One Year Study of Bacterial and Fungal Nosocomial

Infections among Patients in Pediatric Intensive

Care Unit (PICU) in Alexandria
by Ahmed Ahmed El-Nawawy,a Mohamed Mohamed Abd El-Fattah,a Hala Abd El-Raouf Metwally,b
Shahira Salah El Din Barakat,a and Ihab Abdel Rehim Hassana
a
Departments of Pediatrics and bMicrobiology, Faculty of Medicine, University of Alexandria, Egypt

Summary
A 1-year prospective and observational study included all admissions (n^216) until 48 h after discharge
to Alexandria PICU between first of May 2003 and end of April 2004. Cultures for bacteria and fungi
and antibiotic sensitivity tests (19 antibiotic using Bauer-Kirby disc diffusion method) were obtained
(blood, stool, urine and cerebrospinal fluid, if needed) and repeated on suspicion of NIs. All cannulae,

Downloaded from tropej.oxfordjournals.org by guest on January 5, 2011

endotracheal tube (ET) aspirates and tips, nasogastric tubes and different catheters were cultured.
All PICU health care workers (HCWs) were subjected to throat and under-finger nails cultures as
well as inanimate objects, both on bimonthly basis. The referral place (ward or emergency), PRISM III
score, length of stay (LOS) and fate were recorded. Amongst those patients whose age ranged from
1 to 23 months, 23 per cent had NIs with infection rates of 40/1000 days. Significantly high rates
of mortality, LOS and PRISM III score were encountered among patients with NIs (52 per cent
vs 30 per cent; 9.4[4.8 vs 5.4[2.2 days; 14.4[7 vs 11.8[6 respectively). The descending
order of frequency of NIs was blood stream infection (BSI) (47 per cent), urinary tract infection (UTI)
(28 per cent), ventilator-associated pneumonia (VAP) (16 per cent) and meningitis (9 per cent). Gr–ve
bacilli accounted for 76.7 per cent; Grþve cocci 13.3 per cent (with satisfactory sensitivity to cefepime,
imipenem and meropenem) and Candida albicans 10 per cent of all NIs. The rate of NIs/1000 device
days were: 18.7 per cent for BSI, 10.9 per cent for VAP and 25.5 per cent for UTI. Vulnerable age
groups were `6 m for VAP and _6 m for meningitis. Multiple logistic regression analysis identified
LOS, PRISM III score and referral from wards a predictors of NI acquisition (odd ratio and
95 per cent confidence interval: 1.537, 1.423–1.659; 1.073, 1.041–1.105 and 0.269, 0.178–0.406
respectively). Bimonthly cultures for HCWs isolated coagulase–ve Staphylococci, while inanimate
objects isolated diphtheroids and Candida albicans. Conclusion: NIs rate was high (23 per cent)
mainly due to severity of condition on admission as shown by high PRISM III score value, the high
PRISM III score, LOS and referral from wards were predictors of acquisition of NIs and there is a
high incidence of Candida albicans infection (10 per cent of NIs).

Introduction almost certainly a direct consequence of severity of

Nosocomial infections (NIs) remain a significant
problem in the delivery of pediatric intensive care.
They affect the resource needs of hospitalized
patients1 with additional length of stay.2 Nosocomial
infections increase the mortality and morbidity of
affected individuals and expose hospital staff to
increased risk of infection.3 Children under 2 years
of age have the highest infection rates, (up to 25 per
cent).4 Although the risk of nosocomial infection is
Correspondence: Dr Shahira Salah El Din Barakat,
130 Taawinatte Smouha, Andalos St., Alexandria, Egypt.
E-mail 5shahirabarakat@yahoo.com4.
illness, the level of invasive monitoring and proce-
dures, the indiscriminate use of antibiotic and the
length of stay before the first infection (41 week)
greatly influence the risk of nosocomial infection.5
Reports of NI rates in the PICU varied widely,
ranging from 3 to 27 per cent.4,6
To prevent and control these emerging NIs, we
need to increase national surveillance, ‘risk adjust’
infection rates so that interhospital comparisons are
valid, develop more noninvasive infection-resistant
devices and work with health care workers on better
implementation of existing control measures.7,8
The aim of this study was to evaluate the percent
and types of NIs, the mortality rate associated with
NIs, their contribution to PICU mortality and the

ß The Author [2005]. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org 185
doi:10.1093/tropej/fmi091 Advance Access Published on 26 September 2005

possible contribution of health care workers in
PICU to NIs.
Subjects and Methods
This one year prospective and observational study
included data collected on patients admitted to
the PICU between 1st of May 2003 and end of
April 2004. The Alexandria University PICU is a
7-bedded unit in a 200-bedded tertiary health care
teaching hospital. All patients in the PICU were
surveyed for nosocomial infections at all body sites
from the day of ICU admission until 48 h after ICU
discharge. Infections were regarded as nosocomial if
they occurred 48 h or longer after admission to the
PICU with no evidence that the infection was present
or incubating at time of admission.9
This study was performed after obtaining an
informed consent from parents and approval of the
study by Alexandria University Ethical Committee.
For each patient the following data were gathered:

Demography, admission diagnosis, Pediatric Risk
of Mortality (PRISM) III score, referral place,
length of stay (LOS) and outcome.

Different cultures were obtained on admission:
blood, urine, stool, and cerebrospinal fluid
(if needed).

Daily clinical and laboratory search for infection
with recording the site, date of infection and
registration of any performed invasive procedure.
Specimens for recultures differed according to
site (s) of NI: blood, urine, stool, CSF, nasopharynx,
cannulae, endotracheal tube (ET) aspirates and tips,
catheters and nasogastric tubes. All PICU health care
workers (HCWs) were subjected to throat and under-
fingernail swabs for cultures on a bimonthly basis
during the study period as well as the inanimate
objects. Identification of the microorganisms were
done following the standard methods.10 All bacterial
isolates were subjected to antibiotic sensitivity
test using Bauer-Kirby disc diffusion method.11 The
following antibiotics were used in the antibiogram:
Amikacin, Ampicillin, Ampicillin-Sulbactam,
Amoxicillin/clavulanate, Aztreonem, Cefepime,
Cefoperazone, Cefotaxime, Cephradine, Ceftazidime,
Chloramphenicol, Erythromycin, Imipenem,
Meropenem, Penicillin, Pipracillin, Pipracillin/
Tazobactam Trimethoprim/Sulphamethoxazole and
Vancomycin.
The EPI INFO statistical program was utilized for
presentation and statistical analysis of the results.
Descriptive statistical measures included: Count,
percentage, arithmetic mean, standard deviation,
minimum and maximum. Statistical tests included:
2, Student t test, and multiple logistic regression.
The significance level selected was p  0.05. Statistical
analysis was performed with SPSS (version 10).
186
A. A. EL-NAWAWY ET AL.
Results
During the study period, 216 patients were admitted
to the PICU. Sixty four PICU acquired NIs were
identified among 23 per cent of admissions (50/216).
Seventy two per cent of NI patients (36/50) had
only one NI and 28 per cent (14/50) had 2 NIs.
Nosocomial infections contributed to deaths of
12 per cent of total admissions (26/216). The medians
of the overall NI patient and patient-day rates were
29.6 infections per 100 patients and 40 infections
per 1000 patient-days respectively. The percentage
of pneumonia and meningitis among NIs (the first
and second encountered diagnoses) was significantly
higher compared to those patients without NIs.
(44 per cent vs 24.7 per cent, p ¼ 0.0433 and 28 per
cent vs 12.7 per cent, p ¼ 0.0431 respectively).
Table 1 shows the personal and clinical character-
istics of the studied patients. No statistical significant
Downloaded from tropej.oxfordjournals.org by guest on January 5, 2011
difference was shown among patients with and
without NIs regarding age and sex. Referral place
was significantly different among patients with
compared to those without NIs (%ward/%casuality
were 56/44 vs 36/64, respectively, p ¼ 0.0123). A
significantly high mortality rate, LOS and PRISM III
score were encountered among patients with com-
pared to those without NIs (52 per cent vs 30.1 per
cent, p ¼ 0.0045; 9.45  4.78 days vs 5.41  2.24 days,
p 5 0.001; and 14.36  6.96 vs 11.78  5.97, p ¼ 0.011
respectively).
Distribution of nosocomial infections by site in a
descending order of frequency was: bloodstream
infection (BSI) 47 per cent, urinary tract infection
(UTI) 28 per cent, ventilator-associated pneumonia
(VAP) 16 per cent and meningitis 9 per cent.
Table 2 shows pathogen distribution among
nosocomial infections. Gr–ve bacilli accounted for
76.7 per cent of the total organisms isolated from
patients with NIs in PICU:
Klebsiella was the most frequent one (46.7 per
cent) followed by Pseudomonas aeruginosa (16.6 per
cent), Proteus (6.7 per cent) and E. coli (6.7 per cent).
Grþve cocci (Staphylococcus aureus) accounted for
13.3 per cent of the total isolates, while Candida
albicans accounted for 10 per cent of them.
The distribution of the pathogens differed accord-
ing to the type of nosocomial infection; Klebsiella
was the most frequent organism isolated from the
blood (60 per cent), lower respiratory tract (LRT)
(60 per cent) and accounted for 20 per cent of
the organisms colonizing the gastrointestinal tract
(GIT). Klebsiella was also the single pathogen
isolated from CSF of patients with nosocomial
meningitis. Pseudomonas aeruginosa was isolated
from the blood (20 per cent), LRT (20 per cent)
and colonizing the GIT (30 per cent). E. coli and
Proteus were isolated from blood (13.3 per cent),
urinary tract (UT) (33.3 per cent) and NGT (20 per
cent). Staphylococcus aureus was the most frequent
Journal of Tropical Pediatrics Vol. 52, No. 3
A. A. EL-NAWAWY ET AL.

Table 1
Personal and clinical characteristics of the study patients on admission to PICU

Without NIs With NIs Total Test

n ¼ 166 n ¼ 50 n ¼ 216

Age (months)
Min–Max 1–23 2–18 1–23 t ¼ 0.31
Mean  SD 16.17  29.52 14.78  20.01 15.85  27.58 p ¼ 0.756
No % No % No %
Gender
Male 90 54.2 34 68 124 57.4 2 ¼ 2.986
Female 76 45.8 16 32 92 42.6 p ¼ 0.0840
Referral
Ward 60 36.1 28 56 88 40.7 2 ¼ 6.275
Casuality 106 63.9 22 44 128 59.3 p ¼ 0.0123*
Fate
Discharged 116 69.9 24 48 140 64.8 2 ¼ 8.066
Deceased 50 30.1 26 52 76 35.2 p ¼ 0.0045*

Downloaded from tropej.oxfordjournals.org by guest on January 5, 2011

Length of stay (LOS) (days)
Min–Max 3–14.0 3–21 3–21 t ¼ 8.36
Mean  SD 5.41  2.24 9.45  4.78 6.35  3.47 p 5 0.001*
PRISM III score
Min–Max 5–21 7–32 5–32 t ¼ 2.56
Mean  SD 11.78  5.97 14.36  6.96 12.39  6.29 p ¼ 0.011*

NIs: Nosocomial infections.

*Significant, p50.05.

Table 2
Percent distribution of organisms isolated from different sites among patients with PICU-acquired NIs

Organisms Site of isolation

All sites Blood* Urine Resp. tract** CNS# GIT##

(n ¼ 60) (n ¼ 30) (n ¼ 18) (n ¼ 10) (n ¼ 2) (n ¼ 20)

Grþve organisms Staphylococcus aureus 13.3% 6.7% 22.22% 20% 0% 30%

Grve organisms Klebsiella 46.7% 60% 11.11% 60% 100% 20%
Pseudomonas aeruginosa 16.6% 20% 0% 20% 0% 30%
E. coli 6.7% 6.7% 22.23% 0% 0% 0%
Proteus 6.7% 6.6% 11.11% 0% 0% 20%
Fungi Candida 10% 0% 33.33% 0% 0% 0%

*Isolated from blood and/or intravenous devices.

**Isolated from endotracheal tube aspirates and/or tips.
#Isolated from CSF. (other additional 4 cases presented clinically but showed negative CSF culture).
##Isolated from nasogastric tubes (colonization and not included among the NIs.
NIs: Nosocomial infections.
CNS: Central nervous system.
GIT: Gatsrointestinal tract.

colonizing pathogen of the GIT (30 per cent), and the
second most common isolate from the UT (22.2 per
cent), and was also isolated from the blood (6.7 per
cent), and LRT (20 per cent). Candida albicans was
the most important nosocomial pathogen isolated
from UT (33.3 per cent). The rate of nosocomial
infections was 18.7/1000 device days for BSI, 10.9
for VAP and 25.5 for UTIs. Antibiogram sensitivity
showed satisfactory sensitivity of both Gr–ve bacilli
and Grþve cocci to Cefepime and Carbapenems:
imipenem and meropenem.
Table 3 shows the frequency of NIs in different
sites in relation to risk factors. All cases of VAP
(100 per cent) occurred in infants 6 m of age and
all cases of nosocomial meningitis occurred among
infants 56 m. The difference was statistically sig-
nificant compared to thoses patients without NIs
( p ¼ 0.0085 and 0.0263 respectively). On the other

Journal of Tropical Pediatrics Vol. 52, No. 3 187

188

Table 3
Frequency of NIs in different sites in relation to risk factors

Blood Urinary Tract Respiratory tract CSF

Without NIs With NIs Without NIs With NIs Without NI With NIs Without NIs With NIs
(n ¼ 186) (n ¼ 30) (n ¼ 198) (n ¼ 18) (n ¼ 206) (n ¼ 10) (n ¼ 210) (n ¼ 6)

Age (months)
56 m 104 (55.9%) 16 (53.3%) 110 (55.6%) 10 (55.6%) 120 (58.3%) 0 (0.0%) 114 (54.3%) 6 (100.0%)
6 m 82 (44.1%) 14 (46.7%) 88 (44.4%) 8 (44.4%) 86 (41.7%) 10 (100.0%) 96 (45.7%) 0 (0.0%)
2 (p) 0.069 (0.7918) 0.0 (1.000) 6.93* (0.0085) 4.937* (0.0263)
LOS (days)
2–54 70 (37.6%) 2 (6.7%) 70 (35.4%) 2 (11.1%) 72 (34.9%) 0 (0.0%) 72 (34.3%) 0 (0.0%)
4–57 80 (43.0%) 20 (43.0%) 96 (48.5%) 4 (22.2%) 92 (44.7%) 8 (80.0%) 100 (47.6%) 0 (0.0%)
7 36 (19.4%) 8 (26.6%) 32 (16.1%) 12 (66.7%) 42 (20.4%) 2 (20.0%) 38 (18.1%) 6 (100.0%)
2 (p) 31.43* (50.001) 26.02* (50.001) 19.47* (50.001) 24.12* (50.001)
PRISM score
Journal of Tropical Pediatrics

510 78 (41.9%) 80 (40.4%) 80 (40.4%) 10 (55.6%) 88 (41.9%) 2 (20.0%) 90 (42.9%) 0 (0.0%)

10 108 (58.1%) 118 (59.6%) 118 (59.6%) 8 (44.4%) 118 (57.3%) 8 (80.0%) 120 (57.1%) 6 (100.0%)
2 (p) 0.039 (0.8418) 0.112 (0.7384) 2.025 (0.1547) 4.408* (0.0358)

CSF: Cerebrospinal fluid.

LOS: Length of stay.
NIs: Nosocomial infections.
*Significant, p 5 0.05.

A. A. EL-NAWAWY ET AL.
Vol. 52, No. 3

Downloaded from tropej.oxfordjournals.org by guest on January 5, 2011

A. A. EL-NAWAWY ET AL.
hand, BSIs and UTIs occurred in both infants
aged 5 and  6 m with no significant difference.
The higher incidence of nosocomial BSI and
VAP occurred in patients hospitalized at ICU for
4–57 days. The difference was statistically signifi-
cant compared to those without NIs. ( p50.001 and
50.001 respectively). Both UTIs and CNS infections
were mainly recorded among patients hospitalized
at ICU for 7 days. The difference was statistically
significant compared to thoses without NIs.
( p50.001 and 50.001 respectively). There was no
significant statistical difference between patients
with and without NIs regarding the frequency of
infections in different sites and the PRISM III
score on admission except cases with nosocomial
meningitis as all cases of nosocomial meningitis
occurred in those with PRISM III score 10 on
admission ( p ¼ 0.0358).
Multiple logistic regression analysis, including all
the previous data, identified predictors of acquisition
of NIs in PICU to be LOS, PRISM III score
and referral place from the ward (odds ratio (OR)
and 95 per cent confidence interval (CI) 1.537,
1.423–1.659; 1.073, 1.041–1.105 and 0.269,
0.178–0.406 respectively.
Bimonthly results of cultures from staff members,
nurses, and inanimate objects revealed that
coagulase–ve staphylococci was the most frequently
isolated organism from the under-fingernail (80–93.3
per cent) and throat (53–100 per cent) of staff
members and nurses, while Staphylococcus aureus
patients in ICU allows admission of large number
of patients with different diseases and organisms
within the same number of patients’ days; hence
the increased possibility of NIs because of cross-
contamination. This was also observed in a recent
study, where NI patient-days was shown to be
36/1000.16
Nosocomial infections contribute largely to the
mortality of patients in PICU.3 Patients with NIs
in the present study showed a significantly higher
mortality rate compared to those without NIs (52 per
cent vs 30.1 per cent). Studies from developed
countries reported a lower mortality rate. A
European multicenter prospective study showed a
10 per cent mortality rate among patients with NI in
the PICU,12 while another study in the USA showed
a mortality rate of 23.7 per cent.17 The high mortality
rate in the present study compared to others could
Downloaded from tropej.oxfordjournals.org by guest on January 5, 2011
be explained by the severity of underlying illness
as shown by the high mean PRISM III score (14.4)
on admission.
PICU- acquired NIs, in the present study,
increased the average hospital stay by about 6 days.
The impact of NIs on the LOS was clearly shown
in similar studies from developed countries.12,17 On
the other hand, the LOS could be a predisposing
factor for NI: the longer the higher possibility
of acquiring NIs.3 Several studies showed that the
majority of patients develop NI after the second week
of hospitalization.4,17,18 The average duration of
hospitalization of patients with NIs, in the present
and -hemolytic streptococci were isolated less
frequently or not at all at certain months (16.7–0
per cent). Candida albicans was isolated from the
inanimate objects in very limited cultures (5–10 per
cent), while diphtheroids were isolated more
frequently (5–60 per cent).
Discussion
The detection of NIs in 23 per cent of patients
(50/216) admitted to PICU in a large teaching
pediatric hospital indicates that such infections
constitute a significant cause of illness and expense.
In this study, NI patient rate was estimated to be
29.6 per cent and NI patient-days rate was 40/1000.
Recent PICU reports recorded rates of NIs ranging
from 23–45 per cent.12,13 A recent national point
prevalence study in the USA reported, however, a
lower incidence of NI (11.9 per cent).14
Attributing the infection rate to the number of
admission days helps to control for variation in
the average length of patient ICU stay. The present
study reported a high density of incidence (40/1000
patient-days); another study reported a much lower
density of incidence (6.5/1000 patient-days).15 This
means that the denominator in the present study
(LOS) is small and so the density of incidence
increased. In other words, the rapid turnover of
Journal of Tropical Pediatrics Vol. 52, No. 3
study, was much shorter than in other studies (9.5 vs
26, 21 and 15.3 days),4,18,19 which theoretically makes
our patients less susceptible to NIs. However, the
mortality rate associated with NIs, in the present
study, was much higher than those in such studies
because of the severity of their underlying illness. In
the present study, ward referral to PICU constituted
an important risk factor for developing NIs, and
consequently a higher mortality. This was demon-
strated in another study.19
The most frequent sites of infections, in the present
study, were BSI (47 per cent of all NIs), followed by
UTI (28 per cent), VAP (16 per cent) and meningitis
(9 per cent). Several studies also rank BSI as the
most common PICU–acquired NI, however at a
lower incidence (41.3 per cent and 22 per cent).14,20
In the present study, VAP ranked the third while
came in the second order among PICU-acquired
NIs in another study.14 This indicates a proper
implementation of rules related to tube change
and ventilator use in the studied subjects. On the
contrary, UTIs comprised 28 per cent of NIs in the
present study but reached 33 per cent in another
study,21 which means that more attention to practical
guidelines should be implemented to reduce the
incidence of UT nosocomial infections.
Attributing the infection rates to the device
utilization days will allow better intrahospital and
189

interhospital comparisons of ICU infection rates.6
In the present study, nosocomial BSI was 18.7/1000
days of central venous catheterization, VAP was
10.9/1000 days of ventilator use and nosocomial
UTI was 25.5/1000 days of urinary catheterization.
Average rates reported in multicentric report in the
US are 7.1–8.5 for bacteremia, 3.7–6 for respiratory
infections and 4.8–5.4 for urinary infection.14
The most frequent pathogens isolated from
patients with NIs in the PICU were Gr–ve bacilli
(76.7 per cent), followed by Grþve cocci (13.3 per
cent) and fungi (10 per cent). Similar results were
reported by a recent multicenter European study.12
On the contrary, other studies had shown that Grþve
pathogens are more common than Gr–ve ones.3,4,14
Klebsiella and Pseudomonas aeruginosa were the most
frequent pathogens isolated from patients acquiring
VAP in the present study. Candida was the pre-
dominant pathogen isolated from UT followed by
Staphylococcus aureus. On the contrary, Klebsiella
and E. coli were the predominant pathogens recov-
ered from UT in another study while yeast was
infrequently isolated.22
Certain patients were identified to be at higher
risk for NIs than others. Patients with admission
diagnosis of pneumonia and meningitis were more
prone to NI than patients with other diagnoses, in the
present study, and confirmed by another report.4 Age
was also a risk factor for NI; our study and others
showed that children under the age of two years were
at greatest risk of developing NIs.6 Exposure to
lifesaving invasive procedures (intravascular devices,
mechanical ventilation, and urinary catheterization)
was closely associated with development of NIs
as was shown in the present study and confirmed
by others.6,14,18,23,24 Problems arise because these
devices are inserted into the most critically ill patients
at emergency rooms, often under less than ideal
aseptic conditions. A French study showed that the
use of non-invasive ventilation reduced the incidence
of VAP and the risk of other sites NI.25
In the present study, the presence of nasogastric
tube (NGT) is considered a risk factor for develop-
ment of NIs; 56 per cent of patients with NG tube
feeding developed NI in the form of pneumonia
compared to 10.2 per cent of those without NGT
feeding. A Spanish study revealed a strong associa-
tion between NGT feeding and the risk of develop-
ment of multiresistant Klebsiella pneumonia.6
Therefore, rapid removal of nasogastric tubes is
among the guidelines for preventing of nosocomial
pneumonia.
Resistance to antimicrobial agents, an evolving
process in PICUs, is driven by the selection pressure
of heavy antibiotic use in the ICU.26 Problems of
pathogens presently include multiple resistant Gr–ve
bacilli, methicillin-resistant Staphylococcus aureus
(MRSA), and the recently emerged vancomycin-
resistant enterococci (VRE).27,28 The present study
190
A. A. EL-NAWAWY ET AL.
stressed on the marked resistance of the isolated
Gr–ve bacilli to the commonly used antibiotics.
However, they showed statisfactory susceptibility
to cefepime and carbapenems: imipenem and mero-
penem. Several studies confirmed our observation
of the multi- resistant organisms recovered from
patients with NIs.29,30 In agreement with our results,
cefepime and carbapenem were proved to be effective
in the treatment of NIs in other ICUs.31,32 Pediatric
Intensive Care Unit HCWs need to become familiar
with the microorganisms and sensitivity pattern to
antibiotics within their institution by regular review
of surveillance data.
As contacts are the most important and frequent
means of transimission of nosocomial pathogens and
subsequent infections in hospitals,3 the bimonthly
cultures from HCWs and inanimate objects showed
recovery of diphtheroids (5–60 per cent) and coagu-
Downloaded from tropej.oxfordjournals.org by guest on January 5, 2011
lase–ve staphylococci (10 per cent). Cultures taken
from the under fingernail and throat of HCWs
revealed high rate of recovery of coagulase–ve
staphylococci (80–100 per cent) and Staphylococcus
aureus (33.3–20 per cent). The main reservoir of these
organisms is generally colonized children of ICU
from whom others acquire the organism via hospital
personnel.3
In spite of the high rate of recovery of these
relatively avirulent organisms from under fingernail
and throat of our HCWs, these organisms were
infrequently isolated from blood or cannula culture
of patients with BSI. This implies the strict precau-
tions taken by the HCWs to prevent transmission
of these organisms to ICU patients. Hand hygiene
has been singled out as the most important measure
in preventing hospital-acquired infection. Improved
compliance with handwashing was associated with a
significant decrease in overall rates of nosocomial
infection.33
In conclusion, this type of study is necessary to be
performed periodically to evaluate preventive mea-
sures and design control guidelines, the implementa-
tion of which could result in a reduction of NIs and
of the morbidity, mortality and fiscal costs associated
with infections in a PICU.
References
1. Solnim AD, Kurtines HC, Sprague BM, Singh N.
The cost associated with nosocomial bloodstream
infections in the pediatric intensive care unit. Pediatr
Crit Care Med 2001; 2(2): 170–4.
2. Mylatte JM, Graham R, Kahler L, Young BL,
Goodnough S. Impact of nosocomial infection on
length of stay and functional improvement among
patients admitted to an acute rehabilitation unit. Infect
Control Hosp Epidemiol 2001; 22(2): 83–7.
3. Kaoutar B, Joly C, L’Heriteau F, et al. Nosocomial
infections and hospital mortality: a multicentre epide-
miology study. J Hosp Infect 2004; 58(4): 268–75.
Journal of Tropical Pediatrics Vol. 52, No. 3
A. A. EL-NAWAWY ET AL.
4. Gray J, Gossain S, Mrcpath M, et al. Three-year survey
of bacteremia and fungemia in a pediatric intensive
care unit. Pediatr Infect Dis J 2001; 20: 416–21.
5. Singh-Naz N, Sprague BM, Patel KM, Pollack MM.
Risk assessment and standardized nosocomial infection
rate in critically ill children. Crit Care Med 2000; 28(6):
2069–75.
6. Rowin ME, Patel VV, Christenson JC. Pediatric ICU
nosocomial infections: epidemiology, sources and
solutions. Crit Care Clin 2003; 19(3): 473–87.
7. Calil R, Marba ST, Von Nowakonski A, Tresoldi AT.
Reduction in colonization and nosocomial infection by
multiresistant bacteria in a neonatal unit after institu-
tion of educational measures and restriction in the use
of cephalosporins. Am J Infect Control 2001; 29(3):
133–8.
8. Alexis M, McGann KA. Steps to reduce nosocomial
infections in children. Infect Med 2002: 1–9.
9. Emori TG, Culver DH, Horan TC. National Nosocomial
Infections Surveillancce system. Description of surveil-
lance methods. Am J Infect Control 1991; 19: 19–35.
10. Duerden BI, Towneer KJ, Magee JT. Isolation,
description and identification of bacteria. In: Collier L,
Balows A, Sussman M (eds). Topley and Wilson’s
Microbiology and Microbial Infections. 9th edn.
Oxford University Press, London, Sydney, Auckland.
1998; pp. 5–84.
11. Bauer AW, Kirby WM, Sherries JG, Turck M.
Antibiotic susceptibility testing by a standard single
disc method. Am J Clin Path 1996; 45: 493–6.
12. Raymond J, Aujard Y. Nosocomial infections in
pediatric patients: A European multicenter study
group. Infect Control Hosp Epidemiol 2002; 21(4):
260–3.
13. Citak A, Korabocuoglu M, Uesal R, Ugur-Bayals S,
Uzel N. Bacterial nosocomial infections in mechanically
ventilated children. Turk J Pediatr 2000; 42(1): 39–42.
14. Grohskopf LA, Sinkowitz-Cochran RL, Levine GL,
Siegel JD, Jarvis WR. Pediatric Prevention net work.
A national point prevalence survey of pediatric
intensive care unit-acquired infections in the United
States. J Pediatr 2002; 140(4): 391–3.
15. Fayon MJ, Tucci M, Lacroxi J, et al. Nosocomial
pneumonia and tracheitis in a pediatric intensive care
unit. Am J Respir Crit Care Med 1997; 155: 162.
16. Stover BH, Shulman ST, Bratcher DF, Brady MT,
Levine GL, Jarvis WR. Pediatric Prevention Network.
Nosocomial infection rates in US children’s hospitals,
neonatal and pediatric intensive care units. Am J Infect
Control 2001; 29(3): 152–7.
17. Patel JC, Mollitt DL, Tapas JJ. Infections complica-
tions in critically injured children. J Pediatr Surg 2000;
35(8): 1174–8.
18. Urea M, Pons M, Serra M, Lotore C, Palmeque A.
Prospective incidence study of nosocomial infections
in a pediatric intensive care unit. Pediatr Infect Dis J
2003; 22(6): 490–4.
Journal of Tropical Pediatrics Vol. 52, No. 3
19. Goldhill DR, McNarry AF, Hadjanastassiou VG,
Tekkis PP. The longer patients are in hospital before
intensive care admission the higher their mortality.
Intensive Care Med 2004; 30(10): 1908–13.
20. Richard MJ, Edwards JR, Culver DH, Ganyes RP.
Nosocomial infections in pediatric intensive care
units in the USA. National NI Surveillance System.
Pediatrics 1999; 103: e39.
21. Al-Asmary SM, Al-Helali NS, Abdel-Fattah MM,
Al-Jabban TM, Al-Ban AM. Nosocomial urinary
tract infection. Risk factors, rates and trend. Saudi
Med J 2004; 25(7): 895–900.
22. Matlow AG, Wray RD, Cox PN. Nosocomial urinary
tract infections in children in a pediatric intensive
care unit: a follow-up after 10 years. Pediatr Crit Care
Med 2003; 4(1): 74–7.
23. Leone M, Albanese J, Garnier F, et al. Risk factors of
nosocomial catheter-associated urinary tract infection
in a polyvalent intensive care unit. Intensive care Med
2003; 29: 1077–80.
Downloaded from tropej.oxfordjournals.org by guest on January 5, 2011
24. Champs AP, Starling CE. Prospective surveillance
applying the national nosocomial infection surveillance
methods in a Brazilian pediatric hospital. Am J Infect
Control 2002; 30(1): 1–7.
25. Nourdine K, Combes P, Corton MJ, Connamela A,
Ducreux JC. Does non-invasive ventilation reduce the
ICU nosocomial infection risk? A prospective clinical
survey. Intensive Care Med 1999; 25(6): 567–73.
26. Gordon A, Isaacs D. Late-onset infection and the role
of antibiotic prescribing policies. Curr Opin Infect Dis
2004; 17(3): 231–6.
27. Hunt CP. The emergence of enterococci as a cause
of nosocomial infection. Br J Biomed Sci 1998; 55(2):
149–56.
28. Asensio A, Oliver A, Gonzalez-Diego P, et al. Outbreak
of a multiresistant Klebsiella pneumoniae strain in
an intensive care unit: Antibiotic use as risk factor
for colonization and infection. Clin Infect Dis 2000; 30:
55–60.
29. Patzer J, Turner P. Susceptibility pattern of gram–ve
bacteria from a Polish intensive care unit 1999–2000.
Int J Antimicrob Agent 2002; 19(5): 431–4.
30. Anonymous. Intensive care antimicrobial resistance
epidemiology (ICARE) surveillance report, data sum-
mary from January 1996 through December 1997. Am J
Infect Control 1999; 27: 279–84.
31. Toltzis P, Dul M, Ann M, et al. Cefepime use in a
pediatric intensive care unit reduces colonization with
resistant bacteria. Pediatr Infect Dis J 2002; 22: 109–14.
32. Martinez Lacasa J, Garau J. The role of carbapenem in
the treatment of nosocomial infection. Enferm Infect
Microbiol Clin 1997; 15(Suppl 1): 78–85.
33. Lam BC, Lee J, Lau YL. Hand hygiene practices in a
neonatal intensive care unit: a multimodal intervention
and impact on nosocomial infection. Pediatrics 2004;
114(5): 565–71.
191