COMPLICATIONS OF DIALYSIS - PART 9 doc

COMPLICATIONS OF PERITONEAL DIALYSIS IN DIABETIC PATIENTS Eleven percent of diabetics entering renal replacement programs in the United States are treated with some form of peritoneal di

700 Miles and Friedman

predisposition of diabetic patients to access thrombosis

in earlier reports

F Vascular Steal Syndromes

An arteriovenous access (particularly a brachiocephalic

access or a side-to-side radiocephalic fistula) creates a

low-pressure run-off system, which may short-circuit

blood from the palmar arch and ulnar arteries to such

a degree that a steal syndrome results With the

pre-existing, severe medial arterial calcinosis of the ulnar

and digital arteries, which is common in diabetic

di-alysis patients (19), progressive ischemic pain leading

to dry gangrene of one or more fingers may develop

days to weeks after placement of the access

Nonheal-ing wounds of the fingers may also be a manifestation

of vascular steal (20), and in cases of clinical

uncer-tainty, digital pressures of <50 mmHg on noninvasive

vascular studies and arteriography help to confirm the

diagnosis In severe cases of arterial steal syndromes,

the onset may be more acute, within hours of creation

of the access, and signs of acute arterial insufficiency

such as pallor and pulselessness may be seen Ligation

of the distal limb of the radial artery in a side-to-side

radiocephalic fistula or ligation or removal of a

bra-chiocephalic access is necessary to correct the

syn-drome Amputation of one or more digits and even

be-low elbow amputation may sometimes be necessary

G Ischemic Monomelic Neuropathy

The term ischemic monomelic neuropathy was coined

in 1983 by Wilbourn (21) It is a complication of

vas-cular access seen almost exclusively in diabetic patients

(22) and refers to the development of acute pain,

weak-ness, and paralysis of the muscles of the forearm and

hand, often with sensory loss, developing immediately

after placement of an arteriovenous access, usually in

the brachiocephalic or antecubital location The

con-dition results from diversion of the blood supply to the

nerves of the forearm and hand, the ischemic insult

being severe enough to damage nerve fibers but

insuf-ficient to produce necrosis of other tissues Hence,

un-like in vascular steal syndromes, the radial pulse is

usu-ally present, digital pressures normal or only mildly

decreased, and necrosis, ulcers, and gangrene of the

digits are absent The propensity to development of the

complication with brachiocephalic accesses relates to

the fact that the brachial artery constitutes the sole

ar-terial inflow to the forearm and hand and, in the

ab-sence of collateral vessels about the elbow, diversion

of all or most brachial arterial blood through a fistula

or graft results in distal ischemia Nerve conductionstudies are helpful in diagnosing the syndrome, andearly access removal or ligation is necessary to preventpermanent paralysis of the hand Unfortunately, evenwith prompt access closure, paralysis of the hand may

be permanent

H Venous Hypertension

Chronic swelling of the hand, and especially of thethumb (‘‘sore thumb’’ syndrome), related to the pres-ence of the distal segment of the vein used for creation

of the access, may occur in both diabetic and abetic patients Venous hypertension occurs in associ-ation with venous stenosis of the access or a moreproximal stenosis at the level of the subclavian vein,which may have been previously catheterized for tem-porary vascular access Ligature of the distal venouslimb of the fistula or graft will usually correct theproblem

nondi-III BONE DISEASE

Adynamic bone disease is a form of renal phy commonly seen in diabetics, particularly those onperitoneal dialysis (23) It is characterized by low rates

osteodystro-of bone turnover without excess unmineralized osteoidand is associated with parathyroid hormone levels be-low 100 pg/mL Decreased osteoblast proliferation anddefective mineralization contribute to a low rate ofbone formation in diabetic rats (24), and a similarmechanism may underlie adynamic bone disease in hu-mans Diabetic dialysis patients also tend to experiencehigher rates of low-turnover bone disease associatedwith aluminum deposition: reduced bone formationmay allow time for enhanced deposition of aluminum

on the ossification front, and within 1 year of dialysis, aluminum deposition (usually related to use ofaluminum-containing phosphate binders) is observed

hemo-on bhemo-one surfaces in diabetics, and symptoms of bhemo-onepain and fractures related to aluminum bone diseasemay start as early as 2 years after initiation of hemo-dialysis (25) Aluminum bone disease may also be un-masked or accelerated after parathyroidectomy Alu-minum-containing phosphate binders should therefore

be avoided in diabetics, and all diabetics with bonepain and/or fractures should have plasma aluminumlevels measured before and after a single infusion ofdesferrioxamine Aluminum-associated bone disease in

hemodialyzed diabetics responds to a regimen of

vita-min D, calcium, and desferrioxavita-mine

IV DIABETIC RETINOPATHY

Visual loss in diabetic ESRD patients is most

com-monly related to proliferative retinopathy with

associ-ated vitreous hemorrhage and retinal detachment but

may also result from macular edema, glaucoma,

cata-racts, and corneal disease The presence of proliferative

retinopathy is correlated with age of onset and duration

of diabetes, glycemic control, and degree of blood

pres-sure control Heparin use during hemodialysis is no

longer considered a significant contributor to

progres-sion of diabetic retinopathy or to intraocular

hemor-rhage No reports definitively link heparin use on

di-alysis with progression of diabetic retinopathy or visual

loss Indeed, in a study of 112 diabetics followed for

20 months, progression of retinopathy was shown to be

independent of dialysis modality (hemo- or peritoneal

dialysis), while the significant correlation between

blood pressure control and vision preservation was

re-inforced (26) In addition, because of the rarity (0.05%)

of intraocular hemorrhage in diabetics treated with

thrombolytic agents, diabetic retinopathy is not

consid-ered a contraindication to thrombolytic therapy for

acute myocardial infarction (27) Focal or panretinal

laser photocoagulation can reduce the incidence of

se-rious visual loss in patients with proliferative

retinop-athy, and vitrectomy may restore vision in patients with

vitreous hemorrhage

V UNDERNUTRITION

Malnutrition is frequently seen in diabetic hemodialysis

patients, particularly in the presence of intercurrent

ill-nesses Causes of malnutrition in diabetics on

hemo-dialysis include (a) poor glycemic control leading to

gluconeogenesis and catabolism of muscle, (b)

gastro-paresis leading to nausea and vomiting, (c) diabetic

di-arrhea, and (d) underdialysis related to difficulties with

vascular access or to repeated early termination of

di-alysis sessions caused by recurrent hypotension (28) A

diet of 25–30 kcal/kg/day, with 50% of the calories

coming from complex carbohydrates, and protein

con-tent of 1.3–1.5 g/kg/day is recommended for

hemodi-alyzed diabetics In diabetic hemodialysis patients who

develop intercurrent illnesses (e.g., sepsis), early and

intensive nutritional support with enteral or peripheral

parenteral nutrition is necessary Dialysate fluid should

contain at least 200 mg/dL glucose because use of cose-free dialysate results in rapid glucose loss, hypo-glycemia, and production of acute starvation with aci-dosis and hyperkalemia (29)

glu-VI HYPERGLYCEMIA

Insulin requirements after beginning maintenance modialysis vary (30), and it is important to teach pa-tients home glucose monitoring so as to determinechanging insulin requirements Most diabetic patientswith ESRD experience reduction in insulin needs due

he-to decreased renal excretion and catabolism of injectedand endogenous insulin Many diabetics who start di-alysis will no longer need insulin, and some type 2diabetics previously on insulin may achieve glycemiccontrol with a small dose of a short-acting sulfonylureadrug such as glyburide or glipizide, or indeed with nohypoglycemic medications at all The new oral hypo-glycemic agent troglitazone has been approved for use

in type 2 diabetics to decrease insulin requirements orfor use alone or in combination with a sulfonylurea toprevent need for insulin therapy The drug works byincreasing the sensitivity of peripheral tissues to insulinand decreasing hepatic glucose production It under-goes hepatic metabolism and hence does not requiredose reduction in renal failure It is effective in only50% of treated patients, however, and produces mildliver injury in 1.9% of patients and sporadic cases offulminant hepatic failure requiring liver transplantation.Liver function tests should be monitored at the start oftherapy with troglitazone and at monthly intervals forthe first 6 months of therapy, and then every 2 monthsfor the remainder of the first year Rosiglitazone ap-pears to be a safer option with equivalent efficacy.With control of uremia by dialysis, improved appe-tite and weight gain in some diabetics may result inincreased insulin needs In addition, noncompliancewith hypoglycemic medications related to depression

in the often stormy period surrounding dialysis tion may contribute to hyperglycemia Hyperosmolarcoma is uncommon in diabetics on dialysis unless there

initia-is significant residual renal function Ketoacidosinitia-is initia-isalso not frequently seen but may occur in associationwith sepsis or other severe intercurrent illness Bothconditions are managed by low-dose hourly regular in-sulin infusions for blood sugar levels above 450–500mg/dL It is usually possible to achieve rapid control

of blood sugar, and indeed in some patients an initialsingle dose of 10 units of regular insulin may suffice

702 Miles and Friedman

if the patient is not seriously ill The usual fluid

re-plenishment regimens for nonuremic diabetics are of

course not required unless the patient is in shock

Usu-ally, fluid replacement during hemodialysis suffices

VII COMPLICATIONS OF PERITONEAL

DIALYSIS IN DIABETIC PATIENTS

Eleven percent of diabetics entering renal replacement

programs in the United States are treated with some

form of peritoneal dialysis (2) In other countries such

as Canada, the percentage of diabetic ESRD patients

treated by peritoneal dialysis (PD) is much higher,

in-deed, PD is the treatment of choice for diabetics with

ESRD in these countries Physician bias, national

re-sources, patient preference, and the presence of severe

cardiovascular disease are the major factors that

deter-mine the selection of PD over hemodialysis in

diabet-ics The more gentle ultrafiltration afforded by CAPD

will prevent or ameliorate hypotension in diabetics so

prone because of left ventricular dysfunction or

auto-nomic neuropathy

Diabetic hemodialysis patients younger than 60

years have a similar or lower relative risk of death than

diabetics of the same age on CAPD (31,32) In diabetic

CAPD patients older than 60 years, there is a 19%

higher relative risk of death compared with diabetic

hemodialysis patients (26) The higher death rate in

el-derly diabetics treated with CAPD is related to

ad-vanced atherosclerotic cardiovascular and

cerebrovas-cular disease Diabetics are subject to a similar

spectrum and rate of technique-related complications as

in nondiabetics on CAPD, and a discussion of some of

the most common problems of diabetics on PD follows

A Peritonitis

Within the first year of starting PD, 49% of patients

will switch to another modality of renal replacement

therapy (33), while only 37% of hemodialysis patients

change treatment modality during the first year It is

more likely that a CAPD or continuous cyclic PD

(CCPD) patient will switch to hemodialysis (15.6%)

than that a hemodialysis patient will switch to CAPD

(4.4%) (34) The high technique failure rate on CAPD

is due mainly to peritonitis (35) Recurrent peritonitis,

usually caused by Staphylococcus epidermidis or

Staphylococcus aureus, often in association with exit

site infections, is the major disadvantage of CAPD

Peritonitis in diabetic CAPD patients occurs at a rate

of one episode per 11–21 patients per month(26,29,36) Diabetics on CAPD need twice the number

of hospitalization days as nondiabetic CAPD patients(37), and peritonitis accounts for 30–50% of the hos-pitalization days (38) Fungal peritonitis is seen morecommonly in diabetic than in nondiabetic CAPD pa-tients and usually requires removal of the peritonealcatheter There is, however, no overall increased risk

of peritonitis in diabetics over nondiabetics (39), andthe rates of catheter replacement are the same in bothgroups

Because of delayed wound healing, dialysate age and exit site infections may occur in diabetics if anewly implanted Tenchoff catheter is used too early;and we recommend that, if possible, at least 2–3 weekselapse before starting regular CAPD exchanges through

leak-a newly implleak-anted Tenchoff cleak-atheter in leak-a dileak-abetic ternatively, starting nighttime cycling exchanges 1–2weeks after catheter insertion and leaving the abdomendry during the daytime to reduce intra-abdominal pres-sure during ambulation may reduce the risk of dialysateleakage The Moncrief-Popovich peritoneal catheter is

Al-a recently introduced double-cuffed cAl-atheter with Al-anexternal cuff, which is longer than that of the standardTenchoff catheter, and whose external segment (exter-nal cuff and tubing) is buried subcutaneously for 4–6weeks before being externalized and used (40) Be-cause tissue ingrowth into the external cuff has oc-curred during the period of subcutaneous implantation,

it is anticipated that the rates of leaks and peritonitiswill be less with this catheter, and indeed we have hadexcellent results, with no dialysate leaks so far withthis catheter

B Underdialysis

Patients on CAPD tend to have higher levels of bloodurea nitrogen and creatinine than hemodialysis patients,and there is concern about the adequacy of CAPD aslong-term uremia therapy (41) Based on peritonealequilibration testing, a minimum clearance of 6–7 L/day is recommended for patients on peritoneal dialysis.The amount of dialysis delivered on CAPD may have

to be increased with time as residual renal function islost, and peritoneal clearance decreases owing to ad-vanced vascular disease or recurrent episodes of peri-tonitis Microangiopathy and increased vascular per-meability to small and large molecules and resultantincreased diffusive transport of glucose (42) may pro-duce type I ultrafiltration failure in some diabeticpatients

C Undernutrition

Malnutrition may occur in up to 40% of long-term

CAPD patients (43,44) Malnutrition in CAPD-treated

diabetics may occur because of (a) reduced appetite

caused by the large glucose load in dialysate or by

early satiety from increased intra-abdominal pressure

or (b) large protein losses (8–10 g/day) in the dialysate

effluent that may lower serum albumin and total protein

levels (45) Loss of protein through the peritoneal

membrane is increased during episodes of peritonitis

and may worsen with time because of a generalized

increase in permeability related to diabetic

microangio-pathy involving the peritoneal vessels or nausea and

vomiting related to diabetic gastroparesis To maintain

adequate nutrition, CAPD patients should ingest at

least 1.5 g protein/kg/day and between 130 and 150 g

carbohydrates/day

D Hyperglycemia

Blood glucose control may sometimes be difficult in

diabetics on CAPD because of the absorption of a mean

of 182 ⫾ 61 g glucose/day from the peritoneal cavity

(46) A combination of subcutaneous and

intraperito-neal insulin administration usually results in adequate

glucose control, however The total dose of

intraperi-toneal regular insulin required is usually two to three

times the usual subcutaneous total dose Oral

antidi-abetic agents (sulfonylureas with short half-lives and

hepatic metabolism, e.g., glipizide, glyburide) to reduce

the risk of prolonged hypoglycemia may be used in

patients requiring less than 20–25 units of insulin per

day

VIII SUMMARY

Diabetic dialysis patients require extra effort on the part

of nephrologists to prevent and treat macro- and

mi-crovascular disease, to manage intradialytic

complica-tions, and to achieve the usually difficult goal of

main-taining good vascular access A multidisciplinary team

approach is required, and the services of an

experi-enced vascular access surgeon are invaluable As the

epidemic of type 2 diabetes mellitus afflicting

devel-oped countries continues and diabetics comprise an

in-creasing percentage of incident and prevalent ESRD

patients, nephrologists must target the problems

pecu-liar to, or prevalent in the diabetic dialysis population

in order to reduce morbidity and mortality in this

high-risk group

REFERENCES

1 U.S Renal Data System, USRDS 1997 Annual DataReport Bethesda, MD: National Institutes of Health,National Institute of Diabetes and Digestive and KidneyDiseases, 1991

2 Miles AMV, Friedman EA Managing co-morbid orders in the uremic diabetic patient Sem Dial 1997;10:225–230

dis-3 Shideman JR, Buselmeier TJ, Kjellstrand CM dialysis in diabetics Arch Intern Med 1976; 136:1126–1130

Hemo-4 Collins AL, Liao A, Umen A, Hanson G, Keshaviah P.Diabetic hemodialysis patients treated with a high Kt/

V have a lower risk of death than standard Kt/V J AmSoc Nephrol 1991; 2:318

5 Nakamoto M The mechanism of intradialytic sion in diabetic patients Nippon Jinzo Gakkai Shi Jap

hypoten-J Nephrol 1994; 36:374–381

6 Ritz E, Strumpf C, Katz F, et al Hypertension and diovascular risk factors in hemodialyzed diabetic pa-tients Hypertension 1985; 7(suppl II):118–124

car-7 Daugirdas JT Dialysis hypotension: A hemodynamicanalysis Kidney Int 1991; 39:223–246

8 Gotch FA, Keen ML, Yarian SR An analysis of thermalregulation in hemodialysis with one and three com-partment models Trans Am Soc Artif Intern Organs1989; 35:622–624

9 Dorhout Mees EJ Rise in blood pressure during modialysis ultrafiltration: a paradoxical situation? Int JArtificial Organs 1996; 19:569–570

he-10 Ifudu O, Dulin A, Lundin AP, et al Diabetics manifestexcess weight gain on maintenance hemodialysis AmSoc Artif Intern Org 1992; 21:85

11 Jones R, Poston R, Hinestrota A, et al Weight gainbetween dialysis in diabetics Possible significance ofraised intracellular sodium content Br Med J 1980; 1:153–154

12 U.S Renal Data System USRDS 1997 Annual DataReport Bethesda, MD: National Institutes of Health,National Institute of Diabetes and Digestive and KidneyDiseases, 1997

13 Miles AMV, Hong JH, Sumrani N, et al Outcome andcomplications of vascular access placement in elderlydiabetic patients with end stage renal disease J Korean

Am Med Assoc 1996; 2:25–28

14 Taber TE, Maikranz PS, Haag BW, et al Maintenance

of adequate hemodialysis access Prevention of timal hyperplasia ASAIO J 1995; 41:842–846

neoin-15 Gensini GF, Abbate R, Favilla S, Neri Serneri GC.Changes of platelet function and blood clotting in dia-betes mellitus Thromb Hemost 1979; 42:983–993

16 Halushka PV Increased platelet thromboxane J LabClin Med 1981; 97:87–92

17 U.S Renal Data System USRDS 1993 Annual DataReport Bethesda, MD: National Institutes of Health

704 Miles and Friedman

National Institute of Diabetes and Digestive and Kidney

Diseases, 1993

18 Woods JD, Turenne MN, Strawderman RL, et al

Vas-cular access survival among incident hemodialysis

pa-tients in the United States Am J Kidney Dis 1997; 30:

50–57

19 Tzamaloukas AH, Murata GH, Harford AM, et al Hand

gangrene in diabetic patients on chronic dialysis Trans

Am Soc Artif Intern Organs 1991; 37:638–643

20 Redfern AB, Zimmerman NB Neurologic and ischemic

complications of upper extremity vascular access for

dialysis J Hand Surg 1995; 20:199–204

21 Wilbourn AJ, Furlan AJ, Hulley W, Ruschhaupt W

Is-chemic monomelic neuropathy Neurology 1983; 33:

447–451

22 Riggs JE, Moss AH, Labosky DA, Liput JH, et al

Up-per extremity ischemic monomelic neuropathy: a

com-plication of vascular access procedures in uremic

dia-betic patients Neurology 1989; 39:997–998

23 Vincenti F, Arnaud SB, Recker R, et al Parathyroid and

bone response of the diabetic patient to uremia Kidney

Int 1984; 25:677–682

24 Weiss RE, Reddi AH Influence of experimental

dia-betes and insulin on matrix-induced cartilage and bone

differentiation Am J Physiol 1980, 238:E200–E207

25 Andress DL, Kopp JB, Maloney NA, et al Early

dep-osition of aluminum in bone in diabetic patients on

he-modialysis N Engl J Med 1987; 316:292–296

26 Diaz-Buxo JA, Burgess WP, Greenman M, et al Visual

function in diabetic patients undergoing dialysis:

com-parison of peritoneal had hemodialysis Int J Artificial

Organs 1984; 7:257–262

27 Mahaffey KW, Granger CB, Toth CA, et al Diabetic

retinopathy should not be a contraindication to

throm-bolytic therapy for acute myocardial infarction: review

of ocular hemorrhage incidence and location in the

GUSTO-I trial J Am Coll Cardiol 1997; 30:1606–

1610

28 Cheigh J, Raghavan J, Sullivan J, et al Is insufficient

dialysis a cause for high morbidity in diabetic patients

J Am Soc Nephrol 1991; 2:317

29 Davis M, Comty C, Shapiro F Dietary management of

patients with diabetes treated by hemodialysis J Am

Diet Assoc 1979; 75:265–269

30 Davis M, Comty C, Shapiro F Dietary management of

patients with diabetes treated by hemodialysis J Am

Diet Assoc 1979; 75:265–269

31 Maiorca R, Vonesh EF, Cavalli PL, et al A multicenter,

selection-adjusted comparison of patient and technique

survivals on CAPD and hemodialysis Perit Dial Int

1991; 11:118–127

32 Gokal R, Jakubowski C, Hunt L Multicenter study ofoutcome of CAPD and hemodialysis patients NephrolDial Transplant 1986; 1:111–114

33 Held PJ, Port FK, Blagg CR, et al The United Statesrenal data systems annual data report Am J Kidney Dis1990; 16(suppl 2):34–43

34 Yuan ZY, Balaskas E, Gupta A, et al Is CAPD or modialysis better for diabetic patients? CAPD is moreadvantageous Semin Dialysis 1992; 5:181–188

he-35 Nolph KD Continuous ambulatory peritoneal dialysis

as long term treatment for end stage renal disease Am

J Kidney Dis 1991; 17:154–157

36 Scarpioni LL, Balocchi S, Castelli A, et al Continuousambulatory peritoneal dialysis in diabetic patients Con-trib Nephrol 1990; 84:50–74

37 Khanna R, Oreopoulos DG Continuous ambulatoryperitoneal dialysis in diabetics with end stage renal dis-ease A combined experience of 2 North American cen-ters In: Friedman EA, L’Esperance FA, eds DiabeticRenal Retinal Syndrome New York: Grune and Strat-ton, 1986; 363–381

38 Rottemburg J Peritoneal dialysis in diabetics In: Nolph

KD, ed Peritoneal Dialysis Boston: Martinus Nijhoff,1985:365–379

39 Rubin J, Oreopoulos DG, Blair RDG, et al Chronicperitoneal dialysis in the management of diabetics withterminal renal failure Nephron 1977; 19:265–270

40 Moncrief JW et al The Moncrief-Popovich catheter Anew peritoneal access technique for patients on peri-toneal dialysis ASAIO J 1993; 39:62

41 Diaz-Buxo JA Is continuous ambulatory peritoneal alysis adequate long-term therapy for end-stage renaldisease? A critical assessment J Am Soc Nephrol 1992;3:1039–1048

di-42 Lin JJ, Wadhwa NK, Suh H, et al Increased peritonealsolute transport in diabetic peritoneal dialysis patients.Adv Peritoneal Dial 1995; 11:63–66

43 Young GA, Kopple JD, Lindholm B, et al Nutritionalassessment of chronic ambulatory peritoneal dialysispatients: an international study Am J Kidney Dis 1991;17:462–471

44 Rotellar C, Black J, Winchester JF, et al Ten yearsexperience with continuous ambulatory peritoneal di-alysis Am J Kidney Dis 1991; 17:158–164

45 Blumenkrantz MJ, Gahl GM, Kopple JD, et al Proteinlosses during peritoneal dialysis Kidney Int 1981; 19:593–602

46 Grodstein GP, Blumenkrantz MJ, Kopple JD, et al cose absorption during continuous ambulatory perito-neal dialysis Kidney Int 1981; 19:564–567

St Vincent’s Medical Center of Richmond, Staten Island, New York

The care of dialysis patients has been focused first on

survival and then on decreased morbidity

Rehabilita-tion has generally emphasized return to employment or

family and community activities Pregnancy and

child-bearing have been regarded as unfortunate accidents

rather than as goals of treatment Our attention to

gy-necological problems affecting dialysis patients has

been overshadowed by attention to cardiovascular

dis-ease, infection, and other life-threatening complications

of dialysis Pregnancy in dialysis patients is still

un-common and carries a high risk for both mother and

fetus However, with the increasing length of survival

of young dialysis patients and the increasing wait for

transplantation, the problems associated with

child-bearing and contraception have become more

impor-tant The possibility and implications of conception

need to be addressed with each patient While rarely

life-threatening, the problems of sexual dysfunction,

dysfunctional uterine bleeding, and gynecological

in-fections contribute to a diminished quality of life

Gy-necological neoplasms are life-threatening when they

occur, although they are not the most common cause

of death in dialysis patients

I CONTRACEPTION

Early literature reported that only 10% of female

di-alysis patients of childbearing age menstruated (1), but

a more recent report (2) indicates that the frequency of

menses has increased to 42% It is not certain how

many of these women could conceive, but the risks and

possibility of pregnancy and the need for contraceptionshould be discussed with all dialysis patients of child-bearing age

Oral contraceptives offer many advantages for alysis patients Many dialysis patients are estrogen de-ficient, and women with irregular periods may be ex-posed to the effects of unopposed estrogen forprolonged periods of time Estrogen deficiency is added

di-to the many other facdi-tors that contribute di-to bone ease, and unopposed estrogen may increase the risk ofendometrial cancer The use of oral contraceptiveswould not only prevent pregnancy but would treat es-trogen deficiency and allow for regular hormonal cy-cling Oral contraceptives should be used with caution

dis-in hypertensive women and women at risk for boembolic disease These drugs may increase the riskfor lupus flares in women whose end-stage renal dis-ease is secondary to lupus

throm-Mechanical methods of contraception can be used

in women for whom estrogen is contraindicated uterine devices may be associated with increased uter-ine bleeding when patients are heparinized, and an in-crease in the risk of peritonitis would be expected inperitoneal dialysis patients Other mechanical methods

Intra-of birth control are acceptable in dialysis patients

II PREGNANCY

A Frequency of Conception

Fertility is markedly reduced in dialysis patients mates of the frequency of conception in dialysis pa-

Esti-706 Hou and Grossman

tients range from 1.4% per year in Saudi Arabia (3) to

0.5% in the United States (4) and 0.3% in Belgium (5)

The estimates from the United States and Saudi Arabia

are based on surveys that covered only half the women

of childbearing age treated with dialysis, while the

sur-vey from Belgium included a response from all of the

dialysis units in the country Only the report from the

American National Registry for Pregnancy in Dialysis

Patients (NRPDP) included a substantial number of

peritoneal dialysis patients (4) Of note is that the

fre-quency of conception in hemodialysis patients was

two to three times higher than in peritoneal dialysis

patients

The reasons for the rarity of pregnancy in dialysis

patients are not well understood The hormonal

changes in dialysis patients are reviewed elsewhere in

this book Nonhormonal causes of infertility have not

been investigated

It is not clear whether the difference in the frequency

of conception between hemodialysis patients and

CAPD patients is the result of endocrine differences or

is in some way related to peritoneal dialysis itself

Re-current peritonitis might be expected to cause tubal

ob-struction in peritoneal dialysis patients, but if tubal

damage were a major contributor to infertility, an

in-crease in tubal pregnancies would be expected Few

tubal pregnancies have been described in dialysis

pa-tients and none in peritoneal dialysis papa-tients It is

pos-sible that hypertonic dextrose damages the ovum or

that the volume of fluid in the intraperitoneal space

interferes with transport of the ovum from the ovary to

the fallopian tubes

Most pregnancies occur during the first few years on

dialysis, but conception rates as a function of time on

dialysis have not been determined Pregnancy has

oc-curred in women who have been on dialysis as long as

20 years Repeat pregnancies in women who become

pregnant on dialysis are not uncommon In the 318

women whose pregnancies are recorded by the NRPDP,

8 became pregnant twice, 8 became pregnant three

times, and one conceived four times (4) Although it

would be expected that pregnancy would be more

likely in women with regular menses, pregnancy has

been reported in a woman after 9 years of amenorrhea

In contrast to dialysis patients, approximately 12% of

women transplant recipients of childbearing age

be-come pregnant

B Outcome of Pregnancy in Dialysis Patients

In 1980, the European Dialysis and Transplant

Asso-ciation reported 115 pregnancies in dialysis patients

(6) Of those that were not electively terminated, only23% resulted in surviving infants Success rate forpregnancy in dialysis patients has improved since theEDTA report In Saudi Arabia 30% of pregnancies re-sulted in surviving infants The NRPDP recorded 222pregnancies in women who were receiving dialysis atthe time of conception Of the 141 pregnancies thatreached the second trimester, 55% resulted in survivinginfants Eighteen percent of live-born infants died inthe neonatal period; 8.5% of pregnancies reaching thesecond trimester resulted in stillbirth, and 22% resulted

in spontaneous abortion The four induced abortionsdone in the second trimester were done for life-threat-ening maternal problems (three hypertension and onecritical aortic stenosis) rather than for social reasons oranticipated problems

C Maternal Complications

1 Maternal DeathThere have been three maternal deaths reported to theNRPDP One death resulted from lupus cerebritis in awoman who started dialysis after conception Therewere two deaths in women who conceived after startingdialysis, one as a result of hypertension and one fromunknown causes All three infants survived

2 HypertensionHypertension is the most common life-threateningcomplication of pregnancy in dialysis patients Of 57case reports published in the medical literature inwhich blood pressure was noted in a pregnant dialysispatient, only 30% of women were normotensivethroughout pregnancy (7) Sixty percent of women hadblood pressures over 140/90 at some time during preg-nancy, and in 25% the blood pressure exceeded 170/

110 Ten percent were treated with antihypertensivemedications, but blood pressure was not specified

In cases reported to the NRPDP, approximately 80%

of the 68 women for whom blood pressure ments were available either had a blood pressuregreater than 140/90 or required antihypertensive med-ication at some time during pregnancy (4) In over half

measure-of hypertensive pregnant dialysis patients, the bloodpressure exceeds 170/110 Five of these women re-quired intensive care unit admissions in addition to thematernal one death Thirty-eight percent of patientswho developed severe hypertension did so in the firsttrimester In such cases, it was usually possible to con-trol the blood pressure without terminating the preg-nancy Fifty percent of women with severe hyperten-

sion reached blood pressures of >170/110 in the second

trimester These cases were problematic in that

pre-eclampsia could not be excluded and the fetus was still

not viable Three of 16 women required therapeutic

abortion for hypertension Of note, severe hypertension

could be seen as late as 6 weeks postpartum

3 Anemia

A drop in hematocrit is almost invariable in dialysis

patients who become pregnant In pregnancies reported

to the NRPDP, 33% of women treated with

poietin and 77% of women not treated with

erythro-poietin required transfusion (4) Iron stores usually

dropped, but there were several instances in which the

hematocrit dropped despite iron saturation, which

re-mained at acceptable levels

D Prematurity and Growth Restriction

Eighty-five percent of infants born to women who

con-ceived after starting dialysis reported to the NRPDP

were born before 37 weeks gestation (mean gestational

age 32.4 weeks) Thirty-six percent weighed less than

1500 g at birth, and 28% were small for gestational

age Their neonatal course was complicated by

respi-ratory distress and other complications of prematurity

Eleven of 116 live-born infants and 1 stillborn infant

reported to the NRPDP had congenital anomalies (4)

Eleven of 49 infants for whom follow-up data were

available had long-term medical or developmental

problems, most of which appeared to be the result of

prematurity rather than an azotemic intrauterine

envi-ronment The mean gestational age was lower for

in-fants who had long-term problems compared to those

with normal growth and development (30.6 vs 34.3

weeks)

III MANAGEMENT ISSUES IN

PREGNANT DIALYSIS PATIENTS

A Preconception Counseling

Counseling of dialysis patients who are attempting

pregnancy is all but impossible The infrequency of

conception makes it impossible to plan except in

women who have already conceived once on dialysis

Even if a woman is actively trying to become pregnant,

the likelihood of conception is low enough that changes

in dialysis regimen cannot be justified However, folic

acid supplementation should be increased and good

blood sugar control should be achieved in diabeticwomen who are attempting conception

B Diagnosis of Pregnancy

Pregnancy is usually diagnosed late in dialysis patients.Irregular menses are common and abdominal com-plaints are often attributed to other causes A high index

of suspicion is required to make the diagnosis early.Urine tests for human chorionic gonadotropin (hcg) areinaccurate even in women who have residual renalfunction Small amounts of hcg are made by somaticcells, and because the hormone is partially excreted bythe kidney, serum tests for ␤ hcg are sometimes bor-derline or falsely positive in women who are not preg-nant (8) The titers of ␤ hcg may be higher than ex-pected for the stage of gestation The diagnosis andstage of gestation must be confirmed by ultrasound

C Management of Hypertension

Dialysis patients, particularly those on home treatmentmodalities, should measure their own blood pressuretwice daily since severe increases in blood pressure can

be abrupt As with dialysis patients who are not nant, the first line of treatment is correction of volumeoverload Assessment of volume status is difficult be-cause of the expected 9 L increase in total body waterduring pregnancy, but cautious fluid removal should beattempted when hypertension develops If fluid removaldoes not correct blood pressure, pharmacological treat-ment can be started There is experience with a widevariety of antihypertensive medications in pregnancy(Table 1) Of the widely used antihypertensive drugs,only angiotensin-converting enzyme (ACE) inhibitorsand, by inference, angiotensin receptor blockers arestrongly contraindicated in pregnancy

preg-1 Angiotensin-Converting Enzyme InhibitorsThis group of drugs has been associated with oligo-hydramnios, which results in a number of complica-tions Amniotic fluid is necessary for fetal lung devel-opment, and the most serious consequence ofoligohydramnios is pulmonary hypoplasia leading toneonatal death as a result of respiratory failure (9) Oli-gohydramnios also accounts for limb contractures ininfants exposed to the drug Direct pressure of the uter-ine muscle on the fetal skull is thought to result inabnormal calcification of the skull Several instances ofpatent ductus have been described This effect is

708 Hou and Grossman

Table 1 Antihypertensive Drugs Used in Pregnancy

Chronic Hypertension

ACE inhibitors (D) Contraindicated; 2nd and 3rd trimester use associated with pulmonary hypoplasia,

hypocalvaria, renal dysplasia, neonatal anuria, contractures; no known harm in1st trimester

␣ Methyl dopa (C) Safe; 40 year use; careful developmental testing of children at ages 4 and 7; rare

Coombs⫹ hemolytic anemia, rare hepatitis

␤ Blockers (C) Probably safe; fetal bradycardia, hypoglycemia, respiratory depression at birth,

intrauterine growth restriction?↓ fetal tolerance of anoxic stressLabetolol (C) Limited first trimester experience; less bradycardia and growth restriction than

␤ blockersClonidine (C) Probably safe; limited 1st trimester experience

Calcium channel blockers (C) Profound↓ BP with when used with magnesium; limited experience; reserve for

severe hypertensionHydralazine (C) Safe; long experience with use in pregnancy; no↑ birth defects; ineffective as a

single agentMinoxidil (C) Very limited experience; hypertricosis and congenital anomalies in one the infantPrazocin (C) Limited experience; no problems noted

Thiazide diuretics (D) ↑ congenital anomalies with chlorthaldone; subnormal intravascular volume

expansion, neonatal thrombocytopenia, hemolytic anemia,electrolyte abnormalities

Hypertensive Crisis

Hydralazine (C) Used for 40 years without serious side effects

Labetolol (C) Shorter length of use; appears safe

Nitroprusside (C) Fetal cyanide toxicity

Diazoxide (C) Fatal maternal hypotension reported; limit dose to 30 mg boluses;↓ uterine

contraction; neonatal hyperglycemia

thought to be the result of the effects of ACE inhibitors

on prostaglandin metabolism

While exposure to ACE inhibitors in the second and

third trimesters may have serious consequences, no ill

effects have been identified as a result of first trimester

exposure Two studies, one involving 46 infants and

one involving 86 infants, showed no adverse effect of

exposure to ACE inhibitors in the first trimester (10)

In the latter report there were four congenital

anoma-lies, a number that was not significantly different from

the expected three Women with inadvertent first

tri-mester exposure need not be advised to terminate the

pregnancy

There is less experience with angiotensin II receptor

blockers, but it is expected that problems caused by

decreased angiotensin effect will be similar to those

seen in women treated with ACE inhibitors

2 Other Antihypertensive Drugs

a. ␣Methyl Dopa

␣ Methyl dopa has been used in pregnant women for

over 40 years and is still the drug of first choice for

essential hypertension Careful developmental studieshave been done at 4 and 7 years of age in childrenexposed to the drug in utero, and no problems havebeen found (11)

b. ␤Blockers

There are several case reports of neonatal bradycardia,hypoglycemia, and respiratory depression associatedwith␤blockers, but these problems are generally easilymanaged by the neonatologist (12) There are mixeddata concerning whether blockers are associated withintrauterine growth restriction There are reports ofsmall-for-gestational-age infants of mothers treatedwith␤blockers for diseases not usually associated withgrowth restriction (13) There are also data from animalmodels suggesting a decreased ability of the fetus towithstand anoxic stress (14) None of these problemshas turned out to be a major contraindication to the use

of this category of drugs in pregnant humans Fetalbradycardia may make it difficult to interpret antenatalmonitoring, which depends on changes in fetal heartrate

c Labetolol

Labetolol is not associated with fetal bradycardia and

growth restriction and, it is widely used in preference

to ␤ blockers Nonetheless, data on first trimester

ef-fects of the drug are still limited (10) Moreover,

con-trolled studies have not shown it to be superior to other

antihypertensive agents (15)

d Clonidine

Clonidine is a centrally acting ␣2-agonist, which has

been reported in one study to have efficacy and safety

similar to methyl dopa (16) In view of the limited

experience with it, there is no reason to use it in

pref-erence to␣methyl dopa

e Prazocin

No adverse effects on the fetus have been demonstrated

with prazocin, but the experience with it is more

lim-ited than with labetolol,␣methyl dopa, and␤blockers,

and it does not appear to offer any advantage The drug

can be continued in women whose blood pressure is

well controlled on it at the time of conception

f Calcium Channel Blockers

Nifedipine, nicardipine, and verapamil have been used

in severe hypertension They do not appear to be

as-sociated with any increase in congenital anomalies

when used in the first trimester These drugs have been

used for treatment of premature labor in the third

tri-mester Experience with diltiazem is more limited

Cal-cium channel blockers may potentiate the hypotensive

effects and neuromuscular blockade of magnesium and

the interaction should be kept in mind when the drugs

are used in women with a possibility of developing

preeclampsia (17,18) Because of the limited

experi-ence with all members of this group of drugs, their use

is best limited to severe hypertension unresponsive to

other drugs

g Vasodilators

Hydralazine has been used safely during pregnancy for

40 years It is ineffective as a single oral agent but can

be added to a first-line drug if the latter does not

ade-quately control blood pressure The more potent

vaso-dilator, minoxidil has been associated with

hypertri-chosis and congenital anomalies in one case report

(19) It is ineffective unless combined with a diuretic

and a sympatholytic agent

3 Drugs for Hypertensive Emergencies

a Hydralazine

Intravenous hydralazine in doses of 5–10 mg every20–30 minutes is the drug of first choice for hyperten-sive crisis in pregnancy A single study has shown ahigh frequency of malignant ventricular arrhythmias ineclamptic women treated with hydralazine than inwomen treated with labetolol (20) Nine studies com-paring hydralazine with other drugs, most often intra-venous labetolol, have found no advantage of one drugregimen over another (21)

b Labetolol

Intravenous labetolol given either as a 20 mg loadingdose followed by 20 mg every 30 minutes or a 1–2mg/min drip is the second most commonly used regi-men for treating hypertensive emergencies in pregnantwomen There are occasional reports of fetal bradycar-dia, and the newborn should be monitored for hypo-tension

c Diazoxide

There is extensive experience with the use of diazoxide

in pregnancy, but the drug is now primarily of historicinterest In doses of 150–300 mg it has been associatedwith at least one maternal fatality from hypotension It

is also associated with decreased uterine contractionsand neonatal hyperglycemia Its only advantage is along duration of action, which may make it useful in awoman who must be transported with minimal moni-toring capability or when other drugs have failed Itshould be used only in 30 mg boluses every 1–2minutes until the desired blood pressure is reached

D Infections

Pregnant hemodialysis patients are probably at no morerisk of infection than those who are not pregnant, butthe use of antibiotics will be influenced by pregnancy.Most penicillins are safe during pregnancy First-gen-eration cephalosporins such as cephazolin and cephal-exin are safe in pregnancy (10) Cephalosporins with amethyltetrathiazole moiety (cefoperozone, cefotetan,moxalactam, and cefamandole) are usually avoided inpregnancy because studies have shown infertility in an-imals (22) Sulfa drugs, such as sulfamethoxiazole, can

be used in the early part of pregnancy but should beavoided in the latter part of pregnancy because there is

a risk of kernicterus Trimethoprim and sulfamethoxiazole combinations are generally avoided

trimethoprim-710 Hou and Grossman

because of the teratogenicity of folic acid antagonists

In practice, significant increases in congenital

anoma-lies have not been noted with these drugs The

quino-lone antibiotics should be avoided in pregnancy

be-cause they have been associated with weakened

cartilage in young animals Aminoglycosides should

also be avoided because of their association with 8th

nerve damage

E Peritonitis

The anatomic connection between the uterus and the

intraperitoneal cavity raises the concern that peritonitis

will result in intrauterine infection and vice versa

There have been few cases reported of peritonitis in

pregnant CAPD patients In three cases reported,

per-itonitis was followed by labor in two (23,24) One

pregnancy resulted in a premature baby, who survived,

and the other resulted in a stillbirth Five additional

cases of peritonitis have been reported to the NRPDP

There was only one fetal loss, which was remote from

the time of peritonitis There is one report of

postpar-tum Escherichia coli peritonitis requiring removal of

the peritoneal catheter resulting from ascending

infec-tion in a woman with chorioamnionitis (25)

F Erythropoietin

There are limited but reassuring data on the use of

erythropoietin during pregnancy (26) There have been

no reports of teratogenicity Since conception is not

usually expected in dialysis patients, these women are

generally being treated with erythropoietin and are well

into the period of organogenesis when pregnancy is

diagnosed The drug has not been associated with

in-creased difficulty controlling hypertension or with

polycythemia in the infants Dialysis patients who are

not treated with erythropoietin almost always require

transfusions (4) Erythropoietin requirements increase

in pregnancy, and a 50–100% increase in the dose can

be prescribed as soon as pregnancy is diagnosed

G Iron

Serum iron and ferritin usually drop during pregnancy

in dialysis patients, and iron deficiency may play a role

in anemia later in pregnancy The safety of intravenous

iron has not been established, but it has been widely

used There are old data indicating that iron may be

transferred disproportionately to the fetus, and if

intra-venous iron is used, it seems prudent to give it in small

doses to minimize the risk of acute iron toxicity The

outcome of pregnancy in dialysis patients treated withintravenous iron is not different from women who havenot received iron

H Dialysis Regimens

1 Choice of ModalityWhen the first cases of pregnancy in peritoneal dialysispatients were reported, it appeared that the outcomewas better for peritoneal dialysis patients than for he-modialysis (27) With the accumulation of more data,

it has become clear that the apparent superiority ofperitoneal dialysis simply reflected the overall im-provement in outcome for pregnancies in dialysis pa-tients compared to earlier reports

There are theoretical advantages to peritoneal ysis in that there are no rapid metabolic changes andvolume removal is gradual There may be disadvan-tages with difficulty maintaining adequate nutrition forpregnancy There is no reason to switch a woman who

dial-is stable on either hemodialysdial-is or peritoneal dialysdial-is

to another modality because of pregnancy per se Whenstarting dialysis during pregnancy, the usual criteria forchoosing a dialysis modality can be used Peritonealdialysis has been successfully used in women with end-stage renal disease secondary to diabetic nephropathy

It might be relatively contraindicated in nondiabeticwomen who are at high risk for gestational diabetes

2 Hemodialysis

a Intensive Dialysis

The value of intensive dialysis (daily dialysis) in proving the outcome of pregnancy in dialysis patientshas not been established, but there are theoretical rea-sons to support its use Women who begin dialysis dur-ing pregnancy and have residual renal function have abetter pregnancy outcome (75–80% vs 40% infant sur-vival) (4) It is not known whether residual excretoryfunction or endocrine function is responsible for thebetter outcome, but an attempt to lower the fetal ex-posure to metabolic waste products seems reasonable.With daily dialysis, interdialytic weight gains are mod-est and the risk of hypotension with fluid removal isdecreased Polyhydramnios is common in dialysis pa-tients A recent report of 17 pregnancies in dialysis pa-tients from University of Sa˜o Paulo noted polyhydram-nios in 17 (28) Uterine distension associated withpolyhydramnios may contribute to premature labor Ahigh blood urea nitrogen in the fetus that has normalkidneys may cause an osmotic diuresis, which aggra-vates polyhydramnios This hypothesis is supported by

im-Table 2 Dietary Guidelines for Pregnant Dialysis PatientsProtein HD: 1.2 g/kg ideal body wt⫹ 10 g/d

PD: 2.4 g/kg ideal body weight⫹ 10 g/dCalories: 35 kcal/kg⫹ 300 k/cal

Sodium: 3 g/dPotassium: 2–3 g/dPhosphorus: 1200 mg/dCalcium: 1–2 g as phosphate binders; 2 g/d if 2.5 mEq/L

bath is usedVitamin A and E: no supplementFolate: 1.8 mg/d

Vitamin C: 170 mg/dThiamine: 3 mg/dRiboflavin: 3.4 mg/dNiacin: 20 mg/dB6: 5 mgBiotin: 600 mgZinc: 15 mgCarnitine: 330 mg

the observation that a urea diuresis is usual in infants

born to mothers on dialysis A report of 27 pregnancies

in women in Saudi Arabia found a significantly longer

dialysis time in women with successful pregnancies

compared to women with unsuccessful pregnancies (12

h vs 10 h) (3) Limited data from the NRPDP indicate

that dialysis must be increased to at least 20 hours per

week to achieve any improvement in outcome

b Dialysis Bath

If intensive dialysis is used, several adjustments in the

dialysate composition may be necessary Serum

potas-sium may drop if dietary increases in potaspotas-sium do not

offset increased losses, and the dialysate potassium

may need to be raised With daily dialysis,

hypercal-cemia may occur if a bath containing 3.5 mEq/L of

calcium is used, and 2.5 mEq/L is usually preferable

While potassium and calcium in the dialysate are

rel-atively easy to adjust, bicarbonate is more problematic

The dialysate bicarbonate of 35 mEq/L is designed to

offset 2 days of acid production Daily dialysis may

result in excessive bicarbonate gain The situation is

further complicated by the normal respiratory alkalosis

of pregnancy in which the appropriate serum

bicarbon-ate is 18–20 mEq/L rather than 25 mEq/L If serious

alkalosis develops, an individually formulated dialysis

solution may be necessary

c Anticoagulation

In the past, recommendations have been made to

min-imize anticoagulant dose in pregnant women However,

the usual practice in all dialysis patients is to give the

smallest amount of anticoagulation possible An

at-tempt at lowering heparin doses results in the same

problems of clotting of the extracorporeal circuit that

it does in nonpregnant patients Although direct

com-parisons have not been made, clotting problems may

even be increased because pregnancy is a

hyperco-agulable state Heparin does not cross the placenta and

is not teratogenic (29) We recommend that

heparin-free dialysis be limited to women with bleeding

prob-lems Coumadin does cross the placenta, is teratogenic

in the first trimester, and may cause bleeding in the

fetus in the third trimester (10) Women treated with

coumadin either for recurrent access clotting or for

other reasons should be switched to subcutaneous

hep-arin in doses adequate for full anticoagulation

3 Peritoneal Dialysis

In late pregnancy it becomes difficult for a woman to

tolerate her usual exchange volume (27) Volume must

be reduced and the number of exchanges increased Itmay become difficult to maintain even the previouslevel of dialysis To increase the amount of dialysisdelivered, it is necessary to use a combination of day-time CAPD and nighttime CCPD There is not enoughdata to determine whether increasing the amount ofperitoneal dialysis delivered improves outcome

I Calcium and Phosphorus

Thirty grams of calcium are necessary for calcification

of the fetal skeleton If the patient is dialyzed on a bathcontaining more than 3.5 mEq/L of calcium, dialysiseasily provides this amount If she is dialyzed on alower calcium bath, enough calcium should be ab-sorbed from phosphate binders to provide the necessarycalcium if she takes at least 2 g of calcium daily 1,25-Dihydroxyvitamin D preparations, either oral or intra-venous, are usually continued, although their effect inpregnancy is not well understood The placenta doesconvert some 25-OH2D3 to 1,25-OH2D3 (30) Highdoses of 1,25-OH2D3 have been used in one patientwith hypoparathyroidism without ill effects on the fetus(31)

J Nutritional Considerations

Guidelines for nutritional care of a pregnant dialysispatient are still in a state of evolution (Table 2)

712 Hou and Grossman

I Weight Gain

It is almost impossible to prescribe weight gain The

task confronting the health care team is usually to

de-termine how much of observed weight gain is either

pregnancy related increase in soft tissue or increased

total body water appropriate for pregnancy and how

much is fluid overload that should be removed with

dialysis Normal pregnancy is accompanied by an

in-crease in total body water of approximately 9 L, most

of it in the extracellular space Fluid retention occurs

to some extent in dialysis patients A common

obser-vation early in pregnancy and occasionally a clue to

the diagnosis is hypotension in response to the attempt

to remove fluid The patient should be examined

care-fully on a weekly basis for evidence of fluid overload,

and if signs of excessive volume expansion are present,

a careful attempt to remove fluid with dialysis should

be made

2 Protein and Calories

Practically, with intensive dialysis, no protein

restric-tion is necessary during pregnancy, and provision of

adequate protein and calories is frequently a problem

Supplements are often required, and several instances

in which intradialytic parenteral nutrition was used

have been reported (32)

3 Fluid Restriction

Daily dialysis allows for decreasing but not eliminating

a fluid restriction One goal of daily dialysis is to avoid

the need to remove more than 1–2 L during a single

treatment

4 Water-Soluble Vitamins

The requirements for water-soluble vitamins increase

during pregnancy, and increased frequency of dialysis

increases the loss of water-soluble vitamins above the

usual for dialysis patients

Folic acid is necessary for the increased

hematopoi-esis that occurs during pregnancy Preconception

sup-plementation with 0.8 mg of folic acid has been shown

to reduce the risk of neural tube defects (33) The usual

renal diet is frequently low in fruits and vegetables and

thus low in folate The usual supplement for dialysis

patients is 1 mg/day, and we recommend increasing the

supplement to 1.8–2 mg/day

Requirements for all vitamins are increased during

pregnancy and an additional increase is required for

pregnancy with additional increases required in dialysis

patients Vitamin C dose should be increased ciency has been associated with neonatal scurvy

Defi-5 Fat-Soluble VitaminsSupplements of vitamin A are usually prescribed innormal pregnancy but should not be prescribed in preg-nant dialysis patients Excretion is decreased in dial-ysis patients, and it is not removed by dialysis Veryhigh doses of vitamin A have been associated with con-genital anomalies similar to those seen with isoretinoin(34)

Standard preparations of vitamin D have little effect

on dialysis patients, and their presence in vitamin arations has little relevance

prep-Vitamin E supplements are not required in pregnantdialysis patients

6 ZincZinc is necessary for human reproduction, and its de-ficiency has been associated with teratogenesis (34).Later in pregnancy it has been associated with prema-ture birth and atonic uterine bleeding A supplementshould be given particularly to patients taking oral iron

IV OBSTETRIC MANAGEMENT

Care of the pregnant dialysis and transplant patient quires close cooperation between the nephrologist and

re-a perinre-atologist experienced in tre-aking cre-are of womenwith renal disease A referral to a high-risk obstetricianshould be made as soon as pregnancy is diagnosed.Because of the high frequency of severe prematurity, alevel three nursery should be available

A Premature Labor

Prematurity is the greatest cause of morbidity and tality in the infants of women with renal disease Thereare several special considerations to bear in mind withthe most commonly used treatments for premature la-bor Magnesium can be used, but as in its use for pre-eclampsia, extreme caution must be used to avoid mag-nesium toxicity and respiratory depression in womenwith renal insufficiency (35) In dialysis patients, aloading dose can be given and supplemented after eachdialysis treatment or when the magnesium level hasbeen documented to fall below 5 mg/dL Continuousinfusion should not be used in dialysis patients.Indomethacin has been used to treat premature labor

mor-in women with renal disease and may be especially

effective in women with polyhydramnios (26)

How-ever, in women with renal insufficiency or in dialysis

patients with residual renal function, there may be a

loss of renal function, causing hyperkalemia and

re-quiring initiation or increase in dialysis

Premature labor usually occurs early enough that it

is desirable to delay delivery longer than the usual 72

hours that indomethacin is used The fetus should be

monitored for any evidence of right heart strain, and

the mother should be monitored for polyhydramnios

Midtrimester losses are common in dialysis patients,

and there have been a few reports of incompetent

cer-vix Dialysis patients should be monitored for any signs

of cervical shortening or dilatation

B Fetal Surveillance

Because of the increased risk of stillbirth, fetal

sur-veillance should be started as soon as there is a

pos-sibility of survival outside the mother Because of the

risk of precipitating labor, other tests should be used in

preference to oxytocin challenge testing

C Labor and Delivery

Vaginal delivery should be the goal of management,

and cesarean section should be done only for the usual

obstetric indications rather than for renal disease per

se When cesarean section is done in peritoneal dialysis

patients, an attempt should be made to use an

extra-peritoneal approach An attempt to resume extra-peritoneal

dialysis with low-volume exchanges can be made 24

hours after operative delivery If there is leaking from

the incision, the patient should be switched to

hemo-dialysis for 2 weeks

D Management Issues in the Newborn

Infants born to dialysis patients should be observed in

a high-risk setting even if they appear normal Infants

of dialysis patients are born with blood urea nitrogen

(BUN) and serum creatinine equal to the mother’s, and

following birth they experience an osmotic diuresis that

results in volume contraction and electrolyte disorders

unless there is careful monitoring and replacement of

fluid and electrolytes

V DYSPAREUNIA

Some female dialysis patients may experience

dys-pareunia because of estrogen deficiency and resulting

vaginal dryness Dyspareunia resulting from atrophicvaginitis from low estrogen levels can be corrected byintravaginal conjugated estrogens (2–4 g daily) or oralestrogen progesterone compounds A daily dose of0.625 mg of conjugated estrogen and 2.5 mg of med-roxyprogesterone provides enough estrogen to preventdyspareunia If there is breakthrough bleeding on thiscombination, progesterone can be increased to 5 mg.Women treated with intravaginal estrogens should re-ceive progesterone as well because there is substantialsystemic absorption

VI SEXUAL DYSFUNCTION

Fifty percent of female dialysis patients under the age

of 55 are sexually active (2), and a majority of themexperience some sexual dysfunction They suffer fromboth decreased libido and decreased ability to achieveorgasm Treatment with erythropoietin (EPO) appears

to be associated with an improvement in sexual tion, but most of the data collected have been in men(36) Various reasons for sexual dysfunction have beenproposed, including hyperprolactinemia, gonadal dys-function, depression, hyperparathyroidism, and change

func-in body image

Hyperprolactinemia is seen in 75–90% of femaledialysis patients (37–39) The mean serum prolactinlevels in women with sexual dysfunction are higherthan in patients with normal sexual function Treatment

of hyperprolactinemia with the dopamine agonistbromergocriptine has been reported (in limited uncon-trolled studies) to improve sexual function in both menand women on dialysis (40) It has not come into wide-spread use because hemodialysis patients are particu-larly susceptible to the hypotensive effects of this drug.There are no reports of its use in CAPD patients or onthe use of other dopamine agonists Bromocriptineshould be started at a dose of 1.25 mg, and the firstdose should be taken at night Subsequent doses can

be gradually increased Doses of 2.5 mg bid should beadequate to suppress prolactin secretion When cor-rectable physical problems cannot be found, dialysispatients should be referred for sex therapy, as wouldpatients without renal failure

VII HORMONE REPLACEMENT THERAPY

Holley and colleagues found that only 5% of womenaged 55 or greater at the time of starting dialysis werereceiving hormone replacement therapy No firm guide-

714 Hou and Grossman

lines for hormone replacement therapy have been

de-veloped in dialysis patients Hormone replacement

therapy in healthy women slightly increases the risk of

breast cancer while reducing the risk of osteoporosis

and heart disease It is not known whether the risk of

breast cancer would be higher in dialysis patients than

in healthy women, but the risk of heart disease and

bone disease are clearly increased In the absence of

specific data, these risks make treatment of

postmeno-pausal women on dialysis with estrogen-progesterone

cycling a reasonable approach As noted, many women

younger than 55 years of age have estrogen deficiency

Hormone replacement therapy is the standard of care

for premenopausal women who undergo

oophorec-tomy Although it is not general practice, there is every

reason to think that dialysis patients who are estrogen

deficient should receive replacement therapy An

ex-perienced gynecologist may have to try a number of

combinations of these hormones to find a regimen that

does not cause excessive bleeding

There is limited experience with the use of other

treatments for osteoporosis in dialysis patients The use

of alendronate in patients with creatinine clearance of

<35 mL is not recommended by the manufacturer

be-cause of lack of experience, not bebe-cause of known

ad-verse effect A large portion of the drug is excreted by

the kidneys, and a dose adjustment would be necessary

The efficacy of alendronate in decreasing

steroid-in-duced osteoporosis has heightened interest in its use in

renal disease, and its use in dialysis patients is now an

area of active investigation (41)

VIII DYSFUNCTIONAL UTERINE

BLEEDING

A Incidence

Many women develop amenorrhea when the

glomeru-lar filtration rate falls to less than 10 mL/min

Men-struation returns in as many as 50% of premenopausal

women once dialysis is started Over half of women

with end-stage renal disease (ESRD) who menstruate

report hypermenorrhea (2), and 60% have irregular

cy-cles, with similar menstrual abnormalities in

hemodi-alysis and CAPD patients Dysfunctional uterine

bleed-ing is common and is of concern because it may be an

early sign of endometrial cancer Blood loss may lead

to severe anemia even in women treated with EPO,

although the introduction of EPO has made the

man-agement of dysfunctional uterine bleeding substantially

2 Bloody Peritoneal DialysateMenstruation, ovulation, or uterine bleeding from anycause can result in bloody peritoneal fluid in peritonealdialysis patients There is no specific management, andtreatment is rarely necessary unless there is profusebleeding In rare cases, frank hemoperitoneum may oc-cur, requiring suppression of ovulation (42) An asepticperitonitis picture during menstruation or ovulation hasalso been reported (43)

3 AnticoagulationThe lowest possible dosage of heparin should be used

to perform hemodialysis when a woman is ing Heparin-free techniques and citrate anticoagulationalso are available

menstruat-4 Hormone TherapyRecent advances in therapy have facilitated the man-agement of dysfunctional uterine bleeding in womenwith end-stage renal disease

Oral contraceptives remain the safest therapy andthe first-line treatment, although they should not beused if hypertension control is a problem or if there is

a history of deep vein thrombosis There are theoreticalbenefits of using estrogen-progesterone combinations

to prevent uterine cancer and osteoporosis

Medroxyprogesterone acetate (Depo-Provera) can begiven in a dose of 100 mg IM once a week for 4 weeksand then once a month Because many dialysis patientshave a platelet dysfunction, IM injections may result in

hematoma formation Moreover, the half-life of IM

medroxyprogesterone acetate is unpredictable

Medrox-yprogesterone acetate is best reserved for patients with

chronic hypermenorrhea who do not respond to oral

hormonal therapy In women whose hypermenorrhea

does not respond to oral contraceptives or progestins,

gonadotropin-releasing hormone agonists can be used

The dosage is 7.5 mg of long-acting leuprolide acetate

IM, monthly This drug is extremely expensive There

is one report of ovarian hyperstimulation in a patient

on chronic dialysis who received two doses of

leu-prolide acetate (44) It has been postulated that women

with ESRD may be at risk for this complication

be-cause of decreased excretion of the

gonadotropin-re-leasing hormone agonists The use of leuprolide should

be undertaken by a gynecologist familiar with the

prob-lems of patients with end-stage renal disease

In the case of acute excessive blood loss, high-dose

estrogen therapy can be used, giving 25 mg of

conju-gated estrogens IV every 6 hours Bleeding usually

subsides within 12 hours

In setting of acute blood loss when bleeding time is

prolonged, DDAVP can be used as it is in with other

bleeding problems, in a dosage of 0.3 pg/kg in 50 mL

of saline given every 4–8 hours for three to four doses

5 Nonhormonal Treatments

a Laser Ablation

The neodymium (Nd):YAG laser now offers a safe and

effective alternative to hysterectomy With this

tech-nique, the endometrial lining is ablated by vaporizing

all three of its layers Patients are pretreated with either

danazol 200 mg four times daily for 4–6 weeks or with

gonadotropin-releasing hormone agonists The

tech-nique requires a surgeon trained and experienced in

operative hysteroscopy and the use of the Nd:YAG

la-ser The procedure leads to permanent infertility

b Hysterectomy

For postmenopausal women with significant

dysfunc-tional uterine bleeding, hysterectomy is a possible

ap-proach The proposed operation should be carefully

discussed with the patient, and concomitant medical

problems and the risks of surgery should be taken into

consideration With the advent of endometrial ablation

with laser, hysterectomy will now probably be reserved

for women who have bleeding secondary to uterine

fi-broids or to other uterine or pelvic pathology that in

itself warrants the surgery Hysterectomy should be

done only as a life-saving procedure in a

premenopau-sal woman who is a candidate for renal transplantation,because the latter will frequently restore fertility

6 Gynecological Neoplasms

a Benign

Uterine fibroids, or leiomyomata, are extremely mon, occurring in approximately 25% of women overthe age of 30 There is no information about their fre-quency in chronic renal failure In dialysis patientswithout serious comorbidities, the management of uter-ine fibroids is similar to the approach in women with-out renal failure

com-b Incidence of Malignant Tumors

Although it was previously believed that the incidence

of endometrial carcinoma is increased in female ysis patients, several recent studies suggest that breast,endometrial, and ovarian cancers are not increased inthis population

dial-c Screening

Guidelines for breast cancer screening are given similar

to those given for the general population Pap smearsshould be done yearly to screen for cervical cancer indialysis patients Women who have had immunosup-pressive therapy, because of either previous transplan-tation or underlying renal disease, or women withAIDS should have PAP smears every 6 months because

of the increased incidence of cervical cancer in thesepopulations Endometrial cancer usually presents asdysfunctional uterine bleeding, the investigation andmanagement of which have been discussed above.Ovarian cancer usually presents with vague abdominalsymptoms and later as an ovarian mass Abdominal dis-comfort, nausea, and weight loss induced by ovariancancer may initially be misinterpreted as symptoms ofuremia or underdialysis In patients on peritoneal di-alysis, ovarian cancer may present as bloody peritonealfluid, an abnormal peritoneal cell count, or a change inthe color of the fluid A high index of suspicion is nec-essary to detect ovarian cancer at an early and poten-tially curable stage

IX DIAGNOSTIC TESTS FOR GYNECOLOGICAL DISEASE IN DIALYSIS PATIENTS

A Computed Tomography

Intravenous contrast infusion, if needed to perform a

CT scan or angiography, is not contraindicated in a

716 Hou and Grossman

dialysis patient Although the administration of contrast

involves increasing intravascular volume and

osmolal-ity, immediate dialysis following the study can be

per-formed if deemed necessary A patient on peritoneal

dialysis requiring an abdominal CT scan can present

for the examination with dialysis fluid in the abdomen

B Pelvic and Abdominal Ultrasonography

The patient on peritoneal dialysis with a suspected

pel-vic or ovarian lesion should undergo ultrasound

scan-ning of the involved area In those instances where

pel-vic pathological changes cannot be visualized without

distending the bladder, the latter can be filled via a

Foley catheter Hyponatremia will result from

mis-guided attempts by ultrasound personnel to fill the

bladder by having the patient drink water

C Transvaginal Ultrasound

Pelvic abnormalities can be delineated more clearly

us-ing transvaginal ultrasound because of the proximity of

the probe to the pelvic organs and the relatively thin

vaginal vault, which enables the use of higher sound

frequencies and therefore higher resolution On the

other hand, the transabdominal probe will give a more

panoramic view of the pelvis, showing the

interrela-tionship of the major anatomic structures in the pelvic

organs and their pathology The transvaginal probe is

able to furnish a more focused image of the organ of

interest but permits effective imaging to no more than

7–10 cm in depth The transvaginal ultrasound study

is best done while the bladder is empty Since many

patients on dialysis are not able to fill their bladders

unless a Foley catheter is placed and fluid instilled into

the bladder, it makes sense to first perform a

trans-vaginal ultrasound if the pelvic pathology is suspected

and proceed to transabdominal pelvic sonogram if the

information needed cannot be obtained with the

trans-vaginal approach CAPD patients should have the

ab-domen full for transabdominal ultrasound and empty

for transvaginal ultrasound

X MANAGEMENT

The management of gynecological cancers and

non-malignant tumors in women with chronic renal failure

includes surgical excisions and chemotherapy

A Surgery

The general approach to performing surgery in dialysis

patients is discussed elsewhere There are several

ad-ditional points that pertain specifically to gynecologicalprocedures In patients with peritoneal catheters under-going pelvic or abdominal operations, the catheter can

be left in place unless there is bacterial contamination

of the peritoneal cavity When there is a low but surable risk of peritoneal contamination as in a vaginalhysterectomy, 1.0 g of vancomycin hydrochloride and1.0 g of cefoxitin can be administered prophylactically,

mea-IV, just prior to surgery If the patient is known to be

colonized with Pseudomonas, tobramycin 2.0 mg/kg

IV should be added to the prophylactic regimen operatively, the catheter is irrigated with 500 mL ofperitoneal dialysis solution three times daily to main-tain patency Irrigations are decreased to once dailywhen the fluid is no longer bloody The patient can bemaintained on hemodialysis for 10 days to 2 weeksbefore the peritoneal catheter is used again

Post-B Chemotherapy

Use of chemotherapeutic agents in dialysis patients isdiscussed elsewhere Chemotherapeutic agents havebeen given via the intraperitoneal route in patients withnormal renal function who have intraabdominal tumorsand occasionally in peritoneal dialysis patients Intra-peritoneal installation of chemotherapeutic agents re-sults in local concentrations of drugs 10–20 timeshigher than systemic levels as well as high portal veinconcentrations Drugs that have been used intraperito-neally include 5-fluorouracil (5 FU), cisplatinum, cy-tarabine, and doxorubicin 5-FU and doxyrubicin can

be given in the usual doses intraperitoneally, but thedose of cisplatinum should still be reduced to 25% ofthe usual dose

C Transplantation After Curative Resection

of Gynecological Neoplasms

Because immunosuppression increases the risk of mor, most transplant centers wait 2–5 years beforetransplanting a patient who has had a malignancy Earlystages of cervical cancer do not contraindicate trans-plantation, but transplantation after treatment, thought

tu-to be curative, of other tumors must be individualizedaccording to prognosis

XI GYNECOLOGICAL INFECTIONS

Female dialysis patients are subject to the same tions that occur in women without renal disease Somechanges in treatment are required because of the effect

infec-of renal failure and dialysis on drug metabolism

Candida albicans is the most common cause of

vul-vovaginitis Treatment is not affected by either renal

failure of dialysis Similarly, the treatment of

nonspe-cific vaginitis is not changed Metronidazole should be

taken after dialysis There have been rare cases of

fun-gal peritonitis with torulopsis resulting from vaginal

infections in peritoneal dialysis patients

A Chlamydia and Mycoplasma

These organisms are often the cause of nonspecific

vag-initis that does not respond to metronidazole therapy

In addition, they are major causes of infertility and

pel-vic inflammatory disease Treatment is to administer

doxycycline 100 mg daily for 14 days Other

tetracy-clines should be avoided in dialysis patients

Alterna-tive regimens include a single 1 g dose of

azithromy-cin, ofloxacin 150 mg daily for 7 days, or erythromycin

500 mg po qid for 14 days Only ofloxacin and

doxy-cycline also treat gonorrhea Sexual partners should

also be treated

B Genital Herpes

Oral acyclovir has been shown to shorten the intensity

and duration of first-time infections with genital herpes

Acyclovir is normally excreted by the kidney and is

dialyzable When herpetic infection is severe enough

to warrant use of acyclovir, the drug should be given

in a reduced dosage of 200 mg po bid, with the doses

scheduled in such a way that one is normally given

after a dialysis session

C Gonorrhea

In many locations, ceftriaxone has become the initial

drug of choice because of the increased incidence of

penicillin-resistant gonococci The one-time 250 mg

IM dosage is not changed for dialysis patients

Treat-ment with penicillin follows usual dosage regimens

Probenecid, included in the usual regimen to retard

re-nal excretion of penicillin, need not be given when

treating dialysis patients If the patient is allergic to

penicillin, doxycycline in the usual dosage can be

ad-ministered Therapy of resistant strains should be

guided by local information, and sensitivity results

D Syphilis

The treatment of syphilis is unchanged in the dialysis

patient Staff should be aware that secondary syphilis

is highly contagious through blood contact Dialysis

machines should be cleaned with formaldehyde or dium hypochlorite solution after use in a patient withsecondary syphilis

XII CONCLUSION

With the improved survival of women treated with alysis, our care of them should address obstetric andgynecological problems When pregnancy occurs, it re-quires extreme vigilance to minimize the risk to themother and careful management to increase the likeli-hood of a successful outcome With increased under-standing of the causes of infertility in women with re-nal failure and development of management strategiesthat result in healthy mothers and infants, we mayreach the point of actively helping these women at-tempt pregnancy Management of hormonal abnormal-ities, estrogen-replacement therapy, and common gy-necological problems, including dysfunctional uterinebleeding, tumors, and infection, needs to be incorpo-rated into our treatment of these women

di-REFERENCES

1 Perez RJ, Lipner H, Abdulla N, Cicotto S, Abrams M.Menstrual dysfunction of patients undergoing hemodi-alysis Obstet Gynecol 1978; 51:552–555

2 Holley JL, Schmidt RJ, Bender FH, Dumler F, Schiff

M Gynecologic and reproductive issues in women ondialysis Am J Kidney Disease 1997; 29:685–690

3 Souqiyyeh MZ, Huraib SO, Saleh AGM, Aswad S.Pregnancy in chronic hemodialysis patients in the King-dom of Saudi Arabia Am J Kidney Disease 1992; 19:235–238

4 Okundaye IB, Abrinko P, Hou SH A registry for nancy in dialysis patients Am J Kidney Dis 1998; 31:766–773

preg-718 Hou and Grossman

5 Bagon JA, Martens J, Van Roost G, Vernaeve H, De

Muylder X Pregnancy and dialysis Am J Kidney Dis

1998; 31:756–765

6 Registration Committee of the European Dialysis and

Transplant Association Successful pregnancies in

women treated by dialysis and kidney transplantation

Br J Obstet Gynecol 1980; 87:839–845

7 Hou SH Pregnancy in women on haemodialysis and

peritoneal dialysis Baillie`re’s Clin Obstet Gynaecol

1994; 8:481–500

8 Schwartz A, Post KG, Keller F, Molzhan M Value of

human chorionic gonadotropin measurements in blood

as a pregnancy test in women on maintenance

hemo-dialysis Nephron 1985; 39:341–343

9 Hanssens M, Keirse MJNC, Vankelecom F, Van Assche

FA Fetal and neonatal effects of angiotensin converting

enzyme inhibitors during pregnancy Obstet Gynecol

1991; 78:128–135

10 Briggs GG, Freeman RK, Yaffe SJ, eds Drugs in

Preg-nancy and Lactation Baltimore: Williams and Wilkins,

1994

11 Cockburn J, Moar VA, Ounsted MK, Good FJ, Redman

CWG Final report of study on hypertension during

pregnancy: the effects of specific treatment on growth

and development of the children Lancet 1982; 1:647–

649

12 Gladstone GW, Hordof A, Gersony WM Propranolol

administration during pregnancy: effects on the fetus

Pediatrics 1975; 86:962–964

13 Pruyn SC, Phelan JP, Buchanan GC Long term

pro-pranolol therapy in pregnancy: maternal and fetal

out-come Am J Obstet Gynecol 1979; 135:485–489

14 Cottle MKW, Van Petten GR, van Muyden P Maternal

and fetal cardiovascular indices during fetal hypoxia

due to cord compression in chronically cannulated

sheep Am J Obstet Gynecol 1983; 146:678–685

15 Plouin PF, Breart G, Maillard F, Relier JP, Breart G,

Papiernik E, the Labetalol Methyl Dopa Study Group

Comparison of antihypertensive efficacy and perinatal

safety of labetalol and methyldopa in the treatment of

hypertension in pregnancy: a randomized controlled

trial Br J Obstet Gynecol 1988; 95:868–876

16 Horvath JS, Phippard A, Korda A, Henderson-Smart

DJ, Child A, Tiller DJ Clonidine hydrochloride-a safe

and effective antihypertensive agent in pregnancy

Ob-stet Gynecol 1985; 66:634–638

17 Waisman GD, Davis N, Davey DA, et al Magnesium

plus nifedipine: Potentiation of hypotensive effect in

preeclampsia? Am J Obstet Gynecol 1988; 159:308–

309

18 Dynder SW, Cardwell MS Neuromuscular blockade

with magnesium sulfate and nifedipine Am J Obstet

Gynecol 1988; 161:35–36

19 Kaler SG, Patrinos ME, Lambert GH, Myers TF,

Karl-man R, Anderson CL Hypertrichosis and congenital

anomalies associated with maternal use of minoxidil

Pediatrics 1987; 79:434–436

20 Bhorat IE, Naidoo DP, Rout CC, Moodley J Malignantventricular arrhythmias in eclampsia: a comparison oflabetolol with dihydralazine Am J Obstet Gynecol1993; 168:1292–1296

21 Sibai BM Drug therapy: treatment of hypertension inpregnant women N Engl J Med 1996; 335:257–265

22 Hedstrom S, Martens MG Antibiotics in pregnancy.Clin Obstet Gynecol 1993; 36:886–892

23 Jacobi P, Ohel G, Szylman P, Levit A, Lewin M, Paldi

E Continuous ambulatory peritoneal dialysis as the mary approach in the management of severe renal in-sufficiency in pregnancy Obstet Gynecol 1992; 79:808–810

pri-24 Gadallah MF, Ahmad B, Karubian F, Campese VM.Pregnancy in patients on chronic ambulatory peritonealdialysis Am J Kidney Dis 1992; 20:407–410

25 Tison A, Lozowy C, Benjamin A, Usher R, Pritchard

S Successful pregnancy complicated by peritonitis in

a 35-year-old CAPD patient Perit Dial Int 1996; 16:S489–S491

26 Hou SH, Orlowski J, Pahl M, Ambrose S, Hussey M,Wong D Pregnancy in women with end stage renaldisease: treatment of anemia and premature labor Am

J Dis Kidney 1993; 21:16–22

27 Redrow M, Cherem L, Elliot J, Mangalat J, Mishler RE,Bennet WM, Lutz M, Sigala J, Byrnes J, Phillipe M,Hou S, Schon D Dialysis in the management of preg-nant patients with renal insufficiency Medicine 1988;67:199–208

28 Roma˜o JE, Luders C, Kahhale S, Pascoal IJF, Abensur

H, Sabbaga E, Zugaib M, Marcondes M Pregnancy inwomen on chronic dialysis Nephron 1998; 78:416–422

29 Ginsberg JS, Kowalchuk G, Hirsh J, Brill-Edwards P,Burrows R Heparin therapy during pregnancy ArchIntern Med 1989; 149:2233–2236

30 Lester GL: Cholecalciferol and placental calcium fer Fed Proc 1986; 2524–2527

trans-31 Marx SJ, Swart EG, Hamstra AJ, DeLuca HF Normalintrauterine development of the fetus of a woman re-ceiving extraordinarily high doses of 1,25, dihydroxy-vitamin D3 J Clin Endocrinol Metab 1980; 51:1138–1142

32 Brookhyser J, Wiggins K Medical nutrition therapy inpregnancy and kidney disease Adv Ren ReplacementTher 1998; 5:53–63

33 Czeizel AE, Duda´s I Prevention of the first occurrence

of neural tube defects by periconceptional vitamin plementation N Engl J Med 1992; 327:1832–1835

sup-34 Pitkin RM and Committee on Nutritional Status DuringPregnancy and Lactation, Institute of Medicine, Na-tional Academy of Sciences, eds Nutrition DuringPregnancy, Washington, DC: National Academy Press,1990

35 Hussey MJ, Pombar X Obstetric care for renal allograftrecipients or for women treated with hemodialysis or

peritoneal dialysis during pregnancy Adv Ren

Replace-ment Ther 1998; 5:3–13

36 Bommer J, Kugel M, Schwobel B, Ritz E, Barth HP,

Seelig R Improved sexual function during recombinant

human erythropoietin therapy Nephrol Dial Transplant

1990; 5:204–207

37 Lim VS, Kathpalia SC, Frohman LA

Hyperprolacti-nemia and impaired pituitary response to suppression

and stimulation in chronic renal failure: reversal after

transplantation J Clin Endocrin Metab 1979; 48:101–

107

38 Hou SH, Grossman S, Molitch ME

Hyperprolactine-mia in patients with renal insufficiency and chronic

re-nal failure requiring hemodialysis or continuous

am-bulatory peritoneal dialysis Am J Kidney Dis 1985; 6:

245–249

39 Gomez F, De La Cueva R, Wauters JP,

Lemarchand-Be´raud T Endocrine abnormalities in patients

under-going long-term dialysis Am J Med 1980; 68:522–530

40 Muir JW, Besser GM, Edwards CRW, Rees LH, Cattell,

Ackrill P, Baker LRI Bromocriptine improves reduced

libido and potency in men receiving maintenance

he-modialysis Clin Nephrol 1983; 26:308–314

41 Saag KG, Emkey R, Schnitzer TJ, Brown JP, Hawkins

F, Goemaere S, Thamsborg G, Liberman UA, Delmas

PD, Malice MP, Czachur M, Daifotis AG, coid-Induced Osteoporosis Study Group Alendronatefor the prevention and treatment of glucocorticoid-in-duced osteoporosis N Engl J Med 1998; 339:292–299

Glucocorti-42 Harnett JD, et al Recurrent hemoperitoneum in womenreceiving continuous ambulatory peritoneal dialysis.Ann Intern Med 1987; 107:341

43 Poole CL, et al Aseptic peritonitis associated with struation and ovulation in a peritoneal dialysis patient

men-In Khanna R, et al., eds Advances in Continuous bulatory Peritoneal Dialysis Toronto: Peritoneal Dial-ysis Bulletin, 1987

Am-44 Hampton HL, Whitworth NS, Cowan BD pin-releasing hormone agonist (leuprolide acetate) in-duced ovarian hyperstimulation syndrome in a womanundergoing intermittent hemodialysis Fertil Steril1991; 55:429

Gonadotro-45 Nakamura Y, Yoshimura Y Treatment of uterine onyomias in premenopausal women with gonadotropin-releasing hormone agonists In Pitkin RM and Scott JR,

her-ed Clinical Obstetrics and Gynecology 36: 9/93

Automated plasma exchange was originally performed

with centrifugation devices used in blood blanking

pro-cedures These devices offer the advantage of allowing

for selective cell removal (cytapheresis) (1) Plasma

exchange can also be performed with a highly

perme-able filter and standard dialysis equipment, a technique

often referred to as membrane plasma separation

(MPS) (2) A detailed review of the available removal

systems has been provided by Sowada et al (3)

Centrifugal systems utilize G forces to separate the

plasma into its different components Separation of the

plasma can be either intermittent or continuous In the

intermittent system, whole blood is collected into a

re-ceptacle (bowl) and centrifuged down to its plasma and

cellular components After separation, the cellular

com-ponents are resuspended in an appropriate amount of

replacement solution (e.g., albumin, fresh frozen

plasma) and subsequently returned to the patient

Newer devices utilize a continuous flow system in

which the whole blood is processed in an ongoing,

on-line manner

Separation of plasma from the blood’s cellular

com-ponents can also be accomplished by filtration through

a highly permeable membrane This methodology

sep-arates the blood into its cellular and noncellular

com-ponents by subjecting it to sieving through a membrane

whose pores allow the plasma proteins to pass but that

retain the larger cellular elements within the blood

path Configuration of the semi-permeable membrane

can be in a layered flat plate design (3), rolled in a tube(4), or in bundles of hollow fibers (2) The hollow fiberconfiguration can be used with standard dialysis equip-ment with the filter connected to the blood pump andpressure monitoring system while the dialysis machine

is utilized in its ‘‘isolated’’ ultrafiltration mode, whichbypasses the dialysate proportioning system

B Selective Plasmapheresis

Many imaginative techniques have been designed toselectively remove a particular pathogenic substancefrom the plasma, allowing the majority of the plasma

to be returned to the patient, thus minimizing the risks

of depletion coagulopathy and mia (Table 1) (5)

hypogammaglobuline-Cascade filtration or double filtration plasmapheresis

is a selective method of plasma fractionation in whichthe whole plasma separated from the cellular compo-nents is refiltered through a secondary filter with asmaller pore size in order to separate out the larger,unwanted molecules (6) This type of selective removalwill limit the amount of replacement fluid required byallowing most of the smaller molecules, such as albu-min (60,000 daltons), to return to the patient Thismethodology has been used to selectively remove therelatively large␤-lipoproteins (approximately 1 milliondaltons), IgM (900,000 daltons), cryoglobulins, and im-mune complexes

Cryofiltration is a technique in which the removedplasma is subjected to cooling, causing certain patho-genic substances to aggregate, thus increasing theiroverall size and allowing for efficient secondary filtra-

Table 1 Plasmapheresis Techniques

tion (7) The process can be used to selectively remove

cyroglobulins and immune complexes

Selective lipid-removal techniques are employed for

the treatment of hypercholesterolemia and can limit the

loss of non–lipid-containing plasma proteins and the

desirable high-density lipoprotein (HDL) cholesterol

Of these, three have undergone extensive clinical trials

One is an immunoadsorbant system in which plasma is

perfused over sepharose beads coated with antibodies

against low-density lipoprotein (LDL) (8) Another is

a dextran sulfate system by which negatively charged

dextran molecules are covalently bound to the

posi-tively charged apoprotein B lipoproteins (9), and a third

is known as the HELP system and involves the

extra-corporeal precipitation of LDL lipoproteins by

nega-tively charged heparin (10)

C Immunoadsorbant Techniques

There are several commercially available systems for

selective immunoadsorption of a variety of targets

These systems may be designed for nonselective

ad-sorption of immunoglobulins, such as those employing

protein A, or for more selective targets, such as those

mentioned above for the specific immunoadsorption of

LDL cholesterol

Protein A is a 42,000 dalton protein released from

certain strains of Staphylococcus aureus, which can be

used for the ex vivo adsorption of three of the four

classes of IgG (1, 2, and 4) Binding occurs at a

par-ticular site on the heavy chain of the immunoglobulin,

leaving binding sites for complement and antigens

un-affected (11) These devices may work by modulation, with activation of immune modulators, or

immuno-by net removal of immunoglobulin

Selective adsorption of endotoxin can be plished by filters impregnated with polymyxin B, anantibiotic that has the particular propensity to bind en-dotoxin fragments (12)

accom-D Anticoagulation

Regardless of the technique employed, therapeuticplasma exchange (TPE) will normally require someform of anticoagulation in order to avoid clottingwithin the extracorporeal circuit For centrifugal tech-niques, this is often provided by citrate infusions,which bind ionized calcium in the extracorporeal cir-cuit such that the coagulation cascade is impeded Theionized calcium level returns towards its original level

as the blood is returned to the intravascular ment where there are substantial stores of ionized cal-cium and where the citrate will be metabolized to bi-carbonate Rapid infusions of citrate may exceed thepatient’s capability to metabolize citrate and may lead

compart-to hypocalcemia and alkalosis (see below) Membraneplasma separation, using plasma-permeable filters,commonly employs heparin anticoagulation in a man-ner analogous to that used for hemodialysis

II COMPLICATIONS OF TPE

Several reviews have outlined the potential risks ofTPE (13–15), but there have been only a few largeseries that allow the clinician to assess the incidence

of these complications (16–24) Reports from thesenine series, involving more than 15,000 TPE treat-ments, reveal that the most common complications arecitrate-induced parethesias, muscle cramps, and urti-caria (Table 2) (23) Serious complications are reported

at a rate of 0.025–0.2% and include life-threateninganaphylactoid reactions, which are most commonly as-sociated with the use of plasma-containing replacementfluids (e.g., fresh frozen plasma, purified protein frac-tion) (25) The overall incidence of death is 0.05%, butsome of these ‘‘treatment-associated’’ deaths were inpatients with severe preexisting conditions, and theTPE treatment per se may not have been the precipi-tating factor

A Citrate-Induced Hypocalcemia

During TPE, citrate may be infused either as the coagulant or in the fresh frozen plasma (FFP) admin-

anti-Complications During Plasma Exchange 723

Table 3 Percent Decrease inSerum Levels of CoagulationFactors After a SinglePlasma Exchange

Source: Modified from Ref 33.

istered as the replacement fluid Symptoms of

citrate-induced hypocalcemia represent one of the most

common complications and can occur in up to 9% of

treatments (23) The incidence is highest in those

treat-ments utilizing FFP as the replacement fluid, since this

preparation is approximately 15% citrate by volume

Most often the patient will complain of perioral or

dis-tal extremity tingling or paresthesias If severe,

citrate-induced hypocalcemia may be associated with

prolon-gation of the QT interval on electrocardiogram, thus

increasing the risk of cardiac arrhythmia (26,27)

Widely used protocols suggest that citrate toxicity

can be reasonably well controlled with the oral

admin-istration of calcium tablets during the procedure,

re-serving intravenous calcium replacement only for those

who develop symptoms Another conservative

ap-proach involves decreasing the rate of plasma exchange

and decreasing the citrate to blood ratio and

supple-menting with heparin (24,28,29) Others have found

that prophylactic replacement of intravenous calcium

can significantly reduce the incidence of

citrate-induced paresthesias (23,30,31) In an in-depth review

of the topic, Hester et al concluded that the incidence

of citrate-induced hypocalcemic symptoms could be

re-duced if the citrate-infusion rate was limited to between

1.0 and 1.8 mg/kg/min (26) If symptoms occurred

de-spite this limitation, they recommended an infusion of

10 mL of 10% calcium gluconate infused over 15

minutes approximately halfway through the procedure

Kinetic studies have demonstrated that increases in

parathyroid hormone provide an endogenous

compen-satory response to calcium removal during TPE (28),

but patients receiving multiple treatments with albumin

replacement may experience a significant loss of

cal-cium amounting to approximately 150 mg per

treat-ment (31) In contrast, with suppletreat-mentation calcium

balance can be positive

Uhl et al described a case of severe citrate toxicity

when the citrate infusion line became disengaged from

its rotary pump allowing a massive infusion of citrate

into the patient (32) Seven minutes into the procedure,

the patient developed signs and symptoms suggesting

severe hypocalcemia, including muscle spasms, chest

pain, and hypotension Ionized calcium level was 0.64

mmol/L (normal range, 1.18–1.38 mmol/L)

B Coagulation Abnormalities

1 ‘‘Depletion’’ Coagulopathy

Albumin solutions used for replacement fluid are

de-void of clotting factors, and a TPE treatment with

al-bumin as the replacement fluid will result in a depletion

of all coagulation factors, including fibrinogen and tithrombin III (AT-III) (25,33,34) After a single plasmaexchange, the serum levels of most of these factors will

an-be decreased by approximately 40–60% (Table 3) rum levels of these factors rebound in a biphasic man-ner, characterized by a rapid initial increase in the first

Se-4 hours after treatment, followed by a slower increaseover the next few days (23) This dual rate of recoveryrepresents two phenomena: a reequilibration of extra-vascular stores with the intravascular compartment and

a resynthesis of new clotting factors Twenty-four hoursafter treatment, fibrinogen levels are 50% and AT-IIIlevels are 85% of initial levels, while both factors mayrequire 48–72 hours for complete recovery (23) Pro-thrombin time (PT) increases 30% and partial throm-boplastin time (PTT) doubles immediately after a oneplasma volume exchange (31,35) Partial thromboplas-tin time and thrombin time are back to normal range 4hours postpheresis, while prothrombin time normalizes

in 24 hours (25)

When multiple treatments are performed over ashort period (three or more treatments per week), thedepletion in clotting factors is more pronounced andmay require several days for spontaneous recovery(33–35) Under these conditions, the risks of hemor-rhage can be minimized by substituting between 500and 1000 mL (2–4 units) of FFP as the replacementfluid towards the end of the procedure This approach

is most helpful in patients who are immediately surgery (e.g., thymectomy for myasthenia gravis), whohave had a recent renal biopsy (e.g., glomerulonephri-tis), who have active hemoptysis (Goodpasture’s syn-drome or Wegener’s granulomatosis), or in whom there

post-is a desire for the immediate removal of a large-boreintravascular catheter

Complications During Plasma Exchange 725

2 Thrombocytopenia

Thrombocytopenia may result from a loss of platelets

in the discarded plasma, as a result of thrombosis

within the plasma filter, as a consequence of

heparin-induced antiplatelet antibodies, or as a result of a mild

dilutional effect by the infusion of 5% albumin

solu-tion, which is relatively hyperoncotic compared to the

removed plasma (36) With the older centrifugal

ma-chines such as the Haemonetics V-50 (Haemonetics,

Braintree, MA), inefficient separation of the different

plasmatic components resulted in platelet losses with

the discarded plasma, and these treatments have been

associated with decreases in platelet counts of up to

50% The newer centrifugal devices provide more

ef-ficient separation of the plasmatic components and a

more modest loss of platelets Membrane plasma

sep-aration (MPS) can result in a 15% decrease in platelet

count, possibly related to partial thrombosis within the

filter (1,25,31,37) Because heparin is more commonly

used as the anticoagulant, heparin-induced antiplatelet

antibodies are also more likely to occur with MPS

3 Anemia

Posttreatment decreases in hematocrit may result from

hemorrhage associated with the vascular access, from

substantial clotting in the extracorporeal circuit, or

from treatment-related hemolysis Initiation of

treat-ment with a membrane plasma separator is often

as-sociated with a minimal amount of plasma tinting,

which is rarely a cause for significant blood loss and

can be quantified by measuring the free hemoglobin

levels in the collected plasma In most cases, plasma

tinting lasts for only a few seconds; if persistent, the

blood flow should be slowed in order to lower the

transmembrane pressure Hemolysis can also occur in

centrifugal systems as a result of hypotonic priming

solutions Even in the absence of any extracorporeal

losses or hemolysis, hematocrits may decrease by 10%

after each treatment, a phenomenon that may be due

to intravascular expansion related to the use of

rela-tively hyperoncotic replacement fluids (5% albumin)

(31,36,38)

4 Thrombosis

TPE treatments using albumin replacement will cause

a relative depletion of all coagulation factors, including

inhibitors of coagulation such as AT-III In one report,

two episodes of thrombosis were associated with a

postpheresis depletion of AT-III and this deficiency

may have resulted in a hypercoaguable state (39)

Thrombosis has also been associated with the longed use of indwelling vascular catheters (23) Pul-monary embolism, cerebral ischemia, and myocardialinfarction have been reported to occur in associationwith TPE, but the incidence is rare (0.06–0.14%)(18,40) An association with low levels of AT-III isspeculative, especially since these patients will alsohave a concomitant depletion of ‘‘pro’’ coagulant fac-tors (see above)

pro-C Infections

Aside from the infections related to indwelling vascularcatheters, the risk of infection associated with TPE can

be divided into two broad categories: those that may

be the result of a posttreatment depletion of globulins, a situation most likely to occur when thereplacement fluid is mostly albumin, and those that oc-cur as a result of viral transmission from the replace-ment fluid, most likely to occur when the replacementfluid is fresh frozen plasma

immuno-1 Postpheresis InfectionTPE using albumin as the replacement fluid will result

in a predictable decline in levels of immunoglobulinsand complement and may predispose patients to highrates of infection One plasma volume exchange willresult in a 60% reduction in serum immunoglobulinlevels and a net 20% reduction in total body immu-noglobulin stores (31,36) Multiple treatments overshort periods, especially when associated with immu-nosuppressive agents, will yield more substantive de-creases in immunoglobulin levels that may persist forseveral weeks (41,42) Although concentrations of C3and C4 may be reduced by a series of daily treatments,because of rapid resynthesis (short half-lives), levels ofthese proteins rebound within several days CH50 can

be predictably lowered to about 40% of its initial valueimmediately after a given treatment but rebound to pre-treatment values occurs within one day, and even re-petitive daily treatments have a minimal effect on thisparameter (41) Therapeutic plasma exchange with FFPreplacement would not be expected to deplete immu-noglobulin or complement levels

The incidence of infection in patients undergoingTPE varies widely Wing et al compared the incidence

of infection in patients with rapidly progressiveglomerulonephritis (RPGN) who received standardtherapy (steroids and cytotoxic agents) with or withoutplasma exchange (43) The apheresis-treated group had

a higher occurrence of infection, but some of the

con-Table 4 Risk of Transmitted Viral Infections per UnitTransfused in the Mid-1990s

Source: Refs 31–33.

trol cases were taken from retrospective review and

granulocytopenia was present in two of the five

TPE-treated patients who developed infections, a condition

that was more likely the result of the cytotoxic therapy

than from the plasma exchange procedure per se In

other studies of RPGN, 9 episodes of infection were

found in 34 patients treated with TPE and

immunosup-pression, 4 of which resulted in death (44–52) Two of

these nine patients were granulocytopenic (44,50)

Pa-tients treated with TPE for myasthenia gravis appear to

have a lower incidence of infection than those treated

for renal disease (53–57) Of thirty-six patients with

myasthenia gravis treated with TPE in addition to

pred-nisone and azathioprine, only one patient developed an

infectious complication after a mean follow-up period

of 9 months (55)

There has been only one prospective, randomized

trial that has attempted to disassociate the risk of

in-fection associated with TPE and that associated with

the commonly co-administered immunosuppressive

medication In this study, Pohl et al studied 86 patients

with lupus nephritis receiving cyclophosphamide and

steroids, with or without TPE (58) These investigators

found no increase in the rate of infection or in

infec-tion-related deaths in the apheresis-treated group In

patients treated with TPE, the infection rate was 1.22

infections per 200 weeks with three deaths, compared

with 1.15 infections per 200 weeks and four deaths in

the control group Thus, in the only prospective study

in which the effect of TPE could be isolated from that

of drug-induced immunosuppression, treatment with

TPE was not associated with any increased risk of

infection

Although the study by Pohl et al did not find an

increased risk of infection due to the addition of TPE

to an already aggressive immunosuppressive regimen,

there remains the possibility that immunoglobulin or

complement depletion may impair a patient’s ability to

combat an ongoing infection Thus, if a severe

infec-tion develops in the immediate postpheresis period, a

reasonable approach would be to reconstitute normal

immunoglobulin levels with a single infusion of IVIG

(100–400 mg/kg intravenously) (23), similar to the

re-placement dose recommended in patients with hypo- or

agammaglobulinemia (59,60) Because of the relatively

long half-life of IgG (21 days), this approach will

pro-vide normalized immunoglobulin levels for several

weeks, provided there are no further TPE treatments

2 Risk of Viral Transmission

Risk of viral transmission during plasma exchange is

directly related to replacement with FFP or other

plasma-containing solutions (e.g., purified protein tion, cryosupernatant) Albumin preparations aretreated with heat and are considered to be devoid oftransmissible virus (61) The same claims had beenmade for intravenous immunoglobulins, but an out-break of hepatitis C from contaminated IVIG has beendocumented (62), promoting the initiation of newmethodologies for avoiding viral transmission fromIVIG preparations (63) In a review of listed compli-cations from over 15,000 treatments, there was onlyone reported case of latent non-A/non-B hepatitis in-fection (19) Transmission of the human immunodefi-ciency virus (HIV) through therapeutic plasmapheresis

frac-is unlikely and would be anticipated as the result ofinfected FFP (64) There has been a documented case

of HIV transmission to a patient treated with TPE forGuillain-Barre´ syndrome, but the report dates from aperiod in which screening for HIV-infected plasma wasless well established (65) The current incidence oftransfusion-acquired viral infections has declined sub-stantially from the early 1980s and is currently esti-mated as 1/63,000 units for hepatitis B, 1/100,000 unitsfor hepatitis C, 1/680,000 units for HIV, and 1/641,000units for HTLV (66–68) (Table 4) This risk of viraltransmission can be further reduced with the use ofsolvent detergent (SD)–treated plasma SD-treatedplasma is a cell-free, blood group–specific, coagula-tion-active human plasma, which has been treated in amanner to irreversibly inactivate the lipid envelope ofviruses such as HIV1 and 2, HBV, and HCV and hasbeen approved by European guidelines and the U.S.Food and Drug Administration (FDA) (69,70) SinceFFP is normally provided in units of 200–300 mL, asingle plasma volume exchange will involve the infu-sion of 10–15 units Thus, the most compelling indi-cation for the use of SD-treated plasma would be in anoninfected individual who is being treated for throm-

Complications During Plasma Exchange 727

botic thrombocytopenic purpura, in whom a common

treatment protocol may involve multiple treatments and

numerous units of FFP

III COMPLICATIONS RELATED TO

REPLACEMENT FLUIDS

A General Comments

In the United States, albumin is the most commonly

used replacement fluid and, when compared to FFP, has

the advantage of lacking viral transmission and

pos-sessing a decreased risk of anaphylactoid reactions

Disadvantages include a posttreatment coagulopathy

related to the removal of clotting factors and a net loss

of immunoglobulins A 5% concentration of albumin

will provide a reasonable replacement of the oncotic

pressure removed with the patient’s plasma (see

be-low) Some centers prefer to dilute the albumin to

ap-proximately 3.5%, a solution that is hypo-oncotic to

the plasma being removed and may render the patient

more prone to hypotension Dilution of albumin with

sterile water, as opposed to saline, has been associated

with hemolysis and should be avoided (71)

FFP contains all the noncellular components of

nor-mal blood and does not lead to postpheresis

coagulop-athy or immunoglobulin depletion FFP is also

consid-ered essential for the treatment of thrombotic

thrombocytopenic purpura (TTP) since TPE for this

in-dication may be most useful as a means of providing

a missing serum factor (72) Disadvantages include

an-aphylactoid reactions (most often mild, but can be

life-threatening), citrate toxicity, and a small, but persistent

risk of viral transmission Because of these potential

problems, FFP should be avoided except for the

treat-ment of TTP/HUS or when hemorrhagic risks are great

Plasma protein fraction (PPF) contains

approxi-mately 87% albumin and 13%␣- and␤-globulins and

is easier and less costly to prepare than albumin

Al-though difficult to prove, the risks of anaphylactoid

re-action with PPF are probably less than for that of FFP

Nonetheless, PPF has been associated with hypotensive

episodes and circulatory collapse, possibly due to the

presence of prekallikrein activator and bradykinin (61)

As with the use of FFP, concomitant treatment with

ACE inhibitors should be avoided

B Reactions to Protein-Containing

Replacement Fluids

Reactions to plasma-containing fluids (FFP, PPF,

cryosupernatant) are anaphylactoid in nature and are

characterized by fever, rigors, urticaria, wheezing, andhypotension and may eventually progress to laryngo-spasm (73,74) This type of pulmonary distress isclearly in distinction to that of the pulmonary edemathat may accompany fluid overload, in which the pa-tient is often hypertensive and is unlikely to have as-sociated urticaria, wheezing, and laryngospasm In areview of several large series, the reported incidence ofthis type of reaction was between 0.02% and 21% (23)(Table 2) Most reactions are limited to urticaria andrigors, but the potential for life-threatening reactions isunderscored by the list of 42 TPE-associated deathscompiled by Huestis, at least 30 of which were asso-ciated with FFP replacement (75) Similarly, Aufeuvre

et al reported seven deaths in 6200 treatments ing nonhemodynamic pulmonary edema associatedwith FFP replacement (17,40) Sutton et al., in theirreview of over 5000 treatments from the Canadian RedCross, reported 8 patients with severe reactions com-prised of severe urticaria, itching, shortness of breath,and wheezing and noted that all 8 patients had beenreceiving plasma (76)

includ-Human serum albumin consists of 96% albumin andtrace amounts of ␣- and ␤-globulins, and, as opposed

to FFP, anaphylactoid reactions are rare and may beassociated with the formation of antibodies to po-lymerized albumin created by heat treatment or stabi-lization with sodium caprylate (61,77) Recent reportshave suggested that patients taking ACE inhibitors mayalso been prone to an increased risk of ‘‘atypical’’ orhypotensive reactions to albumin (78,79)

Potential mechanisms triggering the anaphylactoidreactions described above include; (1) the presence ofanti-IgA antibodies in a patient who is IgA deficientand who is receiving IgA-containing fluids (i.e FFP,immunoglobulins); (2) contamination of the replace-ment fluid with bacteria, endotoxins or pyrogens; (3)the presence of a prekallikrein activator and bradyki-nin; and (4) the formation of antibodies to polymerizedalbumin (80)

C Management

Because of the relative high incidence of these tions, patients undergoing massive replacement withFFP (i.e for thrombotic thrombocytopenic purpura orhemolytic uremic syndrome) are commonly pretreatedwith 50 mg of diphenhydramine (Benadry) In thosepatients who have already demonstrated a sensitivity toFFP, and in whom FFP replacement is obligatory (i.e.TTP), a successful prophylactic regimen has included

reac-50 mg of prednisone given 13 hours, 7 hours and 1

hour before the treatment, 50 mg of diphenhydramine

given 1 hour before the treatment and 25 mg of

ephe-drine given 1 hour before the treatment (80)

Epineph-rine should be available in the event of a severe life

threatening reaction (laryngeal edema, etc)

IV REACTIONS TO SELECTIVE

REMOVAL TECHNIQUES

A Protein A Columns

Protein A is a 42,000 dalton protein released from

cer-tain strains of Staphylococcus aureus which can be

at-tached to sepharose, collodion charcoal, or silica and

can be used for the ex vivo adsorption of three of the

four classes of IgG (1, 2 and 4) The Prosorba protein

A column is a single-use, nonregenerating system

which is placed in series with a standard plasma

exchange circuit and is FDA approved for idiopathic

thrombocytopenic purpura (ITP)

Secondary effects during use of the Prosorba protein

A column are common In one large series involving

142 patients and 1306 treatments, 79% of patients

ex-perienced at least one episode of toxicity during the

procedure (81) The most common side effects were

fever, chills, and musculoskeletal pains, but more

se-vere reactions, such as hypotension, were also noted

These secondary effects may result from the release of

activated complement products and seem to be the

ba-sis for the recommendation that plasma perfused over

the device not be reinfused into the patient at a rate

exceeding 20 mL/min For similar reasons, the device

should not be used in patients who are currently taking

ACE inhibitors, since these drugs block the degradation

of bradykinins and may result in a severe anaphylactoid

reactions as treated plasma is reinfused into the patient

(78) A recent report suggests that this treatment was

the cause of a systemic vasculitis (82)

B Selective Lipid Removal

There are three conceptually different methods for the

selective removal of cholesterol-containing

lipopro-teins An evaluation of all three of these techniques

found them to be equally biocompatible and equally

efficacious in lowering LDL-associated cholesterol

(83) Two reports, however, have warned of

anaphy-lactoid reactions in patients treated with the dextran

sulfate–based system in whom there was concurrent

treatment with ACE inhibitors (84,85)

V ATYPICAL REACTIONS ASSOCIATED WITH ACE INHIBITORS

Aside from their effect on the angiotensin system, ACEinhibitors will also block the degradation of bradyki-nins and may result in severe anaphylactoid reactions

if a given apheresis procedure results in the activation

of kinins or if there is a high concentration of thesefactors in the replacement fluid Flushing, hypotension,abdominal cramping, and occasionally severe anaply-lactoid reactions have been reported with the use of thedextran sulfate system for selective lipid removal(84,85), and in patients treated with the IMRE Prosorbacolumn (85) A recent report describes such atypicalreactions in those patients receiving albumin replace-ment during standard apheresis (78) In one large review,all (100%) of fourteen patients who were receiving ACEinhibitor therapy during apheresis procedures experi-enced atypical reactions defined as flushing or hypo-tension (decrease of 20 mmHg or more) (79) In con-trast, only 20 (7%) of 285 patients not receiving ACEinhibitors developed atypical reactions (p < 0.001) Theauthors concluded that ACE inhibitors should be with-held for at least 24 hours before apheresis

VI ELECTROLYTE ABNORMALITIES

A Hypokalemia

Five percent albumin solutions are isosmotic to plasmaand contain less than 2 mEq/L of potassium (61) As aresult, a 25% reduction in serum potassium levels mayoccur in the immediate postpheresis period (86), andthere is the potential for hypokalemic arrhythmias dur-ing apheresis and in the immediate postpheresis period.Experience with hemodialysis suggests that this type ofhypokalemic arrhythmia is most likely to occur in thepresence of digoxin and when potassium levels ap-proach 2 mEq/L (87) Considering the above, we fol-low a protocol by which we add 4 mEq/L of potassium

to each liter of 5% albumin

B Alkalosis

Citrate, infused either as the anticoagulant or in thereplacement fluid, will be metabolized to bicarbonateand can result in a substantial alkalemia Formula Bcitrate solution (Fenwal, Baxter, Deerfield, IL) contains

73 mmol/L of citrate, which will yield 219 mmol/L ofbicarbonate, while formula A citrate solution contains

112 mmol/L of citrate which will yield 336 mmol/L ofbicarbonate Fresh frozen plasma contains approxi-

Complications During Plasma Exchange 729

mately 14% citrate by volume In most patients,

post-pheresis bicarbonate levels are unchanged (86) In

pa-tients with renal failure, severe alkalemia may result

from repeated treatments, especially when FFP is used

as the replacement (88) In a most challenging

situa-tion, a patient with hemolytic uremic syndrome may

require massive amounts of FFP while suffering from

severe renal failure In this case, alkalemia may

neces-sitate frequent dialysis in order to remove the excess

bicarbonate Because of this postpheresis alkalemia, if

TPE and hemodialysis are required on the same day, it

is preferable to perform the TPE first in order to allow

the dialysis treatment to correct the citrate-induced

alkalemia

C Aluminum

All albumin solutions are contaminated with between

4 and 24 mmol/L of aluminum (89,90), and repetitive

TPE with albumin may result in significant

accumula-tion of aluminum In patients with severe renal

insuf-ficiency, 60%–70% of infused aluminum is retained

(90) Bone deposition of aluminum has also been

re-ported in a patient with normal renal function (89,90)

VII VITAMIN REMOVAL

Blood concentrations of vitamins B12, B6, A, C, and E

and ␤-carotene have been noted to decline between

24% and 48% after a single TPE treatment, but rebound

to pretreatment levels occurs within 24 hours (91)

Pos-sibly because of their large volumes of distribution

as water-soluble vitamins, folate, thiamine, nicotinate,

biotin, riboflavin, and pantothenate are not significantly

altered by plasma exchange The long-term effects of

repetitive treatments is not known, but net removal of

the protein-bound vitamin B12is approximately 900␮g

per treatment (31), and there is the potential for a

sub-stantial reduction in total body stores after repetitive

treatments

VIII HYPOTENSION

The reported incidence of hypotension during TPE is

1.7% (Table 2), but the actual incidence is probably

higher and dependent on its definition Hypotension

during TPE may occur for a variety of reasons,

includ-ing delayed or inadequate volume replacement,

vaso-vagal episodes, hypo-oncotic fluid replacement,

ana-phylaxis, cardiac arrhythmia, bradykinin reactions

(e.g., reactions to ACE inhibitors), vascular

access–in-duced external or internal hemorrhage, and cular collapse (Table 5) (these issues are discussed indetail in the previous sections) Discontinuous flowplasma exchange systems may be prone to a higherincidence of hypotensive episodes due to intermittenthypovolemia The use of hypo-oncotic replacement so-lutions may also increase the risk of hypotensiveevents A commonly employed preparation of replace-ment fluid is to dilute albumin with an electrolyte so-lution to achieve a concentration of 3.5% albumin Inmost patients this solution is clearly hypo-oncotic tothe patient’s plasma and may predispose them to hy-potension, even when a policy of 1:1 volume replace-ment is rigorously followed An undiluted 5% albuminsolution is less likely to produce a hypo-oncotic hy-povolemia, except when pretreatment plasma volumesare abnormally expanded by a hyperviscosity state(Waldenstrom’s macroglobulinemia)

cardiovas-IX MISCELLANEOUS COMPLICATIONS

Respiratory arrest due to apnea has been reported lowing plasma exchange in patients who had been an-esthetized with succinylcholine (40,92) Succinylcho-line is an anesthetic agent that is metabolized bycholinesterase, and these apneic events were considered

fol-to be the result of abnormally low posttreatment levels

of plasma cholinesterase Cholinesterase levels are duced by 50% immediately after a single treatment(93) Levels less than 30% of normal (approximately

re-1000 U/L) are likely to be associated with decreasedmetabolism of succinylcholine (94,95) Since FFP con-tains normal levels of cholinesterase, depletion of thisenzyme is only expected when albumin or PPF is used

as replacement Anesthetic agents dependent on serumcholinesterase for their metabolism should be used withcaution immediately post–plasma exchange, especiallyafter a series of daily treatments Repletion of cholin-esterase with FFP may be a reasonable approach fortreatments in the immediate perioperative period.Volume-resistant hypotension, bronchospastic dysp-nea, and chest pain may occur secondary to comple-ment-mediated membrane bioincompatibility, similar tothose described during hemodialysis (96) Anaphylac-toid symptoms may also occur due to ethylene oxidesensitivity, which is used as a sterilizing agent (97).The incidence of filter-related leukocytopenia, throm-bocytopenia, and hypo-complementemia is reducedwith more biocompatible membranes, and reactions toethylene oxide can be avoided with adequate priming

of the filter (98) Severe hemolysis has occurred as a

Table 5 Potential Causes for Hypotension During TPE

Delayed or inadequate volume replacement

Vasovagal episodes

Hypo-oncotic fluid replacement

3.5% albumin solutions

Anaphylaxis

Reactions to plasma components in replacement fluids

Anti-IgA antibodies (IgA deficient patient)

Endotoxin-contaminated replacement fluid

Reactions to bioincompatible membranes

Sensitivity to ethylene oxide

Device related: Prosorba protein A column

Cardiac arrhythmia

Citrate-induced hypocalcemia

hypokalemic related (especially in patients on digitalis)

Bradykinin reactions (e.g., reactions to ACE inhibitors)

Hemorrhage

Associated with primary disease (ITP, factor VIII

inhibitors)

Associated with heparin anticoagulation

Associated with vascular access

Guillain-Barre´ syndrome (autonomic dysfunction)

Waldenstrom’s macroglobulinemia (rapid decrease in

plasma volume)

result of inappropriately hypotonic priming solutions

and from excessively high transmembrane pressure

during membrane plasma separation Chills and other

symptoms of hypothermia may be experienced due to

inadequately warmed replacement fluid and can be

avoided by warming the replacement solutions to body

temperature (Table 6)

X DEATHS

In a review of the literature from 1983, Huestis

com-piled a total of 42 deaths associated with TPE (75) Of

these 42, at least 30 were associated with FFP

replace-ment, 6 were identified as occurring with albumin or

PPF, while the replacement solution was uncertain in

the remaining 5 The major causes of death were

car-diovascular, respiratory, and anaphylactic

Unfortu-nately, details about these deaths and their temporal

relationship to the TPE procedure were not noted In

total, Huestis calculated an estimated 3 deaths per

10,000 procedures

In a review of 6200 treatments, Aufeuvre et al ported a total of 7 deaths (17,40) Causes of death in-cluded nonhemodynamic pulmonary edema (FFP re-placement), cardiac dysrhythmia, hemodynamicpulmonary edema, and pulmonary embolism In an ex-tensive review of several large series involving over15,500 treatments, there was a total of 8 deaths, for acalculated incidence of 0.05% (23)

re-XI DRUG REMOVAL

A General Comments

When compared to what is known for hemodialysis orperitoneal dialysis (99), there is little published infor-mation regarding the removal of therapeutic agents byTPE During plasma exchange, alterations in plasmadrug levels are most dependent on the percentage ofprotein binding and the volume of distribution(31,100,101) Thus, a drug with a high percentage ofprotein binding and a relatively modest volume of dis-tribution (<0.3 L/kg) will have the greatest likelihood

of being removed by TPE Using first-order kineticsand assuming the simplest case, the volume of plasmaexchanged during a TPE treatment would have to equal0.7 times the volume of distribution of a drug in order

to remove 50% of its total body burden Thus, a TPEtreatment would have to exchange 7 L of plasma inorder to remove 50% of a drug whose volume of dis-tribution is a modest 0.15 L/kg (approximately 10 L in

a 70 kg patient) As a result, even a drug with a centage protein binding of over 90% would be mini-mally removed if its volume of distribution wasⱖ0.6L/kg (ⱖ42 L in a 70 kg patient) It is therefore notsurprising that a 3 or 4 L TPE treatment is not com-monly used for drug intoxications, despite the fact thatmany drugs are very highly protein bound

per-Drug removal information for several drugs are viewed in the following paragraphs Since there is alarge variability in drug kinetics between individuals,

re-it is recommended that, when possible, all daily drugdosing should be administered after the TPE treatment

B Specific Drugs

Indications for plasma exchange often involve the comitant administration of steroids and immunosup-pressive medication Prednisone has a relatively largevolume of distribution (1 L/kg), and despite a proteinbinding of between 70 and 95% neither it nor its me-tabolite, prednisolone, is significantly removed byplasma exchange (102) Cyclophosphamide is only

con-Complications During Plasma Exchange 731

Table 6 Management Strategies to Treat and Avoid Complications of TPE

mg orally 1 h pretreatment and before pheresisThrombocytopenia Consider membrane plasma separation

Volume-related hypotension Consider continuous flow separation with matched input and output: consider

increasing protein concentration of replacement fluidInfection postpheresis Infusion of intravenous immunoglobulin (100–400 mg/kg)

Hypokalemia Ensure potassium concentration of 4 mmol/L in replacement solution

Membrane biocompatibility Change membrane or consider centrifugal method of plasma separation

Source: Adapted from Ref 23.

12% protein bound, and its volume of distribution is

relatively large (0.8 L/kg), suggesting a minimal

re-moval by TPE (100) Azathioprine is approximately

30% protein bound, with a volume of distribution of

0.6 L/kg, and its removal would also be expected to be

minor

Most aminoglycosides have a low degree of protein

binding (<5%) and a volume of distribution

approxi-mating 0.25 L/kg, thus one would not predict a

sub-stantial removal by TPE Only 7–10% of a 100 mg

dose of tobramycin was removed after two TPE

treat-ments equaling 1700 and 2170 mL (103) Similarly,

only 4–6% of body stores were removed after TPE

treatments of 1725 and 2057 mL (100)

Phenytoin has a variable volume of distribution of

between 0.5 and 1 L/kg and has a substantial

intra-erythrocytic distribution Lui and Rubenstein reported

that approximately 10% of total body stores were

re-moved during TPE treatments of 5.6 and 6.1 L (104),

suggesting that a posttreatment supplement may be

re-quired Depending on initial serum concentrations,

which varied between 8 and 17 ␮g/mL, the total

amount removed ranged from 42 and 93 mg per

treat-ment Data obtained after a 3 L exchange demonstrated

a net 30 mg removal (31)

Digoxin, with an enormous volume of distribution

(5–8 L/kg) and a modest degree of protein binding

(20–30%), is predictably unaffected by TPE (100), but

removal of digibind-bound drug may be enhanced in

patients with renal failure (105) Digitoxin has a greater

degree of protein binding (94%), but its volume of

dis-tribution, although far less than that of digoxin (0.6 L/kg), is still too great to allow for a substantial net re-moval, and TPE has not been found to significantlylower its total body stores (100) Despite the modestnet removal, TPE may still be useful as a treatment forintoxications since cardiac toxicity may be reduced be-cause of the rapid lowering of serum levels (106,107).Twenty-five percent of the active hormone thyroxinecirculates in the intravascular compartment and is 99%bound to serum protein, leading to the use of TPE totreat thyroid storm when conventional methods havefailed (108)

Acetylsalicylic acid is 90% protein bound and has amodest volume of distribution (0.1–0.2 L/kg), and TPEcan remove substantial amounts (100) Although itslarge volume of distribution (2.8–4 L/kg) would sug-gest that net removal would be minimal, TPE has beenreported to reduce the half-life of propranolol (90–96%protein bound) by approximately 75% (109) Vanco-mycin is only 10–50% protein bound and has a volume

of distribution ranging between 0.5 and 1.1 L/kg A oneplasma volume exchange has been found to removeonly 6% of total body stores, with a substantial post-treatment rebound (110)

Cisplatin is 90% protein bound and has an estimatedvolume of distribution of only 0.5 L/kg, suggesting thatTPE may be useful in the management of cisplatinoverdose Two such cases have been reported with TPEresulting in substantial lowering of pretreatment levelsfrom 2979 to 185 ng/mL and from 2900 to 200 ng/mL(111,112)

Theophylline is 55% protein bound and has a

mod-est volume of distribution (0.4–0.7 L/kg), Laussen et

al reported on the use of arterio-venous and

veno-ve-nous (pumped) plasma exchange in three children with

theophylline toxicity, suggesting that plasmapheresis

may be useful in increasing drug clearance and

de-creasing its half-life (113)

REFERENCES

1 Gurland HJ, Lysaght MH, Samtleben W, Schmidt B

A comparison of centrifugal and membrane based

apheresis formats Int J Artif Organs 1984; 7:35–38

2 Gurland HJ, Lysaght MJ, Samtleben W, Schmidt B

Comparative evaluation of filters used in membrane

plasmapheresis Nephron 1984; 36:173–182

3 Sowada K, Malchesky PS, Nose Y Available removal

systems: state of the art In: Nydegger UE, ed

Ther-apeutic Hemapheresis in the 1990s Basel-Karger,

1990:51–113

4 Kaplan AA, Halley SE, Reardon J, Sevigny J One

year’s experience using a rotating filter for therapeutic

plasma exchange Trans Am Soc Artif Intern Organs

1989; 35:262–264

5 Malchesky PS, Kaplan AA, Coo AP, Sadurada Y,

Siami GA Are selective macromolecule removal

plas-mapheresis systems useful for autoimmune diseases or

hyperlipidemia? ASAIO J 1993; 39:868–872

6 Agishi T, Kaneko I, Hasuo Y, Hayasaka Y, Sanaka T,

Ota K, Abe M, Ono T, Kawai S, Yamane K Double

filtration plasmapheresis Trans Am Soc Artif Intern

Organs 1980; 26:406–409

7 Vibert GJ, Wirtz SA, Smith JW, et al Cryofiltration

as an alternative to plasma exchange: plasma

macro-molecular solute removal without replacement fluids

In: Nose Y, Malchesky PS, Smith JW, eds

Plasma-pheresis Cleveland: ISAO Press, 1983:281–287

8 Saal SD, Parker TS, Gordon BR Removal of

low-density lipoproteins in patients by extracorporeal

im-munoadsorption Am J Med 1986; 80:583–589

9 Gordon BR, Kelsey SF, Bilheimer DW, Brown DC,

Dau PC, Gotto AM Jr, Illingworth DR, Jones PH,

Leit-man SF, Prihoda JS, et al Treatment of refractory

fa-milial hypercholesterolemia by low density lipoprotein

apheresis using an automated dextran sulfate cellulose

adsorption system Am J Cardiol 1992; 70:1010–

1016

10 Eisenhauer T, Armstrong VW Schuff-Werner P, et al

Long term clinical experience with

HELP-CoA-Re-ductase inhibitors for maximum treatment of coronary

heart disease associated with severe

hypercholeste-rolemia Trans Am Soc Artif Intern Organs 1989; 35:

580–583

11 Samtleben W, Schmidt B, Gurland HJ Ex vivo and in

vivo protein A perfusion: background, basic

investi-gations and first clinical experience Blood Purif 1987;5:179–192

12 Hanasawa, K, Aoki H, Yoshioka T, Matsuda K, Tani

T, Kodama M Novel mechanical assistance in thetreatment of endotoxic and septicemic shock Am SocArtif Intern Organs Trans 1989; 35:341–343

13 Isbister JP The risk/benefit equation for therapeuticplasma exchange In: Nydegger UE, ed TherapeuticHemapheresis in the 1990s Basel: Karger AG, 1990:10–30

14 Hazards of apheresis (editorial) Lancet 1982; 2:1025–1026

15 Westphal RG: Complications of hemapheresis In:Westphal RG, Kasprisin DO, eds Current Status ofHemapheresis: Indications, Technology and Compli-cations Arlington, VA: American Association ofBlood Banks, 1987:87–104

16 Borberg H Problems of plasma exchange therapy In:Gurland HJ, Heinze V, Lee HA, eds TherapeuticPlasma Exchange New York: Springer-Verlag, 1980:191–201

17 Aufeuvre JP, Morin-Hertel F, Cohen-Solal M, Lefloch

A, Baudelot J Hazards of plasma exchange In: berth HG, ed Plasma Exchange Stuttgart: FK Schat-tauer Verlag, 1980:149–157

Sie-18 Ziselman EM, Bongiovanni MB, Wurzel HA Thecomplications of therapeutic plasma exchange VoxSang 1984; 46:270–276

19 Fabre M, Andreu G, Mannoni P Some biologicalmodifications and clinical hazards observed duringplasma exchanges In: Seiberth HG, ed PlasmaExchange Stuttgart: FK Schattauer Verlag, 1980:143–148

20 Rossi PL, Cecchini L, Minichella G, De Rosa G, fano G, Pieralla L, Testa A, Candido A, Vittorio M,Mango G Comparison of the side effects of therapeu-tic cytapheresis and those of other types of hema-pheresis Haematologica 1991; 76(suppl 1):75–80

Al-21 Sprenger KBG, Rasche H, Franz HE Membraneplasma separation: complications and monitoring Ar-tif Organs 1984; 8:360–363

22 Samtleben W, Hillebrand G, Krumme D, Gurland HJ.Membrane plasma separation: clinical experience withmore than 120 plasma exchanges In: Sieverth HG, ed.Plasma Exchange Stuttgart: FK Schattauer Verlag,1980:23–27

23 Mokrzycki MH, Kaplan AA Therapeutic plasmaexchange: complications and management Am J Kid-ney Dis 1994; 23:817–827

24 Sutton DMC, Nair RC, Rock G, and the CanadianApheresis Study Group Complications of plasmaexchange Transfusion 1989; 29:124–127

25 Flaum MA, Cuneo RA, Appelbaum FR, Disseroth AB,Engel WK, Gralnick HR The hemostatic imbalance

of plasma exchange transfusion Blood 1979; 54:694–702

Complications During Plasma Exchange 733

26 Hester JP, McCullough J, Mishler JM, Szymanski IO

Dosage regimens for citrate anticoagulants J Clin

Apheresis 1983; 1:149–157

27 Olson PR, Cox C, McCullough J Laboratory and

clin-ical effects of the infusion of ACD solution during

platelet-pheresis Vox Sang 1977; 33:79–87

28 Silberstein LE, Naryshkin S, Haddad JJ, Strauss JF

Calcium homeostasis during therapeutic plasma

exchange Transfusion 1986; 26:151–155

29 Huestis DW Complications of therapeutic apheresis

In: Pinada AA, Valbonesis M, Diggs JC, eds

Thera-peutic Hemapheresis Milan: Wichtig Editore, 1986:

179–186

30 Buskard NA, Varghese Z, Wills MR Correction of

hypocalcemic symptoms during plasma exchange

Lancet 1976; 2:344–345

31 Kaplan AA, Halley SE Plasma exchange with a

ro-tating filter Kidney Int 1990; 38:160–166

32 Uhl L, Maillet S, King S, Kruskall MS Unexpected

citrate toxicity and severe hypocalcemia during

apher-esis Transfusion 1997; 37:1063–1065

33 Chrinside A, Urbaniak SJ, Prowse CV, Keller AJ

Co-agulation abnormalities following intensive plasma

exchange on the cell separator Br J Haematol 1981;

48:627–634

34 Gelabert A, Puig L, Maragall S, Monteagudo J,

Cas-tillo R Coagulation alterations during massive

plas-mapheresis In: Sieverth HG, ed Plasma Exchange

Stuttgart: FK Schattauer Verlag, 1980:71–75

35 Kaplan AA Therapeutic plasma exchange for the

ne-phrologist: Semin Dial 1995; 8:294–298

36 Chopek M, McCullough J Protein and biochemical

changes during plasma exchange In: Ulmas J,

Berk-man E, eds Therapeutic Hemapheresis: A Technical

Workshop Washington, DC: American Association of

Blood Banks, 1980:13–52

37 Keller AJ, Chirnside A, Urbaniak SJ Coagulation

ab-normalities produced by plasma exchange on the cell

separator with special reference to fibrinogen and

platelet levels Br J Haematol 1979; 42:593–603

38 Wood L, Jacobs P The effect of serial therapeutic

plasmapheresis on platelet count, coagulation factors

plasma immunoglobulin and complement levels J

Clin Apheresis 1986; 3:124–128

39 Sultan Y, Bussel A, Maisonneuve P, Sitty X, Gajdos

P Potential danger of thrombosis after plasma

exchange in the treatment of patients with immune

disease Transfusion 1979; 19:588–593

40 Aufeuvre JP, Mortin-Hertel F, Cohen-Solal M, Lefloch

A, Baudelot J Clinical tolerance and hazards of

plasma exchanges: a study of 6200 plasma exchanges

in 1033 patients In: Beyer JH, Burgerg H, Fuchs C,

Nagel GA, eds Plasmapheresis in Immunology and

Oncology Basel: Karger, 1982:65–77

41 Keller AJ, Urbaniak SJ Intensive plasma exchange on

the cell separator Effects on serum immunoglobulins

and complement components Br J Haematol 1978;38:531–540

42 Kaplan AA Towards a rational prescription of plasmaexchange: The kinetics of immunoglobulin removal(editorial) Semin Dial 1992; 5:227–229

43 Wing EJ, Bruns FJ, Fraley DS, Segel DP, Adler S.Infectious complications with plasmapheresis in rap-idly progressively glomerulonephritis JAMA 1980;244:2423–2426

44 Lockwood CM, Pinching AJ, Sweny PM, Rees AJ,Pussell B, Uff J, Peters DK Plasma exchange andimmunosuppression in the treatment of fulminatingimmune-complex crescentic nephritis Lancet 1977; 1:63–67

45 Rossen RD, Hersh EM, Sharp JT, McCredie KB,Gyorkey F, Suki WN, Eknoyan G, Reisberg MA Ef-fect of plasma exchange on circulating immune com-plexes and antibody formation in patients treated withcyclophosphamide and prednisone Am J Med 1977;63:674–682

46 Lockwood CM, Pearson TA, Rees AJ, Evans DJ, ters DK, Wilson CB Immunosuppression and plasma-exchange in the treatment of Goodpasture’s syndrome.1976; Lancet 1:711–715

Pe-47 Johnson JP, Whitman W, Briggs WA, Wilson CB.Plasmapheresis and immunosuppressive agents in an-tibasement membrane antibody-induced Goodpas-ture’s syndrome Am J Med 1978; 64:354–359

48 Depner TA, Chafin ME, Wilson CB, Gulyassy PF.Plasmapheresis for severe Goodpasture’s syndrome(abstr) Kidney Int 1976; 8:409

49 Lang CH, Brown DC, Staley N, Johnson G, Ma KQ,Border WA, Dalmasso AP Goodpasture’s syndrometreated with immunosuppression and plasmaexchange Arch Intern Med 1977; 137:1076–1078

50 Rosenblatt SG, Knight W, Bannayan GA, Wilson CB,Stein JH Treatment of Goodpasture’s syndrome withplasmapheresis Am J Med 1979; 66:689–696

51 Walker RG, Dapice AJF, Becker GJ, Kincaid-Smith P,Craswell PW Plasmapheresis in Goodpasture’s syn-drome with renal failure Med J Aust 1979; 1:875–879

52 McKenzie PE, Taylor AE, Woodroffe AJ, Seymour

AE, Chan YL, Clarkson AR Plasmapheresis inglomerulonephritis Clin Nephrol 1979; 12:97–108

53 Pinching AJ, Peters DK, Davis JN Remission of asthenia gravis following plasma-exchange Lancet1976; 2:1373–1376

my-54 Dau PC, Lindstrom JM, Cassel CK, Denys EH, Shev

EE, Spitter LE Plasmapheresis and sive drug therapy in myasthenia gravis N Engl J Med1977; 297:1134–1140

immunosuppres-55 Behan PO, Shakir RA, Simpson JA, Burnett AK, lan TL, Haase G Plasma-exchange combined with im-munosuppressive therapy in myasthenia gravis Lancet1979; 2:438–440

Al-56 Newsom-Davis J, Wilson SG, Vincent A, Ward CD.

Long-term effects of repeated plasma exchange in

my-asthenia gravis Lancet 1979; 1:464–468

57 Winklestein A, Volkin RL, Starz TW, Maxwell NG,

Spero JA The effects of plasma exchange on

immu-nologic factors (abstr) Clin Res 1979; 27:691

58 Pohl MA, Lan SP, Berl T, and the Lupus Nephritis

Collaborative Study Group Plasmapheresis does not

increase the risk for infection in immunosuppressed

patients with severe lupus nephritis Ann Intern Med

1991; 114:924–929

59 Consensus on IVIG Lancet 1990; 1:470–472

60 Haas A Use of intravenous immunoglobulin in

im-munoregulatory disorders Ann Intern Med 1987; 107:

367–382

61 Finlayson JS Albumin products Semin Thromb

He-most 1980; 6:85–120

62 Bjoro K, Froland SS, Yun Z, Samdal HH, Haaland T

Hepatitis C infection in patients with primary

hypo-gammaglobulinemia after treatment with contaminated

immune globulin N Engl J Med 1994; 331:1607–

1611

63 Schiff RI Transmission of viral infections through

in-travenous immune globulin (editorial) N Engl J Med

1994; 331:1649–1650

64 Kiprov D, Simpson D, Romanick-Schmiedl S, Lippert

R, Spira T, Busch D Risk of AIDS-related virus

(hu-man immunodeficiency virus) transmission through

apheresis procedures J Clin Apheresis 1987; 3:143–

146

65 Boucher CA, de Gans J, van Oers R, Danner S,

Goudsmit J Transmission of HIV and AIDS by

plas-mapheresis for Guillain-Barre´ syndrome Clin Neurol

Neurosurg 1988; 90:235–236

66 Lackritz EM, Satten GA, Aberle-Grasse J, Dodd RY,

Raimondi VP, Janssen RS, et al Estimated risk of

transmission of the human immunodeficiency virus by

screened blood in the United States N Engl J Med

1995; 333:1721–1725

67 Schreiber GB, Busch MP, Kleinman SH, Korelitz JJ

The risk of transfusion-transmitted virus invections

The Retrovirus Epidemiology Donor Study N Engl J

Med 1996; 334:1685–1690

68 AuBuchon JP, Birkmeyer JD, Busch MP Safety of the

blood supply in the United States: Opportunities and

Controversies Ann Intern Med 1997; 127:904–909

69 Biesert L, Suhartono H Solvent/detergent treatment of

human plasma—a very robust method for virus

in-activation Validated virus safety of OCTAPLAS Vox

Sang 1998; 74(suppl 1):207–212

70 Klein HG, Dodd RY, Dzik WH, Luban NL, Ness PM,

Pisciotto P, Schiff PD, Snyder EL Current status of

solvent/detergent-treated frozen plasma Transfusion

1998; 38:102–107

71 Steinmuller DR A dangerous error in the dilution of

25 percent albumin (letter) N Eng J Med 1998; 338:

1226–1227

72 Rock GA, Shumak KH, Buskard NA, Blanchette VS,Kelton JG, Nair RC, Spasoff RA, and the CanadianApheresis Study Group Comparison of plasmaexchange with plasma infusion in the treatment ofTTP N Engl J Med 1991; 325:393–397

73 Ring J, Messmer K Incidence and severity of phylactoid reactions to colloid volume substitutes.Lancet 1977; 1:466–469

ana-74 Bambauer R, Jutzler GA, Albrecht D, Keller HE, ler M Indications of plasmapheresis and selection ofdifferent substitution solutions Biomater Artif CellsArtif Organs 1989; 17:9–27

Koh-75 Huestis DW Mortality in therapeutic haemapheresis(letter) Lancet 1983; 1:1043

76 Sutton DMC, Nair R, Rock G, and the CanadianApheresis Study Group Complications of plasmaexchange Transfusion 1989; 29:124–127

77 Stafford CT, Lobel SA, Fruge BC, Moffitt JE, Hoff

RG, Fadel HE Anaphylaxis to human serum albumin.Ann Allergy 1988; 61:85–88

78 Brecher ME, Owen HG, Collins ML Apheresis andACE inhibitors (letter) Transfusion 1993; 33:963–964

79 Owen HG, Brecher ME Atypical reactions associatedwith use of angiotensin-converting enzyme inhibitorsand apheresis Transfusion 1994; 34:891–894

80 Apter AJ, Kaplan AA An approach to immunologicreactions with plasma exchange J Allergy Clin Im-munol 1992; 90:119–124

81 Snyder Jr HW, Henry DH, Messerschmidt GL, et al.Minimal toxicity during protein A immunoadsorptiontreatment of malignant disease: an outpatient therapy

J Clin Apheresis 1991; 6:1–10

82 Case presentation N Engl J Med 1994; 331:792–799

83 Schaumann D, Olbricht CJ, Welp M, et al poreal removal of LDL-cholesterol: prospective eval-uation of effectivity, selectivity and biocompatibility(abstr) J Am Soc Nephrol 1992; 3:392

Extracor-84 Olbricht CJ, Schauman D, Fisher D Anaphylactoidreactions, LDL apheresis with dextran sulphate andACE inhibitors Lancet 1992; 3:908–909

85 Kroon AA, Mol MJTM, Stalenhoff APH ACE itors and LDL-apheresis with dextran sulphate adsorp-tion Lancet 1992; 340:1476

inhib-86 Orlin JB, Berkman EM Partial plasma exchange usingalbumin replacement: removal and recovery of normalplasma constituents Blood 1980; 56:1055–1059

87 Morrison G, Michelson EL, Brown S, Morganroth J.Mechanism and prevention of cardiac arrhythmias inchronic hemodialysis patients Kidney Int 1980; 17:811–819

88 Pearl RG, Rosenthal MH Metabolic alkalosis due toplasmapheresis Am J Med 1985; 79:391–393

89 Mousson C, Charhon SA, Ammar M, Accominotti M,Rifle G Aluminum bone deposits in normal renalfunction patients after long-term treatment by plasmaexchange Int J Artif Organs 1989; 23:664–667

Complications During Plasma Exchange 735

90 Milliner DS, Shinaberger JH, Shurman P, Coburn JW

Inadvertent aluminum administration during plasma

exchange due to aluminum contamination of albumin

replacement solutions N Engl J Med 1985; 312:165–

167

91 Reddi A, Frank O, DeAngelis B, Jain R, Bashruddin

I, Lasker N, Baker H Vitamin status in patients

un-dergoing single or multiple plasmapheresis J Am Coll

Nutr 1987; 6:485–489

92 MacDonald R, Robinson A Suxamethonium apnea

as-sociated with plasmapheresis Anaesthesia 1980; 35:

96 Jorstad S Biocompatibility of different hemodialysis

and plasmapheresis membranes Blood Purif 1987; 5:

123–137

97 Nicholls AJ, Platts MM Anaphylactoid reactions due

to haemodialysis, haemofiltration or membrane plasma

separation BMJ 1982; 285:1607–1609

98 Aeschbacher B, Haeverli A, Nydegger UE Donor

safety and plasma quality in automated

plasmaphere-sis Vox Sang 1989; 57:104–111

99 Bennet WM, Aronoff GR, Golper TA, Morrison G,

Singer I, Brater DC Drug Prescribing in Renal

Fail-ure 2nd ed Philadelphia: American College of

Phy-sicians, 1991

100 Sketris IS, Parker WA, Jones JV Effect of plasma

exchange on drug removal In: Valbonesi M, Pineda

AA, Biggs JC, eds Therapeutic Hemapheresis Milan:

Wichtig Editore, 1986:15–20

101 Jones JV The effect of plasmapheresis on therapeutic

drugs Dial Transplant 1985; 14:225–226

102 Stigelman WH, Henry DH, Talbert RL, Townsend RJ

Removal of prednisone and prednisolone by plasma

exchange Clin Pharmacol 1984; 3:402–407

103 Appelgate R, Schwartz D, Bennett WM Removal oftobramycin during plasma exchange therapy Ann In-tern Med 1981; 94:820–821

104 Liu E, Rubenstein M Phenytoin removal by pheresis in thrombotic thrombocytopenic purpura.Clin Pharmacol Ther 1982; 31:762–765

plasma-105 Rabetoy GM, Price CA, Findlay JW, Sailstad JM.Treatment of digoxin intoxication in a renal failurepatient with digoxin-specific antibody fragments andplasmapheresis Am J Nephrol 1990; 10:518–521

106 Peters U, Risler T, Grabenese B Digitoxin elimination

by plasma separation In: Sieberth HG, ed PlasmaExchange: Plasmapheresis-Plasmaseparation Stutt-gart: FK Schattauer Verlag, 1980:365–368

107 Arsac Ph, Barret L, Chenais F, Debru JL, Faure J.Digitoxin intoxication treated by plasma exchange In:Sieberth HG, ed Plasma Exchange: Plasmapheresis-Plasmaseparation Stuttgart: FK Schattauer Verlag,1980:369–371

108 Ashkar FS, Katims RB, Smoak WM, Gilson AJ.Thyroid storm treatment with blood exchange andplasmapheresis J Am Med Assoc 1979; 214:1275–1279

109 Talbert RL, Wong YY, Duncan DB Propranololplasma concentrations and plasmapheresis Drug IntellClin Phram 1981; 15:993–996

110 McClellan SD, Whitaker CH, Friedberg RC Removal

of vancomycin during plasmapheresis Ann cother 1997; 31:1132–1136

Pharma-111 Jung HK, Lee J, Lee SN A case of massive cisplatinoverdose managed by plasmapheresis Korean J InternMed 1995; 10:150–154

112 Chu G, Mantin R, Shen YM, Baskett G, Sussman H.Massive cisplatin overdose by accidental substitutionfor carboplatin Toxicity and management Cancer1993; 72:3707–3714

113 Laussen P, Shann F, Butt W, Tibballs J Use of mapheresis in acute theophylline toxicity Crit CareMed 1991; 19:288–290

Economic Issues in Dialysis: Influence on Dialysis-Related

Complications in the Managed-Care Era

Theodore I Steinman

Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts

I INTRODUCTION

Economic issues in dialysis must be reviewed in

con-text of the overall changes occurring in our health care

system Increasing and complex pressures in our

pres-ent economic environmpres-ent have wrought changes far

beyond our expectations Medical advances have led to

decreased hospitalizations and less ‘‘need’’ for

hospi-tals, a situation that reflects better care On the other

hand, lower total income for hospitals (after salaries,

benefits, and supplies) results in less money available

to pay for teaching medical and nursing students, for

unrecovered costs of biomedical research, for the

training of tomorrow’s doctors and other providers of

care, and for the care of those for whom there is no

payer The goal of everyone involved in medicine is

to maintain superior quality of care (that has become

our standard) while attempting to decrease the costs

of care Financial pressures have assisted the

forma-tion of systems of care where the economies of scale

and efficiencies of providing care can best be

achieved Collaboration is necessary to strengthen our

delivery system in this era of shrinking health-care

dollars The public is entitled to readily available

health care of the highest quality at the most

reason-able cost Medicine must never compromise its

fun-damental obligation of care of patients and the overall

health of those we serve, regardless of their ability to

pay

II THE BASIS FOR THE GROWTH OF MANAGED CARE

When Medicare and Medicaid were enacted in 1965,

it was anticipated that successive national legislationwould result in universal health insurance coveragewithin one or, at the most, two decades The federalactuaries estimated in those early days that total outlaysfor Medicare in 1990 would come to $10 billion, a farcry from the staggering $180 billion total, which wasreached in 1996 (1) No one was able to predict theexplosive rise of national health-care total expendituresfrom $41 billion in 1965 to $1 trillion in 1996, andthere is an expected doubling of total expenditures toexceed $2 trillion dollars by the year 2007 or shortlythereafter (2) The growth of managed care was fos-tered by dramatic increases in national health-carespending since the passage of Medicare and Medicaid

In 1965 just under 6% of the gross domestic product(GNP) was expended on health care, and this had risen

to almost 14% by 1997 (2) The significant growth indollar expenditures is but a part in a much larger trans-formation in health-care spending patterns

Private health insurance accounted for mately 25% of the national total of $41 billion in 1965.Federal, state, and local governments accounted for25% of the total, and the remaining 50% representedout-of-pocket outlays by consumers (3) Consumer out-of-pocket share for health-care expenditures has de-

approxi-738 Steinman

clined progressively to a current value of around 18%,

and the public has paid little attention to their declining

total financial commitment to expenditures The

gov-ernment (primarily federal) is now contributing about

two thirds of the increased spending Private health

in-surance now pays only about 15% of the total

health-care expenditures, and indemnity insurance is rapidly

falling as a percentage of the financial pie These major

shifts reflect the steep rise of inpatient hospital care

costs, and most of these costs were covered by added

spending by government and employers It is not

sur-prising that most of the population with good private

or public health insurance coverage had little concern

with steeply climbing health-care expenditures Only to

a minor degree (approximately 20%) do personal

out-of-pocket payments become a financial factor With this

anomalous financing arrangement, the model for a

competitive marketplace has only limited applicability

This becomes ever more important based on the fact

that governmental budgets, insurance, and charity

con-tinue to provide access at some level to all persons

irrespective of their ability to pay (4)

Fiscal discipline virtually disappeared from the

health-care marketplace during the first two decades of

Medicare when the principal payers—employers,

pri-vate health insurance, and government—tended to

honor, with few or no questions, all bills submitted for

reimbursement by providers It was not until the

mid-1980s that Medicare began to pay closer attention to

dollar costs At the same time employers and private

health insurance plans looked toward managed care

plans to moderate their annual increases in premium

costs Medicare shifted from a cost-plus reimbursement

for hospital inpatient care to a prospective payment

system It took until the early 1990s for the market to

fully respond to these developments with large-scale

deceleration in employers’ annual contribution to

pre-mium increases In contrast, Medicare outlays continue

to rise at an annual rate of around 10% (1)

Examination of the health-care system infrastructure

leads one to focus on:

1 Delivery of health-care services (hospitals,

di-alysis clinics, physicians, and other health

per-sonnel)

2 Health education

3 Biomedical research

III MANAGED CARE TODAY

In the early to mid-1990s, the annual expenditure data

gave some credence to the advocates of greater reliance

on the competitive marketplace For-profit managedcare organizations (MCOs) expanded rapidly, and theirmedical loss ratio during the 1980s was often in the70–75% range, which meant that they were able toextract 25–30 cents of every premium dollar for mar-keting, expansion, and profits (including high salariesfor management) (5) Shifting potential enrollees fromfee-for-service to managed care can generate a one-time cost savings of 10–15%, but future profits maybecome more elusive In 1994–95 over 90% of for-profit managed care plans were profitable; in 1996–97less than 40% have turned a profit (6)

Equity investors helped generate the shift from for-service into for-profit health maintenance organi-zations (HMOs) since the capital they invested helpedspur the expansion in HMO enrollments A downturn

fee-in the stock market would make it difficult for heavilyindebted for-profit MCOs to survive

Most of the healthy population have already enrolled

in managed care plans, and it will be more difficult forthe for-profit companies to maintain a favorable bottomline in face of the increasing trend toward direct con-tracting between payers and providers, direct negoti-ating with Medicare by provider-sponsored organiza-tions, and the establishment by the teaching hospitals,large academic health centers, and other types ofhealth-care systems to launch independent HMOs orenter into partnerships with established HMOs.HMOs will now be forced to enroll patients withestablished chronic illnesses, and it is evident thatMCOs have little experience in caring for people withend-stage renal disease (ESRD) It is evident thatMCOs/HMOs have only a superficial idea of the truecost of caring for a patient with ESRD, and this willset the stage for a radical transformation in the health-care delivery system Some of the role of the acute carehospital will be progressively peripheralized (7) Mostpatients will require treatment in clinics, at home, and

in specialized ambulatory centers This chronically illpopulation will require access to nurses and other mid-level personnel for follow-up and ongoing care Pa-tients will need to accept increased personal responsi-bility for their own care and well-being There willneed to be a reliance on advice and reinforcement frommembership in support groups Patients’ care will need

to be directed by a management team headed by thephysician, who will in turn direct a considerable num-ber of associates and assistants The nonphysician per-sonnel will actually provide many of the services thatthe patient requires This model is already well estab-lished in the dialysis population, and therefore thenephrology community is well positioned to readily

adapt to changes wrought by MCOs/HMOs The

fol-lowing (4) summarizes the trends that characterize the

U.S health-care agenda:

No great urgency to take action to provide

health-care coverage for all the population

An initial preference to enlarge the scale and scope

of the competitive marketplace to extend its

con-trol over the financing and delivery of health-care

services Vigorous counterefforts via new federal

and state legislation will be the response to limit

the degree of managerial freedom previously

available to managed care plans

New federal legislation reducing the numbers of

present and prospective Medicaid enrollees and

encouraging parallel actions by the states to

con-tain their future outlays for Medicaid

Agreement among the political leadership to support

the Balanced Budget Act of 1997, which

man-dates a substantial reduction in Medicare outlays

in future years The Senate initiative to raise the

age for Medicare eligibility failed, as did the

income-adjusted premium rates for Medicare B

Modest changes in regulations governing private

health-care insurance aimed at facilitating

contin-ued coverage for workers changing jobs and new

federal funding to facilitate coverage for large

numbers of low-income children

With this overall backdrop, we will now focus on

ec-onomic issues specific to dialysis units

IV FINANCES RELATED TO DIALYSIS

Financial management of the entire dialysis process

re-quires expertise that evades most nephrologists Expert

help is needed for the vast majority of nephrologists to

help define all issues of cost, tracking funds

manage-ment, and assistance with contract negotiations

Finan-cial departments of the large dialysis chains generally

have systems in place to provide the above assistance

However, the issue of tracking funds management

var-ies from dialysis chain to dialysis chain True costs of

any system, especially in a hospital setting, is difficult

to understand and follow under the best circumstances

(8)

The issue of cost-effectiveness of test-treatment

strategies needs to be understood by the nephrologist

(9) In the era of cost containment, physicians need

reliable data about specific interventions A necessary

learning step for nephrologists is how to interpret

ec-onomic analyses and estimate their own costs of

im-plementing recommended interventions With the care

of the dialysis patient moving towards national lines that can be outlined in algorithms of care, thenephrologist must understand the cost implications ofeach step in an algorithm (10,11) This is a criticalcomponent of any contract negotiations Every com-ponent of care must have a cost analysis performed sothat maximum utilization of limited financial resources

guide-is employed

Estimates of costs without substantiating data willcreate problems This is especially true in the hospitalsetting, where it is difficult to fully understand truecosts in contrast to filed charges Data on expenditures,start-up costs, and general overhead are frequently ne-glected when examining the bottom line There is aneed for cost data and a standardized protocol so thatmissing data can be detected A bridge between caredelivery and economic analysis is a necessary link (12).More than half of the dialysis provided in the UnitedStates to the ESRD population is done by three verti-cally integrated mega-providers: Fresenius MedicalCare (FMC), Gambro Healthcare, and Total RenalCare Other organizations are attempting to impact themarketplace and become large-scale providers (e.g.,Renal Care Group), and their presence in the market-place is being felt The publicized rationale for con-solidation of dialysis units is to obtain economics ofscale; eliminating duplicity does reduce administrativeoverhead These large national chains have definitelychanged the delivery of care Data management hasimproved, but the doctor-patient interaction has suf-fered In many cases, the doctor-patient relationship hasbecome a nurse-patient or staff-patient relationship, areflection of decreased physician presence in the dial-ysis unit The patient has often become a pawn in the

‘‘return-on-investment’’ game that results from selling

of individually owned dialysis units to one of the tional companies (13) Economic vitality must never beconfused with quality of care as viewed from an out-come standpoint or patient satisfaction as determined

na-by surveys (e.g., Kidney Disease Quality of Life tionnaire—see below)

ques-Increasing regulation of the dialysis industry has curred in a profound fashion over the past two decades,and this is a consequence of government experienceswith escalating costs beyond anticipated projections.The cost increase is related more to providing care formore patients rather than to an increase in cost perpatient In fact, the per-patient costs have declined inreal dollars over the past 20 years, but the governmentnever really fully anticipated the large number of pa-tients who would be candidates for ESRD therapy A

oc-740 Steinman

composite rate-payment schedule for dialysis has led

to cost control as a major factor in patient care

De-clines in reimbursement have led the dialysis industry

to seek individuals skilled in business management to

help run medical operations Larger organizations

ne-gotiate better pricing for supplies Vertically integrated

organizations internally supply their own equipment

and disposables Such an internal supply line

appar-ently results in cost savings since the marketing and

distribution costs can be virtually eliminated Such

ec-onomic factors can sometimes result in the patient

be-ing ‘‘lost in the system.’’

Financial compensation for nephrologists has

dra-matically changed in the past few years A

fee-for-service reimbursement mechanisms is coming to a

close in most parts of the country and will probably

almost totally disappear within the next few years

Cap-itation for global services is being progressively

em-ployed as a method of payment, the attempt being to

encourage routine care that can be captured with a

global fee A modified fee-for-service reimbursement

mechanism may be utilized for circumstances defined

as extraordinary (14) Nephrology services

reimburse-ment can potentially be carved out from a payer/HMO/

MCO as a new method for reimbursement (15)

Ad-vantages of a financial single specialty carve-out

include the following:

1 Payer can transfer risk to the physician group

2 It is easier to negotiate a single-speciality

carve-out with the payer than a global capitation

con-tract for a random population (for all services

to be delivered)

3 Carveouts can increase practice volume and

thus physician income, assuming that the

reim-bursement is adequate

4 Participating single-specialty independent

prac-tice associations (IPAs) can get exclusive

con-tracts with the payer

5 The physician must closely observe his/her

practice efficiency

6 Long-term relationship between physician and

payer is cemented if the physician group is the

first in the market and your group’s performance

has become known to the payers

7 A contract with multispecialty IPAs means

ac-cess to more of the premium dollar because of

a demonstrated track record with payers

There must be detailed explanations of how bonuses

will be allocated within the physician group It is

crit-ical not to provide vague details that are open to

mis-interpretation In a managed care capitated

environ-ment, physicians should be eligible for incentive paybased on improved outcomes, open access to care, andpatient satisfaction results (16,17) Reimbursement in-centives should never be tied to cost savings generated

by denial of services Later in this chapter the issue ofreimbursement for physicians based on outcomes is ex-plained further

An information booklet must be kept in a dialysisunit detailing issues of financial concern involving thepatient (18) Evaluation and management services thatshould be delivered to the patient as part of the monthlycapitation payment (MCP) need to be clearly stated inwritten form for patients to understand This is a criticalreminder to physicians of their responsibilities to thepatient For physicians themselves there must be a sim-ilar booklet that notes the required documentation forevery level of service Physicians need to know howbilling is done within Medicare and other insurer reg-ulations This booklet will need to be updated because

of the rapid changes that are occurring in the area ofreimbursement The steady shift of segments of thepopulation from a fee-for-service to a capitated envi-ronment mandates constant physician informationupdating

There is a demonstrated need to avoid unnecessarytests and treatment when viewed in the context ofquality-adjusted life-year (QALY) (9) Physicians willneed to understand the issues of sensitivity analysiswhen viewing such QALY, which relates to the value

of a specific strategy (test or treatment) as viewed bythe general public (e.g., how much should society payfor an intervention when considering the benefit to thepatient and society?)

V DEMOGRAPHICS AND EXPENDITURES RELATED

TO DIALYSIS

The percent change in Medicare payments for ESRDservices have been declining more than the decline inthe medical Consumer Price Index (CPI), the bottomline being that reimbursement for ESRD care has notbeen keeping pace with medical inflation While there

is still an increase in the medical CPI, it is declining

as compared to previous years Since 1983 the paymentrates for ESRD services have remained essentially thesame With the government attempting to control over-all expenditures, measured in both actual and inflation-adjusted dollars, reimbursement for dialysis care (thecomposite rate) has dramatically declined Despite thecost per patient being less today as compared to 1983,

Table 1 Distribution of Medicare Payments

for ESRD Patients per Year (1995) at Risk

Medicare pays 80% of allowed charges Breakdown

of where payments go for services delivered and the

percentage of total payments per group.

Source: HCFA; USRDS, 1997.

Table 3 Increase in Medicare Payments by Source ofClaim, 1991–1995

Source Annualized percent increase

Source: HCFA; USRDS, 1997.

Table 4 1995 Medicare Component and TotalExpenditures for the ESRD Population

Non-Medicare componentMSP: Employer Group Health PlansPatient out-of pocket obligationsHMO payments

Organ donor acquisition

MSP = Medicare secondary payor.

Source: USRDS, 1997.

Table 2 Medicare MCP Payments by Specialty per

Patient per Year (1995) at Risk

MCP = Monthly capitation payment.

Payment schedule as distribution pieces of the pie for outpatient

services only Percentage change in payments from 1991 to 1995.

Source: HCFA; USRDS, 1997.

ESRD total expenditures is now approaching $14

bil-lion While ESRD comprises only 0.5% of the

Medi-care population, expenditures for this group amount to

5.5% of the Medicare budget There has been a 40%

growth in the number of dialysis facilities since 1990,

increasing from 2200 to approximately 3000

Expenditures for hospital inpatient care and dialysis

outpatient care are almost equal Spending per patient

varies significantly depending on modality of dialysis,

patient age, number of co-morbid conditions, and

dia-betic status Payments to nephrologists are a small

por-tion (approximately 7%) of total patient costs, but the

nephrologist receives the largest proportion of

physi-cian provider payments (as appropriate) (see Tables 1

and 2) The change in Medicare payments as reflected

by an annualized percent increase is noted on Table 3

While the percent increase appears large, the actual

to-tal dollars represents a small increase that is below that

of the adjusted medical CPI Another factor in the totalincrease in expenditures is a decline in the adjustedannual mortality rates (19) In 1989 the U.S RenalData System (USRDS) noted a 24.0% adjusted annualmortality rate By 1996 this had declined to 22.3%.Therefore, with fewer patients dying, the costs of carefor an expanding patient population have increasedoverall Table 4 provides an estimate of the total ESRDcosts for 1995 as analyzed in the 1997 USRDS AnnualReport

In an attempt to control costs in face of a movementtowards managed care, the Health Care Financing Ad-ministration (HCFA) has undertaken a Medicare Cap-itation Demonstration Project The goal is to determine

if cost savings can be achieved and quality of care proved in a capitated environment for patients withchronic illness ESRD is an obvious disease entity to

im-be studied im-because the boundaries of the illness arereadily defined Table 5 provides the HCFA reimburse-ment schedules based on patient age and diabeticstatus This payment is for yearly global services andencompasses all resources needed by the patient, in-cluding physician care in both the inpatient and out-