minimizing hemorrhagic complications indialysis...

Journal of the American Society of Nephrology 961

Minimizing Hemorrhagic Complications in DialysisPatients1

James W. Lohr2 and Steve J. Schwab

J.W. Lohr, Department of Medicine, State University ofNew York at Buffalo and Veterans Administration Med-cal Center, Buffalo, NY

S,J, Schwab, Department of Medicine, Duke UniversityMedical Center, Durham, NC

(J. Am, Soc. Nephrol. 1991: 2:961-975)

ABSTRACTRenal failure is associated with an increased mci-dence of hemorrhage from a variety of sites, partic-ularly In patients undergoing surgical procedures.The primary factors in the pathogenesis of bleedingin renal failure are platelet biochemical abnormali-ties and alterations in platelet vessel wall interac-tions. Hemodialysis improves hemostatic abnormali-

ties in uremia, but the need for heparinization during

the procedure may increase the bleeding risk. Therisk of bleeding may be minimized by using perito-neal dialysis or alternative means to routine heparin-izatlon to prevent clotting in the extracorporeal cir-culatlon during hemodialysis. These include use of

minimal heparin, prostacyclin, regional citrate anti-

coagulation, and no anticoagulation. Continuous ar-teriovenous hemodialysis may also be performedwith regional citrate anticoagulation. There are 5ev-eral nondialytic therapies that may be used to pre-vent or treat hemorrhage in renal failure patients.These include administration of cryoprecipitate, 1-deammno-8-argmnmne vasopressin, estrogens, redblood cells, and erythropoietmn. A clinical strategy tominimize bleeding complications in dialysis patients

is presented.

Key Words: Bleeding. renal failure. ant/coagulation

R ichard Bright first reported bleeding associated

with the uremic state in 1 827 (1). It is now well

established that renal failure causes hemostatic ab-

Received April 16, 1991. Accepted June 12. 1991.

2Correspondence to Dr. J.W. l,ohr, Department of Medicine, VA Medical Center

(151B), 3495 Bailey Avenue, Buffalo, NY 14215.

1046-6673/0205-0961$03.00/0Journal of the American Society of NephrologyCopyright C 1991 by the American Society of Nephrology

normalities that result in an increased incidence ofhemorrhagic complications. These complications are

common in patients with acute renal failure as well

as in those with ESRD. They vary from minor eventssuch as ecchymoses to life-threatening complications

such as pericardial tamponade, gastrointestinal hem-orrhage, and intracranial bleeding. Effective hemo-dialysis may result in correction of the hemostaticabnormalities of uremia, but bleeding episodes arestill frequent in dialyzed patients.

In this discussion, we will review the various bleed-ing problems associated with renal failure and focuson the therapeutic options to minimize hemorrhagiccomplications in high-risk patients.

HEMOSTATIC DEFECTS OF UREMIA

The pathogenesis of bleeding in renal failure is notcompletely understood but is multifactorial. We willdiscuss this topic briefly; the reader is referred tocomprehensive reviews of this topic for further de-

tails (2,3).

The primary factors in the hemostatic abnormali-

ties of uremia are platelet biochemical abnormalitiesand alterations in platelet vessel wall interactions(4). Although the platelet count in uremic patientstends to be lower than that in the normal population(5), it is generally not low enough to account forimpairment of hemostasis. Studies have revealed avariety of biochemical abnormalities that may con-tribute to platelet dysfunction in uremia. ADP andserotonin, compounds that play an important role inplatelet aggregation, have been found to be reduced

in platelets of patients with renal failure (6). In-creased levels of platelet cAMP described in uremiamay also inhibit platelet aggregation (7). Renal failure

has been found to result in abnormalities of plateletarachidonic acid metabolism. Abnormal platelet ag-

gregation has been described in response to arachi-donic acid (8). Thromboxane A2 formation has beenfound to be defective in response to endogenous aswell as exogenous stimuli (9). This may be because

of alterations in the release of arachidonic acid or toa cyclooxygenase defect.

The fact that dialysis partially corrects plateletdysfunction (1 0) has led to the suggestion that uremictoxins may cause platelet defects. Urea, creatinine,guanidosuccmnic acid, phenolic acid, and parathyroidhormone have been studied. High concentrations ofurea have been shown to result in mild platelet ag-

Minimizing Hemorrhagic Complications

962 Volume 2� Number 5’ 1991

gregation defects in normal volunteers ( 1 1 ). Guani-dosuccinic acid (1 2) and phenolic acid (1 3) have beenshown to alter platelet aggregation. High levels ofparathyroid hormone may adversely affect plateletfunction (14). Parathyroid hormone has been shownto inhibit platelet aggregation in response to variousstimuli (1 5, 1 6). However, another study found nocorrelation between secondary hyperparathyroidism(1 7) and alterations in platelet aggregation. The factthat dialysis does not normalize the bleeding time inuremia suggests that uremic toxins are not the sole

factor contributing to the thrombocytopathy.There are various factors which may cause altered

platelet-vessel wall interactions. One compound thathas been proposed to play a major role is prostaglan-

din 12 (PGI2), a prostaglandmn with vasodilatory andantiaggregatory properties (1 8). Increased PGI2 pro-duction has been found in blood vessels from uremicpatients with prolonged bleeding times (1 9). There

may also be a decrease in PGI2 metabolism in uremia(20). Alterations in factor VIII and von Willebrandfactor may be important in the bleeding tendency ofuremia (2). A recent study has shown that endothe-hal-derived relaxing factor, nitric oxide, plays a rolein the bleeding tendency of uremia (21). Administra-tion of N-monomethyl-L-arginine, an inhibitor of ni-tric oxide formation from L-argmnine, normalized

bleeding time when given to uremic rats. Nitric oxidemay interfere with hemostasis by preventing thevasoconstriction that follows vessel injury and by

inhibiting platelet adhesion to vascular endothelium.Anemia may also play a role in altering platelet

adhesion to the endothelium. An increased red cellconcentration in blood has been demonstrated toincrease platelet adhesion to aorta in an in vitro flowsystem (22). In addition, the red cell may have a rolein thrombus formation (23). Livio et al. found a sig-nificant negative correlation between bleeding timeand packed cell volume in uremic patients (24). Pro-longed bleeding times were reduced when uremicpatients were transfused with red blood cells topacked cell volumes of 30%.

Thus, the uremic state is associated with a defectin primary hemostasis. This may be compounded byheparinization during dialysis and by the use of an-

ticoagulant therapy to maintain access patency or to

treat other medical disorders.

HEMORRHAGIC COMPLICATIONS

The most common bleeding problems noted in renalfailure patients are petechial hemorrhages, bloodblisters, and ecchymoses at the site of fistula accesspuncture or temporary venous access insertion.There are more serious internal bleeding complica-tions that occur spontaneously with increased inci-dence in patients on hemodialysis.

Hemorrhagic Pericarditis

Pericarditis occurs as two distinct clinical entitiesin patients with renal failure (25-27). Pericarditismay develop in patients with acute or chronic renalfailure before the initiation of dialysis (uremic peri-

carditis) or in patients who have been maintained onchronic hemodialysis (dialysis-associated pericardi-tis).

The incidence of uremic pericarditis has declined

over the years, reflecting earlier initiation of hemo-dialysis, and studies show an incidence of 2 to 17%

(27-3 1). The incidence of symptomatic dialysis-re-lated pericarditis has remained fairly stable at 0 to1 5% (26-29). Hemorrhagic pericardial effusion may

occur as a complication of pericarditis in either ofthe above settings, although it is more frequent indialysis-related pericarditis. The percentage ofdeaths in dialysis patients due to hemorrhagic peri-

carditis is reported at 3 to 5% (32,33). It has beensuggested that heparin administration during di-

alysis may play a role, but there is no direct evidenceavailable to support this theory.

There is general agreement that uremic pericarditisis an indication for the initiation of dialysis. There isapproximately a 90% response rate to this therapy(33); however, 1 5% of patients may have recurrenceof pericarditis. Dialysis-associated pericarditis re-sponds less uniformly to intense dialysis, with reso-

lution in less than two thirds of patients (26,34). Inmany instances, surgical intervention is required

(26,34). In both types of pericarditis, a method ofavoiding systemic anticoagulation during hemodi-alysis to minimize the risks of hemopericardium isrequired.

GI Bleeding

Gastrointestinal GI bleeding occurs with greaterfrequency and is associated with a higher mortalityin patients with renal failure than in the generalpopulation (35), and upper GI hemorrhage is the sec-ond leading cause of death in acute renal failure (36).The bleeding diathesis of the uremic state (37), aswell as intermittent anticoagulation, may contribute

to the increased incidence of GI bleeding.Peptic ulcers (gastric or duodenal) are the most

common cause of upper GI bleeding (38,39), but hem-orrhagic esophagitis, gastritis, duodenitis, and gas-tric telangiectasias may also cause significant gas-trointestinal bleeding (40-44). Autopsy studies havefound an increased prevalence of peptic ulcer diseasein dialysis patients (40), whereas endoscopic studieshave revealed that the incidence is not increased inthese patients (45).

Hemorrhagic gastritis is a common cause of GIbleeding in patients with untreated renal failure (46).This is less frequent in patients treated with hemo-

Lohr and Schwab


dialysis. In a group of 60 patients on routine hemo-dialysis studied by endoscopy, gastritis was found in22%, nodular duodenitis was found in 60%, andesophagitis was found in 37% (41). Therapy withantacids to increase luminal pH has been shown toreduce the incidence of hemorrhage (47). Antacids,

H2 blockers in doses adjusted for renal failure, andsucralfate have been shown to be effective in healing

peptic ulcer disease (48,49).Angiodysplasia is a small vascular lesion of the GI

mucosa and submucosa which may cause acute orchronic GI hemorrhage (43). These can be located

throughout the GI tract but are more often found inthe upper GI tract in patients with renal failure (42-44). The most common method of diagnosis of angio-

dysplasia is by endoscopic observation. If a lesion isactively bleeding, it should be treated with endo-scopic therapy by electrocautery, heater probe, orlaser. If a patient is not found to have another sourceof bleeding, an angiodysplastic lesion may be as-sumed to be the cause of bleeding even if not activelybleeding. These lesions should be treated if the pa-tient has had recurrent or severe bleeding (43). An-ecdotal studies support the use of progesterone/estro-gen agents to minimize bleeding from these lesions

(50,51).When bleeding is slow, radiological assessment of

the upper GI tract may be performed. When bleeding

is more pronounced, rapid diagnosis with fiberopticendoscopy of the stomach and duodenum is effectivein defining the bleeding site (52).

Bleeding from the lower GI tract may also be due toa number of different etiologies. In the small intes-tine, uremic patients may develop enteritis with dii-fuse mucosal ulceration (53). This is generally re-versed by adequate dialysis. Colonic hemorrhage isnot uncommon in dialysis patients and may be dueto colitis, ischemic disease, perforation, diverticulosiswith hemorrhage, angiodysplasia, or neoplasic

(40,53,54). Uremic colitis may occur in up to 10% ofpatients. Pseudomembranous colitis may occur inpatients on dialysis or after transplantation and may

be life threatening. Ischemic bowel disease may occurwith increased frequency in patients with renal fail-ure, particularly after nephrectomy or after renaltransplantation (55). Diverticulosis may occur withincreased frequency in dialysis patients, particularlypatients with polycystic kidney disease (56). Colonichemorrhage from diverticuli may occur and may belocalized by selective angiography. Infusion of vaso-pressmn or injection of autologous clot or gelfoam intothe bleeding vessel may be effective. If not success-ful, surgical removal of the affected colonic segmentis indicated.

Ulcers of the colon (57) or rectum (58) may hemor-rhage. These occur more commonly in transplantpatients. Diagnosis is made by colonoscopy or arte-riography and generally requires colonic resection.

Although anticoagulation during hemodialysis is

not the cause of these conditions, it is a seriouscomplicating factor. When GI bleeding occurs, treat-ment of the underlying condition is mandatory. Tech-niques that correct the bleeding tendencies of uremiaand minimize mntradialytic anticoagulation are essen-tial.

Intracranial Hemorrhage

Subdural hematoma have been reported to occurinto up to 3% of patients on dialysis (59) and arefrequently fatal. As with other bleeding complica-tions, subdural hematoma were reported with morefrequency before the mid- 1 970s. Etiological factors

include anticoagulation, head trauma, hypertonic di-alysate which may cause brain shrinkage and shear-

ing of vessels, brain swelling during hemodialysis,rapid ultrafiltration for fluid removal, and hyperten-sion (60).

Subarachnoid bleeding may also occur with in-creased incidence in dialysis patients (61 ,62). It hasfrequently has been associated with anticoagulation.

The patient typically has seizures, followed by devel-opment of coma. Computed tomographic scanningcan accurately define the hemorrhage, but mortalityapproaches 100%.

Intracranial bleeding requires either conversion to

peritoneal dialysis or some method of no or regionalanticoagulation with hemodialysis. The durationafter hemorrhage that these techniques should becontinued is unclear, but the minimal duration isprobably 2 wk.

Hemorrhagic Pleural Effusion

The pleural effusion of renal failure may be hem-

orrhagic (63,64). Anticoagulation during dialysis maybe a major factor in causing bleeding in patients withfibrinous pleuritis.

Patients are admitted with pleuritic chest pain.Pleural rubs are common and may be associated with

pericardial rubs. Treatment includes more vigoroushemodialysis or peritoneal dialysis and pleurocen-tesis. The value of nonsteroidal anti-inflammatorydrugs or corticosteroids is not clear (65). Decortica-tion may be required to prevent restrictive lung de-fects in patients with recurrent hemorrhagic pleuri-

tis.

Retroperitoneal

Spontaneous retroperitoneal hemorrhage is a rarecomplication in patients on chronic hemodialysis

(66,67). Predisposing factors include trauma and an-ticoagulation. The presenting symptoms and signsinclude sudden onset of pain in the abdomen, flank,

back, or hip, with an associated drop in blood pres-


964 Volume 2 - Number 5’ 1991

sure. Subsequently, a distended abdomen with lossof bowel sounds may develop. The hematocrit dropsin the absence of any obvious blood loss.

Confirmation of retroperitoneal hemorrhage maybe made by computed tomographic scanning. Adre-

nal function should be assessed. Treatment includesblood transfusions, discontinuation of anticoagula-tion, and bed rest. Surgical exploration is rarely re-quired.

Subcapsular Liver Hematoma

Spontaneous subcapsular bleeding may occur inthe liver of patients on dialysis (68,69). This shouldbe considered in the patient with right upper quad-rant pain and decreasing hematocrit without obviousblood loss. Surgical intervention may be required forevacuation and/or hepatectomy.

Ocular Hemorrhage

Subconjunctival hemorrhage occurs with in-creased frequency in renal failure patients (70). Thismay result from heparinization during dialysis, withValsalva maneuver, or with no obvious cause. Thereis no visual loss, and the hemorrhage generally re-solves without any therapy.

Bleeding may also occur in the anterior chamber of

the eye (71). Intraocular bleeding has been found to

occur in a large percentage of transplant and dialysispatients after cataract surgery. The hemorrhage wasfound to clear, in all instances, without permanent

loss of vision.It has been assumed that anticoagulation associ-

ated with hemodialysis would encourage retinal

bleeding in diabetic patients. However, recent studieshave not supported this conclusion (72,73).

When ophthalmic bleeding complications occur, it

is prudent to avoid systemic anticoagulation. Mini-mum heparin, no heparin, or regional anticoagula-tion techniques are warranted.

Uterine Bleeding

Women developing ESRD usually experience irreg-

ular menses or amenorrhea. However, in manywomen, intermittent uterine bleeding resumes afterinitiation of dialysis with correction of metabolic ef-fects of uremia. One half of patients return to anormal cycle (74). In 10% of women, excessive vagi-nal bleeding may develop requiring gynecological in-tervention.

Most women with excessive uterine bleeding re-spond to curretage followed by administration of ex-ogenous steroids to replace hormone deficiency andstabilize endometrial stroma. Alternative approachesto uterine bleeding include progesterone-impreg-

nated intrauterine devices, intracavitary insertion ofradium, and at last resort, hysterectomy (75).

Surgical Bleeding

The uremic state might be expected to commonlycause hemorrhagic complications intraoperatively orperioperatively, but this is generally not a problem ifhemostasis is diligently maintained. There are cer-

tam clinical situations in which this may pose aproblem. Anticoagulation for dialysis will obviouslyaffect surgery. The effect of routine doses of heparinappear to last for 4 to 6 h after dialysis. An organized

strategy that includes preoperative intense dialysis,transfusion therapy, correction of any bleeding dis-order, and some method of no or regional anticoagu-lation with dialysis is required (76,77).

THERAPEUTIC STRATEGY TO REDUCE BLEEDINGCOMPLICATIONS

Classification of Patients with Increased Bleed-ing Risk

A system for classifying risk of hemorrhage indialysis patients was described by Swartz and Port

(78) and has been employed in most subsequent stud-ies of bleeding risk in dialysis patients.

Bleeding risks were classified as follows: (1 ) “very

high risk,” active bleeding at time of dialysis; (2) “highrisk,” active bleeding stopped for less than 3 days orsurgery or trauma within the previous 3 days; (3)

“moderate risk,” active bleeding stopped for morethan 3 days but less than 7 days, surgery or traumawithin the previous 3 to 7 days, or uremic pericarditis

or pleuritis; and (4) “low risk,” greater than 7 days

after active bleeding, surgery, or trauma.

Dialysis Alternatives in High-Risk Patients

Peritoneal Dialysis. In patients at increased riskof hemorrhage, peritoneal dialysis would appear to

be the dialytic method of choice. Aside from the

bleeding risk associated with the catheter insertion,this technique obviates the risk of further bleedingdue to the administration of anticoagulants.

Peritoneal dialysis has been specifically recom-mended in certain situations such as patients withhead trauma (79), patients with grossly contaminatedperitoneal cavities (80), patients with cardiac insta-bility, and those patients expected to be at increased

risk of bleeding for prolonged periods of time.

Unfortunately, in a large number of patients, theuse of peritoneal dialysis is not feasible, and the riskof peritonitis must be taken into account in each

instance. Peritoneal dialysis may be contraindicatedin patients with open abdominal wounds, with recentintestinal anastomoses, after surgery (such as aorticaneurysm repair) in which the peritoneal membrane

Lohr and Schwab


is not intact, and in hypercatabolic states. These

patients frequently fall into the high bleeding riskgroup. Thus, the development of techniques to avoid

systemic anticoagulation during hemodialysis in pa-

tients with renal failure has been necessary.Hemodlalysis. Routine hemodialysis requires sys-

temic anticoagulation to prevent clotting in the extra-

corporeal circulation. In most patients with renalfailure, this is not associated with bleeding problems.However, in patients with an increased risk of bleed-

ing, as described in the previous section, varioustechniques have been described to limit bleedingcomplications. These include regional heparin anti-

coagulation with protammne reversal, minimal hepa-rim, use of prostacyclmn, regional citrate anticoagula-tion, and hemodialysis without anticoagulation

(Table 1).Regional Heparinization. The earliest method de-

veloped to reduce dialysis-related bleeding complica-tions was regional heparlnization (81-85). A con-

stant infusion of heparin is infused into the inlet lineof the dialyzer. Simultaneously, a constant infusionof protammne sulfate is infused into the outlet portbefore the blood returns to the patient. During di-alysis, samples of blood from the patient from thedialyzer circuit (before return to the patient) are ana-lyzed for clotting times. The infusion pumps are ad-

justed to keep the whole blood-activated clotting timeof the patient nearly the same as baseline and that

of the artificial kidney at 250 s. The protammne dos-age necessary to keep the heparmn neutralized can bedetermined by a protamine titration test (81,83),which determines the lowest protamine/heparin ra-

tb.

Although this method has been employed for over25 yr. the fact that bleeding complications are re-

portedly greater than by the simpler technique ofminimal heparmnization (78) make its disadvantagesstand out: (1 ) An additional infusion pump and inletport are required in the dialysis setup to permit in-fusion of protamine. (2) It is sometimes difficult todetermine proper infusion rates of heparin to achieveextracorporeal but not systemic anticoagulation. (3)Despite a normal clotting time at the end of dialysis,

systemic rebound anticoagulation may begin 2 to 4 hafter completing dialysis and may last up to 10 h.This rebound is caused by the heparin-protaminecomplex being broken down in the reticuloendothe-hal system and heparin reentering the circulation

(86,87). This can be minimized by attempting to limitthe heparin infusion to decrease the WBACT of thereturn blood rather than by increasing the prota-mine. Decreasing the heparin infusion during the

last hour of dialysis or administration of an addi-tional small bolus of protamine 3 h after dialysis mayprevent rebound anticoagulation.

In general, regional anticoagulation with heparinfollowed by protammne reversal has been abandoned

in favor of techniques with less complexity and fewersystemic complications.

Minimal Heparinization. Another method com-monly employed to reduce bleeding complications isthe use of the minimal dose of heparin, which willmaintain anticoagulation in the extracorporeal cir-

cuit (78,88-95) (Figure 1). The modeling of the hep-arm requirements of a patient can be helpful inachieving this goal. A pharmacokmnetic model forcontrolled administration of heparin was first de-

TABLE I . Results of different anticoagulant techniques for dialyzing patients with high bleeding risk

TechniqueNo. of Dialysis

Treatments

BleedingComplications

(%)

Severe DialyzerClotting

(%)

ReferenceNo.

Regional Heparin 29 24 28 82RegionalHeparin 122 19 0 78Minimal Heparin 133 10 0Minimal Heparin 300 5 0 96Minimal Heparin 22 50 0 113Regional Citrate 23 18 0Regional Citrate 15 0 0 108RegionalCitrate 310 0 1 111Regional Citrate 326 0 1 110Minimal Heparin 65 41 6 97Prostocyclin 66 24 19NoHeparin 12 0 0 135NoHeparin 111 0 10 121NoHeparin 100 0 7 123NoHeparin 156 0 5 120No Heparin 262 0 2 122NoHeparin 208 0 6 126

Heparin

Bolua (500-1500 u)

Conatant Infuaion (250-3000 u/hr)

Blood Flow Variable

StandardDialysate

Effluent

Dtalyaate

BTriaodjumCitrate

(0.15)

2. 5-7.Of Blood Flow

Blood Flow200-300 al/am

Calciu,� Free

Dialysate

,.5 al/mm

NaCI

25-50 mlq 15 mm

J

Figure 1. CIrcuit diagrams for hemodialysis by (A) minimalheparin anticoagulation (91), (B) RCA (108), and (C) noheparin anticoagulation (122).


966 Volume 2 - Number 5 ‘ 1991

scribed by Gotch and Keen (91). This was based on afirst-order kinetic model and was further developed

for use in hemodialysis by Farrell et al. (92,93). Thepharmacokinetic model is dependent on two param-eters: the sensitivity of the patient to heparin and

the first-order elimination rate constant. These val-ues can be determined by the use of whole blood-activated clotting times or thrombin clotting timewhole blood activated partial thromboplastin time.

The use of heparmn modeling has permitted a re-duction in the empirically chosen heparin doses of

50 to 70% without causing clotting in the dialyzerand reducing the bleeding complications (93). In a

carefully controlled trial, Swartz and Port demon-strated a reduction in bleeding complications from1 9 to 1 0% by using minimal heparmnization versusregional heparinization with protamine neutraliza-

tion in patients at increased risk of bleeding (78).In a study of 300 dialyses using minimal heparmn-

ization, Swartz found 16 bleeding complications (96).In the highest risk group, bleeding occurred in 26%

of patients. In a subsequent study comparing mini-mal heparmnization to the use of prostacyclmn, ahigher incidence of bleeding complications was notedin the patients receiving minimal heparinization dur-ing dialysis (97). Without the actual pharmacokmnetic

determinations, low-dose heparinization has beenperformed more empirically by frequent measure-ments of clotting times with small variations in ratesof heparmn infusion and/or small boluses. Doses ofheparin can be reduced to ranges of 1 ,500 to 2,500

Effluent U with the use of minimal heparin.� The advantage of minimal heparinization is thatCa, Citrate no additional equipment or alterations in dialysate

are required. With careful monitoring of the arteriallimb pressure, the incidence of partial dialyzer clot-

ting is less than 5%. The major disadvantage is thatminimal heparinization may still result in systemicanticoagulation with resultant risk of bleeding.

Prostacyclin. Although heparmn inhibits gross clot-

ting of the dialyzer membranes, it may aggravate

deposition of fibrin in the membrane and subsequentplatelet aggregation. Prostacyclmn (PGI2), an arachi-donic acid metabolite, is a potent inhibitor of plateletaggregation and vasodilator with an In vtvo half-lifeof 3 to 5 mm. Prostacyclin was first used by Turneyet al. (98,99) to reduce the heparin requirements forhemodialysis in patients at increased risk of bleed-

Standard ing. Subsequently, a number of studies demonstratedDialysate that prostacyclin could replace heparin as the sole

anticoagulant in acute (97,100,101) and chronic

� :��e (97,98, 1 00, 1 02- 1 05) hemodialysis.Prostacyclmn is generally infused at 4 ngJkg/min

into the venous dialysis access before dialysis andinto the dialyzer circuit between the blood pump anddialyzer during hemodialysis. The dose of prostacy-din is then lowered if the patient experiences side

Lohr and Schwab


effects or is increased to a maximum of 8 ng/k�Jminif there is evidence of clotting in the extracorporealcircuit (97,100).

Side effects of prostacyclin infusion include head-

ache, light-headedness, facial flushing at lowerdoses, and hypotension at higher doses (97,100,102,

103). With low-dose prostacyclin (4 to 8 ng/k�Jmin),it has been possible to perform hemodialysis despitevasodilatation (97, 102). Hypotension is less frequent

if bicarbonate rather than acetate buffer is employed(1 03). A prostacyclin analog, 1 5-cyclopentyl-w-pen-tanor 5 (5)-6, 9 metano PGI2, free of hypotensiveeffects, has been used in clinical trials for dialyzeranticoagulation (106).

A recent prospective trial showed that there was a

significant decrease in bleeding complications withprostacyclin compared with minimal heparinizationfor hemodialysis in a high-risk group (97). Althoughthe residual dialyzer volume was lower after the useof prostacyclin, dialyzer efficiency as indicated bychanges in BUN and creatinine were not significantlydifferent between the two groups. Use of prostacyclininfusion prevents the drop in platelet count due toextracorporeal aggregation (1 04) and the rise inplasma free-fatty acid concentrations (1 05) found inhemodialysis with heparin anticoagulation.

The advantage of prostacyclin is that hemodialysiscan be performed without systemic anticoagulation,

and it appears to reduce the bleeding risk compared

with minimal heparinization (97).Disadvantages include expense of the agent, the

potential for clotting the dialyzer, hypotension sec-

ondary to vasodilation, and the other above-men-tioned side effects, which require close patient mon-

itoring. These problems have prevented widespreaduse of this agent.

RCA. The use of regional citrate anticoagulation(RCA) for hemodialysis was first described by Morita

et al. almost 30 yr ago (107). However, only in thelast few years since the description of this technique

by Pinnick et al. (108), has this technique beenwidely used. Anticoagulation of the hemodialysis as-sembly is achieved by the continuous infusion of

isosmotic trisodium citrate solution (1 02 mM) into

the blood withdrawn from the patient (Figure 1). Theinfusion rate is adjusted to maintain an activated

clotting time of greater than 200 s (200 to 300 mL/h). Calcium free dialysate containing acetate or bi-carbonate is circulated through the dialyzer, result-ing in the removal of calcium by complexation with

citrate and dialytic removal. Because the blood re-turning to the patient is calcium depleted, 5% calciumchloride is infused into the venous return line at arate of 0.5 mL/min.

A modification of this technique was developed inwhich a hypertonic trisodium citrate solution (1.6 M)and conventional dialysate containing 3.0 mEq/L

calcium are used (1 09). The citrate solution is admin-

istered by the pump normally used for heparin onthe dialytic assembly. This eliminates the need forhigher filtration rates, calcium free dialysate, andcalcium chloride infusion.

RCA has been successfully employed in the treat-ment of both acute (1 08, 1 1 0, 1 1 1) and chronic(1 07, 1 09, 1 1 1 - 1 1 7) renal failure. In one controlledcomparison of RCA and low-dose heparin in patients

at increased risk of bleeding, the incidence of di-alysis-associated bleeding was significantly more fre-quent in those patients receiving low-dose heparin

(50%) than in those in which RCA was used (18%)(1 1 1). In a group of patients who underwent cardi-othoracic surgery, we found a reduction in sanguin-ous chest tube drainage during the initial postopera-

tive dialysis (1 15).The use of RCA has been shown to maintain or

reduce the activated clotting time in patients duringthe course of dialysis (1 08, 1 09). The efficiency ofdialysis, as determined by changes in creatinine andurea concentrations, is comparable to that foundwith heparin anticoagulation (109).

The advantage of citrate anticoagulation is thatsystemic anticoagulation can be completely avoided.By using the hypertonic citrate technique developedby Von Brecht et al. (109), only one additional line toadd the citrate is required in the extracorporeal cir-cuit. An additional benefit is that red blood cell trans-

fusions may be given through the arterial limb duringthe dialysis procedure.

Disadvantages include the need for additional

equipment, increased ultrafiltration and potentialrisk of hypercalcemia, hypocalcemia, hypernatre-

mia, and metabolic alkalosis due to citrate intoxica-tion (1 1 8). When carefully monitored, the incidence

of complications has been very low in large trials(110,111).

Hemodialysis Without Anticoagulation. Severalgroups have evaluated various techniques for hemo-

dialysis without anticoagulation after Glaser and col-leagues were successful in performing heparin-freehemodialysis in 1979 (1 19). The technique of hemo-

dialysis without anticoagulation involves pretreat-

ment of both the blood lines and the kidney withheparin-containing solution (120-126) (Figure 1).Characteristically, 2,000 to 5,000 U of heparin in aliter of saline are flushed across the kidney and theblood lines. The lines are subsequently flushed clearwith normal saline so that no heparin enters thepatient. Blood flows are rapidly increased to 250 to300 mL/min and are kept at that level for the dura-tion of the treatment. Saline flushes of 25 to 50 mLare used to minimize hemoconcentration during ul-trafiltration across the artificial kidney and to washfibrin strands from the kidney into the bubble trap.These saline flushes are usually performed every 15

A He�arin CAVHD

Anticoagulant Replacement D.alysateHeparin solutions 1 5% D,aneal

(-�4O0 .Jhr) A & B alternating (1000 mI/h,)

i::: � L� �c)

Id)

UIt.ahltrate(effluent dialysate

plus net ultraf�ltrate)

B C,trate CAVHDD.alysate Calcium

Na 117. K 4. Mg 15. 1 mEg/tO mlC) 122.5 mEg/liter; I 40 mI/h,)

Dextrose 2 5%Anticoagulant Replacement zero alkali

4% trisodium citrate solution zero calcium I I centralI � 170 mI/h,) 0 9% saline (1000 ml/hr( �

� � � line

Arterial � � I I � � Venous

catheter �-i.I; (b) � FILTER ‘��jc’� catheter

(dl

Ultrafiltrate(effluent dialys.ate

plus net ultrafiltrete


968 Volume 2 - Number 5 . 1991

mm and as needed if fibrin stranding is noted in theartificial kidney. One-to-one nursing is usually re-quired for this technique to be successful, as the

arterial limb pressure must be closely monitored.By modifications of the technique outlined above,

three large clinical trials have been performed. Sand-ers and colleagues reviewed their retrospective ex-

perlence, predominantly in patients after and beforerenal transplantation with hollow fiber artificial kid-ney and were successful in maintaining solute clear-ances and avoiding the need for anticoagulants (120).

Schwab and coauthors performed a 1 -yr prospectivetrial predominantly in hospitalized intensive care

unit patients with acute renal failure (122). By usingparallel plate kidney, they carried out greater than

92% of their treatments without needing to resort toanticoagulants. Over 50% of the patients were unsta-ble and maintained on vasopressors during thesetreatments. Temporary venous catheters were used

in the majority. Approximately 7% of the patientsrequired conversion to minimal heparin because ofproblems with clotting in the extracorporeal circuit.Caruana and coauthors further modified these tech-niques for outpatient hemodialysis (1 23). Clotted

blood lines in artificial kidneys occurred in less than3% of the episodes, and no other significant morbid-ity was noted in these clinical trials. The effective-ness of no heparin hemodialysis in unstable patients

with significant risk of bleeding is well establishedwith less than 1 0% of the patients needing conver-

sion to mini-dose heparin to complete their treat-ments (120-126).

No heparin hemodialysis has been shown to main-tam urea nitrogen and phosphate clearances as wellas ultrafiltration when compared with anticoagu-lated controls (1 20, 1 22, 1 24, 1 25). In addition, therewere no significant differences in intradialytic hy-

poxemia or platelet degradation rates compared withanticoagulated controls. Conventional dialysis pa-

rameters such as fibrinogen and antithrombin IIIlevels, as well as bleeding times, PT, and PTT wereunchanged during these techniques (1 2 1 - 1 25). Mostimportantly. no episodes of dialysis-related bleedingwere shown in any of the above trials. Recent obser-vations from Duke University have shown no

changes in clotting rates with this technique, despitethe widespread use of erythropoietin resulting inhigher mean hematocrit in chronic hemodialysis pa-tients.

The advantages of no heparin hemodialysis are

that no anticoagulants of any type are required. Nospecific hardware or specific testing procedures arerequired. Specific disadvantages include the need for

one-to-one nursing because of close monitoring ofthe venous and arterial pressure alarms and careful

monitoring of the artificial kidney and the bubbletrap for fibrin stranding or early signs of clotting. In

addition, higher blood flows (250 to 300 mL/min) arerequired as well as a willingness to convert to low-

dose or mini-heparin therapy in approximately 4 to

5% of patients treated. An additional disadvantage isthat transfusion should not be performed duringthese treatments via the hemodialysis circuit becauseof the danger of hemoconcentration.

CAVH/CAVDH. Slow continuous renal replace-ment therapy is being used with increasing fre-

quency in critically ill patients with acute renal fail-ure (127-134). The advantages of this family of ther-apies is continuous physiological ultrafiltration andsolute removal with limited cardiovascular strain.

The technique involves the insertion of an arterialline and a venous return line. Because of the needfor maximum pressure and flow, a femoral or central

arterial catheter is optimal. The pressure gradientfrom the central arterial catheter to the venous linepushes blood through the extracorporeal circuit

through a highly permeable hemofilter where ultra-filtration and solute transfer occur. The technique istermed slow, continuous ultrafiltration, if ultrafiltra-tion is the only procedure performed. If replacementfluid is added, predilution or postdilution so that

plasma water exchange is occurring, the techniqueis considered continuous arteriovenous hemofiltra-tion (CAVH). In addition, if dialysate is dripped in

countercurrent fashion across the hemofilter as inhemodialysis, the technique is termed continuous

arteriovenous hemodialysis (CAVHD) (Figure 2). Be-

cause of the relatively slow blood flows, anticoagu-

Figure 2. Circuit diagrams for heparin CAVHD (A) and citrateCAVHD (B). Sampling ports are marked: peripheral (a).prefilter (b), posifilter (C), and ultrafiltrate (d) (from refer-ence 134 with permission).

Lohr and Schwab


lation across the artificial kidney and blood lines is

of preeminent concern. In general, the procedure hasrequired systemic anticoagulation with heparin, be-cause of the relatively slow blood flows employed andthe likelihood of hemoconcentration. The use of pre-dilution infusion fluid to minimize hemoconcentra-tion routinely allows anticoagulation intensity to bereduced (128). Regional anticoagulation with prota-mine reversal has been reported (1 32) but has similarcomplications to regional anticoagulation in hemo-dialysis and has not been widely accepted. Prostacy-

din was effective as the sole anticoagulant in post-dilutional hemofiltration but was associated withcardiac instability (1 33). In most centers, the use ofpredilution fluid and low-dose continual heparin an-ticoagulation is required for successful continuoustherapy (1 27- 1 3 1). Most centers attempt to keep ac-

tivated clotting times one to one-half times normal orPTT one and one-half times normal when sampling

from the postdilution port with heparin infused into

the first available port in the arterial line (127-131).In most instances, it is possible to avoid completesystemic anticoagulation but anticoagulation of theextracorporeal circuit is required.

Recently, Mehta and colleagues have describedRCA for CAVHD (134) (Figure 2). After using several

different techniques, these authors adopted a tech-nique that involves anticoagulation through the prox-imal arterial port with 4% trisodium citrate at ap-

proximately 1 70 mL/hr with concomitant predilutionsaline or replacement fluid infusion. The authorsremoved calcium from the dialystate because they

noted that, when calcium was present, an increaseddosage of citrate was needed to maintain proper an-ticoagulation. A central venous catheter is employedto maintain a calcium infusion to ensure adequatecalcium administration to prevent hypocalcemia andprevent citrate-induced alkalosis. Several patients

treated with this protocol required administration ofhydrochloric acid via central venous line to correcttheir alkalosis. Although the method was compli-

cated, Mehta and colleagues were able to successfully

perform CAVHD without bleeding complications inthe patients tested. Whether the side effects from

citrate yield a substantial advantage over low-dose

anticoagulation with heparin remains to be estab-lished.

Because of the low blood flow rates employed, noheparin techniques with saline flushes have beendisappointing in CAVHD (134).

Nondialytic Therapy

There are several nondialytic therapies that have

been employed to prevent or treat hemorrhage inpatients with renal failure. These include adminis-tratlon of cryoprecipitate, 1 -deamino-8-arginine va-sopressin (DDAVP), estrogens, red blood cell trans-fusions, and platelet infusion (Table 2).

Cryoprecipitate. The precipitate obtained whenplasma is frozen and thawed (cryoprecipitate) hasbeen shown to be enriched in factor VIII, fibrinogen,and fibronectin. When administered to uremic pa-tients with prolonged bleeding time, cryoprecipitateshortened the bleeding time and reduced hemorrhagein patients who underwent surgery (135). The effectof cryoprecipitate on bleeding time was apparent 1 hafter administration, maximal at 4 to 8 h, and notdetectable after 24 h. Preparation of cryoprecipitatevaries, and some preparations may be less effective.In addition, cryoprecipitate poses a risk of hepatitis

similar to that of whole blood.DDAVP. An alternative to cryoprecipitate is

DDAVP. Infusion of this synthetic analog of antidi-

uretic hormone has been shown to shorten bleedingtimes in uremic patients with hemorrhagic tenden-cies and prolonged bleeding times (136,137). DDAVP

induces an increase in factor VIII release from storagesites into the plasma (1 38). Patients with renal failurehave normal to increased levels of factor VIII activity.

Watson and Keogh found that only the factor VIIIristocetin activity increased significantly afterDDAVP administration, and levels of factor VIII co-

TABLE 2. Alternative agents found to reduce bleeding time in uremia

Agent Dose Onset of Effect Duration of Effect References No.

DDAVP 0.3-0.5 ,�g/kg iv.0.3,�g/kgs.c.3.0 pg/kg intranasal

I h 4-24 h 136-137139140

Cryoprecipitate 10-20 U iv. 1-4 h 4-24 h 135Conjugated estrogen (Pre- 2.5-25 mg orally 2-5 days 3-10 days after 141

mann)0.6 mg/kg iv. daily 6 h

discontinuation9-25 days after 51,142

2.5/1 or 5/0.75 mgdiscontinuation

50Estrogen-progesterone (En-ovid)


970 Volume 2 - Number 5’ 1991

agulant activity and factor VIII related antigen activ-ity remained unchanged (136). It was suggested that

the effect of DDAVP on bleeding times was not re-lated to alteration of factor VIII activity but might be

due to promotion of platelet aggregation.Subsequently, Mannucci et al. showed a consistent

increase in factor VIII coagulant activity, as well as

factor VIII-related antigen and ristocetin cofactor( 1 37). There was also appearance of larger multimersof factor VIII: von Willebrand factor after infusion.

The shortening of bleeding time after i.v. infusionis apparent within 1 h and lasts for 6 to 8 h. Therecommended dose for uremic patients undergoing

surgery is 0.3 to 0.4 U/kg body wt in 50 mL of salineinfused over a 30-mm period. In a recent study, s.c.DDAVP (0.3 ,�g,/kg) was found to be as effective asi.v. DDAVP in shortening the bleeding time in uremia(1 39). The use of intranasal DDAVP, effective in vonWillebrand’s disease and hemophilia, has been re-ported to be of benefit in uremia (140).

Estrogens. Conjugated estrogens have also beenshown to be effective in the treatment of uremicbleeding. The fact that bleeding time was improvedin women with von Willebrand’s disease during preg-nancy led to a trial of conjugated estrogens in uremicbleeding. Daily administration of conjugated estrogenresulted in improvement or normalization of bleedingtime after 2 to 9 days of therapy (141). A subsequentcontrolled study showed that daily administration of

conjugated estrogens for 5 days to patients on main-

tenance hemodialysis with bleeding tendencies re-sulted in a reduction in bleeding time 6 h after initialinjection, reached a maximum effect in 5 to 7 days,and lasted for 14 days (142).

Preliminary studies with estrogens have shownthat they may be effective in treating mucosal bleed-ing from GI telangiectasias, gastritis, or duodenitis

(50,51).The specific mechanisms by which estrogens re-

duce bleeding time in uremia are unclear; however,it may be due to inhibition of vascular prostacyclinsynthesis, known to be increased in uremia. The

dosage employed has been 0.6 mg/kg of conjugatedestrogen infused daily for 5 days. Potential side ef-fects of hormonal therapy include nausea and vom-

iting, loss of libido, gynecomastia, thromboembolic

disease, and risk of malignancy. Further controlledstudies are needed to show the benefits and risks ofestrogen therapy in uremia.

Red Blood Cell Transfusions and Erythropoietin.There appears to be a correlation between low hem-atocrit and prolonged bleeding time in uremic pa-tients. In vitro, it has been shown that platelet adhe-

sion to endothelium of rabbit aorta increases as the

hematocrit increases (22). This beneficial effect ofred blood cells on bleeding time has also been dem-onstrated in vivo by transfusion of washed red blood

cells (24). The bleeding time was shortened by raising

the hematocrit to 30% or above.Thus, the hemostatic defect in severely anemic

patients may benefit from red blood cell transfusionsto a hematocrit of 30%. Recombinant human eryth-ropoietin has also been found to be beneficial inreducing prolonged bleeding times in uremia by cor-

recting anemia (1 43, 1 44). By using low-dose eryth-ropoietin over 1 2 wk at a dose that only partiallycorrected anemia (hematocrit of 27 to 32), bleedingtime was normalized in a group of uremic patientswith bleeding times greater than 15 mm (144).

Platelet Transfusions. Because of the qualitative

platelet defects associated with uremia, transfusions

of normal platelets have been employed for the treat-ment of uremic bleeding or the prevention of intra-

operative and postoperative bleeding. However, it ap-

pears that normal platelets exposed to uremic plasmarapidly develop functional defects (145). Althoughoccasional reports have shown platelet transfusionsto be of value, there is no good evidence to supporttheir use in uremic bleeding.

Future Directions. Several alternative methods toprevent extracorporeal circulation clotting withoutsystemic anticoagulation have been examined.

Low-molecular-weight heparin appears to haveantithrombotic properties comparable to unfraction-

ated heparin but causes less bleeding (146-148).Low-molecular-weight heparin has been used in pre-

liminary clinical studies and may prove to reduce therisk of bleeding when compared with unfractionatedheparin.

Nafamstat mesilate, 5-amidino-2-napthyl p guani-dinobenzoate dimethanesulfonate, is a multienzymeinhibitor which has been employed as an anticoagu-lant for hemodialysis (149). This compound, like

prostacyclin, undergoes rapid metabolism with a bi-ological half-life of under 8 mm. Nafamastat inhibitsenzymes of the coagulation cascade, and its antico-agulant effect was shown to be primarily limited tothe extracorporeal circuit.

A variation of regional anticoagulation with hepa-rim has been studied in which enzymatic degradationof heparin by heparinase is employed before bloodreinfusion (1 50, 1 5 1). Clinical trials have not been

reported.A few studies have looked at the possibility of

producing tubing and dialyzers that are less prone tothrombosis by bonding heparin or other compoundsto the surface (152,153).

CLINICAL STRATEGY FOR MINIMIZING BLEED-ING COMPLICATIONS

Preoperative and high-risk patients with uremiashould have their bleeding time assessed in additionto routine measures of coagulation including PT/PTT

ACTIVE BLEEDING OR HIGH BLEEDING RISK

1) if HCT < 30, RBC transfusions

2) DOAVP, cryoprecipitate or estrogens

/\stable hemodynamics unstable tiemodynamics

/\ /Nperitoneal hemodlalysis peritoneal CAVHdialysis I) no heparin dialysis 1) RCA

2) RCA 2) ? regional heparin3) prostacyclin 3) ? prostacyclin4) minimal heparin

Lohr and Schwab


Figure 3. Guidelines for treatment in renal failure patientswith increased bleeding risk.

and platelet count (Figure 3). Abnormal bleedingtimes should be corrected by the infusion of DDAVP

or cryoprecipitate. Estrogen therapy is warranted ifthe risk of bleeding is likely to be prolonged. Whenpossible, transfusion to a hematocrit of 30 should be

performed to further minimize bleeding tendencies.In elective cases, several days of consecutive hemo-dialysis, not only to minimize uremic bleeding factors

but also to improve fluid and acid-base balance,should be undertaken preoperatively. When dialysisis required in high-risk patients, it should routinely

be performed by the no heparin technique or by oneof the methods for regional anticoagulation (citrate,prostacyclin). If long-term avoidance of anticoagula-tion is required, such as after subarachnoid hemor-rhage, conversion of these patients to peritoneal di-alysis should be considered.

The development of several proven techniques (noheparmn hemodialysis, RCA) has been shown tomarkedly reduce hemorrhagic complications. Each

hospital center should have the capacity to performone of these techniques in the group of patients athigh or very high risk of bleeding. As the risk ofbleeding lessens, conversion of patients at low ormoderate risk of hemorrhage to minimal heparini-zation for anticoagulation during hemodialysis is areasonable step. By using the above scheme, as thepatient’s risk of bleeding lessens, he or she may beconverted to routine systemic heparin anticoagula-tion for hemodialysis.

REFERENCES

1. Bright R: Tabular view of the morbid appear-ance in 100 cases connected with albuminousurine: With observations. Guy’s Hosp Rep1836; 1:380-400.

2. Remuzzi G, Pusineri F: Coagulation defects inuremia. Kidney mt i 988;33:S5l 3-S5 17.

3. Jubelirer SJ: Hemostatic abnormalities inrenal disease. Am J Kidney Dis 1985;5:2l9-225.

4. DiMinno G, Martinez J, McKean M, DeLaRosaJ, Burke J, Murphy S: Platelet dysfunction in

uremia. Multifacted defect partially correctedby dialysis. Am J Med 1 985;79:552-559.

5. Eknoyan G, Wacksman SJ, Glueck H, Will JJ: Platelet function in renal failure. N Engl JMed 1969;280:677-681.

6. Eknoyan G, Brown CH: Biochemical abnor-malities of platelets in renal failure. Evidencefor decreased platelet serotonin, adenosine di-phosphate, and Mg-dependent adenosine tn-phosphatase. Am J Nephrol 1981;1:17-23.

7. Vlachoyannis J, Schoeppe W: Adenylate cy-clase activity and cAMP content of humanplatelets in uremia. Eur J Clin Invest 1982;12:379-381.

8. Remuzzi G, Marchesi D, Livio M, et al.: Alteredplatelet and vascular prostaglandin generationin patients with renal failure and prolongedbleeding time. Thromb Res 1 978; 13:1007-1015.

9. Remuzzi G, Benigni A, Dodesini P: Reducedplatelet thromboxane formation in uremia. Evi-dence for a functional cyclo-oxygenase effect.J Clin Invest 1983;71:762-768.

10. O’Grady JA: Bleeding tendency in uremia.JAMA 1959;l69:1727-l732.

1 1 . Somer JB, Steward JH, Castakli PA: The ef-fect of urea on the aggregation of normal hu-man platelets. Thronib Diath Haemorrh 1968;19:64-69.

12. Horowitz HI, Stein IM, Cohen BD, White JG:Further studies on platelet-inhibiting effect ofquanidosuccinic acid and its role in uremicbleeding. Am J Med l970;49:336-340.

1 3. Rabiner SF, Molinas F: The role of phenol andphenolic acid on the thrombocytopathy and de-fective platelet aggregation of patients withrenal failure. Am J Med l970;49:346-351.

14. Livio M, Benigni A, Remuzzi G: Coagulationabnormalities in uremia. Semin Nephrol1 985;5:82-90.

1 5. Remuzzi G, Dodesini P, Benigni A, et ci.: Para-thyroid hormone inhibits human platelet func-tion. Lancet 198l;2:1321-1323.

16. Benigni A, Livio M, Dodesini P, et at.: Inhibi-tion of human platelet aggregation by parathy-roid hormone: Is cyclic AMP implicated? Am JNephrol 1 985;5:243-247.

1 7. Docci D, Turci F, Delvecchio C, Gollini C,Baidrati L, Pistocchi E: Lack of evidence forthe role of secondary hyperparathyroidism inthe pathogenesis of uremic thrombocytopathy.Nephron 1986;43:28-32.

1 8. Moncada 5, Vane JR: Arachadonic acid metab-olites and interactions between platelets andblood vessel walls. N Engl J Med 1979;300:1142-1149.

1 9. Remuzzi G, Cavenaghi AE, Mecca A, DonatiMB, DeGaetano G: Prostacyclin-like activityand bleeding in renal failure. Lancet1977;2:l 195-1197.

20. Defreyn G, Vergara-Dauden M, Macbin SJ: Aplasma factor in uremia which stimulate pros-tacylin release from cultured endothelial cells.Thromb Res 1980;19:695-700.

2 1 . Remuzzi G, Perico N, Zoja L, Corma D, Mac-corn D, Vigano G: Role of endothelium-derivednitric oxide in the bleeding tendency of uremia.J Clin Invest 1990;86:1768-1771.

22. Turrito VT, Weiss HJ: Red blood cells: Theirdual role in thrombus formation. Science


972 Volume 2 ‘ Number 5’ 1991

1980;207:54 1-543.23. Remuzzi A, Languino LR, Constantini V,

Guardabasso V, DeGaetano G, Dejana E:Platelet adhesion to subendothelium-Effectsof shear rate, hematocrit and platelet count onthe dynamic equilibrium between platelets ad-hering to and detaching from the surface.Thromb Haemostasis 1 985;54:857-86 1.

24. Llvio M, Gott E, Marchesi D, et al.: Uraemicbleeding: Role of anemia and beneficial effectsof red cell transfusions. Lancet 1 982;2: 1013-1015.

25. Kumar S, Lesch M: Percarditis in renal disease.Prog Cardiovasc Dis 1980;22:357-369.

26. Rutsky EA, Rostand SG: Treatment of uremicpericarditis and pericardial effusion. Am J Kid-ney Dis 1987;10:2-8.

27. Comty CM, Cohen SL, Shapiro FL: Pericarditisin chronic uremia and its sequels. Ann InternMed 1971;75:173-183.

28. Ribot S, Frankl HJ, Gielchinky I, et at.: Treat-ment of uremic pericarditis. Clin Nephrol1974;2: 127-130.

29. Young JB, Frommer P. Ayus JC, Alexander11K, Miller RR: Effect of dialysis on pericardialeffusion: A prospective echocardiographicstudy in the end-stage renal failure patient.Circulation 1 980;62: 1 1 1 -186.

30. Wray TM, Stone WJ: Uremic pericarditis: Aprospective echocardiographic and clinicalstudy. Clin Nephrol 1976;6:295-302.

3 1 . Brunner FP, Giesecke B, Hurland HJ, et at.:Combined report on regular dialysis and trans-plantation in Europe 1 974. Proc EDTA1975; 12:3.

32. Comty CM, Shapiro FL: Cardiac complicationsof regular dialysis therapy. In: Maher J, ed.Replacement of Renal Function by Dialysis.2nd Ed. Dordrecht: Kluwer Academic Publish-ens; 1983:595-608.

33. Druecke T, Le Pailleur C, Zingraff J, JungersP: Uremic cardiomyopathy and pericarditis.Adv Nephrol 1980;9:33-70.

34. Rinfrew R, Buselmaier TJ, Kjellstrand CM:Pericarditis and renal failure. Annu Rev Med1980;3 1:345-360.

35. Dinoso VP, Murthy SNS, Saris AL, et at.: Gas-tric and pancreatic dysfunction in patientswith end-stage renal disease. J Clin Gastroen-terol 1982;4:321-324.

36. Kleinknecht D, Jungers P, Chanard J, Bar-band C, Ganeval D: Uremic and nonuremiccomplications in acute renal failure. Evalua-tion of early and frequent dialysis on prognosis.Kidney Int 1972;1:190-196.

37. Carvalho ACA: Bleeding in uremia: A clinicalchallenge. N Engl J Med 1983;308:38-39.

38. Boyle JM, Johnston B: Acute upper gastroin-testinal hemorrhage in patients with chronicrenal disease. Am J Med 1983;75:409-412.

39. Shepherd AMM, Stewart WK, Wormsley KG:Peptic ulceration in chronic renal failure. Lan-cet 1973;1:1357-1359.

40. Vazirl ND, Dure-Smith B, Miller R, MerahimalMK: Pathology of gastrointestinal tract inchronic hemodialysis patients: An autopsystudy of 78 cases. Am J Gastroenterol1985;80:608-61 1.

41. Margolis DM, Saylor JL, Geisse G, et at.: Up-per gastrointestinal disease in chronic renal

failure: A prospective evaluation. Arch InternMed 1978;138:1214-1217.

42. Zuckerman GR, Cornette GL, Clouse RE,Harter HR: Upper gastrointestinal bleeding inpatients with chronic renal failure. Ann InternMed 1985;102:588-592.

43. Dave PB, Romeu J, Antonelli A, Eiser AR:Gastrointestinal telangiectasias. Arch InternMed 1984;144:1781-1�83.

44. Navab F, Masters P, Subramani R, Ortego T,Thompson CH: Angio-dysplasia in patientswith renal insufficiency. Am J Gastroenterol1989;84:1297-1301.

45. Andriulli A, Affi B, Recchia 5, Ponti V, TrioloG, Segoloni G: Patients with chronic renal fail-ure are not at risk of developing chronic pepticulcers. Clin Nephrol 1 985;23:245-248.

46. Jaffe RH, Laing DR: Changes of the digestivetractin uremia. Arch Intern Med 1934;53:851-864.

47. Hastings PR, Skiliman JJ, Bushnell LS, et at.:Antacid titration in the prevention of acutegastrointestinal bleeding. N Engl J MedT978;298: 104 1-1045.

48. Berstad A, Weberg R: Antacids in the treat-ment of gastroduodenal ulcer. Scand J Gas-troenteror i 986; 12:385-391.

49. Marks IN, Wright JP, Denyer M, Garisch JA,Lucke W: Comparison of sucralfate with ci-metidine in the short term treatment of chronicpeptic ulcers. S Air Med J 1980;57:567-572.

50. Bronner MH, Pate MB, Cunningham JT,Marsh WH: Estrogen-progesterone therapy forbleeding gastrointestinal telangiectasias inchronic renal failure. Ann Intern Med1986; 105:371-374.

5 1 . Seth 5, Geier TM: Use of conjugated estrogensto controlgastro-intestinal tractbleedingin twopatients with chronic renal failure. Clin Pharm1 988;7:906-909.

52. Boyle JM, Johnston B: Acute upper gastroin-testinal hemorrhage in patients with chronicrenal disease. Am J Med 1983;75:409-412.

53. Kolodny M, Mushlin A!, Baker WG: Intramuralsmall intestinal hematoma. Arch Intern Med1968; 121:438-445.

54. Sawyer 01, Carvin JF, Codd JE: Colorectalcomplications of renal allograft transplanta-tion. ArchSurg 1978;113:84-86.

55. Dahlberg PJ, Kisken WA, Newconer KL, Yu-tuc WR: Mesenteric ischemia in chronic di-alysis patients. Am J Nephrol 1985;5:327-328.

56. Scheff RJ, Zuckerman G, Harter H, Delmez J,Koehler R: Diverticular disease in patientswith chronic renal failure due to polycystic kid-ney disease. Ann Intern Med 1 980;92:202-204.

57. Mills B, Zuckerman G, Sicard G: Discrete colonulcers as a cause of lower gastrointestinalbleeding and performation in end stage renaldisease. Surgery 1981;89:548-552.

58. Goldberg M, Hoffman GC, Wonbolt DG: Mas-sive hemorrhage from rectal ulcers in chronicrenal failure. Ann Intern Med 1 984; 100:397.

59. Bechar M, Lakke JP, Hem GK: Subdural he-matoma during long term hemodialysis. ArchNeurol 1972;26:513-516.

60. Leonard A, Shaprlo FL: Subdural hematomain regularly hemodialyzed patients. Ann InternMed �975;82:650-658.

Lohr and Schwab


6 1 . Chachati A, Dechenee C, Gordon JP: In-creased incidence of cerebral hemorrhage mor-tality in patients with analgesic nephropathyon hemodialysis. Nephron 1 987;45: 167-168.

62. Barbour BA: Intracranial hemorrhage inchronic dialysis patients. In: Nissenson AR,Fine RN, eds. Dialysis Therapy. Philadelphia:Hanley and Belfus; 1986:202-204.

63. Berger HW, Rammohan G, Neff MS: Uremicpleural effusion: A study in 14 patients onchronic dialysis. Ann Intern Med 1975;82:362-364.

64. Galen MA, Steinberg SM, Lowrie FG: Hemor-rhagic pleural effusion in patients undergoingchronic dialysis. Ann Intern Med1975;82:359-361.

65. Maher JF: Uremic pleunitis. Am J Kidney Dis1987; 10:19-22.

66. Bhasin HK, Dana CL: Spontaneous retroperi-toneal hemorrhage in chronically hemodi-alyzed patients. Nephron 1978;22:322-327.

67. Milutinovich J, Follette WC, Schribner BH:Spontaneous retroperitoneal bleeding in pa-tients on chronic hemodialysis. Ann Intern Med1977;86: 189-192.

68. Borra 5, Kleinfeld M: Subcapsular liver he-matomas in a patient on hemodialysis. AnnIntern Med 1980;93:574-575.

69. Smetana 55, David E, Pelet D, Bar-Khayim Y:Subcapsular live hematoma in a patient onchronic hemodialysis. Nephron 1 987;45:323-324.

70. Burns CL: Opthalmologic aspects of dialysisand transplantation. In: Tilney NL, LazarusJM, eds. Surgical Care of the Renal Patient.Philadelphia: WB Saunders; 1982:170-183.

7 1 . Slusher MM, Hamilton RW: Spontaneous hy-phema during hemodialysis. N Engl J Med1975;293:561.

72. El Shahat Y, Rottembourg J, Bellio P, Gui-mont MC, Rousseie F, Jacobs C: Visual func-tion can be preserved in insulin dependentdiabetic patients treated by maintenance he-modialysis. Proc EDTA 1 980; 17:167-172.

73. Diaz-Buxo JA, Burgess WP, Greenman M,Chandler JT, Farmer CD, Walker PJ: Visualfunction in diabetic patients undergoing di-alysis: Comparison of penitoneal and hemodi-alysis. Int J Artif Organs 1984;7:257-262.

74. Rice GG: Hypermenorrhea in the young hemo-dialysis patient. Am J Obstet Gynecol1973;1 16:539-543.

75. Shirley RI, Tilney NL: Gynecologic and ob-stetnic care of dialysis and transplant patients.In: Tilney NL, Lazarus JM, eds. Surgical Careof the Patient with Renal Failure. Philadelphia:WB Saunders; 1982:163-169.

76. Brenowitz JB, Williams CD, Edwards WS: Ma-jor surgery in patients with chronic renal fail-ure. Am J Sung 1977;134:765-768.

77. Lissoos I, Goldberg B, VanBlerk PJ, MeijersAM: Surgical procedures ion patients in endstage renal failure. BrJ Urol 1973;45:359-365.

78. Swart.z RD, Port FK: Preventing hemorrhagein high-risk hemodialysis. Regional versus low-dose heparin. Kidney Int 1 979; 16:513-518.

79. Tietjen DP, Moore J, Gouge SF: Hemodialysis-associated acute subdural hematoma: Interimmanagement with peritoneal dialysis. Am JNephrol 1987;7:478-481.

80. Hau T, Ahrenholz DH, Simmons RI: Second-any bacterial peritonitis: The biologic basis oftreatment. Curr Probl Sung 1 979; 16:1-65.

8 1 . Gordon LA, Simon ER, Rukes JM, RichardsV, Perkins HA: Studies in regional hepanini-zation. N Engl J Med 1956;255:1063-1066.

82. Maher JF, Lapierre L, Schreiner GE, GeigerM, Westervelt JB: Regional heparinization forhemodialysis. N Eng[ J Med 1963;268:451-456.

83. Lindholm DD, Murray JS: A simplified methodof regional hepaninization during hemodialysisaccording to a predetermined dosage formula.Trans Am Soc Artiflntern Organs 1 964; 10:92-97.

84. Spencer P, Cozzi E, Easterling RE, Penner JA:Regional hepaninization with the hollow fiberartificial kidney. Proc Am Assoc NephrolNurses Technicians 1977;4:69-72.

85. Congdon JE, Kardinal CG, Wallin JD: Moni-toring hepanin therapy in hemodialysis. JAMA1973;226: 1529-1533.

86. Blaufox MD, Hampers CL, Merrill JP: Heparin-ization for hemodialysis. Trans Am Soc ArtifIntern Organs 1966;12:207-209.

87. Hampers CL, Claufox MD, Merrill JP: Antico-agulation rebound after hemodialysis. N Engl JMed 1966;275:776-778.

88. Lee JH, Cocke TB, Conzalez PM: Regional vs.tight hepaninization in hemodialysis. Trans AmSoc Artif Intern Organs 1974;22:239-246.

89. Vogel GE, Kopp K?: Minimal dose hepaniniza-tion for hemodialysis with high bleeding risk.Kidney Int 1975;8:436.

90. Vogel GE, Kopp KF: The conflict between an-ticoagulation and hemostasis during hemodi-alysis. Int J Artif Organs 1978; 1:181-186.

9 1 . Gotch FA, Keen ML: Precise control of minimalheparinization for high bleeding risk hemodi-alysis. Trans Am Soc Artif Intern Organs1977;23: 168-175.

92. Farrell PC, Ward PA, Schindhelm K, GotchFA: Precise anticoagulation for routine hemo-dialysis. J Lab Clin Med 1978;92:164-176.

93. Farrell PC: Heparin modeling. In: NissensonAR, Fine RN, eds. Dialysis Therapy. Philadel-phia: Hanley and Belfus; 1986:69-71.

94. Williman P, Aliga J, Biswanger V: Minimalintermittent heparinization during hemodi-alysis. Nephron 1979;23:l91-193.

95. Shapiro WB, Faubert PF, Porush JG, Chou 5:Low dose heparin in routine hemodialysis mon-itored by activated partial thromboplastin time.Artif Organs 1979;3:73-77.

96. Swartz RD: Hemorrhage during high-risk he-modialysis using controlled heparinization.Nephron 1981 ;28:65-69.

97. Swartz RD, Flamenbaum W, Dubrow A, HallJC, Crow JW, Cato A: Epoprostenol (PGI2 pros-tacyclin) during high risk hemodiaysis-pre-venting furtherbleeding complications. J ClinPharmacol 1988;28:818-825.

98. Turney JH, Fewell MR, Williams LC, ParsonsV, Weston MJ: Platelet protection and heparinsparing with prostacyclin during regular di-alysis therapy. Lancet 1 980;2:2 19-222.

99. Turney JH, Woods HF, Weston MJ: The use ofprostacyclin in extraconporeal circuits. In: Re-muzzi G, Mecca G, de Gaetano G, eds. Hemosta-sis, Prostaglandins, and Renal Disease. New


974 Volume 2 ‘ Number 5 - 1991

York; Raven Press; 1980:353-361.1 00. Zussman RM, Rubin RH, Cato AE, Cocchetto

DM, Crow JW, Tolkoff-Rubin N: Hemodialysisusing prostacyclin instead of heparin as thesole antithrombotic agent. N Engl J Med1981 ;304:934-939.

1 0 1 . Holy B, Saint-Miller Y, Perol JC: Prostacyclinas the sole anti-thrombotic agent for acuterenal failure hemodialysis. Nephron 1983;33:71-78.

102. Dubrow A, Flamenbaum W, Mittxnan N, HallJ, Zinn T: Safety and efficacy of epoprostenal(P012) versus heparin in hemodialysis. TransAm Soc Artif Intern Organs 1984;30:52-54.

103. SmIth MC, Danviriyasup K, Crow JW, et at.:Prostacyclin substitution for heparin in long-term hemodialysis. Am J Med 1982;73:669-678.

104. Addonizio VP, Fisher CA, Bowen JC, Palati-anos GC, Colman RW, Edmunds LH: Prosta-cyclin in lieu of anticoagulation with heparinfor extracorporeal circuration. Trans Am SocArtif Intern Organs 198 1 ;27:304-307.

105. DaviesSJ, HobsonSM, YoungGA, TurneyJH:Acute changes in plasma free fatty acid con-centration during hemodialysis comparing hep-arm with epoprostenol as anticoagulant. ProcEDTA-ERA 1 985;22:32 1-324.

106. Ota K, Kawaguchi H, Takahashi K, Ito K: Anew prostacyclin analogue-an anticoagulantapplicable to hemodialysis. Trans Am Soc As-UIIntern Organs 1983;29:419-424.

1 07. Morita Y, Johnson RW, Dorn RE, Hall DS:Regional anticoagulation during hemodialysisusing citrate. Am J Med Sci 1961 ;242:32-42.

108. Pinnlck E, Wiegmanson TE, Diederick DA:Regional citrate anti-coagulation for hemodi-alysis in the patient at high risk for bleeding.N Engl J Med 1983;308:258-261.

1 09. Von Brecht JH, Flanigan MJ, Freeman RM, etat.: Regional anticoagulation-Hemodialysiswith hypertonic trisodium citrate. Am J KidneyDis 1986;8: 196-201.

1 10. Lohr JW, Slusher S, Diederich D: Safety ofregional citrate hemodialysis in acute renalfailure. Am J Kidney Dis 1989;13:104-107.

1 1 1 . Collart F, Tielemans C, Wens R, Dratwa M:Local experience of regional anticoagulationwith sodium citrate for hemodialysis in pa-tients at risk of bleeding. Proc EDTA-ERA1 985;22:325-328.

1 1 2. Seaton RD, Duncan KA, Pinnick RV, Dieder-Ich DA, Wlegmann TB: Regional citrate anti-coagulation in chronic hemodialysis patients.Trans Am Soc Artif Intern Organs 1983;29:4 14-4 17.

1 1 3. Flanigan MJ, Von Brecht J, Freeman RM, UrnVS: Reducing the hemorrhagic complicationsof dialysis: A controlled comparison of low-doseheparin and citrate anticoagulation. Am J Kid-ney Dis 1987;9:147-153.

1 1 4. Wiegmann TE, MacDougall ML, Diederich DA:Long term comparison of citrate and heparinas anticoagulants for hemodialysis. Am J Kid-ney Dis 1987;9:430-435.

1 15. Lohr JW, Slusher 5, Diederich DA: Regionalcitrate anti-coagulation for hemodialysis fol-lowing cardiovascular surgery. Am J Nephrol1 988;8:368-373.

1 1 6. MacDougall M, Seaton R, Diederich D, Wieg-

mann T: Effects of both composition and dialy-zer membrane on leukocytes and oxygenationduring hemodialysis with citrate anticoagula-tion. Kidneylnt 1984;25:188.

1 1 7. Hocken AG, Hurst PL: Citrate regional antico-agulation in haemodialysis. Nephron 1987;46:7-10.

1 18. Kelleher SP, Schulman G: Severe metabolicalkalosis complicating regional citrate hemo-dialysis. Am J Kidney Dis 1987;9:235-236.

1 1 9. Glaser P, Guesde R, Rouby JJ, Em-in B: He-modialysis without heparin is possible. Lancet1 980;2:579-580.

120. Sanders PW, Taylor H, Curtis JJ: Hemodi-alysis without anticoagulation. Am J KidneyDis 1985;5:32-35.

121. Casati 5, Moia M, Grazianal G, et at.: Hemo-dialysis without anticoagulants: Efficiencyand hemostatic aspects. Clin Nephrol1984;21:102-105.

1 22. Schwab SJ, Omorato JJ, Sharar LR, DennisPA: Hemodialysis without anticoagulation. Oneyear prospective trial in hospitalized patientsat risk for bleeding. Am J Med 1987;83:405-410.

1 23. Caruana RJ, Raja RM, Bush JV, Krammer MS,Goldsmith SJ: Heparin free dialysis: Compar-ative data and results in high risk patients.Kidneylnt 1987;31:1351-1355.

124. Casati 5, Graziani G, Ponticelli P: Hemodi-alysis without anticoagulants in patients withhigh bleeding risk. Int J Artif Organs1 982;5:233-2�6.

125. Ivanovich P, Xu CG, Kwaan HC, Hatbiwala 5:Studies of coagulation and platelet functionsin heparin-free hemodialysis. Nephron1983;33:1 16-120.

1 26. Preuschof L, Keller F, Seeman J, Offerman G:Heparmn free hemodialysis with prophylacticchange of dialyzer and blood line. Int J ArtifOrgans 1988;1 1:255-258.

127. Geronemus R, Schneider N: Continuous arte-riovenous hemodialysis: A new modality for thetreatment of acute renal failure. Trans Am SocArtif Intern Organs 1984;30:610-612.

128. Golper TA, Ronco C, Kaplan A: Continuousarteriovenous hemofiltration: Improvements,modifications and future directions. Semin Dial1988; 1:50-54.

1 29. Kaplan A, Longnecker R, Folkert VW: Contin-uous arteriovenous hemofiltration. Ann InternMed 1984; 100:358-367.

130. Lauer A, Sacaggi A, Ronco A, et at.: Contin-uous arteriovenous hemofiltration in criticallyill patients. Ann Intern Med 1983;99:455-460.

1 3 1 . Mault J, Dechert R, Lees P, et at.: Continuousarteriovenous filtration: An effective treatmentfor surgical ARF. Surgery 1987;10:478-484.

132. Kaplan A, Petrillo R: Regional heparinizationfor continuous arteriovenous-hemofiltration(CAVH). Trans Am Soc Artif Intern Organs1987; 13:312-315.

1 33. Canaud B, Mion C, Arujo A, et at.: Prostacyclin(Epoprostenol) as the sole antithrombotic agentin postdilutional hemofiltration. Nephron1988;48:206-2 12.

134. Mehta R, McDonald B, Aguilar M, Word D:Regional citrate anticoagulation for continuousarteriovenous: Hemodialysis in critically ill pa-tients. Kidney Int 1990;38:976-981.

Lohr and Schwab


135. Janson PA, Jubelirer SJ, Weinstein MJ, Dey-kin D: Treatment of the bleeding tendency inuremia with cryoprecipitate. N Engl J Med1980;303: 1318-1322.

1 36. Watson AJS, Keogh JAB: Effect of 1 -deamino-8-D-arginine vasopressin on the prolongedbleeding time in chronic renal failure. Nephron1 982;32:49-52.

137. Mannucci PM, Remuzzi G, Pusineri F, et at.:Deamino-8-D-arginine vasopressin shortensthe bleeding time in uremia. N Engl J Med1983;308:8- 12.

1 38. Manucci PM, Pareti FF, Ruggeri ZM, CapitianoA: 1 -deamino-8-D-arginine vasopression: Anew pharmacological approach to the manage-ment of haemophilia and von Willebrands ths-ease. Lancet 1977;1:869-872.

1 39. Vigano GL, Mannuci PM, Lattuada A, HarrisA, Remuzzi G: Subcutaneous desmopressin(DDAVP) shortens the bleeding time in uremia.Am J Hematol 1989;31:32-35.

140. Shapiro MD, Kelleher SP: Intranasal deamino-8-DAVP shortens the bleeding time in uremia.Am J Nephrol 1 984;4:260-2&1.

1 4 1 . Liu YK, Kosfeld RE, Marcus SO: Treatment ofuremic bleeding with conjugated oestrogen.Lancet 1 984;2:887-890.

142. Livio M, Mannucci PM, Vigano G, et at.: Con-jugated estrogens for the management of bleed-ing associated with renal failure. N Engl J Med1986;315:731-735.

143. Casati 5, Passerini P, Graziani G, et at.: Hu-man recombinant erythropoietin corrects ane-mia and bleeding tendency in hemodialysis pa-tients. Abstr Int Congr Nephrol 1 987; 10:129.

144. Vigano G, Benigni A, Mecca G: Recombinanthuman enythropoietin (r-HuEPO) to correcturemic bleeding. J Am Soc Neph 1990;! :408.

145. Gardner F!!: Preservation and clinical use of

platelets. In: Williams WJ, Beutler E, ErsleveAJ, Rundles RW, eds. Hematology. 2nd Ed. NewYork: McGraw Hill; 1977:1553-1560.

1 46. Anastassiades E, Lane DA, Ireland H, FlynnA, Curtis JR: A low molecular weight heparin(fragmin) for routine hemodialysis: A crossovertrial comparing three dose regimens with astandard regimen of commercial unfraction-ated hepanin. Clin Nephrol 1989;32:290-296.

1 47. Anastassiades E, Lane DA, Flynn A, CurtisJR: Preliminary evaluation of repeated use ofa low molecular weight heparin in haemodi-alysis for chronic renal failure. Proc EDTA-ERA 1985;22:329-333.

1 48. Goudable C, That HT, Damani RS, et at.: Lowmolecular weight heparin half-life is prolongedin hemodialyzed patients. Thromb Res1986;43: 1-5.

149. Akizawa T, Sato M, Koshikawa S, et at.: Theusefulness of the multi-enzyme inhibitor, na-famstat mesilate, in high bleeding risk hae-modialysis. Proc EDTA-ERA 1985;22:334-338.

150. Langer R, Linhardt RJ, Hoffberg 5, et aL: Anenzymatic system to removing heparin in cx-traconporeal therapy. Science 1 982;2 17:261-263.

1 5 1 . Langer R, Bernstein H, Larsen AK: An enzy-matic approach to anticoagulation control.ASAIO J 1985;8:213-215.

1 52. Schmer G, Teng LNL, Cole JJ, Vizzo JE, Fran-cisco MM, Schribner BH: Successful use of atotally heparin grafted hemodialysis system insheep. Trans Am Soc Artif Int Organs1 976;22:654-662.

153. Tolkoff-Rubin NE, Narini J, Fang LST, RubinRH: Successful hemodialysis of patients at highrisk of hemorrhage using the Ex-Val dialyzer.Dial Transplant 1986;15:125-129.

minimizing hemorrhagic complications indialysis...

Documents