Pediatric Pulmonology

New Parameters for Childhood Ventilator Associated

Pneumonia Diagnosis
_ u
Rana Işg € der, MD,
1 € khan Ceylan, MD,1 Hasan Ag
* Go  ın, MD,1 Gamze Gu
€ lfidan, MD,
2

Yu 2 _
€ ce Ayhan, MD, and Ilker Devrim, MD3
Summary. Purpose: Our aim is to determine whether the presence of soluble triggering receptor
expressed on myeloid cells-1 (s-TREM-1) of bronchoalveolar lavage fluid (BALF), serum
procalcitonin levels (PCT), and Clinical Pulmonary Infection Score (CPIS) have diagnostic value in
children with VAP. Methods: All children followed in pediatric intensive care unit (PICU) who were
mechanically ventilated at least for 48 hr between January 2014 and December 2015
were enrolled into our study. BALF sample was obtained via non-bronchoscopic method from
the children with VAP suspicion (case group) and s-TREM-1 levels were measured. Furthermore
we calculated CPIS and measured serum PCT levels. Same procedures were applied to the
control group who were admitted to PICU without infectious problems and who were not under
antimicrobial therapy. First we compared the case group with the control group and then we
compared the quantitative culture confirmed and non-confirmed VAP cases among themselves.
Results: Case group (n:58) had significant higher PCT and s-TREM-1 levels compared to control
group (n:58). The VAP confirmed cases had higher s-TREM-1, PCT ve CPIS levels compared to
non-confirmed VAP cases. s-TREM-1, PCT ve CPIS variables were found to be independent risk
factors for VAP. The cutoff values for s-TREM-1, CPIS, and PCT, are 281 pg/ml, 6, and 1.9 ng/ml,
respectively. The patients whose s-TREM-1, CPIS, and PCT values above the cutoff levels were
found to have higher cumulative VAP rate. Conclusions: s-TREM-1 of BALF, serum PCT levels,
and CPIS are useful predictors for ventilator-associated pneumonia diagnosis in children. Pediatr
Pulmonol. 2016; 9999:XX–XX. ß 2016 Wiley Periodicals, Inc.

Key words: biomarkers; pediatric critical care; pneumonia; mechanical ventilation.

Funding source: none reported.

1
INTRODUCTION Department of Pediatric Intensive Care Unit, Dr. BehSc et Uz Children’s
Research and Training Hospital, Izmir, Turkey.
Ventilator-associated pneumonia (VAP) should be
2
suspected if the patients who were receiving mechanical Department of Microbiology, Dr. BehSc et Uz Children’s Research and
ventilatory support at least for 48 hr, develop pneumonia.1 Training Hospital, Izmir, Turkey.
In the health care associated infection group, VAP has the 3
Department of Pediatric Infectious Diseases, Dr. BehSc et Uz Children’s
second common place in Pediatric Intensive Care Units Research and Training Hospital, Izmir, Turkey.
(PICUs) or neonatal intensive care units. Ventilator-
associated pneumonia can be seen in 3–10% of children Conflicts of interest: We certify that there is no conflict of interest with any
who were mechanically ventilated.2 VAP is a serious financial and personal relationships with other people or organizations
regarding the our manuscript. None of the authors has any conflict of interest
complication which can be seen in PICUs and it can threaten to disclose.
life and increases the duration of mechanical ventilation
support, length of stay in PICU, and mortality risk.2 Ethical publication statement: We confirm that we have read the Journal’s
In the last 30 years, many definitions of VAP has been position on issues involved in ethical publication and affirm that this report
made but no gold standard was determined yet.2,3 Clinical is consistent with those guidelines.
diagnosis criteria are subjective. Quantitative microbio-
_ uder, MD, Department of Pediatric Intensive
Correspondence to: Rana Işg€
logical culture methods which were used to confirm the Care Unit, Dr. BehSc et Uz Children’s Research and Training Hospital,
VAP diagnose, give results in 48–72 hr. This causes _
Alsancak, Izmir 35210, Turkey. E-mail: [email protected]
treatment delay or unnecessary antibiotic use for
noninfectious processes.4 VAP diagnose and treatment Received 1 March 2016; Revised 18 April 2016; Accepted 26 May 2016.
still remains a difficult challenge for clinicians.3,5,6 DOI 10.1002/ppul.23504
Johanson et al. defined criteria about 40 years ago in Published online in Wiley Online Library
order to improve the VAP diagnosis and later in 1991 (wileyonlinelibrary.com).

ß 2016 Wiley Periodicals, Inc.

2 _Işgu
€ der et al.

Pugin et al. enlarged these criteria and defined a scoring and a soluble form of the receptor, s-TREM-1, may be
system named Clinical Pulmonary Infection Score (CPIS) released into bodily fluids with up-regulation of TREM-1.9
(Table 1).7,8 Initially demonstrated improved diagnostic Recent studies have indicated that s-TREM-1 in bron-
accuracy for VAP, compared to clinical criteria. However choalveolar lavage fluid (BALF) may be of value for VAP
the usefulness of CPIS to discriminate for VAP remains diagnosis.10,17,19,21 On the other hand there are some
controversial.9–12 Except only one of the few pediatric controversial studies.22,23 It should be stated that there are
studies, CPIS was found to be useful in VAP only limited number of studies about the efficiency of
diagnose.13–16 s-TREM in diagnosing VAP in childhood.24,25
Studies are made on various biological markers such as In this study, BAL sample was obtained via non-
procalcitonin (PCT), C-reactive protein (CRP), pro- bronchoscopic method from the children with suspected
inflammatory cytokines, and coagulation cascade ele- VAP and s-TREM-1 levels were measured and compared
ments to facilitate the diagnostic process and to assess the with control group in order to determine its value in VAP
prognosis of VAP. However, PCT and CRP were reported diagnosis. Besides we measured the blood levels of PCT
to be promising biomarkers of only supporting the and calculated CPIS scores, then we assessed the effects
diagnosis.17–20 of these variables in diagnosing VAP.
Triggering receptor expressed on myeloid cells
(TREM), which is expressed on polymorphonuclear MATERIALS AND METHODS
granulocytes and mature monocytes, belongs to an
Study Design and Patients
immunoglobulin superfamily of receptors, first reported
by Gibot et al. as a biomarker of VAP diagnosis. Bacterial The study was approved by local ethic committee and
or fungal infection may activate expression of the receptor, informed consents were obtained from the parents. This
study was conducted in the PICU of a tertiary care children
research and training hospital. All children (1 month to
18 years of age) followed in PICU between January 2014
TABLE 1— Clinical Pulmonary Infection Score (CPIS) and December 2015 who were mechanically ventilated at
Points least for 48 hr were prospectively evaluated.

Temperature Exclusion Criteria Were

36.5 or 38.4 0 Neutropenia (500 neutrophils/mm3), known immune
38.5 or 38.9 1 deficiency, corticosteroid use, (>1 mg/kg prednisolone,
39 or <36.5 2 over 1 month14), HIV infection, alveolar hemorrhage,
Blood leukocytes ml known inflammatory diseases such as as ulcerative colitis,
4,000 or 11,000 0
<4,000 or >11,000 1 vasculitis etc.
Band forms 50% Add 1 For each patient enrolled into the study, a control
point patient was enrolled with similar age, gender, Pediatric
Tracheal secretions Risk of Mortality 3 (PRISM 3) score and duration of
Absence of tracheal secretions 0 mechanical ventilation support (more than 48 hr) before
Presence of nonpurulent tracheal secretions 1
Presence of purulent tracheal secretions 2 BAL sampling. This control patient was neither having
Oxygenation antibiotic therapy nor admitted to PICU with infectious
PaO2/FiO2, mmHg > 240, or Acute respiratory distress 0 problems. In patients meet these criteria, BAL samples
syndrome (ARDS) were taken as soon as possible. Thus we have created a
PaO2/FiO2  240 and no ARDS 2 control group with an equal number of children and
Pulmonary radiography
No infiltrate 0 similar to case group.
Diffuse infiltrate 1 The patients from the case group who had positive
Localized infiltrate 2 quantitative BALF culture (104 colony-forming unit-
Progression of pulmonary infiltrate CFU/ml) were named as confirmed VAP, and the others
No radiographic progression 0 named as non-confirmed VAP.
Radiographic progression (after heart failure and 2
ARDS excluded)
Culture of bronchoalveolar lavage Data Collection
Pathogenic bacteria growth () 0
Pathogenic bacteria growth (þ) 104 CFU/ml 1
In addition to the demographic data of the cases,
Same pathogenic bacteria seen on Gram stain Add 1 PRISM 3 scores and the chief reason for PICU admission
point were noted. Body temperature, PO2/FiO2 ratio, white
Total blood cell (WBC) count, CRP, PCT levels, CPIS,
A score of >6 was considered suggestive of ventilatory associated antibiotic use, s-TREM-1 levels in BALF, quantitative
pneumonia
mini-BAL culture results, total duration of mechanical
Pediatric Pulmonology
 Childhood Ventilator Associated Pneumonia Diagnosis 3

ventilation, duration of mechanical ventilation before catheter is made up of two catheters, one is inside of the
BAL obtaining, length of stay, and death rates in PICU other; and the outer one has a tip made up of polietilen
were assessed. Also age, gender, PRISM 3 score, CPIS, glucose. As it is described before catheter is placed into
WBC, CRP, PCT, s-TREM-1 levels, and quantitative the endotracheal tube until a resistance is felt.26 Then,
mini-BAL culture results of the control group were 20 cc saline is sent and, 2/3 of the saline is aspirated from
recorded. the catheter. Obtained BALF divided into two sterile
parts. One of the specimen was used for microscobic
VAP Definition examination and quantitative culture whereas the other
specimen is used for s-TREM 1 measurement.
For diagnosing VAP, we have used the “Pneumonia
(Ventilator-associated [VAP] and non-ventilator-associated
Sample Processing and Microbiological
Pneumonia [PNEU] Event” named pedVAP surveillance of
Examination
Centers for Disease Control and Prevention (CDC). The
diagnostic criteria under the title “Specific Site Algorithms BALF samples were send to the microbiology
for Pneumonia with Common Bacterial or Filamentous laboratory within 30 min after collection.
Fungal Pathogens and Specific Laboratory Findings, After delivery to the laboratory, all specimens were
abbreviated as PNU2 (Pneumonia 2)” was used to diagnose inoculated on blood agar, chocalate agar, and eosine
VAP.1 metyhlene blue agar plates by using 1 ml sterile
(1) Two or more serial chest imaging test results with at disposable loop and gram and giemsa stained smears
least one of the following: are prepared. Inoculated media were incubated at
378C and chocalate agar plates were also incubated in
 New or progressive and persistent infiltrate. 5% CO2 incubator. After overnight incubation, all plates
 Consolidation. were checked for bacterial growth and incubation period
 Cavitation. was prolonged to 48 hr if no growth was detected.
 Pneumatoceles, in infants 1 year old. Colony counts were made from the plates in which
bacterial growth was obtained and colony counts
(2) At least one of the following findings: 1  104 CFU/ml were accepted diagnostic threshold
for infection.1 Stained smears were examined under oil
 Fever (>38.08C or >100.48F) immersion at 1,000 magnification. Identification of
 Leukopenia (4,000 WBC/mm3) or leukocytosis bacteria isolated from the specimens were performed
(>12,000 WBC/mm3) with automated Vitek 2 identication system (bio-
(3) And at least one of the following: Merieux, France). The antimicrobial sensitivity testing
was performed on Mueller Hilton agar plates by Kirby–
 New onset of purulent sputum or change in character of Bauer’s disc diffusion method and inhibition zones were
sputum, or increased respiratory secretions, or increased measured by using an automatic zone reader (ADAGIO,
suctioning requirements. Bio-Rad, France) according to the European Committee
 New onset or worsening cough or dyspnea or of Antimicrobial Susceptibility Testing (EUCAST)
tachypnea. guidelines.
 Rales or bronchial breath sounds.
 Worsening gas exchange (O2 desaturations, PaO2/FiO2 sTREM-1 Assay
240), increase oxygen requirements or increased
The BALF was centrifuged at 10.000 revolutions per
ventilator demand).
minute at 48C for 30 min, and the supernatant was frozen
If one of the findings above is found, microbiological at 808C until used for s-TREM-1 measurements. A
diagnostic methods were used. We accepted the case as quantitative sandwich enzyme immunoassay (Quantikine
VAP, if we achieve 104 CFU/ml in the culture of BALF Human TREM-1 Immunoassay; R&D Systems, Minne-
obtained via non bronchoscopic mini-BAL catheter, as apolis, MN), was used to detect s-TREM-1 in BALF in
one of the microbiological diagnostic methods that CDC accordance to the manufacturers’ instructions, as
suggests.1 described previously.9

Obtaining BAL Sample Statistical Analyses

BALF was obtained with 1
Combicath Plastimed First of all we compared the case group and the control
Pediatric mini-BAL catheter which is 35 cm in length group then we compared the quantitative culture
and 1.9 mm in diameter, using blind and non- confirmed and non-confirmed VAP groups, among case
broncoscopic technique. The sterile and single sheated group, in terms of all variables.
Pediatric Pulmonology
4 _Işgu
€ der et al.

We used mean and standard deviation for continuous sensitivity and specificity were determined for each of all
variables which were normally distributed and median these three variables (Fig. 1).
and interquartile range for the variables which were The patients, whose CPIS, s-TREM-1, or PCT values
abnormal distributed. Unpaired Student t or Mann–Whitney were higher than the selected cutoff values, according to
U tests used for comparing these variables, according to Kaplan–Meier analysis and Log-rank test, were found to
their distribution. Categorical variables were expressed as have higher cumulative VAP rate (Fig. 2A–C).
percent and compared via Chi-square or Fisher’s exact tests. Correlation analysis was made between the numerical
Spearman analysis has been performed in order to data of the VAP cases and a moderate/mild significant
investigate the correlation between the numerical data of positive correlation was found between s-TREM-1 and
the VAP patients. Multivariate logistic regression analysis both the CPIS and the WBC count. (r ¼ 0.385, P ¼ 0.008;
was undertaken to determine independent risk factors for r ¼ 0.39, P ¼ 0.007, respectively). This correlation was
VAP. s-TREM-1, PCT levels, and CPIS were included into high significant positive between PCT and CRP (r ¼ 0.61,
this analysis as they can be risk factors for VAP P ¼ 0.002) and moderate significant positive between
development. A Receiver Operating Characteristic (ROC) PCT and WBC (r ¼ 0.58; P ¼ 0.006) whereas high
curve was constructed for determination o cutoff value of significant positive correlation was found between CPIS
s-TREM-1, CPIS, and serum PCT for the diagnosis of VAP. and WBC (r ¼ 0.66, P ¼ 0.001).
Kaplan–Meier and Log-rank test were performed in order to The most common causative agent in the quantitative
investigate the effects of cutoff values which were culture confirmed VAP cases was gram-negative bacteria
determined by ROC analysis. For all analyses a two-tailed (Pseudomonas aeroginosa 19.6%, Klebsiella pneumonia
P-value < 0.05 was considered statistically significant. 17.4%, Acinetobacter baumannii 17.4%, Echerichia coli
SPSS 15.0 software was used for all statistical analysis. 15.2%, and Serratia marcescens 8.7%) followed by gram-
positive bacteria (methicillin susceptible Staphylococcus
RESULTS aureus) and the Candida species (C. albicans 6.5%,
C. parapsilosis 2.1%). Non-confirmed VAP group had no
Study group was consisted of 58 case and 58 control culture positive BALFs.
patients, with a sum of 116 individuals. There is no
difference between the case group and control group in
DISCUSSION
terms of age, gender, PRISM 3 score, admission reason
for PICU, and duration of mechanical ventilation before National Healthcare Safety Network represented a
BALF obtaining. Case group had significantly higher surveillance report for PICUs in 2009 and VAP
body temperature, CPIS, WBC, CRP, PCT, and incidence was estimated to be 2.3% for each 1,000
s-TREM-1 levels. In the case group, 34.5% of the patients ventilator days whereas in 2013 the same incidence was
were receiving antimicrobial therapy whereas none of the 0.3%. This incidence was reported as 6.5% by
control group had antimicrobial therapy before BALF International Nosocomial Infection Control Consortium
obtaining (Table 2). (INICC) in 2012 report.27–29 In our hospital, VAP
In the case group, who had confirmed VAP with incidence was found to be 3.9% for each 1,000
quantitative culture analysis with 104 CFU/ml patho- ventilator days in 2015.
gens, PRISM 3 ve CPIS scores, and WBC, CRP, PCT Ventilator-associated pneumonia, is very often nowa-
s-TREM-1 levels were found to be higher than days and it is still an important problem for clinicians due
non-confirmed VAP group. When compared with non- to difficulties during its diagnosis and increasing effects
confirmed VAP group, duration of mechanical ventila- on length of stay, duration of mechanical ventilation time,
tion, length of stay in PICU, and mortality rates are and mortality risk in PICUs.2,5,6 Our study, which has
higher in confirmed VAP group, as they were expected. been conducted in order to facilitate this process reveled
Reason for admission, body temperature, duration of that s-TREM-1 of BALF, serum PCT levels, and CPIS are
mechanical ventilation before BALF obtaining, and very helpful parameters during diagnostic process of
antibiotic therapy rates were similar for both groups pediatric VAP.
(Table 3). Findings like persistent or progressive infiltration
Multivariate logistic regression analysis was performed accompanied by fever or leukocytosis, increased respira-
to determine independent risk factors. PCT, s-TREM-1, tory secretions or increased suctioning requirements and
and CPIS were found to be strong independent risk factors hypoxia are the pediatric VAP criteria of CDC’s PNU2
for the development of VAP (Table 4). definition1; however, they are nonspecific and can be
Area under curve (AUC) values of s-TREM-1, CPIS, easily seen in PICU patients according to systemic
and PCT values were extremely high at the ROC curve so symptoms of another infectious disease, atelectasis, lung
that these variables were assumed to be very effective on edema due to heart failure or non-pneumonic acute
the development of VAP and cutoff values with quite high respiratory distress syndrome.4,9 In support of this view,
Pediatric Pulmonology
 Childhood Ventilator Associated Pneumonia Diagnosis 5
TABLE 2— Characteristics of the Case and Control Groups in Study Population

Variables Case group (n ¼ 58) Control group (n ¼ 58) P-value (% 95 CI6)

Age1 (month) 56.2, 89.9, 3.5–192 66.2, 108.8, 3–193 0.159
Gender (female, n/%) 20/34.5 21/36.2 0.85
PRISM 3 score1 9.5, 10, 2–49 11.5, 15, 2–47 0.35
Reason for admission (n/%)
Acute respiratory failure 15/25.9 7/12.1 0.058
Neuromuscular disease 19/32.8 22/37.9 0.56
Shock 16/27.6 19/32.8 0.54
Drug overdose 5/8.6 3/5.2 0.71
Post-operative 3/5.2 7/12.1 0.03

CPIS2,3 7.75  2.9 2  1.8 <0.001

Body temperature (8C)3 38.4  1.5 36.7  0.3 <0.001
(1.2–2.08)
Laboratory results
White blood cell3  103/ml 17.6  8.5 9.4  1.4 <0.001

C-reactive protein1 mg/dl 9.45, 10.5, 0.33–31 0.33, 0.39, 0.33–1.8 <0.001

Procalcitonin1 ng/ml 8.6, 9.3, 0.05–54 0.12, 1.75, 0.05–3.6 <0.001

s-TREM-11,4 pg/ml 465, 501, 82–1,218 1.35, 5.8, 0–11.3 <0.001

Duration of Mechanical ventilation before BALF5 obtaining1 (hour) 99.5, 56, 58–210 97, 42, 50–207 0.69
Antibiotic therapy ratio (n/%) 27/46.6 0/0 <0.001

Death in Pediatric Intensive Care Unit (n/%) 20/34.5 16/27.6 0.42
1
Median, interquartile range (IQR), minimum–maximum.
2
Clinical Pulmonary Infection Score.
3
Mean  standard deviation.
4
Soluble triggering receptor expressed on myeloid cells-1.
5
Bronko-alveoler lavage fluid.
6
Confidence interval.


P < 0.05.

in our study, there was no difference in body temperature risk factor for VAP diagnosis (Odds ratio: 1.07 [95%CI:
between confirmed and non-confirmed VAP groups. 0.98–1.16]) and we calculated the cutoff value as
In addition, BALF specimens were obtained from 12 1.9 ng/ml (sensitivity 93.5%, specificity 91.7%). Contra-
suspected VAP patients in accordance with these criteria, dictory results and different cutoff values indicates the
but no culture positivity was detected and VAP diagnosis need for further researches.
was excluded. The findings which causes us to suspect Pugin et al. defined a scoring system named Clinical
VAP were later shown to develop from another focus of Pulmonary Infection Score considering body tempera-
infection which accompanied atelectasis or pulmonary ture, WBC, the presence of tracheal secretion, PaO2/FiO2
edema during their follow up. ratio, chest X-ray, culture results of tracheal aspiration in
Over the years, many studies have been made to find a 1991 and they concluded that scores above six were
marker for diagnosis of VAP.10,17,18,30,31 One of them is associated with high likelihood of VAP.8 In the pediatric
PCT. PCT is a peptide precursor of the hormone calcitonin studies with this scoring system, Sachdev et al.13 reported
and is composed of 116 amino acids but has no hormonal that a score of eight had a sensitivity and a specificity of
activity. In normal conditions it is produced by 80%. da Silva et al.14 found that serial CPIS measure-
parafollicular cells of the tiroid and by the neuroendocrine ments could be useful in order to determine the duration
cells of the lung and the intestine but as a response to a and success of antimicrobial therapy in one of their study
proinflamatory stimulus it is produced mainly by the cells and they suggested that it would be useful to use this
of the lung and the intestine.18 Initially it has been shown scoring system as a screening tool of VAP diagnosis in
to be more valuable than CRP in the diagnosis of bacterial another study.15 Conversely Mardganieva et al. reported
infections. However, later it has been reported to be that CPIS is not helpful in pediatric VAP diagnosis.16 In
elevated in different non-infectious cases.18 Duflo et al. our study, we found that CPIS is a very effective
had reported that PCT can be a complementary marker in independent risk factor for VAP diagnosis (Odds ratio:
VAP diagnosis but some studies reporting negative views 4.63, 95%CI: 1.5–14). Setted cutoff value of CPIS score
on this subject are available.32 There are some other 6, has 91.3% sensitivity and specificity 83.3%. Gibbot
studies ongoing about this issue. Su et al. showed that et al. also reported that they found that cutoff value of
PCT levels higher than 0.516 ng/ml can be a marker of CPIS score 6 has an odds ratio of 3, in an adult study.9 As
poor prognosis if it is used together with CPIS in VAP it has seen we have limited knowledge about this score in
cases.10 In our study, we detected PCT as an independent pediatric age group.13–16 Moreover, one of the
Pediatric Pulmonology
6 _Işgu
€ der et al.

TABLE 3— Comparison of the Case Groups, According to the Balf Quantitative Culture Test Confirmation

Confirmed VAP group Non-confirmed VAP group P-value

Variables (n ¼ 46) (n ¼ 12) (%95 CI6)
Age1 (month) 56.2, 89.9, 3.5–192 42.7, 94.8, 4–171 0.88
Gender (female, n/%) 15/32.6 5/41.7 0.73
PRISM 3 score1 12.5, 16, 4–47 8.5, 6, 2–17 0.021

Reason for admission (n/%)
Acute respiratory failure 12/26.1 3/25 0.62
Neuromuscular disease 15/32.6 4/33.3 0.22
Shock 13/28.3 3/25 0.72
Drug overdose 4/8.7 1/8.3 0.54
Post operative 2/4.3 1/8.3 0.053
CPIS2,3 8.6  2.5 4.3  1.2 <0.001
(3.2–5.3)
Body temperature (C8)3 38.6  1.6 37.8  0.8 0.11 (1.7 to
þ0.18)
Laboratory results
PO2/FiO2 ratio1 350, 219, 124–479 406, 257, 176–482 0.29
White blood cell3  103/ml 18.9  9.1 12.4  2.03 0.001

C-reactive protein1 mg/dl 11.2, 11.6, 5.5–31.3 0.38, 6.5, 0.33–11.8 <0.001

Procalcitonin1ng/ml 8.2, 7.8, 1.2–54 0.06, 0.78, 0.05–3.6 0.001

s-TREM-11,4 pg/ml 598, 503, 277–1218 143, 119, 82-303 <0.001

Antibiotic therapy ratio (n/%) 22/47.8 5/41.7 0.7
Duration of mechanic ventilation3 before BALF5 103.7  27 119.3  44.6 0.137 (36.5 to
obtaining (hour) þ5.2)
Total duration of mechanic ventilation (day) 18.7  6.2 8.9  2.8 0.003
(7.3–12.2)
Length of stay in Pediatric Intensive Care Unit3 (day) 21.5  6.4 15.3  5.9 0.002
(3.3–11.5)
Death in Pediatric Intensive Care Unit (n/%) 18/39.1 2/16.7 0.018

1
Median, interquartile range (IQR), minimum–maximum.
2
Clinical Pulmonary Infection Score.
3
Mean  standard deviation.
4
Soluble triggering receptor expressed on myeloid cells-1.
5
Bronko-alveoler lavage fluid.
6
Confidence interval.


P < 0.05.

components of CPIS is tracheal aspiration culture and it Determann et al.17 emphasized the importance of serial
does not give result before 48–72 hr. Contrarily we need s-TREM measurements and find a cutoff value of 20 pg/ml
faster diagnostic methods to confirm VAP diagnosis and (sensitivity 75%, specificity 84%) whereas Huh et al.11
in order to start appropriate treatment. reported the cutoff value of 184 pg/ml (sensitivity 86%,
Gibot et al. were the first investigators to study specificity 90%). Contrarily Hororenko et al.,33 Anand
s-TREM-1 levels from the BALF acquired via mini- et al.,4 Guy et al.,22 and Palazzo et al.23 reported that they
BAL technique in adults and they have found that levels did not find s-TREM beneficial in VAP diagnosis. All the
above 5 pg/ml were very strong independent predictors studies above are the adult studies. There are only two
(sensitivity 98%, specificity 90%, Odds ratio: 41.5).9 pediatric studies and in one of them s-TREM-1 was
measured in the sputum of bronchiectasic children but not
for the purpose of VAP diagnosis.25 The other pediatric
TABLE 4— Independent Risk Factors for the Development of
study investigated the s-TREM for VAP diagnosis from the
VAP According to Multivariate Logistic Regression Analysis BALF and exhaled ventilator condensate of the children
after cardiac surgery. They assessed the microbiological
Variables P-value Odds ratio Confidence interval 95% culture and s-TREM from the BALF collected via feeding
s-TREM-1 1
0.001

3.64 1.7–7.78 tube and they could not find s-TREM levels effective in
PCT2 ng/ml 0.01
1.07 0.98–1.16 VAP diagnosis; however, s-TREM levels above
CPIS3 0.007
4.63 1.5–14 20.58 pg/ml were found to be helpful in VAP diagnosis
1 with a sensitiviy of 50% and specificity of 86% in exhaled
Soluble triggering receptor expressed on myeloid cells-1.
2
Procalcitonin. ventilator condensate.24 Different cutoff values and
3
Clinical Pulmonary Infection Score. contradictory opinions were reported in studies and we


P < 0.05. have limited knowledge in pediatric age group. It is
Pediatric Pulmonology
 Childhood Ventilator Associated Pneumonia Diagnosis 7

Fig. 1. Receiver operating characteristic curve for s-TREM-1, CPIS, and PCT values. Areas under
the curve for s-TREM-1, CPIS, and PCT, are 0.99 (95%CI, 0.97–1), 0.94 (95%CI, 0.81–1), and 0.98
(95%CI, 0.95–1.01), respectively. The cutoff values for s-TREM-1,CPIS and PCT, are 281 pg/ml,
(sensitivity 95.7%, specificity 91.7%); 6, (sensitivity 91.3%, specificity 83.3%); and 1.9 ng/ml
(sensitivity 93.5%, specificity 91.7%), respectively. PCT, procalcitonin; s-TREM, soluble
triggering receptor expressed on myeloid cells-1; CPIS, Clinical Pulmonary Infection Score.

considered that different cutoff values were related to (6,281 pg/ml, 1.9 ng/ml, respectively), had higher cumu-
differences between BALF obtaining and s-TREM lative VAP rates also. The VAP confirmed patients had
analysing techniques and prior antibiotic use.4,17,23 In positive correlations between s-TREM-1 and WBC; PCT
our study, BALF was obtained via Combicath Plastimed and WBC; PCT and CRP; CPIS and WBC. Thus, these
Pediatric mini-BAL catheter. This sterile and single three parameters were shown to have efficient in
sheated catheter is made up of two catheters, one is inside determining infectious processes.
of the other; and the outer one has a tip made up of When the results of the above-mentioned studies and
polyethylene glucose. s-TREM-1 analysis was made by a our study were assessed together, we can say that those
quantitative sandwich enzyme immunoassay technique. clinical signs and laboratory findings neither can be
Control group was consisted of the patients who did not excluded from the diagnosis criteria nor can make
receive antibiotherapy during BALF sampling, and case the diagnose alone itself. The main problem is delaying
group was found to have significant higher levels of the antimicrobial therapy or applying inappropriate
s-TREM-1 than the control group. More importantly, antimicrobial therapy till pending the outcome of
between VAP confirmed and non-confirmed cases by microbiological diagnosing tests. Defining a new scoring
quantitative culture analysis, there was no difference in the system like CPIS, with all the laboratory tests including
utilization rate of antibiotic use; however, s-TREM-1 s-TREM and PCT but excluding microbiological diag-
levels in VAP confirmed patients were found to be nose tests can be the remedy for this problem. So that a
significantly higher than other groups, thus s-TREM-1 has stronger and standard “VAP suspicion” definition can be
been identified as an independent risk factor for VAP created and redundant antimicrobial treatment can be
diagnosis (Odds ratio: 3.64, 95%CI: 1.7–7.78). The cutoff avoided.
value for the s-TREM-1 was found to be 281 pg/ml; with a In the last report of CDC for adult patients, ventilator-
sensitivity of 95.7% and specificity of 91.7%. associated condition (VAC), Infection-related ventilator-
In our study, we found that the patients who had higher associated Complication (IVAC), and Possible Ventilator-
CPIS, s-TREM-1, or PCT levels than the cutoff values Associated Pneumonia (PVAP) with microbiological
Pediatric Pulmonology
8 _Işgu
€ der et al.

Fig. 2. Kaplan–Meier survival curves for patients with (A) s-TREM-1 less than 281 pg/ml and
281 pg/ml (P ¼ 0.019 with Log-rank test); (B) CPIS less than 6 and 6 (P ¼ 0.015 with Log-rank
test); and (C) PCT less than 1.9 ng/ml and 1.9 fL (P ¼ 0.04 with Log-rank test). VAP, ventilator
associated pneumonia; s-TREM, soluble triggering receptor expressed on myeloid cells-1; CPIS,
Clinical Pulmonary Infection Score; PCT, procalcitonin.

confirmation definitions were described under the title believed to form the basis for new researches because
“Ventilator-Associated Events (VAE).”34 Cirulis et al. there are only few pediatric researches.
modified the actual CDC surveillance definition in 2016 On the other hand our study has several limitations. The
for pediatric patients and they reported that the actual and study has relatively small number of patients and it is
the pediatric modified VAE criteria both have poor single centered, non-randomized, non-blinded study. It
sensitivity but good specificity.6 Presently, the efforts of would be better if BALF could be obtained via
improving the VAP diagnosis are ongoing. bronchoscopic method which is more invasive but more
The strength of this study is that it was a prospective, sensitive and specific. These important limitations
controlled study with detailed inclusion and exclusion precludes the generalization of the findings.
criteria. Another agent increasing the power of our study As a conclusion, s-TREM-1 of BALF, serum PCT, and
is that there is no difference between the case and control CPIS are very beneficial in process of VAP diagnosis
group in terms of confounding factors. Our study is however in order to use these results in daily clinical
Pediatric Pulmonology
 Childhood Ventilator Associated Pneumonia Diagnosis 9

practice further randomized, controlled, and blinded 16. Mardganieva EA, Mironov PI, Rudnov VA. Diagnosis and
researches with larger case series are needed. treatment of ventilator-associated pneumonia in children. Anes-
teziol Reanimatol 2006;1:34–38.
17. Determann RM, Millo JL, Gibot S, Korevaar JC, Vroom MB, van
REFERENCES der Poll T, Garrard CS, Schultz MJ. Serial changes in soluble
triggering receptor expressed on myeloid cells in the lung during
1. CDC: Pneumonia (Ventilator-associated [VAP] and non-ventila- development of ventilator-associated pneumonia. Intensive Care
tor-associated Pneumonia [PNEU] Event. pg1-16. Available at: Med 2005;31:1495–1500.
http://www.cdc.gov/nhsn/PDFs/pscManual/6pscVAPcurrent.pdf. 18. Chastre J, Luyt CE, Trouillet JL, Combes A. New diagnostic and
Accessed 26 Feb 2016. prognostic markers of ventilator-associated pneumonia. Curr Opin
2. Northway T, Langley JM, Skippen P. Health care associated Crit Care 2006;12:446–451.
infection in the pediatric intensive care unit: epidemiology and 19. Grover V, Pantelidis P, Soni N, Takata M, Shah PL, Wells AU,
control. Keeping patients safe. In: Fuhrman BP, Zimmerman JJ, Henderson DC, Kelleher P, Singh S. A biomarker panel
editors. Pediatric critical care. Philadelphia: Elsevier Sounders; (bioscore) incorporating monocytic surface and soluble
2011. pp 1358–1360. TREM-1 has high discriminative value for ventilator-associated
3. Klompas M. Does this patient have ventilator-associated pneumonia: a prospective observational study. PLoS ONE
pneumonia? JAMA 2007;297:1583–1593. 2014;9:e109686.
4. Anand NJ, Zuick S, Klesney-Tait J, Kollef MH. Diagnostic 20. Gibot S, Cravoisy A, Dupays R, Barraud D, Nace L, Levy B,
implications of soluble triggering receptor expressed on myeloid Bollaert PE. Combined measurement of procalcitonin and soluble
cells-1 in BAL fluid of patients with pulmonary infiltrates in the TREM-1 in the diagnosis of nosocomial sepsis. Scand J Infect Dis
ICU. Chest 2009;135:641–647. 2007;39:604–608.
5. Grgurich PE, Hudcova J, Lei Y, Sarwar A, Craven DE. 21. Wu CL, Lu YT, Kung YC, Lee CH, Peng MJ. Prognostic value of
Diagnosis of ventilator-associated pneumonia: controversies dynamic soluble triggering receptor expressed on myeloid cells in
and working toward a gold standard. Curr Opin Infect Dis 2013; bronchoalveolar lavage fluid of patients with ventilator-associated
26:140–150. pneumonia. Respirology 2011;16:487–494.
6. Cirulis MM, Hamele MT, Stockmann CR, Bennett TD, Bratton 22. Oudhuis GJ, Beuving J, Bergmans D, Stobberingh EE, ten Velde
SL. Comparison of the new adult ventilator-associated event G, Linssen CF, Verbon A. Soluble triggering receptor expressed on
criteria to the Centers for Disease Control and Prevention Pediatric myeloid cells-1 in bronchoalveolar lavage fluid is not predictive
Ventilator-Associated Pneumonia Definition (PNU2) in a popula- for ventilator-associated pneumonia. Intensive Care Med
tion of pediatric traumatic brain injury patients. Pediatr Crit Care 2009;35:1265–1270.
Med 2016;17:157–164. 23. Palazzo SJ, Simpson TA, Simmons JM, Schnapp LM. Soluble
7. Johanson WG Jr, Pierce AK, Sanford JP, Thomas GD. Nosocomial triggering receptor expressed on myeloid cells-1 (sTREM-1) as a
respiratory infections with gram-negative bacilli. The significance diagnostic marker of ventilator-associated pneumonia. Respir
of colonization of the respiratory tract. Ann Intern Med Care 2012;57:2052–2058.
1972;77:701–706. 24. Matsuno AK, Carlotti AP. Role of soluble triggering receptor
8. Pugin J, Auckenthaler R, Mili N, Janssens JP, Lew PD, Suter PM. expressed on myeloid cells-1 for diagnosing ventilator-associated
Diagnosis of ventilator-associated pneumonia by bacteriologic pneumonia after cardiac surgery: an observational study. BMC
analysis of bronchoscopic and nonbronchoscopic “blind” bron- Cardiovasc Disord 2013;13:107.
choalveolar lavage fluid. Am Rev Respir Dis 25. Masekela R, Anderson R, de Boeck K, Vreys M, Steel HC,
1991;143:1121–1129. Olurunju S, Green RJ. Expression of soluble triggering receptor
9. Gibot S, Cravoisy A, Levy B, Bene MC, Faure G, Bollaert PE. expressed on myeloid cells-1 in childhood CF and non-CF
Soluble triggering receptor expressed on myeloid cells and the bronchiectasis. Pediatr Pulmonol 2015;50:333–339.
diagnosis of pneumonia. N Engl J Med 2004;350:451–458. 26. Papazian L, Thomas P, Garbe L, Guignon I, Thirion X, Charrel J,
10. Su LX, Meng K, Zhang X, Wang HJ, Yan P, Jia YH, Feng D, Xie Bollet C, Fuentes P, Gouin F. Bronchoscopic or blind sampling
LX. Diagnosing ventilator-associated pneumonia in critically ill techniques for the diagnosis of ventilator-associated pneumonia.
patients with sepsis. Am J Crit Care 2012;21:e110–e119. Am J Respir Crit Care Med 1995;152:1982–1991.
11. Huh JW, Lim CM, Koh Y, Oh YM, Shim TS, Lee SD, Kim WS, 27. Dudeck MA, Weiner LM, Allen-Bridson K, Malpiedi PJ, Peterson
Kim DS, Kim WD, Hong SB. Diagnostic utility of the soluble KD, Pollock DA, Sievert DM, Edwards JR. National Healthcare
triggering receptor expressed on myeloid cells-1 in bronchoal- Safety Network (NHSN) report, data summary for 2012, device-
veolar lavage fluid from patients with bilateral lung infiltrates. Crit associated module. Am J Infect Control 2013;41:1148–1166.
Care 2008;12:R6. 28. Edwards JR, Peterson KD, Mu Y, Banerjee S, Allen-Bridson K,
12. Chastre J, Fagon JY. Ventilator-associated pneumonia. Am J Morrell G, Dudeck MA, Pollock DA, Horan TC. National
Respir Crit Care Med 2002;165:867–903. Healthcare Safety Network (NHSN) report: data summary for
13. Sachdev A, Chugh K, Sethi M, Gupta D, Wattal C, Menon G. 2006 through 2008, issued December 2009. Am J Infect Control
Clinical Pulmonary Infection Score to diagnose ventilator- 2009;37:783–805.
associated pneumonia in children. Indian Pediatr 29. Rosenthal VD, Bijie H, Maki DG, Mehta Y, Apisarnthanarak A,
2011;48:949–954. Medeiros EA, Leblebicioglu H, Fisher D, Álvarez-Moreno C,
14. da Silva PS, de Aguiar VE, Fonseca MC. How the modified Khader IA, et al. International Nosocomial Infection Control
Clinical Pulmonary Infection Score can identify treatment failure Consortium (INICC) report, data summary of 36 countries, for
and avoid overusing antibiotics in ventilator-associated pneumo- 2004–2009. Am J Infect Control 2012;40:396–407.
nia. Acta Paediatr 2014;103:e388–e392. 30. Giamarellos-Bourboulis EJ, Zakynthinos S, Baziaka F, Papado-
15. da Silva PS, de Aguiar VE, de Carvalho WB, Machado Fonseca michelakis E, Virtzili S, Koutoukas P, Armaganidis A, Giamarellou
MC. Value of clinical pulmonary infection score in critically ill H, Roussos C. Soluble triggering receptor expressed on myeloid
children as a surrogate for diagnosis of ventilator-associated cells 1 as an anti-inflammatory mediator in sepsis. Intensive Care
pneumonia. J Crit Care 2014;29:545–550. Med 2006;32:237–243.

Pediatric Pulmonology
10 _ u
Işg € der et al.

31. Rea-Neto A, Youssef NC, Tuche F, Brunkhorst F, Ranieri VM, 33. Horonenko G, Hoyt JC, Robbins RA, Singarajah CU, Umar A,
Reinhart K, Sakr Y. Diagnosis of ventilator-associated Pattengill J, Hayden JM. Soluble triggering receptor expressed
pneumonia: a systematic review of the literature. Crit Care on myeloid cell-1 is increased in patients with ventilator-
2008;12:R56. associated pneumonia: a preliminary report. Chest 2007;132:
32. Duflo F, Debon R, Monneret G, Bienvenu J, Chassard D, 58–63.
Allaouchiche B. Alveolar and serum procalcitonin: diagnostic and 34. CDC: Ventilator-Associated Event (VAE) [Internet]. pg1-47. Available
prognostic value in ventilator-associated pneumonia. Anesthesi- at: http://www.cdc.gov/nhsn/PDFs/pscManual/10-VAE_FINAL.pdf.
ology 2002;96:74–79. Accessed 29 Feb 2016.

Pediatric Pulmonology