A prospective study of pyrogenic reactions in hemodialysis patients using bicarbonate dialysis fluids filtered to remove bacteria and endotoxin.

Abstract

Pyrogenic reactions (PR) are a well-recognized complication of hemodialysis and have been associated with dialyzer reuse, high-flux dialysis, and bicarbonate dialysate. However, the roles of bacteria and endotoxin in dialysate for producing PR are not well defined. To determine the effect of removing most bacteria and endotoxin from the dialysate on the incidence of PR, a cohort of chronic hemodialysis patients receiving high-flux, high-efficiency, or conventional hemodialysis at three centers with bicarbonate dialysis fluids that had been filtered with a polysulfone high-flux hemodialyzer was prospectively studied. Unfiltered bicarbonate concentrate had median bacterial and endotoxin concentrations of 479,000 CFU/mL and 39,800 pg/mL, respectively. After filtration of the bicarbonate concentrate at the central proportioner, dialysate had a median 9.2 CFU/mL of bacteria and 17.8 pg/mL of endotoxin. Dialysate filtered at individual proportioning dialysis machines had a median 0.001 CFU/mL of bacteria and 0.19 pg/mL of endotoxin. Nine PR were identified among 303 patients after 28,007 hemodialysis treatments (0.3 PR/1,000 treatments). The rate of PR was similar for the three hemodialysis treatment modalities and for first-use compared with reused dialyzers. Although the PR rate in this study was lower (P = 0.046) than the PR rate of a previous study with unfiltered dialysis fluids (0.7 PR/1,000 treatments), it represents a difference of only 10 PR in over 28,000 treatments. It was concluded that filtration of hemodialysis fluids is efficacious in removing bacterial and endotoxin contamination and can result in a lower incidence of PR in patients receiving high-flux, high-efficiency, or conventional hemodialysis.

Full text

Page 1

I 002
Volume3
‘ Number4
‘ 1992
A Prospective
Hemodialysis
Filtered
Study
Patients
of Pyrogenic
Using
Bacteria
Reactionsin
Bicarbonate
and
Dialysis
?
Fluids
to Remove
David A.Pegues,CarlW.Oettinger,LeeA.Bland2,Jamie C.Oliver, MatthewJ. Arduino,SoniaM.
Aguero, SigridK. McAllister,StevenM.Gordon,MartinS. Favero,andWilliamR. JaMs
D.A.
McAllister.
Program,
ters
Department
lanta,
Pegues.L.A.Bland,
Favero,
M.J.
W.R.
Arduino,
Jarvis,
Infectious
Public
and
SM.Aguero, 5K.
MS.Hospital Infections
National
Disease
Center
Control.
Health
forDiseases,
Service,
Services,
Cen-
U.S.
At-
for Health
ofHuman
GA
C.W.
Renal
Oettinger.
Division,
EmoryUniversity Schoolof
Atlanta,
Medicine,
Department
of
Medicine, GA
J.C.
Oliver, DialysisClinicInc.
, Atlanta,GA
SM.
fectious
Gordon. Emory UniversitySchoolof Medicine,In-
DiseaseDivision,Atlanta,GA
(J. Am.Soc. Nephrol.1992;3:1002-1007)
ABSTRACT
Pyrogenic
plication
reactions
of hemodialysis
(PR) area well-recognized
have
com-
andbeenassociated
with
ate
endotoxin
defined.
bacteria
dialyzer
dialysate.
reuse,high-fluxdialysis,
roles
and
bacteria
PR are
of removing
dialysate
chronic
high-efficiency,
three centers
hadbeen
hemodialyzer
bicarbonate
andendotoxin
39,800pg/mL,
bicarbonate
dialysate
and17.8 pg/mI
at individual proportioning
a median0.001
bicarbon-
However,the ofand
well
most
the
in dialysate
To determine
and
PR, a cohort
receiving
hemodialysis
dialysisfluids
high-flux
studied.
medianbacterial
of 479,000CFU/mL
After filtration
centralproportioner,
CFU/mLofbacteria
Dialysatefiltered
machineshad
for producing
theeffect
from
of
high-flux,
at
that
not
endotoxin theon
incidence
patients
ventional
bonate
polysulfone
tively
had
of
hemodialysis
or con-
bicar-
with
prospec-
concentrate
concentrations
respectively.
concentrate
hada median
ofendotoxin.
with
filtered
was
a
Unfiltered
and
of theat the
9.2
dialysis
CFU/mLof bacteria
I Received
2 Correspondence
February1 7. 1992.
1. A.
Accepted
Bland,
May 6. 1992.
Infections toHospitalProgram,Mailstop CO I,
Centersfor Disease Control,Atlanta,GA30333.
I046-6673/0304-
Journal
Copyright
1002$03.OO/O
American
© 1992 by
of theSociety
American
of Nephrology
Society theof Nephrology
and
among
ments
similar
ties
Although
0.046)
unfiltered
represents
treatments.
0. 19 pg/mL
303
(0.3
for the
andfor first-use
the
than
dialysis
of endotoxin.
patients
PR/1,000
three
Nine PR were
hemodialysis
The
treatment
withreused
study
aprevious
PR/1,000
of only10 PR in over
concludedthat
identified
after
treatments).
hemodialysis
compared
in this
PR rate
fluids
28,007treat-
rate of PR was
modali-
dialyzers.
lower
study
treatments),
PR rate
the
was
(P
with
=
of
(0.7it
a difference
It was
28,000
of hemo- filtration
dialysis
endotoxin
incidence
efficiency,
fluidsis efficacious
contamination
of PR in patients
or conventional
in removing
and
receiving
hemodialysis.
bacterialand
canresult
high-flux,
in a lower
high-
Key
f/on.
Words:
b/carbonate
Hemodialysis, high-flux
endotoxin,
dialysis,pyrogenicreac-
dia/ysate,filtration
P
yrogenic
rigors,
reactions
and/or
treatments.
(PR),characterized
may
byfever,
hypotenslon, complicate he-
in-modiabysis
dude
dialysate
patibility
studies
toxin
membranes
studies,
membranes
It also
stimulation
sate
endogenous
In a previous
wefound
modialysis
ments)
trations
CFU/mL
(1 7). In
cantby
efficiency,
demiobogic
increased
diabysate
Potential sources
which
decreased
ofPR
bacteria and endotoxin,
water,
dialyzer
that
through
high-flux
have
impermeable
hypothesized
of circulating
could
cytokines
study
thatthe
treatments
despitehigh
in thebicarbonate
of bacteria
addition,
by treatment
orhigh-flux
investigations
riskof
bacterial
contaminate
or
of certain
have
passing
productand
membranes
PRmay
biocom-
(1 -9).
from
permeable
Some
endo-suggested
directly
of
arise
more
(10-12).
intact
endotoxin
thattransmembrane
mononuclear
producePR
(16).
of PR atthree
incidenceof
waslow
bacterial and
dialysate
and 380pg/mL
the PR ratedid
modality
hemodiabysis).
in the
PR wasassociated
counts(1),our
the
dialyzers
shown
Other
however, thatdialyzer
(13-16).are to
has been
cells
liberation
by dialy-
endotoxinby theof
dialysis
after
PR/i
endotoxin
centers,
PR
(0.7
26,877
,000
he-
treat-
concen-
19,000(mean,
of
differ
endotoxin)
not signifl-
high-(conventional,
Although
found
with
prospective
epi-
1970s thatan
increased
study

Page 2

Pegueset al
Journalof the American Societyof Nephrology
1003
showed
the
tenial
a high
ing
Since
high-flux
ters
tens
central
abysis
from
prospectively
alysis
or conventional
centers
tiably
used
that
of bicarbonate
andendotoxin
PRmate,
high-flux
that
hemodialysis
hasincreased
also have
proportioner
machines.
thedialysate.
if diabyzerswere carefully
containing
reprocessed,
high
not
patients
usedialysate
concentrations
In hemodialysis
bac-
didresultin
evenmeceiv-
hemodialysis
study
(17).
wascompleted,
performed
threefold.
filters,
individual
remove
Because
a cohort
receiving
hemodialysis
theincidence
andendotoxin
PRratesIn our
the
at
proportion
the
The
atthe
proportioning
and
these
ofchronic
high-efficiency,
treatments
ofPR
freedialysis
previous
of
three
three
dialysate
cen-
cen-almost
installedeither
ondi-
to bacteria
of
endotoxin
changes,we
studiedhemodi-
patientshigh-flux,
atthe three
essen-
were
to compare
bacteria
with
when
fluids
thestudy.
METHODS
StudyPopulation
Thestudypopulation
treatments
February
hemodialysis
by the
theCenters
Human
(Atlanta,
filtered dialysis
beforethe
included
performed
199
centers.
allpatients
between
three
study
Subject
Control
and
April
all
hemodialysis
1 990,
chronic
approved
Board
the
versity
1,
and24,1 , at
The
Human
outpatient
protocolwas
Institutional Review
atfor Disease
Committee
patients
three
initiated.
(CDC)
Emory
dialyzed
week
and
Uni-
with
4
Investigations of
GA).
fluids
was
All were
timesaforwk
study
CaseDefinition
A case
chills
?37.8#{176}C) in
norecorded
thedialysis
of a PR
(visible
was defined
and/or
who
symptoms
as
fever
theonset
(oral
of objective
temperature
andwho
infection
rigors)
a patient
signs
treatment.
wasafebnibe
of
had
orbefore
Case Ascertainment
Oraltemperatures were
dialysis,
obtainedforall
the
patients
dialysis
treatment
San
calibrated
Immediately
session,
with
CA)
fore
Observations
recorded
ingstaff.
were
Dialysis
before
immediately
1
the
TempPbus,
h into
dialysisand
IVAC#{176}3 (Model
thermometer.
thestudy
after
200; an Diego,
Thermometers
retested
of clinical
dialysis
from
daily into
treatment
werebe-
andmonthly.
signs
treatment
eachdialysis
a central
records
andsymptoms
record
treatment
computer
patients
were
nums-
record
system.
with
onthe bythe
Data
entered
forallsus-
3 Use
does
of Health
of trade
notimply
names
endorsement
Human
and commercial
by the Public
Services.
sources
Health
is for
Service
identification
orthe
onlyand
U.s. Department
and
pected
that
of infection
PRwerereviewed
had
the
bya physician
signs
treatment.
to ensure
thepatientnoapparent
dialysis
orsymptoms
before
PreparationofDialysisFluids
Bicarbonate
dialysis
from
every
Thewater
a sand
bow-fiber
Dialysis
endotoxin
fluxdiabyzem
many)
delivery
Filters
cessed
ing system
andwere
uab patient
with
month.
diabysatewas
dialysate
usedexclusively
was
acentral
forall
treatments.
individual
4days
Themanufactured
ingredients
methods
atlocation
bypreviously
at each
described
center
filter,
(18).
treatmentsystem
softener,
contained
andfilter,water carbon hol-
reverse
fluids
by
osmosis
were
placement
Fresenius
bicarbonate
bypreviously
the central
with
(Renatron#{176};
replaced
machines
Renatron#{176}
unit.
filtered to remove
ofa
AG,
concentrate
reported
proportloners
automated
Minntech,
30 uses.
wereremoved
eachweek
bacteriaand
thehollow-fiber
Bad
high-
Gem- (F-80;Homburg,
or
methods
were
inthedialysate
circuits
from
nightly
(19).
repro-
andialyzem
Minneapolis,
Filters
and
and
reprocess-
MN)
after from
reprocessed
replaced
individ-
aeach
DialysisMonitors andDialyzers
Conventional
with
nived
cebbubosic
High-efficiency
fined
ultrafiltration
and
190,
faceareas
anddialysate
tionwith
filtration
land,
dialysistreatments wereperformed
cellubose-de-
Deerfield,
BaxterHealthcane).
hemodialysis
hollow-fiber
of 10 to19 mL/h/mm
respectively
or F-80 [Fresenius
flow rates
of500mL/min
equipped
480; Drake-Wilbock,
hollow-fiber
(CA-
diabyzers
Baxter
containing
1 1 0;
membranes
Healthcare.
(CF-23.08;
high-flux
with
IL) on
andwasde-
as treatmentsdialyzers with
coefficientsHg
?20
Baxter
mL/h/mm
Heabthcare,
? 1 .7m2,
flow
dialysis
controller
Hg. (CA-21 0,CT-
sun- AG]),
mL/min,
conjunc-
an
blood
rates
machines
(Model
of400
in
withultra-
Port-
OR).
MethodofDialyzerProcessing forReuse
Newdiabyzers
After
processed
system
peracetic
disinfectant
ing,the
capped,
nextuse
particular
the total
thattype
diabyzems
were
patient
water
processed
use,
from
oncebeforepatient
rinsed
treat-
use.
with
ment
a
eachthe
the
diabyzers
centers’
with
peroxide-based
Minntech).
removed
roomtemperature
48h).
was
hemodlabysis
divided
used during
were
water
and
acid/hydrogen
(Renaline,
dlabyzers
andstored
(usually
type
number
ofdialyzer
of that
reprocesseda Renatron#{176} with
dialyzer
After reprocess-
machine,
until
mate
the
were fromthe
atthe
withinThe reusefora
of dialyzer
of
definedas ratioof
treatments
number
study
with
new bytheof
typethe period.

Page 3

Cl)
8)
C’)
CC’
0
0
4)
E
zi
PR in PatientsReceivingHemodialysis
1004
Volume3
. Number4
‘
1992
LaboratoryMethods
Hemodialysis
carbonate
lected
each
analysis.
diabysate
after
technique
Membrane
plates
Colonies
microscope
Endotoxin
tumbidimetnic
tem(LAL-5000;
MA)(22).
Renaline
roin/cemnic
fluids (bicarbonate
and
sterile,
formicrobiologic
of bicarbonate
collected
filters.
usedfor
were plated
incubated
werecounted
(magnification.
concentrations
Limulus
Associates
concentrate,
water)
bi-
dialysate.processed
endotoxln-fmee
werecob-
weekly
study
intubes
endotoxin
from
center and
Samples
were
in-line
was
filters
concentrate
located
membrane
microbiologic
onto trypticase
35to 37#{176}C for
withthe
x 15).
were
amebocyte
of Cape
and
andatpoints
The
before
filtration
assays
soy
48
adissecting
the
all(20).
agar
(21). andath
aidof
determined
lysate
Cod,
viaa
assaysys-
Hole, Woods
concentrations
sulfate
were
method
determinedbya fer-
titration(23).
StatisticalAnalysis
Therate
modality,
of PRwascompared
dialyzer
by
reuse
center,by treat-
byment
x2 test,
appropriate.
rate
tered
x2 or
andbystatusthe
Fisherexacttest, orx2 test
was
hemodialysis
previous
appropriate.
fortrend,where
Therate
receiving
from
test,
of PRcompared withthe
in
dialysis
Fisher
patientswithunfib-
usingfluids
exact
our studyby
where
RESULTS
Descriptive Epidemiology
Atotalof303patients
during
black,
yr).
received
study
and
During
throughout
28,277
period;
mean
study,
the
34
dialysis
because
(201 ) or
of dialysis.
patients
The
were fever
and
withPR
were febmlle
median
treatment
120to240
clusteringof
(Figure1 ).
between
[5 of9,947]
[2 of8,349] versus
1]; P
=
0.45).
hemodi-
54%
age
2 1 5 (71%)
entire
(1 1 %)
treatments
a patient’s
thetempera-
alysis
female,
yr
received
54
treatment.
were
temperature
tune
Nine
dence
signs
(22%),
(1 1 %).
all seven
end
start
was
Theme
week,
differ
PR/i
0.2
1,000
treatments
94%
thewere
were
to
thewas57
(range.24 84the
hemodialysis
began
A
excluded
period,
stopped(1 8%)treatment,
of
from
wasnot
?37.8#{176}C at
occurred
of 0.3PR/i
and symptoms
hypotension
Seven(78%)
ofthese
of thetreatment.
ofhemodlalysis
140mm(mange,
wasno
orcenter
significantly
,000treatments
PR/i,000
[2 of9,71
and
total 270
( 1 %)
the analysis
recorded
the
among
treatments.
of
(22%),
patients
patients
The
wasonset
PRninefor anmci-
.000most
(89%),
frequent
PRrigors
nausea/vomiting
hadfever
only
between
onset
only;
atthe
the
PR
time
totheof
mm).
by
rate
PR treatment
of
(Center
versus
Center
the 28,007
shift,
The
centers
PRdidnot
0.5 A,
Center
C,
B,
0.2
dialysis
PR/
Of
4
A,B, ORC
=
CENTER A, B, ORC
3
?[P N ? .
13
? .1
11
.?? .
.??
.
.....? ft?? ?
35 5 404547
Study Week
?? CASES OF PR--SHIFT
1 LI CASES OF PR--SHIFT
2
Figure
1. Casesof PR by study week,treatmentshift, and
center,April1, 1990, to February24, 1991.
treatments
(68%)
flux,
PR
PR/
ciency,
tional,
0.85).
The
ing
rates
210
reuses;
(10.7
reuses;
membranes
Among
inmean
The
Icantby
(0.4
1 ,000
nosignificant
benof dialyzer
reuses,
0.3 PR/i,000
Because
episodes
perature
the study
mediately
98.7%
session
performed
high
2,449
treatment
1 ,000treatments
0.3
0.4PR/i
atthethree
6,518
centers,
(23%)
19,040
were
and
efficiency,
(9%)
modality
were
The
high
wereconventional.
was
of 6,5
of 19.040]
[1 of 2,449];
rateof
bysimilar
181 versus
(highflux,
high
conven-
trend,
0.3
[2
[6
effi-
PR/i .000
,000
versus
x2 test for
P
=
overall
study
of
[i 1.1
mean
period
were
reuse
was
similar
range,
26]),
1 to 30),
27),
rate ofhemodialyzers
per
cellulose
32] and
tniacetate
polysulfone
conventional
reuses;
no significant
of dialyzer.
dialyzers
rate for
[1 of 2,638]
= 0.59).In addition,
inthe rate
increased
treatments
16,743];
P
=
duringdialysis
we determined
foralldialysis
Temperatures
andafterhemodialysis
respectively,
dun-
the10.8uses dialyzer.
acetate
CA-i
Mean
(CA-
[10.8
190)
(iO.0
reusefor
reuses;1 to10
(CT-range.1 tocellulose
reuses;range,(F-80)
range,i toand
(8.7
cellubosic
1 to
difference
(CF-23.08)range.28).
centers,
reuse
PR
different
PR/i
[8 of 25,369];
therewas
type rates
with
by
rate first-usewas notsignif-
thanthe PR reused
versus
dialyzems
0.3
there
the
versus
versus
,000treatmentsPR/
was
P
increase
reuses
,000
[5 of
ofPRas num-
(<5reuses?5
0.4PR/i[4 of
1.0).
identified
the
1 1 ,264J
fever89% of the
of tem-
during
of PR,
recordings
period.
before
of treatments,
in97.8%.
frequency
treatments
wererecorded
in
ati h into
im-
and
the
99.3
and
Microbial
Dialysis
andEndotoxinSurveillanceof
Fluids
Bacterial
fluids
and endotoxin
weekly
concentrations
1 through
in dialysis
orcollected(wk20)every

Page 4

Pegueset al
Journal of theAmericanSocietyof Nephrology1005
other
significantly
of testing
centration
(range,
concentration
(Table
weekthereafter
between
were
in
2 to 34 1 CFU/mL),
was
i).
(wk22to46)didnot
the
differ
resultscenters.Therefore,
mediancombined.
processed
The
water
and
bacterial
58
con-
was
median
CFU/mL
endotoxin
157
the
28pg/mL(range,4 topg/mL)
Filtrationof DialysisFluids
Highconcentrations
present
anddiabysate
i ). The median
tionsfor
were479,000
tively.After
toxin concentrations
mL(Table
filtered bicarbonate
concentrations
CFU/mL
bicarbonate
dialysis
endotoxin
30,200to
mL,respectively
The most
(bothunfiltered
Xanthomonas
nismsisolated
genes,
Pseudomonas
ofbacteria and endotoxin
concen- were
trate
intheunfiltered
used
bicarbonate
three
endotoxin
at thecenters (Table
bacterialandconcentra-
filtrationthebicarbonate
CFU/mL
filtration,
concentrate
and
median
were
samples
concentrate
of bacteria
i 8pg/mL,
dialysate
machinesreduced
concentrations
0.001CFU/mL
(Table
prevalent
and
species.
inthe
Flavobactertum,
species.
before
pg/mL,39,800respec-
endo-
155
bacterial
CFU/mL
dialysate
contained
endotoxin
and
0.4and pg/
1).Allof made from
very bow
and
respectively).
theIndividual
the
in
and
i).
bacteria
filtered)
The
processed
Methylobactertum,
(median,
Filtration
proportioning
bacterial
dialysate
1 ,700
9
andof
at
median
the
from
and
from
pg/to0.2
isolated
were
predominant
from dialysate
Pseudomonas
and
microorga-
were water
AtcaU-
and
TABLE
concentrations
hemodialysis
1.Median microbial
of unfiltered
fluids
and
and
endotoxin
filtered
FluidSpecimen
Bacteria
(CFU/mL)
(range)
Endotoxin
(pg/mL)
(range)
Reverse osmosiswater
58 28
(N=82)
(2-341)(4-157)
Bicarbonate
trate
Unfiltered
Concen-
(N=
83)
479,000
39,800
(9.440-2,360,000)(8,630-227,000)
0.4
(0.0-19)
Filtered
(N=
83)155
(8-1.720)
Bicarbonate
Unfiltered(N=55)
Dialysate
30,2001,700
(556-60,000)(426-10,500)
Filtered
proportioner
I 10)
Filtered
machine
at central9 18
(N
=
(4-433)(4-57)
at dialysis
(N
0.001
(0.0-0.6)
0.2
=
55)
(0.0-3)
The concentrationof
determined
machines
was
the
with
fluids
modalities
with
dialysis
[9 of 28,007]
0.046),
inPR
the
Renalin#{176} usedduringame-
au-processing
tomated
the
We
receiving
sus
of the
all
with
(0.3
of 26,877];
icant
modality
cyclewasfor eachof the
centers;
3.4
In patients
study)
for
2). The
compared
reprocessing
concentration
compared
hemodialysis
atthe
(range.
of
three
mean3.9%to 4.4).
also incidencePR
filtered
(previous
(this yen-
each
over-
unfiltered
three
incidence
unfiltered
PR/i
dialysisstudy)
treatment
of PR
(Table
of filtered
significantly
0.7
there
individual
the use
fluids wasbower
.000 versusPR/i
was
,000 [19
P
although
rate
two
no
treatment
signif-
differenceby
between studies.
DISCUSSION
Theuseof bicarbonate
centers
from
bicarbonate
can
production
bacteria
wasthe
contamination
expressed
contaminated
cause an
pecialby with
diabyzer membranes
dialysis-associated
bicarbonate
sociatedwith
Bacterial
bonatehemodiabysis
diabysate
United
in 1986
concentrate,
in chronic
has
in
unlike
hemo-
dialysis
dramatically
However,
concentrate,
endotoxin
tive
trate
toxin
been
intheStates
to
increased
1989
acetate
growth
of gram-nega-
22% 72%(24).
support theprolific
types
bicarbonate
bacterial
dialysate.
of bicarbonate
and
patient
morepermeable
Inthe
by
high-flux
of PR
contamination
fluids was
and
ofcertain
study,
source
ofthe
theuse
bacteria
adverse
of
(4).
diseases
and
risk
endotoxin
(25).
primary
In this concen-
ofand endo-
Concernhas
that
with
dialysate
endotoxincould
increase inreactions, es-
the usehigh-flux
i 989
the
surveillance
CDC,
dialysis
of
of theuse
dialysate
a higher
and
wasas-
(24).
ofbicar-
substantially re-
TABLE
with
modality
2.Incidence
or
of PR in hemodialysis
dialysis
patients
treatmentfilteredunfiltered fluidsby
Type
Treatment
of
No.of
Treatments
No.of
PR
PR/I
Treatments
.000
Value#{176}
Conventional
Filtered
Unfiltered
HighEfficiency
Filtered
Unfiltered
High Flux
Filtered
Unfiltered
Total
Filtered
Unfiltered
2,474
13,123
I
7
0.4
0.5
1.000
19,245
I 1,345
6
9
0.3
0.8
0.066
6,288
2,409
2
3
0.3
1.2
0.134
28,007
26,877
90.3
0.7
0.046
19
0 Filteredversus unfiltered dialysis fluids.

Page 5

PR in PatientsReceivingHemodialysis
1006
Volume3
. Number4
. 1992
duced
were
sate.
or
were
dialysate
not
toxin
with
The
all
the
mentation
of
water
filtered
with
gram.
(0.3/i
were
tervention.
the
patients
hypotension.
Although
ciated
PR
(when
fenence
duction
ence
though
significantly
significant
dalities
sitive
tion
recorded
marked
had
and
efficiency
from
conventional
Earlier
the
high-flux
stimulation
cause
ences
high-flux,
this
our
alysis
studies
readily
permeable
byfiltration.Bacteriaandendotoxin
bicarbonate
a central
dialysis
colonization
Although
thebacterial
reduced>99.9%
concentrate
toreprocess
this study
Advancement
bacterialand
and <1 ng/mL,
reprocessdialyzers
dialysis
controlleddialyzer
associated witha very
treatments).Most
reactionsand did
Seven(78%)of the
definitionforPR
(22%)developed
that
occasionally
whether
individual
probably
detectedinthe dialy-
preparedfromproportioner
machines,atproportioning
to
the
diabysate,
were
bicarbonate
water
during
for
(AAMI)
CFU/mL
to
bicarbonate
a carefully
was
,000
mild
due
after
bacterial
filter.
inthe
didlines
in sterile
filtration
result and
compared
or diabysate.
dialyzers
endo-
concentrations
unfiltered
processed
three
Association
usedat
centersconsistently
Medical
met
ofInstru-
endotoxinstandards
<200respectively,
(26).
in
reprocessing
lowincidence
of thePR
not require
nine patients
developed
rigorsand
for
usedThe
conjunction
use of
fluids,
pro-
of PR
occurthatdid
medicalin-
who
only;
met
twocase fever
accompanying
the
a
filtration
twofold
with
dialysis
be
incidence
iOPR
overall
(P
=
reduction
(Table 2).
definition
ofpatient
temperatures
changes
occurred
study2.
treatments
49to 68%,
of dialysis
reduction
incidence
fluids
interpreted
of
Inover
incidence
0.046),
inPR
Also,although
forPRwas
treatment
during
inthe
thethree
Thenumber
increased
respectively,
treatments
studieshave
maypass
diabyzers or
ofcirculating
a PR ( 10-12).
intheincidence
high-efficiency,
studywithfiltered
previousstudy
fluids were used.
thatindicate
throughdialyzer
high-flux
fluids
the
our
wasasso-
withinincidence
first
used),
caution.
represented
28,000treatments.
ofPR
therewas
bytypeof treatment
a uniform
usedand
records
thetwo
typeofdialysis
centers
ofhigh-flux
from
whereas
fellfrom49
suggestedthat
throughthe
mayinduce
monocytes
However, no
of PR were
or conventional
dialysis fluids
whenhighly
Thiswould
thatendotoxins
membranes,
membranes
of
comparedthein study
unfiltered
should
in
only
the
werethis
The
a differ-
dif-
me-with
the PR
ofAl-
was reduced
nostatistically
mo-
sen-and
a high
had
study
propor-
three
periods,
all
treatments
atbetweenstudy1
andhigh-
and9 to
the
to 9%.
endotoxmn
membranes
transmembrane
and
significant
found
23%
numberof
in
ofdialysate
thereby
differ-
between
dialysis in
compared
contaminated
support
with
di-
previous
not
the
1 6).
dopass
more
These
even
(1 3-
findings
tional
diseases,
in
ence
herent
definition
CDC
ing
study,
PR
center,
a center.
In this
fluids
tions
undetectable
the
forPR.
associated
reuse
overall
resent
dialysate.
In conclusion,
membrane
abysis
toxin
wasassociated
lowincidence
high-efficiency,
tensthat
strictly
diabyzer
differ from results
the
obtained
of dialysis-associated
association
dialysis(24).
rigorous data
systems.
for the1989
on temperature
to thedefinition
on passive
therate
or morePR
fromthena-
surveillance
which
byCDC
anreported
high-flux
between
This
collection
Although
survey
> 1 00#{176}F dun-
used
self-reporting
ofPR at
hadoccurred
PR
patientsand
reflects
differ-
likely
in passive
less in-
the
the
surveillance
used
of rigors
similar
relied
estimate
whether
for PRby
(1. e.
dialysis)
, onset
was in this
the survey
not
of
and dida given
onlyoneat
study,the filtration
bacterial
ofbicarbonate
endotoxin
dialysis
resulted
in
inand
dlalysate
concentra-
approached thebicarbonate
bevels.
study
finding
with
contribute
rate
a lower
that
patients
case
additional
However,
period
suggests
nineduring
1 i -month metour definition
factors
dialyzer
PR.
may
bicarbonate
Thisthat
treatment
incidence
treatments
rate
hemodialysisand
ofmay
PR
tothe
,000
PR
The
rep-of 0.3
limit
PR/i
for thewith
wefoundthat polysulfone
filter
bacterial
of
reduction
receiving
hemodialysis
hemodiabyzers
standards
high-flux
dialyzems
markedly
used to bicarbonatedi-
fluids
contamination.
reduced
The
a further
inpatients
or conventional
reprocessed
to AAMI
and endo-
usefiltereddialysate
already
high-flux.
in cen-
with
of
in an
PR
carefullyand
adhered
reprocessing.
forwaterusedin
REFERENCES
1 .
Favero
their
Hospital
Brown,
Hindman
NJ,
actions
mural
Petersen
Pyrogenic
tion
Transplant
Man
associated
branes
bicarbonate
1 988;62:24-34.
Taniguchi
dotoxemia
1 990;56:44-49.
Beck-Sague
Outbreak
rogenic
J Nephrob
Gordon
MS.Dialysis-associated
In:Bennett
Infections.
Company;
SH,Favero
Shonberger
during
endotoxin.
NJ,Boyer
reactions
of a dialysis
1 978;52:57-60.
NK, Ciancioni
adverse
andmicrobial
concentrate.
diseasesand
eds.control.JV,Brachman
Ed.
PS,
2nd
1985:267-284.
MS,Carson
Solano
Boston:Little,
and
2.LA,
Pyrogenic
Petersen
LB. JT:
caused
me-
haemodialysis
Lancet
by extra-
1 975;2:732-734.
Carson
inadequate
distribution
3. KM.
from
LA, Favero MS:
dismnfec-
fluid system. Dial
4.
C, Faivre
reaction
JM,
et
high-flux
at.:
Dialysis-
with mem-
liquidcontaminationof
ContnibNephrol
5.T,
in
Katsushima
patients
5,LeeK,
etat.:
Nephron
En-
onhemodialysis.
6.CM,
gram-negative
JarvisWR,BlandLA,
etat.:
py-
Am
ofbacteremiaand
reactions
1990;iO:397-403.
SM,Tipple
inahemodialysiscenter.
7.
M,BlandLA,
etat.:
Pyro-

Page 6

PR in PatientsReceivingHemodialysis
??
..? ? . .
?. ?
Journal of theAmericanSocietyof Nephrology
1007
genic
posable
1988;260:2077-208
Centers
datedwith
modialyzers.
Henderson
donS:
kinhypothesis.
Laude-Sharpe
C,Kazatchkine
tion of
brane passage
Int i990;38:
Haeffner-Cavaillon
C,
etaL:
inghemodialysis.
i218.
Klinkmann
Investigation
permeable
Contnib
Bommer
Urbaschek
fem across
ClinNephrob
Bernick
endotoxemia
ClmnNephmol
Powell
ofplasma
factor-a
alysis
i99i;2:i007-iOi3.
Port FK,
Kluger
tionof
hemodialysis.
122.
Gordon
Pynogenic
ventional,
alysistreatments
containing
reactionsassociatedwiththereuseof dis-
JAMAhollow-fiberhemodiabyzers.
i.
Control:
of disposable
1 986;35:4
Koch
hypotension:
BloodPunif
M, Caroff
M,Haeffner-Cavillon
duringhemodlalysis:
ofintact
i089-1094.
N,
vivoinduction
Kidney
8.
forDisease
reuse
MMWR
LW,
Bacteremia
hollow-fiber
17-418.
Dinarello
asso-
he-
9.KM.CA,Shal-
Hemodialysis theinterleu-
1 983;1:3-8.
1 0.
M, SimmarL, Pusineri
N:
tnansmem-
(LPS).
Induc-
IL-i
endotoxinKidney
i i .
CavaillonJM, Ciancioni
Inof interleukin-
Int
i dun-
1 2 i 2-i 989;35:
12.H,Falkenhagen
the
D,SmollichBP:
ofpermeability
membranes
of
pynogens.
highly
pobysulfone
Nephrol
J,Becker
B:
high
for
i985;46:174-183.
KP,
evidence
flux
i986;27:278-282.
Port F:
despite
i980;i4:i3-i7.
BlandLA,
Intemleukin-i?9
elevation
conditions.
13. Urbaschek
for
polysulfone
R,Ritz
trans-
E,
No endotoxin
membranes.
i 4.
J,Absenceofbacteremia and
contaminated dialysate.
i 5.AC,OettingerCW,
tumor
et at. : Lack
necrosis
hemodi-
Nephmob
and
unfavorable
Am
during
JSoc
16.VanDeKerkhove
The
intemleukin-
KM.Kunkel
the
SL,
MJ:robe of diabysate
1 production
Kidney
in
during
1 987;
stimuba-
clinical
10:118- AmJ Dis
1 7.
SM,Oettinger CW,
patients
BlandLA,
etat.:
reactions
high-efficiency,
inreceivingcon-
orhigh-fluxhemodi-
dialysatewithbicarbonate
highconcentrations ofbacteriaand
endotoxin.
i444.
Oettinger
process
duction:
dialysis
432-437.
Oliver
tenla
dialysis
ficiency
gans
American
Greenberg
Standard
and
American
Arduino
Effects
microbiologic
alysis
1465.
Remillard
sky
using
Levin
Endotoxin
Assay.
European
European
France:
Alter
National
eases
Intern
Bland
bacteriologic
with
Transactions
Association
Instrumentation.
Reuse
ation
mentation;
J Am SocNephrol i992;2:l436-
1 8.
CW,
In
Comparison
population.
Oliver
center
JC: Aneconomical
dialysate
acetate
Organs
new
pro-
large
forbicarbonate
with
Antif
ina
i 989; 13:
i 9.
JC,Bland
endotoxin
fluids
and
i992;i6:l4l-i45.
Public
AE.
?Methods
Wastewater.
Public
MJ,
of incubation
LA, Oettinger
removal
in
flux
CW,
etat.:
Bac-
and frombicarbonate
foruseconventional,
hemodialysis.
high
Artif
ef-
Or-high
20.HealthAssociation.
Clesceni
In:
Trussell RR,
Examination
Ed.
Association;
Aguero
time and
procedures
Microbiol
LS,
of Water
eds.
for
16th
the
Washington, D.C.:
Healthi 985.
Favero 2 1 .
BlandLA, SM,
temperature
MS:
on
sampling
J
for
i ;29:
hemodi-
1462-fluids.Cbini 99
22.JF, GouldMC,
of
Roslansky
endotoxin
assay.
Detection
PF, Novit-
TJ.Quantitation
LALturbidimetnic
J, Novitsky
with
NewYork:
Pharmacopoeia.
Treaty
Council
MJ,Favero
surveillance
In theUnited
Organs
LA,Ridgeway
inproducts
In: Watson
of Bacterial
SW,
TJ, eds.
the
Limulus
Alan
Amoebocyte
1987:197-210.
Vol.3.4.5.
50. Sainte-Ruffine,
1980.
MoyerLA.
dialysis-associated
1989.
Lysate
R. Liss;
23.2ndEd.
Series
of Europe;
MS.
no.
24.BlandLA:
dis-
Artif
of
States,AmSoc
i99i;37:97-i09.
25.MR.AgueroSM:Potential
associatedand
bicarbonate
1987;33:542-545.
for the
endotoxinhazards
concentrate.liquidASAIO
26.Advancement
Recommended
ofMedical
Practice
VA:
for
of
for
Hemodialyzers.
the
1986.
Arlington.Associ-
Instru-AdvancementofMedical