Med. J. Cairo Univ., Vol. 82, No.
2, December: 123-129, 2014
www.medicaljournalofcairouniversity.net
Value of Endotracheal Aspirate Cultures Versus Blood Cultures in
Predicting Sepsis in Ventilated Preterm and Full-Term Neonates
LAILA H. MOHAMED, M.D.*; OLA A. EL-SISY, M.D.**; NERMIN R. MOHAMED, M.D.*;
NAGY A. EL-HUSSIENY, M.D.*** and DALIA B. MOHAMED, M.D.*
The Departments of Pediatrics*,*** & Clinical Pathology**, Faculty of Medicine, Cairo University*,** and
Ahmed Maher Teaching Hospital***
Abstract
Sepsis is a common complication in the NICU. The
incidence of neonatal sepsis is 1-5/1000 livebirth and its
mortality rate is 5-20%. Mechanical ventilation plays an
important role because artificial airways bypass the body’s
defense against inhaled pathogens and offer new routes for
non airborne pathogens.
The goal of our work is to clarify the valuable comparison
of endotracheal aspirate cultures versus blood cultures in
order to predict sepsis in ventilated preterm and full term
neonates. Sixty (60) full term ( >!37 wks) & preterm ( !937 wks)
neonates with respiratory distress and mechanically ventilated
in NICU with clinical evidence of sepsis were selected for
our study. Our selected cases were subjected to clinical
evaluation and to the Hematological Scoring System of Sepsis.
The golden pearl in our study is the blood culture for any
neonatal infant suggested clinically and by laboratory criteria
to be neonatal sepsis.
We found that Endotracheal Aspirate (ETA) culture is not
considered a gold standard as sensitivity and specificity in
predicting sepsis in ventilated newborns is low in comparison
to blood culture results, the rate of culture positivity increased
as the birth weight decreased, gestation week got smaller and
the duration of intubation prolonged.
Because of longer duration of mechanical ventilation,
longer stay in the NICU, increased use of antibiotics, higher
costs for healthcare, and most importantly, increased mortality,
the prevention of VAP is the main priority. In spite of the
great advances in the pathogenesis of VAP, intensivists still
struggle with the prevention strategy.
Key Words: NICU (Neonatal Intensive Care Unit ) – Hema-
tological scoring system of sepsis – VAP (Venti-
lator Associated Pneumonia) – Blood culture.
Introduction
SEPSIS is a common complication in the Neonatal
Intensive Care Unit (NICU). It is the most common
of the smallest and most premature infants, in
Correspondence to: Dr. Nagy A. El-Hussieny, The Department
of Pediatrics, Ahmed Maher Teaching Hospital
123
whom the clinical presentation can be subtle and
nonspecific. The incidence of neonatal sepsis is 1-
5 per 1,000 live births, and its mortality rate is
5%-20%. Disseminated intravascular coagulation
and thrombocytopenia are well-known complica-
tions of sepsis. Hospital-acquired infections (noso-
comial infections) are the most common compli-
cations encountered in the neonatal intensive care
unit [1] .
Nosocomial infections are infections that are
a result of treatment in a hospital or a healthcare
service unit. Infections are considered nosocomial
if they first appear 48 hours or more after hospital
admission or within 30 days after discharge [2] .
Healthcare-associated infections are of impor-
tant wide-ranging concern in the medical field.
They can be localized or systemic, can involve
any system of the body, be associated with medical
devices or blood product transfusions. There are
3 major sites of healthcare-associated infections
(i.e., bloodstream infection, pneumonia, and urinary
tract infection ) with ventilator associated pneumo-
nia (VAP) being the most common form of venti-
lator associated infections (VAI) [3] .
The mortality rate associated with VAP ranges
from 24-50% and can escalate up to 76% based on
specific settings and host-pathogen relationship.
The development of VAP is also associated
with greater hospital mortality rates and longer
lengths of stay in Intensive Care Unit (ICU) and
hospitals. Microorganisms responsible for VAP
may differ according to the population of patients
in the ICU, the duration of hospital and ICU stays,
and the specific diagnostic method(s) used [4] .
Neonatal septicemia is a potentially life-
threatening event. As part of a group of investiga-
124 Value of Endotracheal Aspirate Cultures Versus Blood Cultures
tions to rule out septicemia in newborn infants and
in patients of neonatal intensive care nurseries,
blood culture collection is a common event [5] .
There are few reports about the value of serial
endotracheal tube aspirate culture in predicting
sepsis in ventilated newborns. Researchers reported
50% of sepsis occurrence among infants who had
respiratory tract colonization. They also found that
none of the neonates who had no colonization, had
sepsis. ETA culture is less invasive and results
appear earlier than blood culture [1] .
Many infants are screened for infection and
treated with antibiotics. While others, blood cultures
and other investigations such as a full blood count
were done. The clinical practice is that infants who
are <48h of age at time of blood culture collection
(early septicemia), and who remain clinically well,
cease antibiotic treatment 36h after blood culture
collection. Many neonatal units will wait for blood
cultures to incubate for at least 48h before deciding
to stop antibiotics [6] .
Aim of the work:
This study was conducted to define the pattern
of respiratory tract colonisation in neonates intu-
bated and ventilated for longer than seven days
and to assess the value of serial cultures of endot-
racheal tube aspirates (ETA) in predicting patho-
gens of sepsis in comparison to the results of blood
culture.
Patients and Methods
This prospective study was conducted on 60
full term ( >_ 37 wks) and preterm ( < 37 wks) neonates
who were admitted during their first day of life in
the Neonatal Intensive Care Unit (NICU) of Cairo
university children hospital during the period from
January 2010 to December 2010. All the babies
were presented with respiratory distress and applied
on mechanical ventilation according to NICU
protocol.
Inclusion criteria:
- All ventilated preterm or full term newborns with
prolonged stay on ventilator (more than 7 days).
- All ventilated preterm or full term newborns with
clinical evidence of sepsis e.g.: Poor Moro, poor
suckling reflex, poor capillary refilling, abnormal
tone, convulsions etc , confirmed by labo-
ratory evidence of sepsis according to the Hema-
tological Scoring System for diagnosis of neonatal
sepsis [7] .
Exclusion criteria:
- Full term or preterm newborns on nasal CPAP.
- Full term or preterm newborns with a stay of less
than 7 days on mechanical ventilation.
- Full term or preterm newborns with distress of
non respiratory causes e.g. congenital heart dis-
eases, CNS anomalies, etc
All the neonates were subjected to the following:
• Full maternal history and antenatal care with
stress on any prenatal hazards as pre eclampsia,
PROM, antepartum hemorrhage or intrapartum
fever.
• Detailed perinatal history including gestational
age, mode of delivery and method and mode of
resuscitation.
• Complete general and systemic examination to
confirm inclusion and exclusion criteria.
Score interpretation: Minimum 0, maximum
7, <2 sepsis is very unlikely, >3 sepsis is likely,
the higher the score the greater the likeless of
sepsis.
• Investigations conducted for every case including:
- Laboratory investigations in the form of automated
complete blood count (CBC) with the differen-
tional count done on leishmanina Giemsa stained
peripheral blood film and C-Reactive Protein
(CRP) done using latex agglutination test.
- Imaging studies as Chest X-Ray to confirm the
respiratory cause of distress.
- Cultures:
a- Endotracheal culture (routinely done for in-
cluded cases at day 3 and after day 7 of ven-
tilation) using sterile endotracheal suction
catheter during the technique of sampling
from endotracheal fluid.
3 rd
b- Blood culture (performed in the day of
life) using BACTEC Peds blood culture bottles
and the BACTEC-9050 instrument.
Statistical methods:
Data manegment and analysis were performed
using the Statistical Package for the Social Sciences
(SPSS, version 15). The graphs were done using
Microsoft Word. Numerical data were summarized
using means & standard deviations. Categorial
data were summarized as percentages. Comparison
between groups with respect to numeric variables
was done using the X 2 test and independent t-test,
respectively. The Fisher’s exact test was used when
applicable.
Laila H. Mohamed, et al. 125

The level p<0.05 was considered the cut-off Table (5): Blood culture for studied cases.
value for significance.
N %

Results Blood culture +ve 44 37.3

–ve 16 26.7
All results are illustrated in the following Tables
and Figures.
Table (6): Prevalent organisms detected by blood culture.

Table (1): Hematological scoring system of sepsis (rodwell %

scoring system).
Klebsiella 43.75
Items Abnormalities Score Strept. viridians 18.75
Staph. aureus 12.5
Total white blood <5000 per micro L 1 Blood culture
Staph. epidermidis 6.25
cell count Or >25000 per micro L at Candida 6.25
birth Staph. coag. 6.25
Or >30.000 per micro L at Citrobacter 6.25
12-24h.
Or >21.000 per micro L at
D2 onward Table (7): Early endotracheal tube culture for studied cases.

Total PMNL count No mature PMNL seen on 1 N %

blood film –ve
Early endotracheal culture 34 56.7
Increased or decreased +ve 26 43.3
Immature PMNL Increased 1
Immature/total Increased 1 Table (8): Prevalent organisms detected by early endotracheal
PMNL ratio (I/T) culture.
Immature/mature >0.3 1 N %
PMNL ratio (I/M)
Strept. viridians 1 3.8
Platelets counts <150.000 per micro L 1 Klebsiella 18 69.23
Early endotracheal Staph. coag. 2 7.69
Degenerative >3+ for vacuolization, toxic 1 culture
changes in PMNL granulation or Dohle Pseudomonas 2 7.69
bodies Acinetobacter 3 11.53

Table (9): Late endotracheal culture of studied cases.

Table ( 2): Demographic data of studied cases.
N %
Minimum Maximum Mean SD
Late endotracheal culture –ve 44 73.3
Gestational age 30.00wk. 35.00 33.67 ± 1.36 +ve 16 26.7
BW on admission 1.00 Kg 2.40 1.73 ±0.36
Table (10): Prevalent organisms detected by late endotracheal
culture.
Table (3): Signs of sepsis of studied cases.
N %
Clinical signs of sepsis N %
Staph aureus 1 6.25
Poor Reflexes 27 45 Srept. viridians 2 12.5
Poor capillary filling 12 20 Late endotracheal Klebsiella 10 62.5
culture
Sclerema 15 25 Staph. coag. 2 12.5
Pseudomonas 1 6.25

Table (4): Duration of stay on M.V.

Table (11): Duration of hospital stay of studied cases.
Minimum Maximum Mean SD
Minimum Maximum Mean SD
Duration of stay 7.00 days 69.00 10.45 ±9.14
on MV/days Postnatal age 7.00 103.00 28.37 ± 18.19
126 Value of Endotracheal Aspirate Cultures Versus Blood Cultures

Table (12): Comparison between organisms of early endotra-

Sepsis is very
cheal culture results (in the 3 rd day of life) versus
likely 15%
the results of blood culture in the study.

Blood culture Early ETT

Organism
(+ve) culture (+ve)

Staph. aureus 2 –ve

Srept. viridians 2 1 Sepsis is
Sepsis is very possible 23%
Klebsiella 4 18 unlikely 62%

Candida 1 –ve
Fig. (1): Hematological classification of sepsis for studied
Staph. coagulase 1 2 cases.

Pseudomonas –ve 2

Discussion
Table (13): Sensitivity and specificity of early endotracheal
culture.
Sepsis is a common complication in the Neo-
natal Intensive Care Unit (NICU). It is most com-
Sensitivity Specificity PPV NPV mon in the smallest and most premature infants,
in whom the clinical presentation can be subtle
Early endotracheal 62.5 82.4 62.5 82.4
and nonspecific. The incidence of neonatal sepsis
culture
is 1-5per 1,000 live births, and its mortality rate
is 5%-20%, Disseminated intravascular coagulation
Table (14): Comparison between organisms of late endotracheal and thrombocytopenia are well-known complica-
culture results (in the 7 th of life) versus the results tions of sepsis.
of blood cultures in the study.
Hospital acquired infections (nosocomial infec-
Blood culture Early ETT tions) are the most common complications encoun-
Organism
(+ve) culture (+ve) tered in the neonatal intensive care unit. They
Staph. aureus 2 1 generally manifest 48 hours after hospitalization
or in 48 hours after discharge, especially preterm
Klebsiella 4 10
and low birth weight newborns who are more
Candida 1 –ve vulnerable (20 to 33%) to nosocomial infections.
Staph. coagulase 1 2 Mechanically ventilated babies face a particular
Pseudomonas –ve 1 risk because artificial airways bypass the body’s
defense against inhaled pathogens and offer new
routes for non air-borne pathogens. Intubation
Table (15): Organisms detected only by late ETT. associated lesions of the pharynx and trachea lead
to bacterial colonization by the deterioration of
Organism N % the swallowing reflex and the ciliary functions.
Staph. aureus 2 20.00
Subsequently, these babies may develop pneumonia
and sepsis [1] .
Staph. epedermidis 1 10.00
Srept. viridians 3 30.00 In our study, there were no statistically signif-
icant differences between body weight (BW) &
Klebsiella 3 30.00 incidence of sepsis. Birth weight of cases ranged
Citrobacter 1 10.00 between 1000gm and 2400gm. This is in discor-
dance with [8] who found that the incidence of
Total 10 100.00 sepsis is significantly higher in infants with very
low birth weight (<1000g), at 26per 1000 live
birth, than in infants with birth weight of 1000-
Table (16): Sensitivity and specificity of late endoracheal
culture.
2000g, at 8-9per 1000 live birth.

Sensitivity Specificity PPV NPV In our study, there were no statistically signif-
icant differences between gestational ages (GA)
Late endotracheal 37.5 77.3 37.5 77.3 & incidence of sepsis. Our study agrees with [9]
culture and [10] who reported no statistically significant
Laila H. Mohamed, et al.
difference between sepsis and non septic groups
as regards gestational age.
This is in discordance with [8] who found that
the incidence and severity of sepsis is inversely
related to the gestational age of the infant. Also
our study disagrees with [11] who found that the
premature neonates in particular have a relatively
permeable mucosal surfaces that allow for the
trans-epithelial passage of bacteria and other patho-
gens. Also loss of protective maternal antibodies,
as well as nonspecific alterations in macrophage
phagocytes and clearance of invading pathogens,
impaired T-cell and B-cell responses and altered
production of complement and antibodies. Addi-
tionally, premature infants are at higher risk of
progression to sepsis or severe sepsis and adverse
outcomes [12] .
In our study, the chief clinical presentations
were poor Moro’s reflex and poor suckling as well.
This is in agreement with [13] who described the
same clinical presentations as early signs of sepsis.
On the other hand, our study is in disagreement
with the result of [14] , who reported that chest
manifestations in the form of respiratory distress
are the most common and occurring in up to 90%
of infants with sepsis. Also [15] , found that 90%
of newborns with neonatal sepsis had respiratory
distress in addition to symptoms of gastrointestinal
disturbance and poor perfusion.
In our study, we found 37 cases (62%) with a
Rodwell Hematological Scoring System of sepsis
of <2, denoting absent sepsis, 14 cases (23%) with
a score of 3 denoting that sepsis is likely and 9
cases (15%) with a score of (4-7) denoting that
sepsis is eminent.
In our study, we found a significant correlation
between the duration of stay of neonates in NICU,
prolonged stay on mechanical ventilation & inci-
dence of sepsis. This is in concordance with [16]
who found that ventilator-associated pneumonia
in critically ill patients is associated with longer
ICU stays, and prolonged mechanical ventilation.
Also, this is in concordance with [17] who
concluded that intubation is associated with bacte-
rial colonization of the respiratory tract and, there-
fore, may increase the risk of acquiring an infection.
The infection may prolong the need for mechanical
ventilation and increase the risk of chronic lung
disease. The use of prophylactic antibiotics has
been advocated for all mechanically ventilated
newborns in order to reduce the risk of colonization
and the acquisition of infection.
127
Blood culture is the gold standard method for
isolation of the organisms; blood culture should
be obtained before the initiation of antibiotics [18] .
In this study, blood culture yielded +ve results
in 44 cases (73.3%) of ventilator associated pneu-
monia cases while 16 cases (26.7%) revealed a
culture –ve output. Klebsiella was the most com-
monly detected organism (43.75%), followed by
Strept. Viridians (18.75%), whilest Staph. Aureus
was detected in (12.5%) of cases. Staph. Epider-
midis, Candida, Staph. Coagulase and Citrobacter
were equally detected in (6.25%) of cases each.
This comes in agreement with the study of [19]
in which nearly half of the positive blood cultures
grew Klebsiella pneumoniae. Also, the study done
by [20] revealed that the isolated pathogens included
Klebsiella pneumoniae (47.5%), Pseudomonas
aeruginosa (20%), E.coli (10%), Candida albicans
(10%), Staphylococcus aureus (>7.5%), and En-
terococcus (5%).
Our results are in disagreement with a retro-
spective study done by [21] who studied the prev-
alence of different organisms causing septicemia
and the antibiotic susceptibility pattern. The most
common organism isolated was Staphylococcus
aureus (42.75%) followed by Klebsiella (18.32%),
E.coli (12.21%), Pseudomonas aeruginosa (6.11%);
also Enterobacter spp. Was isolated in (9.23%),
Acinobacter (4.62%), Streptococcal spp. In (7.69%)
and Neisseria gonorrhea in (1.54%).
In our study, 26 cases (43.3%o) showed +ve
early endotracheal tube culture (performed on day
3), while 34 cases (56.7%o) showed –ve early ETT
culture. Among the culture +ve cases, again Kleb-
siella was the most commonly detected organism
(69.23%).
In our study, 16 cases (26.7%) showed +ve late
endotracheal tube culture, while 44 cases (73.3%)
showed –ve late ETT culture. Among the culture
+ve cases, Klebsiella also was the most commonly
detected organism (62.5%) followed by Strept.
Viridians, Staph. Coagluase, each detected in
(12.5%), whilest Staph. Aureus & Psendomans,
each detected in (6.25%) of cases.
In the current study, If blood culture is consid-
ered as gold standard for diagnosis of VAP, accord-
ing to [22] , early endotracheal culture showed
sensitivity of 62.5% and specificity of 82.4% to
blood culture and late endotracheal culture showed
sensitivity of 37.5% and specificity of 77.3% to
blood culture. Therefore early ETT culture showed
higher diagnostic indices than late ETT culture in
128 Value of Endotracheal Aspirate Cultures Versus Blood Cultures
detecting +ve infection (as diagnosed +ve by blood
culture).
This is in concordance with [1] who found that
blood and endotracheal cultures showed the same
organisms only in 17.6% of the patients. There
was no relationship among 86.4% of the patients.
The rate of culture positivity increased as the birth
weight decreased, gestation week got smaller and
the duration of intubation prolonged.
Also [23] , found that the most common bacterial
isolate from ETA of VAP patient was Klebsiella
spp. (32.87%); E.coli and Acinetobacter were the
other two common organisms.
Also, our study is in agreement with [24] who
found that the practice of routine cultures of en-
dotracheal aspirate and cultures obtained from
multiple body sites is an expensive proposition
with low yield. The sensitivity of this test is at best
50% and all studies report a very low positive
predictive value. The specificity of this test is 80%,
hence its role is mainly limited to identifying
infants who are at low risk for sepsis.
This is in discordance with [23] who found that
the endotracheal aspirate of the patients on venti-
lator should be sent routinely for culture sensitivity
and if the patient develops VAP, antibiotics should
be changed as per report.
Conclusion and Recommendations:
• Blood culture is still the gold standard for detec-
tion of organisms in neonates with ventilator
associated pneumonia; Endotracheal tube culture
has a lower sensitivity & specificity.
• Early and appropriate antimicrobial treatment of
patients with VAP significantly improves outcome
of patients.
• Minimize the duration of stay of prematures
especially on mechanical ventilation which sub-
sequently reduces their period of stay in the
NICU and hence reduces the incidence of noso-
comial infections.
• Once mechanical ventilation is indicated, it
should be applied under the direction of a trained
physician who understands the technical aspects
of management of artificial airways, mechanical
ventilators, and humidification systems.
• Universal Standard Precautions of infection
control as described by the Centers for Disease
Control are necessary and should be employed.
• Further studies including large number of patients
are recommended to evaluate the role of endot-
racheal cultures for detection of organisms in
patients with ventilator associated pneumonia,
as tracheal aspirates may be useful track changes
in bacterial flora at neonatal intensive care units.
References
1- BOZAYKUT A., IPEK I.O. and KILIC B.D.: Predicting
neonatal sepsis in ventilated neonates. Indian. J. Pediatr.
Jan., 75 (l): 39-42, 2008.
2- POLLACK and ANDREW: Nosocomial infections, rising
threat of infections unfaced by antibiotics, New York
Times, Feb., 27, 2010.
3- AYESHA MIRZA M.D., CHIEF EDITOR and RUSSELL
W. STEELE: Hospital Acquired Infections, Medscpe
Jornal, 2010.
4- CHANDRAKANTH C., ANUSHREE and VINOD A.:
Incidence of ventilator associated pneumonia, International
Journal of Medical and Clinical Research. ISSN: 0976-
5530 & E-ISSN: 0976-5549, Vol. 1, Issue 2, pp. 11-13,
2010.
5- ANDRESON-BERRY A.L., BELLIG L.L., et al.: Neonatal
Sepsis. WWW.e medicine.Medascape.com/article/978352,
2006.
6- TALLUR S.S., KASTURI A.V., NADGIR S.D. and
KRISHNA B.V.S.: Clinico-bacteriological study of neo-
natal septicaemia in hubli. Indian. J. Pediatr., 67: 169-
174, 2006.
7- RODWELL R.L.: Hematological scoring system. Pediatr.,
112: 761-7, 1988.
8- BELLING L.L.: Neonatal sepsis. J. Pediatr., 155: 120-
126, 2004.
9- MAHIEU L.M., De MUYNCK A.O., DE DOOY J.J., et
al.: Prediction of nosocomial sepsis in neonates by means
of a computer-weighted bedside scoring system (NOSEP
score). Crit. Care Med., 28: 2026-2033, 2000.
10- FERGNANO S., SHARLAND M., KAZEMBE P. and
MWANSAMBO C.: Heath PT. Neonatal sepsis: An inter-
national perspective. Arch. Dis. Child Fetal. Neonatal.,
90: F220-4, 2005.
11- POLITIS I.C.R.: Milk peptides and immune response in
the neonate. Adv. Exp. Med. Biol., 606: 253-69, 2008.
12- STEPHEN B. and THACKER M.D., M.Sc.: Prevention
of perinatal group B Streptococcal Disease November,
19, Vol. 59/No. RR-10, 2010.
13- KLIEGMAN R.M., BEHRMAN R.E., JENSON H.B. and
STANTON B. Eds.: Nelson textbook of pediatrics. Sepsis,
septic shock and systemic inflammatory response syn-
drome. 18 th ed. Philadelphia, Pa.: Saunders Elsevi-
er;133981928-3/0/1608/514.0-B978-1-4160-2450-
7.50178-X- cesec6_4006, 2007.
14- CLOHERTY, ANN R. STARK, ERIC C., JOHN P. and
J. EICHENWALD: Bacterial and fungal infections. Man-
ual of Neonatal Care, Chapter 23: p. 291-295, 2004.
15- KAREN M. PUOPOLO: Epidemiology of neonatal early-
onset sepsis. NeoReviews, 9: e571-e579. DOI: 10.1542/
neo.9-12-e571, 2008.
16- MOHR A.M., SIFRI Z.C., HORNG H.S., et al.: Use of
aerosolized aminoglycosides in the treatment of Gram-
Laila H. Mohamed, et al.
negative ventilator-associated pneumonia. Surg. Infect.
(Larchmt), 8 (3): 349-357, 2007.
17- INGLIS G.D.T., JARDINE L.A. and DAVIES M.W.:
Prophylactic antibiotics to reduce morbidity and mortality
in ventilated newborn infants. Cochrane Database of
Systematic Reviews Issue 3. Art. No.: 14651858.
CD004338. DOI: 10.1002, 2008.
18- SAINI, SOURABHDUTTA, PALLABRAY and ANIL
NARG: Short course versus 7-day course of intravenous
antibiotics for probable neonatal septicemia. Volume: 48,
January, 17, 2011.
19- DZWONEK, AGNIESZKA B., et al.: The role of mannose-
binding lectin in susceptibility to infection in preterm
neonates, Volume 63, Issue 6, pp. 680-685, 2008.
129
20- De BENEDETTI F., AURITI C., D'URBANO L.E., et al.:
Low serum levels of mannose binding lectin are a risk
factor for neonatal sepsis. Pediatr. Res., 61: 325-8, 2007.
21- SHAW M.J.: Ventilator-associated pneumonia. Opin.
Pulm. Med., 11 (3): 236-241, 2005.
22- ROBERTS I. and MARRAY N.: Thrombocytopenia in
the newborn. J. Curr. Opin. Pediatr., 15 (1): 17-23, 2003.
23- SHALINI TRIPATHI, MALIK, AMITA JAIN and NEERA
KOHLI: Study of ventilator associated pneumonia in
neonatal intensive care unit: Characteristics, risk factors
and outcome. January, 5 (l): 12-19, 2010.
24- HARI BHASKAR: Indian. J. Pediatr., 65: 79-84, 1988.