Professional Documents
Culture Documents
Quality Assurance in Dialysis
Quality Assurance in Dialysis
DEVELOPMENTS IN NEPHROLOGY
Volume 36
The titles published in this series are listed at the end of this volume.
Quality Assurance
in Dialysis
Edited by
and
RICHARD S. THUMA
Vice President Quality Assurance,
Baxter Healthcare Corporation, Renal Division
Preface Vll
List of Contributors Xl
v
vi
This is a time in history when the concept of Quality is reaching new highs
in terms of public awareness. Articles describing quality, CQI, quality tools,
critical success factors, failures, and lessons learned appear in local news-
papers, trade journals, scientific periodicals, and professional publications
on a daily basis, yet implementation of a quality system in many hospital
units is approached with caution and the basic tenants of quality systems and
CQI continue to be misunderstood.
In the United States, today, the public debate on healthcare issues rages
on. The application of strategies, such as cost-benefit analysis as a means
for evaluating addition of new technologies to the healthcare cost structure has
not succeeded in curbing the rise in costs of healthcare services.
Because of this focused attention by third-party payers, federal and state
governments, and insurance companies, healthcare organizations are being
pressured to change. One of the strategies for changing involves implementing
quality assurance practices. The focus on quality should produce improvements
in productivity, innovation, and profitability. But, most importantly, the desired
outcome of a quality assurance program is self-improvement.
In its drive to become more productive and more competitive, industry
looked to such Quality Leaders as W. Edwards Deming, and J.M. Juran for
ideas. According to Deming, the way to become competitive is to undergo a
top-to-bottom quality-based transformation of the organization. This quality
transformation will produce the productivity improvements, innovation, and
profitability increases needed. The Deming philosophy emphasizes that quality
is never fully achieved; process improvement is never ending.
But, what is quality? Without defining it, David Garvin makes the point that
"in its original form, quality activities were reactive and inspection-oriented;
today, quality related activities have broadened and are seen as essential for
strategic success."! How can the broad context of quality be applied to the
diverse aspects of ESRD?
Furthermore, although far from a new concept, Continuous Quality Improve-
ment (CQI) has taken its place as a dominant theme in many industries.CQI
Overview of Contents
Chapter one explores the basics of starting a quality program in dialysis. Dr.
Sadler makes the point, which is repeated many times throughout this volume,
that Inspection or top-down, retrospective strategies using a set of specifica-
tions and threshold limits on processes in order to detect failures or defect
levels that exceed the threshold is inappropriate in the dialysis setting. The
pursuit of progressively greater excellence through systematic analysis of
processes and their direct, immediate outcomes is central to CQI.
Dr. Kjellstrand introduces an interesting idea: that, contrary to popular
opinion, poor outcomes in dialysis may be indicative that the physician has
fulfilled his obligation to a larger number of truly sick patients much better
than one whose survival statistics look more favorable. Thus, the insistence
that morbidity and mortality represent indicators of quality of care may actually
be detrimental to quality of care.
In their description of quality systems in a dialysis center, Ms. Prowant
et at. organize their discussion around the eight characteristics of successful
companies identified by Peters and Waterman in their book In Search of
Excellence 2 and apply those ideas to center operation.
The advantages of multi-center operations and application of CQI princi-
ples is discussed in Dr. Berger's chapter. Berger asserts that the notion of
CQI has been so rapidly adopted as "gospel" that it is losing its meaning.
Berger and Lowrie describe critical success factors for implementation of
CQI in a multi -center setting.
Aspects of quality of care in the home dialysis setting is covered by Dr.
Christopher Blagg. In his discussion, he uses the JCAHO Accreditation Manual
for Home Care. 3 It is just as important there be a active QA assessments
of the care provided these patients as for patients dialyzing in a dialysis unit.
Dr. Lee Henderson discusses quality of life and quality of care and their
relationship to membrane selection. Dr. Henderson sees membrane selection
ix
Notes
1. Garvin D.: Managing Quality, New York: The Free Press, 1988.
2. Peters TJ, Waterman RH Jr: In Search of Excellence. New York: Harper & Row, 1982.
3. Held PJ, Brunner P, Odaka M, Garcia JR, Port FK, Galin DS: Five-year survival for
end-stage renal disease patients in the United States, Europe, and Japan, 1982 to 1987.
Am J Kidney Dis 15: 451-457, 1990.
List of Contributors
xi
xii
JOHN H. SADLER
The words, "quality assurance" are familiar to most health professionals, but
not explicitly defined. Starting a formal QA program may seem redundant
(we already do that) or cumbersome (we don't have space or money for staff
who aren't productive) or inappropriate (the Network and surveyors do that).
In fact, good QA practices are not just a method to protect oneself from
critics and government review, but a systematic approach to reorganizing all
practices, assessing and improving them.
Program leaders such as the Medical Director, Nurse Manager, Admin-
istrator, CEO, must first become aware of the importance and utility of QA
throughout facility operations and commit to QA as an operating principle.
The federal structure surrounding ESRD therapy and the recurring and limited
range of dialysis services encourage that awareness and make serial, com-
parable observations accessible. The leaders must then encourage quality
assurance practices as an element of every operation. Properly done, these
practices are the prime instrument for staff education and improving perfor-
mance.
The most important function of quality assurance is self-improvement.
Fulfillment of that function will have byproducts which will document
activities to external agencies, establish a database for analysis and studies,
assist in staff evaluations and cost effectiveness analysis.
This approach to quality assurance is called Continuous Quality Improve-
ment (CQI). That contrasts with the widely practiced (by professional review.
organizations [PROs], federal surveyors and others) or external, "top-down"
QA review which imposes external standards and seeks to discover failures.
External review defines a threshold level of functioning which must, as a
minimum, be met. If it is met, no sanctions will be imposed. If it is exceeded,
no action is required. External review only seeks to protect from inadequate
care. It has no system for promoting good quality of-care, only for punishing
poor quality care. Because of its approach, which is at least suspicious if not
threatening, the response is often resistance, defensiveness, and suspicion in
return, which sours the atmosphere. This atmosphere has undermined many
well meaning efforts at QA and has led others to establish an internal staff,
Those staff directly involved then have a means to make adjustments for
improvement without bureaucratic constraints. Careful records are the source
of accountability. Control of the QA process passes to the person doing the
procedure and seeing its outcome: the one most gratified by improving it.
Aggregating outcomes allows insights which might not be realized as each
single event occurs. Internal critical scrutiny provides the basis for efficiently
making decisions following which the results are reviewed in the same way
as the observations leading to the change. If the change undertaken improves
outcomes or facilitates the process, assessment has led to greater quality of
performance. The practice is no longer a periodic oversight but a continuum
of critical scrutiny and data collection which is part of every activity.
Selection of specific items to track will improve with practice. Those
observations need to be explicit and informative rather than general statements
about the process. For instance, "post dialysis bleeding from (one) or (both)
puncture sites" rather than "access problem." The data needs to be useful by
itself, not a marker that there is a record of an incident to be reviewed. Check
off headings allow easy notation of events and unequivocal entry of which
event. It is better to have 100 specific items which may be noted with a
check or a date than to have 25 items pointing out the need to look back at
the record for details.
As an example, consider that simple but most anxiety producing step for
the patient, fistula puncture. Usually routine but potentially disruptive of
care, it is readily susceptible to analysis through aggregating experience to find
relationships to specific patients, staff members, or shifts. It allows factual
assessment of different products such as needles, their placement or method
of securing. Similarly, review of routine chemistries and of pressures in the
blood circuit should alert the team to spot the possibility of recirculation of
blood. Standing orders may be developed to allow clinical staff to proceed with
tests for recirculation when observations indicate it is likely. This kind of direct
involvement is helpful for learning, for responsibility, and ultimately, for
prompt intervention or understanding of the patient's condition, which results
in better care.
Such an analysis requires time but also assures that each involved clini-
cian shares the fundamental knowledge base on which greater expertise is built,
can assess personal performance as well as that of others, can understand
and perform tests such as those to determine recirculation, and has the aware-
ness that he or she may, by recording the justification, proceed independently
to get information for clarifying or resolving issues. Result: better staff, better
satisfaction, and better patient care.
Similar approaches are equally applicable to peritoneal dialysis. The patient
is in frequent contact and has repetitive visits for routine observation.
Noteworthy observations may easily be places in a format to allow checks
or entry of date of observation of "pain at the end of drainage" or purulent
drainage from exit" as tabulated events as well as a narrative record of the
circumstances. Tracking the patient's weight and blood pressure become part
4
of the matrix for scrutiny over time through being recorded in a format for
easy, repetitive review.
Decentralization of oversight creates opportunities for immediate recogni-
tion of problems and prompt response by the operator, diminishes risks to
that same operator, and allows self-management and pride in achievement.
Since the records are an output of the action, staff evaluation can improve with
good records. The clarity and usefulness of records often improve in such a
system. Hierarchical relationships - the chain of authority - ideally become
resources as well as constraints.
When the steps of each process are evaluated separately, redundant and
ineffective actions can be detected and eliminated. Idiosyncratic practices
may be recognized, corrected where appropriate, or turned into standard
procedure if the unique method offers advantages. Learning and communi-
cating increase, and perfunctory or casual practices which are likely sources
of inconsistent quality of outcomes can be found, and improved by those
employing them. The general practice of critical scrutiny can lead to improved
morale. CQI brings the scope of the review down to a manageable level
where a sense of control in QA is returned to the clinicians.
The possibility of such results from instilling CQI into every aspect of
clinical care is not a certainty. Leading examples, open and constructive
responses to scrutiny, and total fairness to all are elements needed to succeed.
Each person is a player in this QA scheme. It is internal, imbedded in all
practices as part of the process, its outcome, and its goals.
Arvidis Donabedian defined the elements of QA as structure, process, and
outcome. As practiced for years, structure has been assessed as credentials,
physical plant, instruments and devices, numbers of sites, staff, and mate-
rials; none of which gave assurance of how it is used. Process assessment
became routine through review of written procedures, defined responsibili-
ties and documentation of timing and completion of steps; none of which
gave assurance that the process was carried out effectively or efficiently. Lately
most discussion of QA has put down these measures and endorsed outcome
assessment as the gold standard, looking at measured results; complications,
costs, death, or recovery as examples. In the absence of case-mix or complexity
data (structure) and the path to the result (process) no comparisons could be
critically made. Raw outcomes, unadjusted by these factors, allow no rational
review of care on outcome analysis alone.
In health care, all three aspects of care are important in assessing the ultimate
or comparative merit of the service under review. That is true whether the
service is a diagnostic test, medication, or procedure.
With increasing demand for cost effectiveness as well as (or even instead
of) overall effectiveness, cost assessment sometimes threatens to become the
most important measurement of acceptability. Only an effective and realistic
assessment of the quality of care will give confidence to payers, health
professionals, and the public that the service is a good value, worth what it
costs.
5
Human aspects
back and find it; the detail needed to understand that patient is often not
recorded. Memory fails. Subsequent review can't distinguish between indi-
viduals with the same diagnosis but with widely different effects of the disease.
For example, there is usually a major difference between insulin dependent
diabetes mellitus and non-insulin dependent diabetes mellitus, and further great
variation in effects of the disease on individuals within each group. HCFA's
ESRD data does not yet even distinguish the groups, much less the effects
or complications. We need to know. Only the clinical team can get the
information and must get it at the beginning.
If outcomes of care are to be evaluated, first there must be characteriza-
tion of the patient on intake into the program or as soon as possible afterward.
Age, sex, race, primary diagnosis causing ESRD are usually recorded. These
have a bearing on what outcomes are possible. At the same time, secondary
diagnoses and co-morbid states are not routinely tabulated and make a dif-
ference too. For instance, the quality and security of access to the circulation
has a critical bearing on both success and the effort required to reach that
success. Tobacco and alcohol use may limit survival more than renal failure.
Loss of vision, hearing,an extremity or even a close family member may be
severely limiting or require extra measures to avoid deterioration. Such facts
about each patient should be tabulated in easily usable form as characteris-
tics for sorting, not only as part of a narrative.
Had we done a better job of analyzing patient variation we might have come
to objective criteria for adequacy of dialysis treatment which would be
generally acceptable. Including complexity in the description of clinical cir-
cumstances might help to make objective the efficient decisions which
experienced clinicians make through their practice "short cuts."
Implementation
assessed for each patient where specific additional conditions influence the
interpretation, or collectively to see if overall practices are meeting the goal
or require improvement. Multiple outcomes, each reflecting processes directly
affecting the result. Many are amenable to improvement. This is a good
example of a point where forms to display serial values for each patient
reveal the individual's course as well as current status. Further, a spread-
sheet of patients down one side and headings for lab values across the top
allows quick visualization of where this population's metabolic problems are.
Anemia control is now usually readily achieved, but still requires moni-
toring and thought. Even with recombinant erythropoietin, blood losses during
dialysis must be kept to a minimum and unexplained changes require assess-
ment for too little iron, too much aluminum, blood loss or hemolysis. A
single outcome can have multiple interval observations and several procedures
which affect it. Communication of findings and collection of all reports for
effective analysis is important. All clinicians can help each other in quality
assessment by making it easy to get the data together for review.
Water quality is an important element of hemodialysis which has been a
source of many problems. Each water treatment system has multiple values
to be monitored, each of which is a separate output related to one or more
processes in water treatment. If the initial condition of the tap water is not
known, the result of water treatment may be uninterpretable. Data must be col-
lected and acted upon, the outcomes evaluated to see if the process is effective,
and decisions regarding any need for change based on standards derived by
national consensus, reviewed by informed members of the staff.
If the water is contaminated with bacteria, the system must be disinfected,
but he cause of the contamination must be sought. Until the mechanism is
understood, no plan to prevent recurrences is likely to be practical. Although
the standards to be met are determined by outside agencies, the method of
meeting those standards and what resources will be expended in meeting
them is up to the operating staff of the facility.
All of these examples of critical scrutiny in dialysis usually exist in most
facilities, and often not recognized as the central QA function. CQI incorpo-
rates common sense and practical measures to monitor and assess practices
and results. It is often referred to as the "see, think, plan, act" sequence. The
steps often run together quickly since they are part of doing the job. There
is no separation of the observation, its assessment, and implementing the
rational correction.
These measures of quality require that everyone be involved. Routine
recording of clinical observations is the substance of the assessment of the
quality of practice. critical scrutiny requires only the analysis of those obser-
vations with an eye to better results for the patient or more effective practices
for the clinicians. Good records of clinical care are the essential QA document
in a CQI system. Appropriate and timely communication with the physician
or other responsible person completes the loop to action.
All this emphasis on ordinary practices as important to measuring the
9
C.M. KJELLSTRAND
I believe almost all the differences in mortality that exists between different
European regions and the United States, even when age is controlled, are
due to different acceptance criteria or transplantation activity. The reason
dialysis patient mortality is highest in U.S.A. and the Nordic countries is
because they have the highest acceptance to dialysis and the highest trans-
plant rates in the world. If this is true, a worse survival may reflect a better
fulfillment of a nephrologist's duties:
Quality assurance in this chapter is defined as a dialysis unit having a
morbidity and mortality that is comparable to an acceptable norm. The norm
will be derived from large data bases which are constantly updated, provide
much detail and also study the interaction and independence between dif-
ferent factors.
I believe quality comparison is one of the most important and neglected
areas in modern medicine. I will however in this chapter concentrate on the
pitfalls of quality assurance. This is not because of cynicism or a perverse
interest in the flaws of science. Rather, if the problems with quality assur-
ance are not considered the following may result:
1. The old and the poor masses may unnecessarily perish because large groups
of patients who can be successfully sustained may remain untreated and
die too early due to being considered "poor risk".
2. The best physician, humanely striving to take on the old and the sick may
be unjustly punished, while the greedy scoundrels, who ruthlessly select
out the "best cases" and then deny them transplantation, may be rewarded.
3. Propaganda by business looking hard at the bottom line may unneces-
sarily escalate the cost of treatment.
I am purposely avoiding rehabilitation as an outcome parameter because
dialysis rehabilitation status seems to be influenced more by cultural and eco-
nomic factors and by patient motivation, than any factor of dialysis itself [1].
Table I outlines various factors that have been thought to influence the
outcome of dialysis patients. In general, physicians have control and sole
responsibility only for Technical Treatment Related Factors and only some
control over Patient - Treatment Related Factors.
I. PRE-EXISTING
I. Demographic
Age XXX
Sex 0
Race XX
2. Social
Married ?
Family support ?
Area ?
Smoking ?
Alcohol ?
Income ?
3. Diagnosis
Diabetes XXX
Hypertensive nephrosclerosis XX
Systemic Disease XX
PCKD ++
4. Type and duration of renal failure
Acute ?
Intermediate ?
Chronic ?
Late start XX
5. Co-morbid conditions
Chronic heart failure XXX
Arteriosclerotic heart disease XXX
Stroke XXX
Peripheral vascular disease ?
Pulmonary XXX
Malignancy XXX
Gastro-intestinal ?
Hepatic ?
Hypertension XX
II. SELECTION
6. Generous acceptance XXX
High transplant rate XXX
Table I. (Continued).
8. Patient
Malnutrition (low BunlCr-ratio low Bun, low Chol, low Alb,
low BMI, low transferin) xxx
High CA x P04 product XXX
High interdialytic weight gain XXX
Inactivity XXX
Blood pressure control XXX
Age is a very important risk factor. While life expectancy of both young and
middle-aged patients is measured in decades, the mean survival time for
patients above age 70 is only 5 years on dialysis [2-12J. However, the relative
risk of dialysis i.e. the chance of dying on dialysis over a 5 year period when
compared to non-dialysed age matched population decreases with age. Thus
young patients aged less than 45 years encounter a 20 times increased chance
of dying within 5 years when they go on dialysis compared to only a twofold
increase in patients over the age of 75. It is obvious that old age not only
overrides other risk factors but also the very risk of dialysis itself [7J. Sex does
not appear to be a risk factor, thus there is no difference in survival between
men and women in the U.S. [4, 7, 10, 11J. On the contrary, race is of
importance in that non-white patients survive better than white [10, IIJ. The
5 year probability of surviving for black patients was 41.2% versus 37.4%
for white patients, even when adjusted for age, sex and primary disease [11J.
This unexpected result may have to do with the low transplant rate in black
patients and points to an important role for selection in patient survival [13-15].
Diagnosis
Certain diagnoses are associated with a higher death rate. This includes diabetes
and hypertensive nephrosclerosis and some other systemic diseases such as
myeloma and amyloid which appear to be associated with a shorter survival.
On the other hand polycystic kidney disease appears to be associated with a
good outome [2-12J.
There appear to be no good studies of this. Intuitively one would guess that
a fairly short period of uremia may be associated with less chronic meta-
bolic effects and thus with a "stronger body" starting dialysis.
15
Co-morbid conditions
There are many studies that prove that the presence of other diseases may
shorten survival on dialysis. Chronic heart disease, atherosclerotic heart disease,
strokes, peripheral vascular disease, COPD and malignancies have all been
associated with a poor outcome as has severe hypertension [2-11, 21]. Some
of these diseases appear to be additive in their ill effect. For example, in one
study, non-diabetic patients, younger than 45 years, without risk factor, had
a 10% six year mortality, and those with arteriosclerotic heart disease or stroke
had a 20% mortality rate but those with both diseases had a 40% mortality rate
[4]. It is obvious that to evaluate the influence of all these co-morbid factors
and their interrelationships with each other and age requires a large number
of very carefully prospectively examined patients. Such material does not exist
and may never be in existence.
Malnutrition is an extremely important factor in predicting outcome. thus,
patients who have a low BUN/creatinine ratio, or a low BUN, cholesterol,
triglycerides, albumin, BMI or transferrin value have a very high mortality rate
on dialysis [16-18].
It is quite clear that as patients present to dialysis many factors exist which
will predict their survival. It is then equally clear that by careful selection
one can greatly influence survival results. We hypothesized that physicians
who liberally accept many patients per population were likely to take on
many old patients and many patients with degenerative diseases, systemic
diseases, or malnourishment. The survival results of such a physician would
be "poor" when compared to a more selective and fastidious physicians who
treated only those who were young, and except for their kidney disease
otherwise healthy. While this is self-evident, it may also appear equally self-
evident that simple age matching could avoid many of these problems but
things are more complicated than this as will be discussed below. Secondly,
we hypothesized that someone who co-operated with a very active transplant
program would also have poor mortality rates. Over 80% of all patients, who
now start dialysis in the United States and in many other countries with a
high acceptance rate, will remain on dialysis and not be transplanted. The
chance of receiving a transplant is now falling all over the world as dialysis
acceptance rates continue to rise, while transplant rates have levelled or even
declined [19]. Kidneys for transplantation are thus a much more scarce resource
than machines for dialysis. Transplantation removes the very best patients, who
are the young, without other diseases, and who have a long life-expectancy
on dialysis. This results in a worse survival for centres with a high trans-
plant rate when compared to centres where transplantation rates are low.
To investigate this we correlated cumulative 4 year survival in age-matched
16
60 ~-4 __ ~~ __ ~~ __ ~~~~~~~ __ ~~
-USA
60 ~~--~--~~~~--~--~~--~~
20 30 40 so 60 70
PERCENT TRANSPLANTED AT 4 YEARS
Fig. 1. (Top) Linear regression analysis showing the relation between cumulative 4-year
dialysis survival for patients age 15 to 44 and the acceptance rate to dialysis. The more patients
that are accepted, the worse is cumulative survival. (Bottom) Cumulative 4-year dialysis survival
in relation to the percent of patients transplanted at 4 years. The cumulative survival becomes
worse as more patients receive transplants at 4 years. The transplant surgeon takes the young
and healthy who show good survival on dialysis and leaves only the old and those with many
other diseases who have poor survival. When both acceptance rate and transplant rates are
evaluated through multiple regression analysis, over 90% of the differences in cumulative survival
are explained. Thus, differences in survival are not dependent on geographical location, but on
generosity in acceptance and transplant rates.
(From reference 20 reproduced with permission of W. B. Saunders Co.)
18
[23]. The old, the sick and the malnourished will cost more, because they
are more difficult to care for. When they are dumped on public facilities not
only does the survival curve for the fastidious unit shift upwards but the
survival curve of the receiving hospital will shift downwards [22, 23]. This
gives the twice false impression, that quality is better, when in reality selec-
tion is the reason for the improved survival. In a perverse fashion then "better
may actually mean worse". Fastidious "good outcome" physicians have not
fulfilled their obligations to those who need their services particularly urgently.
One can of course argue that "sick" patients should not be treated at all
as the life expectancy is short and the cost relative high. Actually, when counted
in QUALY's hemodialysis for any patient fares very poorly. When the yearly
cost is divided by the expected quality of life and its duration the figures
become very high [24, 25]. The ultimate extension of this of course is that "the
only good patient is a dead patient" and although one can shrug that off as a
meaningless cynicism, when this is brought into the clinical arena it is indeed
a deadening reality for the old, as the easiest of all selection criteria is age,
in itself often used to exclude the old from treatment [26-28].
Technical
This group of factors is the true matter of investigation for quality assur-
ance;
Is the physician performing technically adequate dialysis?
Many factors enter into this equation, and over some the physician has only
partial control as for example late start of dialysis. It has been shown that
patients who come with severe uremia to dialysis may have incurable sec-
ondary metabolic effects that lead to complications and an increased mortality
[29], while those started early do particularly well [30]. To give a patient
insufficient dialysis, for example as expressed as KTIV of less than 0.9 has
been associated with a high morbidity [31, 32]. Similarly intermittent peri-
toneal dialysis is no longer regularly used because of the high complication
rate and the high mortality associated with this procedure [7]. Comparable data
do not exist for home hemodialysis and what differences exist, appear best
explained by different selection criteria [7, 9, 10]. CAPD has the advantage
of an even control of uremia and electrolytes but offers the patient KTIV values
that are insufficient for hemodialysis patients [31, 32]. Fast dialysis has also
been associated with an increased mortality [33] but much of this appears to
be due to the fact that fast sufficient dialysis and short insufficient dialysis
are not separated [34]. Another factor though is that fast dialysis, even when
thought to be adequate, may in reality more easily become inadequate than
slower and longer dialysis. Biocompatibility has been one factor associated
19
with outcome. More biocompatible and open membranes have been thought
to lead to less dialysis arthropathy - amyloid. However, the largest study of
this problem, comparing the 10 year incidence of these diseases between
matched patients on poly acrylonitrile or cellophane membranes has failed to
show any statistical significance [35]. Still, many physicians will use the
open membrane for patients. The higher cost of this, in a system with a capped
budget, results in an erroneously perceived increased quality for some patients
but no dialysis for others.
Blood pressure control during dialysis is an important factor for longevity
as shown many years ago [36]. Reuse, that has resulted in lethal infection,
has invariably been associated with a higher survival or no difference in
survival when compared to no-reuse in rigorous studies [8].
These factors are only partially under a physician's control. For example, a
high calcium x phosphorus product, malnutrition, "giving up", a high intra-
dialytic weight gain and hypertension are factors over which physicians have
only partial control [8]. These factors are only partially influenced by a physi-
cian's actions and patient education, but are much influenced by compliance.
Inactivity in old patients has also been associated with a high mortality rate.
It is not known if inactivity only is the reflection of other underlying factors
that simultaneously shorten life and lead to inactivity, or if it is a risk factor
itself, that can be influenced by education, physical therapy and training
[12].
Conclusions
References
1. Carlson DM, Johnson WJ, Kjellstrand CM: Functional status of patients with end-stage renal
disease. Mayo Clin Proc 62: 338-344, 1987.
2. Degoulet P, Legrain M, Reach I, Aime F, Devries C, Rojas P, Jacobs C: Mortality risk factors
in patients treated by chronic hemodialysis. Nephron 31: 103-110, 1982.
3. Hutchinson T, Thomas DC, Mac Gibbon B: Predicting survival in adults with end-stage renal
disease: An age equivalence index. Ann Int Med 96: 417-423, 1982.
4. Shapiro FL, Umen A: Risk factors in hemodialysis patient survival. ASAIO 6: 176-184,
1983.
5. Vollmer WM, Wahl PW, Blagg CR: Survival with dialysis and transplantation in patients
with end-stage renal disease. N Engl 1 Med 308: 26, 1553-1558, 1983.
6. Neff MS, Eiser AR, Slifkin RF, Baum M, Baez A, Gupta S, Amarga E: Patients surviving
10 years of hemodialysis. Am 1 Med 74: 996-1003, 1983.
7. Hellerstedt BA, Johnson WJ, Ascher N, Kjellstrand CM, Knutson R, Shapiro FL, Sterioff
S: Survival rates of 2,728 patients with end-stage renal disease. May Clin Proc 59: 776-783,
1984.
8. Held PJ, Pauly MV, Diamond L: Survival analysis of patients undergoing dialysis. lAMA
257: 645-650, 1987.
9. Mailloux LU, Bellucci AG, Mossey RT, Napolitano B, Moore T, Wilkes BM, Bluestone
PA: Predictors of survival in patients undergoing dialysis. Am 1 Med 84: 855-862, 1988.
10. Silins J, Fortier L, Mao Y, Posen G, Ugnat A-M, Brancker A, Gaudette L, Wilgle D:
Mortality rates among patients with end-stage renal disease in Canada, 1981-86. CMAl 141:
677-682, 1989.
11. Renal Data System U.S.A., The National Institutes of Health, The National Institute of
Diabetes and Digestive and Kidney Diseases, Division of Kidney, Urologic, and Hematologic
Diseases Bethesda MD. Annual Report 1989.
12. Westlie L, Umen A, Nestrud S, Kjellstrnd C: Mortality, morbidity and life satisfaction in
the old dialysis patient. Trans Am Soc for Artif Intern Organs 30: 21-30, 1984.
13. Kjellstrand CM: Giving life-giving death. Ethical problems with high technology medicine.
Acta Med Scand 725 (Supp): 1-80, 1988.
14. Kjellstrand CM: Racial, sexual and age discrimination in renal transplantation. Arch Intern
Med 148: 1305-1309, 1988.
15. Held PJ, Pauly MV, Bovbjerg RR, Newman J, Salvatierra 0 Jr: Access to kidney trans-
plantation - has the United States eliminated income and racial differences? Arch Intern
Med 148: 2594-2600, 1988.
16. Acchiardo SR, Moore LW, Latour P: Malnutrition as the main factor in morbidity and
mortality of hemodialysis patients. Kidney Int 24: 16:S-199-S-203, 1983.
17. Oksa H, Pasternack A, Pasanen M: Serum urea-creatinine ratio as a prognostic index in
hemodialysis patients. Clin Neph 27: 3:125-130, 1987.
18. Marckmann P: Nutritional status and mortality of patients in regular dialysis therapy.
lInt Med 226: 429-432, 1989.
19. Kjellstrand CM, Shideman J: On the impact of and need for chronic dialysis and renal
transplantation. Trans Am Soc Artif Intern Organs 34: 328-332, 1988
20. Kjellstrand CM, Hylander B, Collins A: Mortality on dialysis - on the influence of early
start, patient characteristics, and transplantation and acceptance rates. Am 1 Kidney Diseases
15: 483-490, 1990.
21. Collins A, Kjellstrand CM, Hull A, Parker T: Mortality on dialysis appears directly
dependent on the generosity of the nephrologist and the activity of the transplant surgeon.
ASATO 36 (Abstr): 16, 1990.
22. Plough AL, Salem SR, Shwartz M, Weller JM, Ferguson CW: Case mix in end-stage renal
disease. N Engl 1 Med, 1432-1436, 1984.
23. Ansell DA, Schiff RL: Patient Dumping: Status, Implications, and Policy Recommendations.
lAMA 257: 1500-1502, 1987.
21
24. Williams A: Economics of coronary artery bypass grafting. Br Med 1 291: 326-329, 1985.
25. La Puma J, Lawlor EF: Quality-adjusted life-years: Ethical implications for physicians
and policymakers. lAMA 263: 2917-2921, 1990.
26. Wetle T: Age as a risk factor for inadequate treatment. lAMA 258: 516, 1987.
27. Lamm RD: Health care as economic cancer. Dialy & Transpi 16: 432-433, 1987.
28. Callahan D: Setting limits - medical goals in an aging society. New York: Simon & Schuster,
1987.
29. Kjellstrand CMN, Evzans RL, Petersen RJ, Shideman JR, von Hartitzsch B, Buselmeier
TJ: The "unphysiology" of dialysis: A major cause of dialysis side effects? Kidney Int 7:
S30-S34, 1975.
30. Bonomini V, Feletti C, Stefoni S, Vangelista A: Early dialysis and renal transplantation.
Nephron 44: 267-271, 1984.
31. Lowrie EG, Laird NM, Parker TF, Sargent JA: Effect of the hemodialysis prescription on
patient morbidity. N Engl 1 Med 305: 1176-1181, 1981.
32. Gotsch FA, Sargent JA: A mechanistic analysis of the National Cooperative Dialysis Study
(NCDS). Kidney Int 128: 526-534, 1985.
33. Kjellstrand CM: Short dialysis increases morbidity and mortality. Contr. Nephroi44: 65-77,
1985.
34. Levin NW: Mortality impacts of shorter dialysis procedures. Kidney Inti 35: 254, 1989.
35. Brunner FP, Brynger H, Ehrich JHH, Fassbinder W, Geerlings W, Rizzoni G, Selwood
NH, Tufveson G, Wing AJ: Case control study on dialysis arthropathy: The influence of two
different dialysis membranes: Data from the EDTA Registry. Nephroi Dial Trans 5: 432-436,
1990.
36. Lundin P, Adler AJ, Feinroth MV, Berlyne GM, Friedman EA: Maintenance hemo-
dialysis. lAMA 244: 38-40, 1980.
CHAPTER 3
Peters and Water found that successful companies have a sound set of beliefs
and values upon which they premise all policies and actions and summarizes
these as "hands on and value driven" [1].
Program integration
Philosoph} of self-care
• Education is initiated 3-6 months prior to the need for chronic dialysis.
• There an unbiased presentation of all treatment options and pros and cons of each.
• Classes supplement one to one sessions.
• Families and significant others are included in the education process.
• Patients meet patients on various ESRD therapies.
• The patient is assessed by the renal team.
• Medical advantages and/or contraindications for a therapy are discussed with the patient.
• The patient is included in the decision-making process.
A recent review of63 patients who selected peritoneal dialysis [4] indicated
that the predominant reason (25%) was to maintain independence, activities
and flexibility in scheduling. Motivation to perform one's own dialysis and
maintain some control was the reason 15% chose peritoneal dialysis and an
additional 12% of the patients cited the ability to continue working. Although
25
the majority of patients chose PD for positive reasons, fourteen percent selected
peritoneal dialysis because of long distances to a hemodialysis unit and an
additional 9% because they felt travel to center hemodialysis or the demands
of home hemodialysis would impose an unacceptable burden on their families.
Administrative support
Philosophy of excellence
Productive companies treat their employees with dignity and respect. Employ-
ees are partners, experts and team members [1].
Table II. Areas of physician knowledge and competence required for a successful peritoneal
dialysis practice.
Peters and Waterman noted that top companies are both centralized and
decentralized. They encourage autonomy at all levels, yet hold a core set of
values [1]. An example of this in the dialysis unit is the professional team
approach which encourages autonomy in patient care within a discipline, yet
strongly promotes the communication and continuity of care among team
members and in a number of settings (the home, outpatient clinic, hospital).
The shared value system discussed above is another example of the tight
properties which contribute to success.
The most successful companies stick close to the central skill or product and
this enables them to perform more effectively than the competition [1]. Because
of the highly specialized nature of peritoneal dialysis programs, few units
branch out into other types of care. This discussion will focus on what is
required to provide a high quality of peritoneal dialysis.
First of all, peritoneal dialysis programs have physical space requirements.
Eighteen centers of excellence for patient education and training practices
identified by Baxter Healthcare had significantly higher technique survival and
patient survival rates than their counterparts. All of these units had space
devoted solely to the home peritoneal dialysis program [9]. Eighty-two percent
of these centers had a PD training room, a separate PD clinic area, and a PD
nurses' office. All of the units also had separate storage and utility areas for
peritoneal dialysis [9].
One of the most basic requirements is a set of standards of clinical practice.
A standard is the yardstick of the quality of a service and Mason states that
nursing standards "define unequivocally what quality care is and provide
specific criteria that can be used to determine whether quality care has been
provided" [20]. A standard describes what should be done and how the patient
will benefit from the care. The American Nephrology Nurses Association
has published standards of clinical practice for nephrology nursing [21].
Appropriate standards can be selected and adapted for use in a particular
dialysis unit.
31
A policy and procedure manual that guides safe practice is also essential.
Policies for all nursing procedures, machine and equipment maintenance,
emergencies, and ma.naging problems will help ensure consistent, safe care.
A unit's standards and policies and procedures are also used by inspectors
and surveyors to evaluate the program. Although clinical policies and proce-
dures are developed primarily by the nursing staff we recommend consultation
with administration regarding legal issues, with the medical director regarding
nursing protocols to manage complications, and with patients regarding the
self-care procedures. An annual review and update of policies and proce-
dures is appropriate. Examples of the types of policies and procedures required
for a peritoneal dialysis program are listed in Table IV.
Table IV. Selected Types of Peritoneal Dialysis Unit Policies and Procedures.
General Procedures
Handwashing
Measuring blood pressure
Quantitative urine collection
Peritoneal dialysis procedures
Exit site care procedures
Exchange procedure (for each system used)
Cycler procedures
Machine set up for closed drain
Machine set up for open drain
Administration of intraperitoneal medication
Catheter adapter change procedure
Peritoneal equilibration test
Quantitative dialysate collection
Protocols for managing problems and complications
Obstruction of flow
Fibrin in dialysate
Contamination of the system
Crack or hole in catheter
Hypervolemia
Peritonitis
Emergency procedures
Cardio-pulmonary arrest
Fire
Hurricane or Tornado
Catheter placement
Peritoneal Dialysis
Acute catheter insertion procedure
Chronic catheter insertion
Current policies
Marking catheter exit site
Preoperative orders
Catheter placement video
Catheter Break-in
Chronic dialysis orders
Peritoneal equilibration test
Exit site infection protocol
Peritonitis protocol
includes the choice of catheter type, determination and marking of the exit site,
and cleansing the abdomen with a disinfectant scrub. Determining the exit
site is usually a joint decision and the surgeon, PD nurse, patient and/or
nephrologist may be involved. Factors to consider for optimal exit site
placement are: avoiding skin folds, the beltline, and scar tissue; and place-
ment where the patient can observe and manipulate that catheter for ease of
exit care [22, 23]. Broad spectrum prophylactic antibiotic therapy is gener-
ally recommended [23, 24].
Catheters can be medically inserted with a trocar or peritonesocopy or
surgically inserted. For optimal results a few general guidelines apply to
either procedure.
• Catheter placement should be limited to experienced surgeons or nephrol-
ogists [24].
• Local anesthesia combined with a sedative is adequate for uncomplicated
insertion [24].
• A lateral or paramedian insertion site is preferred [25].
• The catheter should be soaked in sterile solution prior to insertion to saturate
the cuffs and expel air [23, 24, 26].
• Sutures should not be used at the exit site [23, 26].
• Solution should be infused and drained prior to closure to evaluate catheter
function [24, 26].
• Nonreative, absorbable sutures should be used for the initial incision [23,
24].
• A sterile dressing is applied and the catheter is anchored at the exit site
to prevent movement [24].
The catheter should be immobilized well during healing to avoid tension
and torquing of the catheter. Tight clothing or other external pressure and
trauma to the exit site should be avoided [24].
Ambulatory dialysis should be delayed for 10 to 14 days. During this time
intermittent peritoneal dialysis can be administered with the patient resting
supine and with gradually increasing solution volumes, or the patient may
receive hemodialysis [27].
33
The impact of exit site care procedures on the incidence of exit site infec-
tion has been evaluated post catheter insertion [27-29] and an expert panel has
also made recommendations for post operative exit care [24]. Common
elements of these post operative exit site care procedures are listed in Table
VI.
Table VI. Common elements of post operative procedures for peritoneal catheter exit site care.
The chronic exit site care procedures which have been recommended [24]
and studied [29-33] are more varied. Common elements include cleansing
and drying the exit site and securing the catheter. The ideal cleansing agent
is not known and recommended frequency varies from daily to several times
weekly.
Patient education
• Educational background
• Work experience
• Previous involvement in self-care activities
• General level of health
• Level of cognitive function
• Psychiatric/emotional status
• Current knowledge of PD
• Concerns regarding ESRD and/or dialysis
• Current stressors and symptoms or stress
• Factors that interfere with health care or following the medical regimen
• Level of activity and independence
• Dialysis partner or backup support
• Physical disabilities which could affect learning
• Motivation to learn and perform PD
• Best way to learn
• Best time to learn
• Expectations of PD education program
• Reading test
• Memory test
planning for evaluation. The process continues through the actual implemen-
tation and evaluation phases. Most PD programs use a 1: 1 nurse patient ratio
for initial patient education.
Table VIII lists topics covered by over 90% of 18 facilities evaluated for
the best demonstrated practices in patient education [9]. Lecture and discus-
sion were the primary modes of patient teaching among the best demonstrated
practice centers. Demonstration, return demonstration and simulated problem
solving were also used by all of these facilities [9].
The patient education process is documented a number of ways. An account
of the assessment, goals and progress is recorded in the progress notes or nurses
notes. In addition, annotations are often made on the patient objectives or
education checklist to document a patient's mastery of the subject, that a
topic has been omitted, or that the routine approach or procedure has been
modified. At the completion of training there is a comprehensive evaluation
of the patient's (and/or partner's) knowledge and skills. A variety of testing
methods are used: verbal and written testing, return demonstration of proce-
dures, simulated problem solving.
It takes about 6 days for an average patient to complete PD training. Training
time varies from 5-8 hours per day and ranges from 5-10 days [9].
The process of patient education is ongoing, so review, reassessment of
learning needs and/or teaching take place at almost every patient contact.
35
• Asepsis*
• Handwashing
• Exchange procedure*
• Exit site care*
• Recommended diet, meal planning
• Fluid balance
• Record keeping
• Procedure for system contamination
• Causes of peritonitis
• Peritonitis prevention
• Peritonitis symptoms*
• Reporting peritonitis to unit
• Peritonitis treatment
• Catheter complications
• Supply inventory
• Vital signs
• Laboratory values
• Patient responsibilities
• Communications and call system
Followup care
interactions, the degree of self care, supply inventory and storage, general
management of health, emotional adjustment and dietary practices [40].
Furthermore, home visits to both patients with perceived problems and those
doing well resulted in recommendations for change [40].
As the age of dialysis patients increases, some chronic dialysis patients require
continuous skilled nursing care and reside in long term care facilities. Peritoneal
dialysis programs have successfully taught the staff at nursing homes to
perform PD procedures and manage fluid balance and other aspects of care.
The education outline is similar to that used for patient and family educa-
tion. The patient receiving chronic PD in a long term care facility still requires
37
follow up by the PD staff, and continuing education and support are provided
to the staff.
Patients performing dialysis and managing their ESRD at home require 24 hour
professional staff coverage to guide and assist them in identifying and
managing problems or complications. Support is also essential for newly
trained patients just beginning to dialyze independently at home. Both nursing
and physician must be on call 24 hours/day. Because most questions or
problems are related to dialysis procedures and/or are within the realm of
nursing practice, in most programs the nurse is the initial contact. For medical
problems the nurse consults with or refers to the physician on call. Although
we are aware of programs with only physicians on call, this is not optimal,
because problems other than medical emergencies are frequently not reported
by the patient, or are not dealt with until regular office hours when the nurse
returns.
The traditional CAPD prescription (8-9 liter dialysis solution and 3-4 daily
exchanges) does not provide adequate dialysis for all patients, especially
after loss of residual renal function [43]. In order to determine the optimal
dialysis prescription for an individual patient the physician needs informa-
tion regarding the patient's peritoneal membrane characteristics. The peritoneal
equilibration test [42] measures the dialysate to plasma ratio of creatinine at
0, 2, and 4 hours dwell time, dialysate glucose/dialysate glucose at 0 dwell
time at 2 and 4 hours, and ultrafiltration volume. These values can be compared
to or plotted on published curves to determine if solute transport is average,
high or low. We recommend a peritoneal equilibration test at the time the
patient begins chronic peritoneal dialysis therapy, and repeated when there
are clinical indications that there may be a change in membrane transport
characteristics. Such indicators include an increase or decrease in ultrafiltra-
tion, and an unexplained change in serum chemistries.
Twardowski and colleagues [43, 44] and Diaz Buxo [45] have described
patterns of ultrafiltration and clearances in relation to solute transport. Table
IX indicates the most appropriate or preferred dialysis prescriptions based
on solute transport rates.
Obviously, if physicians are to prescribe the most appropriate therapy based
on peritoneal membrane transport rates, there must be administrative and
nursing support for CCPD, IPD and other cycler therapies. This includes
policies and procedures, availability of machine installation and maintenance,
nurses skilled in operating the cyclers, an educational curriculum and mate-
rials, and billing systems.
38
Table IX. Dialysis prescriptions based on peritoneal equilibration test results [43].
Clinical criteria
Patient "feels well and looks good"
Blood pressure controlled
Good fluid balance
Stable lean body mass
Stable nerve conduction velocities
Absence of uremic symptoms
Anorexia
Dysgeusia
Nausea
Vomiting
Asthenia
Insomnia
Laboratory criteria
Electrolytes within normal range
Serum creatinine < 20 (muscular persons)
< 15 (non muscular persons)
Hematocrit> 25% without EPO or steroids
39
Prevention of peritonitis
Prompt diagnosis and effective treatment of peritonitis are essential for quality
care in a peritoneal dialysis program.
A high percentage of no growth on dialysate cultures [58, 59] during the
early years of CAPD stimulated research to improve the effectiveness of
laboratory cultures. There is abundant evidence that special procedures are
required to culture small numbers of bacteria diluted in dialysate. Culturing
large amounts of fluid [60, 61] using filtration [58, 62] or centrifugation [62,
63] to concentrate the sample, and removal of antibiotics present in the
specimen [61] have been shown to increase the proportion of positive cultures.
Bint et al. [64] suggest that the rate of positive cultures in clinical peritonitis
should exceed 90%.
Williams et at. recommended the use of a cytocentrifuge for the differen-
tial white cell count [65]. The cytocentrifuge concentrates the cells in a small
area on the slide so there are enough cells for an accurate differential count.
Two expert committees have issued recommendations for treatment of
40
peritonitis [61, 66] and reviews of peritonitis treatment has been published [67,
68]. In addition, Twardowski et at. [69] and the North American expert panel
[61] have developed decision trees for the medical management of peritonitis
episodes.
Peters and Waterman found that the successful companies were not para-
lyzed by formal communications and procedures [1]. Employees from the
top down communicated frequently and informally. Systems were simplified
and evaluation of progress was based on a few key numbers. In addition,
employees were encouraged to be innovative, and to experiment. Two of
these qualities particularly apply to peritoneal dialysis programs: monitoring
a few simple numbers, and encouraging innovation and experimentation.
Evaluation of the incidence of peritonitis and characteristics of peritonitis
episodes is a useful indicator of quality within a peritoneal dialysis program
and should be continuously monitored. A simple ratio of peritonitis episodes
over patient months exposure may be used to calculate the peritonitis rate or
life table analysis may be used to determine the probability of the first (or
subsequent) peritonitis episodes [70-72]. It is also of interest to compare
infection rates for each type of peritoneal dialysis therapy and each type of
peritoneal dialysis system which is used. Identifying the portal of entry (or
presumed etiology) may help in identifying trends and developing strategies
to reduce the incidence of peritonitis [59].
The incidence of exist site infection is also a useful indicator of the
effectiveness of catheter placement and exit site care procedures. Although
there is not a uniformly accepted definition of exit site infection in the
literature, infection rates in a single program can be compared over time if
the definition of exit site infection is consistent. Catheter removal and the
reasons for peritoneal catheter removal should also be monitored.
Patient survival and technique survival (the proportion of patients remaining
on peritoneal dialysis therapy) determined by actuarial techniques should
also be monitored as general indicators of program quality.
Innovation and experimentation is probably a hallmark of most successful
PD programs. We believe that all clinical research eventually results in
improved patient care, whether from increased knowledge of physiology,
improvements in patient assessment, development of new procedures, more
effective education, improved documentation and communication, develop-
ment of a new regimen or delivery system, development of new diagnostic
techniques, alternative treatment methods, improved cost effectiveness, or
identification of specific risk factors [73]. Participation in research can also
enhance professional growth and increase job satisfaction resulting in improved
staff retention for non-physician team members.
41
Peters and Waterman also point out that problems with "product" or service
may be offset by caring [1]. On a recent patient satisfaction survey one of
our patients expressed dismay that the satellite unit near her home did not offer
CAPD support, and that she could not be admitted to the local hospital because
none of the nurses were trained to do peritoneal dialysis. In spite of this, she
was quite positive about the dialysis unit, largely due to the "caring" she
received. "I'm well pleased with all the services rendered to me by the nursing
staff and the [other] personnel. ... I couldn't have been treated better."
Measuring customer satisfaction is another hallmark of successful busi-
nesses. We recommend an annual evaluation of the quality of service using
a surveyor questionnaire. This must be written at an appropriate reading
level and printed with a large hold typeface so as many patients as possible
can read it.
Another principle borrowed from good business practice is to conduct loss
reviews. When a patient transfers to another dialysis program the reason
needs to be identified and then the team should evaluate whether the transfer
could have been prevented by changes in the structure or process or the level
of caring.
Summary
References
1. Peters TJ, Waterman RH Jr: In Search of Excellence. New York: Harper & Row, 1982.
2. Boen ST: Integration of continuous ambulatory peritoneal dialysis into endstage renal failure
programmes: Present and future. In Atkins RC, Thomson NM, Farrell PC (eds) Peritoneal
Dialysis (pp 424-429). Edinburgh, Churchill Livingstone, 1981.
3. Baxter Healthcare Corporation: The Best Demonstrated Practices Program: Modality
Selection Practices. Deerfield, IL: Author, 1988.
4. Campbell A: Choosing an appropriate chronic dialysis therapy: A study of decisions by
nephrology staff and patients (abstract). Perit Dial Int 11 (Supp 1): 37, 1991.
5. Ross CJ, Rutsky EA: Dialysis modality selection in the elderly patient with end-stage
renal disease: Advantages and disadvantages of peritoneal dialysis. Peritoneal Dialysis in
the Geriatric Patient, A Supplement to the Advances in Perit Dialysis 6: 11-17, 1990.
6. Hamburger RJ, Mattern WD, Schreiber MJ Jr, Soderblom R, Sorkin M, Zimmerman SW:
A dialysis modality decision guide based on the experience of six dialysis centers. Dialysis
& Transplantation 19(2): 66-69, 84, 1990.
7. Shayman D, King K: Missouri Pre-Dialysis Patient Education Program. Columbia, MO:
Missouri Kidney Program, University of Missouri-Columbia, 1986.
8. Baxter Healthcare Corporation: Predialysis patient education program. Deerfield, IL: Author,
1990.
9. Baxter Healthcare Corporation: The Best Demonstrated Practices Program: Patient
Education and Training Practices. Deerfield, IL: Author, 1987.
43
10. Nolph KD, Prowant BF, Webb J: National conference of continuous ambulatory peritoneal
dialysis. Perit Dial Bull 1: 65-66, 1981.
11. Oreopoulos DG: The peritoneal dialysis nurse: The key to success. Perit Dial Bull 1:
113-114, 1981.
12. Wilkens K (ed): Suggested Guidelines for Nutrition Care of Renal Patients. American
Dietetic Association, 1986.
13. Council on Renal Nutrition NKF, Inc: Guidelines for Estimating Renal Dietitian Staffing
Levels. New York: National Kidney Foundation, 1989.
14. Ducanis AJ, Golin AK: The Interdisciplinary Health Care Team. Germantown, MD: Aspen,
1979.
15. Marram GD, Barrett MW, Bevies EO: Primary Nursing, A Model for Individualized Care.
St. Louis: CV Mosby, 1979.
16. Giovanetti P: Evaluation of primary nursing. Annual Review of Nursing Research 4: 127-151,
1986.
17. Reed SE: A comparison of nurse-related behavior, philosophy of care and job satisfaction
in team and primary nursing. Journal of Advanced Nursing 13: 383-395, 1988.
18. Gardner KG: The Effects of Primary Versus Team Nursing on Quality of Patient Care
and Impact on Nursing Staff and Costs: A Five Year Study. Rochester NY: Rochester General
Hospital, 1989.
19. Joiner C, Van Servellen GM: Job Enrichment in Nursing: A Guide to Improving Morale,
Productivity and Retention. Rockville, MD: Aspen, 1984.
20. Mason EJ: How to Write Meaningful Nursing Standards, 2nd ed. New York: John Wiley
& Sons, 1984.
21. Burrows-Hudson S (ed): Standards of Clinical Practice for Nephrology Nursing. Pitman,
NJ: American Nephrology Nurses' Association, 1993.
22. Copley JB: Prevention of peritoneal dialysis catheter-related infections. American Journal
of Kidney Diseases 10: 401-407, 1987.
23. Suzuki T, Hasuo M, Suzuki T et al.: How to prevent exit site/tunnel infection in CAPD
patients? (abstract). Perit Dial Int 9 (Supp 1): 196, 1989.
24. Gokal R, Ash S, Helfrich BG et al.: Peritoneal catheters and exit-site practices: toward
optimum peritoneal access. Perit Dial Int 13: 29-39, 1993.
25. Helfrich GB, Pechan BW, Alijani MR, Barnard WF, Rakowski TA, Winchester JF:
Reduction of catheter complications with lateral placement. Perit Dial Bull 3 (Supp 4):
S2-S4, 1983.
26. Tenckhoff H: Chronic Peritoneal Dialysis Manual. Seattle: University of Washington, 1974.
27. Copley JB, Smith BJ, Koger DM, Rodgers DJ, Fowler M: Prevention of postoperative
peritoneal dialysis catheter-related infections. Perit Dial Int 8: 195-197, 1988.
28. Schmidt L, Prowant B, Schaefer R, Russ J, Kennedy J, Ryan L, Burrows L, Satalowich
R, Bartelt C: An evaluation of nursing intervention for prevention of postoperative
peritoneal catheter exit site infections (abstract). ANNA Journal 13: 98, 1986.
29. Jenson SR, Pomeroy M, Davidson M, Cox M: Evaluation of dressing protocols that reduce
peritoneal dialysis catheter exit site infections. ANNA Journal 16: 425-431, 1989.
30. Starzomski R: Three techniques for peritoneal catheter exit site dressings. ANNA Journal
11: 9-16, 1984.
31. Prowant BF, Schmidt LM, Twardowski ZJ, Griebel CK, Burrows LK, Ryan LP, Satalowich
RJ: A randomized prospective evaluation of three procedures for peritoneal dialysis catheter
exit site care. ANNA Journal 15: 219-223, 1988.
32. Luzar MA, Borwn C, Bals D, Hill L, Issad B, Monnier B: CAPD exit site care - results
of a randomized multicenter study. Perit Dial Int 10: 25-29, 1990.
33. Fuchs J, Gallagher ME, Jackson-Bey D, Krawtz D, Schreiber MJ: A prospective random-
ized study of peritoneal catheter exit-site care. Dialysis & Transplantation 19(2): 81-84, 1990.
34. Falvo DR: Effective patient education. Rockville, MD: Aspen, 1985.
35. Whitman NJ, Graham BA, Gleit CJ, Boyd MD: Teaching in Nursing Practice: A Professional
model. Norwalk, CT: Appleton-Centry-Crofts, 1986.
44
36. Smith CE (ed): Patient education: Nurses in partnership with other health professionals.
Orlando: Grune & Stratton, 1987.
37. Redman B: The Process of Patient Teaching in Nursing (5th ed.). St. Louis: Mosby, 1985.
38. Baer CL: Principles of patient education. In Lancaster LE (ed.) Core Curriculum for
Nephrology Nursing, 2nd ed. (pp 17-28). Pitman, NJ: American Nephrology Nurses'
Association, 1990.
39. Peritoneal dialysis health history and nursing assessment. Columbia, MO: Dialysis Clinic,
Inc.
40. Ryan L, Prowant B, Schmidt L et al.: Home visit effectiveness for peritoneal dialysis patients.
ANNA Journal 20: 333-6, 1993.
41. Thaler MK, Sasak C: Cooperative nursing care for patients using peritoneal dialysis. ANNA
Journal 15: 237-240, 1988.
42. Twardowski ZJ, Nolph KD, Khanna R, Prowant BF, Ryan LP, Moore HL, Nielsen MP:
Peritoneal equilibration test. Perit Dial Bull, 7: 138-147, 1987.
43. Twardowski ZJ, Khanna R., Nolph KD: Peritoneal dialysis modifications to avoid CAPD
drop-out. In Khanna R, Nolph KD, Prowant B, Twardowski ZJ, Oreopoulos DG (eds)
Advances in Continuous Ambulatory Peritoneal Dialysis/I987 (pp 171-178). Toronoto:
Peritoneal Dialysis Bulletin, Inc., 1987.
44. Twardowski ZI: PET - A simpler approach for determining prescriptions for adequate
dialysis therapy. In Khanna R, Nolph KD, Prowant BF, Twardowski ZJ, Oreopoulos DG
(eds) Advances in peritoneal dialysis, vol 6 (pp 186-191). Toronto: Peritoneal Dialysis
Bulletin, Inc., 1990.
45. Diaz-Buxo JA: Peritoneal permeability in selecting peritoneal dialysis modalities.
Perspectives in Peritoneal Dialysis 5(2): 6-10, 1988.
46. Boen ST, Haagsma-Schouten WQG, Birnie RJ: Longterm peritoneal dialysis and a
peritoneal dialysis index. Dialysis & Transplantation 7: 377-380, 1978.
47. Teehan BP, Schleifer CR, Sigler MH, Gilgore GS: A quantitative approach to the CAPD
prescription. Perit Dial Bull 5: 152-156, 1985.
48. Lysaght MJ, Pollock CA, Hallet MD, Ibels LS, Farrell PC: The relevance of urea kinetic
modeling to CAPD. Trans Am Soc Artif Intern Organs 35: 784-790, 1989.
49. Keshaviah PR, Nolph KD, Van Stone J: The peak concentration hypothesis: A urea kinetic
approach to comparing the adequacy of continuous ambulatory peritoneal dialysis (CAPD)
and hemodialysis. Perit DialInt 9: 257-260, 1989.
50. Gotch FA: Application of urea kinetic modeling to adequacy of CAPD therapy. In Khanna
R, Nolph KD, Prowant B, Twardowski ZJ, Oreopoulos DG (eds) Advances in continuous
ambulatory peritoneal dialysis, vol 6 (pp 178-180). Toronto: Peritoneal Dialysis Bulletin,
Inc., 1990.
51. Brandes IC, Piering WF, Beres JA: A method to assess efficacy of CAPD: Preliminary
results. In Khanna R, Nolph KD, Prowant B, Twardowski ZJ, Oreopoulos DG (eds) Advances
in Continuous Ambulatory Peritoneal Dialysis, vol 6 (pp 192-196). Toronto: Peritoneal
Dialysis Bulletin, Inc., 1990.
52. Teehan BP, Schleifer CR, Brown 1M, Sigler MJ, Raimondo J: Urea kinetic analysis and
clinical outcome on CAPD. A five year longitudinal study. In Khanna R, Nolph KD, Prowant
B, Twardowski ZJ, Oreopoulos DG (eds) Advances in Continuous Ambulatory Peritoneal
Dialysis, vol 6 (pp 181-185). Toronto: Peritoneal Dialysis Bulletin, Inc., 1990.
53. Maiorca R, Cantaluppi A, Cancarini GC et al.: Prospective controlled trial of a Y
connector and disinfectant to prevent peritonitis in continuous ambulatory peritoneal dialysis.
Lancet ii: 642-644, 1983.
54. Churchill DN, Taylor DW, Vas SI, Oreopoulos DG et al.: Peritonitis in continuous
ambulatory peritoneal dialysis: A multi-centre randomized clinical trial comparing the Y
connector disinfectant system to standard systems. Perit DialInt 9: 159-163, 1989.
55. Zappacosta AR, Perras ST: Reduction of CAPD peritonitis rate by ultraviolet light with
dialysate exchange assist device. Dial Transplantation 9: 483, 1988.
45
56. Port FK, Held PJ, Nolph KD, Turenne MN, Wolfe RA: Risk of peritonitis and technique
Failure by CAPD connection technique: A national study. Kidney 1m 42: 967-74, 1992.
57. Baxter Healthcare Corporation. The Best Demonstrated Practices Program: Peritonitis
Management and Antibiotic Therapy Practices. Deerfield, IL: Author, 1987.
58. Rubin J, Rogers WA, Taylor HM, Everett ED, Prowant BF, Fruto LV, Nolph, KD: Peritonitis
during continuous ambulatory peritoneal dialysis. Annals of Internal Medicine 92: 7-13,
1980.
59. Prowant B, Nolph K, Ryan L, Twardowski Z, Khanna R: Peritonitis in continuous
ambulatory peritoneal dialysis: Analysis of an 8-year experience. Nephron 43: 105-109,
1986.
60. Vas SI: Microbiologic aspects of chronic ambulatory peritoneal dialysis. Kidney International
23: 83-92, 1983.
61. Keane WF, Everett ED, Fine RN, Golper TA, Vas SI, Peterson PK: CAPD related peritonitis
management and antibiotic therapy recommendations. Perit Dial Bull 7: 55-68, 1987.
62. Vas VI: Examination of fluids during peritoneal dialysis. Perit Dial Bull 1(1): 2-3, 1980.
63. Knight KR, Polak A: Laboratory diagnosis and oral treatment of CAPD peritonitis. Lancet
ii: 1301-1304, 1982.
64. Bint AJ, Finch RG, Gokal R, Goldsmith HI, Junor B, Oliver D: Diagnosis and manage-
ment of peritonitis in continuous ambulatory peritoneal dialysis. Lancet 1: 845-848, 1987.
65. Williams P, Pantalony D, Vas SI, Khanna R, Oreopoulos DG: The value of dialysate cell
count in the diagnosis of peritonitis in patients on continuous ambulatory peritoneal dialysis.
Perit Dial Bull 1: 59-62, 1981.
66. Keane WF, Everett ED, Fine RN, Golper TA, Vas S, Peterson PK, Gokal R, Matzke GR:
Continuous ambulatory peritoneal dialysis (CAPD) peritonitis treatment recommendations:
1989 update. Perit Dial Int 247-256, 1989.
67. Vas SI: Peritonitis. In Nolph KD (ed) Peritoneal Dialysis, 3rd ed. (pp 261-288). Dordrecht:
Kluwer, 1989.
68. Horton MW, Deeter RG, Sherman RA: Treatment of peritonitis in patients undergoing
continuous ambulatory peritoneal dialysis. Clinical Pharmacy 9: 102-118, 1990.
69. Twardowski ZJ, Nolph KD, Khanna R, Ryan LP, Prowant BF: Peritonitis management in
the CAPD program at the University of Missouri-Columbia: November 1983-0ctober 1984.
In Khanna R, Nolph KD, Prowant B, Twardowski ZJ, Oreopoulos DG (eds) Advances in
Continuous Ambulatory Peritoneal Dialysis (pp 61-65), Toronto: Peritoneal Dialysis
Bulletin, Inc., 1985.
70. D' Apice AJF, Atkins RC: Analysis of peritoneal dialysis data. In Atkins RC, Thomson
NM, Farrell PC (eds) Peritoneal Dialysis (pp 440-444). Edinburgh: Churchill Livingstone,
1981.
71. Corey P: An approach to the statistical analysis of peritonitis data from patients on CAPD.
Perit Dial Bull 1(6): S29-S32, 1981.
72. Pierratos A, Amair P, Corey P, Vas SI, Khanna R, Oreopoulos DG: Statistical analysis of
the incidence of peritonitis on continuous ambulatory peritoneal dialysis. Perit Dial Bull
2: 32-36, 1982.
73. Prowant B: Lack of research by non-physician health care professionals. Perit Dial Bull
8: 11-13, 1988.
74. Hammes BJ, Colvin E: If I Only Knew: A Patient Education Program on Advance Directives.
LaCrosse, WI: Lutheran Hospital Regional Dialysis Center.
75. Colvin E, Hammes BJ: If I Only Knew: A Patient Education Program on Advance Directives.
ANNA Journal 18: 557-60, 1991.
CHAPTER 4
Introduction
The multi-center system poses special challenges and offers special oppor-
tunities for dialysis quality assurance and improvement. Within the frame of
reference defined by the generic Continuous Quality Improvement (CQI)
model, for example, the multi-center system's large patient population and
breadth of clinical experience provide opportunities which are not available
to the single facility for developing and monitoring empirical indicators of
quality. At the same time, however, the implementation and operation of a
quality assurance system cannot be imposed by fiat upon a large and decen-
tralized system of facilities, each with its own physician staff responsible
for patient care. Relatively sophisticated communications and incentives are
required to achieve quality assurance goals in such a multi-center system.
The discussion here is based upon our experience with the ongoing
development and implementation of National Medical Care's (NMC's) dialysis
quality assurance system. As the largest U.S. provider of chronic renal dialysis,
with approximately 385 facilities caring for nearly 30,000 chronic dialysis
patients at the end of 1991, NMC has been able to bring a unique set of
resources to bear upon the problems of dialysis quality monitoring. Most
specifically, the company has maintained a number of closely related and
integrated patient and facility data systems designed to allow description and
analysis of patient characteristics, clinical laboratory values, treatment char-
acteristics, and a variety of outcome measures including morbidity and
mortality.
Periodic routine reports and special analyses derived from these data sources
are provided to practitioners at each individual facility in the NMC system.
Medical Directors and facility clinical staff thereby have the opportunity to
compare their own facility's performance on a variety of measures to that of
the system as a whole. These reports have proven to be powerful tools in
revealing clinical trends not obvious at the level of the individual facility,
provoking review at each facility of the quality of care being delivered, and
evoking facility-specific adaptations to improve outcomes.
The notion of CQI has so rapidly and completely been adopted as quality
assurance gospel in health care that it threatens, in its ubiquity and its gen-
erality, to lose its meaning. It is necessary therefore, in order to assure some
substance in our use of the term, to review briefly what we take to be the
key components of a CQI system. Our presentation is largely consistent with
that contained in the 1990 Institute of Medicine report "Medicare: A Strategy
for Quality Assurance" [1].
CQI, first and foremost, is predicated upon incremental progress toward
improved quality with no prescribed targets or limits. The process does not
end; there is no arbitrary division between acceptable and unacceptable states;
there is simply a continuing search for ways to make outcomes better.
A primary focus on outcomes and patient needs, rather than processes, is
a second critical characteristic of CQI. It is not that processes are unimpor-
tant. Undeniably, understanding of and control over the processes which
comprise the dialysis treatment are critical aspects of dialysis quality control.
An effective quality assurance system must promote such understanding and
control. But processes are not important for themselves. They become
important only because (or when) they are empirically linked to outcomes they
help to create and control. CQI encourages caregivers to review outcomes
and to make those outcomes the basis for a search for the adjustments in the
treatment process which will demonstrably yield clinical improvements.
Third, attention to systems of care rather than individual cases or providers
is equally important. CQI recognizes that results will always manifest them-
selves as statistical distributions, and that from a systemic perspective particular
outliers are far less interesting or important than broad measures of central
tendency and dispersion. If nothing else, focus on outliers is an inefficient
investment of scarce quality assurance resources. For an individual facility,
therefore, CQI would lead to statistical analyses of selected outcomes and their
determinants rather than review of a small sample of case records; for a system
of facilities, these same analyses can be extended to include summary measures
of the overall performance of each facility in the system.
Fourth, CQI focuses on the continuous feedback of new knowledge from
clinical practice as a source for clinical adaptation by practitioners. This
49
contrasts to some other quality assurance models, which impose relatively static
standards of care based upon professional consensus, conventional wisdom,
or specific published doctrine. In CQI, a cycle is established which con-
stantly feeds itself: measured outcomes ~ exploratory dialogue ~ clinical
adaptations ~ improved outcomes, and so on.
Fifth and finally, CQI depends for its power and success on internal control
mechanisms driven by professional norms rather than external regulatory
monitoring. It assumes that caregivers have a commitment to providing care
of the highest possible quality, and works to empower them to improve clinical
standards by providing the best and most recent possible information upon
which to base clinical policies and decisions. It is always and essentially a
collegial rather than an adversarial process.
Externally imposed norms and the power to take punitive action neces-
sarily require defined and stable standards against which to measure specific
practitioners or facilities. They also virtually guarantee that practitioners and
facilities will be forced into a defensive rather than a cooperative stance in
dealing with the quality assurance authority. But those are qualities which
are antithetical to the nature of the CQI process. This truth becomes extra-
ordinarily important when considering public policy interventions to improve
clinical quality. It may be that the public regulatory enterprise necessarily
involves elements which are incompatible with CQI.
We need only look at the government's efforts to improve hospital care
by publishing hospital-specific annual risk-adjusted mortality data to see this
problem in graphic form. Publication of annual hospital mortality rankings
has lead to great public misunderstanding and outcry. As a result, hospital
administrators and practitioners have been forced to expend enormous effort
defending their institutions, through attacks on the government's general
competence and/or specific methodological shortcomings. The need to engage
in this public defense creates resistance to change rather than open partici-
pation in the kind of process of inquiry and adjustment which fosters real
quality improvements. For the vast majority of hospitals (and other types of
providers), private distribution of comparative mortality data would have
more beneficial effects on quality of care than the publication of rankings
by institution.
CQI, which is at heart a set of goals and a set of supporting principles, does
not by itself provide the structure for an effective dialysis quality assurance
program. There is a further need for an operating model which can order the
variables encountered in the exploration of dialysis quality, provide guidance
in the development of the statistical analyses and information sharing processes
required, and help practitioners and facility managers to control those vari-
ables. For NMC, the general operating model of choice has been the one
50
defined for other applications by the Federal Food and Drug Administration's
Current Good Manufacturing Practices (CGMPs) [2], as concretized and
expressed in the form of the Quality Triangle.
For dialysis facilities, accustomed to dealing with Medicare Conditions
of Coverage as interpreted by state survey teams and possibly with State
licensure requirements, CGMPs have sometimes been viewed as extra-
ordinarily arduous and difficult; in 1986, for example, a short-lived proposal
that dialyzer reuse be made subject to CGMP-like requirements was greeted
by nothing short of horror. And it is certainly true that, in the hands of an
unrestrained and largely untrained external agency, CGMPs might in fact
become the excuse for unreasonable regulatory demands. But this is also true
of the chronic renal facility Conditions of Coverage, and some dialysis
facilities would argue that they have experienced unreasonable treatment at
the hands of state surveyors interpreting those familiar standards. The fact is
that CGMPs provide a proven model for ongoing product and process spec-
ification, monitoring and control which is extraordinarily well adapted to
quality assurance in dialysis facilities. The "production" of a dialysis treatment
and the generic manufacturing process addressed by CGMPs have many points
of contact, easily recognized and understood by persons with experience in
both settings.
CGMPs are predicated upon a rigorous systems approach which maintains
that carefully controlled "inputs" and standardized "processes" will yield
predictable and constant "outcomes". They require:
1. Clear specification of the character and quality of inputs to the produc-
tion process;
2. Clear description of the standardized process;
3. The definition and implementation of control systems to monitor inputs and
processes;
4. Carefully controlled documentation of all of these elements to allow for
internal and/or independent audit of process integrity;
5. A demonstration that the finished product meets predetermined performance
standards, either through universal testing or sampling as deemed appro-
priate; and
6. Tracking and investigation of production failures (represented for devices
by failure to pass release testing, customer complaints about released
devices, and/or reported incidents of failure in use) in order to under-
stand, isolate the causes of, and eliminate quality problems.
Each of these enumerated aspects of CGMP controls has a clear analogue
in dialysis. For the dialysis facility, inputs include the patient prescription,
the supplies and equipment used to provide a treatment, and the staff resources
and training available in the facility. Procedural manuals and related rules
define the constant elements of the process of providing a treatment, and the
dialysis flow sheet and patient medical record provide the documentation by
which to assess process control and such immediate physical parameters of
treatment as weight loss. Most facilities exercise careful control over incoming
51
Clinical
Quality
Morbidity
Clinical Failures
Clinical/lab
~~ Process!
Procedure
~-:.~;".".~ .. ~
Lf~~~~~~~~~~~ffi~~~~~~~~~~varlances
Exhibit 1.
Thus, the process or procedure variances found at the base of the triangle
are expected to account for the most numerous adverse observations; they
are antecedent to and by themselves less important than the observations or
events at higher levels, such as clinical laboratory variances, clinical failures,
hospitalizations, or deaths. Deaths and hospitalizations (the two highest levels
of the triangle) are, for the most part, caused by clinical failures, which are
themselves the product of clinical and/or process variances. The specific
variables included in each level of the triangle in Exhibit 1 are exemplary.
Others may readily be added or substituted.
Tracking of variables at the lower levels of the triangle is critical for two
reasons. First, being more numerous, they will be easier to observe; second,
being antecedent and of less severity, their reduction will have a profound
impact on the observed rate of more serious failures tracked at higher levels.
In other words, for example, early identification of a problem such as fistula
recirculation (perhaps through monitoring of the urea reduction ratio) will
prevent underdialysis of a number of patients, development of clinical com-
plications in some subset of those patients, and the need to hospitalize some
further subset.
As is apparent from the discussion above, the linkages between variables
in the lower levels of the triangle and variables in higher levels determines
the utility of the triangle for quality enhancement activities. While the linkages
may in some cases seem obvious, it is in fact the case that most have not
been empirically demonstrated. To a greater degree than is often admitted,
dialysis remains a treatment for which theory is well understood, practice is
often disassociated from that theory, and the links between practice and theory
are only recently being subjected to empirical review.
Intuitively, it seems obvious that a high rate of missed treatments will
eventually translate into some patients receiving inadequate therapy; that failure
to follow (within reasonable limits) a proper dietary regimen will result in
nutritional problems; and that a laboratory value out of the normal range for
a large group of patients will lead in time to predictable clinical complications.
But the nature and importance of those linkages are not well understood, .and
there is little empirical data available which can help us to define critical levels
for variables and/or the degree of association between different antecedent
variables and their probabilistic sequelae.
The National Cooperative Dialysis Study (NCDS) provided a carefully
controlled clinical trial which demonstrated the integrity of important elements
in the logic of the triangle, such as (1) the link between dialysis mechanics
(represented by treatment length) and adequacy of treatment (reflected in target
BUN value), and (2) the link between the adequacy of treatment (BUN control,
subject to adequate nutrition) and therapy failure (hospitalization, death, or
withdrawal from the trial for clinical reasons) [3]. But the degree of therapy
control exercised in that clinical trial cannot practically be duplicated in the
ongoing operations of a normal dialysis facility, and the specific lessons of
the NCDS for that setting are therefore not wholly clear. Dialysis quality
54
The flow of information to and from participating facilities is the most critical
operational element of a CQI program in the multi-facility system. For one
thing, as noted above, full involvement of the facility in a two-way flow of
55
o
~3
-
a:
o
W2
'- --- -. -.- ,--------------------- --- -----------.
..~-~............•.....~............•...........•...........•.... ~ ...................... .
o -.--.~
• ........................
_.-----, ::-.-.-.-•. ..................
,!--". • __ ._.__ ._._----_
,..----------_._._._.
..-:
• ..
01 02 03 04 01 02 03 04 01 02 03 04 01 02 03 04 01 02 03 04
1- 1986 -i 1- 1987 -I I- 1988 -I 1- 1989 -I 1- 1990 -i
Exhibit 2.
4 ...................•................•...•..................•..............•....................................•..................................
o
~3
-
a: •
...........................................................................•.....................................................................
W 2
01 02 03 04 01 02 03 04 01 02 03 04 01 02 03 04 01 02 03 04
I- 1986 - I I - 1987 --! I- 1988 --! I- 1989 --! I- 1990 - I
• =Actual OlE Rallo QUARTER
- - - = Expected OlE Ratio Umlls
- - =Quarter Moving Average
Exhibit 3.
58
mortality (and morbidity) analyses is done for two reasons. First, it is likely
that there are certain irreducible risks associated with some patient char-
acteristics, and failure to account for those risks would skew some facilities'
results. Second, risk-adjustment has become so commonplace that practi-
tioners would likely be tempted to take refuge behind their "sicker patient
group" to explain relatively poor performance if we reported non-adjusted
mortality. It is important to note, however, that risks are not always what
they seem at first look, and it may be that the assumption that a subgroup
of patients (e.g. diabetics) suffers from elevated risk may in fact obscure
the opportunity to realize significant improvements in treatment [11].
3. Analytical or exploratory reports on the relationships among clinical vari-
ables captured in the various data systems. These reports, which are
disseminated throughout the system at irregular intervals as analyses are
performed and refined, have proven to be the most important element in
raising questions about common clinical practice and in suggesting inno-
vations. Exhibit 4, for example, displays findings on the relationship
between URR and patient mortality which have caused reexamination and
upward revision of urea removal targets in many facilities. Findings such
as these, when reported to facilities, provide the challenges to which we
hope practitioners will respond.
~ 1.6
a:
~ 1.4
:;::
~
a: Q)
1.2
1.0
0.8
Exhibit 4.
59
These reports, by design, are never directive. They explore the clinical data,
report the statistical findings, and speculate about the implications. This stance
is critically important. While NMC does as a matter of policy require that
quality assurance system reports be reviewed at regular QIA team meetings,
and the minutes of those meetings are audited for compliance, that is the extent
of the control built into our system. We do not prescribe specific goals or
standards for facilities, or demand that practices be changed in any partic-
ular way in response to these reports. We cannot and do not desire to dictate
physicians' clinical behavior, or to pose as more expert in clinical decision
making than the audience. Efforts to do so would be counterproductive to
the interactive process that the CQI model demands. We do desire to provoke
the audience's interest and to provide stimuli to their critical and evaluative
processes. If physicians are motivated by these reports to explore the
applicability of the findings to their practice, or to investigate for themselves
whether an indicated relationship seems to hold for their patients, the system
is working, and that work will be reflected over time in changes in clinical
behavior and ultimately in the reported data.
References
1. Lohr KN (ed): Medicare: A Strategy for Quality Assurance. Washington, DC: National
Academy Press, 1990.
2. Code of Federal Regulations, Title 21, Part 820.
3. Lowrie EG, Laird NM (eds): The national cooperative dialysis study. Kidney Int. Supplement
No. 13, April 1983.
4. Lowrie EG, Lew NL: Death risk in hemodialysis patients: The predictive value of commonly
measured variables and an evaluation of death rate differences between facilities. Am J
Kid Dis 15(5): 458-482, 1990.
5. Chazan lA, Lew NL, Lowrie EG: Increased serum aluminum: An independent risk factor
for patients undergoing long-term hemodialysis. Arch Intern Med 151: 319-322, 1991.
6. Lowrie EG, Lew NL: The urea reduction ration (URR). A simple method for evaluating
hemodialysis treatment. Contemporary Dialysis and Nephrology February: 11-20, 1991.
7. Lowrie EG, Lew NL, Huang WH: Race and diabetes as death risk predictors in hemodial-
ysis patients. Kidney Int 42 (Suppl 38): 5-22-5-31, 1992.
8. Berger EE, Lowrie EG: Mortality and the length of dialysis. JAMA 265: 909-910, 1991.
9. Held PI, Pauly MV, Diamond L: Survival analysis of patients undergoing dialysis. JAMA.
257: 645-650, 1987.
10. Held PI, Levin NW, Brovbjerg RR, Pauly MV, Diamond LH: Mortality and duration of
dialysis treatment. JAMA 265: 871-875, 1991.
11. Lowrie EG, Lew NL: Commonly measured laboratory variables in hemodialysis patients:
Relationships among them and to death risk. Seminars in Nephrology, 12(3): 276-283, 1992.
12. McClellan W: Applying the 10M perspective to the networks' approach to QA. Nephrology
News and Issues 6(1): 44-46, 1992.
CHAPTER 5
CHRISTOPHER R. BLAGG
Introduction
Home hemodialysis was first used in 1963 in the United States and Great
Britain to reduce the cost of long-term dialysis for chronic renal failure so
that the limited resources available could be used to treat more patients
[1, 2]. Shortly thereafter, home intermittent peritoneal dialysis (IPD) was
developed [3]. In recent years, continuous ambulatory peritoneal dialysis
(CAPD) [4] and various forms of continuous cycling peritoneal dialysis
(CCPD) [5] have generally replaced IPD.
With establishment of the Medicare End-Stage Renal Disease (ESRD)
Program in 1973, almost all U.S. residents who develop ESRD are eligible
for financial support for treatment by dialysis or kidney transplantation. As
of 1990, of 129,800 dialysis patients covered by the Medicare ESRD Program,
109,056 (84.0 percent) were on hemodialysis and 20,744 were on some form
of peritoneal dialysis [6]. Recent years have seen a significant increase in
the number of patients on peritoneal dialysis in the home, so that by 1990 these
accounted for 16 percent of all dialysis patients, and most of these (83.0
percent) were on CAPD. Simultaneously, the number of patients on home
hemodialysis has declined, so that in 1990 this was 2,483. This is 2.3 percent
of all hemodialysis patients, and 1.9 percent of all dialysis patients. Despite
the increase in the number of patients treated by CAPD and CCPD, from
12,189 in 1985 to 19,967 in 1990, at the same time the percentage of all patients
dialyzing at home declined from 19.3 to 17.4.
In 1990, of 22,640 patients dialyzing at home, 2,483 were home hemo-
dialysis patients (11.0 percent of all home dialysis patients), 16,969 were on
CAPD (75.0 percent), 2,998 were on CCPD (13.2 percent), and 190 were listed
as on "home peritoneal" dialysis (0.8 percent). Between 1985 and 1990 the
average annual increase in the number of patients treated by CAPD was 8.6
percent and for CCPD was 25.8 percent. This compared with an annual increase
of 8.9 percent in the total dialysis population, and a decline of 9.0 percent
in the home hemodialysis population. The average annual increase for all
modalities of in-center dialysis was 9.4 percent, while that for all modalities
of home dialysis was 6.7 percent.
Recent years have seen concern that mortality of United States hemo-
dialysis patients is significantly higher than that of patients in Europe, Japan,
and other developed countries [7, 8]. While this may in part reflect the sicker
patient population now being treated, it is also clearly related to the widespread
use of inadequate dialysis in the United States. For example, in 1986-1987,
hemodialysis patients reported to the European Dialysis and Transplant
Association (EDTA) registry on average received 30% more dialysis than
did patients in the United States [9]. This resulted from both shorter dialysis
times and the use of dialyzers with a lower surface area in the United States.
Because of this concern and growing interest in the evaluation and moni-
toring of health care, the Health Care Financing Administration (HCFA), the
Renal Physicians Association and other organizations and agencies have
become involved in assessing quality of care in dialysis and transplant patients.
So far there has been less interest in how this can be applied to patients
dialyzing at home. This chapter will discuss some of these issues.
review screens for use by staff of the ESRD networks in reviewing dialysis
facilities. These are described below. The screens are likely to be modified
with further experience, but this approach undoubtedly will continue to be used
to assess quality of care in dialysis patients.
Federal regulations (42 CFR 405.2138) require the governing body of all
dialysis facilities to adopt written policies on the rights and responsibilities
of patients, and procedures for carrying these out. These must be made
available to patients, guardians, next of kin, sponsoring agencies, representa-
tive payees and the public. Facility staff must be trained and involved in
execution of these policies and procedures.
a. Informed patients. Patients must be fully informed:
• Of their rights and responsibilities and of all rules and regulations gov-
erning patient conduct and patient responsibilities.
• Of services available to them and related charges, including charges
for services not covered by Medicare. These services must include the
option to choose treatment by kidney transplantation and any of the
modalities of dialysis, including home dialysis.
• Of their medical condition by a physician unless medically contra-
indicated as documented in the patient record.
• Regarding the facility's reuse of dialysis supplies, including dialyzers.
• Regarding their suitability for transplantation and home dialysis.
b. Participation in planning. Patients must have the opportunity to partici-
pate in planning their own treatment, including selection of the modality
of care, and to refuse to participate in experimental research. Patients will
be transferred or discharged only for medical reasons, for their own welfare
or that of other patients, or for nonpayment of fees (except as prohibited
by the Medicare program). Patients must be given advance notice to ensure
orderly transfer or discharge.
c. Respect and dignity. Patients must be treated with consideration, respect
and full recognition of their individuality and personal needs, including the
need for privacy during treatment and provision of translation services if
required.
67
Patient care
As for age, home hemodialysis can be used for patients of all ages except
infants provided a family member or other assistant is available. In children,
adolescents and patients likely to be transplanted early, CAPD and CCPD
are particularly useful. At the other age extreme, in Seattle almost 50%
of home hemodialysis patients are aged 60 or older, and there have been
many successful patients in their 70s and 80s. Apart from a greater like-
lihood of medical complications, the most common problem for elderly
home hemodialysis and CCPD patients is lack of a suitable family member
to assist. This can be overcome by providing a trained dialysis assistant.
Patients should not do home hemodialysis on their own, although this
has been done successfully [19]. In contrast, CAPD should always be
done by the patient, making this particularly useful for the single patient.
Ability to learn to do safe home dialysis appears to correlate with the
patient's motivation rather than with their intelligence. Illiterate patients
or those who do not understand English will have more difficulties, but
with illustrations, videotapes and suitable translation such patients can be
trained successfully.
Another important psychological aspect of home dialysis is the degree
of anxiety of the patient and family. Some patients prefer CAPD or CCPD
because of their relative simplicity, but for all modalities of home dialysis
anxiety usually can be relieved by an appropriate approach before and at
the time the patient starts dialysis. Whenever possible the patient should
be introduced to the home dialysis options before they start treatment,
and have the opportunity to meet with and visit successful home dialysis
patients.
Federal regulations require that a dialysis facility's patient care policies specify
the functions carried out by self-dialysis patients with respect to contamina-
tion prevention (42 CFR 405.2140(c)). Also, in the unlikely event dialyzer
reuse is done in the home, regulations require maintenance of records that
can be used to determine whether the procedures for rinsing, cleaning, disin-
fection, preparation and storage of reused items conform to Federal require-
ments (42 CFR 405.2150). This regulation details the conditions for reuse of
hemodialyzers and other dialysis supplies based on recommendations of the
Association for the Advancement of Medical Instrumentation (AAMI) [20].
a. Education. Safety management and infection control for home dialysis
patients require a program designed to educate staff, patients, family
members and, where applicable, dialysis assistants, in safety measures
necessary to minimize the hazards related to home dialysis. The home
dialysis training program should include written and verbal instruction in
basic home safety; safe and appropriate use of dialysis equipment; storage,
handling, delivery and access to supplies and drugs, including needles
70
Quality assurance
Federal regulations describe the qualifications and role of the chief execu-
tive officer (42) CFR 405.2135(c)), and the physician director (42 CFR
405.2161) of a dialysis facility. The former is responsible to the governing
body for operation of the facility, for assuring compliance with applicable laws
and regulations, and for taking action on reports and recommendations of
authorized planning, regulatory and inspection agencies. He or she is also
responsible for carrying out policies and procedures established by the
governing body, including those addressing delivery of care and services and
financial management.
The physician director, who may also be the chief executive, is respon-
sible for the quality of services provided, including home dialysis services,
their availability, and appropriate staffing. He or she is also responsible for
participation in selection of a suitable modality of treatment for all patients,
assuring adequate training of nurses and technicians in dialysis techniques,
assuring adequate monitoring of patients and the dialysis process, assuring
development, availability and implementation of a patient care policy and
procedures manual, and ensuring that teaching materials are continuously
available for patient use whenever self-dialysis training is offered.
Lines of responsibility and accountability in the facility must be clearly
established, including the relationship of the home dialysis program to other
departments and services, the governing body, and, where applicable, the
medical staff. Facility staff must participate in orientation, in-service training,
and continuing education programs.
When any home dialysis-related services are provided by an outside
organization or supplier, there must be a written agreement defining the nature
and scope of the services.
Governing body
care is available to all patients for whom the facility maintains patient care
plans.
Pharmaceutical services
Federal regulation (42 CFR 40S.2163(c» require that a dialysis facility have
either a licensed social worker with a masters degree in social work, or an
80
individual with two years experience as a social worker, one of these years
being in a dialysis or transplant unit, and who has a consultative relation-
ship with a social worker with a masters degree. Such a person, with the support
of the training staff and the patient's physician, can provide support for personal
care and the activities of daily living. For example, with erythropoietin and
coordinated social work, vocational counseling and exercise training, many
patients can be helped improve their daily lives. In those trained early for home
dialysis there is often opportunity for the patient to continue to work.
Unfortunately, some facilities provide much less support than is desirable
and few patients are referred for vocational counseling.
While HCFA sponsored a conference on rehabilitation of ESRD patients
some years ago, this did not lead to any actions. It is to be hoped that current
and proposed efforts by the Institute of Medicine and the American Association
of Kidney Patients will be more successful. Clearly, home dialysis has an
important role to play in providing the best quality of life and greatest
opportunity for rehabilitation for dialysis patients.
Equipment management
Conclusions
Home dialysis, whether home hemodialysis or one of the modalities of
peritoneal dialysis, is an important option that should be available to all suitable
patients. Unfortunately, this is not always the case. However, it is just as
important there be active QA assessments of the care provided these patients
as for patients dialyzing in a dialysis unit. The problems are increased by
the logistical difficulties associated with dialysis remote from the center.
Nevertheless, the results of treatment with home dialysis are reported to be
as good or better than with center dialysis. Careful QA should help to maintain
this record. This is important as we move into what ReIman has called the
"Era of Assessment and Accountability" in medical care [30].
References
1. Merrill JP, Schupak E, Cameron E, Hampers CL: Hemodialysis in the home. lAMA 190:
468-470, 1964.
2. Baillod RA, Comty C, Ilahi M, Konotey-Ahu1u FID, Sevitt L, Shaldon S: Overnight
haemodialysis in the home. Proc Eur Dial Transpl Assoc 2: 99-103, 1965.
3. Tenckhoff H, Shilipetar G, van Paasschen WJ, Swanson E: A home peritoneal dialysate
delivery system. Trans ASAIO IS: 103-107, 1969.
4. Popovich RP, Moncrieff JW, Dechard JB, Bomar JB, Pyle WK: The definition of a novel
portable/wearable equilibrium peritoneal dialysis technique. Abstracts, ASAIO 5: 65, 1976.
5. Diaz-Buxo JA: Continuous cyclic peritoneal dialysis. In Nolph KD (ed) Peritoneal Dialysis
3rd ed. (pp 169-173). Dordrecht, The Netherlands: Kluwer Academic Publishers, 1989.
6. Health Care Financing: Research Report: End Stage Renal Disease, 1990. Department of
Health and Human Services, Health Care Financing Administration, Bureau of Data
Management and Strategy; Office of Research and Demonstration, Baltimore, MD, 1992,
HCFA Pub. No. 03328.
83
7. Hull AR, Parker TF: Introduction and summary: Proceedings from the morbidity,
mortality and prescription of dialysis symposium, Dallas, TX, September 15 to 17, 1989.
Am J Kidney Dis 15: 375-383, 1989.
8. Held PJ, Brunner F, Odaka M, Garcia JR, Port FK, Gaylin OS: Five-year survival for
end-stage renal disease patients in the United States, Europe, and Japan, 1982 to 1987.
Am J Kidney Dis 15: 451-457, 1990.
9. Held PJ, Blagg CR, Liska OW, Port FK, Hakim R, Levin N: The dose of hemodialysis
according to dialysis prescription in Europe and the U.S., 1986-87. Kidney Int 42: Suppl
38, 516-521, 1992.
10. Brook RH, Kosecoff JB: Commentary: competition and quality. Health Affairs, 7: 150-161,
1988.
11. Donabedian A: Evaluating the quality of medical care. Millbank Memorial Fund Quarterly
44 (Suppl): 166-206, 1966.
12. Hopkins A: Measuring the Quality of Care. Royal College of Physicians of London, 1990.
13. Lohr K, Brook R: Quality assurance in medicine. American Behavioral Scientist 27, no 5,
1984.
14. The 1991 Joint Commission. AMHC Accreditation Manual for Home Care. Joint
Commission on Accreditation of Healthcare Organizations, Oakbrook Terrace, IL, 1991.
15. Code of Federal Regulations: Public Health: 42, Parts 400-429. Revised as of October 1,
1990, pp 137-151.
16. 1992 National Listing of Medicare Providers Furnishing Kidney Dialysis and Transplant
Services. Government Printing Office, Washington DC, 1992.
17. Roberts CM: Home dialysis in the haemophilic paraplegic patient. Proc Eur Dial Transpl
Nurses Assoc 7: 62-66, 1979.
18. Roberts CM, Davis B, Pavitt Let al.: Home dialysis training in a blind patient and a deaf
patient. Proc Eur Dial Transpl Nurses Assoc 1: 41, 1973.
19. Baillod RA, Moorhead JF: Review of ten years' home dialysis. Proc Eur Dial Transpl Assoc
11: 68-75, 1974.
20. AAMI Recommended Practice: Reuse of Hemodialyzers. Association for the Advancement
of Medical Instrumentation, Arlington VA, 1986.
21. Eschbach JW, Seymour M, Potts A, Clark M, Blagg CR: A hemodialysis orientation unit.
Nephron 33: 106-110, 1983.
22. Stinson GW, Clark MF, Sawyer TK, Blagg CR: Home hemodialysis training in three
weeks. Trans ASAIO 18: 66-69, 1972.
23. Blagg CR, Liedtke RJ, Batjer JD, Racoosin B, Sawyer TK, Wict MJ, Lawson L, Wilkens
K: Serum albumin concentration-related HCFA quality assurance criterion is method-depen-
dent: revision is necessary. Am J Kidney Dis 21: 138-144, 1993.
24. Sawyer TK, Blagg CR: Home preparation and installation for home hemodialysis. In
Nissenson AR, Fine RN (eds) Dialysis Therapy, 2nd ed, Philadelphia: Hanley & Belfus,
Inc., pp. 52-54, 1993.
25. Held PJ, Garcia JR, Blagg CR, Wolfe RA, Eggers PW, Port FK, Fitzsimmons SC: Trends
in mortality rates among dialysis and transplant patients in the U.S. Abstracts ASN, 51a,
1989,
26. Vollmer WM, Wahl P, Blagg CR: Survival with dialysis and transplantation in patients
with end-stage renal disease. N Engl J Med 308: 1553-1558, 1983.
27. Burton PR, Walls J: Selection-adjusted comparison of life-expectancy of patients on
continuous ambulatory peritoneal dialysis, haemodialysis, and renal transplantation. Lancet
1: 1115-1119, 1987.
28. Evans RW, Manninen DL, Garrison LP, Hart G, Blagg CR, Gutman RA, Hull AR, Lowrie
EG: The quality of life of patients with end-stage renal disease. N Engl J Med 312: 553-559,
1985.
29. Bremer BA, McCauley CR, Rona RM, Johnson JP: Quality of life in end-stage renal disease:
a reexamination. Am J Kidney Dis 13: 200-209, 1989.
30. Reiman AS: Assessment and accountability: the third revolution. N Engl J Med 319:
1220-1222, 1988.
CHAPTER 6
PRAKASH KESHAVIAH
Introduction
According to an Italian proverb "Aqua torbido non lava" i.e. "Dirty water
does not wash clean". This can be most aptly applied to the water used for
hemodialysis. If the water used for hemodialysis is "dirty", the blood of the
hemodialysis patient will not be "washed clean". The hemodialysis patient
is exposed to more water in one year than the normal population in 20 years.
Further, the hemodialysis patient's blood is exposed to this large quantity of
water across a thin non-selective membrane that has none of the "wisdom"
of the gastrointestinal tract of a normal individual. Also, the hemodialysis
patient's ability to excrete harmful toxins from such exposure is limited by
compromised renal function. The combination of large volume of exposure,
non-selective transport across the dialyzer membrane, and compromised renal
function make for a situation that is potentially hazardous to the health and
well-being of the patient on hemodialysis. Unless the water used for hemodial-
ysis is analyzed periodically and subjected to appropriate treatment processes
whose efficacy is monitored regularly, there may be serious risks to patient
well-being that have been well documented in the dialysis literature.
The risks and hazards associated with inadequately treated water are sum-
marized in Table I along with the lowest contaminant levels at which toxicity
has been observed. It should be noted that many of the contaminants asso-
ciated with toxicity in the hemodialysis setting are not regulated by the
Safe Drinking Water Act of 1974 [1]. Also, even for contaminants that are
regulated by the Safe Drinking Water Act, the levels at which toxicity is
observed in hemodialysis are lower than the safe limits established for drinking
water.
Aluminum. Aluminum may not be naturally occurring in the source water
but may be added to the water by the municipal supplier as alum (aluminum
Aluminum 0.06
Copper 0.49
Zinc 0.2
Calcium/Magnesium 88 (Ca++)
Sodium 300
Fluoride 1.0
Nitrate 21 (as N)
Sulphate 200
Chloramines 0.25
Microbiological Contaminants
dental caries. Even at the Safe Drinking Water level of 1 mg/l, prolonged
hemodialysis exposure may result in bone diseases such as osteomalacia and
osteoporosis [19, 20]. Accidental over-fluoridation of the water by the supplier
has been associated with adverse symptoms and one fatality [21].
Nitrate. Contamination of the water supply by fertilizers or excessive bac-
terial contaminants may result in high nitrate levels with consequences of
methemoglobinemia, cyanosis, hypotension, and nausea [22, 23]. Very high
nitrate levels may also cause hemolysis [23].
Sulphate. High sulphate levels in dialysis water have been associated with
nausea, vomiting, and metabolic acidosis [24].
Chloramines. Chloramines are commonly used bactericidal agents in munic-
ipal water treatment. Because they are powerful oxidants, inadequate removal
of chloramines may result in hemolysis, hemolytic anemia (characterized by
Heinz body formation) and methemoglobinemia [25-29].
Microbiological contaminants. There have been documented clusters of
pyrogen reactions associated with high levels of gram negative bacteria in
the water supply [30-33]. Symptoms noted include shaking chills, fever,
hypotension, headaches, myalgia, nausea and vomiting.
The various sources of water may be broadly classified into two major cate-
gories - ground waters and surface waters. In both cases, the water is derived
from the natural water cycle, water evaporating from oceans, rivers, and other
reservoirs into the atmosphere, the vapor condensing and returning to earth
as rain, snow, or hail. Some of this water flows on the earth's surface as streams
and rivers and collects in closed bodies of water such as lakes and ponds.
Precipitated water may also seep into the earth, collecting in underground
reservoirs and aquifers, being classified as ground water.
The quality of surface water depends on the location. In agricultural
locations, surface water may be contaminated by fertilizers and pesticides.
In urban locations, surface water may be contaminated by industrial wastes
and sewage. Unless surface water is appropriately treated for hemodialysis use,
it may pose a grave risk to the well-being of the hemodialysis patient.
Ground water percolates through various layers of soil and its quality may
depend on the geology of that area. Organic contaminants are usually removed
by the percolation of water through the various soil layers. However, inorganic
contaminant levels may increase. Ground water percolated through limestone
is usually hard containing high levels of calcium and magnesium compared
to water percolated through granite.
In 1974, landmark legislation was passed by Congress called the Safe
Drinking Water Act [1] which mandated the surveillance of drinking water
supplies by federal and state governments with the EPA assigned the respon-
sibility of enforcing these regulations. However, these regulations only apply
88
Inorganic
Arsenic 0.05
Barium 1.00
Cadmium 0.01
Chromium 0.05
Lead 0.05
Mercury 0.002
Nitrate (as N) 10.0
Selenium 0.01
Silver 0.05
Fluoride* 1.4-2.4
Organic
Endrin 0.0002
Lindane 0.004
Methoxychlor 0.1
Toxaphene 0.005
2,4-D 0.1
2, 4, 5 - TP (Sivex) 0.01
purposes, it is not, as previously noted, safe for dialysis because of the larger
volume of exposure, exposure to blood across a thin non-selective membrane,
and compromised renal function. More stringent limits than the EPA stan-
dards must be met, and these have been established by the Association for
the Advancement of Medical Instrumentation [36].
Filtration
Filters playa key role in water treatment with a variety of uses from removal
of large particulates to removal of bacteria and viruses. There are basically two
filtration mechanisms - depth filtration and surface filtration. Depth filtra-
tion is typically used for removal of coarse particulate material from water
by allowing the water to permeate the filter matrix and by trapping the
particulates mechanically in channels that are smaller than the particulate or
by adsorptive capture. Some depth filters (e.g. bed filters) consist of multiple
layers of filter media, each layer capable of retaining progressively smaller
particles. The largest particles are removed by the first layers and the smallest
in the final layers of the bed. Bed filters are suited for removing particles up
to 10 j.Lm in size. Surface filters exclude solute at the surface of the filter
primarily by size exclusion, although charge and hydrophobic and hydrophilic
interactions may also be involved. Surface filters are suited to removal of
both dissolved and suspended solutes.
Cartridge filters typically consist of a filter medium with a central drainage
core. Cartridge filters may employ both depth and surface filtration. Cartridges
92
of cellulosic materials and synthetic polymers are available with solid, mesh,
pleated, wound and woven configurations. The cartridge is contained in a
filter having both inlet and outlet ports. Cartridge filters capable of removing
particles as small as 1 11m are available.
Membrane filters employ surface filtration and are usually in the con-
figuration of a thin sheet of porous material with the filter being capable of
retaining solutes that are larger than the pores. Water passes freely through
the membrane pores. As the pores of the filter may not all be of the same
size, the size cutoff of the membrane is not sharply defined and is usually
quoted as the size at which 90% retention is achieved. Cellulosic and poly-
meric membranes are used in water treatment applications. Membrane filters
can be classified as microporous filters, ultrafiltration, and reverse osmosis
filters. Microporous filters are suited to removing particulates, bacteria, and
viruses up to 0.1 11m in size. Ultrafilters are capable of removing endotoxins,
proteins, and solutes as small as 0.001 11m (-200-300 daltons). Reverse
osmosis filters exclude larger organic solutes (-200 daltons) and ionic species
as small as 0.001 11m. Reverse osmosis filters are discussed in greater detail
below.
Filters may be operated in a dead end configuration or with cross flow as
shown in Figure 1. Depth filters are usually used in the dead end configura-
tion whereas membrane filters are commonly used with cross flow. In cross
flow, the tangential flow sweeps away solutes from the surface of the
membrane and reduces filter clogging.
Filters should be sized on the basis of the desired flow rate of water through
the filter. Filters have a finite capacity and need to be replaced or regener-
Permeate Scream
1
Permeate Stream
ated once this capacity is exceeded. The pressure drop across the filter is
monitored and used to determine when the filtration capacity has been reached.
If used beyond this capacity, pressure drops will increase, flow rates decrease,
and filter performance (solute retention) will deteriorate. Depth filters may
be regenerated by back washing, and the use of regenerative chemicals.
Cartridge and membrane filters are usually replaced when the specified
maximum pressure drop is exceeded. In a water treatment system, filters may
become potential sites for bacterial colonization and proliferation. This
potential should be considered in choosing a filter and determining its location
in the water treatment system. Water treatment systems may also require
frequent disinfection or filter replacement to control bacterial proliferation.
bed is exhausted, the second bed is moved upstream and a new bed added
downstream. This arrangement provides an adequate margin of safety.
Because of the high surface area of carbon beds, they provide a good site
for bacterial proliferation, the organics adsorbed by the carbon serving as nutri-
ents for the colonized bacteria. Also the removal of chlorine and chloramines
by the carbon exacerbates the proliferation of bacteria. Suitable water treat-
ment equipment must be used downstream to remove bacteria and endotoxins.
Reverse osmosis
Hydrostatic
Pressure
:water::':':':'il~~
Semipermeable +
Semipermeable
Membrane Membrane
permeate is called product water and the cross flow containing the concen-
trated contaminants is called the reject stream. Typically the product water
to feed water ratio is of the order of 25 to 50% for hemodialysis applica-
tions.
There are 3 major types of reverse osmosis membranes: cellulosic, fully
aromatic polyamide, and thin film composite membranes. Reverse osmosis
as a water treatment process was first demonstrated in the 1950's with
cellulosic membranes (cellulose acetate). These membranes are asymmetric
having a thin, dense layer that provides the solute rejection capability and a
thick porous substructure that provides structural support and strength. It is
because of this substructure that the membrane can withstand high hydro-
static pressures. However, under very high pressures, compaction occurs with
the thin dense layer merging into the porous structure with a reduction in
product water flux. Compaction is accelerated under high temperatures.
Cellulose acetate membranes can only be used with a limited pH range of
4-8 because of their susceptibility to hydrolysis. They are also susceptible
to bacterial degradation. Cellulose acetate membranes have a high water
permeability but poor solute rejection. Aromatic polyamide membranes are
also asymmetric but are resistant to hydrolysis and bacterial degradation. They
can withstand higher temperatures and a wider pH range (4-11), but at
the extremes of this range they are more prone to irreversible membrane
degradation. They have better solute rejection characteristics than cellulosic
membranes but are extremely sensitive to oxidant damage. Water containing
chlorine or chloramine must be treated with activated carbon before being
fed to reverse osmosis modules with polyamide membranes. Thin film com-
posites are made by casting a thin, dense film onto a porous substrate, so
that the materials and manufacturing process for the two layers can be different
and can be optimized for providing the best characteristics of water flux, solute
rejection, pH and temperature tolerance and compaction resistance. The sup-
porting structure is usually polysulphone and the thin dense layer materials
include aromatic polyamide, alkylarylpolymer/polyamide, and polyfurane
cyanate. These membranes are more stable than the aromatic polyamide mem-
branes to oxidant exposure, but exposure to oxidants must still be limited.
The reverse osmosis module contains the reverse osmosis membrane with
provisions for access to feed, product, and reject streams. There are 2 common
module configurations used in hemodialysis - spiral wound modules and
hollow fiber modules. In the spiral wound configuration, a membrane fabric
sandwich is created by attaching two layers of membrane back to back to a
woven fabric like nylon or dacron. This membrane-fabric sandwich is wound
around a central perforated hollow tube in a jelly roll configuration with plastic
mesh used to separate adjacent layers (Fig. 3). This configuration is not
unlike the coil dialyzer configuration used in the early days of dialysis. Product
water flows through the membrane into the woven fabric and from the fabric
into the central hollow core. The plastic mesh controls the feed stream channel
height and improves mixing of the feed stream with 'quasi' turbulence. The
96
o
Perforated Product
Water Tube
Contaminant % rejection*
Cations Ca++ 98
MgH 98
FeH , Fe+++ 98
Na+ 95
K+ 95
Anions HC0 3- 85-95
CI- 95
S04= 98
N0 3- 85-90
* Cellulose acetate membrane with 1500 mgtl TDS feed water, 75% recovery, 400 psi
solids content in the feed water increases, product water recovery (ratio of
product water to feed water flow) decreases. In addition to conductivity
monitoring of rejection, product water and reject stream flow meters, pressure
gauges (pre and post pump as well as product and reject stream pressures)
and low and high pressure switches (loss of water supply and flow path obstruc-
tions) are used to monitor the performance of the reverse osmosis unit. Means
for safe cleaning and disinfection with an interlock switch to safeguard against
accidental institution of these procedures contribute to good unit design. A
periodic high velocity auto-flush feature for removing foulants from the
membrane surface is also a convenient feature.
When water flux or solute rejection begin to diminish, it may be time to
clean the reverse osmosis membrane. The module can be flushed from the feed
to the reject side at a low feed pressure so as to have almost no product flow.
In addition chemical cleaning with cleaning agents such as hydrochloric
acid, citric acid, sodium hydroxide, sodium EDTA, and detergents may be
necessary to restore membrane performance. The chemical stability of a
membrane, type of foul ants in the feed water and materials of construction
dictate the choice of cleaning agent. The manufacturer of the reverse osmosis
device has the responsibility for providing complete instructions for cleaning
and restoring membrane function. Though the capital cost of reverse osmosis
modules is high, with appropriate water pretreatment, suitable operating
conditions, and regular maintenance, cleaning and disinfection, a module life
of several years is feasible making reverse osmosis a highly effective water
treatment process from the solute rejection and cost-effectiveness points of
view.
Ion exchange
In this water treatment process, water flows through a resin bed, the resin
having ion-exchange properties, i.e. able to exchange certain ions on the
resin for certain ions in the feed water. Softeners and deionizers are both
examples of ion exchange devices. In a softener, the resin exchanges sodium
ions for calcium and magnesium ions in the hard water (Fig. 5). Deionizers
may contain cationic or anionic resins or a mixture of the two. The cationic
exchange resin exchanges hydrogen ions for cationic species in the feed
water. Anionic exchange resins exchange hydrogen ions for anionic species
in the feed water. In a mixed bed deionizer, the hydrogen and hydroxyl ions
released from the resin combine to form neutral water (Fig. 6). Ion-exchange
resins are usually in a bead configuration contained in a cylindrical column
(Fig. 7). These resins have a finite ion-exchange capacity and when exhausted,
they need to be regenerated. With softener resins, regeneration is accomplished
using a saturated sodium chloride solution (brine). Softener resin regenera-
tion is usually done on site at periodic intervals in an automated fashion under
timer control. A brine tank containing sodium chloride pellets or crystals is
99
.. 8 8
;x;>...
'f' ~ 'I' ~
Na+ Na+
~
l;;. Na~ ~
Na+ J. ~
Na+
;y 8
.. 98
Na+
~ ~~ ~
Na+ Na+
~ .+'
. 98
~ ~ Ca++
~
Na+ ~ ~
~A.
~
~'"
Fig. 5. Representation of the ion-exchange process used for softening hard water.
Feed Water
Outlet Inlet
~--~--
located adjacent to the resin bed and at appropriate regeneration times, the
saturated brine solution is drawn from the brine tank by a venturi mecha-
nism. The volume of brine drawn depends on the regeneration frequency and
exchange capacity of the resin bed. The frequency of regeneration in turn
depends upon the volume of feed water processed and the hardness of the
feed water.
With deionizer resins, it is more common to have regeneration performed
off-site in a central facility because caustic acids (cationic resins) and alkalis
(anionic resins) are used for regeneration. The central facility will usually pick
up the exhausted exchange tank, leaving a regenerated tank in its place. For
hemodialysis applications, it should be ensured that the central regeneration
facility uses only food-grade chemicals and does not mix resins from indus-
trial users with hemodialysis resins. In industrial applications, deionizers may
be used for recovery of heavy metals or other hazardous contaminants which
may pose a serious threat to the hemodialysis patient. Similarly, for soft-
101
System design
In the previous section, we have described briefly the major water treatment
processes used in hemodialysis applications. A water treatment system may
combine some or all of these processes depending on feed water quality,
seasonal variations, product water flow rates required, and economic factors.
It is beyond the scope of this chapter to get into the details of system design.
However, some broad guidelines for selecting water treatment process com-
binations will be delineated along with some requirements for the water
distribution system used to distribute the product water.
The steps involved in the design of the system area:
1. Determine product water requirements (flow rates, usage factors, appli-
cation (e.g. reuse vs dialysate water).
2. Define produce water quality (e.g. AAMI standards)
102
3. Assess feed water quality including seasonal variations and worst case
scenarios.
4. Determine the concentration reduction ratio for various contaminants for
various water treatment processes based on manufacturers' specifications.
5. Select combinations of water treatment processes taking into considera-
tion economics, product flow rates, and impact of various combinations
of pretreatment requirements and final bacteriological quality.
We will consider an illustrative case. The use of reverse osmosis will yield
a concentration reduction ratio of 0.1 for most inorganic contaminants but is
not as effective in removing small organics, free chlorine, and chloramines.
Further, oxidants like free chlorine may damage aromatic reverse osmosis
membranes. The use of activated carbon may, therefore, be a necessary pre-
treatment to remove free chlorine and chloramines as well as small organics.
As carbon filters may release fines, a sediment filter may be necessary down-
stream of the carbon filter. Also carbon filters and sediment filters may promote
bacterial proliferation. The reverse osmosis membrane may be quite capable
of handling the bacterial burden imposed by the use of activated carbon and
sediment filters. If either because of product flow requirements or economics,
this approach to limiting bacterial contaminants is not feasible, an ultrafilter
may also be required. Further, if the level of an inorganic contaminant like
aluminum is so high as to require a concentration reduction ratio of better
than 0.1, a deionizer may be necessary. From a bacteriological point of view,
locating a deionizer upstream of the reverse osmosis device is preferred.
However, economic reasons may dictate that using deionization to 'polish'
the reverse osmosis product water is the preferred approach. In such cases,
ultrafiltration may be necessary as the last step for maintaining bacteriolog-
ical quality of the product water. If usage requirements are such that the reverse
osmosis device is incapable of satisfying peak product water requirements, a
holding tank may be necessary to meet peak flow requirements, the tank
being filled during periods of lower product flow demand. In such situations,
the bacteriological quality of the product water may suffer because of stagnant
conditions in the holding tank and exposure to the atmosphere. The holding
tank may need to be designed to reduce such exposure with a recirculation
loop through an ultraviolet irradiator to limit bacteriological quality.
As the brief example demonstrates, many factors dictate the design of the
total system and many of these factors may have contradictory requirements
and consequences. Skill and experience with juggling these varying factors
is therefore necessary in designing a safe, reliable, and cost effective system.
The product water distribution system consists of piping and associated fittings
such as valves, pressure and flow gauges, pumps, pressure regulators and seals.
Care must be exercised in the choice of these various components to ensure
103
that the product water quality does not deteriorate during distribution and
that various product water locations can be served at the desired flow rates.
Materials of contamination such as brass, copper, aluminum, and zinc should
be avoided because they are contaminants of known toxicity in hemodialysis.
The piping should be made of more inert materials such as polyvinyl chloride
(PVC), non-pigmented polypropylene, stainless steel, or glass. PVC is widely
available and inexpensive. However, according to some authorities PVC may
not be acceptable as the surface of PVC is considered to be conducive to
bacterial growth. Not only should the materials of construction be relatively
inert, they should also not degenerate with exposure to commonly used dis-
infectants such as bleach. If PVC is used, it should be Type 1 (non-plasticized)
and meet the requirements of the National Sanitation Foundation for potable
water. If solvent welding is used for joints, enough curing time should be
allowed before use, and vigorous flushing of the distribution system to remove
residual solvent is recommended.
In designing the distribution system, care should be taken to ensure that
stagnant areas are avoided because such areas promote bacterial prolifera-
tion. Also, the system should be designed to ensure that all parts of the
system are exposed to cleaning agents and disinfectants where necessary.
The system should be designed to promote high flow velocity, avoiding dead
ends, and long, multiple branching layouts.
be preferred, even if initial capital costs are higher, because of the long term
goal of uniform and appropriate product water quality to safeguard patient
well-being. Prospective vendors should be informed that upon installation of
the system, formal acceptance of the equipment is contingent upon valida-
tion of the entire system including the distribution system in terms of both
product water quality and quantity. Vendors should be responsible for com-
pliance with local plumbing and electrical codes in addition to meeting AAMI
product water requirements. Complete operating manuals should be provided
by the vendor including operator training requirements, monitoring require-
ments, recommended cleaning and maintenance schedules and procedures, and
troubleshooting guidelines.
Even if the water treatment system meets product water requirements at the
time of installation, deviations from these requirements may occur as a
consequence of deterioration of system performance, source water quality
variations, and changes in municipal water treatment practices. It is there-
fore, essential that appropriate monitors of water quality be installed at
appropriate locations of the water treatment system and that samples be drawn
at specified intervals for more detailed laboratory analyses to ensure compli-
ance with the AAMI Water Quality Standard. Appropriate documentation
procedures should be established for long term surveillance of water quality.
In 1988, many municipal suppliers in California switched from free chlorine
to chloramine. Despite all of the advance information provided regarding the
change and its impact on hemodialysis facilities, failure to check product water
for chloramines resulted in an outbreak of hemolytic anemia [29].
System performance may deteriorate with scaling and fouling of reverse
osmosis membranes, exhaustion of components such as activated carbon or
deionizer resin beds, bacterial contamination of the system, etc. Maintenance
schedules and practices should be designed to prevent such deterioration of
system performance. Monitoring will also alert the staff to system performance
deviations from desired levels due to unforeseen circumstances or will indicate
that the monitoring schedules may be inappropriate and may require some
tightening and fine tuning.
In describing the various water treatment processes, we have already
outlined the monitoring requirements for each process. Many of these monitors
are active, on-line monitors such as resistivity monitors, flow, pressure, and
temperature gauges, rejection ratio indicators, etc. However, the final deter-
minant of the adequacy of the water treatment system is the concentration
of various key contaminants in the final product water. These contaminant
levels should be determined at regular intervals using a certified laboratory.
Also other tests such as free chlorine/chloramine levels and bacterial count
105
Concluding remarks
References
1. The Safe Drinking Water Act of 1974, PL 523, 93rd Congress, Dec 16, 1984.
2. Alfrey AC, Mishell JM, Burks J, Contiguglia SR, Rudolph H, Lewin E, Holmes JH:
Syndrome of dyspraxia and multifocal seizures associates with chronic hemodialysis. Trans
Am Soc Artif Intern Organs 18: 257-261, 1972.
3. Alfrey AC: Dialysis encephalopathy syndrome. Ann Rev Med 29: 93-98, 1978
4. Dunea G, Mahurkar SD, Mamdani B, Smith EC: Role of aluminum in dialysis dementia.
Ann Intern Med 88: 502-504, 1978.
5. McDermott JR, Smith AI, Ward MK, Parkinson IS, Kerr DNS: Brain-aluminum concen-
tration in dialysis encephalopathy. Lancet 1: 901-903, 1978.
6. Elliot HL, Dryburgh F, Fell GS, Sabet S, MacDougall AI: Aluminum toxicity during regular
haemodialysis. Br Med J 1: 1101-1103, 1978.
7. Pierides AM, Edwards WG, Cullum UX, McCall JT, Ellis HA: Hemodialysis encephalopathy
with osteomalacic fractures and muscle weakness. Kidney Int 18: 115-124, 1980.
8. Short AI, Winney RJ, Robson JS: Reversible microcytic hypochromic anaemia in dialysis
patients due to aluminum intoxication. Proc Eur Dial Transpl Assoc 17: 226-233, 1980.
9. McGonigle RJS, Parsons V: Aluminum-induced anaemia in haemodialysis patients. Nephron
39: 1-9, 1985.
10. Schreeder MT, Favero MS, Hughes JR, Petersen NJ, Bennett PH, Maynard IE: Dialysis
encephalopathy and aluminum exposure: An epidemiological analysis. J Chronic Dis 36:
581-593, 1983.
11. Matter BJ, Pederson J, Psimenos G, Lindeman RD: Lethal copper intoxication in hemo-
dialysis. Trans An Soc Artif Intern Organs 15: 309-315, 1969.
12. Ivanovich P, Manzler A, Drake R: Acute hemolysis following hemodialysis. Trans Am
Soc Artif Intern Organs 15: 316-318, 1969.
13. Manzler AD, Schreiner AW: Copper-induced hemolytic anemia. A new complication of
hemodialysis. Ann Intern Med 73: 409-412, 1970.
14. Gallery EDM, Blomfield J, Dixon SR: Acute zinc toxicity in haemodialysis: Br Med J 4:
331-333, 1972.
15. Freeman RM, Lawton RL, Chamberlain MA: Hard-water syndrome. N Engl J Med 276:
1113-1118, 1967.
16. Evans DB, Slapak M: Pancreatitis in the hard water syndrome. Br Med J 3: 748, 1975.
17. Drukker W: The hard water syndrome: A potential hazard during regular dialysis
treatment. Proc Eur Dial Transpl Assoc 5: 284-287, 1969.
18. Nickey WA, Chinitz VL, Kim KE, Onesti G, Swartz C: Hypernatrernia from water softener
malfunction during home dialysis (letter). JAMA 214: 915, 1970.
19. Jowsey J, Johnson WJ, Taves DR, Kelly PJ: Effects of dialysate calcium and fluoride on
bone disease during regular hemodialysis. J Lab Clin Med 79: 204-214, 1972.
20. Lough J, Noonan R, Gagnon R, Kaye M: Effects of fluoride on bone in chronic renal failure.
Arch Pathol 99: 484-487, 1975.
21. Anderson R, Beard JH, Sorley D: Fluoride intoxication in a dialysis unit - Maryland.
Morbidity and Mortality Weekly Report 29: 134-136, 1980.
22. Carlson DJ, Shapiro FL: Methemoglobinemia from well water nitrates: A complication of
home dialysis. Ann Intern Med 73: 757-759, 1970.
23. Salvadori M, Martinelli F, Comparini L, Bandini S, Sodi A: Nitrate induced anemia in home
dialysis patients. Proc Eur Dial Transpl Assoc 21: 321-325, 1984.
24. Comty C, Luehmann D, Wathen R, Shapiro F: Prescription water for chronic hemo-
dialysis. Trans Am Soc Artif Intern Organs 20: 189-196, 1974.
25. Yawata Y, Howe R, Jacob HS: Abnormal red cell metabolism causing hemolysis in uremia.
A defect potentiated by tap water hemodialysis. Ann Intern Med 79: 362-367, 1973.
26. Eaton JW, Kolpin CF, Swofford HS, Kjellstrand CM, Jacob HS: Chlorinated urban water:
A cause of dialysis-induced hemolytic anemia. Sciences 181: 463-464, 1973.
107
27. Kjellstrand CM, Eaton JW, Yawata Y, Swofford H, Kolpin CF, Buselmeier TJ, von
Hartitzsch B, Jacob HS: Hemolysis in dialyzed patients caused by chloramines. Nephron 13:
427-433, 1974.
28. Botella J, Traver JA, Sanz-Guajardo D, Torres MT, Sanjuan I, Zabala P: Chloramines, an
aggravating factor in the anemia of patients on regular dialysis treatment. Proc Eur Dial
Transpl Assoc 14: 192-199, 1977.
29. Safety alert: Chloramine contamination of hemodialysis water suppliers. Food and Drug
Administration, Department of Health and Human Services, February 19, 1988.
30. Favero MS, Petersen NJ, Carson LA, Bond WW, Hindman SH: Gram-negative water bacteria
in hemodialysis systems. Health Lab Sci 12: 321-334, 1975.
31. Lauer J, Streifel A, Kjellstrand C, DeRoos R: The bacteriological quality of hemodialysis
solution as related to several environmental factors. Nephron 15: 87-97, 1975.
32. Blagg CR, Tenckhoff H: Microbial contamination of water used for hemodialysis. Nephron
15: 81-86, 1975.
33. Robinson PJA, Rosen SM: Pyrexial reactions during haemodialysis. Br Med J 1: 528-530,
1971.
34. Stacha JH, Pontius FW: An overview of water treatment practices in the United States.
J Am Water Works Assoc 76 (Oct): 73-85, 1984.
35. Environmental Protection Agency, Office of Water Supply - National Interim Primary
Drinking Water Regulations, U.S. Government Printing Office, Washington D.C. 1978.
36. American National Standard for Hemodialysis Systems. (RD-5) Association for the
Advancement of Medical Instrumentation, Arlington, VA, 1982.
37. Keshaviah P, Luehmann D, Shapiro F, Comty C: Investigation of the Risks and Hazards
Associated with Hemodialysis Systems (Technical Report, Contract 223-78-5046). U.S.
Department of Health and Human Services, Public Health Service, Food and Drug
Administration, Bureau of Medical Devices, Silver Spring, MD, June 1980.
38. Favero MS, Carson LA, Bond WW, Petersen NJ: Factors that influence microbial
contamination of fluids associated with hemodialysis machines. Appl Microbiol28: 822-830,
1974.
39. Favero MS, Petersen NJ: Microbiological guidelines for hemodialysis systems. Dial
Transplant 6: 34, 1977.
40. Man NK, Ciancioni C, Faivre JM, Diab N, London G, Maret J, Wambergue FP: Dialysis-
associated adverse reactions with high-flux membranes and microbial contamination of liquid
bicarbonate concentrate. Contr Nephrol 62: 24-34, 1988.
41. Klinkman H, Falkenhagen D, Smollich BP: Investigation of the permeability of highly
permeable poly sulfone membranes for pyrogens. Contr Nephrol 46: 174-183, 185.
42. Dinarello C: The biology of interleukin 1 and its relevance to hemodialysis. Blood Purif
1: 197-224, 1983.
43. Lonnemann G, Koch KM, Shaldon S: Induction of interleukin 1 from human monocytes
adhering to hemodialysis membranes. Kidney Int 31: 238, 1987.
CHAPTER 7
LEE W. HENDERSON
Introduction
of adequacy such as KtIV, for urea and plasma albumen concentration, treat-
ment time, etc.
At present we turn to urea kinetic modeling and the results of the National
Cooperative Dialysis Study (NCDS) [1] to provide a clinically qualified
quantitative definition of adequate hemodialysis! (there is no comparable study
addressing peritoneal dialysis). The derivative analysis by Gotch and Sargent
[3] has provided us with the now familiar "Domain Map" (Fig. 1) in which
normalized dietary protein intake is plotted against predialysis blood urea
nitrogen concentration with isopleths drawn out of the origin like spokes
from the hub of a wheel that show the amount of therapy rendered in terms
I
~KtIV
0.7 0.8
I I ,
140
- 0.9
120
- 1.5
20 ""inadequate excessive
I
...
protein protein
intake I intake
o ~ I
0.5 0.7 0.9 1.1 1.3 1.5 1.7
Fig. 1. A Domain Map showing the interralationship between predialysis blood urea nitrogen
concentration, protein catabolic rate and the dimensionless parameter KtlV (modified from
reference 3).
111
...
~ ""- ...... ---- ..
.E
.;
.....
.....0
0.0
~ '\ LowTAC
:.0 0.8
~ '\
urea
,.Q 'I.
0
~ 0.6 \
'I.
0.4
0.2
0.0 ......-'-..I....I....I.~..-.I...I......I.....-.I...I.............I...I.-'-...&..II....I.-'-......I....I.-'-......
0.4 0.6 0.8 1.0 1.2 1.4 1.6 0.4 0.6 0.8 1.0 1.2 1.4 1.6
Normalized protein catabolic rate, gldaylkg of body wt
Fig. 2. The probability of failure is here defined as any patient who died, withdrew for medical
reasons or was hospitalized prior to 24 weeks in the experimental phase of the protocol. The
date of failure was the date of death, withdrawal or first hospitalization, whichever occurred first.
The solid curves were calculated from NCDS data. The dashed lines represent 95% confidence
limits. The impact of treatment time (Td) may be read vertically for low and high time average
concentration of urea (T AC urea) and horizontally for the impact of T AC urea on probability of
failure. The probability of failure approximately doubles when going from long treatment time
to short treatment time for both low T AC ure• (0.1 to 0.2) or high TAC ure• (0.3 to 0.5).
standard treatment time patient groups, points up the reduced dietary protein
intake in the short treatment group. This raises the question of malnutrition
and its role in determining a satisfactory treatment outcome. It remains to
be determined what the respective contributions are of those factors that
cause death in association with malnutrition vs. those that do so in asso-
ciation with retained, slowly diffusing toxins (i.e. middle molecules) and
whether these factors interrelate.
The importance of treatment time as being a significant prescription variable,
of course reaches back to the work of Scribner and Babb on middle
molecules and the square meter hour hypothesis [15]. Treatment time taken
in the context of a constant Kt product is surrogate for middle molecule
clearance, i.e. clearance of molecular weight species that are larger than
the conventionally considered uremic toxins and smaller than albumen
114
(> 500 < 60,000 daltons). The line of reasoning that makes these delibera-
tions relevant to our topic of membrane selection is as follows. If time may
be considered surrogate for middle molecule clearance, then membrane selec-
tion for a more open transport structure will also enhance middle molecule
removal and may translate into the potential for shorter treatment time or more
adequate (read lower mortality) treatment. In considering the limitations placed
on extrapolations from the NCDS, both urea and its surrogate solutes and
time and its surrogate solutes must be carefully considered.
Nutritional Status as an Index of Adequacy: It is important to note a
recent observation by Lowrie (Fig. 3), that identifies a powerful inverse
correlation between plasma albumen concentration and death rate in the ESRD
population under treatment with conventional hemodialysis [12]. Teehan
confirms this observation for patients on CAPD [16]. Albumin concentration
is widely believed to be one objective, if not very sensitive, measure of nutri-
11 I---------------------------------~//A----_f
10
9
8
7
6
5
4
3
2
1
o
GT4.5 4.0-4.5 3.5-4.0 3.0-3.5 2.5-3.0 LT2.5
Fig. 3. Relative risk of death is plotted again sed serum albumin concentration with the
reference concentration of 1 being the 4 to 4.5 g/dl. Those with serum albumin concentrations
of greater than 4.5 (OT 4.5) were not significantly different from the reference standard. The
powerful inverse correlation for the crude data is even more significant when the ratios are
adjusted for the case mix predictors of death.
115
Table 1 lists the experimental parameters and their test ranges used in the
NCDS [1]. I will specifically explore the following treatment prescriptions
which employ different membranes that fall outside the test ranges noted in
Table 1 and indicate how these prescriptions may affect the quality of the
treatment rendered:
A. Large area and/or highly permeable membrane
B. High flux membrane
C. The peritoneal "membrane"
Dialysis length, min 274 271 274 268 278 200 275 190
±1O ±8 ±1O ±6 ±13 ±15 ±15 ±l7
TAC urea, mg/dl 52.0 51.1 51.4 87.8 50.8 54.7 52.0 93.6
±3.9 ±4.1 ±4.2 ±121.6 ±3.7 ±5.3 ±3.1 ±7.2
Midweek BUN, mg·dl 74 76 75 108 72 77 75 ll5
±6 ±6 ±6 ±14 ±6 ±7 ±6 ±9
Dialyzer urea clearance, 158 171 161 77 162 206 165 107
mUmin ±41 ±41 ±36 ±20 ±32 ±30 ±30 ±21
PCR,glkg 1.04 1.04 0.96 1.08
1"11
...cu
1"11
>< 100
:3
....
,Q =
e
I'll ....
cu
S;::;
:uc.. s
-
-;
~
10
~
U
10 10,000
Fig. 4. The log of the average cell wall mass transfer coefficient (permeability times area) is
plotted against the log of solute molecular weight. Data points are represented by symbols
and have been taken from a variety of workers. The shaded area represents the domain of
hemodialyzer clearance for 1 sqare meter cuprophan membrane. Clearance across the cell wall
exceeds that for the dialyzer for solutes small in molecular weight, such as urea and creati-
nine. Uric acid, B12 and inulin are measured to fall significantly below dialyzer clearance.
118
for adequacy, one can assume a larger clearance of middle molecules and hence
a dialysis prescription that is "more adequate" than the same KtlV obtained
using cuprophane and solely diffusive transport. The fraction of overall mass
transport that is convective in mechanism may be considered to be surrogate
for treatment time. I speculate that a suitable mix of shortening treatment
time sustaining urea removal and increasing convective transport would result
in a treatment that is comparably adequate to conventional full time hemo-
dialysis. If adequacy is comparable, the shorter treatment time by definition
offers a higher quality of therapy.
At present there are only a few prospective studies that give clear qualifi-
cation of this treatment strategy. See, for example, studies such as that of
Channard et al. [18] using the more open complement kind PAN membrane
in which patients (n = 31) treated for 9.3 ± 0.2 hrs/wk are contrasted with a
control group (n = 31) treated for 16.2 ± 0.3 hrs/wk with cuprophane. A
significantly lower hospitalization rate and number of hospitalization days
for dialysis related complications was achieved with the PAN membrane in
spite of carrying a higher pretreatment blood urea nitrogen concentration
1.16 ± 0.05 mmolll for short treatment vs. 1.01 ± 0.03 mmolll for conventional
treatment. There are likely 2 mechanisms at work here, i.e. differences of
complement activation (this will subsequently be discussed) and the aug-
mentation of middle molecule clearances achieved with the PAN membrane.
Von Albertini et al. [24] have clinically applied the bench study of Cheung
et al. [25] in a small number of patients (n = 18) to deliberately exploit the
relationship (therapeutic equivalency) of time and middle molecules. They have
achieved an average of 2 hours in treatment 3 times per week for a year or
more, using two modules of high flux membrane and flow rates that are
nearly twice the common values (Fig. 5). His patients have not shown the
cardiovascular instability, considered limiting by most, for removal of excess
body water (3.6 ± 1.2 1) in a 2 hour treatment time frame with hemodialysis.
His rehabilitation rate for these patients was very high at > 90%. This extra-
ordinarily high flux treatment has shown the expected urea rebound that is
predicted when the limiting mass transfer resistance is the cell wall and not
the dialysis membrane [9]. His modeling studies support a more than adequate
status of clearance for not only urea but for the index solute Inulin (5,200
daltons) as well.
At present the only safe recommendation for this form of treatment, as noted
by Von Albertini, is to use a KtlV urea of 1.2 as a minimum value. This will
ensure that the mass of urea removed in spite of urea disequilibrium at high
clearance rates is adequate and larger and/or more slowly diffusing solutes that
are removed is comparable to, or greater than with conventional therapy [26].
An additional point of note, commented on above, may be drawn from
the recent work of Guttierez et al. [27, 28]. Their work implicates comple-
ment activation and cytokine release in the increase of catabolism noted by
Borah et al. [29], and Farrell et al. [30], in response to hemodialysis. The
extra catabolism that occurs on the days of dialysis with cuprophane results
120
~
Blood Pump
~-
500
1025
----.. Drain
Net UF Pump 1000
I
-
900 Differential
375 Pressure
[ Control
900
--
1000
Pyrogen
-
1000
Dialysate
-... 475
Filter
Fig. 5. Flow diagram and flow rates in mUmin for a hemodiafiltration circuit (von Albertini).
A pair of high flux membranes is used in series with blood and dialysis fluid flowing in
countercurrent manner. High flux membrane I is driven by the differential pressure controller and
blood pump to ultrafilter from blood to dialysate in the amount of 125 mUmin. Sterile pyrogen
free diluting fluid from the pyrogen filter and fluid cycling device is delivered to high flux
membrane II and restores all but 25 ml/min of backfiltration to the blood path. Diffusive
transport occurs in both membranes I and II.
weekly, using 1.0 to 1.5 m2 cuprophane membrane and conventional flow rates
of blood (250-350 mllrnin.) and dialysis fluid (500 mllrnin.). The amount of
hemofiltration given was calculated with the formula that one third of the
total body water needed to be ultrafiltered at each treatment [26]. This
would mean that a 70 kg man with 42 I of total body water would exchange
25 liters of ultrafiltrate for diluting fluid, assuming a 2 I removal of excess
body water. Urea clearance in post dilution hemofiltration equates to the
volume of ultrafiltrate i.e. in this example the total volume of plasma water
cleared of urea would be 25 1. This of course, provides a KtlV for urea of
only 0.6; a figure that by NCDS standards should produce major morbidity
promptly. Figure 6 is a plot of mortality for these patients over the ten year
follow up. This may be compared with similar mortality curves for his
hemodialysis population and for that reported by the EDTA and French Dialysis
Registries [26].
His hemodialysis population was dialyzed for 5-6 hours thrice weekly, with
1.0 m2 cuprophane. This would likely ensure an adequate KtlV by NCDS
standards although weight data and dietary protein intake for these populations
is lacking. The patients in Figure 6 were randomly assigned to hemodialysis
100
....... (132)
~ " '''(115)
\ \ (115) ....
1974-1978
--
':\ (91) , ... (108)
\\ ......
90 ~104)
-.....
\\ .........
~
\. \ . ........... (72)
!IS \ \. ...........
>
"-
\ \ ........ (46)
t:
::s
\\
\ \ ..... (3) HF
"
CJ)
80
'\.
"\ \DP \
'- EDTA \
... HD
72 (11)
T
i i i
1 2 3 4 5 6 7 8
Time in Years
Fig. 6. Survival data for patients starting regular hemofiltration or hemodialysis treatment between
1974 and 1978. For comparison, data taken from both the EDTA and Diaphane registries for
the same time period are shown. See text for further discussion.
122
or hemofiltration and "poor risk" patients were excluded from the study, i.e.
patients with "complications such as, diabetes mellitus, cancer, and systemic
or severe cerebrovascular or cardiovascular disease" [31]. The message here
is that quality assurance quidelines for treatment with hemodialysis cannot
be mindlessly applied to a therapy where the selection of a more hydrauli-
cally permeable membrane alters urea's surrogate status for uremic solute
removal.
1000 DowHFAK
Model 3
1M 2
100 ~------~~~------~~----~
"-
" "-
/\
Pet\oneal membrane
Inulin
10
t Sucrose
1 Uric Acid
ere thune
Fig. 7. Log log plot of the permeability area product ("clearance") versus solute molecular
volume. See text for further discussion of clearance. Note the crossover point between the
peritoneal membrane and the cellulosic membrane for solutes in the 5,000-6,000 dalton range.
dependent than large ones when clearance is being examined. Hence, urea
clearance for the peritoneal membrane falls well below that for cellulosic
membrane. It is important to note that peritoneal clearance as it is commonly
measured and described is different from the clearance term for the native
kidney or that for the artificial kidney that operates by diffusion (hemodial-
ysis) or that for convective transfer (hemofiltration) or any combination of
the two (hemodiafiltration). The common formula for peritoneal clearance
pD Vlt == C
where
C == clearance (mllmin.)
D == dialysate concentration (mg/dl)
P == plasma water concentration (mg/dl)
Vlt == volume (ml) of spent dialysate per exchange time in (minutes)
has in the numerator a mass of solute (DV) that has been removed and is
present in the spent dialysate over a given exchange time. (Note: Urea in the
dialysate is directly measured and this is not subject to errors that occur in
124
10
09
08
·s
:::
07
] 06
~
-~
u
::: 05
...
r;I
r;I
~
04
P
"-
<:0 03
p...
02
01
00
20 60 90 120 160 200 240 300 360 420 480
Time in minutes
Fig. 8. Stylized curves for the equilibration ratio between dialysate and plasma plotted against
time for a CAPD patient with average transport characteristics.
125
For larger solutes the transport rate across the peritoneal membrane may
be comparable to or higher than that for cellulosic membrane. It is the rela-
tionship or relative concentration that urea bares to other more toxic solutes
that establishes its surrogate status. Furthermore, the NCDS employed a
schedule of three times weekly treatment with each treatment ranging (by study
group) from 3.0 to 4.5 hours (Table I). CAPD is, of course, a continuously
applied therapy. This having been said, I note for you that patients consid-
ered to be adequately treated on CAPD using clinical criteria (we have no
NCDS for CAPD) have a measured KtlVof 1.6-1.7 per week [38]. Dividing
this by 3 to equate it to the amount of hemodialysis offered on a thrice weekly
schedule identifies that CAPD at 0.5-0.6 is dramatically lower than, i.e. half,
that required for adequate hemodialysis (1.0-1.2).
Dr. Keshaviah emphasizes this point (i.e. the relevance of treatment
application time in a recent publication [39]. Table II is taken from that pub-
lication and casts clearance in terms of liters per week for solutes of different
molecular weight contrasting CAPD with two commonly used hemodialysis
membranes. CAPD, because of its continuous nature, showed higher net weekly
clearance for solutes as low in molecular weight as vitamin B 12 when con-
trasted with 1.5 m2 cuprophane (8 micron) applied for four hours, three times
weekly.
Table II. Weekly clearances of peritoneal and hemodialysis membranes (units = liters per week).
treatment is surrogate for middle molecule removal [6, 13]. With refer-
ence to Table II, one notes that with the selection of the peritoneal membrane
that is more open to the diffusive transport of larger solutes coupled with
the continuous nature of the CAPD prescription that urea's surrogate status
with regard certainly to the larger molecular weight toxins would be
radically different than for hemodialysis, i.e. more middle molecules
removed/gram of urea removed. I have noted previously the surprising
lack of attention paid to this relationship in the published results of the
NCDS. Recent analysis of data from the USRDS registry provides powerful
evidence that short prescription time correlates with early mortality and
likely offers one explanation of the higher death rate present in the U.S.
dialysis population than that in Europe [2, 14].
- Thirdly, as stated above, for a patient with a reasonably permeable
peritoneal membrane, one may well see that the dialysate to plasma con-
centration ratio for urea may achieve a value of unity for the last hour or
two of a six hour dwell time, i.e. no more movement of urea from blood
water to dialysate. This equilibration of urea does not preclude the continued
loss into the dialysate of other less swiftly diffusing toxic solutes. Again,
the surrogate status that urea holds for more toxic solutes of slower
diffusivity is abrogated by this event.
- Fourthly, as previously noted, for non complement activating synthetic
membranes, the choice of the peritoneal membrane for treatment exempts
the patient from a thrice weekly activation of complement with it's release
of the powerful inflammatory mediators, interleukin-1 and tumor necrosis
factor which appears to result in a burst of protein catabolism with hemo-
dialysis that is not present with CAPD [29, 30]. Recent work by the
Bergstrom Group [41] points out further that the relationship between
protein catabolic rate plotted on the axis and Kt/V urea on the ordinate
showed a markedly steeper slope than that for hemodialysis. This is
interpreted to show that increasing dialysis dose, as measured by Kt/V
urea has a "more salutory" effect on appetite in peritoneal dialysis than in
hemodialysis.
There are some common features between CAPD and hemofiltration that are
instructive and may explain a clinically satisfactory outcome at remarkably
low Kt/V values:
- Both employ a complement kind membrane and hence a reduced catabo-
lism of protein on dialysis days.
- Both have a disproportionately high clearance of middle molecules when
compared with urea.
- Both techniques are conducted using only sterile pyrogen free solutions -
unlike routine hemodialysis.
- Both techniques show less reduction in residual renal function over time
[31,42].
There are interesting points of difference that will require further study before
fully understanding their implications:
127
From the above discussion it is apparent that one may not isolate selection
of the membrane from the other components of the treatment prescription when
examining the quality of therapy rendered. At present we have no hard data
on the comparative importance of:
A. The two physician variable components of the treatment prescription,
namely, treatment time and the elements comprising artificial kidney
clearance rate (i.e., membrane area, and permeability, the flow rates of
blood and dialysate, and the respective components of convection and
diffusion) and
B. Removal of the two broad solute domains i.e. conventional size « 500 d)
or middle molecule (> 500 d and < 60,000 d) size, which are affected
differently, depending upon the choice of membrane and technique.
One may only guide patient therapy by offering, at minimum, the amount
of therapy given by the best treatment group of the NCDS. This is not to
say that more treatment might not be desirable. Urea kinetic modeling coupled
with an understanding of the influence of treatment time on solute clearance
profile is required in order to know when this minimum is achieved. Where
uncertainty exist for a given prescription about middle molecule clearance it
is well to make actual measurements of a test solute such as inulin or vitamin
B 12 in order to resolve that uncertainty.
One may approach urea kinetic modeling using any of a number of sim-
plifications of the original approach that vary more in style than substance.
Not all patients will need to be modeled as empirical wisdom about average
patients requiring an average prescription is quit strong. It is only where
significant variations in patient parameters of height, weight protein cata-
bolic rate or the treatment parameters that determine clearance, that is, A.
and B. above, that depart from the common experience that formal modeling
will be required to assure the delivery of an high quality treatment.
There is even more uncertainty here as there has, as yet,4 been no prospec-
tive clinical qualification of urea kinetic modeling or for that matter any
129
Notes
1. Clinical assessment such as changes of lean body mass, subjective well being, complica-
tion rate and the like, coupled with serial measurements of blood urea nitrogen, creatinine,
calcium phosphate, CO 2 content, plasma proteins, hemoglobin, etc. are still more commonly
used to assess treatment adequacy worldwide than is urea kinetic modeling. The quality
of treatment rendered using these parameters is sufficiently high to make subtle
distinctions in quality impossible to detect at the clinic level simply because the number
of study subjects required to show significant trends likely exceeds that available to even
the director of a large clinic. I note, for your persuasion to this point, the surprise
engendered in the U.S. dialysis community by the recent identification of a substantially
increased mortality rate in the U.S. dialysis patient population as contrasted with those of
Europe and Japan when natinal population bases were examined [2].
2. The popular use of 58-60% of body weight in kg or nomogram values involving height,
weight, sex, to compute total body water will obviate this problem but has problems of
its own. A malnourished patient will be assigned a lower body water volume and hence
make his calculated KtlV artifactually high.
3. I put fractional clearance in quotes as it is not strictly that. Note that fractional clearance
cannot exceed. KtlV can, of course, exceed unity, but this is because the same volume
of blood passes through the dialyzer repeatedly and is cleared again and again. Said another
way, if KtlV is truly a fractional clearance, a value of unity indicates that the entire
space of distribution for urea has been emptied. This of course does not happen with a
KtlV = 1.
4. A North American Multicenter trial to prospectively qualify urea or creatinine kinetic
modeling as a suitable predictor of morbid/mortal events is now being conducted under
the guidance of David Churchill, M.D. [39].
130
References
23. Channard J, Brunois JP, Melin JP, Lavaud S, Toupance 0: Long term results of dialysis
therapy with a highly permeable membrane. Artif Organs 6: 261-266, 1982.
24. von Albertini B, Barlee V, Bosch JP: High flux hemodialfiltration: Long term results.
J Am Soc Nephrol 2 (abstract): 354, 1991.
25. Cheung AC, Kato Y, Leypoldt IK, Henderson LW: Hemodiafiltration using a hybrid
membrane system for self - generation of diluting fluid. Trans Am Soc for Art Int Organs
28: 61-65, 1982.
26. von Albertini B: High-efficiency hemodialysis: An overview. Contrib to Nephrology 61:
37-45, 1988.
27. Gutierrez A, Alvestrand A, Wahren J, Bergstrom 1: Effective in vivo contact between
blood and dialysis membranes on protein catabolism in humans. Kidney Int 38: 487-494,
1990
28. Gutierrez A, Alvestrand A, Bergstrom 1: Membrane selection and muscle protein catabolism.
Kidney Int 42: S86-S90, 1992.
29. Borah MF, Shoenfeld P, Gotch FA, Sargent JA, Wolfsom M, Humphreys MH: Nitrogen
balance during intermittent dialysis therapy of uremia. Kidney Int 14: 491-500, 1978
30. Farrell PC, Hone PW: Dialysis induced catabolism. Am J Clin Nutr 33: 1417-1422, 1980.
31. Quellhorst E: Long-term survival. In Henderson LW, Quelhorst EA, Baldamus CA, Lysaght
MI (eds). Berlin: Springer-Verlag, p 221, 1986.
32. Rippe B, Stelin G: Simulations of peritoneal solute transport during CAPO. Application
of the two-pore formalism. Kidney Int 35: 1234-1244, 1989.
33. Nolph KD, Twardowski Z: The peritoneal dialysis system, in Nolph KD (ed) Peritoneal
Dialysis (pp 13-27). Kluwer Academic Publishers, 3rd Edition, 1989.
34. Dedrick RL, Flessner MF, Collins 1M, Schultz IS: Is the peritoneum a membrane? ASAIO
J 5: 1-8, 1982.
35. Henderson LW: The problem of peritoneal area and permeability. Kidney Int 3: 409-410,
1973.
36. Babb AL, Johansen PI, Strand MI, Tenckhoff H, Scribner BH: Bi-directional permeability
of the human peritoneum to middle molecules. Proc of the European Dial and Trans Assoc
10: 247-262, 1973.
37. Henderson LW, Nolph KD: Altered permeability of the peritoneal membrane after using
hypertonic peritoneal dialysis fluid. J of Clin Invest 48: 992-1001, 1969.
38. Nolph KD: Quantitating peritoneal dialysis delivery: A required standard care. Seminars
in Dial 4(3): 139-141, 1991.
39. Keshaviah P: Urea kinetic and middle molecule approaches to assessing the adequacy of
hemodialysis and CAPO. Kidney Int 43: S28-S38, 1992.
40. Keshaviah PR, Nolph KD, VanStone JC: The peak concentration hypothesis: A urea kinetic
approach to comparing the adequacy of continuous ambulatory peritoneal dialysis and
hemodialysis. Peritoneal Dialy Bul 9: 257-260, 1989.
41. Bergstrom 1, Alvestrand B, Lindholm B, Tranaeus A: Relationship between KtlV and
protein catabolic rate (PCR) is different in continuous peritoneal dialysis (CPO) and
haemodialysis (HD) patients. JASN 2 (abstract): 358, 1991.
42. Lysaght MJ, Vonesh E, Gotch F, Ibels L, Keen M, Lindholm B, Nolph KD, Pollock CA,
Prowant B, Farrell PC: The influence of dialysis treatment modality on the decline of
remaining renal function. Trans Am Soc Artif Intern Organs 27: 598-604, 1991.
43. Wolf AV, Remp DG, Kiley IE, Currie GO: Artificial kidney function: Kinetics of hemo-
dialysis. J Clin Invest 30: 1062-1070, 1951.
44. Smith HW: Principles of renal physiology. Oxford University Press (pp 25-35), 1956.
45. Churchill ON, Taylor W, Members of the CANUSA Peritoneal Dialysis Study Group:
Canada-USA (CANUSA) Multicentre Study of Peritoneal Dialysis Adequacy; Description
of the study population and preliminary results. Adv Peril Dial 8: 88-92, 1992.
46. Blake PG, Oreopoulos DG: Total creatinine clearance is not a good predictor of patient
outcome in CAPO JASN 2 (abstract): 358, 1991.
CHAPTER 8
Introduction
To physicians who are concerned with the impact that end-stage renal disease
and hemodialysis have on the day-to-day life and psychological well-being
of patients and their families, a major part of the "quality" in quality assur-
ance consists of patient quality of life.
The patient with end-stage renal disease faces a debilitating and disrup-
tive chronic illness, and a treatment regimen that is itself a complex, demanding
and incessant intrusion into one's personal and social life. Dialysis patients are
required, at the very least, to stoically endure pain, discomfort, fatigue and
deprivation on a daily basis, year after year. And most dialysis patients
heroically struggle to continue to fulfil their roles as spouses, parents, friends,
workers and community members even as they are nudged by the healthy
around them into a twilight world of the " ... marginal person - here today,
gone tomorrow [1]." Thus it is not surprising to find that aspects of quality
of life, often referred to as 'psychosocial factors', have been of concern in
nephrology for many years. 1
The measurement of quality of life has become an integral part of clinical
research in many areas. As technological and pharmacological advances have
increased the efficacy of modern medicine, concomitant research efforts have
gone beyond the assessment of mortality and morbidity to include the dimen-
sions of physiological, psychological and social functioning. Likewise, as a
growing proportion of the health care system is involved with the management
of long-term chronic diseases and conditions, the focus of research has widened
to include the patient's physical comfort, psychological health, social and
vocational activity, and the impact on the patient's family. In short, the extent
to which medical treatment is consciously directed to help patients live com-
fortable, productive and satisfying lives is the extent to which clinical research
into that treatment has included measures of patient quality of life.
In addition, there is a growing awareness among medical practitioners that
knowledge of, and sensitivity to, quality of life issues can contribute to
successful treatment outcome. For example, minimizing and ameliorating
L.w. Henderson and R.S. Thuma (eds.), Quality Assurance in Dialysis, 133-149.
© 1994 Kluwer Academic Publishers.
134
It may be helpful to briefly sketch the outlines of how the issue of quality
assurance, from the point of view of patient quality of life, could be addressed
in a pragmatic way within a busy dialysis unit. The foundation would be the
ongoing collection of psychosocial data from patients in a regular and
systematic fashion, in much the same way that medical and physiological
data is routinely collected today. In order to minimize the burden on medical
staff, the bulk of this information could be collected through patient-completed
questionnaires. In centre dialysis units, for example, most questionnaires could
be filled out before or during dialysis sessions. The role of unit staff would
be limited, in most cases, to the distribution and collection of questionnaires
with a quick check of completion rates. In order to minimize the burden on
the patients, questionnaires need to be short, clear, and easy to read and answer.
Data collection could occur relatively infrequently, perhaps quarterly, with
the option of larger-scale assessments on an even less frequent cycle.
Scheduling would likely be done on an individual patient basis, keyed to the
data of admission to the unit. In order to maximize the utility of the infor-
mation gathered, the measures chosen should meet high standards of validity
and reliability; as enumerated later, a substantial number of accredited quality
of life instruments are, in fact, currently available.
The routine collection of quality of life information will generate a data base
135
with several valuable applications. The individual dialysis unit will instantly
have an empirical snapshot of their caseload, a descriptive cross-section
captured in the range of scores, typical values, characteristics of particular
subgroups, and values associated with individual patients of interest. By
matching measurements from questionnaires with personal knowledge of
individuals, unit staff will immediately begin to make the rather esoteric
numbers derived from the quality of life scales more meaningful, with certain
types of problems, for example, becoming associated with specific ranges of
values on particular scales. In addition, through relatively simple statistics
the inter-relationships between the different aspects of quality of life, and
the relationships between quality of life and other characteristics, can be
investigated at the local level, in response to local interests and concerns. Once
a series of measurements have been taken, it becomes possible to track indi-
viduals, subgroups, and the caseload as a whole over time. With experience,
the emergence of individual problems can be followed and even anticipated
through change in a patient's scale scores, thus facilitating early preventa-
tive interventions. Likewise the effects of therapeutic interventions over time
can be assessed. As long-term data builds up, it will be possible to trace the
course of psychosocial adaptation to treatment from initial adjustments
onwards, and to relate individual or subgroup patterns to other characteris-
tics, therapeutic milestones, and events in the unit. In this way, high-calibre
quality of life data, collected systematically and regularly over time, could
provide valuable supplemental information useful in diagnostic and therapeutic
decision-making in the dialysis unit.
When many dialysis units are routinely collecting quality of life data, their
utility is further enhanced. Individual units can compare patient and perfor-
mance data on a regional, national and international basis. Data on patients
with rare conditions or problems can be combined for more powerful analyses.
Comparisons between different types of units - rural versus urban, short-hours
versus long hours, reusers versus single use units, for example - become
possible on an ad hoc basis, facilitating answers to questions that today can
only be addressed through expensive "extracurricular" research projects.
All of these potential benefits from quality of life assurance, however,
assume that patient quality of life can be assessed in meaningful, valid, reliable
and efficient ways. It is to these points that we now tum, beginning with matters
of conceptual definition, followed· by measurement issues.
Conceptual definitions
Quality of life
ness of the concept, however, means that each individual will define the quality
of their own life in a unique way, with different priorities and different weights
differentially assigned to the various aspects of personal existence. In order
for the concept to have utility in scientific research and clinical practice, this
multiplicity of personal meanings must be transcended through conceptual and
operational definitions which have universal applicability2 at the same time
as they accommodate the richness of individual variation.
The first step in delimiting the concept is to note that we are concerned here
with patient quality of life, as distinguished from the quality of life of a general
population. The latter notion comes from within the tradition of the general
social survey in the field of sociology, and is concerned with the broad
assessment of the nature and grading of life along such dimensions as attitudes
towards self and others, racial and ethnic groups, women, family life, work,
social and leisure activities, personal economic outlook, national economic
outlook, and various political issues [2-4]. While the notion of patient quality
of life takes a more functional orientation within a narrower focus, a great deal
of theory and methodology is imported directly from these large scale social
surveys.
The concept of patient quality of life can be derived almost directly from
the World Health Organization's definition of health as "[a] state of complete
physical, mental and social well-being and not merely the absence of disease
or infirmity [5]." When we consider the quality of life of a dialysis patient,
by definition disease is present. The central focus is thus the level of physical,
mental and social functioning of the individual in the presence of kidney
disease. Patient quality of life " ... represents the functional effect of an illness
and its consequent therapy upon a patient, as perceived by the patient [6]."
The main dimensions of patient quality of life are embedded in the WHO
definition of health: physical functioning, social functioning and psycho-
logical functioning. While specific components of these dimensions may be
chosen to fit particular research questions or clinical concerns, there are several
aspects of functioning that are of primary interest and importance with dialysis
patients (see Fig. 1).
Physical functioning
IQUALITY OF LIFE I
PHYSICAL II PSYCHOLOGICAL I SOCIAL I
- Physiological - Depression - Psychosocial
Functioning - Denial Stress
- Disease Stress - Anxiety - Social-leisure
- Dialysis Stress - Social Activities
- Sleep Adequacy Introversion - Social Support
- Fatigue - Self Depreciation - Marital/Family
- Hypochondriasis Functioning
- Vocational
Functioning
Psychological functioning
Social functioning
~easurennentissues
Measurement approach
Disease-specific instrumentation
shivering; back pain; chest pain; and hypotension. Four additional items
were included for the EPO project: dyspnea; wheezing; perspiration; hives,
rash. Patients are scored via a Symptom Count (0 to 11) or the number
of incidents that occurred during the dialysis session, and via a Total
Discomfort score (0 to 99) equalling the sum total of distress reported
from all items. Average per-item stress can also be computed.
(6) Fatigue rating scale: This scale measures the mean level of fatigue
experienced by patients in the time period between dialysis sessions.
Scores can range from 0 to 9.00.
(7) Basic personality inventory: The full Basic Personality Inventory is a
psychological profile composed of 12 scales, each consisting of 20
True-False items, with a corresponding score range of from 0 to 20 [24].
While the orientation of the inventory as a whole is towards the
measurement of psychopathology, we have found six of the scales to be
appropriate for use with populations that are psychologically normal:
depression, denial, anxiety, social introversion, self-depreciation and
hypochondriasis.
The Somatic Symptoms Distress Scale is included as Figure 2 in order to
illustrate some of the characteristics of these instruments. The most efficient
format is the self-completed questionnaire: data can be gathered with minimal
additional work by clinical staff. In order to maximize the return rates and
completion rates it is important to word the instructions and questions in
simple, concrete terms, and to utilize a layout that patients with mild visual
impairments find easy to read and respond to. As can be seen in the SSDS,
we prefer to separate questions on the occurrence of symptoms/stressors from
questions on the level of discomfort/stress, as this seems to minimize missing
data. The 9-point response scale which we use for level of discomfort or
stress allows for a wide range of variability within the between patients, and
facilitates the sensitivity of the instrument to treatment differences and change
over time.
It has been our experience that when patients report a reduction, for example,
in the overall level of dialysis-related discomfort, this may be due to the fact
that they are affected by a fewer number of symptoms, or due to a lessening
of the level of discomfort for certain symptoms, or to some combination of
the two. Hence scoring the SSDS involves taking a symptom count (the number
of Yes's), computing a total discomfort score (by summing the level of
discomfort for individual items with No-symptom not present weighted 0), and
calculating an intensity score by dividing total discomfort by symptom count.
Subsequent analysis will involve an assessment of change/difference in these
three scale scores, as well as an examination of each individual items to
locate the individual aspects of dialysis affected.
142
Circle YES if you did experience the symptom and indicate the level of
discomfort.
DISCOMFORT SCALE
very
low low moderate
f---+---+I--+---II
2 3 4 5 6 7 8 9
Nausea No Yes
D Itching No Yes
D
Vomiting No Yes
D Shivering No Yes
D
Headache No Yes
D Back Pain No Yes
D
Muscle Cramps No Yes
D Chest Pain No Yes
D
Dizziness No Yes
D Hypotension
(Low blood pressure)
No Yes
D
Tingling of No Yes
D
extremities Other (specify):
D
Fig. 2. Somatic symptoms distress scale (SSDS).
The reliability of a scale refers to its consistency internally and its stability
over time. In a scale with high internal consistency, all component items are
tapping slightly different aspects of the same underlying phenomenon and
are thus highly intercorrelated. Internal consistency reliability is generally
expressed by Cronbach's Alpha, a conservative measure that ranges from 0
to +1. A value of 0.60 is taken by convention as the minimum level for research
purposes, while a value of 0.90 may by required for diagnostic applications.
Stability over time is evaluated by test-retest reliability using the correlation
143
between scores obtained over a short period during which the measured
phenomenon is assumed not to have changed [25-29]. In a recent assess-
ment of the instruments listed earlier [22], all scales except the Physiological
Index (an inventory of largely unrelated items), attained Cronbach's alphas
between 0.60 and 0.88, while test-retest coefficients ranged between 0.38
and 0.92.
Validity centres on the question of whether a scale actually measures the
concept it is intended to measure. Validity assessment tends to be a contin-
uous process as new data and experience are accumulated under different
conditions. The validity of a scale is usually determined relative to the specific
use(s) for which it is intended; hence different types of validity assessments
correspond to different research applications. The concept "quality of life",
when used in clinical research, often takes the role of a secondary treatment
outcome supplemental to morbidity and mortality [30]. The validity assess-
ment of a quality of life instrument is thus normally concerned with
construct validity: the adequacy with which the scale represents the underlying
theoretical concept, namely a specific domain of quality of life. Generally
this is demonstrated by showing that the scale scores correlate in logical
directions with other variables. Again, the instruments listed above have been
shown to be correlated in expected directions with related variables, and have
successfully discriminated between treatment groups, at statistically signifi-
cant levels [22].
It should be noted that the use of quality of life measures to assess quality
assurance in dialysis units represents a new application for these instruments,
for purposes other than those typically associated with clinical research. Hence
the validity and reliability of the scales in the ongoing monitoring of patients'
coping with ESRD and dialysis, for establishing local, regional, national and
even international norms of patient functioning, and for providing supplemental
input into diagnostic and treatment decisions, remains to be established through
actual clinical applications.
# patients %
Canada 69 23.0
Denmark 5 1.7
England 30 10.0
France 21 7.0
Germany 33 11.0
Italy 72 24.0
Sweden 26 8.7
U.S.A. 44 14.7
Total 300 100.0
145
Table II reveals that interference due to disease and the demands of dialysis
with work and housework (vocation) and vacations, are the most common
complaints affecting 39% of respondents. Interference with social life, and
financial and sexuality problems are also relatively common. The mean stress
for an item is a measure of the seriousness of the complaint. Interference
with vacation, sexuality issues, vocational interference, social interference,
financial problems and childcare interference all generated average levels of
stress, for those affected by the item, at levels towards the high end of the
scale. Data on the scale scores indicates that 78.4% of these patients were
affected by at least one item of the scale, with a sample-wide average of
2.37 items; the mean total stress level is 12.32 in a possible range of 0
through 81.
The primary disease-related stressor is feeling weak and tied, which affects
three quarters of the sample, although the actual amount of stress this causes
is moderate (Table III). The general health items (up and down health, being
physically ill) affect half the sample, with itch, blood pressure problems,
fluid retention and sleep disturbance also relatively common. For those
affected, the worst stressor is the impairment of sexual functioning; the mean
stress level for sleep problems, being weak and tired, fear of death, fluid reten-
tion and up/down health also fall towards the high end of the response scale.
Almost every patient (97%) reports at least one disease-related stressor, with
the average being 6 symptoms; the overall mean total stress is approximately
30 in a range of 0 to 126.
Meaningful Interpretation of BPI scores requires some standards from
known groups. Several population norms are presented in the scale manual
[24]. In general, dialysis patients tend to score higher than the population at
large on denial (a constructive defense mechanism), depression and hypo-
146
chondriasis, and close to the general population on the other measures (Table
IV).
The most commonly reported symptoms during dialysis are itch, hypoten-
sion and muscle cramps, while itch and back pain cause the most severe
discomfort; however, the mean level of stress for all items are towards
the low end of the response scale (Table V). Ninety percent of patients
experience at least one physical symptom during dialysis, with an average
of 2 symptoms causing a mean total stress of 8.15 within a potential range
of 0-99.
It is interesting to compare the three stress/distress scales via the mean
per-item score for the scale as a whole: disease stress seems to be the most
severe intrusion into patient's lives (4.60) followed by psychosocial stress
(3.94) and dialysis distress (3.80). It should also be noted that the standard
Conclusions
regular basis. This data collection is not part of an external research project,
but is added to the routine record keeping of the dialysis unit. Step three
consists of integrating the analysis of quality of life data into the flow of
therapeutic decision making within the unit - the screening of patients into
appropriate treatment modalities, the identification and diagnosis of
psychosocial problems, the choice of intervention, and the deployment of
preventative measures. The sharing of data between centres constitutes
step four. Through this interchange, individual units can compare caseload
characteristics and experiences at an empirical level. Of even greater poten-
tial is the creation of a growing quality of life database that can facilitate
the investigation of unusual problems and special populations, and can support
general research into the interrelationship between physical, psychological and
social functioning, and the determinants of high quality of life for dialysis
patients.
Quality assurance in dialysis represents the consolidation of ethical, prac-
tical and scientific goals in a unified perspective. The promise of quality of
life assurance is this: through the careful and systematic attention to dialysis
as delivered and dialysis as received, medical practitioners have a signifi-
cant contribution to make not only to the quantity of years remaining for the
renal patient, but also to the quality of that time.
Notes
1. The earliest paper known to the authors is: Schreiner GE. Mental and personality changes
in the uraemic syndrome. Med Ann DC 28: 316-323, 362, 1959.
2. Universally applicable in the sense of generally meaningful within the bounds of western
industrial culture; the term assumes a shared world-view and basic commonality of values
that would not be transferrable between major cultures.
References
1. Calland CH: Iatrogenic problems in end-stage renal failure. New Engl J Med 287: 334,
1972.
2. Campbell A, Converse PE, Rodgers WL: The Quality of American Life: Perceptions,
Evaluations, and Satisfactions (New York: Russell Sage Foundation), 1976.
3. Converse PE, Dotson JD, Hoag WJ, McGee WH III: American Social Attitudes Data
Sourcebook, 1947-1978 (Cambridge, Mass: Harvard University Press), 1980.
4. Campbell A: The Sense of Well-Being in America: Recent Patterns and Trends (New York:
McGraw-Hill), 1981.
5. Fries JF, Spitz PW: The hierarchy of patient outcomes. In Spilker B (ed) Quality of Life
Assessments in Clinical Trials (New York: Raven Press, Ltd.), pp 25-35, 1990.
6. Schipper H, Clinch J, Powell V: Definitions and conceptual issues. In Spilker B (ed) Quality
of Life Assessments in Clinical Trials (New York: Reven Press, Ltd.), pp 11-24, 1990.
7. Guyatt GH, Bombardier C, Tugwell PX: Measuring disease-specific quality of life in clinical
trials. CMAJ 134: 889-895, 1986.
8. Spitzer WO, Dobson AJ, Hall J et al.: The QL-index. J Chronic Dis 34: 585-597, 1981.
9. Churchill DN, Torrance D, Taylor DW, Barnes CC, Ludwin D, Shimizu A, Smith EKM:
149
Measurement of quality of life in end-stage renal disease: The time trade-off approach.
Clin Invest Med 10: 14-20, 1987.
10. Bergner M, Bobbit RA, Carter WB et al.: The sickness impact profile. Med Care 19:
787-805, 1981.
11. Morrow GR, Chiarello RJ, Derogatis LR: A new scale for assessing patients' adjustment
to medical illness. Psychol Med 8: 605-610, 1978.
12. Guyatt GH, Jaeschke R: Measurements in clinical trials: Choosing the appropriate approach.
In Spilker B (ed) Quality of Life Assessments in Clinical Trials (New York: Raven Press,
Ltd.), pp 37-46, 1990.
13. Lindsay RM: A comparison of CAPO and haemodialysis in adaptation to home dialysis.
In Moncreith J, Popovich R (eds) CAPD Update - Continuous Ambulatory Peritoneal
Dialysis (Masson, NY: Modern Problems in Kidney Disease), pp 171-179, 1981.
14. Lindsay RM, Oreopoulos 0, Burton H, Conley J, Richmond J, Wells G: The effect of
treatment modality (CAPO vs. haemodialysis) in adaptation to home dialysis. In Atkins
RC, Thomson NM, Farrel PC (eds) Proceedings of the Pan Pacific Conference on Peritoneal
Dialysis (Edinburgh: Churchill Livingstone Press), pp 385-394, 1981.
15. Lindsay RM, Richmond J, Burton H, Conley J, Clark WF, Linton AL: Is home dialysis
too stressful? Controversies in Nephrology III, 395-405, 1981.
16. Wai L, Richmond J, Burton HJ, Lindsay RM: The influence of psychosocial factors on
survival of home dialysis patients. Lancet 2: 1155-1156, 1981.
17. Richmond JM, Linday RM, Burton HJ, Conley J, Wai L: Psychological and physiological
factors predicting the outcome on home haemodialysis. In Cluthe R (ed) Clinical Nephrology
(Frieburg, West Germany), pp 109-113, 1982.
18. Burton HJ, Kline SA, Lindsay RM, Heidenheim AP: The relationship of depression to
survival in chronic renal failure. Psychosomatic Medicine 48(3/4): 261-269, 1986.
19. Lindsay RM: Comparison of patient quality of life in short dialysis and conventional dialysis.
In Man NK, Mion C, Henderson LW (eds) Blood Purification in Perspective: New Insights
and Future Trends. (ISAIO Press), pp 203-204, 1987.
20. Lindsay RM, Heidenheim AP, Spanner E, Burton HJ, Lindsay S, LeFebvre JMJ: A multi-
centre study of short hours (SH) dialysis using AN69S: Preliminary results. ASAIO
Transactions 37(3): M465-M467, 1991.
21. Heidenheim AP, Lindsay RM: The impact of erythropoietin and Dialyzer geometry on patient
quality of life. Submitted to Clinical Nephrology
22. Lindsay RM, Heidenheim AP: Measurement of quality of life in international trials involving
dialysis patients and the use of recombinant human erythropoietin. In Bauer, Koch, Scigalla,
Wieczorek (eds) Clinical Applications of Erythropoietin (New York: Marcel Dekker Inc.),
pp 81-921, 1993.
23. Strauch MR, Lipke R, Schafheutle R, Nachbauer B, Stauch-Rahaiiser G: A standardized
list of somatic criteria for comparative assessment of regular dialysis therapy patients.
Artificial Organs 2 (Suppl): 370-372, 1978.
24. Jackson OJ: The Basic Personality Inventory (Port Huron, Michigan: Research Psychologists
Press), 1976.
25. Nunally JC: Psychometric Theory 2nd ed. (New York: McGraw-Hill), 1978.
26. Carmines EG, Zeller RA: Reliability and validity assessment (Beverly Hills: Sage
Publications), 1979.
27. Selltiz C, Wrightsman LS, Cook SW: Research methods in social relations 3rd ed. (New
York: Holt, Rinehart, Winston), 1976.
28. Helmstadter GC: Principles of Psychological Measurements (New York: Appleton-Centory-
Crofts), 1964.
29. Lemke E, Wiersma W: Principles of Psychological Measurement (Chicago: Rand McNally),
1976.
30. Spilker B: Introduction. In Spilker B (ed) Quality of Life Assessments in Clinical Trials New
(York: Raven Press), p 3, 1990.
CHAPTER 9
L.W. HENDERSON
Introduction
Reuse, since its early introduction by Shaldon [1] as a way to make treat-
ment more economical, has been at the center of controversy. As of this writing
there is a good deal more emotion than fact available on this matter. The issues
that enter the debate are both economic and scientific. I shall not provide a
detailed exploration of them all but rather reference the interested reader to
the source material. I do this as it is clear that one cannot, with the present
incomplete (and conflicting) information, make a clear statement about the
most fundamental concern about reuse; namely, the impact on treatment as
measured by comparative mortality rate. This is a sad commentary on both
U.S. regulatory agencies and North American nephrology, given that in 1990
reuse in one form or another was employed in 75% of the approximately
110,000 patients on treatment with hemodialysis in North America. About half
the reused dialyzers were disinfected manually and half with automated
systems. Formaldehyde was used in 54%, peracetic acid-hydrogen peroxide
in 39%, and glutaraldehyde in 7%.
There are many questions about-reuse. For example, when considering reuse
of cellulosic membrane (Le., the cheapest and most commonly used membrane):
- Does the transport performance of the dialyzer deteriorate with reuse and
if so how may it best be addressed quantitatively? [2]
- Does a reused dialyzer that is "scrubbed fully clean" with bleach have meta-
bolic disadvantages to the user due to restoration of complement activation?
[3]
- Does the new dialyzer contain sufficient manufacturing "by products" e.g.
particulates, cotton linters, residual ethylene oxide, etc. to warrant some
form of pre treatment rinse and, if so, should it contain protein to reduce
compliment activation? [4, 5, 6]
- If a small amount of denatured protein is allowed to remain on the
membrane following reuse in order to reduce complement activation, will
it act as an antigen when it finds its way into the body with subsequent
untoward effects? The reported "reuse syndrome." [7]
For the synthetic membranes that in general are far more active adsorbers of
plasma constituents than the cellulosic membranes:
- Does bleach, the preferred agent to remove all adsorbed protein, etch the
very thin discriminating skin of an asymmetric membrane like polysulphone
so that protein leak results? [8]
- Does failure to use bleach in these protein adsorbing membranes put the
patient at risk from repeated small infusions of potentially antigenic
("foreign") protein denatured by whatever other sterilant is selected?
- Does adsorption of useful plasma proteins that are in short supply (e.g.
erythropoietin, insulin, other?) trade off positively or negatively with the
concern about not using a bleach reprocessing solution? [9]
For all membranes:
- Does the repeated minor re infusion of the chosen sterilant have a long term
morbid or mortal outcome? e.g. cancer.
- Does the patient prevail in his or her wish not to participate in a reuse
program if the economic circumstances of the unit deteriorate as a result?
These many questions become moot if the practice of reuse results in higher
mortality. For this central question we have reports from the Urban Institute
using U.S. Department of Health and Human Services, Health Care Financing
Agency Medicare (HCFA) records and the National Institutes of Health (NIH)
identifying a reduced survival with certain forms of reuse. These are power-
fully large data bases that must command our attention. As these studies
have not, as yet, been published, I am going to offer a summary taken from
recent discussions, presentations and draft documents on the subject.
For the Urban Institute study data from 1,121 free-standing dialysis units (859)
from units using membranes with low hydraulic permeability like curpo-
phane membrane) representing 85% of all free-standing units in 1990 were
examined (i.e., approximately 45,000 patients for 1990 (some 34,000 using
conventional low-flux dialyzers) and a similar number for 1989.
• Patients treated in free-standing dialysis units that employ conventional
non-reused dialyzers had lower mortality on average than did patients treated
in units that used selected germicides for dialyzer reprocessing (analysis
comparing units).
• Patients treated in dialysis units that used peracetic acid-peroxide or
glutaraldehyde as a disinfectant experienced higher mortality, on average
than did those patients treated in dialysis units that used low flux dia-
lyzers that were not reused. For peracetic acid-peroxide manual systems,
the mortality was higher by 15 percent; for Renalin automated systems,
the mortality was higher by 11 percent; for glutaraldehyde systems (both
manual and automated) the mortality was higher by 15 percent, on average
(analysis comparing patients).
153
This study was a "case mix" study which matched approximately 3,100 reuse
and non-reuse patients from 280 dialysis units and "arrayed them according
to a number of medical and demographic characteristics." They showed
increased mortality only when dialyzers were disinfected manually with
peracetic acid-hydrogen peroxide. Automated use did not result in higher
mortality, nor did use of formaldehyde.
In summary, there are internal conflicts in the two studies. Both are retro-
spective and draw from a large sample size. The relative risk of death citation
is sufficiently small so as to be transparent to any given unit director. There
is sufficient variability in the data between units reporting their results to
raise questions in my mind as to the clinical significance of these small average
differences. To further add to the confusion, as earlier study from the USRDS
data base showed a reduced relative risk of death (0.88) for large dialysis
units that had been long term reusers, as contrasted with smaller units that
did not reuse (patient number = 4661) [9].
Conclusions
We must conclude that there is a crying need for more information in the
reuse domain in order to upgrade the quality of the therapy we offer. A specific
set of recommendations for safe reuse is not appropriate at this time, nor is
a recommendation for or against this practice.
References
1. Shaldon SR, Silva R, Rosen SM: Technique of a refrigerated coil of preservation hemo-
dialysis with femoral venous catheterization. Brit Med J 2: 411, 1964.
2. AAMI: Recommended practices for reuse of hemodialyzers (Arlington, VA, Association for
the Advancement Instrumentation), 1986.
154
3. Henderson LW, Chenoweth DE: Immune response to reuse: anaphylatoxins and IgE. In
Deane N, Wineman RJ, Bemis JA (eds) Guide to Reprocessing of Hemodialyzers (Boston:
Martinus Nijhoff Publishers), p 151, 1986.
4. Ogden DA: New dialyzer syndrome. New England Med. 302: 1262, 1980.
5. Ogden DA: Clinical response to new and reprocessed dialyzers. In Deane N, Wineman
RJ, Bemis JA (eds) Guide to Reprocessing of Hemodialyzers (Boston: Martinus Nijhoff
Publishers), p 87, 1986.
6. Pearson FC, Bohon J, Lee, W et 01.: Comparison of chemical analyses of hollow-fiber
dialyzer extracts. Artif Organs 9: 291-298, 1984.
7. Yudis M, Sirota RA, Stein HD: Dialyzer hypersensitivity reactions associated with reuse.
Am Soc Artif lnt Organs 14 (Abstr): 59, 1985.
8. Donahue PR, Ahmad S: Dialyzer permeability alterations by reuse. 1 Am Soc Neph 3 (Abstr):
363, 1992.
9. Cheung AK, Hohnholt M, Leypoldt JK: Hemodialysis membrane biocompatibility: The case
of erythropoietin. 1 Blood Purif9: 153-164, 1991.
10. Held PJ, Pauly MV, Diamond L: Survival analysis of patients undergoing hemodialysis.
lAMA 257: 645-650, 1987.
CHAPTER 10
DAVID N. CHURCHILL
The terms quality assurance and multicenter clinical trials encompass many
elements and the relationships between these elements are complex. The
principles of quality assurance can be applied to the research methodology
used in clinical trials. In turn, the results of these clinical trials can be used
to establish standards for clinical care. Finally, the efficacy of the applica-
tion of those standards in improving clinical care can be evaluated in the
context of a clinical trial.
The term quality assurance has been variously defined. Two definitions
appear to capture the spirit of this concept [1]. The first states that "quality
assurance is a program of systematic evaluation to ensure excellence in health
care". The second is that "the essence of quality assurance consists of the
establishment of standards, evaluation of those standards and implementa-
tion of actions that ensure adherence to those standards".
Use of the term standard, in turn, requires definition. Again, there are several
definitions. The American Nursing Association defines a standard as "an
agreed upon level of excellence; an established norm". Another definition is
"the level of performance considered acceptable by one having authority in the
situation or by those having authority in maintaining such performance levels".
Dependant on the objectives of the group developing the standards, the standard
may represent minimal expectations or it may reflect optimal performance.
A variety of similar but not synonymous terms have been used to describe
the concept of standards. These include standards, guidelines and clinical
practice parameters. These terms each have an inherent sense of flexibility
or, conversely, rigidity.
The title of this section implies that standards exist and can be applied to
clinical trials. A recent analysis of the process by which guidelines or
standards are generated illustrates the heterogenous nature of that process
[2]. Differences exist in the definition of objectives, methods of guideline
development, implementation and evaluation.
L. W. Henderson and R.S. Thuma (eds.), Quality Assurance in Dialysis, 155-166.
© 1994 Kluwer Academic Publishers.
156
Rather than replicate the efforts of Audet et al. [2] and review the process
of guideline or standards development for clinical trials, I have selected one
representative example. The members of the Department of Clinical Epidemi-
ology and Biostatistics at McMaster University in Hamilton have developed
a set of guidelines for the critical appraisal of clinical articles advocating a
specific treatment [3].
The objective defined was to provide, for the reader of the results of clinical
trials, a set of guidelines for the critical appraisal of data published in the
medical literature. The objective, applying the criteria of Audet et al. [2], is
broad rather than specific. A broad objective makes evaluation of the process
difficult.
The method of guideline development was similar to the second model
described by Audet et al. [2]. The process relied on internal group judge-
ment with little use of outside experts to review and evaluate the pertainent
literature. The group consisted of the members of the Department of Clinical
Epidemiology and Biostatistics (clinicians, research methodologists and
biostatistians). The guidelines developed were submitted for peer review
followed by publication.
Implementation involves dissemination, training of individuals in the proper
use of the guidelines and monitoring. Dissemination has been internal and
external. Internal dissemination has been accomplished by incorporation of the
principles of critical appraisal in both the undergraduate and postgraduate
programs in the Faculty of Health Sciences at McMaster University. External
dissemination has been through the Annual Workshop on How to Teach Critical
Appraisal offered since 1981.
Linton and Peachey [4] stated that "Physicians like to believe that they make
logical clinical decisions based on analysis of known data, but this is often
not the case. Clinical policies have been developed on an ad hoc basis from
the considerations and decisions of hundreds of physicians acting individually.
Textbooks and medical journals publish this fragmentary information, and other
contributions come from continuing medical education, institutional policies,
the example of influential physicians and the often biased information provided
by pharmaceutical detailing". They also state that "honest self-evaluation would
suggest that the individual physician data base and analytic abilities are strained
even by apparently trivial clinical problems".
The standard, defined as an informed critical appraisal of published medical
literature pertaining to a specified clinical problem, incorporates the concept
that the published article must be scientifically credible and that the results
be applicable to the clinical problem which stimulated the physician to evaluate
this article. The former deals with the internal validity of the article, the
latter with external validity or generalizability. There are 6 questions or guides
for the physician appraising the article [3]. These evaluable components of
the standard are the criteria (Table I).
The first criterion is that the study deal with human subjects who have
been truly randomly allocated to one of several specified interventions. This
157
Table I. Criteria to evaluate a publication addressing the efficacy of a clinical intervention [3].
Scientific credibility
1. Were the subjects truly randomly allocated to the specified treatments?
2. Were all clinically relevant outcomes reported?
3. Were both clinical and statistical significance considered?
Clinical applicability
4. Were the study patients recognizable?
5. Was the clinical intervention described in sufficient detail to permit replication in clinical
practice?
6. Were all study patients accounted for at the conclusion of the clinical trial?
is considered the gold standard for clinical trials. The methodologic weak-
nesses inherent in less rigorous designs should relegate them to use in two
circumstances. The first is when a randomized controlled trial is not possible
and with a clear statement of the biases in the design. The second is to generate
hypotheses to be tested using the randomized clinical trial methodology. With
the exceptions stated, the randomized clinical trial design is necessary but may
not be sufficient in that the subsequent 5 criteria must also be satisfied.
Before proceeding to the other criteria, the alternative research designs and
their weaknesses will be considered.
The clinical trial hierarchy starts with the case report and the case series.
Observations from these studies may generate hypotheses for more rigorous
examination. Cross-sectional surveys permit examination of relationships
between variables and permit hypothesis generation. The case-control study
design is highly subject to bias [5] with different studies often reaching
different conclusions for the same research question. A more common research
design compares clinical outcomes in a treatment group after an intervention
to a group which has not had the intervention. These can be further divided
into a one group design in which the outcomes in the treatment group are
compared to historical controls, a one group design in which outcomes before
and after the iintervention are compared and finally a design in which the
outcomes before and after the intervention are compared to outcomes in a
contemporaneous group which has not received the intervention [6, 7].
The one group before-after intervention design has been used extensively
in the evaluation of pharmacologic therapy for prevention of recurrent kidney
stones. The biases inherent in this design include co-intervention, regression
to the mean, systematic measurement error and natural history of the disease.
These biases will almost always favour the intervention under evaluation and
might lead the investigator to conclude that the intervention was effective when
it was not [6, 7]. The use of a contemporaneous but non-equivalent compar-
ison (i.e. not randomly allocated) group strengthens the research design but
introduces a selection bias [6, 7]. One can attempt to statistically adjust for
prognostic factors which are potential determinants of clinical outcomes and
which are unequally distributed between groups. Unfortunately, there may
158
There are few formally developed clinical standards in the field of Nephrology.
There are, however, generally accepted treatment practices and these gener-
ally accepted approaches may be considered informal standards or guidelines
for medical practice. Although an explicit application of the criteria evaluating
clinical trials addressing these generally accepted treatment practices was
not incorporated into development of standards, an implicit evaluation process
probably did occur.
What are some generally acceptable treatments? For renal transplantation,
the use of cyclosporine is an effective and widely accepted intervention. For
hemodialysis, the use of prescriptive dialysis based on urea kinetic model-
ling is widely accepted. In the area of continuous ambulatory peritoneal dialysis
(CAPD), the concept of flush before fill either with a disposable disconnect
system or with a disinfectant-multiple use system is considered superior to
conventional systems.
For the transplantation example, a randomized clinical trial of cyclosporine
in cadaveric renal transplantation was performed by the Canadian Multicentre
Transplant Study Group [12]. In that multicentre trial, 209 recipients of
cadaveric renal transplants were treated either with cyclosporine and pred-
nisone or with standard therapy which included prednisone and azothiaprine.
Graft survival at 1 year was 80.4% in patients receiving cyclosporine and
64.0% in patients receiving standard therapy. The authors concluded that
160
clinical centers. Other therapy used for the standard therapy group included
anti-lymphocyte globulin, plasmapheresis, leukopheresis, graft irradiation and
cyclophosphamide. This was not a blinded study and one cannot exclude
differential general care favoring the cyclosporine group.
The final criterion that all patients be accounted for at the conclusion of
the trial was met.
Prior to this publication, single center uncontrolled trials had suggested
that cyclosporine was superior to standard therapy but suffered from the
methodologic problems of all studies without random allocation to treatment
group. Multicenter studies permit enrollment of adequate numbers of patients
within a reasonable time to execute a study with acceptable statistical power
to detect clinically important effects as well as evaluate the effect of other
prognostic factors and their interactions. An additional strength is that the
generalizability of the results is enhanced if the treatment effect is constant
across centers. In this example, 11 of the 12 centers demonstrated better results
with cyclosporine.
This multicenter clinical trial satisfies the criteria for a methodologically
sound study of an intervention. In addition, the multicenter nature of the
study enhances generalizability.
For hemodialysis, the use of urea kinetic modelling is based on the report
of the National Cooperative Dialysis Study (NCDS) in 1981 [14]. The authors
concluded that if the dietary intake of proteins and other nutrients were
adequate, a mid-week pre-dialysis BUN value of 71-73 mg/dl was asso-
ciated with less morbidity than if this value were 105-109 mg/dl. Does this
publication meet the criteria for scientific credibility and clinical applicability?
Were the subjects truly randomly allocated to the specified treatments? There
was random allocation to one of 4 treatment groups, two producing mid-
week pre-dialysis BUN values of 120 ± 10 mg/dl and two producing values
of 70 ± 10 mg/dl. The randomization procedure resulted in no statistically
significant differences among groups with respect to age, history of heart
disease, hypertension, peripheral vascular disease, pulmonary disease, gastro-
intestinal disease or pre-randomization rates of hospitalization.
Were all clinically relevant outcomes reported? One outcome evaluated was
removal from the study for death or medical reasons, the latter defined by
any medical diagnosis on the exit form. A second outcome was hospitaliza-
tion related to uremia. Those not due to uremia (e.g. vascular access problems)
were excluded prospectively. There were only 3 deaths among the 151 patients
randomized. There were 32 patients withdrawn for medical reasons (i.e. due
to uremia). There were 66 withdrawn for other reasons (completed protocol,
patient preference and transplantation). The distribution of these other reasons
was not provided. This was an unblinded study. One might expect a bias in
attributing symptoms to uremia if the BUN value were known to the inves-
tigatory making that decision. For example, if 2 patients had anorexia, one with
a pre-dialysis urea of 70 mg/dl and another with a value of 120 mg/dl, one
would predict that the patient with the higher BUN would be more likely to
162
have that symptom attributed to uremia and be withdrawn form the study. A
similar bias could affect definition of hospitalization due to uremia despite
the authors attempts to prevent bias. An additional outcome of interest which
was not addressed was quality of life &/or uremic symptoms using a reliable
validated questionnaire.
Were both clinical and statistical significance considered? The statistical
significance was evaluated by comparing survival to the event of interest
(i.e. withdrawal or hospitalization) among the 4 groups. In the 2 high BUN
groups, 45 and 62% had been withdrawn by 1 year follow-up compared to
18 and 6% in the 2 low BUN groups. A similar effect was seen for hospital-
ization. The clinical effects were dramatic.
Were the study patients recognizable? The patients were reasonably well
described. They were between the ages of 18 and 70, were treated in a dialysis
center and had a maximum creatinine clearance of 3 mllmin. Those with cancer,
diabetes mellitus, uncontrolled hypertension, revers able renal failure, systemic
disease or unstable cerebro or cardiovascular disease, clinically important
pulmonary or hepatic disease were excluded. There was a stabilization period
of 3 months on standard therapy followed by random allocation to the defined
treatment groups. The duration of ESRD therapy was not provided nor was
there data about race and underlying renal disease. If these were long-term
hemodialysis patients, the prevalence-incidence bias [5] would be operative
and the results less applicable to new dialysis patients. Information regarding
baseline nutritional status would have been helpful.
Was the clinical intervention described in sufficient detail to permit repli-
cation in clinical practice? This criterion was satisfied. The dialysis membranes
used were cuprophane. Although not explicitly stated, the dialysate was
presumably acetate. The dialysis times were 4.5-5.0 hours for the long and
2.5-3.5 hours for the short dialysis. The target BUN values pre mid-week
dialysis were defined. The mathematical model used to achieve these targets
was referenced. All patients were accounted for at the conclusion of the
study.
This was, like the cyclosporine study, a multicenter effort. There were 8
centers in the United States in this study. Although not commented upon in
the results, center was a co-variate in the time to event analyses and was
presumably not a significant factor. There were subsequent analyses of these
data which re-inforced the importance of a minimally acceptable dialytic
removal of small molecules, more commonly expressed as KTIV [15]. The
changes in dialysis technology, particulary with respect to membranes with the
ability to remove potentially toxic middle molecular weight molecules, has
introduced some doubt about the applicability of the urea kinetic modelling
derived from use of cuprophane membranes. Nevertheless, the contribution
of this study to the development of quantitative dialysis prescription is a major
one.
For CAPD, a major clinical problem has been infection, particularly peri-
tonitis. For standard CAPD systems similar to that described by Oreopoulos
163
et al. [16], about 2/3 of peritonitis episodes are due to organisms thought to
gain access to the peritoneal cavity through the lumen of the tubing used to
connect the bags of dialysate to the peritoneal dialysis catheter. The use of a
flush before fill sequence appears to have gained general acceptance as an
effective mode of preventing these infections. There are 2 multicenter ran-
domized clinical trials which have demonstrated the efficacy of the use of a
flush before fill technique with disinfection of the re-usable connecting set
[17, 18].
The first study was reported by Maiorca and colleagues in 1983 [17]. They
found that use of a Y connector disinfectant set was associated with a peri-
tonitis rate of 1 episode per 33 patient-months compared to 1 episode per
11.3 patient months for those using the conventional system.
Were the subjects truly randomly allocated to the specified treatments? There
was random allocation of 30 new CAPD patients to a standard method
and 32 to the Y connector disinfectant set. The randomization produced no
imbalances with respect to age or sex.
Were all clinically relevant outcomes reported? The outcomes of interest
were months of follow-up per episode of peritonitis and time to first episode
of peritonitis. There were only 2 deaths, 1 in each group. There were 11
episodes in 8 patients of accidental infusion of disinfectant into the abdom-
inal cavity, apparently without serious clinical sequelae. Technique survival
is commented upon but the small numbers preluded any statistical evalua-
tion. There was no comment regarding impact on hospitalization rates.
Were both clinical and statistical significance reported? The probability
of remaining peritonitis free for one year was 60% for the Y connector dis-
infectant set compared to 20% for the standard set. Expressed as months of
follow-up for episode of peritonitis, the results were 1 per 33 patient-months
for the former and 1 per 11.3 patient-months for the latter. Both were
statistically significant and clinically important. These findings were dramatic
and far exceeded previous reports in the literature. There are no data in the
methods indicating the reason for selection such a small sample size and this
weakened the impact of this study.
Two of the 3 criteria addressing clinical applicability of the results were
met. The intervention was well described and all patients were accounted
for at the conclusion of the study. However, the patients were not described
other than for sex ratio and mean age. The results section suggests that addition
of drugs to the dialysate was an exclusion criterion.
In this 2 centre study, there was random allocation to alternative forms of
treatment, appropriate determination of relevant outcomes and consideration
of both statistical and clinical importance. Failure to describe the expected
differences between the treatment alternatives, the Type I and Type II error
probabilities in sample size calculation decreased the credibility of the dramatic
difference reported. The patients were not described in sufficient detail to
justify unreserved application to another population of CAPD patients. The
proportion of dialysis patients treated with CAPD in Italy was small compared
164
description of the patients, both eligible and ineligible, made the results more
generalizable to patients in North America.
Selection of these 3 generally accepted treatments in transplantation,
hemodialysis and peritoneal dialysis has been arbitrary. There are other
generally accepted treatments and there may be some who would not agree
that those selected represent treatment standards. The selection of the 4
randomized clinical trials which appear to have played an important role in the
acceptance of these treatments was to illustrate the degree to which these
trials meet the criteria for a standard of excellence in the execution of clinical
research.
References
10. Braitman LE: Confidence intervals assess both clinical significance and statistical signifi-
cance. Ann Intern Med 114: 515-517,1991.
II. Clinical Efficacy Reports. Philadelphia: American College of Physicians; 1987 and ongoing.
Available from the American College of Physicians, Independance Mall West, Sixth Street
at Race, Philadelphia, PA 19106-1572.
12. Canadian Multicentre Transplant Study Group: A randomized clinical trial of cyclosporine
in cadaveric renal transplantation. N Engl J Med 309: 809-815, 1983.
13. Canadian Multicentre Transplant Study Group: A randomized clinical trial of cyclosporine
in cadaveric renal transplantation; analysis at 3 years. N Engl J Med 314: 1219-1225,
1986.
14. Lowrie EG, Laird NM, Parker TF, Sargent JA: Effect of the hemodialysis prescription on
patient morbidity. N Engl J Med 305: 1176-1181, 1981.
15. Gotch FA, Sargent JA: A mechanistic analysis of the National Cooperative Dialysis Study
(NCDS). Kidney Int 28: 526-534, 1985.
16. Oreopoulos DG, Khanna R, Williams P: Continuous ambulatory peritoneal dialysis - 1981.
Nephron 30: 293-303, 1982.
17. Maiorca R, Cantaluppi A, Cancarini GC, Scalamonga A, Broccoli R, Graziani G, Brasa
S, Ponticelli C: Lancet 2: 642-644, 1983.
18. Canadian CAPD Clinical Trials Group: Peritonitis in continuous ambulatory peritoneal
dialysis; a multi-centre randomized clinical trial comparing the Y connector disinfectant
system to standard systems. Peritoneal Dial Int 9: 159-163, 1989.
19. Department of Clinical Epidemiology and Biostatistics, McMaster University Health Sciences
Centre: How to read clinical journals; VI. To learn about the quality of clinical care. Can
Med Assoc J 130: 377-381, 1984.
CHAPTER 11
ROBERT W. STEINER
Introduction
The basics of quality assurance for renal transplant programs are as follows:
(1) regular meetings of relevant transplant program personnel; (2) identifica-
tion of areas of concern; (3) monitoring of transplant program performance
in those areas; (4) initiation of changes to improve performance; (5) evalua-
tion of changes made in response to quality assurance concerns. Confidentiality
of proceedings is a practical requirement for transplant quality assurance
meetings to insure that program deficiencies will receive full and frank analysis.
Any quality assurance activity must be focused on selected issues and rely
on the judgment of the transplant team to direct and prioritize its efforts. As
the purpose of all quality assurance activities is patient benefit, it should be
relatively easy to integrate meaningful quality assurance activities into a renal
transplant program.
The best in quality assurance feature recognition of each individual's desire
to contribute to a better medical outcome, thereby improving morale and
factually educating caregivers as performance is constructively evaluated.
The process defines the broad goals of the organization and helps to meet
those goals by systematically addressing the activities and structure of the
organization with the goal of improving outcome.
The worst in quality assurance results in paperwork and other clerical efforts
which consume designated resources and produce little improvement. The staff
focuses narrowly on statistical goals or other descriptive activity, e.g., moti-
vated by fear of criticism or of regulatory sanctions; the staff primarily attempts
to collect data to document lack of deficiency - overall system improvement
and innovation are not directly addressed. Program changes which are made
are not well evaluated prospectively or retrospectively for efficacy.
The following sections are written from the perspective of continuous quality
improvement, that is, they attempt to set out a schema for program evalua-
tion in terms of program goals.
The overall goal of a renal transplant program can be stated simply: to
maximize long term patient and allograft survival with a minimum of patient
morbidity. The scope of related quality assurance activity can be divided into
eight areas: (a) preoperative recipient screening, (b) management of the living
donor, (c) cadaver kidney quality, (d) surgical care, (e) postoperative medical
care, (f) immunosuppression, (g) long term graft survival, (h) long term medical
care, and (i) patient education and psychological well-being. At present, good
programs often differ significantly in their protocols and practices relative
to the above categories. The following sections will attempt to set out
important issues in quality assurance. To document fully or take a position
on the numerous and important controversies which are addressed is beyond
the scope of this chapter. In many cases, the precise risks and benefits
associated with many of the alternative approaches which are discussed below
have not been quantified.
Preoperative screening
100,........ _______________ _
-MM"'......."',,"""'.:.......................................
=---
--------------- ,
.--.. ----
c: ........
Q)
-
~ 80
--
a.. < 5 years •••••••••••• n = 41
Q) 5-18 years . - - - - - n = 320
19-45 years _ _ _ n=3740
-( lj
a: 46-65 years n = 2342
(lj > 65 years - - - - n = 169
>
.~ 60
:::J
Cf)
40~1~--~~--~~--L--L--L--L--L-~~
o 1 2 3 4 5 6 7 8 9 10 11 12
Month After Transplant
Fig. 1. UNOS patient survival in the first year post transplant. Survival of patients older than 65 is significantly less than survival of patients
19-45 years of age. Most deaths occur in the first 6 months after transplantation. (From reference [4]) - .)
V.l
-
.....
-.l
.j::1..
40
o 2 3 4 5 6 7 8 9 10 11 12
Month After Transplant
Fig. 2. UNOS graft survival at one year based on percent panel reactive antibody (PRA). The outcome for PRA 0-20% differs from that for
PRA 80-100% (p < 0.01). (From reference [4])
175
oriented to procedures which are relatively likely to promote patient well being.
Requirements for extensive preoperative screening may discourage the patient
from seeking transplantation or may result in long delays. Screening proce-
dures such as gastroesophageoduodenoscopy, colonoscopy, cystoscopy, and
arteriography have a low but definite morbidity and add to patient inconve-
nience. These procedures also add to the cost of transplantation. It is difficult
to evaluate the utility of admittedly low yield screening procedures which
are recommended to rule out unlikely and/or inapparent problems in dialysis
patients. For example, the possible occurrence of an undetected preoperative
bladder carcinoma in a transplant recipient may not justify a protocol to
cystoscope candidates routinely. Often, dialysis patients and referring nephrol-
ogists have the choice of several competing local renal transplant programs
and may not choose a program with extensive preoperative requirements. In
this way, competition among competing programs and the proclivity of many
potential recipients to avoid lengthy screening also may limit extensive
screening protocols.
The fundamentals of preoperative screening can be divided with some
overlap into: (a) achieving a complete and current history and physical
examination, (b) cardiac evaluation, (c) detecting chronic infection, (d)
detecting occult neoplasm, (e) characterizing patients immunologically and
identifying immunologically high risk patients, (t) anticipating anatomic
problems, (g) anticipating special medical problems, and (h) assessment of
patient compliance.
A physician knowledgeable in renal transplantation (usually a member of
the transplant team) should generate and review the basic intake patient profile:
medical history, physical examination, and laboratory testing, which should
include detailed head and neck examination, pelvic and rectal examination,
EKG, chest x-ray, and frequently PAP testing and mammography. Referring
physicians should be educated to keep their patient information current at
the transplant center. The discovery of, e.g., a previously unreported history
of active duodenal ulcer, an unexplained change in a cardiogram, or a new
deep-seated foot infection in a diabetic potential recipient just prior to
cadaveric transplantation will often disqualify that recipient, add to graft
cold ischemia time, and unnecessarily strain staff resources or contribute to
inappropriate preoperative decisions. These last minute surprises may be
unavoidable to some extent, but their occurrence should prompt a review of
preoperative evaluation and communication.
Cardiac disease is a major cause of morbidity and mortality for dialysis
patients [5, 6] and for transplant recipients [7]. Many patients may require
no further screening beyond a routine EKG. Asymptomatic patients with
cardiac risk factors such as age, a smoking history, hypertension, and hyper-
lipidemia should have noninvasive testing with a stress, stress thallium or
persantine-thallium study. Many dialysis patients do not have the endurance
to complete cardiac exercise testing [8, 9]. The thallium-persantine cardiac
scan, which does not require exercise, is helpful by most reports [10, 11],
176
but may miss significant disease [12]. Usually patients with positive nonin-
vasive testing go on to angiography and possible coronary bypass grafting
or angioplasty. The strongest indications for thorough cardiac evaluation are
diabetic status and history of MI or angina, but experienced programs may
differ on indications for detailed cardiac evaluation, as discussed in the last
section of this chapter. Selective testing for diabetic recipients based on other
risk factors has been proposed [13, 14].
Patients with certain cardiac risk factors (e.g., arrhythmias or coronary artery
disease) will require special postoperative care. The occurrence of signifi-
cant cardiac morbidity post-operatively should suggest a review of intake
protocols and physician communication. Cardiac events such as myocardial
infarction or sudden death in the first post transplant year may also be a
reflection of failure of pre transplant cardiac screening.
Immunosuppression will exacerbate preexisting chronic infections such as
otitis media, dental abscesses, prostatitis, chronic pyelonephritis, occult
tuberculosis, and osteomyelitis. These abnormalities are usually detected by
intake physical examination and must be corrected before transplantation.
Preoperative cholecystectomy for asymptomatic cholelithiasis or sigmoidec-
tomy for diverticulosis are not required in most programs. The identification
of infection post-operatively that could have been a chronic pretransplant
condition should be viewed as a possible failure of pretransplant screening.
Chronic hepatitis Band C may be exacerbated by immunosuppression but
this condition is not generally considered to be a contraindication to trans-
plantation [15, 16]. Individual pretransplant counseling in patients with
hepatitis may be indicated.
The presence of malignant neoplasm in the recipient is an absolute con-
traindication to renal transplantation. Basic screening for neoplasm includes
chest X-ray and stool guaiac test. Abnormalities in the basic intake profile such
as a thyroid or prostatic nodule, abnormal liver function tests (even at times
in patients with "hepatitis"), or a positive stool guaiac test should be pursued.
A variety of neoplasms may develop in long term dialysis patients [17],
including renal neoplasms [18] but screening ultrasound or CT studies of native
kidneys are not required by many programs. Neoplasms arise de novo with
increased frequency in rental transplant recipients [19], but the emergence
of solid tissue neoplasm in the first year post transplantation suggests a failure
in preoperative screening.
Patients must be accurately tissue typed for HLA and DR antigens, and
the amount of preformed antibody to those antigens quantified (% panel
reactive antibody or PRA). Patients who are at increased immunologic risk
for acute or accelerated rejection are those with PRA greater than 50-80%
and candidates for repeat transplants, especially when previous transplant
survival has been brief [20, 21, 22]. These patients should be identified
prospectively. Many programs believe that these patients should receive induc-
tion therapy with OKT3 or ALG, as discussed later.
Preoperative assessment of anatomic problems must be carried out by
177
Living donorship has become an even more attractive alternative for transplant
programs with growing waiting lists and a shortage of cadaveric donors
[26, 27]. The goals of living donorship (related or unrelated) are (a) a safe
nephrectomy, (b) the absence of current and future renal disease in the donor,
(c) no disease communicated via the allograft, and (d) living kidney donation
which is informed and uncoerced. Donor evaluation protocols differ among
different programs but include a thorough history and physical, laboratory
screening (including HIV, CMV, hepatitis C and HBAG testing), IVP, and renal
arteriography. Cardiac stress testing should be undertaken in older donors
178
and should be at least as complete as for recipients with similar risk factors.
In general, the presence of any renal disease rules out donorship. The presence
of low grade hematuria which has been completely evaluated in a young
potential donor also can present a difficult decision, as it often has a benign
prognosis [28]. The finding of borderline hypertension also presents a diffi-
cult decision as to donor acceptability. Within limits, potential donors can
elect to donate in borderline cases after detailed discussion. Occasionally,
psychosocial evaluation can uncover unacceptable family dynamics involving
coercion or other unusual incentives to a potential donor. An attempt should
be made to follow living donors postoperatively for several years to assess
the adequacy of preoperative evaluation.
Cadaver kidney quality will become more of an issue as cadaveric waiting lists
lengthen. Cadaver kidneys will increasingly be taken from older donors, donors
with hypertension, and donors with possible acute ATN [29]. UNOS and
regional organ procurement agencies may at times offer cadaveric kidneys from
donors of marginal quality, to be accepted for transplantation at the discre-
tion of the individual program. Poorer cadaveric kidney quality may result
in a higher postoperative dialysis rate, a lower one year graft survival, and poor
quality of graft function long term [29, 30, 31]. These same complications may
of course arise from factors which are independent of cadaver kidney quality:
deficient preservation techniques, increased warm or cold ischemia time [32]
cyclosporine toxicity, postoperative ATN, early allograft rejection, inade-
quate immunosuppression, anatomic problems from the surgery itself, and poor
patient compliance. Complications from allograft harvest which are not strictly
donor dependent include shortened or traumatized renal arteries, veins, or
ureters, inadequate preservation of ureteral blood supply, failure to identify
or preserve multiple renal arteries, and prolonged warm ischemia time.
Surgical care
The chief goals in postoperative medical renal transplant care are (a) man-
agement of known preoperative risk factors, (b) safe and effective dialysis
support, (c) infection management, (d) management of immunosuppression,
and (e) patient psychological well-being. Preoperative risk factors can be
divided into general risks and patient-specific factors. General risks stem
from the prevalence of cardiac disease in the dialysis production [5, 6, 8-14],
the effect of age on post transplant mortality and morbidity [37-45], and the
increased incidence of fluid, electrolyte, and hemorrhagic complications.
Volume expansion immediately postoperatively in an effort to help allograft
function [46, 47], may increase left ventricular work and risk myocardial
ischemia [48], cardiogenic pulmonary edema [49], and noncardiogenic pul-
monary edema with OKT3 [50].
Preoperative patient-specific conditions that might lead to postoperative
therapeutic urgencies include subtle preoperative glucose intolerance, par-
oxysmal cardiac arrhythmia, chronic seizure disorder, diverticulosis, and drug
allergy. The need for postoperative dialysis support varies widely. Hemodialysis
associated hypotension, bleeding (with heparin), and sudden loss of access can
be minimized by planning and attention to detail. Post transplant peritoneal
dialysis poses special risks of infection with immunosuppression, dialysate
leakage through the wound, and hyper-glycemia. Patient-specific problems
include dialyzer allergies, CAPD equipment incompatibilities, and reduction
of peritoneal volume after transplantation in small CAPD patients. The
therapeutic goal in this area is communication of the details of each patient's
preoperative status and continual postoperative awareness of same by the
medical team.
Prophylaxis and treatment for infection that arises in the immediate post
transplant period involve several areas. Cytomegalovirus (CMV) infection is
particularly common and can be particularly severe in seronegative recipi-
ents who receive kidneys from seropositive donors and in patients who receive
increased immunosuppression [51-53]. CMV prophylaxis is indicated in
high risk groups [54, 55] and is employed by some centers in all patients in
the postoperative period. Acyclovir, gancyclovir, or immune globulin are
used for CMV prophylaxis [54]; Gancyclovir is indicated for invasive CMV
disease [52] but not necessarily for viremia alone, which is much more common
[56]. Acyclovir is effective prophylaxis and treatment for herpes simplex
infection [57, 58]. Other factors which influence the occurrence and severity
of infection are delays in diagnosis and in decreasing or discontinuing immuno-
180
Immunosuppression
CSA levels tend to fluctuate early in the post-transplant period and cannot
be predicted accurately for a given patient from a standard intravenous or
oral dose [72]. Patients who receive daily cyclosporine (CSA) as initial therapy
must have accurate, daily trough levels, which can be monitored by one
of several techniques [73]. Although the precise value of CSA levels is
unsettled, in the absence of antilymphocyte (induction) therapy, adequate CSA
and steroid dosage is critical in the early post transplant period, to minimize
early graft loss [74].
The amount of time and effort spent on quality assurance activities varies
widely among renal transplant programs in the United States. Financial
considerations or lack of overt program commitment to quality assurance
may limit quality assurance activities. The present standard of practice in
this area is therefore somewhat vague and broad. A proposed basic plan for
the organization and activity of a renal transplant quality assurance program
is outlined in Table II. The categories for data collection closely relate to
the eight areas outlined in the section "Scope of Program Evaluation" above.
The emphasis given by such a quality assurance committee to data from each
area would vary from program to program and would depend on individual
program strengths and weaknesses and designated priorities. Other data
categories such as patient death, readmissions, early graft failure, and reop-
erations, will fall into the data categories already proposed in Table I. Although
the process of continuous quality improvement postulates that all areas of a
transplant program are subject to review and improvement by the transplant
team, emphasis and priority depend on available resources and perceived needs
in these different areas.
The difficulty in setting out specific protocols or standards for adequate
quality care of the transplant patient is illustrated by the following exercise.
Eight renal transplant programs in the United States were surveyed by tele-
phone. Two of these programs performed 80 to 100 transplants per year, and
the rest between 100 and 200 transplants per year. A physician or a trans-
plant coordinator who was knowledgeable in program practices and procedures
was asked a series of standardized questions concerning preoperative evalu-
ation procedures for potential candidates for renal transplantation and
postoperative care. The results are displayed in Table III in the text. Some
A. Committee members: medical and surgical staff members and transplant coordinators; floor
nursing or administration where appropriate.
B. Meet regularly (monthly or bimonthly).
C. Gather and analyze data on:
(l) medical complications (cardiac, infectious, and dialysis related, related to inadequate
pretransplant evaluation, physician or communication, or to postoperative care).
(2) surgical complications (arterial or venous thrombosis, ureteral breakdown or stenosis,
wound infection or dehiscence, bleeding, biopsy complications).
(3) early allograft dysfunction (diagnosis, cadaver donor selection).
(4) rejection (incidence, efficacy of baseline immunosuppression, diagnosis and treatment
for rejection, complications of treatment).
(5) patient psychological status.
(6) long term complications (neoplastic, cardiac, infections, patient compliance).
(7) long term allograft function.
D. Institute changes where appropriate.
E. Evaluate effect of changes.
184
Table III. Patient evaluation and management: a survey of 8 large renal transplant programs.
Infection prophylaxis:
Transplant CMV positive kidneys into CMV negative recipients?: Yes 7; No 1
Routinely prophylax with acyclovir: Yes 4; No 1; High risk patient only 1
Immunoglobulin routinely used for CMV prophylaxis?: No 7; Occasionally 1
Trimethoprim-sulfamethoxazole prophylaxis for PCP or UTI: Yes 6; No 2
Cardiac evaluation:
Cardiac stress testing in a hypothetical 45-year-old male on hemodialysis for two years without
risk factors: Usually 2; Occasionally I; No 5
Diabetic patients: noninvasive "stress" testing (exercise tolerance test or persantine thallium scan):
Yes 8; No 0
Estimated number of patients in whom pretransplant cardiac bypass grafting or angioplasty would
be recommended: 5% or less 6; 5-10% 1
(All programs require baseline EKG and would pursue additional cardiac testing if EKG abnormal
or if patient had suggestive symptomatology.)
cols were not tabulated, but given the present day differences of opinion in
the literature, one would expect to see differences in immunosuppression
protocols if a similar survey were conducted. These results reinforce the
conclusion that acceptable practice may vary considerably in renal transplant
programs at the present time.
Acknowledgments
Nicholas Halasz, M.D. reviewed this manuscript and made helpful suggestions.
Dorothea Rehman diligently aided in manuscript preparation.
Note
1. The data in the UNOS 1988 & 1989 Annual Report on the Scientific Registry and the
Organ Procurement and Transplantation Network have been supplied by transplant
facilities and Scientific Registry subcontractors who continually upgrade and supply new
information to UNOS. The data reflect the database on April 7, 1990 and are subject to
change. The interpreting and reporting of data are the sole responsibility of the authors.
References
1. Berwick DM: Sounding board. Continuous improvement as an ideal in health care. New
Engl J Med 320(1): 53-56, Jan 1989.
2. Renal Transplants and Renal Transplant Centers: Annual Report on the U.S. Scientific
Registry for Organ Transplantation and the Organ Procurement and Transplantation Network,
1988 and 1989. UNOS, Health Resources and Services Administration, and the Division
of Organ Transplantation, p 40 Aug 1990.
3. Accreditation Manual for Hospitals: The Joint Commission on Accreditation of Healthcare
Organizations. Chicago, 1990.
4. Transplant Outcomes: Annual Report on the U.S. Scientific Registry for Organ
Transplantation and the Organ Procurement and Transplantation Network 1988 and 1989.
UNOS, Health Resources and Services Administration, and the Division of Organ
Transplantation, pp 69-74, Aug 1990.
5. Parfrey PS, Harnett JD, Barre PE: The natural history of myocardial disease in dialysis
patients. J Amer Soc Neph 2(1): 2-12, 1991.
6. Rostand SG, Gretes JC, Kirk KA et al.: Ischemic heart disease in patients with uremia
undergoing maintenance hemodialysis. Kidney Int 16: 600-611, 1979.
7. Hill MN, Grossman RA, Feldman HI et al.: Original investigations. Changes in causes of
death after renal transplantation, 1966 to 1987. Amer J Kidney Dis XVII(5): 512-518,
May 1991.
8. Philipson JD, Carpenter BJ, Itzkoff J et al.: Evaluation of cardiovascular risk for renal
transplantation in diabetic patients. Am J Med 81: 630-634, 1986
9. Holley JL, Fenton RA, Arthur RS et al.: Thallium stress testing does not predict cardio-
vascular risk in diabetic patients with end-state renal disease undergoing cadaveric renal
transplantation. Am J Med 90: 563-570, 1991.
10. Brown KA, Rimmer J, Haisch C: Noninvasive cardiac risk stratification of diabetic and
nondiabetic uremic renal allograft candidates using dipyridamole-thallium-201 imaging
and radionuclide ventriculography. Amer J Card 64: 1017-1021, Nov 1989.
186
11. Kletter K, Derfler K, Balcke P et al.: Thallium-20 1 dipyridamole (DPM) stress and
redistribution scintigraphy in hemodialysis (HD) patients and renal transplant recipients
(RTRs): 6 year follow-up study, J Am Soc Neph 1 (Abstr): 292, 1990.
12. Marwick TH, Steinmuller DR, Underwood DA et al.: Ineffectiveness of dipyridamole
spect thallium imaging as a screening technique for coronary artery disease in patients
with end-stage renal failure. Transplantation 49(1): 100-103 Jan 1990.
13. DeVault GA Jr, Wynn SW, Swanson S et al.: The impact of routine coronary screening
in diabetic candidates for renal transplantation. (Abstract) 10th Annual Meeting, American
Society of Transplant Physicians, Chicago, 1991.
14. DeVault GA Jr, Wynn SW, Holman JM Jr et al.: Coronary risk assessment limits the need
for routine coronary screening in diabetic candidates for renal transportation. (Abstract) 10th
Annual Meeting, American Society of Transplant Physicians, Chicago, 1991.
15. Parfrey PS, Forbes RDC, Hutchinson TA et al.: The clinical and pathological course of
hepatitis B liver disease in renal transplant recipients. Transplantation 37(5): 461-466
May 1984.
16. Huang C, Lai M, Fong M: Hepatitis B liver disease in cyclosporine-treated renal allograft
recipients. Transplantation 49(3): 540-544, 1990.
17. Port FK, Ragheb NE, Schwartz AG et al.: Neoplasms in dialysis patients: A population-
based study. Amer J Kidney Dis XIV(2): 119-123, Aug 1989.
18. Ishikawa I, Saito Y, Shikura N et al.: Ten-year prospective study on the development of
renal cell carcinoma in dialysis patients. Amer J Kidney Dis XVI(5): 452-458 Nov 1990.
19. Penn I: The changing pattern of posttransplant malignancies. Transplantation Proceedings
23(1): 1101-1103 Feb 1991.
20. Iwaki Y, Iguro T, Terasaki PI: Effect of sensitization on kidney allografts, in Terasaki PI
(ed) Clinical Kidney Transplants 1985 (Los Angeles, CA: UCLA Tissue Typing Laboratory),
pp 139-145, 1985.
21. Perdue ST: Risk factors for second transplants, in Terasaki PI (ed) Clinical Kidney
Transplants 1985 (Los Angeles, CA: UCLA Tissue Typing Laboratory), pp 191-203,
1985.
22. Almond PS, Matas AJ, Gillingham K et al.: Risk factors for second renal allografts
immunosuppressed with cyclosporine, Transplantation 52(2): 253-258, Aug 1991.
23. Holley JL, Shapiro R, Lopatin WB et al.: Obesity as a risk factor following cadaveric
renal transplantation. Transplantation 49(2): 387-389, 1990.
24. Garvin PJ, Castadeda M, Linderer R et al.: Management of hypercalcemic hyperpara-
thyroidism after renal Ttransplantation. Arch Surg 120: 578-583, 1985.
25. Dunn J, Golden D, Van Buren CT et al.: Causes of graft loss beyond two years in the
cyclosporine era. Transplantation 49(2): 349-353 Feb 1990.
26. Bay WH, Hebert LA: The living donor in kidney transplantation. Ann Int Med 106(5):
719-727, 1987.
27. Bertolatus JA: Renal transplantation for the nephrologist: Living donor kidney transplan-
tation: What did we learn during the 1980s? What should we learn during the 1990s? Am
J Kid Dis XVII: 596-599, 1991.
28. Mohr DN, Offord KP, Owen RA et al.: Asymptomatic microhematuria and urologic disease.
A population-based study. JAMA 256(2): 224-229, July 1986.
29. Rosenthal JT, Miserantino DP, Mendez R et al.: Extending the criteria for cadaver kidney
donors. Transplant Proc 22(2): 338-339, 1990.
30. Rao KV, Kasiske BL, Odlund MD et al.: Influence of cadaver donor age on post-
transplant renal function and graft outcome. Transplantation 49(1): 91-95, 1990.
31. Leunissen KML, Bosman FR, Nieman FHM et al.: Amplification of the nephrotoxic effect
of cyclosporine by preexistent chronic histological lesions in the kidney. Transplantation
48(4): 590-593, 1989.
32. Finn WF: Prevention of ischemic injury in renal transplantation. Kidney Int 37: 171-182,
1990.
33. Tilney NL, Kirkman RL: Surgical aspects of kidney transplantation. In Garavoy MR,
187
to prevent cytomegalovirus disease in renal transplant recipients. New Engl J Med 317:
1049-1054, 1987
56. Van Den Berg AP, Van Der Bij W, Wan Son WJ et al.: Cytomegalovirus antigenemia as
a useful marker of symptomatic cytomegalovirus infection after renal transplantation - a
report of 130 consecutive patients. Transplantation 48(6): 991-995, 1989.
57. Wade JC, Newton B, Flournoy N et al.: Oral acyclovir for prevention of herpes simplex
virus reactivation after marrow transplant. Arner Intern Med 100: 823, 1984.
58. Tomlanovich SJ, Sabatte-Caspill, J, Melzer J et al.: The incidence and impact of herpes
simplex virus infections in the first month following renal transplantation. Transplantation
Proceedings 21(1): 2091-2092, Feb 1989.
59. Rubin RH, Tolkoff-Rubin NE: Infections: The new problems. Transplant Proc 21(1):
1440-1445, 1989.
60. Maddux MS, Veremis SA, Bauma WD et al.: Effective prophylaxis of early post-
transplant urinary tract infections (UTI) in the cyclosporine (CSA) era. Transplantation
Proceedings 21(1): 2108-2109, Feb 1989.
61. Higgins RM, Bloom SL, Hopkin JM et al.: The risks and benefits of low-dose cotrimoxa-
zole prophylaxis for pneumocystis pneumonia in renal transplantation Transplantation 47(3):
558-560, 1989.
62. Shield CF: Use of OKT3 as prophylaxis in cadaveric renal transplantation. Transplantation
Proceedings 21(6) (Suppl 2): 15-18, 1989.
63. Monaco AP: Renal prophylaxis with orthoclone OKT3 in the United States. Transplantation
Proceedings 21(6) (Suppl 2): 7-13, 1989.
64. Melzer JS, Husing RM, Feduska NJ et al.: How to improve results for second renal allo-
grafts. Transplantation Proceedings XX(3) (Suppl 3): 176-179, June 1988.
65. Schroeder TJ, First MR, Mansour ME et al.: Prophylactic use of OKT3 in immunologic
high-risk cadaver renal transplant recipients. Arner J Kidney Dis XIV(5) (Suppl 2): 14-18
Nov 1989.
66. Benvenisty AI, Cohen D, Stegall MD et al: Improved results using OKT3 as induction
immunosuppression in renal allograft recipients with delayed graft function. Transplantation
49(2): 321-327, 1990.
67. Cohen DJ, Benvenisty AI, Cianci J et al.: OKT3 prophylaxis in cadaveric kidney trans-
plant recipients with delayed graft function. Am J Kidney Dis XIV(5) (Suppl 2): 19-27,
Nov 1989.
68. Michael HJ, Francos GC, Burke JF et al.: A comparison of the effects of cyclosporine versus
antilymphocyte globulin on delayed graft function in cadaver renal transplant recipients.
Transplantation 48(5): 805-808, 1989.
69. Pearson RC, Johnson RW, Bakran A et al.: A prospective study of prophylactic ATG
versus cyclosporine in regrafted and highly sensitized renal allograft recipients. Trans-
plantation 50(6): 1061-1063, 1990.
70. Todd PA, Brogden, RN: Muromonab CD3. A review of its pharmacology and therapeutic
potential. Drugs 37: 871-899, 1989.
71. Cockfield SM, Preiksaitis J et al.: Is sequential use of ALG and OKT3 in renal trans-
plants association with an increased incidence of fulminant posttransplant
Iymphoproliferative disorder? Transplantation Proceedings 23: 1:11 06-11 07, 1991.
72. Kahan BD: Individualization of cyclosporine therapy using pharmacokinetic and pharma-
codynamic parameters. Transplantation 5: 457, 1985.
73. Kahan BD, Shaw LM, Holt D et al.: Consensus document: Hawk's Cay meeting on
therapeutic drug monitoring of cyclosporine. Clin Chern 36(8): 1510-1516, 1990.
74. Kahan BD, Mickey R, Flechner SM et al.: Multivariate analysis of risk factors impacting
on immediate and eventual cadaver allograft survival in cyclosporine-treated recipients.
Transplantation 43(1): 65-70, 1987.
75. Salomon D, Brunson M, Vans ickier J et al.: A retrospective analysis of late renal graft
function: Correlation with mean cyclosporine levels and lack of evidence for chronic
cyclosporine toxicity. Transplantation Proceedings 23(1): 1018-1019, 1991.
189
76. Wrenshall LE, Matas AJ, Canafax DM et al.: An increased incidence of late acute rejec-
tion episodes in cadaver renal allograft recipients given azathioprine, cyclosporine, and
prednisone. Transplantation 50(2): 233-237, 1990.
77. Vanderwerf BA, Serota AI: Low-dose cyclosporine for cadaveric renal transplantation.
Transplantation 45(2): 320-323, 1988.
78. Delmonico FL, Conti D, Auchincloss H Jr et al.: Long-term, low-dose cyclosporine treat-
ment of renal allograft recipients. Transplantation 49: 899-904, 1990.
79. Lewis RM, Van Buren CT, Radovancevie B et al.: Impact of long-term cyclosporine
immunosuppressive therapy on native kidneys versus renal allografts: Serial renal function
in heart and kidney transplant recipients. J Heart and Lung Transplant 10(1): 63-70, Jan/Feb
1991.
80. Brinker KR, Dickerman RM, Gonwa TA et al.: A randomized trial comparing double-
drug and triple-drug therapy in primary cadaveric renal transplants. Transplantation 50(1):
43-49, 1990.
81. Kramer NC, Peters TG, Rohr MS et al.: Benefician effect of cyclosporine on renal trans-
plantation. Transplantation 49(2): 343-348, 1990.
82. Hall BM, Tiller DJ, Hardie I et al.: Comparison of three immunosuppressive regimens in
cadaver renal transplantation: Long-term cyclosporine, short-term cyclosporine followed
by azathioprine and prednisolone, and azathioprine and prednisolone without cyclosporine.
New Engl J Med 313(23): 1499-1507, 1988.
83. Helling TS, Wood BC, Nelson PW et al.: Elective conversion from cyclosporine to aza-
thioprine in sensitized patients following cadaveric renal transplantation. Am J Kidney Dis
XV(2): 137-140, 1990.
84. Schulak JA, Mayes JT, Moritz CE et al.: A prospective randomized trial of prednisone versus
no prednisone maintenance therapy in cyclosporine-treated and azathioprine-treated renal
transplant patients. Transplantation 49(2): 327-332, 1990.
85. Reisman L, Lieberman KV, Burrows L et al.: Follow-up of cyclosporine-treated pediatric
renal allograft recipients after cessation of prednisone. Transplantation 49(1): 76-80, 1990.
86. Zanker B, Walz G, Wieder KJ et al.: Evidence that glucocorticosteroids block expression
of the human interleukin-6 gene by accessory cells. Transplantation 49(1): 183-185, Jan
1990.
87. Furue M, Ishibashi Y: Differential regulation by dexamethasone and cyclosporine of human
T cells activated by various stimuli. Transplantation 52(3): 522-526, 1991.
88. Griffin PJA, Gomes Da Costa CA, Salaman JR: A controlled trial of steroids in cyclosporine-
treated renal transplant recipients. Transplantation 43(4): 505-508, 1987.
89. Meyers BD, Sibley R, Newton L et al.: The long-term course of cyclosporine-associated
chronic nephropathy. Kidney Int 33: 590-600, 1988.
90. Mihatsch MJ, Thiel G, Ryffel B: Histopathology of cyclosporine nephrotoxicity. Transplant
Proc 20 (Suppl 3): 759-771, 1988.
91. Kasiske BL, Heim-Duthoy K, Venkateswara R et al.: The relationship between cyclosporine
pharmacokinetic parameters and subsequent acute rejection in renal transplant recipients.
Transplantation 46(5): 716-722, 1988.
92. Mathew TH: In-depth review. Recurrence of disease following renal transplantation. Am J
Kidney Dis XII(2): 85-96, 1988.
93. Stevenson JA, Dumke A, Glassock RJ et al.: Thrombotic microangiopathy: recurrence
following renal transplant and response to plasma infusion. Am J Nephrol 2: 227-231,
1982.
94. Hebert D, Sibley RK, Mauer SM. Recurrence of hemolytic uremic syndrome in renal
transplant recipients. Kidney International 30: S-51-5-58, 1986.
95. Cameron JS, Senguttuvan P, Hartley B et al.: Focal segmental glomerulosclerosis in fifty-
nine renal allografts from a single centre: Analysis of risk factors for recurrence. Transplant
Proc 21(1): 2117-2118, 1989.
96. Bruan WE: Long-term complications of renal transplantation. Kidney Int 37: 1363-1378,
1990.
190
97. Mahony JF, Caterson RJ, Pollock CA et al.: Coronary artery disease is the major late
complication of successful cadaveric renal transplantation. Clin Transplantation 4: 129-132,
1990.
98. Kasiske BL, Tortorice KL, Heim-Duthoy KL et al.: Lovastatin treatment of hypercholes-
terolemia in renal transplant recipients. Transplantation 49(1): 95-100, 1990.
99. Glicklich DG, Tellis VA, Quinn T et al.: Comparison of captopril scan and doppler ultra-
sonography as screening tests for transplant renal artery stenosis. Transplantation 49(1):
217-219, 1990.
100. Palmer BF, Dawidson I, Sagalowsky A et al.: Improved outcome of cadaveric renal trans-
plantation due to calcium channel blockers. Transplantation. 52(4): 640-645, 1991.
101. Kirk AJB, Omar I, Dark JH: Long-term improvement in renal function using nifedipine
in cyclosporine-associated hypertension. Transplantation 50(6): 1061, Dec 1990.
102. Shapiro R, Carroll PB, Andreas GT et al.: Adrenal reserve in renal transplant recipients with
cyclosporine, azathioprine, and prednisone immunosuppression. Transplantation 49(5):
1011-1013, 1990.
CHAPTER 12
RICHARD S. THUMA
Introduction
CQI
Engineering
Redesign
Resolve
wlvendor
Raw Mafl
Component
Inventory
+ +
glllllllX B!!~Ul M!!!!tlOgs
• Field Problems Review
• Revlaw of returned product analysis results
• Review of rework Issues
• Revlaw of QC Data
Fig. 4. A tabulation of factors that influence the quality strategies used in the manufacture of peritoneal dialysis solutions.
......
10
--.J
198
of the container are joined, using a sealing process. The integrity of the
seal is a function of the method chosen for forming the seal as well as
the composition and thickness of the container material. Some semi-
rigid containers are blow molded in one piece so seal integrity is an
issue only when access ports are added to the container.
(4) Solvent constraints - solvents are used routinely to seal medication and
access ports to the container. Most manufacturers use cyclohexanone, a
compound whose general use is coming under some scrutiny in some parts
of the world; cyclohexanone is prohibited in Japan, so suppliers in Japan
use Ultrasonic (RF) sealing strategies - which are usually more expen-
sive and difficult to control than solvent systems.
(5) Leachables - because the PD solutions are in contact with the container
for potentially long periods of time (most manufacturers put two or more
years of dating on their products), the plastic formulation must be shown
to be non-toxic and the long-term leachability of plastic components and
plasticizers must be taken into account in the design of the container
and choice of materials.
PVC has been used for many years in the production of medical devices
and as containers for solutions. This is due partly to its excellent bio-
compatibility, ease of processing, and several other characteristics. Most
of the medical grade plasticized PVC is consumed in the manufacture
of containers, gloves, and tubing. In the manufacture of flexible bags
for solutions, DEHP is the only plasticizer used routinely.
DEHP is the most widely used plasticizer in flexible PVC for medical
applications and its toxicological properties have been examined most
extensively. The acute toxicity of DEHP is very low.
When receiving fluids stored in PVC containers, patients may be
exposed to plasticizers; PVC tubing used either for administration of fluids
or for dialysis may also contribute to the exposure.
Data on DEHP exposure in patients undergoing medical treatment show
a wide variation. One contributor to this variation is explained by the
fact that the reliability of the analytical methods differ. However, recent
evidence estimates exposure by Hemodialysis patients at 50-370 mgt
treatment [2, 3] and Peritoneal Dialysis patients at 0.1-20 mg/day
[4, 5]. It should be noted that even the highest exposures of DEHP in
dialysis patients provides a significant margin of safety when compared
with levels of exposure which produced no effect in animal studies [6].
(6) Pyrogens - both the solutions and components must be non-pyrogenic
so pyrogen testing is a part of every manufacturers process. The pyrogen
test most frequently employed is the Limulous Amebocyte Lysate Test
(LAL test), an in-vitro test for bacterial endotoxins. The test detects the
presence of bacterial endotoxins by the formation of protein clots.
Extensions of the test include use of enzyme-linked reagents that cleave
chromophores in the presence of the endotoxins and allow a direct
colorimetric measure of the presence of the pyrogenic substances.
199
From a quality assurance point of view, the LAL test requires appro-
priate reference materials as positive controls, proper training of personnel,
and, perhaps most important of all, pyrogen-free glassware and equipment
used in the assay.
The LAL test will detect only bacterial endotoxins and will therefore
be completely ineffective at detecting chemical pyrogens, therefore, in
some areas of the world, the rabbit pyrogen test is still performed.
(7) Trace elements - can be a concern, especially aluminum, and must be
minimized. This is usually accomplished through control of water used
to make the solutions and very tight control of the analyte raw mate-
rials.
(8) Bag clarity - the end user must be able to determine acceptability of
the solution (cloudy solutions may signal bacterial contamination or
presence of particulate matter). Since this is facilitated by the clarity of
the bag, the formulation of the container is important. As mentioned
earlier, most blow-molded and semi-rigid containers are not transparent,
thus making it difficult for the end user to determine solution acceptability.
B. Filling Process
(1) Water quality - pH, trace element content, and pyrogenicity all affect the
final product. Water processing and testing is a critical step in the
manufacturing process. Location of the manufacturing plant is determined
by a number of economic and logistical factors, including the avail-
ability of good water.
In the manufacturing process, the pH of the solution is lowered to about
5.5 to prevent caramelization of the glucose during sterilization.
(2) Filter and filter testing - filter testing is required before use and is accom-
plished by using a pressure test (bubble test). This assures filter pack
integrity. The objective of filtration is not sterilization but rather, by
passing the product through a series of filters of decreasing pore sizes,
elimination of particulate matter.
(3) Fill volume variations - the filling process uses automated filling machines
that may be fed manually or automatically. Checks on fill volume are
carried out by weighing samples at set intervals. Process control charts
(X-bar and R-charts) are common methods used to track fill variations
during the filling process. Filled container weight variations that fall
outside the control limits for the process trigger immediate corrective
action.
(4) Manufacturing environment - filling generally takes place in a controlled
access, class 100,000 or class 10,000 clean room to control bioburden
and particulate matter. Access to the room is usually restricted and may
require donning special lint-free uniforms, shoe covers, head covers, and
gloves.
200
C. Overpouch
D. Sterilization
Here, many of the same issues covered in previous sections apply: choice of
packaging materials, load patterns on pallets, sealing of packing materials,
labeling, storage effects and pest control.
Hemodialyzer manufacturing
For the purposes of this discussion, only hollow fiber (capillary) dialyzers
will be considered. Conceptually, the dialyzer is a tubular container
separated into two compartments: a dialysate compartment and a blood com-
partment. A semipermeable membrane, made up of a bundle of thousands of
small, hollow fibers, separates the two. Access to the two compartments is
through two pairs of ports.
In hemodialysis, the blood flows from the patient into a port at one end
of the dialyzer and is channeled through the hollow fibers and back to the
patient. The dialysate solution flows around the outside of the fibers.
A number of advantages and disadvantages of hollow-fiber dialyzers can be
articulated [7] today with most attention focused on the membrane materials.
In general, membrane materials are classified into one of three types: cellu-
loses (cuprammonium cellulose, cuprammonium rayon), modified celluloses
(cellulose acetate and triacetate), and synthetics (polyacrylonitrile, PAN,
poly sulfone, polymethylmethacrylate). The various membrane types differ in
their physiologic effect. For purposes of this discussion, it is only important
to recognize that the membrane type determines many of the manufacturing
and quality control strategies.
Similar to the manufacture of PD solutions, there exist many factors that
influence the quality and production strategies used to produce dialyzers. A
number of these factors are tabulated in Figure 6 and described in more detail
below.
The manufacturing process flow for dialyzers can be generalized into the
steps tabulated in Figure 7. A number of these process steps are only relevant
for highly vertically integrated producers.
1:5
N
A••emble Flnalln.pactlon
Raw Mat'l. Fiber Membrane Potting, Cutting Packaging. Storage and
and Sterilization
Components Winding Treatment Leak Te.tlng and Cleaning Shipping
o Component o Fiber Type o Lumen Ruld o Header Attach- o Type of o Sterile Barrier o Sampling Plan
Variation o Breaks o Solvent Type ment Method Compound Choice o Pyrogen
o Defects o Winding Method o Method o Tip Protector o Potting Method o Sterilization Testing
oVendor o Number of o Bundle Insertion Material o Polymerization Method o Leak Testing
Approvals spools Method o Stress Cracking Time o Effect on Mat'l o PM Testing
o Incoming oTension Control o Potting Cap - o Tip Protector • Cure Time o Safety o Performance
Manufacturing Inspections • Process Speed Plastic Type Fit • Cutting Method oLoad Pattems Testing
Isaue. o Pilot Lots • Bundle Wrap o Dryvs. Wet o Cutting Blade o Cycle Variations • Storage
o Injection molding o Pressure Parameters o Lumen Fluid Conditions
Decay vs. Gas o Durometer of o Degassing Time • Carton Sealing
Composition. Cured Potting o Pest Control
Compound • Distribution
Method
oTraceability
oOverallGMP
Compliance
o Leachables oHeader Leaks o Toxicology o Sterility o Pyrogens
Therapy o Plastic Blends o Tip Protector Fit o Leachables o Brittle Fibers • Performance
Issu•• o Pyrogens o Sterility of Blood • EtO Absorbtion o EtO Residuals
o Trace Path • Blood Leaks oBioburden
Elements o Fiber Leaks • Rattened Fibers
o PM Tolerances
Fig. 6. Tabulation of factors that influence the quality strategies used in the manufacture of dialyzers.
203
A. Raw materials
and source at least some raw materials and components from outside
vendors.
A dialyzer is made up of several basic components (see Fig. 8), including
a plastic case, headers or end caps, a hollow fiber bundle, potting compound,
and tip protectors. The heart of the dialyzer is the fiber bundle. Since the
performance of the dialyzer must meet certain AAMI guidelines, manufac-
turers must carefully qualify each lot of fiber used in the manufacturing
process.
When manufacturers use vendors to supply fiber, often the receiving and
inspection process includes physiologic testing and actual production of a pilot
lot of dialyzers using various lots of fiber furnished by the vendor.
Manufacturer's have a choice with respect to the fiber raw material, if it
is to be "wound" at the dialyzer manufacturing location, it is received as spools,
in which case the process of inspection includes identification, certification,
and permeability testing. On the other hand, if winding capability is missing
at the manufacturing site, it will be received as bundles of fibers already wound
and cut into the proper size for insertion into the dialyzer case. When fiber
bundles are received, inspection includes checking length and diameter of
the bundle as well as permeability or porosity.
As far as the rest of the dialyzer componentry is concerned, materials that
make up the case and headers (the rigid parts) are critical and must be of
appropriate materials to withstand the stresses of the manufacturing process
including the sterilization method (i.e., EtO, gamma, and steam). Also, the
chemicals used by dialysis clinics to process dialyzers for reuse may affect
some case and other component materials dramatically. Polycarbonate is the
most common case material but some manufacturers use acrylic. The choice
of materials is somewhat limited by the potential toxicity of the materials
and the potential for leachables to affect the dialysate and blood.
Other critical aspects of the case include the port connectors - both dialysate
and blood; both need to conform to international standards. There are two types
205
of dialysate connectors: Hanson and Walter; the difference between the two
being the Walter connector incorporates an internal locking system with a
check-valve (most often found in European dialysis clinics).
The quality strategy in manufacturing revolves around the choice of
Statistical Process Control (SPC) vs. inspection by attributes after manufac-
ture.
B. Winding process
Fber
opools
C. Membrane treatment
D. Membrane insertion
Once any treatment of the fiber bundles has been completed, they are inserted
into the dialyzer case, either manually or by some automated process. After
insertion, plastic potting caps are put in place. The purpose of the cap is to
207
create a barrier at each end of the dialyzer to contain the potting compound
during the potting process. These caps are either partially retained or removed
after potting.
E. Potting
F. Cutting
After encapsulation and curing, a cutting process is used to reopen the end
of the membrane - essentially by slicing off the ends of the potting cap. This
is generally done using knives which either perform shear or radial cuts. In
radial cutting, blades are spun at a high rate of speed, like a radial saw blade,
and applied to the dialyzer end. In shear cutting, a blade is pressed down on
208
the dialyzer potting cap ends, slicing it off. There is no question that the
spinning radial blade method for cutting is a more rapid process but the
blades tend to generate additional particulate matter which must be removed
in later processing steps.
In the cutting process, factors affecting the quality of the cut and, conse-
quently, the fiber, are:
• number of passes made by the cutting blades
• blade geometry
• blade rotation speed
• penetration speed
• blade angle
All of these parameters, are affected by the durometer (or flexibility) of the
cured urethane potting compound, type of membrane used, and the size of
the bundle. Common defects that influence quality of the final product include
flatness and membrane adherence to the urethane. If cutting results in severe
flattening of the fibers, it can influence the fit of the header. Also, dull or
high speed cuts cause fibers to pull away from the urethane.
The choice of potting compounds also affects the cutting process. There
is some evidence that aromatic urethanes cut better.
G. Cleaning
Once cut, the dialyzer must be cleaned before the caps (headers) are applied
to the ends. The purpose of cleaning is both to remove particulate matter
and other by-products of the manufacturing process and remove lumen fluid.
Similar to the cleaning process described previously, this step is accom-
plished using solvents, usually one of the following: Freon, Freon-alcohol
azeotrope, alcohol + glycerol, water (for synthetics), or inert gas (if no lumen
fluid is present).
The actual cleaning process uses either a bath or a flushlflow through
method. Cleaning the blood path is the primary objective. Both methods are
multi-stage. No matter what solvent system is used, critical to the process is
the reclaim procedure (for both economic and environmental reasons).
At present most manufacturers use Freon, which is scheduled for phase-
out by 1995. This already presents substantial problems for dialyzer
manufacturers because there is no "drop-in" replacement solvent for the freon
washing system.
H. Assembly
Once cleaned the headers and tip protectors are assembled to the case con-
taining the potted fiber bundles.
Headers are attached using three differing methods: torqued screw-on, ultra-
209
1. Leak testing
All manufacturers use a 100% leak testing process. The effectiveness of the
testing depends on the type of leak testing selected. There are two basic
types: (1) pressure decay testing, where about 20 PSI filtered air is applied
to the dialyzer and a change in pressure with time determined; and (2) gas
composition testing. The gas composition method is considered to be signif-
icantly more sensitive in leak detection.
Pressure decay testing options include (1) pressure blood side; (2) pressure
dialysate side; (3) negative pressure the blood side; (4) negative pressure the
dialysate side. Whatever leak testing strategy is chosen by the manufacturer,
it should be understood that leak testing does not totally prevent leakers.
J. Packaging
There are two strategies for the packaging of the finished dialyzer: either
(1) the package provides sterile barrier; or (2) the package is only a dust
cover - the tip protectors provide sterility assurance.
The type of packaging selected by the manufacturer is determined by both
the sterilization method and whether the package will represent a sterile barrier.
K. Sterilization
There are three common method for sterilization. Each has its own set of
advantages and disadvantages.
Gamma irradiation is fairly common and is very effective. However, irra-
diation has a negative effect on most components of the dialyzer - it tends
to make some fibers more brittle and can cause discoloration of the plastic case
and other components. This effect is related to the total sterilizing radiation
210
L. Final testing
Once the dialyzer has been sterilized, the manufacturer, using a statistical
sampling plan, performs some testing on the finished product. This testing
usually includes pyrogen testing using the LAL method, leak testing, blood
path monitoring (wherein the dialyzer is flushed with a known volume of
fluid and the effluent examined for particulate matter), performance testing,
including UFR and clearance measurements.
Conclusion
issues. These factors can be generally classified as falling into one or more
of the following categories: biology, pyrogenicity, toxicity, chemistry, and
general manufacturing. It is only by applying quality and manufacturing
technology that the manufacturer can provide a truly quality product to the
patient.
References
1. Juran JM, Gyrna FM Jr, Bingham RS Jr: Quality Control Handbook, fourth ed. (Manchester,
MO: McGraw-Hill), 1988.
2. Pollack GM et al.: Circulating concentrations of di(2-ethylhexyl)pthalate and its
deestrified pthalic acid products following plasticizer exposure in patients receiving
haemodialysis. Toxicol Appl Pharmacol 79: 257-267, 1985.
3. Ganning AE et al.: Regular exposure from plasticizer can have detrimental effects on health.
Liikartidningen 81: 4389-4392, 1984.
4. Henderson IS et al.: Factors affecting plasticizer content of unused CAPD fluid. Peritoneal
Dialysis Bulletin Suppl 7(2): 539, 1987.
5. Nassberger Let al.: Exposure of patients to phthalates from polyvinyl chloride tubes and
bags during dialysis. Nephron 45: 286-290, 1987.
6. European Chemical Council (CEFIC) draft position paper on Medical Applications of
Plasticised PVC, 1993.
7. Van Stone JC: Hemodialysis Apparatus. In Daugirdas JT, Ing TS (eds) Handbook of Dialysis
(Brown: Little, Brown and Co), 1988.
CHAPTER 13
GAIL S. WICK
While Deming and Juran [3, 7] focus heavily on the use of statistical tech-
niques in total quality management, they also emphasize management's
leadership role as a key factor in success. State of the art equipment, abundant
supplies and pleasant environment facilitate provision of quality products
and services, but ultimately the humans making and using products and
providing services determine the success or failure of all quality manage-
ment efforts. As Hoesing and Roey [2] point out, employees' behavior or
performance determine patient clinical and satisfaction outcomes. These
outcomes drive organizational profit (market share and margin). Clearly then,
quality management (quality assurance, control and improvement) is an
employee driven concept.
Although employees ultimately determine success or failure of quality
efforts, the fact that "workers are responsible for only 15% of quality problems
and the system for the other 85%" illustrates the challenging task that man-
agement faces. It must create a system in which quality management activities
can flourish and succeed rather than flounder and die [7]. The purpose of
this chapter is to focus on the responsibility and role of management in leading
personnel to and through quality management in the health care setting.
Corporate culture must embrace and nurture the concept of quality manage-
ment and do this through judicious implementation, evaluation and continuous
improvement. While quality is a total company commitment, it must come
from the top down.
Quality management
Quality is built into the process by developing and promoting standards that
reflect the organization's definition of quality services. Standards are supported
by policies while procedures promote positive outcome. In other words,
standards state what the organization believes quality is. Standards can also
be used to monitor, evaluate and for improvement [2]. Policy states clear
expectations of what can and is to be done; and procedures define the exact
steps or process which will achieve the desired outcome in an efficient manner.
The role of a company's employees is to assist by (1) defining quality through
the development of unit specific standards, (2) identifying indicators to measure
quality, (3) determining procedures to achieve it and (4) developing a futur-
istic plan of continual improvement based on statistical analysis and problem
solving.
Management is responsible for the education of the entire organization on
how to "do it right" and then leading it into a commitment. Mission, values
and guiding principles must be established early, because they form the basis
for all other activities. An organization's mission statement communicates
the broad vision and purpose of the organization, and gives employees
direction and focus to pursue their tasks to achieve this mission. The mission,
along with guiding principles of operation, clarifies the important roles that
all employees of the organization have in accomplishing this mission. From
this mission statement, goals which clearly communicate how to get, maintain
and measure quality can be collaboratively established [7]. When an organi-
zation clearly states and consistently acts on its commitment to quality, then
pride, security, loyalty and commitment will increase.
As with any concept, there are barriers to implementation and success. Deming
[2] outlines certain barriers that result in quality failure. These barriers include:
and emphasis on numbers, not quality; turning out products (services) quickly
rather than properly; turning a deaf ear to suggestions; too much time spent
on rework; using poor or improper tools and difficulties with incoming
materials. These barriers can be found within the dialysis setting and must
be identified in order to guard against them. Are staff mix and staff/patient
ratios set regardless of patient acuity? Do caregivers rush to get patients on
and off dialysis to meet "quotas" in lieu of proper assessment, following proper
procedure and clinical trouble shooting? Are preventive care and patient
satisfaction conscious goals? Have staff meetings and inservices been cut from
the schedule for perceived lack of time thus hindering important communi-
cation and educational opportunities? Are unnecessary complications and
deficiencies occurring due to the implied need to focus on mechanics and
quotas rather than the total health care picture? Is the staff apathetic and
dispirited? Are tools for quality management and care such as statistical
tools, state of the art equipment, adequate supplies and qualified personnel
available to accomplish quality goals? Management must identify barriers
and eliminate them through effective leadership and provision of adequate
resources.
215
Juran [3] states that the goal of leadership activity is to (1) improve the
system through input from the team and (2) to facilitate consistency in
performance so that over time, employees perform with fewer variations from
the established process. In other words, the goal is for everyone to do what
he is supposed to and for the team to continually evaluate and improve the
process in order to get the desired results. This can only occur if manage-
ment selects the right team, prepares team members for their roles within
the organization and provides them with a work environment that is con-
ducive to the provision of quality services.
Key human resource issues in quality management are therefore (1) the
selection process, (2) training and education activities and (3) team develop-
ment. Provision of quality services must be built into the system. Human
resources must be maximized and self leadership capabilities must be
promoted.
Once the manager has made a commitment to an employee through the hiring
process, the challenge is to provide the employee with the necessary knowl-
edge, skills and resources to meet, and hopefully exceed, the quality standards
established by the institution. Education and training lay the groundwork for
positive patient outcome and practitioner performance, both of which directly
affect bottom line as well as quality. An additional benefit is that a clear
understanding of expectations and indicators of acceptable performance
provides focus and direction for personnel. Managers and subordinates can
collaboratively measure and monitor performance as well as outcome.
Initial training and education for a direct care giver in the dialysis setting
should include, but not be limited to the following:
• overview of chronic and acute renal failure with indications for therapy,
including the principles and practice of dialysis
• nutritional and psychosocial considerations
• manifestations of renal failure
• medical complications of dialytic therapy
• patient teaching needs and techniques
• specific procedures and rationale for the steps
• the interrelatedness of technique, assessment and preventative measures
to quality care
Regardless of exact content, nationally recognized standards of practice such
as those of the American Nephrology Nurses Association and the American
Association of Medical Instrumentation, should serve as the basis for educa-
tional programs. These standards can also be useful guidelines to coach,
counsel and appraise personnel involved in direct patient care as well as the
technical activities of dialysis.
Quality management requires a cooperative atmosphere where departmental
goals enhance the overall efforts of the organization. For this reason, team
members must understand the role and relationship of all departments to the
quality management effort. Orientation of support and non-direct care per-
sonnel should include a basic introduction to ESRD, its treatment, financial
aspects of the program as well as challenges to quality management. Direct
care givers, in tum, must understand the specific role and importance of other
departments in the quality effort.
Because quality management is so important to overall patient outcome
and program success, employees must be taught additional techniques and
skills. These should include problem solving skills, creative brain storming,
communication skills, time management and data collection and analysis.
All are important factors in a successful employee driven quality manage-
ment program. The principles and practice of care identify (1) what aspects
of care are high volume, high risk or problematic and (2) reasonable indica-
tors for monitoring quality. Data collection and statistical analysis come next,
followed by activities which require the ability to identify problems, imple-
218
Work environments that foster quality services and products demand the highest
standard of care and work. They are consistent in promoting quality man-
agement and strive to keep employees satisfied. They provide for clear and
frequent two way communication and have a committed team that is
dedicated to the concept of quality.
In corporation of quality management programs has an added bonus of
helping create a success oriented work environment. Three steps to incor-
poration of quality management programs include (1) careful and honest
assessment of the current situation, (2) analysis of the answers and gearing
up for action, and (3) development of programs which give employees
something to work toward [8]. Quality management carried out by self-
managed teams can provide most, if not all of the conditions employees feel
are most important in a work environment.
Because people like to work for and with the best, they expect certain stan-
dards and rules which promote the best service or product. Helping to establish
the standards fosters ownership. Much of the time, the goals and standards
set by employees are as high if not higher then those of management. In
addition, people forced to carry out standards and goals that they had no
hand in establishing, do so half heartedly and without uniformity [7]. This can
be very damaging to quality efforts because quality requires total commit-
ment throughout the organization as well as uniformity in quality performance.
Self imposed goals are an important ingredient for successful self leader-
ship. They help employees establish self direction and priorities for immediate
work tasks and long term career achievements [5]. If employees have been
carefully selected and adequately educated and trained, they are likely to choose
goals which promote quality as well as support the principles of the organi-
zation. When employees establish goals they are more likely to strive to
make the process and outcome successful. Active involvement in the direc-
tion of the organization also self guards against boredom and loss of focus
on the ultimate goal - quality. Good questions to ask in assessing status and
setting new goals include:
• What did we do right this week?
• What do we want to accomplish within the next 90 days?
• How can we do it?
• What can we do to make things better?
Once agreed upon, goals must be clearly communicated and followup
monitored.
In addition to establishing standards and goals, employees can and should
be involved in regularly answering the following:
220
Teamwork
is elected for each team and has the responsibility for facilitating meetings,
communication and problem solving as well for his/her normal work respon-
sibilities. A pay differential accompanies the added leadership responsibilities.
The purpose of the self-managed team is give control of performance and
outcome of efforts to employees who have been prepared philosophically
and educationally for such responsibilities. The responsibilities include such
areas as budget preparation, scheduling, data collection and analysis and
problem solving. Communication within and between teams is facilitated
through specific channels. Short, weekly team meetings provide the formal
arena for communication and group problem solving, while daily communi-
cation among team members is possible because of close work proximity.
Coordinators become involved in team discussions only if necessary for
specific problems or situations. Communication between teams and the three
team levels is maintained through posted written agendas and minutes of all
meetings along with regularly scheduled leadership level meetings. Information
sharing is free and open with the exception of those matters which are truly
personal. Motivation and discipline come from within the teams, thus the
term self-leadership. They do not come from management. Problems and
quality issues are identified and solved by the teams rather than by manage-
ment. Solutions are therefore employee owned, which increases motivation
to make the plan of correction and subsequent improvement succeed. In
actuality, when practiced successfully, the self-managed team concept addresses
almost all the qualities that people want most from their job [1] which are
to:
• work for efficient managers
• think for themselves
• see the end result of their work efforts
• be assigned interesting work
• be informed
• be listened to
• be respected
• be recognized for the efforts
• be challenged
• have opportunities for increased skill development
Quality is the end product of the coordinated efforts of all departments
within an organization. Self-managed transdepartmental teams with well estab-
lished channels for communication is a mechanism for serious consideration.
The role of management is to influence the team members to provide quality
and to create a positive work environment themselves rather than for man-
agement to issue directives or to do it themselves [5]. In other words, control
lies with employees rather than management. Since quality control is an
employee driven activity, the organizing, directing and monitoring functions
of self-leadership and self-managed teams should promote successful quality
programs if adequately planned for and implemented. The enthusiasm and
motivation necessary to assure quality within a dialysis unit can come from
222
Job satisfaction
Summary
References
1. Dell T: An Honest Day's Work: Motivating Employees to Give Their Best (Los Altos,
California: Crisp Publications), 1989.
2. Hoesing H, Kirk R: Common sense quality management. JONA 20, Number 10, October
1990.
3. Juran JM: Juran on Leadership for Quality, An Executive Handbook (New York: The Free
Press), 1989.
4. Leeds D: Smart Questions - A New Strategy for Successful Managers (New York: McGraw
Hill), 1987.
5. Manz C, Sims H: Super-Leadership (New York: Prentice Hall Press), 1989.
6. Peters T, Austin N: A Passion for Excellence (New York: Random House), 1985.
7. Walton M: The Deming Management Method (New York: Putnam Publishing Company),
1986.
8. Wick G: Stimulating peak performance in the dialysis setting. NRAA Journal 11, Number
I, Spring/Summer, 1990.
CHAPTER 14
The clinical records in dialysis vary from the official medical record of the
patient to a multitude of unofficial, yet vital, forms and tracking systems.
All of these allow the clinician to monitor the many variables associated
with management of acute and chronic dialysis patients. Quality manage-
ment improvement in the clinical dialysis setting requires the tracking and
trending of data which can be used to identify and analyze areas for poten-
tial change which will bring about a new and hopefully a better level of care
for the renal patient.
Documentation systems
In order to implement and maintain data systems which will provide mean-
ingful information for continuous quality improvement, the clinical records
must be evaluated for their relevancy to the services being delivered and for
the monitoring of those services through the quality improvement process. This
evaluation process is ongoing and action-oriented. Data collection systems
should support the evaluation process and be designed to facilitate easy
identification of the trends, actions, and outcomes. Finally, the data collec-
tion systems should support activities which will lead to future improvements
in service.
Continuous improvement of health care services is being promoted by
regulatory and advisory agencies in health care today. One such entity is the
Joint Commission of Accreditation for Health Care Organizations (JCAHO),
which is implementing a plan of change over the next five years [1]. This
plan will require hospitals, and therefore hospital-based dialysis units, to
demonstrate that components of the Continuous Quality Improvement CCQI)
[2] model are integrated into the quality monitoring process in use in the
facility. This process was developed by W. Edwards Deming and Walter
Shewhart [3-5]. The essential elements of their theories are: (1) Establish a
constancy of purpose, (2) Utilize statistical analysis for evaluation of the quality
of services, (3) Examine the variations found in the statistical analysis, (4) Use
the knowledge gained from the analysis to correct the systems and bring
about improvement in the quality of the service.
In the future, it is reasonable to expect mandatory guidelines will be
developed based on national standards, that will require both hospital-based
and free-standing dialysis units to meet criteria which will measure the quality
of patient outcomes. It is also reasonable to expect that the process of
measuring these standards will be based on the principles of continuous quality
improvement.
Measurement
Searching for better ways to provide dialysis treatment is not new to dialysis
professionals who have been involved with the research and application of
the technology. What is new, however, is the concept of measuring the rela-
tionship of clinical response of the patient to technological advances.
Conceptually, both in industry and in the healthcare setting, measurement of
quality has focused on the methods and the products used in the treatment
of patients as well as the quality of the environment where patient care was
given. These methods for measuring quality focused on the process and
structure of the organization. Now the emphasis of measuring quality is focused
on the patients' response to the health care intervention. The emphasis on
outcomes has added a new dimension to health care quality measurement [6-7].
Previously, quality assurance plans for health care concentrated on counting
the presence or absence of specific criteria. Outcome based quality monitors
encourage the use of tracking information for the purpose of identifying trends
on which decisions regarding the quality of the care being delivered can be
made. The additional concept of continuous improvement then becomes the
driving force for elevating the level of care beyond established industry
standards.
Continuous Quality Improvement acts as a catalyst. The catalyst is the
challenge Berwick refers to as a continuous reach for opportunities to improve
all aspects of health care [8]. This dimension stimulates action to intervene
and improve patient care and decrease risk through system changes and
improvement in practice.
A search for opportunities to improve care should not require intensive
detective work to uncover meaningful trends in the care being given. Trends
should be easily identified from the data collection process. In addition, the
people who collect the data are usually the patient care staff who are invested
in improving the outcome of what they do for patients. Since time allotted
for patient care is the priority in the clinical setting, data collection systems
should not detract from the caregiver's primary responsibility. The data systems
should support the care given both from a legal perspective, as well as from
a treatment perspective. Therefore, the data that is recorded should be concise,
accurate, timely, legible, factual and clinically pertinent [9]. The data
227
collection method developed for the clinical dialysis setting should promote
these criteria. Additionally, it should be streamlined to achieve efficiency in
the daily work of all patient care providers.
Analysis of documentation
The Clinical setting also influences the type of data necessary for a quality
recording system. Location and acuity level are influencing factors in the
planning for documentation systems. The patient who is acutely ill and
receiving treatment in an intensive care unit will have many core data elements
which are similar to the stabilized patient and the home dialysis patient. On
the other hand, each of these groups of patients have a set of data which is
unique to their setting and care requirements. In addition to the location and
acuity of the patient population, the need to interface with other health care
providers also influences the design of the documentation system. In the
hospital setting, the dialysis record must be part of the centralized medical
record system. In the free-standing dialysis clinic, mechanisms must be in place
for the communication of clinical data between the center and other health
systems which provide support to the nephrology team for the care of the
patient. The home dialysis record must support an even broader network of
health care providers such as the patient's local physician, community health
agencies and home equipment and supply companies.
Finally, the data should be reviewed for its accuracy and validity. The facts
relating to the patient's care should be recorded in a timely manner and in
an organized format so that there is a clear chronological picture of what
has happened to the patient during the course of the treatment by each provider.
The people who use the documentation system are the best individuals to
ask for input when designing a new form or an entire system. If the system
is simple, user-friendly, flexible and designed with future improvements in
mind such as computerization, the system will be used. The forms them-
selves should be organized and printed so as to guide the user in a sequential
and efficient manner from the assessment of patient care through planning,
229
implementation and evaluation. This process then sets the data collection in
a logical sequence for monitoring continuous quality improvement oppor-
tunities.
MEDICAL RECORD
Recommended Optional
Intake Admissions Data Chart Checklist
Dialysis Consent Pre Dialysis Checklist
Medication Record Problem List
Long Term Care Plan Health Maintenance Record
Modality Selection Transfusion Record
Pre and Post Dialytic Assessment Research Protocols
Modality Selection
Intradialytic Flow Sheet
Dialysis Adequacy
Patient Care Plan by Dialysis Team Individualized
230
Recommended (Continued)
Laboratory and Diagnostics
Medical Directives (Living Wills)
Narrative Records
Patient Bill of Rights
Patient Grievance Procedure
Viral Surveillance
Infection Rates
History and Physical
Discharge Summary
Facility records
FACILITY RECORDS
Recommended Optional
Water Treatment Records Mission Statement
Reuse Records Research Protocols
Machine Maintenance Records Patient Classification System
Credentialing Records Career Ladder Records
Performance Appraisals Continuing Education
Morbidity Mortality Data Annual Report
Patient Standards of Care Operational Directives
Policy and Procedures Organizational Chart
Infection Control
Staffing Patterns
Required Individual Orientation and Inservice
Records
Facility Mandatory Inservice Records: Fire/Safety,
CPR, Infection Control
Quality Program
Occurrence Reports
Staff Meeting Records
Hazardous Waste Program
Medical ByLaws
Governing Board Minutes
On Call Records
Home Supply Records
Billing Records
231
Both of these lists of records found in the dialysis setting have many
common data elements. Because the number of reporting requirements have
grown over time, much of the information is repeated from form to form.
Reducing the amount of recording by staff through an organized and effi-
cient information system is the first step in ensuring quality of care in the
dialysis unit.
Information systems which streamline the process of recording, interfacing
and integrating the data will lead to better and more efficient use of the records
by staff in their care of the patient. Without a streamlined information system,
quality programs become unmanageable, discouraging and eventually ignored.
Once the information systems for recording and maintaining patient data
has been established, these records should be key tools or sources for quality
monitoring data [16].
The most common format for monitoring quality in the clinical setting today
is based on the JeARO 10 Step model [17]. Adaptions of this model to
incorporate more continuous improvement principles has already occurred and
over time the model will evolve as the concept becomes refined in the health
care setting. The relevance of this model and of the Deming industrial model
or the adaption of Deming for health care by Batalden and Vorlicky is that
there is a definitive evaluation plan [4]. This plan will guide an organization
to establish standards of quality, monitor the organization's ability to uphold
the standards and ultimately to improve upon the standards.
The goals of quality monitoring are:
1. To assure identification and correction of system wide and unit based
problems.
2. To promote care that is cost effective and clinically effective.
3. To promote optimal patient care through the ongoing assessment of all
aspects of care, together with the correction of identified problems.
4. To identify and reduce risk factors while at the same time managing known
risks.
5. To meet the requirements of external agencies.
In the future new models will be developed to approach quality improve-
ment, but the basic structure of the JCARO model is simple and easy to
implement in a relatively short period of time.
The JCARO 10 Step model can be compared to other models already in
place in the health care field. The nursing process is one example where a
systematic approach of problem analysis can be used to determine the care a
patient will need and the patient's response to the care. In research, the problem
solving process is orderly and disciplined to acquire reliable information and
to analyze that information in an orderly manner.
All of these models or processes focus on improving quality of care and
232
INDICATORS
STRUCTURE INDICATOR
The medical record organization reflects the chronological factual occurrences
of the patient from admission to discharge.
PROCESS INDICATOR
Documentation in the medical record is meaningful, significant issues are iden-
tified and management of the patient is consistent with the findings.
OUTCOME INDICATOR
The quality of care provided to the patient is clearly documented.
DIALYSIS CLINICAL RECORD INDICATOR
The clinical record documentation will reflect individualized and appropriate
medical treatment.
b. The final diagnosis coincides with the findings and treatments docu-
mented.
c. Plans for follow-up care and discharge instruction are pertinent to
the care of the patient.
CHRONIC CRITERIA
1. The medical record reflects that appropriate ESRD treatment options
have been discussed with the patient and family and that they have
participated in the choice of treatment.
2. The medical record reflects assessment of the patient prior to initiation
of dialysis treatment in the following areas:
a. Access
b. Volume - target weight
c. Potential infection
d. Laboratory values
3. The medical record reflects ongoing assessment of the patient receiving
chronic dialysis therapy in the following areas:
a. Understanding of treatment modality
b. Understanding of present health problems
c. Coping mechanisms
d. Rehabilitation
4. The medical record reflects participation by the patient and his ability to
make decisions in the planning of his own ESRD care.
HOME DIALYSIS CRITERIA
1. The medical record reflects ongoing clinical assessment of the appropriate-
ness of the ESRD modality of choice.
2. The medical record reflects ongoing assessment of the patient's and/or
family's satisfaction with the treatment modality.
3. The medical record reflects appropriate referrals to community resources.
quality monitoring can be achieved with efficiency because the data will be
readily retrievable. Quality monitoring can take the route of examining the
structure of the organization to determine if the organization itself promotes
quality outcomes for patients. Another route of quality monitoring is to look
at the process of the organization to determine if the systems are in fact
supportive of quality care. The emphasis, however, should be focused toward
the outcome of the patient. Patient outcome results should demonstrate that the
interventions produced desired results.
Conclusion
The importance of the clinical records in the dialysis setting cannot be empha-
sized enough. As demonstrated through the various examples in this chapter
of how to establish a quality monitoring program the core element is a valid
clinical record. Successful implementation of the continuous improvement
model requires an understanding of the theory of quality management and how
these concepts integrate with medical practice and the influence of the
environment on that practice. Continuous quality improvement is a concurrent,
dynamic, action oriented process based on objective analysis of valid data. This
model simultaneously evaluates quality, assesses, prevents and reduces risk,
implements cost-effective resource management measures, and identifies and
combats potential problems. A well organized, streamlined, interactive clinical
record system is the key to a successful continuous quality improvement
program.
References
1. The Joint Commission Guide to Quality Assurance, 1988: The Joint Commission on
Accreditation of Healthcare Organizations, Chicago.
2. Conway-Welch C: Entering a new era of quality care. ANNA Journal 16(7): 469-471,
Dec 1989.
3. Deming WE: Out of crisis. Cambridge, MA: Massachusetts Institute of Technology, Center
for Advanced Engineering Studies, 1989.
4. Walton M: The Deming management method (New York: Putnum Publishing Group), 1986.
5. Vlchek DL, Day LM: A quality improvement management model for renal care. Nephrology
News & Issues Special Supplement.
6. Milton 0: Challenges of quality assurance program evaluation in a practice setting. J Nurs
Qual Assur 2(4): 25-34, 1988.
7. McCormick KA: Future data needs for quality of care monitoring, DRG consideration, reim-
bursement and outcome measurement, Image: Journal of Nursing Scholarship 23(1): 29-32,
Spring 1991.
8. Berwick OM: Continuous improvement as an ideal in health care. American Journal of
Medicine 320(1): 53-56, Jan 5, 1989.
9. Professional Liability Program, Farmers Insurance Group of Companies, Keeping the Record
Straight: Guidelines for Charting, QRC Advisor 6(3): 7-9, Jan 1990.
10. Pollak VE, Peterson OW, Flynn J: The computer in quality control of hemodialysis patient
care. QRB, 202-210, June 1986.
238
11. West E: Designing information systems to increase quality care. Computers in Healthcare,
Sept 1990.
12. Kaiser LR: Anticipating your high tech tomorrow. Healthcare Forum, 12-20, NovlDec 1986.
13. McDonald IC, Newton GA: The patient flow management model: A process for quality
assurance. The Health Rec Manage 10(3): 32-43, 1990.
14. Mogli GD: Role of medical records in quality assurance program. Am Ro 30(4): 11-15,
Nov 1989.
15. Chasteen IE: For the record. Dental Assisting, 23-26, Sept/Oct. 1987.
16. Donabedian A: The quality of care: How can it be assessed? JAMA 260(12): 1743-1748
Sept 23/30, 1988.
17. Joint Commission on Accreditation of Healthcare Organizations, 1988: Accreditation Manual
for Hospitals, Chicago.
18. Hexum 1M: Monitoring standards instead of problems. J Nurs Qual Assur 1(3): 8-13,
1987.
19. Laffel G, Blumenthal D: The case for using industrial quality management science in
health care organizations. JAMA 262Z(20): 2869-2873, Nov 24, 1989.
20. Bednar B, Neff M: Preparing for inspection: A tool to maximize quality and minimize
risk. ANNA Journal 17: 159-164, April 1990.
21. Rajki KL, Feltman BA, Smeltzer CH: Assessing the quality of nursing care in a dialysis
Unit. ANNA Journal 12(1), Feb 1985.
22. Schyve PM, Prevost IA: From quality assurance to quality improvement. Psychiatric Clinics
of North America 13(1): 61-71, March 1990.
23. Louden TL: Customer perception counts in quality assurance, Hospitals 84, Jan 20, 1989.
24. ANNA, Quality Assurance for Nephrology Nursing, First ed, 1989.
25. Characteristics of Clinical Indicators. QRB 330-339, Nov 1989.
26. Parker 1: Reduction in ESRD reimbursement rate: Identifying research priorities and quality
indicators. ANNA Journal 17(2), April 1990.
27. Getting Ready for Continuous Quality Improvement: Fundamentals for Department
Managers. American Hospital Association, 20 June 1991.
28. Van Valkenburgh D: Continuous quality improvement (CQI) of current quality - assur-
ance programs. Dialysis & Transplantation 20(9), Sept 1991.
CHAPTER 15
Background
In 1986, when the focus on quality assurance was emerging in dialysis, the
Renal Division of Baxter Healthcare Corporation initiated a quality improve-
ment-support program for peritoneal dialysis, the Best Demonstrated Practices
Outcome measurement
100 .-;:~
90 ~_ _
.- ~~
.!-
c •
,! ~ .~~ ------ .......
.-. .,.
"'-,s
'0 • 60 •
#.= 50 - - - - ._ _ _ _
:.,s
40 ---.
.~
'!i "'- 30
EE
o:::I u0 20
10
O~--------------------------------~
o 6 12 18 24 30 36
InlelV,1 numb.,
Still on CAPD
40.3%
Morbidity measures
The PRA forms the basis for problem evaluation, prevention and practice
improvement. Quantification of treatment outcomes facilitates comparison
within and to other dialysis units. The PRA thus provides a basis to assess
program strengths and weaknesses, to target areas for improvement, and assess
the impact of practice changes through re-analysis.
Print Materials
BDP Meetings
Meetings based upon the BDP process have been held in several countries.
Following presentations on clinical topics, meetings focus on workshop
sessions. Emphasis is placed on exploring practical management techniques
244
and potential solutions to key issues. Participants reexamine current issues and
practices, and provide the basis for individual quality assurance linkages.
Suggestions for future research are outlined, as well as practical ideas to
evaluate.
Specific practice improvements are generated for individual problems by
professionals who have managed the complication area well, based on demon-
strated low drop out from PD in specific practice areas in the PRA. Centers
with low dropout due to the complication use their own experience as a catalyst
for practice brainstorming with other professionals. Specific practices are
shared, in search of "best" practices. The professionals (1) clearly define PD
complications for their group (2) identify potential solutions/new ideas for
PD patient management and (3) evaluate and support implementation of
practice changes.
At a Canadian BDP meeting in December 1992, BDP program partici-
pants met to understand Canadian issues in patient and center management
practices. Meeting objectives included:
1. Share Canadian BDP Program results.
2. Share current PD issues and Practices.
3. Understand Canadian issues in PD patient and center management prac-
tices.
4. Share information from Canadian Centers of Excellence in each Practice
area.
S. Evaluate the Canadian BDP program effectiveness as a quality assur-
ance/assessment tool [11].
The attendees consisted of medical and nursing professionals from 12 Canadian
centers. The centers represent 7 provinces, 16% of the total Canadian dialysis
centers, 39% of all Canadian PD patients. The average PD usage is 49%.
Outcome/results are discussed later.
Professionals compared their individual center results to Canadian averages
and identified individual strengths and weaknesses in 4 clinical areas:
Peritonitis, Catheter issues, Adequacy, and Psychosocial/Non compliance.
Workshop sessions were held to discuss clinical practices and develop
recommendations for 'best' practices in these areas. Each group then presented
their recommendations to the entire audience. Through this process, program
participants established Canadian best demonstrated practices. Recommenda-
tions assist the QA teams in each center in improving practices and evalu-
ating quality on an on-going basis. A summary publication was developed
and distributed to all Canadian dialysis centers. Thus, Canadian practices
of excellent centers was disseminated to all Canadian centers as a reference
guide and forum for sharing "Best Demonstrated Practices".
At one U.S. meeting, peritonitis was explored as a reason for therapy
transfer. Of note, 29% of patients transferring to HD due to peritonitis did
so because of an unwillingness to continue CAPD, in the absence of medical
contra-indications. This finding highlighted the need to understand the
psycho-social aspect of therapy transfer. A similar finding in the Dec '92
245
A success model
A large university based CAPD program identified their dropout issues through
data analysis in the PRA. Catheter problems were identified as a primary cause
of early dropout. The data provoked investigation of catheter practices; it
was determined that variation existed in catheter implantation procedures.
Following a meeting with surgeons, nephrologists and CAPD nurses to review
the latest in catheter insertion practices, a common protocol was developed
for catheter insertion which improved early catheter success.
A small, rural, free-standing CAPD program improved their training
246
The collection of data from individual dialysis units has permitted a macro-
analysis of a large number of dialysis facilities in several countries. The U.S.
database-analysis and a Canadian BDP meeting data are contained below:
U.S.: As of June 1991, 356 U.S. centers had submitted data for a patient
retention analysis of 21,497 patients over a 10 year period. These data
represent unique views into PD use due to its longevity and large number of
patients. Figure 5 compares the aggregate patient and technique survival for
the 21,997 patient. Several points are noteworthy. First, according to this
data, the CAPD patient has a 50% chance of technique success after 5.5 patient
years. Secondly, in this data base, 362 patients had been successfully managed
on the therapy for over 6 years.
Canada: As a part of a multi-center BDP meeting in December 1992, 13
centers submitted data for a PRA for 1,990 patients spanning a 10 year period.
Figure 1 shows the aggregate patient and technique survival. At 3 years, 80%
of patients were maintained on PD excluding death and transplant. The
Canadian patient has a 50% chance of technique success after 6.2 patient years,
and 40% were still on the therapy at ten years. Transfer to HDS occurred
most frequently in the first six months of PD treatment (30%). Transfer
247
100
90 - - Patient Survival
70
iij
> 60
..
~
::J
C
50
..
8....
Q.
40
30
20
10
0
.. ., ..,
0 <D CII
-
CI)
CII
<D
Mon1hs on PO
o
<D
<D
'"
Fig. 4. Patient and technique success, lifetable, U.S. Database. The PD patient has a 50%
probability of remaining on PD at 5.5 years.
10%
21%
32%
23% o Catheter
• PsychologlcaVNon-compllance
Other Medical
Peritonitis
Inadequate Dialysis
Fig. 5. Reasons for modality transfer, U.S. database. Peritonitis is the prmary factor for transfer
to hemodialysis over the entire data base.
248
steadily declined over the scope of the analysis, with death and transplant
increasing:
Peritonitis (30%) accounts for the major reason for therapy change over
the ten year period (Fig. 6). Other medical problems account for the second
largest reason for leaving (23%). In the first six months of use, other medical
factors accounts for the largest reasons for changing therapy (30%). This
18%
23%
o Catheter
• PsychologicaVNon-compliance
• Other Medical
D Peritonitis
Inadequate Dialysis
Fig. 6. Reasons for drop-out to hemodialysis - Canadian data. Peritonitis accounts for 30% of
the reasons for modality change over the entire period.
249
data supplemented individual center data and served as a basis for therapy
practice analysis in the December BDP meeting.
Conclusion
Baxter developed the BDP program with the goal of increasing PD technique
success by improving the clinical and training practices associated with the
modality. Results support the merits of this type of clinical management process
in PD. The BDP database did not show significant outcome differences in
diabetes, cardiovascular disease, and age, as having significant influence on
PD technique success. Rather, more subjective elements, such as professional
commitment to PD, quality practices, ongoing commitment to improvement,
and a full line of dialysis therapeutic options appear to positively affect program
success.
The use of BDP has been varied. In those centers with strong focus on
quality, BDP has supported outcome measurement and change. There are
successful reports of its use in individual centers as a means of providing
annual analytical feedback. The large database analysis has also successfully
provided practitioners a basis for discussion of outcomes and practices in BDP
meetings.
BDP's success lies in its two part approach to problem discovery and
change. It has increased the confidence of physicians and nurses in their clinical
practices and programs. The process has provided access to successful prac-
tices and has enhanced communication among professionals. The challenge
is to understand the dynamics behind success and failure and apply the proper
means to correct problems. Professionals using the program report that BDP
supplies the process; however, it is the individual's commitment that is the
underlying success factor.
Embedded in the BDP program are several quality improvement concepts
and tools, including: Personal involvement, goal setting, education, commu-
nication, striving for excellence, team-work, specific quality improvement
projects, and an on-going improvement process. Results are never instanta-
neous; ongoing attention, review and commitment to the process are required.
BDP provides a framework for quality improvement, assisting professionals
with their quality assurance efforts.
References
1. Holden A, Graumer G: Best demonstrated practices program promoting CAPO patient reten-
tion in the United States. Advances in Continuous Ambulatory Peritoneal Dialysis, aI86-191,
1987.
2. Best Demonstrated Practices, Peritonitis Management and Antibiotic Therapy Practices
Module. Baxter Healthcare Corporation, 1987.
3. Best Demonstrated Practices, Catheter and Exit Site Practices. Baxter Healthcare
Corporation, 1987.
4. Best Demonstrated Practices, Patient Education and Training Module. Baxter Healthcare
Corporation. 1987.
5. Best Demonstrated Practices, Modality Selection Practices Module. Baxter Healthcare
Corporation. 1987.
6. Keane WF et al.: Continuous ambulatory peritoneal dialysis (CAPO) peritonitis treatment
recommendation: 1989 update. Peritoneal Dialysis International 9: 247-256, 1989.
7. Keane W et al.: Peritoneal dialysis-related peritonitis treatment recommendations. Peritoneal
Dialysis International 13: 1, 1993.
8. Oreopoulos DG et al.: Peritoneal catheters and exit site practices: Current recommendations.
Peritoneal Dialysis Bulletin 7: 130-137, 1987.
251
9. Gokal R et al.: Peritoneal catheters and exit-site practices: towards- optimum peritoneal
access. Peritoneal Dialysis International 13: 1, 1993.
10. Hamburger RJ et al.: A dialysis modality decision guide based on the experience of six
dialysis centers. Dialysis & Transplantation 19: 65-69, 84, 1990.
11. Suart Chatterley: J. Best Demonstrated Practices - The Canadian Results, Dec. 1992.
12. Finkelstein F et al.: Initiatives in Peritoneal Dialysis: Where do we go from here? Peritoneal
Dialysis International 11: 274-278, 1991.
Subject index
253
254
1. J.S. Cheigh, KH. Stenzel and A.L. Rubin (eds.): Manual of Clinical Nephrology o/the
Rogosin Kidney Center. 1981 ISBN 90-247-2397-3
2. KD. Nolph (ed.): Peritoneal Dialysis. 1981 ed.: out of print
3rd revised and enlarged ed. 1988 (not in this series) ISBN 0-89838-406-0
3. AB. Gruskin and M.E. Norman (eds.): Pediatric Nephrology. 1981
ISBN 90-247-2514-3
4. O. Schuck: Examination o/the Kidney Function. 1981 ISBN 0-89838-565-2
5. J. Strauss (ed.): Hypertension, Fluid-electrolytes and Tubulopathies in Pediatric
Nephrology. 1982 ISBN 90-247-2633-6
6. J. Strauss (cd.): Neonatal Kidney and Fluid-electrolytes. 1983 ISBN 0-89838-575-X
7. J. Strauss (ed.): Acute Renal Disorders and Renal Emergencies. 1984
ISBN 0-89838-663-2
8. LJ.A. Didio and P.M. Motta (eds.): Basic, Clinical, and Surgical Nephrology. 1985
ISBN 0-89838-698-5
9. E.A. Friedman and C.M. Peterson (eds.): Diabetic Nephropathy. Strategy for Therapy.
1985 ISBN 0-89838-735-3
10. R. Dzurik, B. Lichardus and W. Guder: Kidney Metabolism and Function. 1985
ISBN 0-89838-749-3
11. J. Strauss (ed.): Homeostasis, Nephrotoxicy, and Renal Anomalies in the Newborn.
1986 ISBN 0-89838-766-3
12. D.G. Oreopoulos (ed.): Geriatric Nephrology. 1986 ISBN 0-89838-781-7
13. E.P. Paganini (ed.): Acute Continuous Renal Replacement Therapy. 1986
ISBN 0-89838-793-0
14. J.S. Cheigh, KH. Stenzel and AL. Rubin (eds.): Hypertension in Kidney Disease. 1986
ISBN 0-89838-797-3
15. N. Deane, R.J. Wineman and G.A. Benis (eds.): Guide to Reprocessing of
Hemodialyzers. 1986 ISBN 0-89838-798-1
16. C. Ponticelli, L. Minetti and G. D'Amico (eds.): Antiglobulins, Cryoglobulins and
Glomerulonephritis. 1986 ISBN 0-89838-810-4
17. J. Strauss (ed.) with the assistence of L. Strauss: Persistent Renalgenitourinary
Disorders. 1987 ISBN 0-89838-845-7
18. V.E. Andreucci and A Dal Canton (eds.): Diuretics. Basic, Pharmacological, and
Clinical Aspects. 1987 ISBN 0-89838-885-6
19. P.H. Bach and E.H. Lock (eds.): Nephrotoxicity in the Experimental and Clinical
Situation, Part 1. 1987 ISBN 0-89838-997-1
20. P.H. Bach and E.H. Lock (eds.): Nephrotoxicity in the Experimental and Clinical
Situation, Part 2. 1987 ISBN 0-89838-980-2
21. S.M. Gore and B.A Bradley (eds.): Renal Transplantation. Sense and Sensitization.
1988 ISBN 0-89838-370-6
22. L. Minetti, G. D' Amico and C. Ponticelli: The Kidney in Plasma Cell Dyscrasias. 1988
ISBN 0-89838-385-4
23. AS. Lindblad, J.W. Novak and KD. Nolph (eds.): Continuous Ambulatory Peritoneal
Dialysis in the USA. Final Report of the National CAPD Registry 1981-1988. 1989
ISBN 0-7923-0179-X
Developments in Nephrology
24. V.E. Andreucci and A. Dal Canton (eds.): Current Therapy in Nephrology. 1989
ISBN 0-7923-0206-0
25. L. Kovacs and B. Lichardus: Vasopressin. Disturbed Secretion and its Effects. 1989
ISBN 0-7923-0249-4
26. M.E. de Broe and l.W. Coburn (eds.): Aluminum and Renal Failure. 1990
ISBN 0-7923-0347-4
27. K.D. Gardner Jr. and 1. Bernstein (eds.): The Cystic Kidney. 1990
ISBN 0-7923-0392-X
28. M.E. de Broe and G.A. Verpooten (eds.): Prevention in Nephrology. 1991
ISBN 0-7923-0951-0
29. T.A. Depner (ed.): Prescribing Hemodialysis. A Guide to Urea Modeling. 1991
ISBN 0-7923-0833-6
30. V.E. Andreucci and A. Dal Canton (eds.): New Therapeutic Strategies in Nephrology.
Proceedings of the 3rd International Meeting on Current Therapy in Nephropology
(Sorrento, Italy, 1990). 1991 ISBN 0-7923-1199-X
31. A. Amerio, P. Coratelli and S.G. Massry (eds.): Turbulo-Interstitial Nephropathies.
Proceedings of the 4th Bari Seminar in Nephrology (April 1990). 1991
ISBN 0-7923-1200-7
32. M.G. McGeown (ed.): Clinical Management of Renal Transplantation. 1992
ISBN 0-7923-1604-5
33. C.M. Kjellstrand and J.B. Dossetor (eds.): Ethical Problems in Dialysis and Transplan-
tation. 1992 ISBN 0-7923-1625-8
34. D.G. Oreopoulos, M.F. Michelis and S. Herschorn (eds.): Nephrology and Urology in
the Aged Patient. 1993 ISBN 0-7923-2019-0
35. E.A. Friedman (ed.): Death on Hemodialysis: Preventable or Inevitable? 1994
ISBN 0-7923-2652-0
36. L.W. Henderson and R.S. Thuma (eds.): Quality Assurance in Dialysis. 1994
ISBN 0-7923-2723-3