Peritoneal Dialysis International, Vol. 34, pp. 162–170 0896-8608/14 $3.00 + .

00
doi: 10.3747/pdi.2013.00275 Copyright © 2014 International Society for Peritoneal Dialysis

original articles

32 YEARS’ EXPERIENCE OF PERITONEAL DIALYSIS–RELATED PERITONITIS

IN A UNIVERSITY HOSPITAL

Sadie van Esch,1 Raymond T. Krediet,2 and Dirk G. Struijk1

Nephrology1 and Renal Unit,2 Academic Medical Centre, Amsterdam, Netherlands

♦  Background:  Peritonitis in peritoneal dialysis (PD) group 4. The overall peritonitis rate and the gram-positive
patients can lead to technique failure and contributes to and gram-negative peritonitis rates showed a time-trend
infection-related mortality. Peritonitis prevention and opti- of decline. However, the duration of antibiotic treatment
mization of treatment are therefore important in the care for increased over time, with groups 3 and 4 having the longest
PD patients. In the present study, we analyzed the incidence duration of treatment, accompanied by a higher percentage
of peritonitis, causative pathogens, clinical outcomes, and of antibiotic switch. Increased resistance to cephradine was
trends in relation to three major treatment changes that found for coagulase-negative Staphylococcus.
occurred from 1979 onward: use of a disconnect system since ♦  Conclusions:  Peritonitis rates declined significantly
1988, daily mupirocin at the exit-site since 2001, and exclu- over the years because of several changes in PD treatment.
sive use of biocompatible dialysis solutions since 2004. However, the need to change the initial antibiotic increased
♦  Methods:  In this analysis of prospectively collected data, because of diminished antibiotic susceptibility rates over
we included peritonitis episodes from the start of PD at our time. Nevertheless, the cure rate was high and remained
center in August 1979 to July 2010. Incident PD patients stable during the entire period analyzed, and the death rate
were allocated to one of four groups: Group 1 – 182 patients remained low. Consequently, peritonitis is a manageable
experiencing 148 first peritonitis episodes between 1979 complication of PD that cannot be considered a contraindica-
and 1987, before the introduction of the disconnect system; tion to this mode of renal replacement therapy.
Group 2 – 352 patients experiencing 239 first episodes of
peritonitis between 1988 and 2000, before implementation Perit Dial Int 2014; 34(2):162–170 www.PDIConnect.com
of daily mupirocin application at the catheter exit-site; epub ahead of print: 01 Mar 2014   doi:10.3747/pdi.2013.00275
Group  3  – 79 patients experiencing 50 first peritonitis
episodes between 2001 and 2003, before the switch to KEY WORDS: Infection; peritonitis; peritonitis
biocompatible solutions; and Group 4 –118 patients experi- outcomes.
encing 91 first peritonitis episodes after 2004. Cephradine
was used as initial antibiotic treatment.
♦  Results:  In 32 years, 731 adult patients started PD, and
2234 episodes of peritonitis in total were diagnosed and C hronic peritoneal dialysis (PD) treatment without the
occurrence of peritonitis is impossible. Although most
cases of peritonitis follow a rather benign course with
treated. Of those episodes, 88% were cured with medical
treatment only, and 10% resulted in catheter removal. In 3% appropriate antibiotic treatment, some episodes are com-
of the episodes, the patient died during peritonitis. Median plicated by hospitalization and temporary or permanent
time to a first peritonitis episode increased from 40 days loss of the PD catheter. Severe and long-lasting peritonitis
for group 1 to 150 for group 2, 269 for group 3, and 274 for probably leads to peritoneal membrane failure and hence
Correspondence to: S. van Esch, Nephrology, Academic the need to discontinue PD and switch to hemodialysis. In
Medical Centre, Meibergdreef 9, Amsterdam  1105  AZ a recent study in 709 incident PD patients participating
­Netherlands. in the Netherlands Cooperative Study on the Adequacy
s.vanesch@amc.uva.nl of Dialysis, infections were the reason for a switch to
Received 24 October 2013; accepted 31 December 2013 hemodialysis in 10% – 18% during the first 3 years of PD,
162
PDI march  2014 - Vol. 34, No. 2 32 YEARS’ EXPERIENCE OF PD-RELATED PERITONITIS

depending on duration of treatment (1). Moreover, peri- and culture. Peritonitis was defined as the presence of
tonitis contributes to a significant portion of infectious- two of three criteria: namely, clinical symptoms, effluent
related mortality in PD patients (2). Peritonitis prevention cell count exceeding 100 cells/μL, and a positive culture.
and optimization of peritonitis treatment are therefore Those criteria were developed by Vas and colleagues
important aims in the care of PD patients. (3) and are confirmed in the current guidelines of the
In 1979, chronic PD was started as a renal replace- International Society for Peritoneal Dialysis (4). Except
ment therapy in our center. In the years thereafter, for the first cell count after presentation, all counts
three important treatment changes were realized. First, were performed in peritoneal effluent after the long
we switched to disconnect (twin bag) systems in 1988; dwell. Patients who were not admitted to the hospital
second, application of mupirocin at the catheter exit were instructed to bring the first clear night bag to the
site was implemented in 2001; and third, biocompatible outpatient clinic. Initial treatment consisted of a first-
dialysis solutions have been used in all patients since generation cephalosporin, which was combined with
2004. Since 1979, 731 adult patients have been started gentamicin when the patient was clinically ill and needed
on PD and a total of 2234 episodes of peritonitis have hospitalization. Cephradine was the first-generation
been diagnosed and treated. The aims of present study in cephalosporin used. It was replaced by cephalothin after
this cohort were to analyze the incidence of peritonitis, cephradine was removed from the market in 2008. When
the causative pathogens, the clinical outcomes, and the hypersensitivity reactions occurred, vancomycin was
trends in the above findings in relation to the three major given. Antibiotic treatment thereafter could be adjusted
treatment changes. according to the resistance of the causative organism.
Treatment duration was 1 week after cultures became
METHODS negative and cell counts reached less than 100  cells/
μL. The protocol was published in 1985 and has not been
PATIENTS changed since (5).

All patients 18 years of age and older who were treated STATISTICAL ANALYSIS
with PD at the Academic Medical Center from 1979 to July
2010 were included in the study. From the day PD was To analyze possible associations between categorical
started, all peritonitis data were collected prospectively data, the chi-square test was applied. For continuous
in a large database. The data include patient character- variables, a Kruskal–Wallis or Mann–Whitney test was
istics, start and duration of every peritonitis episode, used. Results are expressed as means and standard devia-
causative organisms, dialysate leukocyte counts, type tions, or medians and ranges.
and duration of antibiotic treatment, antibiotic resis- For the analysis of clinical outcomes and ­peritonitis
tance, and outcome of the peritonitis, which could be ­incidence in relation to the three major changes in PD treat-
cure, change of PD catheter, transfer to hemodialysis, ment, 4 groups of incident PD patients were formed:
or death. The time in months from PD start to the first
• Group  1 consisted of 182 patients who started PD
episode of peritonitis was computed. Primary kidney
between 1979 and 1987—thus, before the introduction
disease was classified into 9 main categories ­according
of disconnect systems. These 182 patients experienced
to the codes of the European Renal Association  –
148 first peritonitis episodes.
European Dialysis and Transplant Association. Patients
• Group  2 consisted of 352 patients who started PD
with two or more pathogens isolated from the cul-
between 1988 and 2000—that is, before implementa-
tures in a peritonitis episode were considered to have
tion of daily mupirocin at the catheter exit-site. They
polymicrobial peritonitis. A relapse of peritonitis was
experienced a total of 239 first peritonitis episodes.
defined when the same organism or a culture-negative
• Group 3 contained 79 patients who started PD between
peritonitis episode was identified within 4 weeks
2001 and 2003, before the switch to the sole use of
of completing appropriate antibiotic therapy for a
biocompatible solutions, and they experienced 50 first
previous episode.
peritonitis episodes.
• Group 4 consisted of the remaining 118 patients, who
PERITONITIS DIAGNOSIS AND TREATMENT PROTOCOL
started PD after 2004 and who experienced 91 first
peritonitis episodes.
In patients with any combination of cloudy effluent,
abdominal pain, and fever, a sample of the effluent was The Jonckheere trend test was used to analyze lin-
obtained for a cell count with differential, Gram stain, ear trends in the peritonitis incidence over time, and

163
VAN ESCH et al. march  2014 - Vol. 34, No. 2 PDI

the Kendall tau was used to calculate the correlation PERITONITIS INCIDENCE AND OUTCOMES
­coefficient. All statistical analyses were performed using
SPSS for Windows (version 19.0: IBM, Armonk, NY, USA). No peritonitis occurred in 203 of the 731 patients
A p value less than 0.05 was considered significant. (28%). The remaining 528 patients experienced 2234
peritonitis episodes in total. Median time from the
RESULTS start of PD to a first peritonitis episode was 128 days
(range: 0 – 3450 days). Relapsing peritonitis accounted
GENERAL for 301 of the 2234 episodes (13%). Of the remaining
1933 episodes, cultures identified a single organism in
Between August 1979 and July 2010, 731 patients 18 71% and polymicrobial infection in 21%; 9% of episodes
years of age and older (57% men) started treatment with were culture-negative. The peritonitis rate declined
PD. Median age at PD start was 52 years (range: 18 – 87 to 1.4 episodes per patient–year in 1989 (1 year after
years). Diabetic nephropathy was the cause of renal failure the introduction of the twin-bag system) from 4.4
in 22% of the patients. Table 1 provides further details. episodes per patient–year in 1979. The rate further
Death was the reason for PD discontinuation in 38% of declined to 0.95 episodes per PD–year in 2010, as shown
the patients, followed by switch to hemodialysis (29%) in Figure 1.
and renal transplantation (28%). In 3% of patients, Most of the peritonitis episodes (88%) were cured
PD was able to be stopped because of return of kidney by antibiotic treatment. Catheter removal was required
function, and 3% of the patients transferred to another in 10% of the episodes. In 3%, the patient died during
hospital and were lost to follow-up. Causes of death were peritonitis. Analysis of cause of death in those episodes
classified as cardiovascular (29%), infectious complica- showed that 45% (n = 30) died from peritonitis, and 6%
tions (22%), treatment refusal (11%), and unidentified (n = 4) died from other infections. Death from cardio-
(35%) when the patient died suddenly at home, probably vascular causes during peritonitis occurred in 33% (n =
from cardio- or cerebrovascular causes. 22) of the patients.

TABLE 1
Demographic and Baseline Clinical Characteristics of Incident Peritoneal Dialysis (PD) Patients Treated
During the Study Period

Patient group
1 2 3 4
Twin bag Twin bag,
Single Twin and mupirocin, and
Characteristic bag bag mupirocin biocompatible Overall

Time period 1979–1987 1988–2000 2001–2003 2004–2010 1979–2010

Patients (n) 182 352 79 118 731
Sex (men:women) 96:86 213:139 43:36 67:51 419:312
Mean age (years) at PD start 49±15 52±16 53±17 54±17 52±16a
Cause of ESRD (%)
Glomerulonephritis 26 16 11 12 17a
Interstitial nephritisb 19 10 10 3 11a
Cystic kidney disease 6 7 5 7 7
Congenitalc 1 1 6 3 2a
Renal vascular disease 8 19 17 21 16a
Diabetic nephropathy 21 23 18 21 22
Other multisystem diseases 9 9 11 12 10
Other or unknown 11 14 22 21 15a
Modality (APD:CAPD) 1:181 19:332 25:54 52:21 105:625a

ESRD = end-stage renal disease; APD = automated PD; CAPD = continuous ambulatory PD.
a p < 0.05 (analysis of variance).
b Includes pyelonephritis, drug induced nephropathy, and urolithiasis.
c Other congenital and hereditary kidney diseases.

164
PDI march  2014 - Vol. 34, No. 2 32 YEARS’ EXPERIENCE OF PD-RELATED PERITONITIS

CAUSATIVE PATHOGENS and Pseudomonas (32%) resulted in the greatest catheter

loss, followed by S. aureus. Culture-negative peritonitis
Table 2 shows the distribution of causative pathogens in resulted in 4% of catheter loss. Catheter removal was
all peritonitis episodes. Most of the peritonitis episodes were required in 9% of non-relapsing peritonitis episodes and
caused by gram-positive organisms (57%). The organism in 14% of relapsing peritonitis (p < 0.01).
most often cultured was coagulase-negative Staphylococcus
[CNS (30%)], followed by S. aureus, Streptococcus species, ANALYSIS OF PERITONITIS EPISODES IN RELATION TO
and Enterococcus species. In 11% of all peritonitis episodes, THE THREE TREATMENT CHANGES
gram-negative organisms such as Pseudomonas species,
Escherichia coli, and others were found. Fungi were the Table  1 shows the baseline characteristics of the 4
cause of peritonitis in 1% of episodes. patient groups. Over the years, patients were older at
Figure 2 shows PD catheter removal for all peritonitis the start of PD, and more patients started automated PD
episodes, by organism, but with the exclusion of poly- than continuous ambulatory PD. The causes of end-stage
microbial peritonitis. Peritonitis caused by fungi (38%) renal disease were different in the groups.
Median time in days to a first peritonitis episode was
40 for group 1, 150 for group 2, 269 for group 3, and 274
for group 4 (p < 0.01). Figure 3 shows the Kaplan–Meier
analysis for time to a first peritonitis episode in the
patient groups. Table 3 summarizes peritonitis outcome,
antibiotic treatment, and antibiotic switches in incident
patients with a first peritonitis. Compared with group 2,
the single-bag group (group 1) had a lower percentage
of catheter removal. Median duration of antibiotic treat-
ment increased over time, with the longest duration of
treatment observed in groups 3 and 4. Those longer times
were accompanied by a higher percentage of antibiotic
switch, as Table 3 shows.
Trends in resistance to flucloxacillin by S. aureus did
Figure 1 — Peritonitis incidence per year. ISPD = International not change significantly over time. The resistance to
Society for Peritoneal Dialysis. cephradine by CNS increased to 5.8% in group  3 from
0.8% in group 1. The susceptibility of Pseudomonas to
TABLE 2 aminoglycosides showed no change.
Distribution of Causative Pathogens in Peritonitis

Organism [n (%)]

Overall 2234 (100)

Gram-positive
CNS 681 (30)
Staphylococcus aureus 342 (15)
Streptococcus sp. 149 (7)
Enterococcus sp. 32 (1)
Others 88 (4)
Gram-negative
Pseudomonas sp. 53 (2)
Escherichia coli 44 (2)
Others 152 (7)
Fungi 24 (1)
Mycobacterium 2 (0.1)
Others 16 (1)
Polymicrobial 429 (19)
Culture-negative 222 (10) Figure 2 — Peritoneal dialysis (PD) catheter removal, by orga­
nism. CNS = coagulase-negative Staphylococcus; S. aureus =
CNS = coagulase-negative Staphylococcus. Staphylococcus aureus.

165
VAN ESCH et al. march  2014 - Vol. 34, No. 2 PDI

Table 4 compares peritonitis rates between the groups. species was observed. A time-trend of decline was found
The rates of S.  aureus and S.  epidermidis peritonitis for the rate of relapsing peritonitis (p ≤ 0.01).
declined significantly over time (p = 0.01). After the sole
use of biocompatible dialysis solutions, gram-negative DISCUSSION
peritonitis attributable to E. coli declined (p = 0.03); how-
ever, no decrease in peritonitis caused by Pseudomonas In the present study, 32 years’ experience of 2234
PD-related peritonitis episodes in a large university
­hospital were analyzed. Over those years, three impor-
tant changes were made in the care of PD patients:
specifically, use of a disconnect system since 1988, daily
mupirocin at the exit-site since 2001, and use only of
biocompatible dialysis solutions since 2004.
A significantly declining trend was observed in the
overall peritonitis rate, and in episodes caused by both
gram-positive and gram-negative organisms. Our find-
ings confirm those of other, much smaller studies of the
dramatic effects that the use of disconnect systems,
which started in 1988, have had on the peritonitis inci-
dence (6–10). The decreases were especially marked for
CNS and S. aureus peritonitis.
The introduction of daily mupirocin application at the
catheter exit site in 2001 led to a further reduction in
S. aureus infections, not only at the exit-site, but also
as peritonitis. A Cochrane analysis by Strippoli et al.
Figure 3 — Time to first peritonitis per group, maximal follow- (11) published in 2004 reported that nasal mupirocin
up of 5 years. reduced exit-site and tunnel infection, but not ­peritonitis.

TABLE 3
Peritonitis Outcomes and Antibiotic Treatment Duration and Switch in the Study Patientsa

Patient group
1 2 3 4
Twin bag Twin bag,
Single Twin and mupirocin, and
bag bag mupirocin biocompatible
Variable (n=148) (n=239) (n=50) (n=91)

Time period 1979–1987 1988–2000 2001–2003 2004–2010

Peritonitis outcome (%)
Cure 94 86 92 90
Catheter removalb 3.4 12.1c 4.0 8.8
Death during peritonitis 2.7 2.1 4.0 1.1
Antibiotic treatmentb
Median duration (days) 13 14 16d 16e
Switch (%) 30 41f 70g 67h

a Groups include only incident patients experiencing their first peritonitis.

b p < 0.05.
c p < 0.01 compared with group 1.
d p = 0.02 compared with groups 1 and 2.
e p < 0.01 compared with groups 1 and 2
f p = 0.04 compared with group 1.
g p < 0.01 compared with groups 1 and 2.
h p < 0.01 compared with groups 1 and 2.

166
PDI march  2014 - Vol. 34, No. 2 32 YEARS’ EXPERIENCE OF PD-RELATED PERITONITIS

TABLE 4
Time Trends for Peritonitis in the Study Patientsa

Mean peritonitis rate (episodes/PD year), by patient group

1 2 3 4
Twin bag Twin bag,
Single Twin and mupirocin, and
bag bag mupirocin biocompatible
Variable (n=182) (n=352) (n=79) (n=118)

Time period 1979–1987 1988–2000 2001–2003 2004–2010

All episodes 3.22 1.30 0.89 0.95b
Gram-positive episodes 2.39 0.87 0.71 0.65b
Staphylococcus aureus 0.79 0.23 0.05 0.03b
S. epidermidis 1.23 0.35 0.35 0.13b
Gram-negative episodes 0.51 0.51 0.28 0.30b
Escherichia coli 0.20 0.20 0.21 0.04c
Pseudomonas species 0.06 0.06 0.02 0.06
Culture-negative 0.47 0.21 0.10 0.14
Relapsing 0.32 0.11 0.04 0.15b

a Groups include only incident patients experiencing their first peritonitis.

b p ≤ 0.01.
c p = 0.03.

However, that conclusion was based on only one trial was not associated with a lower overall peritonitis inci-
performed in 1996 with intranasal mupirocin. It can be dence. The literature contains conflicting data about
hypothesized that compliance by patients is higher for top- treatment modality and risk of peritonitis. Some studies
ical than for intranasal mupirocin. Recently, Xu et al. (12) showed that automated PD is associated with a lower
published a systematic review of studies that compared peritonitis risk (14,15); other studies found no differ-
mupirocin treatment with placebo or no treatment. Those ence (16–18).
authors also found that mupirocin treatment was effective Despite the decline in peritonitis incidence, our center
in preventing exit-site infection and peritonitis. often failed to achieve the goal set by the International
The total abandonment of conventional dialysis solu- Society for Peritoneal Dialysis (4) of a peritonitis rate
tions in 2004 was associated with a further reduction no more than 1 episode every 18 months (0.67 epi-
in the rates of gram-positive and E. coli peritonitis, but sodes per year at risk). That finding might be due to the
no change in the incidence of culture-negative, gram- ­continuation of PD after a few episodes of uncompli-
negative, and overall peritonitis was observed. Recently, cated peritonitis. In addition, the low socio-economic
Cho et al. performed a systematic review of randomized status of many of our patients might have played a role.
controlled trials to evaluate the benefits and harms of However, some papers report the same high incidence
biocompatible solutions compared with conventional of peritonitis (10,19,20). Reporting bias cannot be
solutions in PD patients (13). That review showed that excluded as a potential factor in lower peritonitis rates,
the use of biocompatible PD solutions might result in because authors tend to publish success rather than
better preservation of urine output and residual renal failure, as hypothesized by Davenport (10). In contrast
function and might improve inflow pain. A decline in to the improved results for peritonitis incidence, the
peritonitis episodes was not found, nor an improvement duration of antibiotic treatment for a peritonitis episode
in technique and patient survival. However, it must be increased to 16 from 13 days, and that increase was
appreciated that outcomes in the various studies of this accompanied by a higher percentage of antibiotic switch.
issue have been highly inconsistent. The results of our Because the initial antibiotic was changed according to
long-term observational study suggest that biocompat- the resistance of the causative organism, we examined
ible solutions might have some protective effect against resistance for the two major causes of peritonitis in
some, but not all micro-organisms. our series: CNS and S.  aureus. Increased resistance to
Since 2004, more patients started with automated cephradine was found for CNS; S. aureus was not tested.
PD than with continuous ambulatory PD, a change that Many other authors showed increased resistance of CNS
167
VAN ESCH et al. march  2014 - Vol. 34, No. 2 PDI

to methicillin, and consequently to first-generation similar results in 250 patients who experienced an
cephalosporins (9,21–23). With respect to flucloxacillin, episode of peritonitis in the 30 days before death,
S. aureus showed no difference in the rate of resistance except that the percentage of deaths attributable to
over time. All isolates were susceptible to vancomycin. peritonitis was lower at 28%. The authors of that paper
The International Society for Peritoneal Dialysis (4) touched on the problem of reporting a cause of death.
recommends a first-generation cephalosporin as the Because no standard definition of peritonitis-associated
initial empiric antibiotic for gram-positive coverage, mortality has yet been established, the approach to a
but underlines the importance of looking at the local diagnosis of peritonitis-associated death is variable
history of sensitivities for the organisms causing peri- and subjective. As a consequence, the prevalence rates
tonitis. Given our observations in the present study presented in the literature vary. Systemic inflammation
of significant changes in the antibiotic susceptibility may predispose to cardiovascular disease, as reported
of the main gram-positive pathogen causing perito- by others (33,34). Although peritonitis cannot always
nitis, re-evaluation of our current treatment protocol be considered a systemic inflammation, the potential
is indeed warranted. for inflammation could be an additional explanation
Over the years, our cure rate has been high and for the high percentage of cardiovascular causes of
stable, and the rate of death during peritonitis has death during peritonitis.
remained low. Catheter removal was stable over the The present study has a number of limitations. First, it
years, with the exception of a lower catheter removal is a single-center study, which makes the results difficult
rate in the early years of PD. We speculate that the lat- to apply to other centers. However, in a single-center
ter observation is related to the high incidence of CNS study, data are more complete, and accuracy is more eas-
peritonitis, which often has a relatively mild clinical ily checked than is possible with registry data. Second,
course and usually does not require catheter removal. although the peritonitis episodes were thoroughly
In contrast, catheter removal rates were significantly prospectively documented in an extensive peritonitis
higher when peritonitis was caused by a single gram- database, data collection was not part of a general pro-
negative organism or fungus than when the cause was spective study in which all imaginable parameters were
a single gram-positive organism, which is consistent collected. Hence, we were not able to distinguish the
with results in other studies (23–26). Peritonitis caused possible effects of nutrition status, patient education,
by fungi (38%) and Pseudomonas (32%) resulted in and socio-economic status on peritonitis incidence and
the greatest catheter loss. Our relatively low catheter outcome. Also, data about exit-site infections were not
removal rates in fungal peritonitis contrast with those of prospectively collected.
most studies (27–30) and with the International Society One of the strengths of this study is the large number
for Peritoneal dialysis recommendation, which advises of peritonitis episodes that were collected during 32
catheter removal immediately after fungi are identified years of PD practice. Another is that our initial antibi-
by microscopy or culture (4). Our group always fol- otic treatment protocol has not changed over the years.
lowed another policy (31), in which medical treatment Moreover, the percentage of culture-negative episodes
with intraperitoneal amphotericin and flucytosine was was just 9%. That observation underlines the high stan-
initiated and then continued for 4 weeks. In Candida dard for culture methods in our medical microbiology
albicans peritonitis, this treatment is effective in 60% laboratory, making the culture results very accurate.
of cases. It is not appropriate for fungal peritonitis
caused by other fungi or yeasts, however. A pitfall with CONCLUSIONS
the above regimen is that the dialysis effluent may
remain turbid because of a high, amphotericin-induced The practice of PD with respect to peritonitis has
cell number, which could erroneously suggest that the improved signif icantly over the years because of
medical treatment is not effective. Ignorance of that implementation of a disconnect system and daily
fact and unjustified fear of inducing peritoneal mem- mupirocin application at the exit site. A clear positive
brane damage might be reasons for considering primary effect of the use of biocompatible solutions could not
catheter removal. be found. Still, our peritonitis rates are higher than
In almost half the patients who died during peri- guideline recommendations and need to be improved.
tonitis (46%), the cause of death was assigned to the Moreover, because of decreased rates of antibiotic
peritonitis. Other infections and cardiovascular prob- susceptibility that developed over time, our initial
lems were the most important causes in 6% and 33% empiric antibiotic treatment—consisting solely of a first-­
of deaths respectively. Boudville et al.(32) reported generation cephalosporin—declined in ­effectiveness.

168
PDI march  2014 - Vol. 34, No. 2 32 YEARS’ EXPERIENCE OF PD-RELATED PERITONITIS

Yet the cure rate was and has remained stable and high, 12. Xu G, Tu W, Xu C. Mupirocin for preventing exit-site infec-
and the rate of death during peritonitis was and has tion and peritonitis in patients undergoing peritoneal
remained low. Those observations show that perito- dialysis. Nephrol Dial Transplant 2010; 25:587–92.
nitis is a manageable complication of PD that cannot 13. Cho Y, Johnson DW, Badve SV, Craig JC, Strippoli GF,
be considered a contraindication to this mode of renal ­Wiggins KJ. The impact of neutral-pH peritoneal dialysates
with reduced glucose degradation products on clinical
replacement therapy.
outcomes in peritoneal dialysis patients. Kidney Int 2013;
84:969–79.
DISCLOSURES 14. de Fijter CW, Oe LP, Nauta JJ, van der Meulen J, Verbrugh
HA, Verhoef J, et al. Clinical efficacy and morbidity
The authors have no financial conflicts of interest ­associated with continuous cyclic compared with con-
to declare. tinuous ambulatory peritoneal dialysis. Ann Intern Med
1994; 120:264–71.
REFERENCES 15. Sanchez AR, Madonia C, Rascon–Pacheco RA. Improved
­p atient/technique survival and peritonitis rates in
1. Kolesnyk I, Dekker FW, Boeschoten EW, Krediet RT. Time- patients treated with automated peritoneal dialysis
dependent reasons for peritoneal dialysis technique when compared to continuous ambulatory peritoneal
failure and mortality. Perit Dial Int 2010; 30:170–7. dialysis in a Mexican PD center. Kidney Int Suppl 2008;
2. Pérez Fontán M, Rodriguez–Carmona A, Garcia–Naveiro R, (108):S76–80.
Rosales M, Villaverde P, Valdes F. Peritonitis-related mor- 16. Oo TN, Roberts TL, Collins AJ. A comparison of peritonitis
tality in patients undergoing chronic peritoneal dialysis. rates from the United States Renal Data System data-
Perit Dial Int 2005; 25:274–84. base: CAPD versus continuous cycling peritoneal dialysis
3. Vas S, Low D, Layne S, Khanna R, Dombros N. Microbiologi- ­patients. Am J Kidney Dis 2005; 45:372–80.
cal diagnostic approach to peritonitis in CAPD patients. In: 17. Nessim SJ, Bargman JM, Austin PC, Nisenbaum R, Jassal
Atkins RC, Thomson NM, Farrell PC, eds. Peritoneal Dialysis. SV. Predictors of peritonitis in patients on peritoneal
Edinburgh, UK: Churchill Livingstone; 1981: 264–71. ­dialysis: results of a large, prospective Canadian database.
4. Li PK, Szeto CC, Piraino B, Bernardini J, Figueiredo AE, Clin J Am Soc Nephrol 2009; 4:1195–200.
Gupta A, et al. Peritoneal dialysis–related infections 18. Martin LC, Caramori JC, Fernandes N, Divino–Filho JC,
recommendations: 2010 update. Perit Dial Int 2010; ­Pecoits–Filho R, Barretti P on behalf of the Brazilian Perito-
30:393–423. [Erratum in: Perit Dial Int 2011; 31:512] neal Dialysis Multicenter Study BRAZPD Group. Geographic
5. Boeschoten EW, Rietra PJ, Krediet RT, Visser MJ, Arisz and educational factors and risk of the first peritonitis
L. CAPD peritonitis: a prospective randomized trial of episode in Brazilian Peritoneal Dialysis study (BRAZPD)
oral versus intraperitoneal treatment with cephradine. J patients. Clin J Am Soc Nephrol 2011; 6:1944–51.
Antimicrob Chemother 1985; 16:789–97. 19. Kavanagh D, Prescott GJ, Mactier RA on behalf of the
6. Daly C, Campbell M, Cody J, Grant A, Donaldson C, Vale Scottish Renal Registry. Peritoneal dialysis–associated
L, et al. Double bag or Y-set versus standard transfer peritonitis in Scotland (1999–2002). Nephrol Dial Trans-
systems for continuous ambulatory peritoneal dialysis in plant 2004; 19:2584–91.
end-stage renal disease. Cochrane Database Syst Rev 2001; 20. Moraes TP, Pecoits–Filho R, Ribeiro SC, Rigo M, Silva MM,
(2):CD003078. Teixeira PS, et al. Peritoneal dialysis in Brazil: twenty-five
7. Korbet SM, Vonesh EF, Firanek CA. Peritonitis in an urban years of experience in a single center. Perit Dial Int 2009;
peritoneal dialysis program: an analysis of infecting 29:492–8.
pathogens. Am J Kidney Dis 1995; 26:47–53. 21. Berns JS, Tokars JI. Preventing bacterial infections and
8. Holley JL, Bernardini J, Piraino B. Infecting organisms in antimicrobial resistance in dialysis patients. Am J Kidney
continuous ambulatory peritoneal dialysis patients on the Dis 2002; 40:886–98.
Y-set. Am J Kidney Dis 1994; 23:569–73. 22. Kan GW, Thomas MA, Heath CH. A 12-month review of peri-
9. Zelenitsky S, Barns L, Findlay I, Alfa M, Ariano R, Fine toneal dialysis–related peritonitis in Western Australia:
A, et al. Analysis of microbiological trends in peritoneal is empiric vancomycin still indicated for some patients?
dialysis–related peritonitis from 1991 to 1998. Am J Kidney Perit Dial Int 2003; 23:465–8.
Dis 2000; 36:1009–13. 23. Kim DK, Yoo TH, Ryu DR, Xu ZG, Kim HJ, Choi KH, et al.
10. Davenport A. Peritonitis remains the major clinical compli- Changes in causative organisms and their antimicrobial
cation of peritoneal dialysis: the London, UK, peritonitis susceptibilities in CAPD peritonitis: a single center’s expe-
audit 2002–2003. Perit Dial Int 2009; 29:297–302. rience over one decade. Perit Dial Int 2004; 24:424–32.
11. Strippoli GF, Tong A, Johnson D, Schena FP, Craig JC. 24. Bunke CM, Brier ME, Golper TA. Outcomes of single orga­
Antimicrobial agents for preventing peritonitis in perito- nism peritonitis in peritoneal dialysis: gram negatives
neal dialysis patients. Cochrane Database Syst Rev 2004; versus gram positives in the Network 9 Peritonitis Study.
(4):CD004679. Kidney Int 1997; 52:524–9.

169
VAN ESCH et al. march  2014 - Vol. 34, No. 2 PDI

25. Troidle L, Gorban–Brennan N, Kliger A, Finkelstein F. experience. J Infect 2004; 48:96–101.

Differing outcomes of gram-positive and gram-negative 30. Miles R, Hawley CM, McDonald SP, Brown FG, Rosman
peritonitis. Am J Kidney Dis 1998; 32:623–8. JB, Wiggins KJ, et al. Predictors and outcomes of fungal
26. Krishnan M, Thodis E, Ikonomopoulos D, Vidgen E, Chu peritonitis in peritoneal dialysis patients. Kidney Int
M, Bargman JM, et al. Predictors of outcome following 2009; 76:622–8.
bacterial peritonitis in peritoneal dialysis. Perit Dial Int 31. Struijk DG, Krediet RT, Boeschoten EW, Rietra PJ, Arisz L.
2002; 22:573–81. Antifungal treatment of Candida peritonitis in continuous
27. Goldie SJ, Kiernan–Troidle L, Torres C, Gorban–Brennan ambulatory peritoneal dialysis patients. Am J Kidney Dis
N, Dunne D, Kliger AS, et al. Fungal peritonitis in a large 1987; 9:66–70.
chronic peritoneal dialysis population: a report of 55 32. Boudville N, Kemp A, Clayton P, Lim W, Badve SV, Hawley
episodes. Am J Kidney Dis 1996; 28:86–91. CM, et al. Recent peritonitis associates with mortality
28. Wang AY, Yu AW, Li PK, Lam PK, Leung CB, Lai KN, et al. among patients treated with peritoneal dialysis. J Am Soc
­Factors predicting outcome of fungal peritonitis in peri- Nephrol 2012; 23:1398–405.
toneal dialysis: analysis of a 9-year experience of fungal 33. Rost NS, Wolf PA, Kase CS, Kelly–Hayes M, Silbershatz
peritonitis in a single center. Am J Kidney Dis 2000; H, Massaro JM, et al. Plasma concentration of C-reactive
36:1183–92. protein and risk of ischemic stroke and transient ischemic
29. Prasad KN, Prasad N, Gupta A, Sharma RK, Verma AK, attack: the Framingham study. Stroke 2001; 32:2575–9.
Ayyagari A. Fungal peritonitis in patients on continuous 34. Fine A. Relevance of C-reactive protein levels in peritoneal
ambulatory peritoneal dialysis: a single centre Indian dialysis patients. Kidney Int 2002; 61:615–20.

170