Available via license: CC BY-NC-ND 4.0
Content may be subject to copyright.
Research
Adding positive airway pressure to mobilisation and respiratory techniques
hastens pleural drainage: a randomised trial
Elinaldo da Conceição dos Santos
a,b
, Juliana de Souza da Silva
b
, Marcus Titus Trindade de Assis Filho
b
,
Marcela Brito Vidal
b
, Moisés de Castro Monte
c
, Adriana Cláudia Lunardi
a,d
a
Master and Doctoral Program in Physical Therapy, Universidade Cidade de São Paulo;
b
Department of Biological and Health Sciences, Universidade Federal do Amapá;
c
Department of Physical Therapy, Faculdade de Macapá, Macapá;
d
Department of Physical Therapy, School of Medicine, Universidade de São Paulo, São Paulo, Brazil
KEY WORDS
Pleural effusion
Chest drain
Respiratory care
Physical therapy
Clinical trial
ABSTRACT
Questions: In patients with a collection of fluid in the pleural space, do mobilisation and respiratory
techniques: shorten the drainage period and length of hospital stay; improve respiratory function and
oxygenation; and prevent pulmonary complications? Does the addition of positive airway pressure to this
regimen further improve the effects? Design: Randomised controlled trial with three intervention arms,
concealed allocation, intention-to-treat analysis and blinded assessment. Participants: One hundred and
fifty-six inpatients with a fluid collection in the pleural space and with chest drainage in situ. Intervention:
Participants received usual care and were randomly assigned to: a control group that also received sham
positive airway pressure (4 cmH
2
O) only (Con); an experimental group that received incentive spirometry,
airway clearance, mobilisation and the same sham positive pressure (Exp1); or an experimental group that
received the Exp1 regimen except that the positive airway pressure was 15 cmH
2
O (Exp2). Treatments were
provided three times per day for 7 days. Outcome measures: Days of chest tube drainage, length of hospital
stay, pulmonary complications and adverse events were recorded until hospital discharge. Costs in each
group were estimated. Results: The Exp2 group had shorter duration of chest tube drainage and length of
hospital stay compared with the Exp1 and Con groups. In addition, the Exp2 group had less antibiotic use
(18% versus 43% versus 55%) and pneumonia incidence (0% versus 16% versus 20%) compared with the Exp1
and Con groups (all p,0.01). The groups had similar rates of adverse events (10% versus 2% versus 6%,
p.0.05). Total treatment costs were lower in the Exp2 group than in the Exp1 and Con groups. Conclusions:
In patients with a fluid collection in the pleural space, the addition of positive pressure to mobilisation and
respiratory techniques decreased the duration of thoracic drainage, length of hospital stay, pulmonary
complications, antibiotic use and treatment costs. Registration:ClinicalTrials.gov NCT02246946. [dos Santos
EC, da Silva JS, de Assis Filho MTT, Vidal MB, Monte MC, Lunardi AC (2020) Adding positive airway
pressure to mobilisation and respiratory techniques hastens pleural drainage: a randomised trial.
Journal of Physiotherapy 66:19–26]
© 2019 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under
the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Introduction
Pleural effusion is a collection of excessive fluid in the pleural space,
which can have many aetiologies.
1
Each year, pleural effusion is
diagnosed in approximately 1.5 million people in the USA.
2
Among
people diagnosed with a pleural effusion, mortality at 1 year differs
with the aetiology: 25% if fromrenal failure, 46% if from hepatic failure,
50% if from cardiac failure and more if from malignancy.
3
After trauma,
blood (haemothorax) and other fluid may also collect in the pleural
space.
4
Some pleural fluid collections can be managed with aspiration of
the fluid via an intercostal needle –a procedure known as
thoracentesis.
5
However, British Thoracic Society guidelines
recommend the insertion of an intercostal tube attached to a closed
drainage system in malignant pleural effusions, empyema, traumatic
haemopneumothorax and after thoracic surgery.
6
Depending on its volume and aetiology, excessive pleural fluid
can cause dyspnoea and pleuritic pain. The presence of the
intercostal tube causes additional pain, which is not always able to
be controlled. Furthermore, various complications can develop
during the period that the intercostal tube is in situ. The most
common complication is drain blockage, which occurs in about 8%
of small-bore tubes (16 F) and 5% of large-bore tubes (20 F).
6
The next most common complications include infection of the
pleural fluid collection (empyema), movement of the tube into a
poor position and injury to internal structures.
6
For all these
reasons, facilitating rapid drainage of the pleural fluid collection is
an important aim of management.
Lung expansion techniques have been proposed as one group of
interventions that could be used to hasten the drainage of a pleural
fluid collection and thereby reduce the opportunity for complications
from the drainage tube.
7
A randomised controlled trial involving
Journal of Physiotherapy 66 (2020) 19–26
https://doi.org/10.1016/j.jphys.2019.11.006
1836-9553/© 2019 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/
licenses/by-nc-nd/4.0/).
journal homepage: www.elsevier.com/locate/jphys
104 inpatients with pleural effusions with or without an intercostal
drainage tube showed that the addition of breathing exercises and
mobilisation to other usual care reduced the severity of pleural
effusion, as judged by blinded assessment of chest radiographs, and
reduced length of stay by a mean of 12 days (95% CI 8 to 16).
7
Although those results for lung expansion techniques were very
promising,
7
few clinical practice guidelines for pleural effusion have
incorporated that evidence,
8
presumably because of some limitations.
Although the trial was generally very rigorous, some participants who
did not complete the allocated intervention were excluded, which is
contrary to the recommendation to analyse by intention to treat.
9
The
authors did not test whether the effects differed between the
subgroups managed with or without an intercostal drainage tube.
Furthermore, although the substantial effect on length of stay
suggests that the time spent with the intercostal drainage tube in situ
was probably also shorter, this was not measured. Perhaps most
importantly, the evidence comes from a single trial
7
so an attempt to
replicate the findings is warranted.
In addition to breathing exercises and mobilisation, another
intervention that might expand the lungs and promote drainage is
positive airway pressure, which can be appliednon-invasively via a face
mask. We hypothesised that increasing the intra-pleural pressure
would promote drainage and reabsorption of the pleural fluid
collection, thereby hastening the recovery of respiratory function,
permitting earlier removal of the chest drain and shortening the
hospital stay. However, the only trial of positive airway pressure for
pleural effusion examined continuous positive airway pressure (CPAP)
in participants with tuberculous pleural effusions and no drainage
tubes.
10
Therefore, the effects of positive airway pressure in patients
with intercostal tubes for pleural drainage are unknown.
Some indirect evidence suggests that positive airway pressure may
have additional advantages. It may help to prevent or treat respiratory
complications in patients after thoracic surgery,
11
where
end-operative placement of an intercostal drainage tube is routine. For
example, in patients undergoing lung resection, non-invasive
ventilation significantly improved gas exchange, spirometry and
length of stay.
12
In patients who had already developed
acute hypoxaemic respiratory insufficiency after lung resection,
non-invasive ventilation significantly improved arterial oxygenation
and respiratory rate, and significantly reduced reintubation and
mortality.
13
Positive effects have also been noted in other thoracic
surgical procedures such as coronary artery bypass grafting
14
and
resection of lung cancer.
15
However, it is important to remember that
many of these patients may primarily have pleural gas with some
resolving post-surgical bleeding and oedema in the pleural space,
rather than an extensive pleural fluid collection. Furthermore, the
beneficial effects after thoracic surgery may occur via mechanisms
that are unrelated to the pleural space, such as more efficient gas
exchange. Therefore, although these findings are promising, it would
be hazardous to extrapolate from these studies to draw conclusions
about the effects of positive airway pressure in the overall population
of patients receiving drainage of fluid via an intercostal tube.
Therefore, the research questions for this randomised controlled
trial were:
1. In patients with a collection of fluid in the pleural space, do
mobilisation and respiratory techniques: shorten the drainage
period and length of hospital stay; improve respiratory function
and oxygenation; and prevent pulmonary complications?
2. Does the addition of positive airway pressure to this regimen
further improve the effects?
Method
Design
This study was a randomised controlled trial with concealed
allocation, blinded assessors and intention-to-treat analysis. At two
university-affiliated tertiary hospitals in Macapá, Brazil, inpatients
with a collection of fluid in the pleural space and an intercostal drain
were invited to participate. Inpatients who met the eligibility criteria
and agreed to participate had the following data collected: personal
and anthropometric information; history of smoking; cause of the
pleural fluid collection; type of drainage; and tomographic
confirmation of a pleural fluid collection. All participants continued to
receive usual medical and nursing care, and were immediately
randomised to one of three groups: the control group (Con) received
no additional active interventions; the first experimental group
(Exp1) received breathing exercises, an airway clearance technique
and mobilisation; and the second experimental group (Exp2)
received the same interventions as Exp1 with the addition of periods
of CPAP.
Randomisation was performed by a researcher not involved in
patient selection, evaluation or intervention. The random sequence of
allocations was computer-generated, with each allocation placed in a
corresponding sequentially numbered opaque envelope. The
envelopes were sealed and stored securely where only the
researcher responsible for the randomisation had access. This
researcher informed only the physiotherapist who administered the
interventions of each participant’s random allocation.
16
This
physiotherapist remained blinded to the assessment information
being collected on the participants.
The randomly allocated interventions were commenced within 24
hours of insertion of the drainage system, and were continued three
times per day for 7 consecutive days. On the first day of the
intervention and after the intervention protocol was completed (on
the day after the intervention ended), pulmonary function and
peripheral oxygen saturation were reassessed. In cases where the
drain was removed before completing the intervention protocol
period, the reassessments were conducted immediately before the
drain was removed to homogenise the analyses. Every 24 hours while
the chest drain remained in situ, the drainage rate was quantified and
lung expansion was evaluated by a physician blinded to the
intervention groups. At this assessment, the physician decided
whether the drain should be removed or not. Treatment tolerability
and adverse events were also recorded for all patients. In addition,
patients were evaluated daily for the presence of pulmonary
complications and to estimate treatment costs until hospital
discharge. If participants presented any pulmonary complications,
they received individualised treatment according to their condition.
Participants
Between December 2014 and February 2017, inpatients with a
pleural fluid collection at either hospital were approached
consecutively, advised about the study, screened for eligibility
(if willing) and invited to participate in the study (if eligible). The
inclusion criteria were: being aged 18 years; having a diagnosis of a
fluid collection in the pleural space confirmed by computer
tomography; and having an intercostal thoracic drainage system in
situ for ,24 hours. Exclusion criteria were drowsiness, restlessness,
treatment refusal, haemodynamic instability, shock (systolic blood
pressure ,90 mmHg), facial trauma, ineffective cough, swallowing
impairment, vomiting, upper gastrointestinal bleeding, acute
myocardial infarction in the past 48 hours, or bullous emphysema.
17
Intervention
Usual nursing and medical care were provided to all participants
in all groups. Where positive airway pressure was provided, the active
and sham pressure levels were based on a previous study that tested
different values and evaluated the resulting lung expansion via
thoracic computer tomography.
18
Con group
Participants in this group breathed with CPAP of 4 cmH
2
O
(sham)
18
via an oronasal mask attached to a bedside ventilation unit
a
for 30 minutes, in order to have a similar intervention period for all
groups and guarantee blind assessment (ie, presence of equipment in
the room and the impression of the mask on the patient’s face).
20 dos Santos et al: Positive pressure for pleural drainage
Exp1 group
Participants in this group undertook: incentive spirometry
b
,
performing five sets of 20 repetitions; airway clearance manoeuvres
using a high-frequency oscillator
c
, performing five sets of 10
repetitions; breathing with CPAP of 4 cmH
2
O (sham)
18
via an oronasal
mask attached to a bedside ventilation unit
a
for 5 minutes while
sitting on a chair; and walking a distance of 100 metres.
Exp2 group
Participants in this group received all the same interventions as
those provided to the Exp1 group, except that the CPAP was not a
sham. Participants breathed with CPAP of 15 cmH
2
O (active)
18
via an
oronasal mask attached to a bedside ventilation unit
a
for 30 minutes
while sitting on a chair.
Outcome measures
Primary outcome
The primary outcome was the duration of thoracic drainage,
quantified in days until drain removal. The criteria for drain removal
were 200 ml fluid drainage in 24 hours and complete lung
expansion assessed by chest radiograph.
19
The decision about the
time of drain removal was made by a physician blinded to the
participant’s group. The physician verified the following criteria every
day during the period that drainage was in situ: drainage volume in
millilitres over 24 hours, and lung expansion assessed by chest
radiograph.
Secondary outcomes assessed during the treatment period
Pulmonary function was evaluated by spirometry following the
performance and acceptability criteria previously established by the
European Respiratory Society and American Thoracic Society.
20
Spirometry assessments were performed prior to the first
intervention and at the end of the scheduled intervention protocol
(Day 8). In cases where the criteria for drain removal were met before
the end of the protocol, spirometry was performed immediately
before the drain was removed. The spirometry measurements were
performed by an assessor blinded to the participant’s group. The
variables forced vital capacity (FVC) and forced expiratory volume in
1 second (FEV
1
) were calculated as a percentage of the predicted
values for the Brazilian population.
21
Discharge
Day 8
Baseline
People admitted and referred for assessment of eligibility (n = 468)
Excluded (n = 312)
.did not meet inclusion criteria (n = 309)
.declined to participate (n = 3)
Measured spirometry and oxygenation
Randomised (n = 156)
(n = 52) (n = 52) (n = 52)
Exp2 group
.usual care
.mobilisation
.respiratory techniques
.positive airway
pressure (15 cmH₂O)
Con group
.usual care
.sham positive airway
pressure (4 cmH₂O)
Exp1 group
.usual care
.mobilisation
.respiratory techniques
.sham positive airway
pressure (4 cmH₂O)
Loss to follow-up
.death (n = 1)
Measured spirometry, oxygenation, treatment tolerability (satisfaction and discomfort), and adverse
events
(n = 51) (n = 51) (n = 49)
Loss to follow-up
.discharged with
no authorisation
(n = 1)
Loss to follow-up
.discontinued
trial (n = 2)
.death (n = 1)
Measured duration of pleural drainage, length of hospital stay, treatment costs, and
antibiotic use
(n = 51) (n = 51) (n = 49)
Figure 1. Design and flow of participants through the trial. Note that data obtained before loss to follow-up were included in the analysis for some outcomes.
Exp2 = experimental group 2, Exp1 = experimental group 1, Con = control group.
Research 21
Peripheral oxygen saturation (SpO
2
) was assessed daily until
completion of the intervention protocol or removal of the thoracic
drain. Measurements were taken by an assessor blinded to the
intervention group. Oximetry was performed with the participant
breathing ambient air (ie, without oxygen supplementation) for at
least 10 minutes.
Treatment tolerability was assessed using participant ratings of
satisfaction and discomfort on a visual analogue scale ranging from
0 to 10 points. Participants were requested to report their satisfaction
with their randomly allocated respiratory interventions using scores
where 0 points indicated complete dissatisfaction and 10 points
indicated complete satisfaction with the treatment. Participants were
also instructed to respond about discomfort regarding the randomly
allocated respiratory interventions received, where 0 points indicated
no discomfort with the oronasal mask or flow and 10 points indicated
unbearable discomfort.
Adverse events were recorded during the positive airway pressure
treatment periods based on daily visual evaluation and questioning of
participants. Whenever an adverse event was detected, it was
monitored and reported to the ethics committee. In cases of
confirmed air leaks or aerophagia, the respiratory care session was
interrupted for 24 hours and the participant was re-evaluated the
next day.
Secondary outcomes assessed until discharge from hospital
Length of hospital stay was tallied as the total number of days in
hospital.
Treatment costs were estimated based on established values of
overnight hospital stay (R$500.00), antibiotic use (R$313.49), physio-
therapy session (R$150.00), equipment and accessories for positive
airway pressure (R$2000.00 divided by the number of participants in
theExp2group)forthewholelengthofhospitalstayofeachparticipant.
Pulmonary complications were assessed by a physician who was
blinded to each participant’s group, considering the following
events: pneumonia (chest radiograph with pulmonary infiltrate
associated with two of the following signs: purulent sputum,
hyperthermia .38.8 C, or increase in baseline leucocyte count
.25%),
22
atelectasis (abnormal chest radiograph associated with
acute respiratory symptoms)
23
and hypoxaemia (SpO
2
,85%
associated with respiratory symptoms).
24
In addition to pulmonary
complications, the lung entrapment rate was also verified by a
physician who was blinded to the intervention groups. Lung
entrapment was defined as contralateral displacement of the
mediastinum
25
verified by chest radiograph or computer tomo-
graphy, or reaccumulating pleural effusion within 24 to 72 hours.
26
Need for thoracotomy was also recorded until hospital discharge.
Data on antibiotic use were extracted from the medical prescrip-
tion records from the inclusion of the participant in the study until
hospital discharge. The type of antibiotic used was not evaluated.
Data analysis
It was calculated that a sample of 132 participants (43 per group)
was required, based on the primary outcome of thoracic drainage
duration. Two days was nominated as the smallest worthwhile effect,
anticipating a standard deviation of 3 days,
27
and requiring statistical
power of 80% and setting alpha at 5%. Incorporating an allowance for
an anticipated 15% loss to follow-up (eg, deaths, refusal to continue
participation in the study and hospital transfer),
27
resulted in a
required sample size of 156 patients (52 per group).
All analyses were performed by a researcher who was not
involved in the study participants’assessments and interventions.
Statistical analysis was performed following the intent-to-treat
principle.
9
The Kruskal-Wallis ANOVA was used to test the
continuous variables, namely: FVC, FEV
1
, SpO
2
, treatment costs and
treatment tolerability, which were not normally distributed. Pairwise
chi-squared tests were applied to the categorical variables
(pulmonary complications, adverse events and antibiotic use), with
between-group comparisons reported as absolute risk reduction (95%
CI). Kaplan-Meier survival analysis was used to test the time variables
(duration of thoracic drainage and length of hospital stay). A signif-
icance level of 5% (p,0.05) was adopted for all statistical analyses.
Results
Flow of participants, therapists, centres through the study
Four hundred and sixty-eight patients were referred for
evaluation and selected for possible inclusion in this study. Of these,
309 were ineligible and three refused to participate, so 156 patients
were included in the study. Figure 1 presents the flowchart of
enrolment and monitoring of the research participants. Although the
study allowed for an anticipated 15% loss to follow-up based on
existing research,
28
a 2% loss to follow-up occurred, with two deaths
and one unauthorised hospital discharge.
Compliance with the study protocol
No ineligible participants were randomised. All participants
commenced on the correctly designated intervention. Treatment was
interrupted in three patients who presented clinical complications
unrelated to the proposed intervention (hypovolaemia and cardiac
arrhythmia). No assessors were accidentally unblinded during the
study. Assessors guessed group allocations correctly 17% of the time,
which was less than the 33% predicted by chance alone; this suggests
that blinding was well preserved.
Baseline characteristics of the participants
At baseline, all groups were similar with respect to the anthro-
pometric and demographic characteristics, as well as to pulmonary
function, SpO
2
and the variables related to thoracic drainage (Table 1).
Effect of the intervention
Duration of thoracic drainage
The total drainage time ranged from 2 to 37 days: 2 to 19 days in
the Exp2 group, 2 to 37 days in the Exp1 group and 2 to 16 days in the
Con group. In the Exp2 group, drainage duration was shorter
compared with that in the other groups (Table 2). Over time, the
following approximate probability values for participants remaining
on thoracic drainage for 7 days were observed: 7% in the Exp2 group,
Table 1
Baseline anthropometric and demographic characteristics of the participants (n = 156).
Variables Exp2
(n = 52)
Exp1
(n = 52)
Con
(n = 52)
Gender, n male (%) 46 (88) 45 (87) 46 (88)
Age (yr), median (IQR) 32
(23 to 38)
27
(23 to 34)
27
(23 to 34)
BMI (kg/m
2
), median (IQR) 25
(22 to 28)
25
(22 to 28)
24
(21 to 27)
Smoking, n (%) 24 (46) 27 (52) 25 (48)
FVC (% pred), median (IQR) 61
(44 to 80)
57
(45 to 72)
64
(51 to 79)
FEV
1
(% pred), median (IQR) 41
(30 to 64)
46
(38 to 60)
46
(39 to 57)
SpO
2
(%), median (IQR) 97
(96 to 98)
97
(95 to 98)
97
(95 to 98)
Cause of pleural effusion, n (%)
trauma 48 (92) 50 (96) 48 (92)
pneumonia 2 (4) 1 (2) 3 (6)
neoplasia 2 (4) 1 (2) 1 (2)
Type of drainage used, n (%)
unilateral 50 (96) 50 (96) 50 (96)
bilateral 2 (4) 2 (4) 2 (4)
Pain scale (0 to 10), median (IQR) 6
(5 to 8)
7.5
(5 to 9)
7
(5 to 8)
BMI = body mass index, Con = control group (usual care plus sham positive pressure),
Exp1 = experimental group 1 (usual care plus mobilisation, respiratory interventions
and sham positive pressure), Exp2 = experimental group 2 (usual care plus
mobilisation, respiratory interventions and positive pressure), FEV
1
= forced
expiratory volume in 1 second, FVC = forced vital capacity, SpO
2
= peripheral oxygen
saturation, % pred = percentage predicted.
22 dos Santos et al: Positive pressure for pleural drainage
17% in the Exp1 group and 21% in the Con group; these probabilities
were maintained up to 16 days (Figure 2).
Length of hospital stay
The total length of hospital stay varied from 2 to 46 days, and it
was also shorter in the Exp2 group when compared with the other
groups (Table 2). Over time, the approximate probabilities for
participants remaining hospitalised for 7 days were as follows: 9% in
the Exp2 group, 23% in the Exp1 group and 31% in the Con group;
these probabilities were maintained up to 33 days (Figure 3).
Treatment costs
The estimates of costs ranged from R$1,000.00 to R$29,980.00 per
participant during hospital stay. Comparison between groups showed
that the Exp2 group presented a significantly lower cost compared
with the other groups (Table 2).
Pulmonary function and oximetry
These outcomes tended to improve in all groups from baseline until
Day 8 (Table 3). The interventions compared in this study did not have
clear differences in their effects on pulmonary function and oximetry.
Tolerability
No inability to tolerate the positive airway pressure was noted in
any participants, so there was 100% adherence during each
participant’s period of participation in the trial. Discomfort reported
by the patients was also similar between groups (Table 3).
Pulmonary complications
Pulmonary complications were less common in the Exp2 group
compared with the Con group (absolute risk reduction (ARR) 0.18,
95% CI 0.06 to 0.31) and compared with the Exp1 group (ARR 0.16,
95% CI 0.04 to 0.28). This effect seemed to arise mainly through a
reduction in pneumonia. Pneumonia was less common in the Exp2
group compared with the Con group (ARR 0.20, 95% CI 0.09 to 0.32)
and compared with the Exp1 group (ARR 0.16, 95% CI 0.05 to 0.28). No
difference was found in the rates of atelectasis and hypoxaemia
between groups (Table 4).
Use of antibiotics
The Exp2 group presented less need for antibiotic use compared
with the Con and Exp1 groups (Table 4). The ARRs were 0.35 (95% CI
0.17 to 0.51) and 0.25 (95% CI 0.08 to 0.41), respectively.
Adverse events
All groups showed similar rates of air leaks and aerophagia
(Table 4). Individual participant data are available in Table 5 on the
eAddenda.
Discussion
The results show that the combination of intermittent CPAP of 15
cmH
2
O and mobilisation and respiratory care reduces the duration of
chest drainage, the length of hospital stay, pulmonary complications,
use of antibiotics, and treatment costs. In addition, the use of CPAP of
15 cmH
2
O did not produce a higher rate of adverse events or lower
tolerability compared with CPAP of 4 cmH
2
O (without therapeutic
effect). Despite its other beneficial effects, the addition of CPAP had
no clear effect on lung function recovery and oxygenation.
This study also showed that CPAP of 15 cmH
2
Oaddedto
mobilisation and respiratory physiotherapy interventions (Exp2) was
more effective than mobilisation and respiratory physiotherapy alone
(Exp1). The benefits were observed on the same outcomes (duration of
chest drainage, the length of hospital stay, pulmonary complications,
use of antibiotics, and treatment costs) and had a similar magnitude as
in the comparison against the Con group. This suggests that the
intermittent CPAP was mainly driving the observed benefit.
It is important to gauge whether these benefits are large enough
to be clinically worthwhile. Even without well-established estimates
of the smallest worthwhile effect for these comparisons, it seems
reasonable to conclude that most of the estimates have confidence
intervals that include both trivial and worthwhile effects. For
example, the estimates of the reductions in complications generally
and in pneumonia specifically all have confidence intervals around
the ARRs that span from ,10% to around 30%. The reduction in risk of
requiring antibiotics is stronger (ARR 0.35, 95% CI 0.17 to 0.51), so this
is arguably a clinically worthwhile benefit in its own right. The
clinical utility of the estimated effect on treatment cost is more
difficult to interpret because the data were analysed with a test for
non-parametric data. However, the difference in medians of R$1,396
between the Exp2 and Con groups is reassuring, given the other
clinical benefits obtained.
Previous studies conducted on participants having open lung
resection have shown divergent results from those of this research.
Danner et al
29
evaluated 21 participants divided into two groups: one
group received non-invasive ventilation (NIV) with an inspiratory
positive airway pressure of 16 cmH
2
O, whereas the other group
underwent conventional respiratory physiotherapy without NIV. The
authors found no between-group difference for duration of chest tube
drainage. Similarly, Garutti et al
28
and Perrin et al
12
assessed the use
Table 2
Median (IQR) of each group, pairwise differences in ranks, and statistical significance of the pairwise comparisons.
Outcome Groups Difference in ranks
(pvalue)
Exp2
(n = 51)
Exp1
(n = 52)
Con
(n = 49)
Exp2 versus
Con
Exp2 versus
Exp1
Exp1 versus
Con
Duration of drainage
(d)
4
(3 to 4)
4
(3 to 6)
5
(3.5 to 7)
32.25
(,0.05)
23.59
(,0.05)
8.66
(.0.05)
Length of hospital stay
(d)
4
(3 to 4)
5
(3 to 7)
6
(4 to 9.5)
35.89
(,0.05)
24.36
(,0.05)
11.52
(.0.05)
Treatment cost
(R$)
2671.41
(2021.41 to 3321.41)
3933.96
(2680.00 to 6528.21)
4067.45
(2440.47 to 6537.69)
24.33
(,0.05)
32.64
(,0.05)
8.32
(.0.05)
Con = control group (usual care plus sham positive pressure), Exp1 = experimental group 1 (usual care plus mobilisation and respiratory interventions and sham positive pressure),
Exp2 = experimental group 2 (usual care plus mobilisation and respiratory interventions and sham positive pressure), R$ = Brazilian real.
eganiard no gniniamer fo ytilibaborP (%)
Time (days)
0 5 10 20 3025
100
80
60
40
0
20
15 35
Exp1
Con
Exp2
*
Figure 2. Kaplan-Meier curve for duration of thoracic drainage.
*p,0.001.
Research 23
of NIV with lower inspiratory positive airway pressures (7 to 12
cmH
2
O) for surgical patients. They also observed no difference in
duration of chest drainage between intervention and control
(conventional treatment) groups. Perhaps the difference detected in
the present trial occurred due to the combination of two factors: the
absence of surgical trauma and the use of CPAP at high pressure. We
believe that respiratory physiotherapy interventions, which included
airway clearance techniques and mobilisation, in addition to
pulmonary expansion help patients recover more quickly and stay
more active during their hospital stay. This increased activity along
with the use of CPAP can accelerate the elimination of excess pleural
fluid, with consequent faster resolution of pleural fluid collection and
shorter duration of chest drainage.
Reduced duration of chest drainage associated with a lower rate of
pneumonia in the Exp2 group presumably contributed to the 2-day
reduction in the length of hospital stay. A randomised trial
30
conducted with surgical patients also showed shorter length of
hospital stay for patients receiving CPAP of 10 cmH
2
O when
compared with conventional respiratory care. In contrast, a
systematic review showed that non-invasive ventilation did not
reduce the length of hospital stay in surgical patients with cancer.
15
This difference is most likely associated with the severity of clinical
conditions of cancer patients. The shorter length of hospital stay and
duration that antibiotics were used resulted in a reduction in the
estimated treatment costs.
In the Exp2 group, despite the need to purchase positive pressure
equipment and accessories and for physiotherapy assistance three
times a day, costs were reduced compared with those in the Con
group. This finding is consistent with evidence that adequate
application of positive airway pressure seems to present good
cost-effectiveness in some clinical situations, such as the
postoperative period of thoracic surgeries,
14
although further
evidence about cost-effectiveness is needed.
31
Although more robust
evidence is still required, the results of the current study indicate that
investing in positive pressure equipment and hiring physiotherapists
can actually reduce costs instead of increasing them.
The current results show that the addition of CPAP at 15 cmH
2
O
reduced the rate of pneumonia and, consequently, the use of antibiotics
in non-surgical patients with drained pleural effusions. Zarbock et al
32
evaluated 468 surgical patients randomly allocated into two groups;
patients who received CPAP of 10 cmH
2
Ohadalowerrateof
pneumonia compared with a no-intervention control group. A Cochrane
review
15
found no studies that have tested the efficacy of non-invasive
positive airway pressure in the prevention of complications after lung
resection and evaluated the use of antibiotics. One recent study of
patients undergoing myocardial revascularisation showed no
difference in antibiotic use between patients who received and those
who did not receive postoperative non-invasive ventilation.
33
The cur-
rent results revive this important aspect in the treatment of patients
with chest drainage and reinforce the need for further studies assessing
this outcome. It is believed that optimised pulmonary expansion
through positive pressure facilitates blood perfusion
34
with possible
transport of immune cells facilitating pulmonary defence.
22
Despite the effectiveness of the interventions provided to the Exp2
group in this study, patient comfort and adherence to treatment are
also important aspects to be considered. In this study, patients in all
groups reported similar levels of satisfaction and tolerable discomfort in
response to the use of CPAP of 4 and 15 cmH
2
O. Possibly, factors like the
use of a padded silicone oronasal mask as the interface and intermittent
application facilitated good tolerance even at the higher pressure of 15
cmH
2
O. In agreement with these results, Stéphan et al assessed
tolerability by means of a comfort score in 830 patients, with half of
them receiving non-invasive ventilation, and found that 17% reported
slight, 29% acceptable and 53% good comfort, with no difference
observed when compared with patients only using nasal highflow
oxygen.
35
In another study conducted on 66 patients undergoing non-
invasive ventilation, the rate of intolerability to treatment was 3.5%.
36
Regarding adverse events, the rates of air leaks and aerophagia
were similar in all groups in the current study. Liao et al
37
compared
surgical patients who received NIV of 10 cmH
2
O with patients who
underwent usual care, and also found similar rates of air leaks in both
groups: 17% versus 11%, respectively. In another trial of patients who
underwent lung resection and then random allocation of CPAP of 8.5
cmH
2
O, no occurrence of air leaks was observed.
36
Therefore, all of
latipsoh
n
igniniamerfoytilibaborP (%)
Time (days)
0 5 10 20 3025
100
80
60
40
0
20
Exp1
Con
15
Exp2
35
*
Figure 3. Kaplan-Meier curve for length of hospital stay.
*p,0.001.
Table 3
Change scores for pulmonary function and peripheral oxygen saturation, and final scores for discomfort and satisfaction. Median (IQR) for groups and the statistical significance of
the Kruskal-Wallis ANOVA.
Outcome Groups p
value
Exp2
(n = 51)
Exp1
(n = 51)
Con
(n = 49)
FEV
1
(% pred), median (IQR) 15
(3 to 28)
6
(22to17)
9
(22to17)
0.23
FVC (% pred), median (IQR) 14
(5 to 33)
22
(216 to 9)
26
(227 to 8)
0.34
SpO
2
(%), median (IQR) 0.0
(21.0 to 1.0)
0.0
(21.0 to 2.0)
1.0
b
(0.0 to 2.0)
0.77
Discomfort (0 to 10), median (IQR) 6.5
a
(3.8 to 8.0)
7.0
a
(4.0 to 8.0)
6.0
(4.0 to 7.5)
0.37
Satisfaction (0 to 10), median (IQR) 10.0
a
(10.0 to 10.0)
10.0
a
(9.0 to 10.0)
10.0
(9.0 to 10.0)
0.11
Con = control group (usual care plus sham positive pressure), Exp1 = experimental group 1 (usual care plus mobilisation and respiratory interventions and sham positive pressure),
Exp2 = experimental group 2 (usual care plus mobilisation and respiratory interventions and positive pressure), FEV
1
= forced expiratory volume in 1 second, FVC = forced vital
capacity, SpO
2
= peripheral oxygen saturation.
a
n=50
b
n=51
24 dos Santos et al: Positive pressure for pleural drainage
these results show that, unlike what many professionals fear in
clinical practice, this type of intervention is safe and well tolerated by
patients, even with a pressure of 15 cmH
2
O.
Despite the benefits observed in this study, the groups did not mark-
edly differ in pulmonary function recovery and oxygenation. Most likely,
the performance of patients during spirometry was influenced by the high
level of pain. Furthermore, SpO
2
was never outside the normal range.
The present study had some limitations. Most participants required
their intercostal drains because of thoracic trauma, so the results may
be more generalisable to haemorrhagic pleural effusions and
haemothoraces rather than other pleural effusions. However, the high
adherence of participants to the protocol and the good effect of adding
CPAP of 15 cmH
2
O to more traditional respiratory interventions may
indicate that the treatment can be effective in various types of pleural
fluid collection. Another limitation concerns the estimation of
treatment costs, which was not performed by individualised
cost-effectiveness analysis. The estimates were determined by the cost
of each procedure and based on a chart provided by the hospital. The
hospitalisation cost is estimated by the intervention costs added to the
length of hospital stay, but the amount paid to the hospital is the same
between 2 and 10 days of hospitalisation.
38
Data on the size of the
drainage tubes were not recorded but it is assumed that randomisation
would have produced similar distributions in the three groups.
In conclusion, the results of this study indicate that non-invasive
positive airway pressure of 15 cmH
2
O added to mobilisation and
respiratory care for patients with a collection of fluid in the pleural
space reduces the duration of chest drainage, length of hospital stay,
pulmonary complications, use of antibiotics and treatment costs. This
type of intervention showed good tolerability by the patients and a
low rate of adverse events; therefore, it can be safely integrated into
clinical practice.
What was already known on this topic: Fluid and/or blood
can accumulate in the pleural space due to a range of conditions.
An intercostal tube attached to a closed drainage system is often
used to drain the pleural space. The existing evidence is unclear
about whether positive airway pressure applied non-invasively at
the mouth assists resolution of the fluid collection.
What this study adds: In patients with a chest tube drainage
system in situ, bouts of continuous positive airway pressure via a
face mask combined with mobilisation and respiratory techniques
decreases the duration of thoracic drainage, length of hospital
stay, pulmonary complications, antibiotic use and treatment costs.
This intervention was well tolerated with few adverse events, so it
can be safely integrated into clinical practice.
Footnotes:
a
Müller, Engemed, Brazil.
b
Respiron, NCS, Mexico.
c
Shaker, NCS, Mexico.
eAddenda: Table 5 can be found online at https://doi.org/10.1016/j.
jphys.2019.11.006.
Ethics approval: This project was approved by the Research Ethics
Committee of the Universidade Cidade de São Paulo (process number
793.133). All participants gave written informed consent before data
collection began.
Competing interest: Funders had no role in the execution,
analysis, interpretation of data or decision to present the results in
this study and were only involved with funding. No other competing
interests are declared.
Sources of support: The present study was funded by the National
Council for Scientific and Technological Development (CNPq), project
number 442709/2014-5. Funding received from the aforementioned
agency was fundamental in the purchase of equipment to conduct the
interventions proposed in this study.
Acknowledgements: We greatly thank the participants and the
clinical staff who were involved in the project.
Provenance: Not invited. Peer reviewed.
Correspondence: Adriana Cláudia Lunardi, Master’s and Doctoral
Programs in Physical Therapy, Universidade Cidade de São Paulo, São
Paulo, Brazil. Email: adriana.lunardi@unicid.edu.br
References
1. Light RW. Pleural effusions. Med Clin North Am. 2011;95:1055–1070.
2. Light RW. Pleural diseases.5
th
ed. Philadelphia, USA: Lippincott Williams & Wilkins;
2007.
3. Walker SP, Morley AJ, Stadon L, De Fonseka D, Arnold DT, Medford AR, et al.
Nonmalignant pleural effusions: a prospective study of 356 consecutive unselected
patients. Chest. 2017;151:1099–1105 .
4. Liu F, Huang YC, Ng YB, Liang JH. Differentiate pleural effusion from hemothorax
after blunt chest trauma; comparison of computed tomography attenuation values.
J Acute Med. 2016;6:1–6.
5. DeBiasi EM, Pisani MA, Murphy TE, Araujo K, Kookoolis A, Argento AC, et al.
Mortality among patients with pleural effusion undergoing thoracentesis. Eur
Respir J. 2015;46:495–502.
6. Havelock T, Teoh R, Laws D, Gleeson F. Pleural procedures and thoracic ultrasound:
British Thoracic Society pleural disease guideline 2010. Thorax. 2010;65:i61–i76.
7. Valenza-Demet G, Valenza MC, Cabrera-Martos I, Torres-Sánchez I, Revelles-
Moyano F. The effects of a physiotherapy programme on patients with a pleural
effusion: a randomized controlled trial. Clin Rehabil. 2014;28:1087–1095.
8. New South Wales Agency for Clinical Innovation - Respiratory Network, Australia.
Consensus Guideline: Pleural Drains in Adults.https://www.aci.health.nsw.gov.au/
__data/assets/pdf_file/0018/201906/PleuralDrains_Guideline-021116.pdf. Accessed
5 September, 2019.
9. Elkins MR, Moseley AM. Intention-to-treat analysis. J Physiother. 2015;61:165–167.
10. Oliveira JF, Mello FCQ, Rodrigues RS, Boechat AL, Count MB, Menezes SL. Effect of
continuous positive airway pressure on fluid absorption among patients with
pleural effusion due to tuberculosis. Braz J Phys Ther. 2010;14:127–132.
Table 4
Number (%) of complications, lung entrapment, thoracotomy, antibiotic use and adverse events, and absolute risk reduction (95% CI) for pairwise comparisons.
Outcome Groups Absolute risk reduction (95% CI)
Exp2
(n = 51)
Exp1
(n = 51)
Con
(n = 49)
Exp2 relative to
Con
Exp2 relative to
Exp1
Exp1 relative to
Con
Complications, n (%) 1 (2) 9 (18) 10 (20) 0.18
(0.06 to 0.31)
0.16
(0.04 to 0.28)
0.02
(20.13 to 0.17)
pneumonia, n (%) 0 (0) 8 (16) 10 (20) 0.20
(0.09 to 0.32)
0.16
(0.05 to 0.28)
0.04
(20.11 to 0.19)
atelectasis, n (%) 1 (2) 2 (4) 0 (0) 20.02
(20.10 to 0.05)
0.02
(20.07 to 0.11)
20.04
(20.13 to 0.04)
hypoxaemia, n (%) 0 (0) 0 (0) 0 (0) 0.00
(20.07 to 0.07)
0.00
(20.07 to 0.07)
0.00
(20.07 to 0.07)
Lung entrapment, n (%) 3 (6) 7 (14) 8 (16) 0.10
(20.04 to 0.23)
0.08
(20.04 to 0.20)
0.02
(20.12 to 0.16)
Thoracotomy, n (%) 3 (6) 5 (10) 7 (14) 0.08
(20.04 to 0.20)
0.04
(20.08 to 0.16)
0.04
(20.09 to 0.17)
Antibiotic use, n (%) 9 (18) 22 (43) 27 (55) 0.35
(0.17 to 0.51)
0.25
(0.08 to 0.41)
0.10
(20.09 to 0.28)
Adverse events, n (%) 5 (10) 1 (2) 3 (6)
a
20.04
(20.16 to 0.08)
20.08
(20.19 to 0.22)
0.04
(20.05 to 0.14)
Con = control group (usual care plus sham positive pressure), Exp1 = experimental group 1 (usual care plus mobilisation and respiratory interventions and sham positive pressure),
Exp2 = experimental group 2 (usual care plus mobilisation and respiratory interventions and sham positive pressure).
a
n=50
Research 25
11. Chiumello D, Chevallard G, Gregoretti C. Non-invasive ventilation in postoperative
patients: a systematic review. Intensive Care Med. 2011;37:918–929.
12. Perrin C, Jullien V, Venissac N, Berthier F, Padovani B, Guillot F, et al. Prophylactic
use of noninvasive ventilation in patients undergoing lung resectional surgery.
Respir Med. 2007;101:1572–1578.
13. Auriant I, Jallot A, Herve P, Cerrina J, Le Roy Ladurie FR, Fournier JL, et al.
Noninvasive ventilation reduces mortality in acute respiratory failure following
lung resection. Am J Respir Crit Care Med. 2001;164:1231–1235.
14. Matte P, Jacquet L, Van Dyck M, Goenen M. Effects of conventional physiotherapy,
continuous positive airway pressure and non-invasive ventilatory support with
bilevel positive airway pressure after coronary artery bypass grafting. Acta
Anaesthesiol Scand. 2000;44:75–81.
15. Torres MF, Porfirio GJ, Carvalho AP, Riera R. Non-invasive positive pressure
ventilation for prevention of complications after pulmonary resection in lung
cancer patients. Cochrane Database Syst Rev. 2015;9:CD010355.
16. Schulz KF, Chalmers I, Hayes RJ, Altman DG. Empirical evidence of bias. Dimensions
of methodological quality associated with estimates of treatment effects in
controlled trials. JAMA. 1995;273:408–412.
17. Schettino GPP, Reis MAS, Galas F, Park M, Franca S, Okamoto V. III Consenso
brasileiro de ventilação mecânica: ventilação mecânica não invasiva com pressão
positiva. J Bras Pneumol. 2007;33:S92–S105.
18. Dos Santos EDC, Pontes Campos AE, do Carmo OF, Lunardi AC. Comparing high and
low levels of continuous positive airway pressure on lung aeration in patients with
pleural drainage: A feasibility study for a randomized controlled trial. Physiother
Res Int. 2019;24:e1753.
19. Paydar S, Ghahramani Z, Ghoddusi Johari H, Khezri S, Ziaeian B, Ghayyoumi MA,
et al. Tube thoracostomy (chest tube) removal in traumatic patients: what do we
know? What can we do? Bull Emerg Trauma. 2015;3:37–40.
20. Miller MR, Hankinson J, Brusasco V, Burgos F, Casaburi R, Coates A, et al.
Standardisation of spirometry. Eur Respir J. 2005;26:319–338.
21. Pereira CAC. Espirometria- Diretrizes para testes de função pulmonar. J Bras
Pneumol. 2002;28:S1–S82.
22. Duggan M, Kavanagh BP. Pulmonary atelectasis: a pathogenic perioperative entity.
Anesthesiology. 2005;102:838–854.
23. Evans AL, Gleeson FV. Radiology in pleural disease: state of the art. Respirology.
2004;9:300–312.
24. Sakai RL, Abrão GM, Ayres JF, Vianna PT, Carvalho LR, Castiglia YM. Prognostic
factors for perioperative pulmonary events among patients undergoing upper
abdominal surgery. Sao Paulo Med J. 2007;125:315–321.
25. Huggins JT, Doelken P, Sahn AS. The unexpandable lung. F1000 Med Rep. 2010;2:1–3.
26. Na MJ. Diagnostic tools of pleural effusion. Tuberc Respir Dis. 2014;76:199–210.
27. Lunardi AC, Cecconello I, Carvalho CR. Postoperative chest physical therapy
prevents respiratory complications in patients undergoing esophagectomy. Rev
Bras Fisioter. 2011;15:160–165.
28. Garutti I, Puente-Maestu L, Laso J, Sevilla R, Ferrando A, Frias I, et al. Comparison of
gas exchange after lung resection with a Boussignac CPAP or Venturi mask. Brit J
Anaesth. 2014;112:929–935.
29. Danner BC, Koerber W, Emmert A, Olgemoeller U, Doerge H, Quintel M, et al.
Non-invasive pressure support ventilation in major lung resection for high risk
patients: does it matter? Open J Thoracic Surg. 2012;2:63–71.
30. Kindgen-Milles D, Müller E, Buhl R, Böhner H, Ritter D, Sandmann W, et al.
Nasal-continuous positive airway pressure reduces pulmonary morbidity and
length of hospital stay following thoracoabdominal aortic surgery. Chest.
2005;128:821–828.
31. Cabrini L, Plumari VP, Nobile L, Olper L, Pasin L, Bocchino S, et al. Non-invasive
ventilation in cardiac surgery: a concise review. Heart Lung Vessel. 2013;5:137–141.
32. Zarbock A, Mueller E, Netzer S, Gabriel A, Feindt P, Kindgen-Milles D. Prophylactic
nasal continuous positive airway pressure following cardiac surgery protects from
postoperative pulmonary complications: a prospective, randomized, controlled
trial in 500 patients. Chest. 2009;135:1252–1259.
33. Westerdahl E, Urell C, Jonsson M, Bryngelsson L, Hedenström H, Emtner M. Deep
breathing exercises performed 2 months following cardiac surgery a randomized
controlled trial. J Cardiopulm Rehabil Prev. 2014;34:34–42.
34. Chen Y, Yeh M, Hu H, Lee CS, Li LF, Chen NH, et al. Effects of lung expansion therapy
on lung function in patients with prolonged mechanical ventilation. Can Respir J.
2016;ID 5624315:1–7.
35. Stéphan F, Barrucand B, Petit P, Rézaiguia-Delclaux S, Médard A, Delannoy B, et al.
High-flow nasal oxygen vs noninvasive positive airway pressure in hypoxemic
patients after cardiothoracic surgery a randomized clinical trial. JAMA.
2015;313:2331–2339.
36. Roceto LS, Galhardo FD, Saad IA, Toro IF. Continuous positive airway pressure
(CPAP) after lung resection: a randomized clinical trial. Sao Paulo Med J.
2014;132:41–47.
37. Liao G, Chen R, He J. Prophylactic use of noninvasive positive pressure ventilation
in postthoracic surgery patients: A prospective randomized control study. J Thorac
Dis. 2010;2:205–209.
38. Ministério da Saúde (Brazil). Sistema de Gerenciamento da Tabela de Procedimentos,
Medicamentos e OPM do SUS (SIGTAP) [Internet]. Brasília: Secretaria de Atenção à
Saúde. http://sigtap.datasus.gov.br/tabela-unificada/app/sec/inicio.jsp. Accessed 24
February, 2017.
26 dos Santos et al: Positive pressure for pleural drainage