Aljama P: Fibrinolytic activity during hemodialysis: A biocompatibility-related phenomenon. Kidney Int 41:S213-S216, 1993 (suppl) 20. Kolb G, Nolting C, Eckle I, Muller T, Lange H, Havemamr K: The role of membrane contact in hemodialysisinduced granulocyte activation. Nephron 5764-68, 1991 21. Lohr JW, Schwab SJ: Minimizing hemorrhagic complications in dialysis patients. J Am Sot Nephrol 2:961-975, 1991 22. Ward DM, Mehta RL: Extracorporeal management of acute renal failure patients at high risk of bleeding. Kidney Int 41:S237-S244, 1993 (suppl) 23. Akizawa T, Koshikawa S, Ota K, Kazama M, Mimura N, Hirasawa Y: Nafamostat mesilate: A regional anticoagulant for hemodialysis in patients at high risk for bleeding. Nephron 64:376-381, 1993 24. Kerr PG, Mattingly S, Lo A, Atkins RC: The adequacy of fragmin as a single bolus dose with reused dialyzers. Artif Organs 18:416-419, 1994 25. Eberst ME, Berkowitz LR: Hemostasis in renal disease: Pathophysiology and management. Am J Med 96:168179, 1994 26. Nakamura Y, Tomura S, Tachibana K, Chida Y, Marumo F: Enhanced fibrinolytic activity during the course of hemodialysis. Clin Nephrol 38:90-96, 1992 27. Ireland HA, Boisclair MD, Lane DA, Thompson E, Curtis JR: Hemodialysis and heparin. Alternative methods of measuring heparin and detecting activation of coagulation. Clin Nephrol 35:26-33, 1991 28. Sanders PW, Taylor H, Curtis JJ: Hemodialysis without anticoagulation. Am J Kidney Dis 5:32-35, 1985 29. Suki WN, Bonuelous RD, Yocom S, Conlin CA, Crater JE, Silas LM, Wright JA, Kelly CA: Citrate for regional anticoagulation. Effects on blood PO*, ammonia, and aluminum. ASAIO Tram 34:524-527, 1988