EPA No. 560/6-81-009
Assessing the Feasibility of Epidemiologic
Research on DEHP Exposure Among
Renal Dialysis Patients
July 1981
Final Report
U.S. Environmental Protection Agency
Office of Pesticides and
Toxic Substances
Washington, D.C.
-------
EPA No. 560/6-81-009
July 1981
ASSESSING THE FEASIBILITY OF EPIDEMIOLOGIC
RESEARCH ON DEHP EXPOSURE AMONG
RENAL DIALYSIS PATIENTS
Ira Marks, M.P.H.
Leslie Euinton, M.B.Ch.B, D.I.H., F.R.S.H.
George Shreiner, M.D.*
Dorothy Wellington, Ph.D.
Pauline Wagner, M.S.
Allen Sinclair, R.N.
Carole Steele, A.R.T.
JRB ASSOCIATES, INC.
8400 Westpark Drive
McLean, Virginia 22102
*Georgetown University Medical School
PROJECT OFFICER
JANE KELLER
TASK MANAGER
JASON TOTH
HEALTH AND ENVIRONMENTAL
REVIEW DIVISION
OFFICE OF PESTICIDES
AND TOXIC SUBSTANCES
WASHINGTON, D.C.
-------
DISCLAIMER
This project has been conducted with Federal funds from
the Environmental Protection Agency under contract
number 68-01-6280 by JRB Associates, 8400 Westpark
Drive, McLean, VA 22102. The content of this publi-
cation does not necessarily reflect the views or
policies of the U.S. Environmental Protection Agency,
nor does mention of trade names, commercial products,
or organizations imply endorsement by the U.S.
Government.
-------
ACKNOWLEDGMENTS
The authors wish to thank the many talented researchers who helped them
in this assignment. In particular, Dr. George Shreiner of Georgetown
University Medical School, Dr. Jack Mandel of the University of Minnesota
School of Public Health, Dr. Eric Marks at the National Heart, Lung, and Blood
Institute, Ms. Michael McMullan and Ms. Sandra Kappert of the Health Care
Financing Administration, DHHS, Dr. Patricia Buffler of the University of
Texas School of Public Health, Dr. Joseph Fraumeni and Dr. William Hoover of
the National Cancer Institute, DHHS, and Dr. Taichi Sugimoto of the University
of South Carolina.
-------
ABSTRACT
Di-2-ethylhexylphthalate (DEHP) is a plasticizing agent used in polyvinyl
chloride (PVC) to give the polymer its flexibility and softness, and may
account for up to 40 percent of the final weight of the PVC product. Although
insoluble in water, it is soluble in organic solvents and oils.
A two-year bioassay study performed by the National Cancer Institute
found that DEHP produced hepatocellular carcinomas and neoplastic nodules in
male and female rats, and adenomas and hepatocellular carcinomas in male and
female mice.
One human population receiving relatively high exposures to DEHP is
dialysis patients. Dialysis patients receive DEHP exposure from two sources,
blood stored in PVC blood bags and tubing used in dialysis treatment. The
Environmental Protection Agency (EPA) conducted a preliminary investigation
into the suitability of using this group for an epidemiologic study because of
their documented exposure to DEHP. As part of this assessment, the Health
Care Financing Administration, Department of Health and Human Services (DHHS),
End Stage Renal Disease Medical Information System (ESRD MIS) data base was
evaluated to determine its usability in the conduct of such a study.
The end stage renal disease (ESRD) patient population was found not
to constitute a viable population for an epidemiologic investigation to
determine the health effects of DEHP exposure. The bases for this conclusion
were the complexities of end stage renal disease and the multiple factors
involved in the treatment of the disease. In addition, the ESRD MIS data base
was found to lack the completeness, consistency, and accuracy necessary to
perform meaningful analyses other than demographic analysis.
IV
-------
CONTENTS
Page
ABSTRACT iv
Chapter 1
INTRODUCTION 1
Chapter 2
END STAGE RENAL DISEASE (ESRD): PATHOLOGIES,
TREATMENTS AND PATIENT POPULATION 7
2.1 The Pathological Processes of End Stage
Renal Disease. 7
2.1.1 Some Primary Causes of ESRD 9
2.1.2 Diseases Commonly Associated with ESRD 11
2.1.3 Conditions Associated with ESRD 13
2.1.4 The Complicated Picture of ESRD Patients 15
2.1.5 Course and Prognosis of ESRD Patients 17
2.2 Renal Dialysis 18
2.2.1 Hemodialysis 18
2.2.2 Peritoneal Dialysis 30
2.2.3 Continuous Ambulatory Peritoneal Dialysis 35
Chapter 3
TOXICOLOGICAL FACTORS OF DEHP EXPOSURE IN RENAL
DIALYSIS PATIENTS 38
3.1 Review of DEHP Toxicity in Animals 39
3.2 Biodistribution 41
3.3 Relevant Animal Studies 44
3.4 Extrapolation to Human Health Effects 45
3.5 Toxicological Environment of Dialysis Patients .46
3.6 Conclusion: Feasibility of Assessing DEHP
Exposure Among Dialysis Patients 48
Chapter 4
AN ASSESSMENT OF THE UTILITY OF ESRD MIS DATA FOR
EPIDEMIOLOGICAL RESEARCH 50
4.1 The End Stage Renal Disease Medical Information
System: An Overview 50
-------
Page
4.1.1 The Flow of Information 52
4.1.2 The ESRD MIS Patient Population 59
4.1.3 THE ESRD Patient History and
Treatment Plan 60
4.1.4 The ESRD Outpatient Dialysis
Service Information 63
4.1.5 The ESRD Transplant Tissue
Typing Information 65
4.1.6 The ESRD Death Notification 66
4.1.7 The ESRD Center Patient Listing 66
4.1.8 The ESRD Facility Cost Questionnaire 68
4.1.9 Problems with the ESRD MIS Data:
A Summary 68
4.2 Assessment of Epidemiologic Research
Potential of ESRD MIS Data 69
4.2.1 Specificity 69
4.2.2 Accuracy 70
4.2.3 Completeness 71
4.2.4 Statistical Methodology 72
4.2.5 Summary Conclusion. . 73
Chapter 5
AN EXAMINATION OF AN EPIDEMIOLOGICAL STUDY WITH ESRD PATIENTS ... 75
5.1 The Problem of Causality in the ESRD Patient Population. ... 76
5.2 Analytic Approaches 79
5.2.1 Multiple Causal Paths 80
5.2.2 Indiscriminate Exposure Measure 80
5.2.3 Dominating Competitive Risks 82
5.2.4 Size of Study Population 84
REFERENCES
VI
-------
CHAPTER 1
INTRODUCTION
Di-2-ethylhexylphthalate (DEHP) is a phthalate ester prepared from the
reaction of pthalic acid with the alcohol, 2-ethyl hexanol. Phthalate esters
are primarily used as plasticizing agents in polyvinyl chloride (PVC) to give
the polymer its flexibility and softness. These plasticizers may account for
up to 40 percent of the final weight of the PVC product. DEHP and its isomer,
dioctyl phthalate (DOP), are currently the most widely used plasticizers o€
this type.
Although DEHP is insoluble in water, it is soluble in organic solvents
and oils. Because DEHP is dispersed in the polymer matrix and not chemically
bound, it can be easily leached out by solubilizing substances. Human blood
plasma is one such substance.
The alkyl phthalate esters have uniform physical characteristics.
Generally, they are colorless or lightly colored oily liquids with little, if
any, odor; they have relatively high boiling points and low vapor pressures;
and they are lipophilic, with negligible solubility in water (Graham 1973).
Although these compounds possess molecular dipoles, they do not form strong
intermolecular hydrogen bonds (Solomons 1980). As a result, they have lower
boiling points and lower water solubilities than acids or alcohols of
comparable molecular weight.
DEHP is used in resinous and polymeric coatings for food packaging (CFR
121.2514), in paper and paperboard for contact with aqueous and fatty foods,
in closures for sealing gaskets, in beverage cups and linings, in containers
for the storage of blood and intravenous solutions, and in intravenous tubing.
DEHP is also used in vacuum pumps as an organic pump fluid and as a substitute
for polychlorinated biphenyls (PCS) in liquid-filled capacitors. It is also
registered with the Environmental Protection Agency (EPA) for use as a
miticide in orchards.
-------
Dialkyl phthalates are released into the environment during each maior
stage of handling: manufacture, transportation/distribution, compounding/
processing into products, consumer use of products, and disposal. The
materials balance analysis for phthalate esters produced in the United States
in 1977 showed that 10 percent of the phthalate esters produced were released
to the environment (Arthur D. Little 1979). More specifically, 18,900 kkg of
phthalate esters were released to the air and 31,500 kkg of phthalate esters
were released to surface water during manufacture, processing, transport, and
consumer use of products. In addition, 453,800 kkg of phthalate esters were
discarded at the end of products' useful lives: 440,200 kkg to landfills and
13,600 kkg to incinerators. Jerque (1973) and Kotzias (1975) have reported
substantial leaching of phthalates from landfills to ground water.
Three sources of consumer exposure to phthalates have been documented.
In the first, Tomita (1977) reported increased DEHP and di-butyl-phthalate
(DBF) concentrations in blood following the ingestion of foods which were
stored or wrapped in plastics or other wrappings containing phthalates. In
the second, a series of studies has clearly documented DEHP concentrations in
tissues and blood of patients and former patients who were exposed to DFHP
during medical treatment. For example, Jaeger and Rubin (1972) reported
concentrations ranging from non-detectable to 91.5 (u)g/g in lung tissues
obtained at autopsy from 13 patients who died after receiving blood trans-
fusions. Hillman et al. (1975) reported significantly higher DEHP concentra-
tions at autopsy in neonatal heart tissues of 17 infants who had umbilical
catheterization alone or with blood products. Other authors (Gibson et al.
1978, Lewis et al. 1978, Rubin et al. 1976) followed the kinetics of dis-
appearance of DEHP after blood transfusions. The studies document that
patients are subject to phthalate exposure as a result of leaching from
plastic products into the blood.
i
The third documented source of consumer exposure is through the volatil-
ization of phthalates from plastic materials used in automobiles. While the
reliability of the data indicating the airborne concentrations generated is
open to question, exposure in an unventilated car may conceivably range
3
upwards to 3 rag/m on hot days.
-------
Poole and Wibberly (1977), Mes and Campbell (Mes et al. 1974, Mes et al.
1976), and Overturf et al. (1979) have reported DEHP and/or DBF concentrations
in tissues from humans who had no reported medical or occupational exposure to
phthalates. Poole and Wibberly reported mean concentrations of 0.06 +^0.2
(u)g/g in placentae from 10 women who gave birth to normal babies. Mes and
Campbell (Mes et al. 1974) reported concentrations of DBF that ranged from 0
to 1.0 (u)g/g and concentrations of 0.01 to 4.0 (u)g/g in adipose tissue
samples taken from accident victims at autopsy. In a later study, Mes and
Campbell (Mes et al. 1976) reported that DEHP concentrations in three samples
of adipose tissue obtained at autopsy ranged from 0.64 to 1.11 (u)g/g.
Overturf et al. (1979) reported DBF and DEHP concentrations in the triglycer-
ide fraction of the kidney cortex and medulla in 4 of 17 kidney samples taken
at autopsy. Ono et al. (1975) found serum blood levels of DEHP of 389 ppb in
patients dialyzed for six hours and up to 1010 ppb of phthalic acid (metab-
olite of DEHP) after 5 hours of in vitro dialysis.
Mishkel et al. (1979) analyzed autopsy tissues of seven patients (six of
whom had been on long-term dialysis and one who had received a massive blood
transfusion) and four controls. The authors stated that adipose tissue of the
seven patients contained most of the DEHP (up to 460 (u)g/g tissue, no further
data provided); bone marrow, brain, and peripheral nerve tissue showed
"moderate accumulation" (up to 100 (u)g/g tissue, no further data provided);
and lesser amounts were found in the kidneys, liver, lymph nodes, spleen,
heart, and lungs.
An investigation into DEHP exposure was conducted by the National Cancer
Institute to determine the toxicity and possible carcinogenicity of di-2-
ethylhexyl-phthalate. Preliminary reports presently under review suggest that
DEHP may be carcinogenic. The two-year chronic feeding study of this plasti-
cizer found the development of hepatocellular carcinomas and neoplastic
nodules in male and female rats, and hepatocellular carcinomas and adenomas in
male and female mice.
In concurrence with current research trends, study was needed to
substantiate these bioassay findings with epidemiologic investigations of
-------
health effects in human population. One population identified for study were
patients receiving dialysis for end stage renal disease (ESRD). Reasons for
selecting this population included:
• The exceptional risk of exposure to DEHP leached into the
blood during dialysis treatment
• A dichotomous distribution of the ESRD population receiving
hemodialysis or peritoneal dialysis
• The availability of a large data base maintained by the
Department of Health and Human Services, Health Care
Financing Administration (HCFA), the End Stage Renal Disease
Medical Information System (ESRD MIS)
• Interest in performing epidemiologic research on ESRD
patients for effects associated with DEHP.
In response, the Epidemiology Branch of the Office of Toxic Substances,
EPA conducted a study to determine the following:
• The feasibility of utilizing renal dialysis patients as a
population for investigation of DEHP
• Evaluation of the ESRD MIS as a data base suitable for
accessing information required to perform appropriate
analyses
• If the above population and data base were found satis-
factory, to design a protocol for an epidemiologic study of
the effects of DEHP on end stage renal disease patients.
In compliance with this direction, the following activities were
conducted:
• Contact and interview with the Branch Chief of ESRD MIS
• Contact with a university consulting group, who had per-
formed preliminary analyses of ESRD MIS data, to obtain
information on the manipulation they had performed to
correct for inadequacies in the data base
• Review of the natural history of pathologies associated with
end stage renal disease
-------
• A site visit of a dialysis treatment center to obtain
information on the state-of-the-art in dialysis treatment
modalities
• Consultation with nephrologists .and epidemiologists at the
National Heart, Lung, and Blood Institute, the National
Cancer Institute, the Georgetown University Medical School
Nephrology Department, the University of Texas School of
Public Health, and the University of Minnesota School of
Public Health.
The information gained from these activities and discussed in chapter 2
suggests that ESRD patients are not a satisfactory population for an epidemio-
logic study of the effect of di-2-ethylheyxl-phthalate. The basis of this
conclusion is the multifactorial aspects of this patient population which
include the broad array of pathologies leading to and associated with ESRD, as
well as the effects of the different treatment modalities and their corre-
sponding components. Chapter 2 begins with a discussion of contributory
components of end stage renal disease, the pathologic process, and the effects
of ESRD on the constitution of the body. The second section discusses the
three dialysis modalities used in treating ESRD patients. Each modality is
described in terms of the process involved, the components of each, the
effects attributed to each of these components, and the criteria for selection
of one modality over another.
Chapter 3 describes toxicological factors in renal dialysis patients
resulting from DEHP exposure. These factors could conceivably produce toxic
effects in the ESRD population which may not be representative of those seen
in normal animals and humans. The non-unique toxic effects of DEHP make it
unlikely that DEHP could be identified as either a primary or synergistic
cause of toxic symptoms in dialysis patients. In addition, the generally poor
health of dialysis patients, the multiplicity of their drug therapies and-
their inadvertent exposures to both chemicals and disease because of dialysis
could disguise or act syngeristically with DEHP to produce toxic effects.
-------
Chapter 4 concludes that the End Stage Renal Disease Medical Information
System (ESRD MIS) lacks the accuracy, consistency, and completeness necessary
to perform any meaningful analyses other than demographic. The conclusion is
based upon input provided by the Health Care Finance Administration (HCFA) End
Stage Renal Disease Medical Information System (ESRD MIS) Branch and the
Environmental Epidemiology Branch of the National Cancer Institute. This
information is presented in Sections 4.1, "An Overview of the ESRD MIS," and
4.2, "An Assessment of the Epidemiologic Research Potential of ESRD, MIS
Data.
Chapter 5 uses a biostatistical approach to discuss the criteria required
to derive inferences from an epidemiologic study, i.e., the type and quality
of the information which is necessary to determine causal association between
exposure to a factor and development of a specific disease outcome. The
discussion concludes that ESRD patients would not constitute a viable study
group for an epidemiologic investigation of the carcinogenic effect of
exposure to DEHP for two reasons: 1) the inability to measure exposure to
DEHP separately from exposure to carcinogens and possible etiologic factors;
and 2) the extremely high, combined, competitive risk of mortality or
morbidity from causes other than primary cancer of the liver or other
malignancies.
-------
CHAPTER 2
END STAGE RENAL DISEASE: PATHOLOGIES,
TREATMENTS, AND PATIENT POPULATION
L. Euinton, M.B.Ch.B., D.I.H., F.R.S.H.,
G. Schreiner, M.D., A. Sinclair, R.N., and
I. Marks, M.P.H.
"End stage renal disease" is an aggregate term representing
a complex cluster of systemic, pathological processes, with
additional contributory and associated conditions. This chapter
reviews the systemic complexity of end stage renal disease and
the problems of multiple exposures associated with the modalities
of dialysis.
2.1 THE PATHOLOGICAL PROCESSES OF END STAGE RENAL DISEASE
End stage renal disease, commonly called chronic renal failure (CRF), may
occur as a result of many pathological conditions. Each of these conditions
has its own group of associated pathologies. The major classes of macro-
pathology include: a) vascular diseases; b) primary glomerular diseases;
c) primary tubulointerstitial nephritis; and d) miscellaneous causes, includ-
ing stones, tumors, cysts, and infiltrative diseases. In addition, there are
clusters of uremic symptoms which are common to a given level of renal failure
which may be experienced regardless and distinct from the pathology of the
underlying kidney disease. The time, course, and specific morphologic
changes, as well as the stages, of these pathological conditions are highly
variable from patient to patient. As a result, ESRD patients should not be
described as a uniform population, even if controlled by sex and age.
Renal failure develops when disease or trauma compromises the ability of
the kidney to remove toxic materials from the blood, and maintain fluid,
electrolyte, and acid-base balances. The condition may be acute or chronic.
Acute renal failure is caused by processes extending over days or weeks and
potentially results in recovery from uremia. Chronic renal failure occurs
from processes extending over weeks or months, is usually characterized by
progression, and exhibits a high degree of variability in the slope or rate of
-------
progression. Rapidly progressive disease may evolve from health to renal
failure in one to three years, while slowly progressive disease may extend
over 30 to 40 years. For functional purposes, it is clinically useful to
divide patients into three phases:
•• Phase 1, from 100 percent glomerular filtration rate to 20 percent of
normal
• Phase 2, from 20 percent of normal to 5 percent of normal
• Phase 3, less than 5 percent of normal.
It is largely the patients in Phase 3 who have been designated by the Govern-
ment as exhibiting end stage renal disease.
The characteristics associated with chronic renal failure are: nitrogen
retention; screened by elevations of serum urea, creatinine, acidosis;
screened by pH and depression of serum bicarbonate, anemia; screened by HCT
and hemoglobin depression, hyperphosphatemia and hypocalcemia except in hyper-
parathyroid or other hypercalcemic etiologies; and other electrolyte distur-
bances, including normal or reduced sodium, normal or elevated serum
potassium.
Urinary volume is often relatively fixed between one and four liters per
day, and does not respond to variations in water intake. One practical and
clinical definition of renal failure is when the daily urine output is the
maximum urine output. Urinary osmolarity is usually fixed close to that of
plasma (300 to 320 mOsm/l). The findings on urinalysis depend on the nature
of the underlying disease, but broad (especially waxy) casts are often promi-
nent in advanced renal insufficiency of any cause. The hematologic picture is
that of a normochromic, normocytic anemia of moderate severity.
The outcome of urinalysis depends on the nature of the underlying dis-
order and superimposed complications. Progression of underlying chronic renal
disease is generally not susceptible to specific treatment, although slowing
of the rate of progression may be a practical goal in certain diseases.
Oliguria, progressive hyperkalemia, and pericarditis are often manifestations
of a preterminal state. Even in these situations, however, dialysis and
transplantation may improve the outlook.
-------
A renal injury occurs in patients with immune complex deposition where
the antigen is endogenous, as in systemic lupus erythematosus (SLE) and neo-
plasia associated with glomerular disease, but may be exogenous in the case of
serum sickness, bacterial, viral, and protozoa-associated glomerular nephri-
tides. Disease from antibodies to glomerular basement membrane glycoprotein
occurs in the nephritides of antilymphocyte serum treatment, Goodpasture1s
Syndrome, and some rapidly progressive forms of idiopathic cresentic
glomerular nephritis. Antibodies to other active glomerular antigens occurs
in some forms of membranous glomerulopathy and antibody to planted glomerular
antigens occurs in SLE, drugs, and post-streptococcal glomerulonephritis.
Inflammatory vasculitis involving the glomerulus may occur in periarteritis
nodosa, Wegner's granulomatosis, and other vasculitidies.
2.1.1 Some Primary Causes of ESRD
The purpose of this section is to distinguish some of the different
causes of kidney disease which may progress to end stage renal disease, or
CRF. The pathological process is incited at the glomeruli, at the tubulo-
interstitial areas, or in the renal arteries, according to the etiology. By
the time the stage of chronic renal failure is reached, all of these three
sites are involved to some degree. The signs, symptoms, or laboratory tests
which may distinguish these conditions by diagnosis are not detailed. But,
plainly, the clinical and biological parameters will vary at the stage of CPF
according to the etiology and pathophysiology of the renal disease. These
parameters may be further complicated by the effect of the disease on other
organs, and still further complicated in those patients whose renal
involvement is secondary to systemic disease.
Chronic Tubulointerstitial Nephropathy
Chronic tubulointerstitial nephropathy (TIN) includes all those chronic
kidney disorders in which generalized or local changes in the tubulointersti-
tial area predominate over glomerular or vascular lesions. Since some tubulo-
interstitial changes are associated with all renal diseases, this distinction
may be difficult to identify.
-------
Other conditions associated with TIN include:
• Effects related to the use of certain drugs, including analgesics;
sulfonamides, penicillin, methicillin, and their homologs; diuretics
such as furosemide, thiazide, phenytoin (dilantin), etc.
• Effects related to heavy metal toxins, including lead, cadmium,
uranium, copper, mercury, bismuth, thallium, arsenic, and iron.
• Oxalate deposition from ethylene glycol, methoxyflurane, and
anesthetic agents, primary hereditary oxaluria and small bowel
disease.
• Uric acid nephropathy from gout and hematologic disorders, especially
primary and secondary hyperparathyroidism, milk alkali syndrome,
sarcoid, neoplasia and multiple myeloma.
• Effects related to malignancy, in which renal interstitial spaces may
be invaded by proliferative malignant cells in leukemia and lympho-
sarcoma. The cortex is involved more than the medulla (Merck et al.
1.977).
Chronic tubulointerstitial renal disease is due to ascending bacterial
infection. Obstructive uropathy, such as strictures, stones, reflux, myo-
neurogenic disease, etc., is usually present. Even with advanced disease,
proteinuria is usually < 1 gm/sq m/day. The urinary sediments tend to be
scanty, but renal epithelial cells, granular casts, and occasionally white
blood cell casts are found, especially in analgesic nephropathy (Willis et al.
1972).
The functional features differentiating tubulointerstitial nephritis are:
proteinuria less than 2 gm, scant urinary sediment, sodium wasting, and
anemia which is disproportionately severe to the level of nitrogen retention;
a lesser degree of hypertension, hyperchloremic acidosis, and markedly
elevated uric acid for the degree of nitrogen retention; and increased urine
volume with decreased maximum osmotic U/P ratio (concentrating ability is-one
of the most sensitive early signs).
Hereditary Renal Disease
In polycystic renal diseases, the bilateral cysts cause enlargement of
the total renal size, while reducing, by compression, the functioning renal
tissue. Chronic infection frequently is superimposed and contributes to the
10
-------
progressive loss of renal function. There is also a high associated incidence
of intracranial aneurysm, and hypertension is common. Death is usually due to
uremia or the complications of hypertensive cardiovascular disease. About 10
percent of patients die of intracranial hemorrhage from rupture of aneurysms.
Polycystic patients make particularly favorable chronic dialysis patients
with a five year survival in the range of 80 percent. They tend to have less
anemia than other patients with chronic renal disease and lower blood
pressures.
Medullary cystic disease, sometimes called juvenile nephronothisis, is a
diffuse nephropathy, either genetic or congenital in origin. Retarded growth
and evidence of bone disease are common in children. These patients have
anemia out of proportion to their degree of renal insufficiency.
Medullary sponge kidney is a tubular ectasia or dysplasia resulting in
congenital cystic dilation of the collecting tubules, leading to urinary
stasis and nephrocalcinosis. People with this condition often develop
infections and can soon develop mild renal insufficiency.
Hereditary nephritis is a familial disorder characterized by hematuria,
renal functional impairment, nerve deafness and, on occasion, ocular
abnormalities (Merck et al. 1977).
2.1.2 Diseases Commonly Associated with ESRD
Auto-immune collagen vascular diseases which involve the kidney include
generalized lupus erythematosus (GLE) and polyarteritis nodosa (PAN).
Generalized lupus erythematosus clinically shows a variable combination of
fever, an erythematous rash (especially of the face), arthritis, pleurisy,
enlarged lymph glands and spleen, albuminuria and hematuria, signs of endo-
carditis and anemia. An important diagnostic laboratory finding is the
presence of "lupus-cells" (LE cells) in the blood and bone marrow. These are
polymorphonuclear lymphocytes containing rounded masses of phagocytosed
11
-------
altered desoxyribonucleic acid, essentially similar to the haematoxylin-
staining bodies seen in the kidneys and other organs. The blood serum often
contains excessive gamma globulin and special serological studies give evi-
dence of complex auto-immunity changes, with the formation of antibodies
against the patient's own nucleoproteins, anti-DNA, and so injuring the cells
of many organs.
Polyarteritis nodosa is a rare febrile illness, probably allergic in
nature, with diffuse or nodular acute inflammation of many arteries, chiefly
medium-sized and small arteries especially in the viscera. The vessel walls
show areas of necrosis, fibrinous exudate, accumulation of neutrophil and
eosinophil polymorphonuclear leucocytes, plasma cells, lymphocytes and macro-
phages. Some arteries show multiple small aneurysms, with multiple small
infarcts in the organs, and later organization of the thrombi. Veins adjacent
to the inflamed arteries show phlebitis. According to the main distribution
of the lesions, the symptoms are highly varied. They may be predominantly
alimentary, cardiac, cerebral, muscular, or renal. The disease is often
fatal, from emaciation, anemia, hemorrhages, and infarction of the heart,
brain, or other organs, but spontaneous recovery sometimes occurs. Steroids
are useful in some cases of vasculitis, and cytoxan is specially effective in
the subgroup known as Wegner's Granulomatosis (Willis et al. 1972).
In several other diseases, there is strong evidence that sensitization of
the tissues to bacterial or other antigenic substances plays an important
part. Because all show prominent lesions in connective tissue, including
"fibrinoid" degeneration or necrosis, they have been spoken of collectively
and vaguely as the "collagen diseases"—including GLE, PAN, and such rare,
possibly kidney-affecting diseases as scleroderma and dermatomyositis (Willis
et al. 1972).
In addition, two metabolic disease are commonly associated with end stage
renal disease:
• Diabetes Mellitus: The complications of untreated diabetes include
ketosis, resulting from impaired fat metabolism; hypercholesterolemia
and xanthomatosis; arterial atheroma with consequent thrombosis and
gangrene of the limbs; diabetic retinitis, which also is secondary to
12
-------
changes in retinal arteries; increased susceptibility to infections,
i.e., tuberculosis, and certain renal complications, i.e., glomerular
sclerosis of Kimmelstiel and Wilson; and medullary necrosis,
especially of the papillae (Willis et al. 1972). Diabetic nephro-
pathy, its pathogenesis, histopathology, and clinical course, was
described by D'Elia et al. (1974) from a study of 49 patients of whom
27 percent were blind.
Amyloid Degeneration: It is probable that in amyloid disease the con-
nective tissue protein, perhaps collagen and elastin as well as chon-
dromucoid, undergoes direct conversion into amyloid, of which, how-
ever, there is also a great accumulation causing enlargement of the
affected organs. The generalized disease usually results from a long-
standing bacterial infection, most often chronic supperative infec-
tions, tuberculosis, rheumatoid arthritis, hereditary syphilis,
various types of neoplasms including Hodgkin's disease, multiple
myeloma and chronic infective endocarditis. A peculiar form of amy-
loidosis is present in occasional cases of plasma-cell myelomatosis.
In the kidneys, deposits occur in the glomeruli, and also around the
tubules and in the walls of the larger vessels (Willis et al. 1972).
2.1.3 Conditions Associated with ESRD
The pathology of ESRD includes a number of conditions which inhibit the
normal function of various systems within the body. The following is a dis-
cussion of three of these conditions, hypertension, uremia, and immuno-
suppression, and their associated effects.
Hypertension
During the chronic stage of CRF, there is a progressive rise of blood
pressure. In malignant cases of essential hypertension, the progressive renal
damage and azotemia are secondary to rapidly mounting blood pressure. The
renal changes generally resemble those of nephrosclerosis but with severe and
characteristic lesions of the arteries, namely a peculiar swelling or fibri-
noid necrosis of the walls of the arterioles (especially the afferent
glomerular arteries), and a cellular fibrous endarteritis of the larger inter-
lobular arteries. Of special interest are the vascular lesions which are not
confined to the kidneys, but occur in the arterioles of many other tissues.
This fact suggests that malignant hypertension is a primary vascular lesion
with widespread lesions in the arterial system and serious secondary ischemic
effects on the kidneys.
13
-------
Thus, renal ischemia leads to hypertension through stimulation of.
reninangiotensin resulting in the production of Angiotensin II, a potent
vasoconstrictor produced in the body. Ischemia may be the result of a host of
obstructive diseases of arteries, including:
• Arteritis, embracing all five inflammations of arteries
•' Atheroma or atherosclerosis, a degenerative change in the intima
• Medial calcification or Monckeburg's sclerosis, a degenerative change
in the media
• Hypertensive sclerosis, a generalized arterial disease associated with
high blood pressure
• Mechanical constriction of arteries
• Local arterial spasm including fibromuscular dysplasia
• Arterial embolism (Willis et al. 1972).
Uremia
Renal insufficiency is a measurable decline in kidney function. Renal
failure is a sufficient degree of decline to induce biochemical abnormalities
in the serum and loss of the kidney's homeostatic function. Uremia is reached
when renal failure becomes symptomatic.
The first manifestations of uremia may be lassitude, fatigue, and often
decreased mental acuity. Neuromuscular features include coarse muscular
twitches, peripheral neuropathies with sensory and motor phenomena, muscle
cramps, and convulsions (usually the result of hypertensive encephalopathy).
Gastrointestinal manifestations, such as anorexia, nausea, vomiting,
stomatitis, an unpleasant taste in the mouth, are almost uniformly present.
Malnutrition leading to generalized tissue-wasting is a prominent feature.of
chronic uremia. The skin may develop a yellow-brown discoloration, and
occasionally urea from sweat may crystallize on the skin as uremic frost.
Pruritis is an especially uncomfortable feature in some patients. Hyper-
tension is often present, as are signs of congestive heart failure (Merck
et al. 1977).
14
-------
Uremia has been repeatedly shown to be immunosuppressive according to
Matas et al. (1975) and Birkeland (1976), who also confirms that there is an
increased incidence of cancer in uremic patients. The mode of connection
between these established factors is being pursued.
Immunosuppression
The kidney is frequently injured by those disorders in which the immune
system produces antibodies to an endogenous antigen. Several mechanisms for
developing an immune response to autoantigens are recognized. Genetic factors
play a part. Immunosuppressive agents have been developed which suppress all
immunologic reactions and the metabolism of rapidly dividing cells. For this.
reason, overwhelming infection is the leading cause of death in transplant
recipients. Nevertheless, carefully selected and administered immunosuppres-
sive treatment has been primarily responsible for the present success of clin-
ical transplantation.
Evidence for the occurrence of immune responses to a variety of human
tumors is increasing. Tumor-specific (or tumor-associated) transplantation
antigens have been demonstrated in most experimental animal tumors and in
several human neoplasms. It seems likely that the presence of these surface
markers on neoplastic cells would allow their recognition by immunocompetent
host cells as well as their reaction with antibodies directed against immuno-
genic surface configurations. The significance of such recognitions and reac-
tions in the pathogenesis and conduct of tumors is currently the object of
intensive laboratory and clinical investigations (Merck et al. 1977).
2.1.4 The Complicated Picture of ESRD Patients
The interaction of the pathologies described above has additional effects
on the body which contribute to the further deterioration of the overall body
systems. Thus, apart from the clinical and biologic parameters affected by
the ESRD-induced uremia, patients may suffer a very wide range of symptoms,
signs, and complications from the underlying disease process which, to some
extent, has either caused or contributed to the failure of kidney function.
Thus, hypertension leads to cardiac hypertrophy, then congestive heart
15
-------
failure, and a strikingly increased risk for cerebral vascular accident and
myocardial infarction.
The obstructive diseases of arteries, whether associated with hyperten-
sion or not, may impair the glomerular and/or tubulointerstitial function of
the kidneys, while having ischemic effects elsewhere, i.e., coronary atheroma,
diabetic gangrene of limbs, retinal exudates and papilledema, etc.
Kidneys damaged by toxic agents may be accompanied by toxic effects else-
where, such as the anemia of chronic lead poisoning, and liver damage from any
of many toxic agents.
Anemia is seen with uremia, regardless of the underlying renal disease.
Unlike the anemia of other chronic disorders, however, decreased body iron
stores are not a consistent feature. The anemia associated with ESRD-induced
uremia is characterized by a shortened red blood cell survival and diminished
erythropoietin production with a subnormal marrow response. In the very rare
patient, renal arteriolar microvascular disease may be associated with hyper-
tension, marked hemolysis, and increased reticulocyte production. This so-
called hemolytic-uremic syndrome is generally associated with widespread
platelet consumption and occlusion of the microvasculature, and convincingly
demonstrates a dissociation between renal excretory and endocrine functions
(that is, decreased glomerular filtration, but with increased erythropoietin
and renin production). The mechanism of the anemia of ESRD, or CRF, is
discussed in detail by Fisher (1980).
Spontaneous regressions of human neoplasms have encouraged interest in
the immunologic therapy for neoplastic diseases. Present immunotherapy in
human neoplasia is based on recent advances in knowledge of humoral and cel-
lular immunity, immunosuppression, human transplantation antigens, and immuno-
logic tolerance, a state in which a substance normally capable of inducing an
immune response fails to do so (Merck et al. 1977).
Cancer developing after renal transplantation is a well-documented event
that has clinical importance and immunologic implications. Montie (1977) sug-
gests that recent data do not support that cancer after a transplant is caused
16
-------
by suppression of the immune surveillance system. According to Perm (1972),
the renal transplant patient has approximately 30 to 100 times greater risk of
developing cancer than otherwise apparently similar individuals in the general
population. This increased incidence is not reflected in a uniform increase
in the incidence of all types of cancers, but is evident primarily in two
areas: 1) epithelial tumors of the skin, lip, and cervix account for 47 per-
cent; and 2) solid lymphomas account for 23 percent. The lymphomatous tumors
in transplant patients have an unusual predilection for central nervous system
involvement and a poor response rate to therapy.
Further data supporting the likelihood that a patient with ESRD is more
prone to develop cancer than a patient without renal disease, and that those
with glomerulonephritis or phenacetin nephropathy are particularly prone, were
provided by Sutherland et al. (1977). - However, data pertaining to increased
cancer risk in nontransplanted patients are not this definitive, and this is
an area in need of further investigation.
2.1.5 Course and Prognosis of ESRD Patients
Dialysis patients have an average survival time of seven to 10 years
after the onset of end stage renal disease, or CRF.* An average cross
sectional group of dialysis patients will have an average annual mortality of
10 percent. This tends to increase with the mean age of the patient group.
During this time, the course and prognosis of ESRD patients varies enormously
according to the:
• Nature of the renal disease, and the stage and possibly
the duration of the uremic state
• Nature and extent, possibly modified by treatment, of
generalized disease or specific organ disease, and the
complications which may arise from these
• Immunosuppressive effects of uremia itself or of
immunosuppressive drugs administered to the patient
*An interview with Nancy Blackburn, R.N., and Raphael J. Osweroff, M.D., of
the Dialysis Center of Northern Virginia, January 20, 1981.
17
-------
• High risk of malignancies, some 30 to 100 times that in
the general population, known to exist in ESRD patients
• Increasing age of the cohort
• The prevalence of epidemic infections, such as hepatitis
and the opportunity for opportunistic infections in
suppressed patients
• Further complications in dialysis patients of the
dialysis, itself, and the concominant treatment.
2.2 RENAL DIALYSIS
Renal dialysis is the process in which an artificial mechanism is
utilized to accomplish the biological cleansing functions normally done by the
kidney. There are three treatment modalities used in treating end stage renal
disease, or chronic renal failure. They are:
• Hemodialysis
•• Peritoneal dialysis
• Continuous ambulatory peritoneal dialysis (CAPD).
Each of these modalities differ in the equipment utilized, the complications
associated with each, and the conditions which warrant the use of one over
another. At present, the most widely used modality is hemodialysis. However,
the criteria used for the selection of one modality over another involve a
number of factors, both medical and psycho-social in nature. Some of these
are listed in table 2-1.
2.2.1 Hemodialysis
Patients receiving chronic (ESRD) maintenance hemodialysis require fbur-
to eight-hour treatments three times a week to maintain a state of well-being.
The hemodialysis process reduces the BUN and creatinine by approximately 50
percent, and corrects metabolic acidosis and hyperkalemia. These patients are
also advised to restrict their fluid intake to 1,000 cc's per 24 hours and are
usually placed on a special low-protein diet. It has been reported that ESRD
patients gain a variable amount of weight per 24 hours between treatments,
18
-------
CRITERIA Hemodialysis Peritoneal CAPD
Physical competency x
Intellectual competency x
Mental retention x
Reliability . x
Socio-economic level x
Arteriovenous shunt competency x
Arteriovenous graft competency x
Patient choice x
Psychological stability x
Hyperkalemia, acute life threatening x
Acute renal failure with recovery
potentials x
Lack of sophisticated equipment x
Age x
Motivation x
Home support x x
Cooperation x x
Cardiovascular system competency x
Small children x
Older patients (60 years or older) x
Home dialysis for patients living alone x
Patients refusing blood transfusions x x
Holding for transplantation and
fistula maturation x
TABLE 2-1. CRITERIA FOR SELECTION OF DIALYSIS MODALITY
19
-------
depending upon their urine output and their level of fluid intake. Hemodialy-
sis may be accomplished in the hospital, in an out-patient clinical setting,
or even at home (Guyton 1956).
Hemodialysis involves a mechanical "kidney machine" consisting of three
essential components:
• The exchanger, a membrane unit where the body's blood, carrying the
waste and toxic materials of the body, comes into direct contact with
the infused fluid (dialysate), its purpose being to take on the body's
waste and toxic materials, thereby cleansing the blood
• A tubing system utilized for the conduct of the patient's blood and
dialysate to and from the membrane component
• A dialysate delivery system, comprised of the machine itself and
dialysate, composed of a buffering component, usually acetate or
bicarbonate, with electrolytes and glucose.
Each of these components has the potential of exposing the patient to contami-
nants, producing adverse iatrogenic reactions (table 2-2).
Membrane Unit
Hemodialysis units basically fall into three fundamental categories,
based on the type of membrane used in the unit. These three categories are:
• Parallel plate or sandwich design
• Coil
• Hollow-fiber devices.
The underlying purpose of these membranes is to provide an influencing factor
on the transport of solutes from the blood to the dialysate. The successful
function of the particular membrane will assist the clinical team in maintain-
ing the patient in metabolic homeostasis.
Each membrane has certain mechanical differences which influence the
physiological dialyzing of a patient. Some of these characteristics are high-
lighted in the following paragraphs.
20
-------
COMPLICATION
Peritonitis
Tunnel infection
Hernia
Anemia
Electrolyte inbalance
Weight
Low back pain
Cramps, abdominal and extremities
Shoulder pain
Chest pain
Cardiac arythmia
Excessive fluid loss
Fluid overload
Blood tinged drainage
Congestive heart failure
Os teodys trophy
Hypertension
Hydrothorax
Atelectasis
Protein loss
Hyperosmolar coma
Hypernatremia
Hypotension
Hemorrhage
Hypokalemia
Air embolism
Pyrogenic reactions
Hemolysis
Seizures
Thrombosis
Hepatitis
Neuropathy
Hyperuricemia
Constipation
Menstrual dysfunction
Infertility
Sexual problems
Insomnia
Dialysis dementia
Sclerota peritoneum
Leukopenia
Dermatitis
Arteriosclerosis
Encephalopathy
Hemodialysis
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Peritoneal
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
CAPD
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
TABLE 2-2. POTENTIAL COMPLICATIONS PER MODALITY
21
-------
COMPLICATION
Hemodialysis Peritoneal CAPD
Osteomalisia
Hypercalcemia
Anorexia
Gastroenteritis
Pulmonary edema
Cardiac atheroma
Myocardial infarction
Cerebral thrombosis
Intracranial hemorrhage
Cardiac tamponode
Sepsis
Lymphoma, non-Hodgkins
Death
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
TABLE 2-2. POTENTIAL COMPLICATIONS PER MODALITY (continued)
22
-------
Plate or sandwich design dialyzers consist of parallel sheets of cello-
phane sealed at the edges between grooved polypropylene membrane supports.
Blood passes one way between the sheets, while dialysate passes in the oppo-
site direction outside of the membranes. The arterial-venous pressure gradi-
ent is sufficient to circulate the blood through the membranes without a pump
(Hampers et al. 1973). Ultrafiltration is the removal of solute-containing
fluid by means of a transmerabrane pressure differential. This may be accom-
plished either by raising the internal hydrostatic pressure of the blood or
applying a negative pressure to the dialysate compartment. In practice,
usually both are used and the ultrafiltration pressure gradient becomes the
sum of the positive and negative pressure.
In coil dialyzers, tight concentric coils of cellophane tubing are sepa-
rated by screening or mesh spacers. Blood is pumped through the inside of the
tubing, and dialysate circulates through the coil outside. Ultrafiltration is
accomplished by partially obstructing the outflow of blood from the coil with
a screw clamp, thus increasing the blood pressure within the coil (Hampers et
al_. 1973).
Hollow-fiber dialyzers consist of a bundle of hollow cellulose fibers
200mm in diameter through which the blood passes. Dialysate is circulated
around the outside of these fibers in a countercurrent fashion. This dialyzer
is compact and efficient because of the high ratio of surface area to blood
volume. Like the parallel sheet dialyzer, blood flow resistance is low.
The membrane unit is a potential medium for introducing factors that may
result in an adverse patient reaction. Mechanical trauma imposed on the red
cell by the artificial kidney itself results in considerable loss of membrane
lipoid and premature red-cell senescence (Kjellstrand 1978).
Studies have shown Beta thromboglobulin (BTG) levels significantly ele-
vated in patients on chronic hemodialysis. Beta thromboglobulin, a platelet-
specific protein, can be utilized to determine the clinical stability of the
platelets whose integrity are essential to the clotting mechanism. It is not
liberated from the platelet during exposure to heparin, but is liberated dur-
23
-------
ing the release reaction coming in uremia. So, the increased BTG levels are
not due to turbulent flow within the vascular access, but may be due to plate-
let aggregation within the hemodialyzer during the first hour of dialysis.
This effect is observed with the Travenol CF series hollow fiber (cuprophane)
equipment, but not with the Cordis-Dow artificial kidney (regenerated
cellulose) (Adler et al. 1979).
Dialyzers with cellulose membranes were found to cause marked leukopenia
in a study conducted by J. Shin, while its occurrence was significantly less
in non-cellulose membrane dialyzers. The extent of white blood cell decrease
seemed to correlate inversely with an increase in the ultrafiltration rate per
membrane area of dialyzer. However, there are recent reports that indicate
that leukopenia and hypoxia are possible unrelated effects of hemodialysis
(Shin 1978).
In recent years, there has been a discussion of the reuse of membranes
and the effect on the hemodialyzed patient. Membranes are intended for single
use, so reused membranes lack the protection from contamination originally
provided by the manufacturer (Hampers et al. 1973).
Because there is no definite policy on the reuse of membranes, it is dif-
ficult to ascertain the number of patients being dialyzed with reused mem-
branes. At the same time, it would be difficult not to ascribe certain clini-
cal manifestations to the reuse of dialyzers (Anonymous 1980). Thus far,
reductions in dialyzer efficiency and adverse patient reactions, including
fever and infection, have been demonstrated following reuse. In addition, the
practice of storing dialyzers filled with diluted formaldehyde can result in
an increased number of patient reactions to formaldehyde (Willingmyre 1979).
"Agents used to sterilize dialyzers, to reduce bacterial counts in the
fluid pathways of artificial kidney machine, water treatment, and water or
dialysate distribution systems have included sodium hypochlorite, chlorine
dioxide, benzelkonium chloride, 27 percent sodium chloride, and formaldehyde"
(Kjellstrand 1978). These compounds have been associated with various
clinical manifestations of hemolysis, hepatitis, fever, and even death.
24
-------
Reported adverse effects from the reuse of dialyzers can be categorized
into the following areas:
• Increased risk of infections from improperly sterilized
membranes
• Pyrogenic or hemolytic reactions from the residues of
sterilants and other substances such as formaldehyde
•• Deterioration of dialyzer performance parameters
• Immunological reactions resulting from formaldehyde-
residual blood interactions (Anonymous 1980).
.Significant residuals of formaldehyde and toxic reaction products of ethylene
jglycol were identified in the flood compartment of the dialyzer after proper
preparation for dialysis and untoward patient reactions were attributed to
these residues (Willingmyre 1979). However, the synergistic reaction of these
residues1 and products on the various multitudes of oral, intravenous, and
intramuscular medications received by the patient might also result in
unexpected, untoward clinical manifestations. Some positive benefits to
reusing a dialyzer include less blood loss, and probably less activation of
complement, which may be one of the factors associated with the production of
leukopenia during the first hour of dialysis.
Tubing System
The dialysis access and extracorporeal blood components include:
• Arteriovenous shunts for access to the patient's blood systems
• Tubing for transporting blood from the filtration membrane back to the
patient.
Access to the circulation for hemodialysis is generally obtained with an
arteriovenous (AV) shunt or fistula. An arteriovenous shunt consists of
Teflon vessel tips inserted into the radial artery and cephalic vein (or other
accessible vessels in an upper or lower extremity). These vessel tips are
linked by a silicone rubber cannula or tube external to the skin. Two sec-
tions of the cannula are joined by a connecting piece of Teflon and are sepa-
rated at the connector for dialysis. After each dialysis session, the cannula
loop is rejoined with a new Teflon connector.
25
-------
Subcutaneous arteriovenous fistulas avoid the recurrent infections and
clots associated with shunts. They permit a freer lifestyle for bathing and
swimming, and have largely replaced the arteriovenous shunts for vascular
access. The radial artery is anastomosed to the cephalic vein in an end-to-
end, end-to-side, or side-to-side fashion. The forearm veins dilate,
eventually arterialize, and are suitable for repeated puncture which may be
used with a single needle alternating pump device or two needles placed in the
opposite directions within the flowing stream, the proximal needle being used
for intake and the venous needle for outflow (Morgan 1973). Plastic tubing is
utilized as the pathway to transport the blood between the exchange membranes,
the machine, and the patient.
Plastic tubing contaminants, such as plasticizers, have been implicated
as causing severe and complex reactions. Plastics used in the dialysis proce-
dure contain unreacted materials capable of being leached from the plastic
when in contact with lipotrophic components of the blood. These unreacted
materials can provide the source of the hapten. Phthalic anydide (PA),
phthalic acid derivatives, and diphenylmethane dusocynate (MDI) will cova-
lently bind to proteins (i.e., in the blood) and can act as an antigen causing
immunologic disease under appropriate circumstances. PA may cause asthma, and
MDI may cause asthma or hypersensitivity pneumonitis. PA, MDI, or chemicals
similar in their capability to react with autologous proteins of the patient
could result in an immune response (Patterson et al. 1980). The large number
of variables that have an impact on the dialyzed patient makes analysis of the
various factors and their relevance to clinical disease complex.
Plasticizers used in the manufacture of dialysis tubing, blood clot
traps, and in all devices involving the external circulating of blood may be
toxic to patients. The plasticizers and stabilizers employed are: phthalic
acid esters, organic tin compounds, epoxidized soya bean oil, dibasic C. to C,
acids, and sobacta compound. It was found that certain patients exhibit a
clinical illness simulating "viral hepatitis" shortly after beginning dialysis
with new PVC tubing. The hepatitis disappeared when the tubing was changed
(Kjellstrand 1978).
26
-------
In response to the controversy over the leaching of di-2-ethylhexyl-
phthalate, both Dow Chemical Medical Division and Renal Systems Corporation
have manufactured tubing for dialysis that does not contain DEHP as its
plasticizer. This alternative tubing has been commercially available for at
least five years.
Another risk to dialysis patient related to plastic is the introduction
into the body of plastic fragments from the tubes. Kletschka et al. wrote
that "spalled plastic particles are undoubtedly responsible for much of the
heart (i.e., subendocardial necrosis 'Stone Heart Syndrome1), brain, and other
organ damage...." (Kjellstrand 1978).
Dialysate Composition
The composition of the dialysate used in hemodialysis depends on clinical
requirements. However, the components are usually water, electrolytes, glu-
cose, and added nutrients and medications. The accuracy of the dilution is
checked by measuring the conductivity, osmolality, or the chloride content.
Water is the largest component utilized in the dialysate. There are
extreme geographic variations in potable water supplies regarding hardness,
arsenic, and iron content. There are also various methods used to "purify"
water: distillation, softness, de-ionization, and reverse osmosis. Extremely
pure water requires complicated equipment with more chances of breakdown and
contamination. Also, the availability of the proper equipment is somewhat
dependent upon the facilities' and/or patients' economic competency. Some of
the compounds commonly found in water include: chlorine, copper, nitrate,
chlorazine, fluoride, and aluminum (from alum treatment). The latter has been
implicated in the onset of a devastating complication called dialysis dementia
(Kjellstrand 1978). The advent of alum treatment of water with its resulting
increase in aluminum content has been implicated in cluster outbreaks of
dialysis dementia in Athens, Newcastle Upon Tyne, Chicago, and some other
cities with a high aluminum content in the water supply.
Some new diseases have been identified which may have genetic predis-
positions aggravated by environmental exposure. An example would be iron
27
-------
myopathy. which is seen in patients with iron overload from too prolonged
treatment of anemia, excessive hemolysis, multiple transfusions, or high
levels of'iron in the water and food supply. The disease is associated with
high blood ferritin levels and deposition of iron in the proximal muscles of
the legs. HLA-typing suggests some increased risk associated with the same
groups that have been implicated in the disease, hetnochromatosis.
Chlorine compounds are found in dialysis water and are known to cause
moderate to severe hemolytic processes. Chlorine occurs in two principle
forms, chlorine and chloramine. Chloramine is the best known toxic element,
but both are strongly suspected of causing pathological effects on the reticu-
loendothelial system. One significant outcome of an impaired reticuloendo-
thelial system is the eventual requirement for blood transfusions (Kjellstrand
1978).
Patients on maintenance hemodialysis are almost invariably anemic. Three
mechanisms of red-cell injury are identified as complications of hemodialysis.
Oxidant red cell destruction has been found after exposure to dialysate con-
taining such agents as copper, chlorazine, or nitrate. Thermal red-cell
injury has been observed, when overheated dialysate was used, with resultant
spherocyte formation, increased osmotic fragility, and red-cell destruction
(Kjellstrand 1978). Drug-induced hemolysis on a hypersensitivity basis has
also been seen.
Medications Used by Hemodialysis Patients and Other Factors
Patients on maintenance hemodialysis are exposed to a multitude of sub-
stances which may be toxic, or invoke hypersensitivity. Oral drugs, intra-
venous medications containing stabilizers and preservatives, blood products,
chelating agents in transfused blood, traces of chemicals, such as formalde-
hyde and ethylene oxide plasticizers from tubing or bags of intravenous
fluids, particles of plastics, particles of glass from drug vials, and even
particles of rubber from stoppers of multiple dose vials are just a few
potential hazards (Hoy etal. 1979).
28
-------
Ethylene oxide from EO-sterilized equipment can cause systemic allereic
reactions, thiouram leached from rubber has caused allergic contact sensi-
tivity in a group of maintenance dialysis patients, and heparin can cause
eosinophilia. Anaphlactoid reactions due to nonimmune complex serum protein
aggregation have been identified, and a model for this may exist in mainte-
nance dialyzed patients exposed to reinfusion of plasma proteins altered after
adherence to and aggregation on the d'ialyzer membrane. The presence of cir-
culating free DNA during dialysis, and the increased prevalence of antinuclear
antibody in maintenance dialyzed patients attest to reinfusion of and sensiti-
zation to nuclear material as a result of white cell disruption on the mem-
brane. Formaldehyde reacts with and alters amino acids, proteins, nucleic
acids, nucleosides, nucleotides, nucleoproteins, and red cell membranes,
rendering them more antigenic (Patterson et al. 1980).
Patients have been reported to experience a clinical manifestation
labeled as "new-dialyzer syndrome" which may often go unrecorded. This
occurs: 1) because of the complexity of factors that may cause reactions
during hemodialysis, including the blood tubing, the dialyzer, the water or
concentrate used to make the dialysate, or the intravenous solution used to
fill the extracorporeal blood circuit; and 2) because of the variety of agents
that could potentially cause adverse reactions, including particulates,
plasticizers, stimulant residuals or reaction products, pyrogens, or bacteria.
Most dialysis facilities are not equipped to pursue the exhaustive evaluation
necessary to define a causative agent in the occasional acute reaction
(Ogden 1980).
It is not known whether the body burden of aluminum increases in patients
with normal renal function who are given aluminum-containing antacids over an
extended time, such as patients with chronic peptic-ulcer disease. However,
tissue burdens of aluminum were found to be markedly altered in uremic
patients on dialysis receiving aluminum-containing, phosphate-binding gels.
Brain gray-matter aluminum was higher in all patients with the dialysis-
associated encephalopathy syndrome than any of the control subjects or the
uremic patients on dialysis who died of other causes. With the exception of
29
-------
aluminum, other trace-element abnormalities have not been consistently
observed in patients who died of the dialysis encephalopathy syndrome. There
is some indirect evidence that this neurologic syndrome may be related to
aluminum toxicity (Alfrey 1976).
As part of the hemodialysis process, the patient's blood is heparinized
to prevent clotting in the extracorporeal circuit. Heparin causes a reduction
of the clotting mechanism, particularly in hemodialysis, and is contraindi-
cated in patients who require surgical procedures and have a faulty clotting
mechanism.
These and other agents could all cause sensitization, especially in
patients with an atopic diathesis. These profound pathological systemic
"shock waves" on the human organism are only a small part of the overall clin-
ical manifestations displayed by the ESRD patient, who usually has a primary
clinical syndrome with farreaching compounding "shock waves" of its own. This
overall complex situation makes identification of any single cause-effect
relationship very cumbersome and scientifically questionable.*
2.2.2 Peritoneal Dialysis
Acute peritoneal dialysis requires an invasive surgical procedure.
procedure is accomplished by inserting a temporary sterile plastic catheter
into one of the pelvic gutters and irrigating the peritoneum with sterile
solutions. A small stab wound is made in the midline, one-third of the way
from the umbilicus to the symphysis pubis. Upon proper placement of the
catheter, one liter of solution is infused rapidly and allowed to drain
immediately to test the adequacy of drainage. A purse-string suture is placed
in the skin around the puncture-site to minimize leakage, a sterile dressing
is applied, and the catheter is taped in place and connected to a sterile
closed infusion-drainage system.
*Blackburn and Osweroff interview, 1981, p. 17.
30
-------
The irrigation schedule is usually organized into inflow, dwell, and out-
flow periods. Inflow of two liters of solution is usually accomplished within
five to ten minutes. The fluid is allowed to remain in the abdomen for a
30-tninute dwell period, and is then drained by gravity into a sterile'closed
system over a period of 15 to 20 minutes. Thus, a single two-liter exchange
requires about one hour. Shortening the dwell-time enhances the removal of
small molecules, such as urea, but increases the amount of dialysis solution
needed. Duration of treatment varies from 24 to 72 hours, depending on
clinical indications. Due to the duration of treatment time and because this
process is a surgical procedure, peritoneal dialysis is usually accomplished
in a hospital setting. Peritoneal dialysis removes fluid efficiently, will
not precipitate bleeding, and avoids sacrificing vessels that may later be
needed for maintenance dialysis (Guyton 1956).
The placement of a permanent peritoneal catheter (Tenchoff catheter) for
chronic peritoneal dialysis also involves a surgical procedure. Placement of
a permanent catheter is done by making an incision in the midline and putting
the catheter into the peritoneal cavity. These permanent catheters have a
felt cuff which is placed above the peritoneal membrane. The felt cuff is
sealed by growth of fibrous tissue around it, thus closing the hole to the
peritoneal cavity. The catheter is directed out subcutaneously along the
abdominal wall two to three inches from the point at which it is inserted into
the peritoneal cavity. It then exits from the peritoneal wall skin site.
Chronic peritoneal dialysis involves two-liter exchanges generally
speaking every half hour. It requires 40 to 60 hours of the two-liter, half-
hour exchanges weekly depending upon the patient's residual renal function.
Peritoneal dialysis can easily be performed at home. It is felt that the
technique is much less complex than hemodialysis.
The main problem with chronic peritoneal dialysis is infection in the
peritoneal cavity (peritonitis). This infection is usually due to poor tech-
nique in handling the peritoneal catheter. With use of the chronic indwelling
peritoneal catheter (Tenchoff catheter), the occurrence of peritoneal
infection has decreased to one to two infections per six to twelve months.
31
-------
Peritoneal dialysis equipment includes:
• Peritoneal catheter as the entrance/exit mechanism into the peritoneal
cavity
• Tubing system used to conduct the ingress and egress of the dialysate
solution
\
• Dialysate component that primarily consists of water, electrolytes,
glucose, and added medication as required.
Each of these components, along with an occasionally utilized rotary pump,
provide potentialities to the patient similiar to those present in hemo-
dialysis for the introduction of contaminants and adverse iatrogenic
reactions.
There are several peritoneal dialysis machines. One is the reverse
osmosis peritoneal dialysis machine, which uses a concentrate of peritoneal
dialysate similar to that used in hemodialysis. The machine dilutes the
concentrate similarly to the hemodialysis machine, and produces a dialysate of
composition similar to plasma. Pure water is produced by a reverse osmosis
component in this peritoneal dialysis machine. Another type of peritoneal
dialysis machine is called the peritoneal cycler, which is a simple machine
containing a heating element to heat the peritoneal dialysate fluid and does
not use concentrate. This machine uses two-liter bags of peritoneal dialysate
and generally can hold eight bags to run four half-hour cycles.
Factors Involved in Selection of Peritoneal Dialysis
Problems concerning the selection of patients for peritoneal dialysis
include the establishment of a blood access, the poor or unstable cardiovas-
cular condition of the patient, and psycho-social factors. Peritoneal
dialysis is less traumatic to the patient's cardiovascular system than hemo-
dialysis, because there is no reduction in circulating volume or strain on the
cardiac output. However, overfilling of the abdomen with elevation of the
diaphram can reduce ventilation and increase the symptoms of congestive heart
failure. These and other selection criteria, listed in table 2-1, cause many
clinicians to prefer peritoneal dialysis to other modalities.*
*Blackburn and Osweroff interview, 1981, p. 17.
32
-------
Peritoneal Catheter and Equipment
A significant complication of peritoneal dialysis has been both septic
and aseptic peritonitis. Septic peritonitis results when nonhomogenic bac-
teria are introduced into the peritoneum, while aseptic peritonitis is
attributed to the use of equipment sterilants.
In many instances, sterilization of machines utilized for peritoneal
dialysis is carried out with formaldehyde and acetic acid. The Gandhi et al.
(1979) study review found 15 to 30 percent of peritonitis occurring in
patients on peritoneal dialysis to be characterized as aseptic in nature. The
cause of the peritonitis was ascribed to occasional failure of the reverse-
osmosis membrane adequately to remove endotoxin from the input water. Gandhi
et al. (1979) also found evidence to incriminate the dialysis concentrate as a
source of endotoxin. Their most significant find was that both formaldehyde
and phenols pass through reverse-osmosis membranes. In addition, they pointed
out that these compounds are irritants, and could conceivably cause a sterile
peritonitis.
Peritoneal dialysis patients, like their hemodialysis counterparts, must
be provided a dialysate with controlled bacterial levels. While peritoneal
dialysis patients also must be provided a sterile pyrogen-free dialyzing
fluid, the complexity of the extracorporal circuits and the frequency of the
patients' exposure to them affords great opportunity for the common, but
complex, problems of sepsis and pyrogen reactions in the dialysis population.
Tubing System Component
The tubing utilized for peritoneal dialysis is similar to that of the
tubing used in hemodialysis. However, it is agreed that there is little
contact between blood and peritoneal dialysate or peritoneal dialysis tubing,
except for the indwelling Tenchoff permanent peritoneal catheter. This
catheter may contact structures in the peritoneal cavity, including the
omentum and the intestines, as well as the site of catheter placement. It is
possible under certain circumstances that blood may contact the Tenchoff
catheter, if the layer of peritoneum covering the omentum and the intestines
is exposed.
.33
-------
The peritoneal dialysis patient is exposed to approximately 3 to 5 mg. of
DEHP per treatment day (Harris 1981).
Dialysate Component
The composition of a typical commercially available peritoneal dialysis
solution mimics the electrolyte composition of interstitial fluid, but differs
in that acetate replaces bicarbonate. Heparin, sodium and, if needed, potas-
sium chloride also may be added to peritoneal dialysate.
Glucose, which is absorbed from the peritoneum, can be used as an osmotic
force to balance the oncotic pressure in the plasma and to produce ultrafil-
tration in higher concentrations. All peritoneal dialysis solutions are
prewartned to 37 degrees Celsius to promote efficient diffusion and prevent
cooling of the patient with resulting vasoconstriction. The latter may induce
abdominal pain, abdominal cramps, and also impair the efficiency of the
diffusion exchange.
The same dialysate problems faced by the hemodialysis patient are
generally true for the peritoneal dialysis patient, although in peritoneal
dialysis the dialysate enters the peritoneal cavity, providing a much slower
absorption ratio for contaminants than the blood route used by the hemo-
dialysis patient.
Medications Used by Peritoneal Dialysis Patients and Other Factors
Like hemodialysis patients, peritoneal dialysis patients also face an
onslaught of medications and other additives during their dialysis. Medica-
tions may be added directly to the two-liter bags of peritoneal dialysate
fluids, especially for those patients on a cycler or who are receiving acute
peritoneal dialysis with a manual technique.
Patients who are on a reverse osmosis peritoneal machine use concentrate,
and thus do not often have additives added to their solution. However, the
effect of the burdens on their systems is of no less consequence than that
discussed earlier in the hemodialysis section (table 2-2).
34
-------
2.2.3 Continuous Ambulatory Peritoneal Dialysis^
Continuous ambulatory peritoneal dialysis (CAPD) utilizes the continuous
presence of dialysate fluid in the peritoneum, 24 hours per day, seven days
per week. Dialysate is drained from the abdomen and replaced with fresh solu-
tion three to five times each day. The schedule of solution exchange is
adjusted to meet individual patient needs. However, daytime exchanges usually
last from four to six hours and the overnight exchange from 8 to 12 hours.
CAPD is a closed system consisting of the peritoneal cavity, a chronic
indwelling Tenchoff catheter, a 42-inch-long connecting tugin, and the collap-
sible dialysate bag (Fraedrick etal. 1980). The empty dialysis solution bag
and tubing are folded compactly and carried in the clothing or in a small
fabric pouch until time for drainage of the dialysate solution. Drainage is
accomplished by gravity, and the filled bag is disconnected and discarded.
Then, another bag is attached from an elevated position for the inflow cycle.
No machinery is required.
CAPD has been found to be a better dialysis modality for treatment of
acute renal failure for patients who have severe cardiovascular disease and
for those who have had cardiovascular intervention. This is because hemodi-
alysis often results in a negative influence on the often critical hemodynamic
situation of the patient.
Intermittant dialysis, especially hemodialysis, seems to be disadvanta-
geous because of its relatively short and intensive elevation of perenal sub-
stances. By comparison, CAPD guarantees a continuous exchange of metabolites
and volume, comparable to physiological renal performance.
A variation of CAPD called CCPD, or continuous cyclic peritoneal
dialysis, is employed when daytime exchanges are inconvenient or impossible
for the CAPD patient because of working conditions. Suitable clearances of
nitrogenous wastes of metabolites are obtained by using a single all-day
peritoneal exchange, and then connecting the catheter to a continuous cycling
machine during the sleeping hours in order to increase the efficiency of the
nocturnal period.
35
-------
Factors involved in Selection for CAPD
CAPD requires strict aseptic conditions in exchanging the dialysis
solution and connecting tubing. If this is violated, peritonitis is sure to
follow (Fraedrick et al. 1980).
CAPD is a relatively new technique with a highly selective criteria and,
therefore, limited general use. Current estimates suggest its suitability for
from 20 to 35 percent of the ESRD population. The patient goes through an
extensive evaluation process and, if selected as a CAPD patient, undergoes a
teaching program which addresses the understanding of peritoneal dialysis,
dialysate solutions, cycle times, dietary requirements, medications, fluid
balance, techniques of monitoring weight and vital signs, procedures, compli-
cations, catheter care, and procedures for ordering supplies (Fraedrick et al.
1980).
CAPD patients tend to have higher hematocrits, improved energy, and have
a more desirable lifestyle because of the freedom from the machine, the
ability to travel, take holidays, vacations, etc. Because of its relative
simplicity, flexibility, and recent significant increases in the availability
of supplies, technology, and third-party payment, CAPD will probably become
the dialysis modality of choice for patients who meet the criteria. Table 2-1
lists these and other selection criteria per modality.
The processes, materials, and toxic substances to which the CAPD patients
are exposed are the same as those in patients receiving peritoneal dialysis
(table 2-2). Many of these also are common to the hemodialysis patient. The
basic exception would be the length of exposure and the individual patient's
level of involvement in their own treatment modality. However, peritoneal
dialysis and CAPD patients were initially found to exhibit a high incidence of
peritonitis.
Lately, there has been an appreciable decline in the incidence of peri-
tonitis, which is due possibly to a stricter adherence to the criteria for
sterile conditions required during the preparation and changing of the
dialysis bags by CAPD patients. There has also been improvement in the
36
-------
adapter technology with the introduction of the teflon connectors. Large
current programs are reporting total incidence of cloudy return fluid in the
range of one episode per 6-8 patient months of CAPD experience. There are
currently over 2,500 patients on CAPD in the United States, 120 in the group
monitored by the New York State Institute and 70 in a single program in the
Washington D.C. area.
37
-------
CHAPTER 3
TOXICOLOGICAL FACTORS OF DEHP EXPOSURE IN RENAL DIALYSIS PATIENTS
P. Wagner, M.S., and I. Marks, M.P.H.
Because DEHP produces toxic effects similar to those
produced by a host of other chemicals and medical conditions, it
is unlikely that DEHP can be identified as either a primary or
secondary cause of toxic symptoms in dialysis patients. While
the human effects of ingesting DEHP have been reported to be
minimal, animal studies have shown DEHP to produce mild liver
dysfunction. Preliminary results of a two-year bioassay suggest
DEHP may be a hepatocarcinogen. However, dialysis patients are
generally in poor heal.th and subject to a multiplicity of drug
therapies. The symptoms of disease plus the toxic effects of any
one or a combination of drugs may confound, mask, or act
synergistically with DEHP to produce similar toxic effects.
This chapter addresses the toxicity of di-2-ethylhexylphthalate (DEHP) in
humans and the feasibility of an epidemiologic investigation of mortality and
morbidity caused by such exposure in a population of renal dialysis patients.
DEHP is a plasticizer commonly used in blood storage bags, medical tubing, and
other medical paraphernalia used in dialysis treatment. DEHP is a phthalate
ester prepared from the reaction of phthalic acid with the alcohol, 2-ethyl
hexanol. Phthalate esters are primarily used as plasticizing agents in poly-
vinyl chloride (PVC) to give the polymer its flexibility and softness, and may
account for up to 40 percent of the final weight of the PVC product. DEHP and
its isomer, dioctyl phthalate (OOP), are currently the most widely used
plasticizers (Autian 1973). The chapter discusses the following topics:
• Animal toxicity of DEHP
• Biodistribution of DEHP
• Relevant animal studies
• Extrapolation to human health effects
• Synergistic effects of DEHP with contaminants or drugs
• Confounding effects of medical problems in end stage renal disease.
The likelihood of identifying DEHP effects in an epidemiologic study of
dialysis patients is then assessed.
38
-------
Although DEHP is insoluble in water, it is soluble in organic solvents
and oils. Because DEHP is not chemically bound, but only dispersed in the
polymer matrix, it can be leached out by solubilizing substances. Human blood
plasma is such a substance. The solubility of DEHP in human blood plasma has
been reported to approach 5 mg/ml (Miripol et al. 1976). Dialysis patients
are known to receive relatively high exposures to DEHP leached from the
plastic tubing used in dialysis treatment.
3.1 REVIEW OF DEHP TOXICITY IN ANIMALS
Acute toxicity data for DEHP indicates that the compound is relatively
nontoxic when exposure results from a one-time dose. The values for the LD50
range from 2 to 31 g/kg for intraperitoneal injections in rats to 128 g/kg for
intraperitoneal injections in mice. The oral LD50 for rats and rabbits is
reported to be 30 g/kg (Rubin et al. 1973). Humans have reportedly ingested
up to 10 g DEHP and experienced either mild diarrhea or no effects (Shaffer et
al. 1945).
Longer term studies on experimental animals have shown some decrease in
growth patterns and alteration in liver and kidney weights, when the animals
were fed doses of DEHP from 200 to 3400 mg/kg (Shaffer et al. 1945, Carpenter
et al. 1953, Nikonorow et al. 1973).
Sherman rats were fed 0.02, 0.06, and 0.20 gm/kg DEHP in their diets for
13 weeks. Growth retardation was reported and ratios of liver and kidney
weight to body weight were increased, but no significant histopathologic or
hematologic changes were noted (Shaffer et al. 1945).
In a three-month study on rats, a dose of 3.4 g/kg/day showed only weight
loss and diarrhea. A 12-month study, using the same dosage, produced a
decreased weight gain and liver enlargement (Nikonorow et al. 1973). In a
14-week study on dogs, one animal gavaged with 5 gm/kg/day DEHP showed some
chronic cholecystitis. However, this animal as well as the others in the
experiment had normal hematology and organ weights (Harris et al. 1956). In a
study where rats were dosed IV with 250 to 300 mg/kg DEHP dissolved in
39
-------
surfactants, the animals showed acute respiratory distress and hemorrhagic
congestion (Schultz et al. 1975).
More recent studies have focused on the biochemical measurement of the
toxic effects of DEHP exposure. These studies have shown lowered serum
cholesterol levels and inhibitory effects on hepatic lipid biosynthesis. Male
Sprague-Dawley rats were fed diets containing one and two percent DEHP for 21
days (Bell et al. 1978). The animals were then killed and the tissues
analyzed. The liver weights, expressed as a percentage of body weights, were
increased 70 percent above control values. Serum cholesterol levels were
significantly lowered (78 percent of control value) by the seventh day of
exposure to DEHP and this low value was maintained throughout the 21-day
study. Liver cholesterol values in the rats fed DEHP did not differ signif-
icantly from the values for the controls.
The investigators noted an inhibitory effect on hepatic lipid biosyn-
14
thesis as determined by decreased incorporation of C-acetate into hepatic
lipids. An inhibition of phospholipid and triglyceride biosynthesis was
postulated to be related to an impairment of fatty acid esterification.
Livers from rats fed 0.5 percent DEHP for nine days showed reduced uptake of
3
H-oleate. These effects confirmed earlier results by Bell and Nazir (1976).
In the earlier study, the investigators examined the effect of 0.5 per-
14
cent dietary DEHP on the incorporation of C-acetate in lipids by rat kid-
14
neys. They reported a significant decrease of incorporation of C-acetate
only in the triglyceride and sterol ester + hydrocarbon fraction. In another
experiment (Bell et al. 1976), the kidneys of rats fed one percent DEHP in
their diet for 18 days did not differ significantly in their ability to
14
incorporate C-acetate into total lipids. The authors interpreted the exper-
imental results as an ability of the kidney to normalize alterations in lipid
metabolism.
Preliminary and unreviewed reports from the National Cancer Institute,
Information Division, suggest that DEHP may be carcinogenic. A two-year
40
-------
chronic feeding study of DEHP produced hepatocellular carcinomas and neo-
plastic nodules in male and female rats, and adenomas and hepatocellular
carcinomas in male and female mice. However, the doses used in this study
were high—6,000 and 12,000 ppm for rats and 3,000 and 6,000 ppm for mice.
Extrapolation of the effects seen at those dose-levels to patients receiving
doses orders of magnitude lower is tenuous.
3.2 BIODISTRIBUTION
Biodistribution studies in vivo have been reported where radioactive DEHP
was administered either orally or by IV. Because DEHP is insoluble in water,
many of these studies used a variety of solubilizing or emulsifying agents.
These agents appear to have affected the distribution of DEHP in the tissue.
Stein et al. (1974) reported DEHP accumulated in significant amounts in
the heart and epididymal fat pads when fed to rats at 0.1 percent in the diet.
Schultz and Rubin (1975) reported that 200 mg/kg DEHP emulsified in albumin
was found in high concentrations in the liver, moderate concentrations in the
lung and spleen, and low concentrations in the fat and kidney. However, mice
dosed at 115 mg/kg IV with DEHP in serum had high concentrations of the
chemicals in the liver and kidney and low concentrations in the lung, spleen,
and fat (Thomas et al. 1978, Waddell). Rats dosed at 600 mg/kg IV with DEHP
solubilized with oleic acid were reported to have high concentrations of the
chemical in the liver and lung, but low concentrations in the spleen and fat.
In addition to the effect of the solubilizing and emulsifying agents on
distribution patterns, they also appear to produce toxic effects themselves
and with DEHP. Rubin (1975) studied the effects of intravenous doses of DEHP
solubilized in a 13 percent aqueous solution of detergent on the blood
pressure in rats. In some cases, doses at 80 to 100 mg/kg resulted in blood
pressure falling below 30mm Hg, causing death. Doses of 40 mg/kg resulted in
a drop in blood pressure of 18 to 27 mm Hg. However, detergent alone will
cause a hypertensive reaction when injected IV and DEHP produces a hypotensive
effect administered by the same route. Thus, the drop in blood pressure, when
the solubilized DEHP is administered, is the resultant of the two contrary
41
-------
effects. The ability of DEHP to induce hypotension is greater than the hyper-
tensive effect of the detergent.
DEHP is thought to metabolize to water-soluble acids, alcohol, and
ketones that are excreted in the urine of normal animals and humans (Albro et
al. 1973, Daniel et al. 1974). The clearance is about 60 to 90 percent in 24
hours (Rubin et al. 1976).
DEHP has been found in the adipose tissue, liver, brain, heart, lung and
spleen of experimental animals and humans (Rubin, et al. 1973, Thomas et al.
1978, Waddell, Stein et al. 1974). No data has been found on the retention
and storage of the metabolites by either nephrectomized experimental animals
or dialysis patients. However, in one study (Chen et al. 1978), nephrecto-
mized dogs were injected IV with 225 mg/kg DEHP. Analyses of serum showed
that levels of DEHP in the nephrectomized dogs were higher for a longer period
of time than in control dogs. The 72-hr and 96-hr serum concentrations of
DEHP were similar in both groups. The investigators interpreted this finding
as likely evidence of storage and/or metabolism of DEHP by nephrectomized
dogs, but postulated no metabolic pathway. The dogs were killed after four
days and their tissues analyzed for DEHP. The highest tissue concentration
was found in the lungs in both nephrectomized and control dogs.
Human kidneys from 17 persons dying of causes other than renal failure
were analyzed for DEHP (Overturf et al. 1979). Extent and duration of
exposure to DEHP was unknown in each case. Two kidneys were diagnosed as
nephrosclerotic; the remaining 15 were normal. Both nephrosclerotic kidneys
contained measurable amounts of DEHP in the cortical and medullary tissues,
but DEHP was found in only two of the normal kidneys. The authors presented
no explanations for their findings. Any conclusions based on these findings
are difficult, since extent or duration of exposure to DEHP is unknown. The
chronic toxicity of the metabolites is not known, since in normal circum-
stances they are rapidly excreted.
Stern et al. (1977) demonstrated that the disposition and effects of DEHP
can be dependent upon dosage and the form of administration. In a series of
42
-------
experiments, DEHP was emulsified with a surfactant, or solubilized in ethanol
and added to plasma, or leached by plasma from polyvinyl chloride plastic.
Groups of rats (15 each) were then injected with 2.5 mg/kg of DEHP in each
form, including the DEHP-surfactant form undiluted with plasma. Kinetic
studies performe<
rapid clearance.
14
studies performed with. C-DEHP showed that the leached DEHP had the most
The kinetics followed a biphasic pattern, with the half-lives for the
second phase showing the leached form as having a significantly reduced half-
life over the other administered forms (31 minutes vs. 263, 83, and 18]). An
examination of tissue residues showed a greater retention by the liver, lungs,
and spleen of the non-leached forms of DEHP than the leached form. The inves-
tigators postulated that the excretion rates of DEHP would probably be differ-
ent since organ retention of the compound differed. Analyses of urine and
feces showed that in the first seven hours after infusion the leached DEHP was
excreted at approximately twice the rate of the emulsified compound, but from
14 to 44 hours the rates were approximately the same.
The investigators conclude that experiments designed to elucidate the
effects of DEHP in humans should parallel actual human doses and dosage forms
as closely as possible. Using doses or dosage forms not encountered by man
may well introduce artifacts into the evaluation which will inhibit or distort
a valid conclusion.
In summary, animal studies, primarily feeding studies, using doses of
DEHP of 20 to 5,000 mg/kg daily produced the following effects: reduced
weight gain, increased liver and kidney weights, lowered serum cholesterol,
inhibition of hepatic lipid biosynthesis, and a transitory decrease in the
renal lipid biosynthetic process. However, as shown by Stern et al. (1977),
routes of exposure and doses not consistent with actual exposure oftentimes
lead to erroneous interpretations, as well as altering experimental results.
For these reasons, the only studies considered to be relevant to dialysis
patients are those where the DEHP dose is actually leached from P^c tubing or
plastic bags. These relevant studies are discussed below.
43
-------
3.3 RELEVANT ANIMAL STUDIES
Miripol et al. (1976) injected rats with plasma containing 1.0 and 3.7 ing
DEHP/kg twice a week for 63 days. The authors state that a dose of 3.7 mg/kg
is equivalent to what a 70 kg human would receive in an exchange transfusion
of 21-day-old blood. Rats injected with serum or not injected at all were
used as controls. Rats from each group were killed after eight injections and
19 injections, or 63 days after the last injection. No significant differ-
ences were seen for the following parameters in rats treated with DEHP: sur-
vival rate, growth rate, general behavior, hemograms, serum chemistry values,
liver functions, absolute or relative organ weights, plasma and tissue levels
of DEHP, or gross and microscopic pathology.
The effects of DEHP on rhesus monkeys undergoing chronic transfusion
therapy were studied by Jacobson et al. (1977). Immature rhesus monkeys were
divided into the following groups: three monkeys were transfused weekly with
15 ml of platelet-rich plasma stored in PVC; two monkeys were transfused
biweekly for six months with 15 ml of platelet poor plasma stored in P^C; two
monkeys were transfused weekly for six months with 15 ml of platelet-rich
blood stored in polyethylene; three monkeys were not transfused. The monkeys
received dosages of DEHP of 6.6 to 33 mg/kg. This amount compares favorably
(2.1 to 27.5 mg/kg) to the amount of DEHP received by human pediatric patients
monitored for one year.
Serum chemistries, histopathology, and liver-spleen scans were used as
diagnostic tools. Serum chemistries obtained at four-month intervals were
normal throughout the course of the experiments. Liver-spleen scans showed
decreased activity ratios in three monkeys transfused with PVC-platelet-rich
plasma after three months of treatment, which persisted 14 months after
termination of the experiment. Plasma disappearance curves for BSP follow a
single, exponential function for immature primates and human children.
Adults, both human and primate, exhibit a biexponential function.
Children with asymptomatic liver disease, for example cystic fibrosis accom-
panied by hepatomegaly, will follow a biexponential function. In this present
study, BSP transport was used as an indicator of liver abnormality. Three of
44
-------
the five monkeys who received the PVC-platelet-rich blood demonstrated biexpo-
nential disappearance curves after one year of transfusion therapy. Fourteen
months after cessation of treatment, the three monkeys still had biexponential
disappearance curves. Liver biopsies showed abnormal histologic findings in
four of the five monkeys transfused with PVC-platelet-rich plasma. Followup
biopsies at five and 14 months after cessation of treatment showed diffuse
portal lymphotocytic infiltration, hepatic necrosis, and evidence of
binucleate cells." The effects seen in PVC-plasma were not seen in the control
animals.
In summarizing the effects seen in monkeys, only the BSP clearance and
the liver-spleen scan are routinely performed on living human subjects.
Abnormal results for these procedures can be caused by a variety of clinical
conditions and/or exposure to numerous chemicals. For example, ab-normal BSP
clearance can occur with infectious hepatitis or portal cirrhosis.
3.4 EXTRAPOLATION TO HUMAN HEALTH EFFECTS
Extrapolation of animal data on health effects to humans should be done
with care. Factors to be considered include:
• Do the animals metabolize the chemical in a manner
similar to humans?
• Are the biodistribution and bioaccumulation patterns
similar?
• Are the toxic signs the same?
• Are dose-response curves similar?
• Are the animal experiments designed so that the exposure
route and dosage are valid for human exposures?
Obviously, the answers to the above questions should be "yes" in so far as
possible.
However, other more basic problems are also present. Test animals of the
same species and strain are generally extremely similar genetically, espe-
cially rats and mice. These two species, used widely to test chemicals, have
45
-------
little innate resistance to diseases, develop many varieties of spontaneous
tumors, and may be chronically infected with respiratory diseases or hepati-
tis. Extrapolation of data derived from these genetically-similar animals to
humans, who are genetically dissimilar, should be done carefully.
In addition, test animals are kept in controlled environments, and fed
controlled diets. Humans, on the other hand, are exposed to a variety of
environments and differ widely in their dietary habits. Humans also expose
themselves to drugs and pollutants which may or may not act synergistically.
Finally, humans appear to have detoxification mechanisms that are more sophis-
ticated than those of most animal species. A toxic effect seen in animals may
not necessarily be present in all humans similarly exposed.
3.5 TOXICOLOGICAL ENVIRONMENT OF DIALYSIS PATIENTS
Dialysis patients are often in poor health. They may have a variety of
other clinical conditions, including hepatitis, diabetes, cancer, cardiovas-
cular disease, arthritis, etc. A survey conducted of hospitalized chronic
hemodialysis patients in 27 dialysis centers in Michigan (Mayor et al. 1979)
showed that 11.4 percent of the patients carried the Hepatitis B surface
antigens distributed in 21 of the 27 dialysis units. Antibodies for the
Hepatitis B antigen was found in 31 percent of the patients from selected
units. The survey also showed that 26 percent of the Hepatitis B antigen
carriers had not been previously identified. Some of this information is
addressed in detail in Task Report No. 1.
Dialysis patients are also routinely exposed to multiple drug therapies.
Heparin, an anticoagulant, is routinely given to dialysis patients. Gelfand
et al. (1978) have reported that free thyroxine levels rise after heparin
administration for hemodialysis. Heparin is also thought to antagonize the
action of ACTH, insulin, and corticoids. Dialysis patients that are candi-
dates for renal transplants may be treated with immunosuppressant drugs such
as azathioprine, which may cause liver damage characterized by elevated alka-
line phosphatase levels and slightly elevated bilirubin. Patients treated
with immunosuppressants may ultimately develop cancer and be treated with
46
-------
chemotherapeuti.es such as methotrexate, which is a hepatotoxin and may cause
liver atrophy, necrosis, cirrhosis, and periportal fibrosis. In addition,
dialysis patients, may be routinely administered drugs such as tranquilizers,
barbiturates, and analgesics which may or may not be synergistic with DEHP and
routine drug therapies.
DEHP is reported to show synergistic effects with pentobarbital and
methaqualone (Seth et al. 1977). Since these compounds are CNS depressants,
synergism may exist with other CNS depressants as well.
In addition to the various clinical conditions that affect dialysis
patients and the toxic effects caused by routine drug therapies, the dialysis
patient is exposed to other toxic chemicals present in the dialysis equipment
and assorted support chemicals. The PVC tubing used in hemodialysis units has
calcium, zinc, and tin added as stabilizers (Geertz et al. 1974). The concen-
trations of these stabilizers range from 110 to 31 ppm for zinc to approxi-
mately 0.3 ppm for tin and 102 to 2.5 ppm for calcium.
These metals may be leached from the medical devices during the dialysis
procedures. Ionized calcium leached into blood or serum may cause such
problems as soft tissue calcification and depression of phosphorus levels in
the blood. Cadmium, because of its chemical similarity to zinc, generally
occurs as a trace impurity in zinc and zinc products. Cadmium is a nephro-
toxin and is stored in the body in the kidneys. Its effects are cumulative.
The dialysis machine itself may release copper, which is reported to cause
hemolysis and fever (Kjellstrand 1978). Sterilizing agents used on membranes
and filters commonly contain formaldehyde (Kjellstrand 1978, Orringer et al.
1976, Sandier et al. 1979, Fassbinder et al. 1979). The exposure of dialysis
patients to the formaldehyde in these solutions has been implicated in
hemolysis due to the induction of anti-N or anti-N-like antibodies (Sandier et
_al_. 1979, Fassbinder et al. 1979).
The water used in dialysis may also contain numerous impurities
(Kjellstrand 1978). Because of the extreme geographic variations, the trace
impurities in water naturally vary greatly from location to location.
47
-------
Contaminants can also vary because of industrial activity which may discharge
chemicals into water used as municipal water supplies. The efficiencies of
individual water treatment plants is also a consideration in the levels and
types of trace contaminants in water used in dialysis treatment. Most water
treatment facilities add chlorine or chloramine to the untreated water.
Chloramine may cause methemoglobin formation with further oxidation to Heinz
Bodies. Heinz Bodies may accumulate and predispose erythrocytes to destruc-
tion by the spleen (Kjellstrand 1978). Chlorine may be contaminated with
carbon tetrachloride, a hepatotoxin and suspect human hepatocarcinogen.
Organic chemicals present in untreated water may undergo chlorination to form
chlorinated compounds such as chloroform, a hepatotoxin and suspect human
carcinogen.
Trace metals found in drinking water include aluminum, arsenic, iron,
cadmium, zinc, and copper. The. actions and interactions of these metals with
each other or the other chemicals dialysis patients are exposed to are not
known. Aluminum, however, is suspected of causing dialysis dementia
(Kjellstrand 1978). The reasons are four-fold:
•> Body storage of aluminum is higher in dialyzed uremic
patients as compared to dialyzed patients for other
causes
• Compartmentalization of aluminum differs in dialyzed
uremic patients from that of other dialyzed patients
• The occurence of the syndrome can be related to naturally
occuring aluminum-contaminated water
• The syndrome is preventable by removing aluminum.
3.6 CONCLUSION: FEASIBILITY OF ASSESSING DEHP EXPOSURE AMONG DIALYSIS
PATIENTS
In summary, a review of toxicity data for DEHP suggests the compound may
produce some generalized liver abnormalities in normal experimental animals.
Dialysis patients are exposed to DEHP due to its leaching into their blood
from plastic tubing and bags used in dialysis. Although data show the major
amount of the compound is metabolized to water-soluble products and excreted
48
-------
fairly rapidly in the urine, no data are available on these processes in
either nephrectomized animals or dialysis patients.
Thus, it is possible that the compound and/or its metabolites may be
stored in various tissues of the body or be excreted in other ways. Since
excretion of the metabolites in urine is not possible in dialysis patients,
the pharmacokinetics or metabolic pathways may be altered. Thus, it is
conceivable that the toxic effects seen in normal, animals and humans may not
be representative of the effects produced in this special population.
Further confounding factors are the general poor health of dialysis
patients, the multiplicity of drug therapies to which they are exposed, and
inadvertent exposures to both chemicals and disease because of the dialysis
treatments. Any one of these factors could mask, mimic, or act syngeristic-
ally with DEHP to produce toxic effects. Finally, given the general, non-
unique toxic effects of DEHP, it is not likely to be possible to identify this
chemical as either a primary or synergistic cause of liver abnormalites or
other toxic symptoms of dialysis patients.
49
-------
CHAPTER 4
AN ASSESSMENT OF THE UTILITY OF ESRD MIS DATA
FOR EPIDEMIOLOGIC RESEARCH
I. Marks, M.P.H., C. Steele, A.R.T., and
D. Wellington, Ph.D., etal.*
This chapter assesses the utility of the Health Care
Financing Administration's (HCFA) End Stage Renal Disease
Medical Information System (ESRD MIS) data base for conducting
epidemiologic research on persons with chronic renal failure.
The discussion is presented in two sections. In the first
section, ESRD MIS data are discussed in terms of the elements of
the ESRD MIS, the data collection system and its intended uses.
The second section addresses the difficulties involved in
considering ESRD MIS data as a basis for epidemiologic research.
Based on these data, the report concludes that the ESRD MIS lacks
the accuracy, consistency, and completeness necessary to perform
any meaningful analysis other than demographic analysis.
Since its advent in 1973, the Health Care Financing Administration's End
Stage Renal Disease Medical Information System has undergone a number of major
changes as to where, when, and how data are collected and stored. As a result
of these system changes, the type and quality of the data has been greatly
compromised. The purpose of this document is to describe the ESRD MIS data
base as it existed in the years 1977 through 1979, and to determine what type
of analyses can be performed given the specificity, completeness, consistency,
and accuracy of the data files.
4.1 THE END STAGE RENAL DISEASE MEDICAL INFORMATION SYSTEM: AN OVERVIEW
The End Stage Renal Disease Program was established by Public Law 92-603
in 1972. From 1973 to 1977 the ESRD MIS was administered by the Public Health
*The generous assistance of Michael McMullan and Sandra Kappert of the Health
Care Financing Administration, DHHS, is gratefully acknowledged. Drs.
Fraumeni and Hoover of the National Cancer Institute, DHHS, also provided
helpful advice and assistance during the research for this chapter.
50
-------
Service. In 1977, the Health Care Financing Administration (HCFA) was
assigned responsibility for the Program. Beginning in 1973, patient history
data along with billing information were collected under the auspices of the
Medicare Program. In 1976, the Medical Information System and its inclusive
components were added. The Program is currently administered by the Health
Care Financing Administration, Department of Health and Human Services.
The ESRD MIS, a component of the Program, is a nationwide administrative
reporting system with a limited amount of general medical information
included. The system is used as a management tool to predict trends, distri-
bution,, and utilization of services, and to assist in assessing performance
under legislative and regulatory mandates. The system also functions as a
locating system and may be useful as an index to assist in identifying certain
populations. In its present form, the ESRD MIS is of only limited value as a
source of accurate statistical incidence data.
Prior to 1973, registries were maintained by the American College of
Surgeons (the Renal Transplant Registry) and Research Triangle Institute (the
National Dialysis Registry). Historical data from these registeries were
transferred to form the basis of the ESRD MIS in 1973. Major modifications in
the system were instituted after 1976, and additional revisions are currently
underway.
During the period, 1977, 1978, 1979, and January to June 1980, eleven
documents comprised the MIS and its complementary systems including the
following:
• ESRD Patient History and Treatment Plan
• ESRD Outpatient Dialysis Service Information
• ESRD Transplant Tissue Typing Information
• ESRD Death Notification
• ESRD Center Patient Listing
• Inpatient Hospital and Skilled Nursing Facility Admission
and Billing, SSA-1453
51
-------
• Provider Billing for Medicare and Other Health Services,
SSA-1483
• Carrier Payment Records
• ESRD Facility Survey
• ESRD Cost Questionnaires
•• ESRD N'otices of Approval for Rendering ESRD
Services/Demographic and1 Reimbursement Parameters.
The following subsections of this section address the flow of information
to the ESRD MIS, describe the patient population, detail the content and
accuracy of the data elements and summarize the problems with the ESRD MIS
dataset.
4.1.1 The Flow of Information
The Social Security Administration (SSA) identifies and accretes ESRD
patients to their file. When an ESRD patient applies for Medicare benefits, a
Form SSA-2728, Chronic Renal Disease Medical Evidence Report, is completed and
sent to the SSA. The information on this form is utilized in making an enti-
tlement decision. The form remains in the patient's disability folder at SSA,
but its information is not conveyed to ESRD MIS.
At the time of first Medicare inpatient or outpatient billing, an ESRD
Patient History and Treatment Plan, HCFA-2742 (figure 4-1), should be
completed and sent to the ESRD MIS. For most patients, this occurs once.
However, certain circumstances can result in subsequent entries for a single
individual. If, for example, a dialysis patient is transplanted, losing
Medicare benefits, and later rejects, returning the dialysis, another history
must be submitted at the time of the first billing for any new entitlement
period.
The ESRD Outpatient Dialysis Service Information form (figure 4-2) is
submitted each time a bill is prepared for outpatient services; for most
maintenance dialysis patients, this would be once a month.
52
-------
Figure A-.1.. ESRD PATIENT HISTORY AND TREATMENT PLAN
ESRD PATIENT HISTORY AND TREATMENT PLAN
. DiSlASE MEDICAL INFORMATION STTEH
OM Or
HEALTH IMSUAAMC* CiAlM
4. PATIENTS COUNTY Of HCSIOCMCI.*
t. DATE OF BISTH
7. DATE OF RCTOOT
l_ _l
0«» Yr.
t PHOV1OER NAME ANOADO«C!S CITY AND STATII
I. PROVIDER NUMBER
1O. INTERMEDIARY MJMUR
i O
3O
30
40
|
5O I
a. miM
01 O >i»
ID a c
030 0
04-Q
as O
06 a
O7Q
080
090
IOO Obancm*
11 O
13 O
M O
15 O
UCTHOOS OF OlACMOmC
CONFIRMATION
1 O
3D
S Q
IS. IPTM£PATICNTKAXHAOAMAl.>CMANCrOIIIF*
HAUONANCT B CUKIUMTLT HttXtNT tbntrrmr
Irm
II. IUKO)CAL HSTOKr H>~X tU
17. TMANVUUtT MSTORY
1. _ _ Nwnratt
TREUMCKT PUD
It. DATE Of F4RST04HOM1C
MAUTTQUMCE OUU.TBS
LEVEL PKKJR TO FlfUT OlAlYm
1 . "1*
za. FUESEMT ot ALTBS ACCXS
3D
30
32. IF NOT A TMAMVUUTT
6 O On»f
33. FUUINCO TTK OF OIALYU
1 0
2.a
9 a
HJumtO CTALTia it I IINU
01 O
03 a
03 O
t IF PATIENT 14 NOT PLAN-
NED FOR HOME MALYSX
01 O F>
CUKKENT HEALTH
CTATUKO
TTATVIS TMKEE MONTHS
PKIOR TO FIR
a Or •• «*3- U^C- *M: 30 CFH.AOS. l«mo« 3133).
nr in M *•.-.> An (X 1974 a UAC «""•« Cm >wr bl.
771 Otominr JSA-2743)
UA/EIO
-------
FIGURE 4-2. ESRD PATIENT DIALYSIS SERVICE INFORMATION
BUD OUTPATIENT DIALYSIS SERVICE INFORMATION
END STAGE RENAL DISEASE MEDICAL INFORMATION SYSTEM
Form Aporov?d
OY3 Ko.6
LAST NAME
FIRST
I Ml
2. SEX
DM OF
3. HEALTH INSURANCE CLAIM NU.V.BER
4. PATIENTS COUNTY OF RESIDENCE*
5, STATE
I — —
6. DATE OF BIRTH
"Mpr "Day" ~Yr.~
7. STATEMENT COVERAGE PERIOD
from_ _!__ —1_ _ thru__ —1-, _ I —
~VrT
"MO. ""&,
ay
Yr.
B. KYtOVIDER SAME AND ADDRESS (CITY AND STATE)
9. PROVIDER NUMBER
10. INTERMEDIARY NUMBER
n. DCHECK HERE IF THE ATTACHED BILL INCLUDES NO CHARGES FOR DIALYSIS TREATMENTS OR SUPPLIES, AND SKIP TO ITEM s.
III. Total .
Dialysis .
Sessions
c ____
h
12. DIALYSIS SETTINGS (Check the patient'* usual tatting. H his imiiHstting changed
d-jrirtg.tha.month, chock both. If the patisnt wzs di»>yzed in any tettirtg on a temporary
beat, check the appropriate block (s) in the temporary column alto).
D Check h<£» rf treatment wes provided unoW «n alHnduwve rule, and do not
compute IV .nd V below. , .^ „. Temp.
(Check) .
OUTPATIENT
Full care' aO
Partial care2 fD gl
Self-care' kD lD m
Self-care training p D q D r '
HOME
Self-care (Volunteer,
or no assistant). "D vD w
Partial care (Paid
«3>ranr/ x D y D z •
1 — Patient performs few, if any, ineasaracJe dialysis tasks.
2 — Patient performs some, but not all. tasks.
3 — Patient performs all tasks without facility professional care.
IV. Charge
Per
Dialysis
d
I
n
t-
V. Total
Dialysis
Charges
e .
i.
o.
t
ROUTINE LABORATORY TESTS
13. IN-UNIT CHARGES S
14. OUT-OF-i-VT CHARGES $.
HOME DIALYSIS INFORMATION
15. METHOD OF EQUIPMENT PAYMENT
a D Rental b D Purchase
16. TOTAL PURCHASE PRICE S
17. PURCHASE OR RENTAL
MONTHLY CHARGE $
18. INSTALLATION CHARGE $
19. DISPOSABLE SUPPLIES *
2CL CHANGES IN USUAL DIALYSIS SETTING
(H there has been u change in the patient's
treatment plan tuch that heAhe now has m new
-usual or regular treatment tatting, indicate both
the previous and the new totting by cheeking
the appropriate "FROM"and TO"columns
below).
DIALYSIS SETTING
Outpatient-Full care
Outpatient-Partial care
Outpetient-Self-care
Outpatient-Self-care
training
Home-Self-care
Home Partial care
FROM
.D
cD
eD
gQ
i D
kD
ro
bD
dD
j D
i D
21. REASON IS) FOR CHANGE IN USUAL DIALYSIS
SETTING (Check all applicable reasons for change
indicated in Item 20).
• D Medical complications
b .D Self-care training initiated
c D Self-care training successfully completed
d D Patient unable to complete training
e D Change in home support
•f D Patient refuses to continue in present setting
g D Other
22. REASON(S) FOR TEMPORARY
OIALYS4S SETTINGS (Check below all
masons for any dialytis treatments
Indicated in Item 12 which were given
in m temporary setting; La., not the
patient's usual setting).
a D Medical complications
b O Change in home support
c D Mechanical failure of equipment
d D Patient on personal or job-related
travel
e Q Patient or assistant temporarily
unable to complete training
f O Patient receiving additional self-
care training
g O Patient waiting for availability of
space in another facility
h D Other
NUMBER OF DIALYSIS ACCESS PROCE-
DURES (New or revised) DURING THIS
PERIOD (Check none or enter number).
* DrJone
b _ Vascular
e _ Peritoneal
23. CURRENT DIALYSIS PRESCRIPTION
(Usual tatting)
_______ Hours per session
24- TYPE OF DIALYSIS (Check one)
a LJ Hemodialysis only
b Q Peritoneal dialysis only
c O Both (Indicate number of treatments).
• Hemodialysis
Peritoneal dialysis
NUMBER OF UNITS OF BLOOD GIVEN THIS
PERIOD (Enter number of units given of each
type. If number of units is unknown, enter "X"
. t appropriate category).
a D None
b _____ Frozen
c Leukocyte poor (Birffy coat free)
d _____Other
e Unknown type
27. PATIENT TREATMENT STATUS (Check one)
a D Dialysis only; not a transplant candidate
b D Transplant candidate
c D Transplanted, off dialysis
d D Transplanted, on support dialysis
e D Expired — _l_ —!_ _
Mo. Day Yr.
28. REMARKS
NOTE: *tf patient residence is not in a specific county, enter incorporated city or township.
This report is required by law (42. U.S.C. 426; 20 CFR 405, Section 2133). Individually identifiable patient information will not be disclosed except as provided
for in the Privacy Act of 1974 (5 U.S.C. 5520:45 CFR Part 5a).
Form SSA-2743 (10-76)
SSA/ESRD
Department of Health; Education, and Welfare
-------
The Inpatient Hospital and Skilled Nursing Facility Admission and Billing
form, SSA-1453 (figure 4-3), is submitted to fiscal intermediaries for
reimbursement of services rendered. Information from this form is sent to the
ESRD MIS through the Medicare masterfile.
Transplant Tissue Typing Information (figure 4-4) is submitted within two
weeks from the date of transplant.
o
A Death Notification (figure 4-5) is submitted when a death has occurred,
regardless of the place or cause of death.
Compliance was a major problem within the system as it existed from 1977
to 1979. Information requested by HCFA for the ESRD MIS was not required by
the Social Security Administration in order to receive Medicare benefits. If
bills were submitted to SSA for a patient, they were paid, regardless of
whether or not a Patient History and Treatment Plan had been submitted to
HCFA.
The ESRD MIS management at HCFA had asked that the fiscal intermediaries
police the system by requiring that a Patient History and Treatment Plan be
submitted for each patient prior to the intermediaries' payment of the first
bill. This procedure was in effect until two years ago. Unfortunately, it
became unwieldy and impractical.
Until January 1981, supplemental forms were sent in with bills by inter-
mediary or carrier. Communication between the intermediaries is poor and this
creates additional data problems. For example, an intermediary would handle
one State in which a patient lived; another intermediary would be responsible
for the contiguous State in which the patient was dialyzed. The MIS is
dependent on the intermediary to supply specific information that may or may
not be accurate. The intermediary pays on the basis of the bill. They may
adjust this payment, if there is a change, but do not take the time to adjust
the supplemental document. Billing records are maintained separately from a
patient's medical record.
55
-------
31
If.'PATIEKT HOSPITAL AND SKILLED NURSING FACILITY ADMISSION AND BILLING
HOiriTAl. AND MEDICAL If.'S'JRANCt-EINLriTS—SOCIAL SECURITY ACT
:. f-f.-f •: » :i
IMI
iniurjr.;c CHIT. nu-r.a
~. r^t.e-.t j i^3'«J .'E:-ss: .-.L-.-n=s:. Cirr. State. ZIP Code)
S. Da:e of birtr '6. Mdcil re::rc nurr.aer
.£. Provider name and address (City and Slate)
9. Provider number i 10. ArtenOin: pnyvi;ian
11. Dates cf c.uii:?y:n; stay
FROM i
:12. Qualifying and other prior stay information
THRU
If you have otrvtr hea!t.". insurance or if your State Medical Assistance Agency will pay part of your medical expenses and you want information
stiout this ciaim released to Uvem upon their request, complete items 13 and 14.
13. Insuring organization and .' or Suit agency name and.cddrns
14. Policy and / or m«4ica> as»tunce number
kase to tne Social S<
reou«t tra'. paymen
scunty Administration or its intermediaries or earners any information needed tor mis or a relates Mcsjisare claim. 1
t of amnoriietf benefits be made on my behalf.
r— ; Contained in j Signature (PatJent or authorized representative) (Signature by mark must be witnessed)
1 — ! provider's record j
16. Aam.ttmg oxgncses (II employment r*l»te4, also give namt
and address of employer)
Renal Transplant Cadaveric
18, Surgical procedures (Show Cxrt of Mdi)
Renal Transplant Cadaveric ( )
IS. STATEMENT OF SERVICES RENDERED
e*«<:
£""**
>Rtft 1 Pint!
Accommodation
B. 1-Bed
C. 2-3-i Bed
D. 5
FOR
H
0
f
1
A
CULT
or ra=rr SeC»
E. Intensive an
helnnXKX!
Days
T. Coronary are |
G- 1
pint
ToulCharrn
Do net u»«
17. Oicnarge or current diagnoses
(a) Primary
Renal Transplant Caaaveric
(b) Secondary
Ho~a~+r*iCt*l.
R«te §,^55SE^r^S«gS!s5S*:fiS^
H. Operating room
1. Anesthesia
J. Outpatient services
K. Blood asmmisrmran
L Pharmacy
M. Radiology
K. Laboratory
0. Medics, surjicai and central supplies
P. Physical therapy
Q. Occupational therapy
R. Speecr. oitnology
S. Inhalation tfwnsy
T. Otner (Describe.' 4
1
HDx (5, 123.00
PDx ?250.00 |
-
1
Coocverie Kidnev Accuisition 1 ... 1 1 i
U. TCTA-.S
! T--- i i !
V. Inaitien*. dedur.iSie ' :
W. B
X. C
Y. T
2S. 1
aoc oeauctibie
c:n$i.-ri-.ce
™A_ Di3UC~!C-NS
ms. e
days (
) ( ) •''].'•
1
eet-*y tMt t-e r«u:rec physicia.'.'s certification and recertificatien* are on file.
Si;nifjrt c' nv::e- re:-esenative
rurcn* "*? tBU»r>P» ^^
ZI. Datt S uar»m»* e( 22. Dvi
1 ' 1
1 1
23. bvta *clnw car* «nd«4 24. D«l
1 1
1 1
i i
25. ritiontmtus
A. p«t« tti»cA«ni*d B. O«u of e«
i i I
**• ^J*"J^J**"V" I?- N»«.e»w»W
PIP
Curte
IDcncx UM
n«« K9«ca
1
1
1
i I
« UR ncttc* umi^a
1 1 •
1 1
la b«n«ftn •mj4«s>»tf
1 1
. p'.stm
-
j;. Rt.mourvrm.nt.moun: S
FOR INTERMEDIARY US£
From Thni
II II
1 ' 1 II
3i. Appr»v«4 9y
U. Non. M. ban
1 Dai. tocn^e
\ : ;_
Social
N'CTi: This is an incomplete bill. Dato entries are for example only.
FIGURE 4-3. INPATIENT HOSPITAL
BILLING.
-------
ESRD TRANSPLANT TISSUE TYPING INFORMATION
END STAGE RENAL DISEASE MEDICAL INFORMATION SYSTEM
Form Approved
OMB No. 066-R-0098
1. PATIENT'S LAST NAME FIRST ' Ml 2. HEALTH INSURANCE CLAIM NUMBER
I
I
I
3. PATIENT'S COUNTY OF RESIDENCE * 4. STATE 5. DATE OF BIRTH 6. TRANSPLANT HOSPITAL PROVIDER NUMBER
~~ ~Mo~ "bay" Yr7~
7. PROVIDER NAME AND ADDRESS (CITY AND STATE)
8. NUMBER OF PREVIOUS TRANSPLANTS
(Enter zero if none)
9. DATE OF TRANSPLANT
~~Mo~ "bay" ~Yr~
11. AGE .,
1
Recipient
2
Donor
RACE
Recipient Donor —
White 01 D 07 D
Black . 02 O 08 D
Oriental 03 D 09 D
Am. Indian 04 D 10 D
Other 05 D 1 1 D
Unknown 06 D 12 D
15. CROSSMATCH RESULTS
1 D Not Performed
2 D Positive
3 D Negative
4 O Equivocal
12. SEX
Recipient Donor
Male 1 D 3D
Female 2 D 4 D
14. BLOOD TYPE
Recipient Donor
0 1 D 5 D
A 2.D 6 D
8 3 D 7 D
AB 4 D 8 D
16. MIXED LYMPHOCYTE
CULTURE
1 D Not Performed
2 D Positive
3 O Negative
4 D Equivocal
Date Performed 1
Mo. Yr.
10. DONOR RELATIONSHIP
1 D Monozygotic twin 5 D Child
2 D Dizygoiic twin 6 D Cadaver
3 D Sibling 7 D Other (Specify)
4 D Genetic Parent
HL-A TISSUE TYPING (Circle antigens detected)
17. RECIPIENT
A Series
1 A 1 2 3 9 10 11 28 29
AW 23 24 25 26 19 30 31 32 33 34 36 43
B Series
„ B 5 7 8 12 13 14 18 27
* BW 15 16 17 21 22 35 37 38 39 40 41 42
C Series
" CW 12345
18. DONOR
A Series
1 A 1 2 3 9 10 11 28 29
AW 23 24 25 26 19 30 31 32 33 34 36 43
B Series
B 5 7 8 12 13 14 18 27
BW 15 16 17 21 22 35 37 38 39 40 41 42
C Series
3
"CW 12345
19. REMARKS
NOTE: * H patient residence is not in a specific county, enter incorporated city or township.
This report is required by law (42. U.S.C. 426; 20 CFR 405. Section 2133). Individually identifiable patient information will not be disclosed except as provided
for in the Privacy Act of 1974 (5 U.S.C. ^520; 45 CFR Part 5a).
i HC FA-600-1 (9-78)
ESRD MIS COPY
Department of Health. Education, and Welfare
Health Care Financing Administration
FIGURE 4-4. ESRD TRANSPLANT TISSUE TYPING INFORMATION.
-------
ESRD DEATH NOTIFICATION
END STAGE RENAL DISEASE MEDICAL INFORMATION SYSTEM
Form Approved
OME Ne. 066-R-007E
1. PATIENT'S LAST NAME
FIRST
Ml
2. HEALTH INSURANCE CLAIM NUSY.3ER
3, PATIENTS COUNTY OF RESJDENCE*
I 4. STATE
S. DATE OF BIRTH
Mo. Day Yr.
6. DATE OF DEATH
Mo. Day Yr.
7. PROVIDER NAME AND ADDRESS (CITY AND STATE)
8. PROVIDER NUMBER
9. PLACE OF DEATH (Check one)
1 D Hospital 3 D Home
2 D Dialysis facility 4 D Other
1.1. CAUSES OF DEATH (Place number from the Lift of Causes in the spaces pr
10. WAS AN AUTOPSY PERFORMED?
1 D Yet
2 D No
Secon
dMV Ca
user
LIST OF CAUSES
01 Pericarditis
(Including car-
diac tamponade)
02 Myocardial in-
farction, acute
03 Cardiac (Other
than 01 or 02}
04 Cerebrovascular
(Including spon-
taneous subdural
hematoma)
05 Embolism, air
06 Embolism,
pulmonary
07 Gl hemorrhage
08 Vascular access
hemorrhage
09 Hemorrhage
(Other ttun O4.
07, or 081
10 Pulmonary
infection
11 Septicemia
12 Viral hepatitis
13 Jnfection (Other
than JO. It,or 12)
14 Hyperkalemia
15 Pancreatitis
16 Malignancy
~17 Withdrawal from
•dialysis
18 Suicide
19 Accidental death,
treatment related
(Other than OS)
20 Accidental death not
treatment related
21 Unknown cause
22 Other (Specify in
Remarks)
1Z IF A MALIGNANCY WAS PRESENT AT DEATH, INDICATE THE YEAR DJAGNOSED. SITE, ANDTYPE OF EACH PRIMARY
1.
2.
Yr.
Sin
Yr.
She
Type
Type
13. IF DECEASED RECEIVED A TRANSPLANT
1. Date of most recent transplant
Mo.
Day
Yr.
2. Was kidney functioning (patient off dialysis) prior to death?
1 D Yes 2 D No 3D Unknown
3. Did transplant patient resume outpatient chronic maintenance
dialysis prior to death? J
1 D YM
2 D No
14. REMARKS
SIGNATURE
DATE
*!' patient residence is not in a specific county, enter incorporated city or township.
This report is required by law (42. U-S.C. 426; 20 CFR 405. Section 2133). Individually identifiable patient information will not be disclosed except as provided
for n. the Privacy Act of 1974 (5 U.S.C. 5520; 45 CFFt Part Sal.
FORM HCFA-600-2 (9-78)
ESRD MIS COPY
FIGURE 4-5. ESRD DEATH NOTIFICATION.
Department ol Health. EOucanon. and Welfare
Hulth C«r» Finjncing Adminiilnlion
-------
4.1.2 The ESRD MIS Patient Populatiori
Only those eligible for Medicare benefits are included in the ESRD MIS.
This excludes certain groups such as veterans, Federal employees, and the
persons without sufficient earnings to be eligible for social security
benefits. These groups comprise, perhaps, 10 percent of the entire end stage
renal disease patient population in the United States. Consequently, only 90
percent of the ESRD patient population is eligible for inclusion in the ESRD
MIS.
Two groups of patients constitute the ESRD MIS patient population. Tn-
unit patients are those patients who are treated in a facility either with
hemodialysis, peritoneal dialysis, or both. In-home patients are those
treated at home. In-home patients account for 13 percent of the end stage
renal disease patient population. It is assumed that they also account for 13
percent of the ESRD MIS patient population.
It is not known whether these in-home patients are on hemodialysis,
peritoneal dialysis, or continuous ambulatory peritoneal dialysis (CAPD).
ESRD MIS data from 1973 to the present confirm only that the patient is
receiving service at home, but not the type of service received. This is
because cost information concerning supplies is listed, but information as to
type of supplies is not requested. The type of dialysis listed in the
database is the type of treatment received when home treatment begins. If the
type of dialysis treatment changes, the change may not be reported.
Continuous ambulatory peritoneal dialysis (CAPD) patients are not
identified, as such, whether in-home or in-facility. Prior to September 1979,
CAPD was not recognized as an authorized treatment modality and was not paid
for by Medicare. Therefore, whatever authorized treatment modality the
patient started on continued to be listed in the ESRD MIS data base. If a
patient started treatment on hemodialysis and then switched to CAPD, treatment
remained listed as hemodialysis. If a patient began treatment on CAPD, it
would be listed as peritoneal dialysis or hemodialysis to obtain Medicare
subsidies. Beginning in April 1981, ESRD MIS will be modified to recognize
and pick up CAPD patients.
59
-------
An additional problem with the dataset is that the patients listed may or
may not be end stage renal disease patients. Several explanations are offered
for this problem.
One explanation is that service providers initially assumed that a
Patient History and Treatment Plan should be submitted to HCFA if a patient
ever received dialysis. Because of this belief, acutely ill patients,
requiring short-term dialysis, were often required to fill out the form and
were introduced into the system data base.
Another difficulty is that the first design of the Patient History and
Treatment Form was entitled "Medicare Patient History." This title led many
patients already enrolled in Medicare, but not on dialysis, to complete and
submit this form. These forms were routinely entered into the system data
base.
Consequently, HCFA does not know what percentage of the approximately
6,000 questionable patients included in the dataset were ever actually
dialyzed, and what percentage are "legitimate" long-term dialysis patients.
These erroneous records are currently being cleared off the system database by
HCFA.
4.1.3 The ESRD Patient History and Treatment Plan
Limited medical information is included in the ESRD MIS. The Patient
History and Treatment Plan contains the following data elements:
• Patient name
• Sex
• Health insurance claim number, name, and address
• Intermediary number
• Race (entered by observation only).
All identifiers, such as patient name, health insurance number, provider
number, name and address, and the intermediary number were deleted from the
60
-------
dataset. HCFA believes the remaining patient demographic data provided are
reliable. The medical information included primary disease, methods of
diagnostic confirmation, complicating conditions which are or were present,
and surgical and transplant histories.
Completeness
The History and Treatment Plan is the only form that provides information
concerning cause of end stage renal disease. Unfortunately, these data are
not updated. As mentioned earlier, the only instance where a second history
is submitted occurs when a patient requires additional dialysis after a "suc-
cessful" transplant rejects and the patient therefore re-enters the system.
Transplants occuring between July 1973 and October 1977 were considered
successful if the patient remained off dialysis for a period of 12 months.
After the 12-month point, Medicare benefits were terminated and no additional
patient records were supplied to the system data base. Because 12 months has
been found to be too short a period for tracking transplant patients, HCFA has
modified the entitlement policy so that transplant patients are now followed
for a 36-month period.
Another problem is that transplants before 1973 are not known. If a pre-
1973 transplant patient rejects and requires a return to dialysis, the date of
first chronic maintenance dialysis entered on the Patient History and Treat-
ment Plan will most likely be listed as the date of return to treatment,
rather than the date of treatment initiation preceding transplant.
If a complicating medical condition develops while the patient is under-
going dialysis, that information is rarely captured. The only instance in
which such a complication would be routinely identified occurs if: the con-
dition occasions the admission of the patient to a hospital or a skilled
nursing facility, and the complicating condition is listed on the bill (Form
SSA-1453) as the principal or secondary diagnosis. A second instance in which
a complicating medical condition, that develops after dialysis is initiated,
might be picked up occurs when a patient dies. If a Death Notification (Form
HCFA-600-2) is submitted, it may or may not list—as a primary or secondary
cause of death—a condition which developed subsequent to completion of the
61
-------
History and Treatment Plan. This must be considered an unreliable collection
mechanism.
Additional data include:
• Date of first chronic maintenance dialysis
• Highest serum creatinine level prior to first dialysis
• Present dialysis access
• Patient treatment classification
• If the patient is not a transplant candidate, the reasons
why.
The planned dialysis modality and treatment setting are requested. If the
patient is not planned for home dialysis, an explanation is requested. The
patient's current health status (level of activity) and employment status also
are recorded.
Since it is not required to receive Medicare benefits, often the Patient
History and Treatment form is not submitted to HCFA. In 1978, the percent of
new patients not submitting forms, by network, ranged from a low of 26 percent
unreported in one network to a high of 87 percent unreported in another, with
a nationwide average of 53 percent not submitting this report.
Accuracy
Without identifying patients and reviewing individual medical records,
many of the fields on the Patient History and Treatment Plan are unreliable.
The medical information that is requested is limited and the accuracy of the
responses obtained is questionable.
If the patient has had a malignancy or if a malignancy is currently pres-
ent, the year diagnosed, site, and type of each primary malignancy is
requested. According to Dr. Joseph Fraumeni, Jr., of the National Cancer
Institute, Environmental Epidemiology Branch, these data are considered
unreliable. To be considered valid, the information must be verified by
reference to individual medical records. It is possible to ascertain, with
62
-------
some degree of confidence, the population's age, sex, and race. The fact of
\
dialysis and the type of dialysis is fairly reliable, though there are
problems with these figures due to the inability to report CAPD patients, and
the in-home patient factor.
It would be necessary to review individual medical records in order to
determine anything as specific as medications the patient is receiving,
patient smoking and alcohol history, cause of end stage renal disease for
many, and complicating conditions present for many more. It would also be
necessary to review medical records to validate presence and type of malig-
nancy with any degree of accuracy. The date of first chronic maintenance
dialysis is also a questionable data field, since pre-1973 transplants were
not reported.
4.1.4 The ESRD Outpatient Dialysis Service Information
The ESRD MIS Outpatient Dialysis Service Information is completed at the
close of each billing period and is attached to the Provider Billing for
Medical and Other Health Services (Form SSA-1483), shown in figure 4-6.
Information requested, in addition to patient and provider identifying data,
includes:
• Dialysis setting (outpatient or home) and number of sessions
• Changes in the usual dialysis setting
• Reason(s) for change
• Reason(s) for temporary dialysis settings.
The number of sessions is used in calculating the change. Medicare will pay
for 39 treatment sessions per month. If more than 39 sessions are indicated,
the Medicare intermediary will review the medical record before determining
payment.
Accuracy
The average number of sessions reported is 31; however, this field is not
considered reliable. It is felt that, while the bill is accurate (what the
intermediary bases payment on), intermediary payers make no effort to update
63
-------
PROVIDER BILLING FOR MEDICAL AND OTHER HEALTH SERVICES
MEDICAL INSURANCE BENEFITS— SOCIAL SECURITY ACT
NOi|C!L- Anyone »'ho rriKrepres-nts or Calcifies essential information requested by this form may upon conviction be subject to fine and imprisonment under Federal law.
Fo'm Approved
OMB No. 72-R0738
^^vatient's last name IFirst name IMI
^ 1 1
1 1
3. Patient's address (Street number, City, State, ZIP Code)
6. Provider name and address (City and State)
7. Provider number
8. Medical record number
2. Heattn insurance claim number
4. Date of birth 5. Sex
! ! DM
DF
9. Type of service
A, [H Inpatient C. CH Other (Specify;
B. 1 1 Outpatient
If you have other health insurance or if your State Medical Assistance Agency will pay part of your medical expenses and you want informa-
tion about this claim released to them upon their request, complete items 10 and 11.
10. Insuring organization and/or State agency name ond address
11. Policy end/or medical assistance number
12. Patient's Certification, Authorization to Release Information, and Payment Request. I certify that the information given by me in applying
for payment under Title XVIII of the Social Security Act is correct. I authorize any holder of medical or other information about me to
release to the Social Security Administration or its intermediaries or carriers any information needed for this or a related Medicare claim.
I request that payment of authorized benefits be made on my behalf.
Signature (Patient or authorized representative) (Signature by mark must be witnessed)
i—I Contained in
1—I provider's record
Date
13. Nature of illness or injury
D
Check here if illness or injury
was connected with employment
Do not use
this space
14. Surgical procedures
15". Statement of services
A. Clinic visit ( )
Covered Charges
16.
Statement
Covers
Period
First service
I
Last service
i
I
B. Emergency room (
C. Laboratory
17.
Information
A. Pints
furnished
B. Pints
replaced
C. Pints
Not Replaced
D. Charge
per pint
E. Patient
paid for
deductible
D. Radiology
18. Professional component (hospital inpatients)
A. Pathology
E. Pharmacy
B. Radiology
19. Other professional
. component
F. Blood
20. Date benefits exhausted or
HH plan terminated
I i
I I
21. Patient paid (Excluding 17E)
G. Ambulance
22. I certify that the required physician's
certification is on file.
" H. Physical therapy
23. Dote received
I I
FOR INTERMEDIARY USE ONLY
I. Other (Specify)
24. Verified Patient Liability
A. Blood deductible B. Cash deductible
C. Coinsurance
25. Payment Distribution
Provider
J. TOTAL
brks:
Patient
26. Date approved
I I
! I
1 1
FIGURE 4-6. PROVIDER BILLING FOR MEDICAL AND OTHER HEALTH SERVICES.
FORM SSA-1483
Department of Health, Education, and Welfare
Racial &»eurirv Ariminifttr»tion
-------
the supplemental information. Consequently, HCFA believes that a relatively
fixed number of sessions is reported for each patient, regardless of actual
changes in treatment activity.
Since the current dialysis prescription is specified on the outpatient
form, this figure should represent the average number of hours per dialysis
session that patients were dialyzed in their current setting. However, this
field is usually not filled in correctly and HCFA believes that the original
prescription is entered repeatedly month after month. If the prescription has
changed, it may be noted in the patient's medical record, but not changed in
the patient's billing record. There is no way to edit this field for correct-
ness; consequently, it will be eliminated from the ESRD MIS data base as of
April 1981.
An additional problem in this area involves the hours dialyzed per ses-
sion, which in practice vary greatly from session to session. If patients
have consumed alcohol or more than their usual amount of fluids, they will
need to be dialyzed longer. The only way to obtain accurate information as to
length of individual treatment sessions is to consult individual patient medi-
cal records.
Additional outpatient dialysis information collected includes the type of
dialysis. On in-unit patients, this information is probably fairly accurate.
But, as discussed earlier, if a home patient has changed dialysis methods,
this information will not be picked up. Information concerning the number of
dialysis access procedures during each monthly reporting period, the number of
units of blood given during this period, and the patient's current treatment
status is also collected. Though the Outpatient Dialysis Service Information
form is usually submitted, it is often inaccurate because it is not updated.
4.1.5 The ESRD Transplant Tissue Typing Information
The ESRD Transplant Tissue Typing Information (Form HCFA-600-1) provides
transplant data. Information concerning transplants prior to 1973 was not
collected.
65
-------
Unreported forms are. a problem in gaining this information. Tn 1978, the
percent unreported, by network, ranged from a low of 30 percent unreported in
one network to a high of 100 percent unreported in another, creating a nation-
wide average of 42 percent unreported transplants.
4.1.6 The ESRD Death Notification
The ESRD Death Notification (Form HCFA-600-2) contains identifying data,
date and place of death, and whether or not an autopsy was performed. Primary
and secondary causes of death also are collected. If a malignancy was present
at death, the year diagnosed, site, and type of each primary malignancy is
requested.
If the deceased received a transplant, information as to date of most
recent transplant, whether the kidney was functioning prior to death, and
whether the patient resumed outpatient chronic maintenance dialysis prior to
death is requested. This form was implemented only in 1977.
In 1978, the national average for unreported Death Notification was 68
percent. The percent unreported, by network, ranged from a low of 19 percent
to a high of 85 percent. While the fact of death is known, the cause of death
is known for an average of only 32 percent of the population. Some additional
causes of death could be collected, if an end stage renal disease patient died
in a hospital and the cause of death was the reason for admission. Conse-
quently, this is an unreliable collection mechanism.
Additional data reported on this form, such as malignancy information,
would have to be verified from the individual patient medical record to be
considered reliable.
4.1.7 The ESRD Center Patient Listing
The ESRD Center Patient Listing (figure 4-7) was provided without iden-
tifiers, so this additional information could only be considered as reliable
as the data listed on the other input documents, since it is a compilation of
these documents.
66
-------
I.S'II) C"MI IIH.MTIAI - HI 1)1. M LtftSU) TO AUTIIMH 1 /I I) I'l KSOIINI'L (lUI T
111 ruin nan : i/- Ji-i"»
iin: i-o . UHtll'li "i i>'H T.r
»rn>i,r::s: i SA A?
siM.r HI.HAI. nisi ASI MLHICAI. Hit iiniiAi ion stsiu
III TAI IMF IJI OF lirALHI. UJIICA1 Hill. Aril) Ul_l F Al'f.
i. INT OF f nn siAGr
(uscAsr LIVING niAiY:;ir,
I'Al.t N».
T)
I— I
O
JS
171
O
H
I—I
M
H
f
M
H
I—I
O
l ir i-l i ,..'r
i;i mr.Ai'.r
T~inr Uiniiii ii — rnni — ITKTI
/ii' UIIMH c
y CACL UIALT:;IT; — nncrmir
iti'F. OH IHI'ATHENT THAIISI rtiiiri- IK/.IIR- THAU- coimr.cr
mrri - inntr.c - nTtrr-rrnncn sirm—L-rro-wNK — BS--SHOMN-
\-rt
s>n pism
rcs——
Him) I'cul SA so Snr 11
YES
lit HO rCTfl ort »>U ill
in t'Aiir.-::
MLMO ITKI SA SO SIlT H
YER
III.I1U IT_R I on oil all T II
inno I'EKI SA sn SOT 11
x x x x x YES
i.| HA' .!•.«. i
i
i in
SA
so
—srr
Sl)
Sill II
sin ii
SflT II
Sill II
sm ii
x x x x
ji i( H M
x x x x
X X X X
x
x
x
YES
YES
YES
• ci.i.ri r li ii !'r:
I'lioi t :
nftfl.
:_J.^(ia(u(rJ,
-------
4.1.8 The ESRD Facility Cost Questionnaire
The Facility Cost Questionnaire, which during the 1970s was comoleted
annually, is the only document in the MIS that contains any type of informa-
tion concerning types of machines and reuse of coils. Since June 1980, the
Facility Cost Questionnaire has been completed semi-annually.
The questionnaire is designed to provide a limited amount of information.
Specific information relating to machine use includes:
• Number of fixed-position and portable machines
• Number of coil and parallel flow machines
• Information on whether coils are reused and, if they are,
how often (on the average).
In order to determine anything as specific as brands of machines, types of
membranes, amount of tubing, treatment of water, and sterilants used, it would
be necessary to contact individual facilities.
4.1.9 Problems with the ESRD MIS Data: A Summary
The various elements reported in the ESRD MIS have decreasing reliabili-
ties. Age, sex, and race data are reliable. Information concerning primary
cause of, disease, complicating conditions, and malignancies is not very reli-
able. To get reliability, it would be necessary to verify these data by
reviewing individual patient medical records. Fact and type of dialysis is
fairly reliable, although, as stated earlier, problems exist with these
figures.
Fact of death is known, but cause of death, when identified, would need
to be verified by reviewing patient medical records. Patient and provider
identifiers were not specified, thus making it impossible to return to indi-
vidual medical records.
Poor compliance in reporting to the ESRD MIS presents a major problem.
According to HCFA, the percentage of unreported forms increases with each
68
-------
previous year of operation. JRB was provided with 1978 figures. Conse-
quently, it is likely that a higher percentage of unreported forms occurred in
1977. Reporting rates may have improved slightly for 1979, and HCFA believes
they are continuing to improve.
The data available in the ESRD MIS, as it existed during the years 1977
through 1979, are limited and must be qualified. Poor compliancy rates and
inaccurate reporting are major factors that must be considered. In addition,
sufficient detail is not collected by the ESRD MIS. The information necessary
to conduct an epidemiologic study is not available through the ESRD MIS alone.
4.2 ASSESSMENT OF EPIDEMIOLOGIC RESEARCH POTENTIAL OF ESRD MIS DATA
In the determination of causality, the relevant competing factors to
which the selected population is exposed were identified:
• Copper and zinc from the filtration membranes, the
component action of which causes leukopenia and hypoxia.
In addition, the trace element of cadmium, a carcinogen,
is a contaminant of zinc and copper compounds.
• Sterilizing agents, such as formaldehyde, that have been
directly correlated with hemolysis and recently have been
shown to cause cancer in test animals.
• Water used in dialysate and containing fluoride, which
has been demonstrated to produce bone disease.
• Chloramine, which causes hemolysis.
• Nitrate-nitrite, which also is linked with hemolysis and
other synergistic effects.
• Heparin, which is linked to the development of ostopenia.
4.2.1 Specificity
The Patient History and Treatment Plan provides medical information on
the first day of treatment, primary cause of ESRD, complicating conditions,
type of dialysis, method of diagnosis, dialysis access, surgical history, and
history of malignancy, in addition to relevant demographic information. For
the most part, information of this type would not be specific enough to draw
69
-------
conclusions as to the extent of the condition, the severity of the condition,
and the impact of the condition on the constitutional factors of the patient.
Medical information available from the ESRD Outpatient Service Informa-
tion form includes type of dialysis, setting for dialysis treatment, reasons
for selection of dialysis setting used, and current dialysis prescription.
Again, the specificity of the information is limited to checking the type of
modality, which for the home-dialysis population is not fully known, and the
prescription, which can provide only approximate information as to how long
and how many treatments a patient received. However, the form does not
include the type of medications used, the additions of concentrate to the
dialysate, or information relating to any changes in the composition of the
dialysate, itself.
The reason for change of setting does specify "medically unfit" as a
factor for selection of dialysis setting, but does not provide information on
the type of medical complication, the severity of the condition, whether the
condition is acute or chronic, or if it is associated with end stage renal
disease or an accompanying concomitant disease.
Medical information from the ESRD Death Notification provides primary and
secondary causes of death, if an autopsy .was performed, transplant informa-
tion, and information on malignancy. Information recorded provides the "fact
of" these variables, but does not provide information specific enough to
describe the factors which precipitated the events listed or the outcomes of
these events.
4.2.2 Accuracy
The Health Care Financing Administration has supplied information
pertaining to the accuracy of the data within the ESRD MIS. It is HCFA's
belief that information pertaining to the demographic characteristics (age,
sex, race), date of initial chronic maintenance dialysis treatment (with the
exception of pre-1973 transplants), fact of dialysis, fact of death, fact of
70
-------
transplant, and type of dialysis is accurate. Information which they feel is
greatly lacking in accuracy include:
• Cause of death
l
• Prescription data (number of-hours of dialysis treat-
ment), a field which is being removed from the system in
April 1981
•• Type of dialysis modality used by the home-treatment
population
• Data pertaining to malignancy (both patient treatment and
history and death notification).
Information pertaining to primary cause of end stage renal disease and
complications must be verified by consulting the individual patient's medical
^
records. Information pertaining to dose, needed to perform an analysis assoc-
iating dose response to outcome, is unavailable, since prescription informa-
tion is considered to be invalid. Since cause of death and malignancy data
are considered to be unreliable without verification from medical records,
there can be no basis for information pertaining to outcome. The ESRD MIS
database records the prevalent conditions stated above; however, the ESRD MIS
is not designed to collect information relating to incidence of disease during
the treatment period, unless the condition requires hospitalization and in
this case only billable information is captured. Therefore, adjustment from
competing causes of death not specified in the Patient History and Treatment
Plan cannot be accomplished.
4.2.3 Completeness
A major factor in determining the generalizability of research findings
is the representativeness of the population data being analyzed. The ESRD MIS
has had great difficulty in aquiring the compliance of the provider popula-
tion. In 1978, the ESRD MIS received national averages of only 47 percent of
initial Patient History and Treatment Plans, and only 32 percent of the Death
Notification forms. Network compliance ranged from a high of 74 percent to a
low of 13 percent for the Patient History and Treatment Plan form and 81 per-
cent to 15 percent for the Death Notification form. In light of this, it will
71
-------
be extremely difficult to describe the study population in terms of represen-
tativeness.
4.2.4 Statistical Methodology
Many questions arise concerning the analysis of ESKD MIS data over and
above the quality of the data, itself.
The salient characteristic of this dataset is that the patients are under
different lengths of observation time, the duration dependent upon factors
other than the event of interest, i.e., the development of a malignancy. This
requires that the response variable not be a dichotomous either/or variable on
the individual patient, hence not a prevalence rate for the treatment group.
The response variable here is time from .first treatment to development of
malignancy, and the methodology of failure time (survival) is applicable.
Since so many covariates must be adjusted for, it is clear that a mul-
tiple regression model is required. An assumption must be made about the
underlying distribution of times to the event, and the justification for this
assumption will influence the nature of the analysis and the outcome of the
study: i.e., if exponential, can the hazard function be assumed to be
constant over time? If a distribution-free method is used, such as proposed
by Mantel (1966) or Cox (1972), then is the assumption of a constant relative
risk over time justified? It would appear not, since the longer the exposure
to hemodialysis, the greater the risk versus those on peritoneal dialysis.
Although the sample size is large, the number of covariates is poten-
tially exceptionally large, and there are dangers associated with the esti-
mation of many parameters by the maximum likelihood method used for these
models. These techniques can produce inconsistent estimates which require
careful investigation to assure that they are reasonable. Additional terms,
the interactions, would also require investigation.
Some of the covariates could be time-dependent, and would require special
treatment.
72
-------
A basis would need to be established for assuming model effects were
either additive or multiplicative.
How will the treatment groups be represented—each separately, hence by
two dummy variables, or hemodialysis plus combined versus peritoneal dialysis,
and why? What about the heterogeniety in the proportions of hemodialysis and
peritoneal dialysis received by the individuals in the combined group?
Given the very large difference between the hemodialysis and the peri-
toneal dialysis groups in sample size, how will the contrast in the precision
of their measurements be adjusted for?
Will the latency periods for the development of different types of malig-
nancies be covered by the relatively short observation time on each patient?
Since the competitive risks of death from other diseases are very great
in this study, how will they be adjusted for in the response measure of time
to developed, malignancy?
What will be done with observations containing missing data? Most of the
applicable techniques of statistical analysis have no procedures for dealing
with missing data. Will the final, cleaned-up dataset be truly representative
of the population of interest, and equally representative of each treatment
group individually?
4.2.5 Summary Conclusion
After considering the limitations of the dataset and carefully assessing
the likelihood of fruitful results from future investigations along these
lines of research, it was found that the ESRD MIS is not suitable for the type
of analyses required to determine the additional risk to ESRD patients. This
decision is based upon the following reasons:
• The quality of the data to be used, according to both the HCFA
Branch Chief of ESRD MIS and Drs. Fraumeni and Hoover of NCI,
lacks the accuracy, completeness, and consistency necessary to
perform any meaningful clinical analysis.
73
-------
• This type of analysis has the danger connected with "fishing":
if enough testing is done, it is highly probable that some
variables might correlate. However the validity of such
findings will suffer from the inaccuracies of the data used to
determine them.
• The understandable reluctance of the Health Care Financing
Administration to provide ESRD MIS patient identifier data,
which would be required to perform a future comprehensive
epidemiologic investigation of the ESRD patient population.
As noted elsewhere in this volume, it is perhaps feasible to use end
stage renal disease patients as an epidemiologic study population, if appro-
priate attention were to be devoted to addressing rigorously the complex of
biomedical conditions and confounding factors represented by the pathological
processes., treatment regimens, dialysis modalities, and multiple exposures
involved. Useful information on some of the competing disease etiologies
might result from such a study. But at the present time and with the knowl-
edge and detection capabilities currently available, mortality or morbidity
incidence attributable to DEHP exposure would not be a fruitful subiect for
such an investigation.
Furthermore, the End Stage Renal Disease Medical Information System data
base does not offer an adequate route of approach to epidemiologic research on
the patient population. HFCA program officials recognize and concur in this
judgment. The dataset provided offers the potential for demographic
analysis, but very little beyond demographics.
If epidemiologic evidence of the health consequences of chronic exposure
to DEHP is sought, it must be sought in a population less confounded than
systemically ill end stage renal disease patients.
74
-------
CHAPTER 5
AN EXAMINATION OF AN EPIDEMIOLOGICAL
STUDY WITH ESKD PATIENTS
D.. Wellington, Ph.D., and. I. Marks, M.P.H.
A great many factors are involved in the development and
treatment of end stage renal desease. An investigation of
causality in populations with renal failure must include an
examination of variables involved in the pathological processes.
Close examination of these factors requires both an understanding
of the disease and treatment, and the limitations inherent in
uncontrolled field investigations. Evaluation of the factors
associated with the ESRD population concludes that ESRD patients
may not constitute a viable study group for an epidemiological
investigation of the carcinogenic effect of exposure to DEHP for
two reasons: 1) the inability to measure exposure to DEHP
separately from exposure to carcinogens and possible etiologic
factors; and 2) the extremely high combined competitive risk of
mortality or morbidity from causes other than primary liver
cancer and other malignancies.
The limitations of an epidemiologic study of ESRD patients include the
multiple disease processes leading to chronic renal failure and the various
treatments of renal disease which may, in themselves, lead to the development
of primary liver cancer and other malignancies. This chapter discusses these
causal factors, how they affect the design of an epidemiologic study of ESRD
patients, and the problems associated with that design from an epidemiologic
standpoint.
Renal failure leading to dialysis is not a single-entity disease. Renal
failure is a trauma to the renal system resulting in irreversible damage. The
therapeutic approach to the replacement of homeostatic functions usually
provided by functioning kidneys is dialysis. Dialysis, without regard to the
specific modality, presents multiple toxic factors, in addition to the
disease, for which the body must compensate. It is feasible to envision the
total force of ESRD mortality as simply the sum of interdependent competing
factors.
75
-------
Several aspects of the origin and natural history of chronic renal
failure, or end stage renal disease, complicate an investigation of the
determinants of another disease, such as cancer, in an ESRD patient
population. These aspects include the multifactorial nature of etiology. End
stage renal disease can occur from a pathologic process which affects the
function of the kidneys, or as a secondary condition, resulting from such
diseases as diabetic nephropathy or lupus erythematosus which may contribute
to chronic renal failure. Each of these conditions has a particular affect on
a number of body systems. The interaction of multiple conditions can be
either additive or multiplicative, exacerbating other conditions or involving
other systems which previously were functioning in a relatively normal manner.
The multiplicity of these factors and their effects upon the development of
disease vary between patients, making it increasingly difficult to assemble a
population which meets a standard selection criteria.
Another aspect is the long latent period. End stage renal disease is a
long and degenerative condition. Its natural history includes a great many
pathologies which contribute to the progression of the disease and its com-
plications. In such cases., it is difficult to link antecedent events to
outcome.
The final aspect is the differential effect of factors on the incidence
and course of the disease. Factors may act differently at various stages of
the disease. For example, patients who suffer from end stage renal disease,
and later develop diabetes as a result of aging and other factors not directly
related to the primary renal failure, differ from patients who develop chronic
renal failure as a result of a hypertensive condition caused by diabetes.
Therefore, it is important when conducting a epidemiological study to
differentiate between different stages and categories of disease to adjust for
competing causes of morbidity and mortality.
5.1 The Problem of Causality in the ESRD Patient Population
The determination of causal associations can be obscured by a number of
confounding variables. Therefore, a number of observational studies must be
performed to establish proof of a causal relationship in any association.
76
-------
Interpretation of such an association must be conducted in a systematic
manner. The Advisory Committee to the Surgeon General, U.S. Public Health
Service, defined five criteria that should be fulfilled to establish a causal
relationship. These five criteria have been generally adopted as a test of
causation:
•' Consistency of association
• Strength of association
• Specificity of association
• Temporal relationship of association
• Coherence of association (Mausner et al. 1974).
The potential for obtaining each of these five criteria in the ESRD patient
population is addressed in the following paragraphs.
Consistency of Association requires that an association determined in one
study persist in testing under other circumstances, with other study popula-
tions, and in some cases, different countries. The more often the particular
association appears under diverse circumstances, the more likely it is causal
in nature. One should be aware, however, that the same bias (i.e., systematic
error) could exist in multiple studies and produce an apparent artifactual
consistency (Mausner et al. 1974).
, The use of renal dialysis patients as a study population entails an
increased danger of having a systematic error producing a spurious associa-
tion. ESRD patients suffer from a number of systemic pathologies and are
exposed to a great number of agents in the course of their treatment.
Therefore, when examining causal associations in this group, it is necessary
to be sure the cause is not attributable to the effect of some unadjusted
factor.
The bias described here is called indirect association. Indirect asso-
ciation in a renal dialysis population could be the effect of a factor or
factors, other than the one identified, which cause the determined effect.
For example, in attempting to identify DEHP as a cause of cancer in ESRD
patients, the investigator must rigorously adjust for a multitude of patho-
logic and iatrogenic variables other than this plasticizer before the causal
link can be substantiated.
77
-------
Strength of Association refers to the ratio of disease rates for those .
with and without the hypothesized factor. The likelihood of a causal relation
is strengthened if a dose-response effect (gradient) can be demonstrated.
Thus, with increasing levels of exposure to the factor, a corresponding rise
in the occurrence of the disease is found (Mausner et al. 1974).
Renal dialysis patients utilize three different treatment modalities,
each of which exposes the individual to a complex of agents. In addition to
dialysis, the ESRD population receives a high number of blood infusions and
medications. The main source of contamination from DEHP, other than dialysis,
comes from blood stored in PVC blood bags. Therefore, all three dialysis
patient groups—hemo, peritoneal, and continuous ambulatory peritoneal—are
exposed to the agent DEHP.
As a patient's condition declines, a greater number of dialysis
treatments may be needed and a greater number of transfusions are likely to be
required. This factor makes it difficult to determine if the development of a
condition such as cancer is more evident in more severe ESRD patients, or is
caused by exposure to the plasticizer.
Specificity of Association measures the degree to which one particular
exposure produces one specific disease (Mausner et al. 1974). ESRD patients
are exposed to a great number of potentially toxic agents, each of which has a
determined singular potential effect as well as many undetermined effects.
In addition, ESRD patients exhibit a large range of pathologies which
vary with the nature of the underlying disease entity. Due to the multiple
factors associated with the ESRD patients, it is difficult—if not
impossible—to determine which of these factors is the sole preceding agent
responsible for the manifestation of an outcome such as cancer.
Temporal Relationship of Association requires that exposure to the
putative factor must antedate the onset of disease, and must allow for any
necessary period of induction and latency (Mausner et al. 1974). The disease
process which brought about chronic renal failure will antedate the exposure
78
-------
to any treatment agent to which the individual is exposed, and may itself
influence the onset of cancer.
Without a successful transplantation intervention, the end stage renal
disease patient has a seven- to ten-year maximum life expectancy after the
onset of chronic renal failure. This period is too short a timeframe to
expose and induce the development of a chronic disease such as cancer.
Consequently, the patient may not live long enough to deve.lop the specific
disease.
Coherence of Association infers biological plausibility, which may have
been established in animal models (Mausner et al. 1974).
There are a great many difficulties associated with the extrapolation of
animal data to human populations, in general. This situation is further
complicated by consideration of a severely ill study population, such as
patients suffering from'chronic renal failure. The metabolism of substances
such as DEHP in ESRD patients differs from that in healthy individuals. In
healthy individuals, most of the DEHP is excreted within 24 hours. Renal
dialysis patients cannot excrete, and store this substance in their tissues.
Studies to date have not used animals with chronic renal failure over a
large enough time frame to make them even somewhat comparable to ESRD
patients.
5.2 Analytic Approaches
Since chronic diseases develop over a prolonged time interval, the etio-
logic study of their conditions requires an analysis of the events which occur
during this time. For ESRD patients, such analysis must include the problems
of confounding factors associated with the various treatment modalities for
end stage renal disease. The following paragraphs examine these problems from
a statistical standpoint. Each factor and its potentiality as a cause of
cancer is presented through a discussion of each issue, its effect and the
problems associated with correcting for it.
79
-------
5.2.1 Multiple Causal Paths
Figure 5-1 presents a flow chart of potential development of cancer in an
ESRD patient in which the broken lines represent paths which may lead to a
particular treatment and the solid lines indicate probable physical results.
The possible interrelationships between renal disease and associated treatment
therapies, and the likelihood of additional diseases, are presented in a
simplified manner, but are sufficient to illustrate the critical confounding
of factors which could be responsible for primary liver cancer or other types
of malignancy.
It has been suggested that the sensitivity of ESRD patients to primary
liver cancer or other cancer would be advantageous in determining efficiently,
with respect to time and sample size required, the development of these
diseases in a study population. However, the case is not simply that these
patients are more sensitive to malignancy in the way that mice are bred to be,
but that they are subject to several causative factors which can lead to
cancer by interconnected etiological paths. Discrimination among these
factors would be almost impossible, particularly in a retrospective study. As
shown in figure 5-1, the renal disease process itself, through attendant
uremia and protein-calorie malnutrition, leads to immunosuppression and hence
to increased cancer risk. In addition, the treatment therapies include
transplantation, which involves immunosuppressive therapy both before surgery
and after (frequently for an indefinite period). Thus, multiple routes
emanate from both the disease process and the treatment to immunosuppression,
which in turn increases the risk of malignancies of all types.
Treatment combinations of nephrectomy, transplantation, and blood
transfusions increase the incidence of hepatitis B, which is a risk factor in
hepatic cancer. At the same time, the treatment process of hemodialysis
itself also increases the risk of hepatitis B and, hence, of hepatic cancer.
5.2.2 Indiscriminate Exposure Measure
An even more difficult, in fact prohibitive, problem involved in differ-
entiating among the causative factors is the acute exposure during dialysis
80
-------
PRE-ON9W OF
RENAL DISEASE
ST-ONSET OF
RENAL DISEASE
Kidney Cancer
Other Predisposing
Diseases
Other Cancer •*'
Immunosuppressive Therapy
Renal
Disease
Transplantation
A
/
Immunosuppressive
Therapy
Memo-..
Peritoneal
CAPO
-^•Cancer
Exposure to DEHP
Exposure to
Formaldehyde
Exposure to
Testosterone
Derivatives
Liver Cancer
Other Cancer
(Carrier May Be Unaware)
Liver Cancer
Immunosuppression
Cancer
Protein-Calorie
Malnutrition
• May result in this physical effect
Simply pre-dates
• May lead to this treatment
Figure 5-1. POSSIBLE INTERRELATIONS BETWEEN CANCER (LIVER AND OTHER), RENAL DISEASE AND ITS TREATMENT
-------
simultaneously to several known or potential carcinogenic agents: to DEHP
through the PVC-tubing; to formaldehyde through sterilizing agents; to
zinc/cadmium contaminants through membranes; to testosterone derivatives which
are specific for hepatic cancer; to direct entry into the blood stream of
fluoride, chloramine, nitrate-nitrite, aluminum, and other trace metals in the
water (Kjellstrand 1978); and to complement activation from membranes which
produces the immunosuppressant state of leukopenia (Shin 1978). Moreover,
other confounding exposures undoubtedly occur in the dialysis process with
effects which are as yet undiscovered.
Thus, the measure proposed to determine the exposure to DEHP of ESRD
patients—duration on dialysis—is simultaneously the surrogate measure for
several other possible carcinogenic factors, as shown in table 5-1. Even the
specific cancer of interest, primary liver cancer, can result from at least
three factors: 1) DEHP and testosterone derivatives, whose individual measures
are identically determined by duration of any type of dialysis; 2)hepatitis B,
whose incidence is closely related to the duration of hemodialysis by the
process itself; and 3) the duration itself of any type of dialysis because of
its positive correlation with the number of blood transfusions and, hence,
with the risk of hepatitis.
It is not possible, as had been proposed, to use patients on peritoneal
dialysis as a control group vis-a-vis those on hemodialysis in order to
represent differentiation of degree of exposure to DEHP, since there is
actually very little exposure difference between them. Peritoneal dialysis
involves exposures of 3 to 5 mg, while hemodialysis involves exposures of 4 to
8 mg. (Harris 1981). Thus, no method presents itself by which an end result
such as hepatic or other cancer could be traced specifically to exposure to
DEHP in ESRD patients.
5.2.3 Dominating Competitive Risks
Although several factors related to the disease process and to its
treatment could lead to an increased risk of cancer in ESRD patients, their
extremely high risk of death from vascular diseases and from infectious
82
-------
TABLE 5-1. FACTORS FOR WHICH DIALYSIS DURATION IS AN IDENTICAL, SURROGATE
MEASURE
The duration on dialysis is an identical surrogate measure of the
exposure to the following potential etiologic factors in the development of
cancer:
1. Exposure related to the dialysis treatment process itself:
a. DEHP
b. Fo rma1d ehyd e
c. Testosterone derivatives
d. Zinc/cadmium contaminates
2. Exposure directly reflecting the passage of time:
a. Factors due to the duration and progression of renal disease:
• Immunosuppressive effects of uremia
• Immunosuppressive effects of protein-calorie malnutrition
• Increased risk of hepatitis due to additional blood transfusions
b. Factors due to the duration of post-transplantation therapy:
• Immunosuppression
c. Factor due to aging:
• Increased risk of malignancy.
83
-------
diseases dominates their mortality model, and would reduce drastically the
probability of observing cancer development in them.
In a study of causes of mortality in a single year among patients on
dialysis (Moorhead 1973), 50.5 percent of the deaths were due to vascular
disease, 19.4 percent to infectious disease, and only 2.0 percent to malignant
disease. Clearly, intervention by deaths from more immediate risks would
severely censor the observation of cancer deaths, particularly of relatively
infrequent liver cancer deaths, in this patient population. Moreover,
statistical adjustment for competitive risks so disproportionately greater
than the mortality risk of interest could obscure the estimate of the latter's
probability of occurrence.
5.2.4 Size of Study Population
As with any study using a selected population, consideration must be
given to size of the group to be observed in order to determine the study
method to be used. Two of the more common study methods in this area are
cohort and case-control.
Cohort Study
The 1970 annual incidence of neoplasia for the 45 to 70 year age group in
the U.S. was approximately 623 per 100,000. If the asbestos model is the
appropriate one, then a five-fold increase in incidence is the largest that
would be expected in the exposed population. The sample size required to
detect a significant increase (one-tailed test) over the control group with a
type I error of .05 and power of .90 (Snedecor and Cochran 1967) would be
approximately 500 person-years. However, if just a doubling of the incidence
could be expected, then a sample of 4,100 would be needed; and if only a 50
percent increase in cancer occurred, it would require a sample of 13,650 to
detect the difference.
The incidence of primary liver cancer in the U.S. population is less than
three per 100,000, and therefore much larger sample sizes are needed to detect
significant differences when the actual elevation of risk is of the orders
84
-------
cited above. For example, if a five-fold increase in risk existed in the
study population, 107,500 person-years would be required to detect it with the
significance level and power stated above. If the increased risk of primary
liver cancer were less than five-fold, for example two-fold, the required
sample size would be 860,000 person-years.
Moreover, these sample sizes are only for a simple one-tailed test of
difference between proportions or incidence without the classification which
would be necessary for factor-adjustment, which would increase the sample size
required.
Case-Control Study
Case-control studies with measured exposure are more statistically
sensitive (powerful) than cohort analysis and thus do not require sample sizes
as large for the same detection level. However, in these data, differences in
DEHP exposure between cases and controls would be indeterminate, since the
only measure available—time on dialysis—measures the potential of multiple
factors indiscriminately.
It is possible that an epidemiologic study protocol could be developed
that would address successfully some segment of the end stage renal disease
patient population, relying on rigorous segregation of comparable elements of
the study population and adequate attention to the multitude of confounding
factors. However, it is unlikely, given the chronic illness of the patient
population and the limited prognosis for survival, that such an epidemiologic
study would succeed in determining either association or causality of so
subtle a contaminant as DEHP relative to cancer mortality in the population.
85
-------
REFERENCES
Adler AJ, Lundin AP, Friedman EA, Berlyne GM. 1979. Effect of hemodialysis
on plasma beta thromboglobulin levels. American Society Artificial Internal
Organs. 25:347-50.
Albro PW, Thomas R, Lishbein L. 1973. Metabolism of diethylhexyl phthalate
by rats. Isolation and characterization of the urinary metabolites. J.
Chromatography. 76:321-30.
Alfrey AC. 1976. The dialysis encephalopathy syndrome. The New England
Journal of Medicine. 294:184-88.
Anonymous. 1980. Single-use devices should not be used. Hospitals.
54:157-60.
Arthur D. Little, Inc. 1979. Risk assessments of priority pollutants:
phthalate esters. Draft report submitted to the US Environmental Protection
Agency, Monitoring and Data Support Mission.
Autian J. 1973. Toxicity and health threats of phthalate esters: review of
the literature. Environmental Health Perspect. 4:3-26.
Bell FP, Patt CS, Brundage B, Gillies PJ, Phillips WA. 1978. Studies on
lipid biosynthesis and cholesterol content of liver and serum lipoproteins in
rats fed various phthalate eaters. Lipids. 13:66-74.
Bell FP, Nazir FJ. 1976. Effect of dietary di-2-ethylhexyl phthalate
on lipid biosynthesis in selected tissues from the rat, in vitro. Lipids.
11:216-21.
Birkeland SA, M.D. 1976. Uremia as a state immune deficiency. Scandinavian
Journal of Immunology. 5:107-15.
Carpenter CP, Weil CS, Smyth HF, Jr. 1953. Chronic oral toxicity of
di (2-ethylhexyl) phthalate for rats, guinea pigs and dogs. Arch. Indust.
Hyg. 8:219.
Chen WS, Kerkay J, Pearson KH, Paganini EP, Nakamoto S. 1978. Tissue
analysis of plasticizer in dogs. Proc. Dialysis Transplant Forum.
Daniel JW, Bratt H. 1974. The absorption, metabolism and tissue distribution
of di (2-ethylhexyl) phthalate in rats. Toxicology. 2:51-65.
D'Elia JA, M.D., Kalany A, M.D., Malanick C, R.N., Christlieb RA, M.D.,
Weinrough LA. 1974. Diabetic nephropathy. Genitourinary Diseases :47-53.
Fassbinder W, Frei U, Koch KM. 1979. Haemolysis due to formaldehyde-induced
anti-N-like antibodies in haemodialysis patients. Klin Wochenshr. 57:673-79.
-------
Fisher JW. 1980. Mechanism of anemia of chronic renal failure. Nephron.
25:106-11.
Fraedrick G, Scholz R, Mulch J, Leber HW, Hehrlein FW. 1980. Continuous
peritoneal dialysis (CPD)-an alternative to hemodialysis in acute renal
failure after cardiovascular operations. Cardiovascular Surgeon. 7.8:246-48.
Gandhi VC et al. 1979. Aseptia peritonitis in patients on maintenance
peritoneal dialysis. Nephron. 24:252-59.
Geertz W, Dyer K, Johnson D, Miller B, Taylor B, Sharp W. 1974. Polyvinyl
chloride biomedical products: toxicity testing and identification of
additives. Transplant American Society Artificial Internal Organs. 20:42-46.
Gelfand MC, Rodelas R, McAnally JF, Novello A, Winchester JF. 1978. Heparin
associated hemodialysis hyperthroidism (HHH): a physiologically significant
phenomenon. Proc. Dialysis Transplant Forum: 255-58.
Gibson TP, Briggs WA, Boone BJ. 1978. Delivery of di-2-ethylhexyl phthalate
to patients during hemodialysis. J. Lab. Clin. Med. 87(3):519-524.
Graham PR. 1973. Phthalate ester plasticizers—why and how they are used.
Environmental Health Perspect.
Guyton AC. 1956. Textbook of medical physiology. 3rd edition. W.B.
Saunders Co. Philia., Pa.: 534-39.
Hampers CL, Schupak E, Lowrie E, Lazarus J. 1973. Long-term hemodialysis.
2nd edition. Grune & Stratton Inc. New York, N.Y.: 20-25.
Harris, GW, Baxter, Travenol Laboratories, 1981. Di (2-ethylhexyl) phthalate
(DEHP) in medical devices. A talk before the National Toxicology Program
Conference on Phthalates, Washington, D.C.
Harris RS, Hodge HC, Maynard EA, Blanchet HJ, Jr. 1956. Chronic oral
toxicity of 2-ethylhexyl phthalate in rats and dogs. AMA Arch. Ind. Health.
13:259-64.
Hillman LS, Goodwin S, Sherman W. 1975. Identification and measurement of
plaasticizer in neonatal tissues after umbilical catheters and blood products.
New England J. Med. 292(8):381-6.
Hoy WE, Castero RVM. 1979. Eosinophilia in maintenance hemodialysis
patients. Journal of Dialysis. 3:73-87.
Jacobson MS, Kevy SV, Grand RF. 1977. Effects of a plasticizer leached from
polyvinyl chloride on the subhuman primate: a consequence of chronic
transfusion therapy. J. Lab. Clin. Med. 89:1075-79.
Jaeger JR, Rubin RJ. 1972. Migration of a phthalate ester plasticizer from
polyvinyl chloride blood bags into stored human blood and its localization in
human tissues. New England J. Med. 287:1114-18.
-------
Jerque MA. 1973. PCBs, phthalates, and other organic compounds in
groundwater. University of Oklahoma. (Ph.D. dissertation). In Versar, Inc.
1979. Non-aquatic fate: phthalate esters. Washington D.C.: US Environmental
Protection Agency (draft report).
Kotzias D, Kelin W, Korte F. 1975. Ecological Chemistry: CVI: occurrence
of xenobiotics in ground water of waste deposits. Chemosphere. 4(5):301-6.
Kjellstrand CM, Chairman. 1978. Toxicity of materials and medications used
in dialysis. Panel Conferance in Transplant American Society Artificial
Internal Organs. 24r764-69.
Lewis L, Fletchner T, Kerkay J, Pearson KH, Nakamoto S. 1978. Bis
(2-ethylhexyl) phthalate concentrations in the serum of hemodialysis patients.
Clin. Chem. 24:741-46.
Lilienfeld AM, Lilienfeld DE. 1980. Foundations of epidemiology. 2nd
edition. Oxford University Press: 194.
Massry SG, Glasscock RJ. Textbook in nephrology. Uremia by G.E. Schreiner.
Elsevier North-Holland, Inc. New York (in press).
Matas AJ, Simmons RL, Kjellstrand CM, Buselmeier TJ, Najarian JS. 1975.
Increased incidence of malignancy during chronic renal failure. The Lancet:
883-85.
Mausner JS, Bahn AK. 1974. Epidemiology, an introductory test. W.B.
Saunders and Co. Philadelphia-Toronto-London: 101-02.
Mayor GH, Hourani MR, Greenbaura DS, Pallerson MJ. 1979. Prevalence of
hepatitis B in 27 Michigan hemodialysis centers. AJPH. 69:581-83.
Merck, Sharp, Dolme. 1977. Merck manual of diagnsois and therapy. 13th
edition. Research Laboratories. Rahway, N.J.
Mes J, Coffin DE, Campbell DS. 1974. Di-n-butyl and di-2-ethylhexyl
phthalate in human adipose tissue. Bull. Environ. Contain. Toxicol.
12:721-25.
Mes J, Campbell DS, 1976. Extraction efficiency of polychlorinated biphenyl,
organchlorine pesticides and phthalate esters from hyuman adipose tissue.
Bull. Environ. Contain. Toxicol. 16(1):53-60.
Miripol JE, Garvin PF, Stern IJ, Wallin RF. 1976. Toxicity of components of
plastic having contact with blood. NTIS.
Mishkel M, Nazir D, Shimiza A. 1979. Di-2-ethylhexyl phthalate (DEHP)
accumulation in dialyzed uremic patients. Kidney Intern. 15(6):712-13.
Montie JE. 1979. The significance of cancer after renal transplantation.
The Journal of Urology. 122:298-99.
Moorhead JF. 1973. Dialysis Transplantation Nephrology. Pitman Medical:
xxxiii-v.
-------
Morgan AP. 1973. Access to the circulation. Long-term hemodialysis. 2nd
edition. Grune & Stratton, Inc: 40-63.
Nikonorow M, Mazur H, Piekacz H. 1973. Effect of orally administered
plasticizers and polyvinyl chloride stabilizers in the rat. Toxicol. Appl.
Pharmacol. 26:253-59.
Ogden DA. 1980. New dialyzer syndrome. The New England Journal of Medicine.
302:1262-63.
Ono K, Tatsukawa R, Wakimoto T. 1975. Migration of plasticizer from
hemodialysis blood tubing. Preliminary report. J. Am. Med. Assoc.
234:948-49.
Orringer EP, Mattern WD. 1976. Formaldehyde-induced hemolysis during chronic
hemodialysis. New England Journal of Medicine: 1416-20.
Overturf ML, Druhet RE, Liehr JG, Kirkendall WM, Caprioli RM. 1979.
Phthalate esters in normal and pathological human kidneys. Bull. Environ.
Contam. Toxicol. 22:536-42.
Patterson R et al. 1980. Antibodies in heraodialysis patients against
hapten-protein and hapten-erythrocytes. Journ. Lab. Clin. Med. 96:347-55.
Penn I, Starzl TE. 1972. Malignant tumors arising de novo in
immunosuppressed organ transplant patients. Transplantation. 14:407-17.
Poole CF, Wibberly DG. 1977. Determination of bis (2-ethylhexyl) phthalate
in human placenta. J. Chromatography. 132(3):511-18.
Rubin RJ, Schiffer CA. 1976. Fate in humans of the plasticizer, di-2-
ethyhexyl phthalate arising from transfusion of platelets stored in vinyl
plastic bags. Transfusion. 16(4):330-35.
Rubin RJ. Jaeger JR. 1973. Pharmacologic and toxicologic effects of bis
(2-ethylhexyl) phthalate (DEPH) and other plasticizers. Environ. Health
Perspect. 3:53-59.
Rubin RJ. 1975. Toxicologic evaluation of the migration of a plasticizer
di (2-ethylhexyl) phthalate (DEHP) from vinyl plastics. NTIS.
Sandier SG, Sharon R, Bush M, Stroup M., Sabo B. 1979. Formaldehyde-related
antibodies in hemodialysis patients. Transfusion. 19:682-87.
Schultz CO, Rubin RJ, Hutchins GH. 1975. Acute lung toxicity and sudden
death in rats following intravenous administration of the plasticizer di
(2-ethylhexyl) phthalate solubilized with tween surfactants. Toxicol. Appl.
Pharmacol. 33:514-25.
Seth PK, Srivastava SP, Agarwal DK, Mushtaq M. 1977. Interaction of di
(2-ethylhexyl) phthalate (DEHP) with pentabarbitone and methaqualone. Bull.
Environ. Contamin. and Toxicol. 17:727-32.
-------
Shaffer CB, Carpenter CP, Smyth HF, Jr. 1945. Acute and subacute toxicity of
di (2-ethylhexyl) phthalate with note upon its metabolism. J. Ind. Hyg. and
toxicol. 27:130-35.
Shin J. 1978. A study of hemodialysis luekopenia using various dialyzers.
Journal of Dialysis. 4(l):51-62.
Snedecor GW, Cochran WG. 1967. Statistical Methods. Iowa State University
Press: 221-22.
Solomons TWG. 1980. Organic Chemistry. 2nd edition. New York: John Wiley
and Sons.
Stein MS, Caasi PI, Nari PP. 1974. Influence of dietary fat and di-2-
ethylhexyl phthalate on tissue lipids in rats. J. Nutrition. 104:187-91.
Stern IJ, Miripol JE, Izzo RS, Lueck JD. 1977. Physicochemical aspects of
the extraction in blood and the disposition in rats of di (2-ethylhexyl)
phthalate plasticizer. Toxicol. Appl. Pharamaco. 41:507-22.
Sutherland GA, Glass J, Gabriel R. 1977. Increased incidence of malignancy
in chronic renal failure. Nephron. 18:182-84.
Thomas JA, Darby TD, Wallin RF, Garvin PJ, Martis J. 1978. A review of the
biological effects of di (2-ethylhexyl) phthalate. Toxicol. Appl. Pharmaco.
45:1-27.
Tomita I, Nakamura Y, Yasi Y. 1977. Phthalic acid esters in various foods
and biological materials. Ecotoxicol. Environ. Saf. l(2):275-87.
Willis RA, Willis AT. 1972. Principles of pathology and bacteriology.
Appleton-Century-Crofts. London, Butterworth: 289-310.
Willingmyre GT. 1979. Communications. Artificial Organs. 3:281.
------- |