Heart & Lung xxx (2016) 1e8

Contents lists available at ScienceDirect

Heart & Lung

journal homepage: www.heartandlung.org

Efficacy of two noninvasive weaning strategies in intubated patients

with chronic obstructive pulmonary disease: A meta-analysis and
indirect treatment comparison
Pei-Jun Li, PhD a, Ting Wang, PhD a, Jun Xiao, PhD b, Fa-Ming Jiang, PhD a, Jian Luo, PhD a,
Chao-Li Shi, MD a, Guan-Jian Liu, PhD c, Zong-An Liang, MD a, *
a
Department of Respiratory and Critical Care Medicine, West China School of Medicine/West China Hospital, Sichuan University,
No. 37 Guoxue Road, Chengdu, Sichuan 610041, PR China
b
Department of Intensive Care Unit, West China School of Medicine/West China Hospital, Sichuan University,
37 Guoxue Road, Chengdu 610041, PR China
c
Chinese Evidence-Based Medicine Center, West China School of Medicine/West China Hospital, Sichuan University,
37 Guoxue Road, Chengdu 610041, PR China

a r t i c l e i n f o a b s t r a c t

Article history: The purpose of our indirect comparison was to explore the optimal switching time to noninvasive
Received 7 October 2015 ventilation for further weaning in patients with chronic obstructive pulmonary disease (COPD) under-
Received in revised form going invasive mechanical ventilation. A comprehensive literature search was performed to identify
26 December 2015
randomized controlled trials comparing noninvasive weaning at spontaneous breathing trial (SBT) failure
Accepted 29 December 2015
Available online xxx
after meeting simple weaning criteria or at the pulmonary infection control window (PIC window) with
conventional invasive weaning in COPD patients. Using conventional invasive weaning as a bridge, we
indirectly compared the two noninvasive weaning strategies using the Bucher approach. Noninvasive
Keywords:
Chronic obstructive pulmonary disease
weaning at SBT failure after meeting simple weaning criteria was associated with an extended duration
Noninvasive ventilation of endotracheal mechanical ventilation (standardized mean difference 1.90, 95% CI 1.27e2.53, P < 0.001)
Weaning compared with noninvasive weaning at the PIC window. No significant differences in mortality or the
Meta-analysis rate of ventilator-associated pneumonia were observed. Our study suggests that the PIC window may be
Indirect treatment comparison a promising switching time for noninvasive weaning in COPD patients.
Ó 2016 Elsevier Inc. All rights reserved.

Introduction ventilator time is dedicated to the weaning process.5 To reduce

Prolonged invasion ventilation may increase the risk of
ventilator-associated pneumonia (VAP) and airway trauma,1e3
which are closely linked with increased morbidity and mortality.4
Discontinuing mechanical ventilation presents a challenging
period. Esteban et al have analyzed the evolution of mechanical
ventilation practices over time among a large and diverse group of
patients with respiratory failure and concluded that 40% of the total
Abbreviations: CIs, confidence intervals; COPD, chronic obstructive pulmonary
disease; ETMV, endotracheal mechanical ventilation; ICU, intensive care unit; PIC
window, pulmonary infection control window; RCT, randomized controlled trial;
RR, risk ratio; SBT, spontaneous breathing trial; SMD, standardized mean differ-
ence; VAP, ventilator associated pneumonia.
* Corresponding author. Tel.: þ86 18980601259; fax: þ86 028 85422377.
E-mail address: za_liang319@163.com (Z.-A. Liang).
intubation-associated complications and mortality in critically ill,
mechanically-ventilated patients, investigators have explored the
effect of noninvasive weaning, which involves early extubation
before meeting conventional weaning criteria and the direct
application of noninvasive ventilation. A recent Cochrane system-
atic review suggested that noninvasive weaning may reduce mor-
tality and the rate of VAP predominantly in intubated patients with
chronic obstructive pulmonary disease (COPD) compared with
conventional invasive weaning.6
Choosing the appropriate time to switch to noninvasive
ventilation plays a crucial role in promoting successful noninva-
sive weaning. Nava et al first conducted a randomized controlled
trial (RCT) on the application of noninvasive weaning in intu-
bated COPD patients, which considered spontaneous breathing
trial failure after meeting simple weaning criteria (Pre-SBT suc-
cess) as the switching time.7 Pre-SBT success means meeting the

0147-9563/$ e see front matter Ó 2016 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.hrtlng.2015.12.008
2 P.-J. Li et al. / Heart & Lung xxx (2016) 1e8

Records identified through

Identification
database searching
(n=586)

Records after duplicates

removed
(n=395)
Screening

Records screened through Records excluded

titles and/or abstracts (n=374)
(n=395)

Full-text articles excluded (n=15)

Full-text articles assessed
Not RCTs (n=3)
for eligibility
Eligibility

(n=21) Not all of COPD (n=5)

Not relevant comparisons (n=3)
Meta-analyses and reviews (n=4)

Studies included from Additional studies included from

database searching relevant reference lists
(n=6) (n=2)
Included

Studies included in quantitative

synthesis (meta-analysis)
(n=8)

Fig. 1. Study flow diagram. COPD, chronic obstructive pulmonary disease; RCT, randomized controlled trial.

simple weaning criteria as assessed by oxygenation status,
mental status, cough ability, and clinical condition, but failing an
SBT. The trial concluded that sequential noninvasive weaning
during this period may decrease the weaning time and reduce
mortality and the rate of VAP compared with conventional
invasive weaning.
Similar clinical benefits have also been demonstrated with
another noninvasive weaning strategy of adopting the pulmonary
infection control window (PIC window) as the switching time for
intubated COPD patients.8 The PIC window involves significantly
reduced radiographic infiltrations, decreased respiratory secre-
tions, a normal or near-normal body temperature, and a normal
leukocyte count. The exact and optimal time as a generally accepted
boundary of standard switching, however, has not been
determined. To date, there are no published RCTs on a head-to-head
comparison of clinical efficacy of the two noninvasive weaning
strategies.
Indirect treatment comparison has become a promising method
to resolve such issues, when there is no direct evidence from ran-
domized trials.9,10 If direct evidence of both A versus C and B versus
C are available, an adjusted indirect comparison of A versus B will
be conducted using the same intervention C as a bridge.11,12
Therefore, we searched RCTs comparing noninvasive weaning at
the two switching times with conventional invasive weaning in
intubated patients with COPD. Then, a relevant adjusted indirect
comparison was performed to explore the optimal time for
noninvasive weaning to further reduce mortality and the rate of
VAP.
 P.-J. Li et al. / Heart & Lung xxx (2016) 1e8 3

Table 1
Main characteristics of the included studies.

First author, year Patients (noninvasive/ Inclusion criteria (patients) Switching timing of noninvasive weaning Interventional mode Outcomesa
invasive)

Numbers Age (y) Male% Noninvasive Invasive

Nava, 19987 25/25 69/67 NA ETMV due to COPD exacerbation SBT failure after meeting simple PSV PSV 1-2, 4
weaning criteria
Wang, 20058 47/43 68/70 60/74 ETMV due to COPD exacerbated Pulmonary infection control window BiPAP SIMV þ PSV 1-5
by pulmonary infection
El-Shimy, 201316 20/20 65/66 NA ETMV due to COPD exacerbation SBT failure after meeting simple BiPAP SIMV þ PSV 1-2, 4
weaning criteria
17
Mishra, 2014 25/25 60/61 80/72 ETMV due to COPD exacerbation SBT failure after meeting simple BiPAP PSV 1-4
weaning criteria
18
Prasad, 2009 15/15 58/61 80/60 ETMV due to COPD exacerbation SBT failure after meeting simple BiPAP PSV 1-4
weaning criteria
Zou, 200619 38/38 72/72 76/74 ETMV due to COPD exacerbated Pulmonary infection control window BiPAP SIMV þ PSV 1-5
by pulmonary infection
Wang, 200420 14/14 61/65 79/71 ETMV due to COPD exacerbated Pulmonary infection control window BiPAP SIMV þ PSV 1-4
by pulmonary infection
21
Zheng, 2005 17/16 73/72 65/63 ETMV due to COPD exacerbated Pulmonary infection control window BiPAP PSV 1-4
by pulmonary infection

BiPAP, bilevel positive airway pressure; COPD, chronic obstructive pulmonary disease; ETMV, endotracheal mechanical ventilation; NA, not available; PSV, pressure support
ventilation; SBT, spontaneous breathing trial; SIMV, synchronized intermittent mandatory ventilation; y, years.
a
Outcomes were intended for meta-analysis and indirect treatment comparison: 1. mortality; 2. the rate of ventilator associated pneumonia; 3. duration of endotracheal
mechanical ventilation; 4. duration of intensive care unit stay; 5. reintubation rate.

Methods selection process included screening through titles and/or abstracts

and further full-text evaluation. Disagreement was resolved by
Search strategy consulting a third investigator.

We searched PubMed, EMBASE, and the Cochrane Central Reg-

Data extraction and quality assessment
ister of Controlled Trials without language restrictions using the
following search terms: (“chronic obstructive pulmonary disease”
Two authors independently extracted the data regarding gen-
OR “COPD”) AND (“noninvasive ventilation” OR “non-invasive
eral information (first author and year of publication), baseline
ventilation” OR “noninvasive positive pressure ventilation” OR
characteristics of the patients, switching time of noninvasive
“non-invasive positive pressure ventilation” OR “NPPV” OR “NIPPV”
weaning, interventional mode of mechanical ventilation, and effi-
OR “positive-pressure ventilation” OR “positive-pressure respira-
cacy outcomes. The quality of the included studies was assessed by
tion”) AND (“weaning” OR “ventilator weaning” OR “wean” OR
the Cochrane Risk of Bias Tool for RCTs.13 The risk of bias was rated
“extubation”). The last search was updated during August 2015. In
as low risk, unclear risk, or high risk. Any disagreement was
addition, the reference lists of relevant studies were examined for
resolved by consensus.
other possible eligible articles missed by the electronic search.

Study selection
We prespecified the following inclusion criteria: (1) RCTs, (2)
adult patients with COPD requiring endotracheal mechanical
ventilation (ETMV), (3) intervention with noninvasive weaning at
Pre-SBT success or at the PIC window, (4) control intervention with
conventional invasive weaning, and (5) outcomes including at least
one of mortality and the rate of VAP. Studies only containing ab-
stracts published were excluded.
Two authors independently evaluated all of the potentially
eligible articles according to the study inclusion criteria. The study
Statistical analysis
The primary outcomes were mortality and the rate of VAP. The
secondary outcomes included the duration of ETMV, the duration
of intensive care unit (ICU) stay, and the reintubation rate. A meta-
analysis was conducted to pool the risk ratio (RR) with the Mantel-
Haenszel method for dichotomous data and standardized mean
differences (SMD) with the inverse variance method for continuous
data using Review Manager 5.3 software (Cochrane Collaboration,
Oxford, UK). A random-effects model was selected to incorporate
covariate adjustment to mitigate overspecification.14 The results
were presented with 95% confidence intervals (CIs), and statistical

Table 2
Quality assessment of the included randomized controlled trials.

First author, year Random assignment Allocation concealment Blind evaluation for outcomes Incomplete outcome data Selective reporting
Nava, 19987 Unclear risk Low risk High risk Low risk Low risk
Wang, 20058 Unclear risk Unclear risk High risk Low risk Low risk
El-Shimy, 201316 Unclear risk Unclear risk High risk Low risk Low risk
Mishra, 201417 Low risk Low risk High risk Low risk Low risk
Prasad, 200918 Low risk Unclear risk High risk Low risk Low risk
Zou, 200619 Low risk Unclear risk High risk Low risk Low risk
Wang, 200420 Unclear risk Unclear risk High risk Low risk Low risk
Zheng, 200521 Low risk Unclear risk High risk Low risk Low risk
4 P.-J. Li et al. / Heart & Lung xxx (2016) 1e8

7
16
17
18

7
16
17
18

Fig. 2. Effect of noninvasive weaning at spontaneous breathing trial failure after meeting the simple weaning criteria (Pre-SBT success) versus conventional invasive weaning on
mortality and the rate of ventilator-associated pneumonia (VAP).

significance was defined as a P-value less than 0.05. All statistical
tests were two-sided. We utilized I2 statistics for calculating sta-
tistical heterogeneity (threshold P < 0.1) and defined that I2 values
of 25%, 50%, and 75% represented low, moderate, and high het-
erogeneity, respectively.15 The sensitivity analysis was performed
by sequentially excluding one study to explore any sources of
heterogeneity. A funnel plot was used to detect the publication bias
for mortality.
We applied summary estimates of efficacy outcomes from the
meta-analysis to indirect treatment comparison through STATA/SE
12.1 software (StataCorp LP, College Station, TX) using the methods
described by Bucher et al.11 Treating conventional invasive weaning
as a bridge, we indirectly compared the clinical efficacy of nonin-
vasive weaning at Pre-SBT success versus noninvasive weaning
during the PIC window in intubated adults with COPD.
Results
Study characteristics
We identified 395 articles through the electronic databases after
duplicates were removed, of which six studies met the pre-
determined inclusion criteria (Fig. 1).7,8,16e19 Two additional studies
published in Chinese were also included from the reference lists of
relevant articles.20,21 The main characteristics of all of the included
studies that enrolled a total of 397 intubated adults with COPD are
summarized in Table 1. A total of four studies evaluated the efficacy
of noninvasive weaning at Pre-SBT success versus conventional
invasive weaning,7,16e18 while the remaining four studies compared
noninvasive weaning during the PIC window with conventional
invasive weaning.8,19e21 The mean age of the patients ranged from
58 to 73 years, and the proportion of males varied from 60% to 80%
according to the available data. The etiology of initial ETMV for all
participants was an acute exacerbation of COPD. All of the eligible
studies reported mortality, the rate of VAP, and the duration of ICU
stay. Only two studies did not evaluate the duration of ETMV.7,16
The reintubation rate was only reported in two studies regarding
noninvasive weaning during the PIC window.8,19 All of the included
studies provided diagnostic criteria for VAP, of which three studies
defined VAP as the presence of new and persistent (>48 h) lung
infiltrates on chest radiography combined with fever and total
leukocyte count >10,000/ml after 48 h on a ventilator.16e18 Based on
this definition, two additional studies added the microbiological
evidence of pulmonary infection to diagnose VAP,7,20 and the
remaining three studies also reported the changes in the charac-
teristics of respiratory secretions.8,19,21
The simple weaning criteria in the four studies met the
following common features: (1) adequate oxygenation reflected by

17
18

7
16
17
18

Fig. 3. Effect of noninvasive weaning at spontaneous breathing trial failure after meeting the simple weaning criteria (Pre-SBT success) versus conventional invasive weaning on the
duration of endotracheal mechanical ventilation (ETMV) and the duration of intensive care unit (ICU) stay.
 P.-J. Li et al. / Heart & Lung xxx (2016) 1e8 5

8
19
20
21

8
19
20
21

Fig. 4. Effect of noninvasive weaning during the pulmonary infection control window (PIC window) versus conventional invasive weaning on mortality and the rate of ventilator-
associated pneumonia (VAP).

a PaO2  60 mm Hg or SaO2  88% on an FiO2 of 40%, (2) satisfactory
neurologic status, and (3) adequate cough and clinical stability. The
common characteristics of the PIC window in the remaining four
studies were as follows: (1) significantly reduced radiographic in-
filtrations, (2) significantly decreased respiratory secretions, and (3)
body temperature below 38  C and/or leukocyte count below
10  109/L.
The quality assessment of the included studies is shown in
Table 2. Overall, all of the included trials were considered to be of
moderate to high quality.
Effect of noninvasive weaning at Pre-SBT success versus
conventional invasive weaning
As indicated in Fig. 2, the pooled analysis revealed that nonin-
vasive weaning at Pre-SBT success significantly reduced mortality
(RR 0.46, 95% CI 0.27e0.78, P ¼ 0.004) and the rate of VAP (RR 0.17,
95% CI 0.06e0.46, P < 0.001) without statistical heterogeneity in
intubated patients with COPD, compared with conventional inva-
sive weaning. The former also exhibited a significantly decreased
duration of ETMV (SMD 0.49, 95% CI 0.93 to 0.04, P ¼ 0.03) and
duration of ICU stay (SMD 0.80, 95% CI 1.11 to 0.48, P < 0.001)
without statistical heterogeneity (Fig. 3).
Effect of noninvasive weaning at PIC window versus conventional
invasive weaning
As shown in Fig. 4, noninvasive weaning during the PIC window
was associated with a statistically significant lower mortality (RR
0.38, 95% CI 0.17e0.84, P ¼ 0.02) and rate of VAP (RR 0.33, 95% CI
0.18e0.59, P < 0.001) than conventional invasive weaning in
intubated patients with COPD, with no statistical heterogeneity.
Additionally, the former technique showed a significant advantage
over the latter in terms of duration of ETMV (SMD 2.00, 95%
CI 2.92 to 1.07, P < 0.001) and duration of ICU stay (SMD 1.24,
95% CI 2.06 to 0.42, P ¼ 0.003) (Fig. 5); however, the statistical
heterogeneity was high for both the duration of ETMV (I2 ¼ 86%,
P < 0.001) and the duration of ICU stay (I2 ¼ 86%, P < 0.001). After
exclusion of the study by Wang et al (2005) in the sensitivity
analysis,8 prominently decreased heterogeneity was observed for

8
19
20
21

8
19
20
21

Fig. 5. Effect of noninvasive weaning during the pulmonary infection control window (PIC window) versus conventional invasive weaning on the duration of endotracheal me-
chanical ventilation (ETMV) and the duration of intensive care unit (ICU) stay.
6 P.-J. Li et al. / Heart & Lung xxx (2016) 1e8

Table 3 Discussion
Results of indirect comparison of noninvasive weaning at Pre-SBT success versus
noninvasive weaning at PIC window.
In the current meta-analysis and indirect treatment comparison,
Indicators RR/SMD 95% confidence interval P-value we identified eight RCTs comparing noninvasive weaning at Pre-
Mortality 1.21 0.46e3.16 0.70 SBT success or during the PIC window with conventional invasive
The rate of VAP 0.52 0.16e1.67 0.27 weaning, or against each other, in intubated adults with COPD.
Duration of ETMV 1.90 1.27e2.53 <0.001
Noninvasive weaning at either Pre-SBT success or the PIC window
Duration of ICU stay 0.79 0.14e1.44 0.02
was found to be superior to conventional invasive weaning in terms
RR (risk ratio) was used for assessing mortality and the rate of VAP (ventilator-
of mortality, the rate of VAP, the duration of ETMV, and the duration
associated pneumonia), and SMD (standardized mean difference) was applied for
the duration of ETMV (endotracheal mechanical ventilation) and the duration of ICU
of ICU stay. Indirect comparison of noninvasive weaning at Pre-SBT
(intensive care unit) stay. Statistically significant results are displayed in bold. success versus noninvasive weaning at the PIC window showed no
PIC window, pulmonary infection control window; Pre-SBT success, spontaneous significant differences for mortality and the rate of VAP, although
breathing trial failure after meeting the simple weaning criteria. the former seemed to be accompanied by a significantly extended
duration of ETMV and duration of ICU stay.
Similar to several previous meta-analyses,6,22,23 our meta-
both the duration of ETMV (I2 ¼ 0%, P ¼ 0.92) and the duration of
analysis suggests that intubated patients with COPD may be
ICU stay (I2 ¼ 48%, P ¼ 0.15). Furthermore, similar benefits favoring
ideally suited to noninvasive weaning. COPD exacerbations are
noninvasive weaning during the PIC window were found for both
often accompanied by respiratory muscle failure,24 which is
the duration of ETMV (SMD 2.39, 95% CI 2.83 to 1.94,
responsible for difficult weaning and prolonged ETMV. Noninvasive
P < 0.001) and the duration of ICU stay (SMD 1.59, 95% CI 2.15
ventilation is bound up with the ability of noninvasive positive
to 1.02, P < 0.001).
pressure ventilation to relax the respiratory muscle, increase tidal
volume, and reduce intrinsic positive end-expiratory pressure,25,26
and thus facilitates liberating COPD patients from invasive venti-
Outcomes of indirect treatment comparison lation. A reduced duration of ETMV, in turn, is associated with a
lower risk of VAP rate,1e3 as endotracheal tubes contribute to
The results of the indirect comparison of noninvasive weaning bacterial colonization and aspiration of contaminated secretions.27
at the two switching times are presented in Table 3. No significant Finally, decreasing the rate of VAP may decrease the duration of the
differences were observed for mortality (RR 1.21, 95% CI 0.46e3.16, ICU stay and reduce the risk of death in critically ill patients.4
P ¼ 0.70) or the rate of VAP (RR 0.52, 95% CI 0.16e1.67, P ¼ 0.27). Because ICU stay is one of the major determinants of hospitaliza-
Compared with noninvasive weaning at the PIC window, nonin- tion costs in COPD exacerbations,28 a longer ICU stay may lead to
vasive weaning at Pre-SBT success was associated with an increased significantly higher total costs. From this angle, noninvasive
duration of ETMV (SMD 1.90, 95% CI 1.27e2.53, P < 0.001) and weaning compared with conventional invasive weaning is a cost-
duration of ICU stay (SMD 0.79, 95% CI 0.14e1.44, P ¼ 0.02). Suffi- effective strategy for COPD exacerbations.
cient data were not available for indirect comparison of the rein- Formulating the time to switch to noninvasive ventilation is a
tubation rate. key issue in the administration of noninvasive weaning. Our study
summarized and defined the common characteristics of Pre-SBT
success and the PIC window in the included trials. The PIC win-
Publication bias dow is not a well-known concept in the application of noninvasive
weaning for COPD exacerbations. The assessment for the PIC win-
A funnel plot was created to evaluate publication bias for mor- dow mainly consists of radiographic infiltrations, respiratory se-
tality in the pooled analysis. The individual effect estimates showed cretions, body temperature, and leukocyte count, which potentially
a nearly symmetric distribution around the overall treatment ef- considers stable clinical status. The appearance of the PIC window
fect, indicating that there may be minimal publication bias in the in intubated patients with COPD exacerbated by pulmonary infec-
included studies (Fig. 6). tion indicates that the patients’ major problem has converted to
ventilator insufficiency from insufficient secretion drainage. During
this period, the acute phase of disease is resolved and the patients’
condition becomes stable. Based on this principle, Wang et al have
suggested that adopting the PIC window as the switching time for
noninvasive weaning may facilitate advanced extubation and thus
reduce the rate of VAP and mortality in intubated patients with
COPD exacerbated by pulmonary infection.8 Notably, prompt
recognition of the PIC window may be a time-consuming and
challenging procedure for medical and paramedical staff, as the
judgments for significantly reduced radiographic infiltrations and
respiratory secretions rely on close monitoring via chest X-rays in
addition to observation and sophisticated professional standards.
The present indirect comparison indicated that noninvasive
weaning during the PIC window significantly decreased the dura-
tion of ETMV in COPD exacerbations compared with noninvasive
weaning at Pre-SBT success; however, we did not observe any
significant differences in terms of the rate of VAP and mortality
between the two noninvasive weaning strategies. The inconsis-
Fig. 6. Funnel plot to evaluate publication bias for mortality. PIC window, pulmonary
tency of diagnostic criteria for VAP across trials should be taken into
infection control window; Pre-SBT success, spontaneous breathing trial failure after account. Combining new or progressive radiographic infiltrates
meeting the simple weaning criteria. with fever, purulent or increased secretions of airway, leukocytosis,
 P.-J. Li et al. / Heart & Lung xxx (2016) 1e8
and microbiological evidence of respiratory secretions after 48 h on
ventilator is used to diagnose VAP; however, these criteria are often
subjective and ambiguous, and have a high degree of interobserver
variability.29 We thus imagined that over-diagnosis for VAP might
occur in the included trials regarding the PIC window, which might
be responsible for the lack of superiority for noninvasive weaning
at the PIC window over that at Pre-SBT success in terms of the VAP
rate, although a decreased duration of ETMV was observed with the
former.
The duration of noninvasive ventilation after extubation in the
noninvasive weaning group may have an important role in mor-
tality. Noninvasive ventilation was applied continuously except
during meals and for expectoration in the four included trials
regarding Pre-SBT success,7,16e18 whereas the two studies on the
PIC window only claimed that using noninvasive ventilation
initially for over 2 h would be accepted,19,21 and the remaining two
did not provide specific statements.8,20 The shortage of noninvasive
ventilation after reaching the PIC window may lead to poor efficacy
and an increase in mortality, which may explain the result that a
reduced duration of ETMV for noninvasive weaning at the PIC
window was not followed by a lower risk of mortality compared to
noninvasive weaning at Pre-SBT success.
There are several potential limitations to be addressed in the
current study. First, the power of indirect comparison is inferior to
that of RCTs; however, in the absence of head-to-head trials, this is
a well-established and accepted method.9,10 Second, the PIC win-
dow is a less familiar and relatively subjective concept, although it
represents a novel approach for noninvasive weaning in COPD
exacerbations. Third, the heterogeneity of the populations should
be noted. The cause for acute exacerbation of COPD in the four
trials on the PIC window was pulmonary infection, while the
remaining four trials regarding Pre-SBT success did not mention
specific causes. Notably, respiratory infection is responsible for
acute exacerbation in the majority of COPD patients,30 which may
reduce the effect of this heterogeneity on our analysis to a large
extent. Additionally, the meta-analysis of duration of ETMV and
duration of ICU stay showed high statistic heterogeneity compared
to noninvasive weaning at the PIC window to conventional inva-
sive weaning. This finding may be related to variations among
studies in compliance with mechanical ventilation and the defi-
nition of ICU discharge criteria. Finally, no sufficient data on the
reintubation rate were available, which makes it difficult to eval-
uate the relationship between advanced extubation and the
reintubation rate.
Conclusions
The results of the present study together with the results of
previous meta-analyses provide strong evidence supporting
noninvasive weaning in intubated patients with COPD. Given the
limitations to available data, however, whether the PIC window is a
much more suitable switching time to noninvasive weaning than
Pre-SBT success in COPD patients exacerbated by pulmonary
infection is difficult to assess. Large well-designed RCTs are needed
to further explore the optimal time for noninvasive weaning in this
select group of COPD patients exacerbated by pulmonary infection.
Acknowledgment
We thank Professor Dong-Tao Lin (College of Foreign Languages,
Sichuan University) for providing language help.
7
References
1. Fagon JY, Chastre J, Domart Y, et al. Nosocomial pneumonia in patients
receiving continuous mechanical ventilation: prospective analysis of 52 epi-
sodes with use of a protected specimen brush and quantitative culture tech-
niques. Am Rev Respir Dis. 1989;139:877e884.
2. Tobin MJ. Mechanical ventilation. N Engl J Med. 1994;330:1056e1061.
3. Boles JM, Bion J, Connors A, et al. Weaning from mechanical ventilation. Eur
Respir J. 2007;29:1033e1056.
4. Heyland DK, Cook DJ, Griffith L, Keenan SP, Brun-Buisson C. The attributable
morbidity and mortality of ventilator-associated pneumonia in the critically ill
patient. Am J Respir Crit Care Med. 1999;159:1249e1256.
5. Esteban A, Ferguson ND, Meade MO, et al, VENTILA Group. Evolution of me-
chanical ventilation in response to clinical research. Am J Respir Crit Care Med.
2008;177:170e177.
6. Burns KE, Meade MO, Premji A, Adhikari NK. Noninvasive positive-pressure
ventilation as a weaning strategy for intubated adults with respiratory fail-
ure. Cochrane Database Syst Rev:CD004127, http://dx.doi.org/10.1002/
14651858.CD004127.pub3, 2013;12.
7. Nava S, Ambrosino N, Clini E, et al. Noninvasive mechanical ventilation in the
weaning of patients with respiratory failure due to chronic obstructive pul-
monary disease: a randomized, controlled trial. Ann Intern Med. 1998;128:
721e728.
8. Wang C, Zhang Q, Cao Z, Wei L, Cheng Z, Liu S, Collaborating Research Group
for Noninvasive Mechanical Ventilation of Chinese Respiratory Society. Pul-
monary infection control window in treatment of severe respiratory failure of
chronic obstructive pulmonary diseases: a prospective, randomized controlled,
multi-centred study. Chin Med J. 2005;118:1589e1594.
9. Song F, Altman DG, Glenny A-M, Deeks JJ. Validity of indirect comparison for
estimating efficacy of competing interventions: empirical evidence from pub-
lished meta-analyses. BMJ. 2003;326:472.
10. Edwards S, Clarke M, Wordsworth S, Borrill J. Indirect comparisons of treat-
ments based on systematic reviews of randomised controlled trials. Int J Clin
Pract. 2009;63:841e854.
11. Bucher HC, Guyatt GH, Griffith LE, Walter SD. The results of direct and indirect
treatment comparisons in meta-analysis of randomized controlled trials. J Clin
Epidemiol. 1997;50:683e691.
12. Wells GA, Sultan SA, Chen L, Khan M, Coyle D. Indirect Evidence: Indirect
Treatment Comparisons in Meta-analysis. Ottawa: Canadian Agency for Drugs
and Technologies in Health; 2009.
13. Higgins JPT, Green S. Cochrane Handbook for Systematic Reviews of Interventions
(Version 5.1.0). London: The Cochrane Collaboration, http://handbook.
cochrane.org; 2011. Accessed 20.07.15.
14. Cooper NJ, Sutton AJ, Morris D, Ades A, Welton NJ. Addressing between-study
heterogeneity and inconsistency in mixed treatment comparisons: application
to stroke prevention treatments in individuals with non-rheumatic atrial
fibrillation. Stat Med. 2009;28:1861e1881.
15. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in
meta-analyses. BMJ. 2003;327:557.
16. El-Shimy WS, Barima MA, Abo El-Magd GH, Mansour SA. Non invasive venti-
lation versus synchronized intermittent mandatory ventilation with pressure
support in weaning of COPD patients: comparative study. Egypt J Chest Dis
Tuberc. 2013;62:159e166.
17. Mishra M, Chaudhri S, Tripathi V, Verma AK, Sampath A, Chauhan NK. Weaning
of mechanically ventilated chronic obstructive pulmonary disease patients by
using non-invasive positive pressure ventilation: a prospective study. Lung
India. 2014;31:127e133.
18. Prasad SB, Chaudhry D, Khanna R. Role of noninvasive ventilation in weaning
from mechanical ventilation in patients of chronic obstructive pulmonary
disease: an Indian experience. Indian J Crit Care Med. 2009;13:207e212.
19. Zou S, Zhou R, Chen P, et al. Application of sequential noninvasive following
invasive mechanical ventilation in COPD patients with severe respiratory
failure by investigating the appearance of pulmonary-infection-control-win-
dow. J Cent South Univ. 2006;31:120e124.
20. Wang X, Du X, Zhang W. Observation of the results and discussion on the
timing of transition from invasive mechanical ventilation to noninvasive
ventilation in COPD patients with concomitant acute respiratory failure.
Shandong Med. 2004;44:4e6.
21. Zheng R, Liu L, Yang Y, Guo F, Huang Y, Qiu H. Prospective randomized
controlled clinical study of sequential non-invasive following invasive me-
chanical ventilation in patients with acute respiratory failure induced COPD.
Chin J Emerg Med. 2005;14:21e25.
22. Burns KE, Adhikari NK, Keenan SP, Meade M. Use of non-invasive ventilation to
wean critically ill adults off invasive ventilation: meta-analysis and systematic
review. BMJ. 2009;338:b1574.
23. Zhu F, Liu Z, Long X, et al. Effect of noninvasive positive pressure ventilation on
weaning success in patients receiving invasive mechanical ventilation: a meta-
analysis. Chin Med J. 2013;126:1337e1343.
24. Organized jointly by the American Thoracic Society, the European Respiratory
Society, the European Society of Intensive Care Medicine, and the Société de
Réanimation de Langue Française, and approved by ATS Board of Directors,
December 2000. International consensus conferences in intensive care medi-
cine: noninvasive positive pressure ventilation in acute respiratory failure. Am J
Respir Crit Care Med. 2001;163:283e291.
8 P.-J. Li et al. / Heart & Lung xxx (2016) 1e8

25. Appendini L, Patessio A, Zanaboni S, et al. Physiologic effects of positive end-
expiratory pressure and mask pressure support during exacerbations of chronic
obstructive pulmonary disease. Am J Respir Crit Care Med. 1994;149:1069e1076.
26. Nava S, Bruschi C, Rubini F, Palo A, Iotti G, Braschi A. Respiratory response and
inspiratory effort during pressure support ventilation in COPD patients.
Intensive Care Med. 1995;21:871e879.
27. Craven DE, Steger KA. Epidemiology of nosocomial pneumonia: new per-
spectives on an old disease. Chest. 1995;108:1Se16S.
28. Chen YH, Yao WZ, Cai BQ, et al. Economic analysis in admitted patients with
acute exacerbation of chronic obstructive pulmonary disease. Chin Med J.
2008;121:587e591.
29. Klompas M. Interobserver variability in ventilator-associated pneumonia sur-
veillance. Am J Infect Control. 2010;38:237e239.
30. Papi A, Bellettato CM, Braccioni F, et al. Infections and airway inflammation in
chronic obstructive pulmonary disease severe exacerbations. Am J Respir Crit
Care Med. 2006;173:1114e1121.