Australian Critical Care 35 (2022) 527e534

Contents lists available at ScienceDirect

Australian Critical Care

journal homepage: www.elsevier.com/locate/aucc

Research paper

Inspiratory muscle training in intensive care unit patients: An

international cross-sectional survey of physiotherapist practice
Ellie Hearn, PT, Bachelor of Physiotherapy (Hons) a, b, Rik Gosselink, PT, PhD c,
Nicole Freene, PT, PhD a, Ianthe Boden, PT, PhD d, e, Margot Green, PT, BAppSci
(Physiotherapy) f, Bernie Bissett, PT, PhD a, g, *
a
Discipline of Physiotherapy, University of Canberra, Australia; b Physiotherapy Department, John Hunter Hospital, Newcastle, Australia; c Department of
Rehabilitation Sciences, KU Leuven, Belgium; d Physiotherapy Department, Launceston General Hospital, Launceston, Australia; e School of Medicine,
University of Tasmania, Australia; f Canberra Hospital and Health Services, Canberra, Australia; g Physiotherapy Department, Canberra Hospital, Australia

article information a b s t r a c t

Article history: Background: Inspiratory muscle training is safe and effective in reversing inspiratory muscle weakness
Received 27 April 2021 and improving outcomes in patients who have experienced prolonged mechanical ventilation in the
Received in revised form intensive care unit (ICU). The degree of worldwide implementation of inspiratory muscle training in such
3 August 2021
patients has not been investigated.
Accepted 7 August 2021
Objectives: The objectives of this study were to describe the current practice of inspiratory muscle
training by intensive care physiotherapists and investigate barriers to implementation in the intensive
Keywords:
care context and additionally to determine if any factors are associated with the use of inspiratory muscle
Physiotherapy (techniques)
Breathing exercises
training in patients in the ICU and identify preferred methods of future education.
Intensive care Method: Online cross-sectional surveys of intensive care physiotherapists were conducted using
Critical care voluntary sampling. Multivariate logistic regression analysis was used to identify factors associated with
Mechanical ventilation inspiratory muscle training use in patients in the ICU.
Results: Of 360 participants, 63% (95% confidence interval [CI] ¼ 58 to 68) reported using inspiratory
muscle training in patients in the ICU, with 69% (95% CI ¼ 63 to 75) using a threshold device. Only 64%
(95% CI ¼ 58 to 70) of participants who used inspiratory muscle training routinely assessed inspiratory
muscle strength. The most common barriers to implementing inspiratory muscle training sessions in
eligible patients were sedation and delirium. Participants were 4.8 times more likely to use inspiratory
muscle training in patients if they did not consider equipment a barrier and were 4.1 times more likely to
use inspiratory muscle training if they aware of the evidence for this training in these patients. For
education about inspiratory muscle training, 41% of participants preferred online training modules.
Conclusion: In this first study to describe international practice by intensive care therapists, 63% reported
using inspiratory muscle training. Improving access to equipment and enhancing knowledge of inspi-
ratory muscle training techniques could improve the translation of evidence into practice.
© 2021 Australian College of Critical Care Nurses Ltd. Published by Elsevier Ltd. All rights reserved.

1. Introduction least 24 h found inspiratory muscle weakness was almost twice as

prevalent as limb muscle weakness (63% and 34% respectively, with
Up to 84% of patients who are mechanically ventilated in the 21% having both).1 This weakness is detectable early during ICU
intensive care unit (ICU) develop inspiratory muscle weakness.1 A admission (i.e., within 18 h of commencement of mechanical
cross-sectional study of 76 patients mechanically ventilated for at ventilation)2 and is likely a direct consequence of mechanical
ventilation.3,4 Risk factors for inspiratory muscle weakness in me-
chanically ventilated patients include older age, sedation, infection,
systemic inflammation, and severe illness.3,5e8
* Corresponding author at: Discipline of Physiotherapy, University of Canberra Inspiratory muscle training (IMT) is a multidisciplinary strategy
Hospital, Office 8, Level 2, Guraguma Street, Bruce ACT 2617, Australia. Tel.: þ61 2
that can be used by ICU clinicians to reverse inspiratory muscle
6206 8657.
E-mail address: Bernie.Bissett@canberra.edu.au (B. Bissett). weakness and enhance patient recovery from prolonged

https://doi.org/10.1016/j.aucc.2021.08.002
1036-7314/© 2021 Australian College of Critical Care Nurses Ltd. Published by Elsevier Ltd. All rights reserved.
528 E. Hearn et al. / Australian Critical Care 35 (2022) 527e534

mechanical ventilation. This targeted training requires a multidis- offered. Participants were informed that consent could be with-
ciplinary approach, whereby medical and nursing colleagues work drawn by ceasing the survey at any time without consequence.
closely with physiotherapists to minimise sedation and optimise The online surveys comprised questions concerning participant
alertness such that patients can actively participate in focused characteristics, the use of IMT in the ICU, potential barriers to its
training of their breathing muscles as part of a comprehensive use, the methods of assessment of MIP, as well as details regarding
ventilator-weaning plan.9,10 IMT can be provided to mechanically adverse events. Participants were also asked about their knowledge
ventilated patients using either a mechanical (spring-loaded) or an of the available evidence pertaining to IMT in patients in the ICU.
electronic device that provides a titratable load on inspiration.10 Key aspects of the survey were modelled on the telephone survey
IMT improves inspiratory muscle strength,11e14 enhances quality conducted in France in 2013e2014.19 Additional questions were
of life (QOL),13 and may improve weaning outcomes in these included regarding IMT dosage, physiotherapist knowledge of ev-
patients.11,14e16 A 2015 systematic review found IMT improves idence for IMT, and preferred learning approaches for IMT
maximal inspiratory pressure (mean difference 7 cmH2O, 95% education.
confidence interval [CI] ¼ 5 to 9) and facilitates weaning success Data regarding participant characteristics and use of IMT were
(risk ratio [RR] ¼ 1.34, 95% CI ¼ 1.02 to 1.76) with potential re- analysed using descriptive statistics, presented as a proportion of
ductions in ICU length of stay.16 An updated systematic review in respondents. Incomplete responses (<100% questions completed)
2018 concluded IMT improves inspiratory muscle strength (mean were included where data were available. Free-text responses
increase 6 cmH2O, 95% CI ¼ 5 to 8 cmH2O) and may be associated regarding IMT protocols using threshold devices (repetitions/sets/
with a mean reduction in weaning time of 3.2 days (95% intensity) were categorised as being a strength training approach
CI ¼ 0.6e5.8 days).12 (four to five sets of six to 10 breaths per day, 5 days per week with
IMT is safe and feasible in patients undergoing mechanical the device set to the highest pressure setting that could be
ventilation,17 and it has been recommended that ICU physiothera- consistently opened during inspiration and progressed daily)14 or a
pists should use a strength-based training protocol with an IMT nonestrength-based training approach. Themes arising from free-
device that provides either a threshold or resistive load that can be text comment boxes were grouped by a single researcher and
accurately titrated.10 Physiotherapists can measure maximum verified by another researcher before analysis.
inspiratory pressure (MIP) as a baseline assessment to determine if To determine which factors were associated with IMT use in ICU
a patient is likely to benefit from IMT,18 and this MIP value can be patients, associations between all variables were assessed using
used to establish a starting pressure for threshold training.10 The Spearman's rho with significance level set at p < 0.05. Multivariate
current preferred strategy for measuring MIP in patients in the ICU logistic regression analysis with manual backward stepwise elim-
is not known and may be important for accurate IMT prescription. ination was used, commencing with all potential predictor vari-
Despite the evidence for improved outcomes in patients who ables to identify the most parsimonious model for predicting the
have experienced prolonged mechanical ventilation, IMT is re- use of IMT by ICU physiotherapists. The criterion of p < 0.10 was
ported to be infrequently used by ICU physiotherapists.19 In France used to determine which variables were retained. All quantitative
in 2015, only 5% of ICU physiotherapists used evidence-based statistics were conducted using IBM SPSS Statistics for Windows,
threshold IMT in clinical practice, while only 16% routinely Version 26.0. Armonk, NY: IBM Corp.
measured inspiratory muscle strength.19 Since 2015, new rando-
mised trials,11,13 systematic reviews,12,16 and clinical guidelines9,10 3. Results
have been published supporting the use of IMT in patients in the
ICU. It is unknown if IMT implementation by physiotherapists has 3.1. Participants
changed in this time, in France or elsewhere.
The primary objective of this study is to describe contemporary In total, 360 ICU physiotherapists completed the survey. Com-
IMT use by ICU physiotherapists worldwide, including barriers and plete data sets were obtained for 83% of survey responses (Fig. 1).
factors associated with IMT application. Secondary objectives Partial responses were included where data were available. An
include ascertaining physiotherapists’ familiarity with the current overview of participant characteristics can be found in Table 1, with
evidence for IMT in patients in the ICU and their preferred educa- a comprehensive description of individual countries included in
tion method to enhance the translation of evidence into practice. Table E1(supplementary file). In total, 63% (95% CI ¼ 58 to 68)
(n ¼ 225) of participants stated they used IMT in patients in the ICU.

2. Methods
3.2. IMT use by location of practice
An online cross-sectional survey was conducted between
November 1st and December 16th, 2019 (Australia and New Zea- IMT application varied between regions (Fig. 2). South America
land) and September 9th and October 7th, 2020 (international). The had the highest rate of IMT use amongst responders (82% [n ¼ 23])
minimum number of survey responses was not set for this study. followed by Asia (76% [n ¼ 55]). More detailed regional information
Ethical approval was obtained from the University of Canberra was available for Oceania only (54% [n ¼ 57] of responders using
Research Ethics Committee on 01/10/2019, project number HREC- IMT), where there was wide variability in usage by state and ter-
2170 (Australia and New Zealand) and on 09/04/2020, project ritory (Figure E2b, supplementary file).
number HREC-4440 (International).
The target population for this study was physiotherapists who 3.3. Inspiratory muscle training techniques
had treated patients in an ICU in the 6 months before completing
the survey. There were no exclusion criteria. Multiple physiother- Techniques used by physiotherapists to specifically strengthen
apists from a single centre were eligible to participate. Participants the inspiratory muscles in patients in the ICU varied. Of those who
were invited to complete the online survey using voluntary sam- reported using IMT, 69% (95% CI ¼ 63 to 75) (n ¼ 153) reported
pling via email networks and social media sharing. To facilitate using a threshold device (Fig. 3), with 79% (n ¼ 114) of these using a
snowball sampling,20 the email invited participants to share the spring-loaded (mechanical) IMT device and 40% (n ¼ 58) using an
link with colleagues. No incentives encouraging participation were electronic tapered flow resistive IMT device. Less than half of
 E. Hearn et al. / Australian Critical Care 35 (2022) 527e534 529

Total Australian/New Zealand Survey Total International Survey Responses

Responses (n=111) (n=409)

Not Eligible (n=56)

Not a currenta ICU
physiotherapist (n=30)
Working in Australia or New
Zealand (n=26)
Not Eligible (n=5)
Not a current ICU physiotherapist in
Australia or New Zealand (n=5)

Survey ceased prior to completion of

demographic questions (n=99)

Eligible Australian/New Zealand Survey Eligible International Survey Responses

Responses (n=106) (n=254)
Complete Responses (n=98) Complete Responses (n=202)
Incomplete Responses (n=8) Incomplete Responses (n=52)

Total Combined Eligible Survey Responses

(n=360)

Complete Responses (n=300)

Incomplete Responses (n=60)

Fig. 1. Flow of participants through study. aTreated patients in the ICU within the previous 6 months. ICU ¼ intensive care unit.

participants (44% [95% CI ¼ 36 to 52]) (n ¼ 65) who used a 82]) (n ¼ 169) and/or tracheostomised (74% [95% CI ¼ 68 to 80])
threshold IMT device used a strength-based protocol. (n ¼ 166) while half of the participants (50% [95% CI ¼ 43 to 57])
(n ¼ 112) applied the technique in intubated patients. More than
3.4. Assessment of MIP half of the participants (57% [95% CI ¼ 51 to 63]) (n ¼ 128) used IMT
in recently weaned patients. IMT was used for patients who were
MIP was assessed by 64% (95% CI ¼ 58 to 70) (n ¼ 145) of par- never mechanically ventilated by 24% (95% CI ¼ 18 to 30) (n ¼ 54) of
ticipants who used IMT in the ICU. Measurement of the negative participants. While most participants who used IMT in the ICU
inspiratory pressure on the ventilator was the most common stated that IMT was continued after ICU discharge (61% [95% CI ¼ 51
method used (48% [95% CI ¼ 39 to 57]) (n ¼ 54), followed by an to 71]) (n ¼ 87), this continuation was frequently described as
electronic IMT device (46% [95% CI ¼ 37 to 55]) (n ¼ 52) and therapist dependent and not completed routinely and/or was
electronic manometer (34% [95% CI ¼ 25 to 43]) (n ¼ 38). Occlusion poorly monitored.
of the inhalation valve >20 s was used as a method of MIP
assessment by 21% (95% CI ¼ 13 to 29) (n ¼ 25) of participants. This 3.6. Adverse events during IMT
occlusion method was used less commonly by ICU physiotherapists
in Oceania (3% [n ¼ 1]) compared with Europe (45% [n ¼ 13]). In total, 34% (95% CI ¼ 26 to 42) (n ¼ 50) of all participants who
In participants who did not measure MIP, 83% (95% CI ¼ 76 to 90) used threshold IMT devices reported experiencing adverse events
(n ¼ 66) stated a lack of access to equipment was a barrier. during IMT. There was a large difference in adverse event rates
Furthermore, nine participants (9% [95% CI ¼ 3 to 15]) cited a lack of between Oceania and other continents. In Oceania, five participants
knowledge or training as a reason for not measuring MIP, whilst (12% [95% CI ¼ 78 to 98]) who used threshold IMT devices reported
three participants (4% [95% CI ¼ 0 to 8]) did not believe measure- an adverse event during IMT. These included anxiety-related
ment of MIP was necessary in clinical practice. tachycardia, hypertension, and tachypnoea; desaturation; and
bradycardia which was ongoing and not isolated to during IMT. No
3.5. Patient selection for IMT details on severity for any event were provided.
In all other regions, adverse events were experienced by a
IMT was most commonly used for patients in the ICU who were greater proportion of ICU physiotherapists who used threshold IMT
difficult to wean from mechanical ventilation (76% [95% CI ¼ 70 to devices (45% [95% CI ¼ 35 to 55]) (n ¼ 45). Twenty-six participants
530 E. Hearn et al. / Australian Critical Care 35 (2022) 527e534

Table 1
Participant characteristics.

Characteristic All participants (n ¼ 360)a Participants who used IMT (n ¼ 225)b Participants who did
not use IMT (n ¼ 130)b

Age (y), n (%)

<21 1 (0) 1 (100) 0 (0)
21e30 118 (33) 35 (30) 82 (70)
31e40 159 (44) 70 (45) 86 (55)
41e50 57 (16) 39 (70) 17 (30)
51e60 19 (5) 13 (68) 6 (32)
>60 6 (2) 4 (67) 2 (33)
Maximum qualification, n (%)
Entry level 144 (40) 79 (55) 64 (45)
Grad certificate 16 (4) 13 (81) 3 (19)
Grad diploma 29 (8) 21 (72) 8 (28)
Masters 109 (30) 69 (64) 39 (36)
PhD 46 (13) 32 (74) 11 (26)
Otherc 16 (4) 11 (69) 5 (31)
ICU experience (y), n (%)
<1 37 (10) 13 (35) 24 (65)
1e5 116 (32) 42 (36) 73 (64)
6e10 87 (24) 33 (38) 53 (62)
>10 120 (33) 42 (36) 75 (64)
Type of ICUd, n (%)
Medical 171 (48) 110 (65) 60 (35)
Surgical 178 (49) 113 (64) 64 (36)
Cardiac 125 (35) 78 (63) 46 (37)
Neurosurgical 116 (32) 78 (67) 38 (33)
General 217 (60) 128 (60) 86 (40)
Respiratory 144 (40) 95 (66) 48 (34)
Adult 201 (56) 116 (58) 83 (42)
Paediatric 31 (9) 23 (74) 8 (26)
Neonatal 20 (6) 16 (80) 4 (20)
Number of ICU beds, n (%)
1e5 16 (4) 6 (40) 9 (60)
6e10 65 (18) 28 (43) 37 (57)
11e15 52 (14) 20 (39) 31 (61)
16e20 66 (18) 24 (38) 39 (62)
20e30 69 (19) 26 (38) 43 (62)
>30 92 (26) 26 (28) 66 (72)

Partial responses included where possible; hence, the sum of the number (n) of participants who use IMT and those who do not use IMT may not equal the number (n) of ‘all
participants’ in that row.
ICU ¼ intensive care unit; IMT ¼ inspiratory muscle training.
a
Percentages in this column reported as a proportion of the total sample.
b
Percentages in these columns reported as a percentage of those who did or did not use IMT.
c
Other responses: PhD candidate (n ¼ 3), doctor of physical therapy (n ¼ 2), masters modules (n ¼ 2), diploma specialist (n ¼ 1), cardiorespiratory specialist (n ¼ 1),
associate (n ¼ 1), doctor of physiotherapy (n ¼ 1), master of science (n ¼ 1), did not specify (n ¼ 4).
d
Multiple responses possible; hence, total greater than 100%.

Fig. 2. IMT use in patients in the ICU by location (continent). ICU ¼ intensive care unit; IMT ¼ inspiratory muscle training.
 E. Hearn et al. / Australian Critical Care 35 (2022) 527e534 531

Fig. 3. Techniques nominated by ICU physiotherapists as methods of IMT (n ¼ 223). ICU ¼ intensive care unit; IMT ¼ inspiratory muscle training. Multiple responses possible hence
total greater than 100% Other: Spontaneous breathing trials (n ¼ 2), adjusting pressure support on the ventilator (n ¼ 2), decreasing pressure support and encourage deep breathing
in intubated patients (n ¼ 1), electrostimulation (n ¼ 1), breath-stacking devices (n ¼ 1), decreasing pressure during continuous positive airway pressure (CPAP) (n ¼ 1), 15- to 20-s
occlusions with one-way inspiratory valve in children (n ¼ 1), upside down incentive spirometer with resisted expiration (n ¼ 1), Bubble PEP (n ¼ 1), forced expiratory technique
(n ¼ 1), did not specify (n ¼ 4).

reported changes in oxygen saturation; 10 participants reported CI ¼ 46 to 64]) (n ¼ 70), and being unaware of the technique (27%
tachypnoea; 11 reported dyspnoea; seven participants reported [95% CI ¼ 20 to 36]) (n ¼ 35) (Fig. E4, supplementary file).
cardiovascular changes; and two reported detachment of lines. The three most common barriers to implementing IMT in
Additionally, there was one report each of coughing, postural hy- eligible patients in the ICU were sedation (62% [95% CI ¼ 57 to 67])
potension, bradycardia in small children, and anxiety in adults, and (n ¼ 186), access to equipment (46% [95% CI ¼ 40 to 52]) (n ¼ 149),
five participants did not specify. Three participants (3%) described and competing demands within a limited treatment session (41%
events that required medical intervention: changes in blood pres- [95% CI ¼ 35 to 47]) (n ¼ 122), as shown in Table E2 (supplementary
sure requiring medication and atrial fibrillation requiring medical file).
intervention; a heart attack with ST elevation, low blood pressure
and tachycardia after training at high intensity (80% of MIP); and 3.8. Factors associated with IMT use in patients in ICU
desaturation requiring a bronchoscopy after 10 min when using the
technique to induce sputum in a patient with polytrauma. There were significant but only weak correlations between IMT
use and 18 factors including equipment [r ¼ 0.33], awareness of
evidence (not aware [r ¼ 0.23]), perceived relative importance of
3.7. Barriers to IMT use in patients in the ICU IMT [r ¼ 0.25], and location (see Table E3). No other factors had a
significant correlation with IMT use. These factors were used to
The three most common reasons for not using IMT in patients develop a model for predicting IMT use. In the final logistical
who had experienced prolonged mechanical ventilation were a lack regression model, seven factors were associated with the use of IMT
of equipment to facilitate training (65% [95% CI ¼ 57 to 73]) in patients in the ICU (Table 2). The Hosmer and Lemeshow test
(n ¼ 83), a lack of equipment to evaluate treatment (55% [95% demonstrates the model is of good fit to the data, and the Nagel-
kerke R Square test shows the model explains 42% of the variability
of the data.
Table 2 If participants did not consider lack of equipment as a barrier to
Logistic regression model for factors associated with use by ICU physiotherapists.
the use of IMT in the ICU, they were 4.8 (95% CI 2.67 to 8.74) times
Variable Odds ratioa 95% CI more likely to use IMT. Participants who reported that they were
Barriers to IMT use b aware of the evidence for IMT were 4.1 (95% CI 1.69 to 10.11) times
Equipment 4.831** 2.67 to 8.74 more likely to use IMT in the ICU. Participants from South America
Contraindications 0.251* 0.10 to 0.66 and China were much more likely to use IMT in the ICU than par-
Evidence for IMT in the ICUb e 'not aware' 4.138* 1.69 to 10.11
ticipants from Canada (Table 2).
IMT importanceb e 'just as' important 0.389* 0.21 to 0.71
Location (Continent)b e South America 0.290* 0.09 to 0.96
Location (Country)b e Canada 14.904* 1.79 to 124.24 3.8. Participant familiarity with evidence and perception of benefit
Location (Country)b e China 0.126** 0.03 to 0.50
of IMT
HosmereLemeshow goodness of fit (p) 0.933
Nagelkerke R Square 0.424
Participant knowledge of the evidence varied considerably
CI ¼ confidence interval; ICU ¼ intensive care unit; IMT ¼ inspiratory muscle
training.
(Table 3). In total, 39% (95% CI ¼ 32 to 46) (n ¼ 69) of participants
a
Odds ratio <1.0 greater odds of using IMT, odds ratio >1.0 less odds of using IMT. were familiar with the available evidence for IMT in patients in the
b
Reference yes, *p < 0.05, **p < 0.001. ICU. Participants who described themselves as ‘unaware’ of the
532 E. Hearn et al. / Australian Critical Care 35 (2022) 527e534

Table 3
Participant perceptions of evidence and benefits of IMT in patients in the ICU.

Participant perceptions All participants Participants who used Participants who did
(n ¼ 332) IMT (n ¼ 208) not use IMT (n ¼ 124)

Strength of the available evidence,a n (%)

Unaware 38 (11) 12 (32) 26 (68)
Evidence does not support IMT in the ICU 10 (3) 6 (60) 4 (40)
Weak 75 (23) 47 (63) 28 (37)
Moderate 164 (49) 109 (67) 55 (33)
Strong 45 (14) 34 (76) 11 (24)

Participant perceptions All participants Participants who used Participants who did
(n ¼ 300) IMT (n ¼ 186) not use IMT (n ¼ 114)

Which patients benefit most from IMT,b n (%)

MIP <30 cmH2O 96 (32) 64 (67) 32 (33)
MIP 30e60 cmH2O 108 (36) 80 (74) 28 (26)
MIP >60 cmH2O 6 (2) 2 (33) 4 (67)
Otherc 16 (5) 13 (81) 3 (19)
Unsure 74 (25) 27 (37) 47 (64)
IMT importance,d n (%)
Just as 188 (63) 134 (71) 54 (29)
Not as 88 (29) 34 (39) 54 (61)
More than 24 (8) 18 (75) 6 (25)

The number of total participants varies owing to incomplete surveys at different stages.
ICU ¼ intensive care unit; IMT ¼ inspiratory muscle training;MIP ¼ maximum inspiratory pressure.
a
Participant perceived strength of the available evidence for IMT in patients in the ICU.
b
Evidence regarding patients in the ICU who benefit most from IMT after prolonged ventilation, where 30e60 cmH2O is ideal.18.
c
Other responses: High work of breathing but preserved inspiratory muscle function (n ¼ 4); weak and moderate assuming low enough resistance to train weak patients
(n ¼ 3); weaning failure and any degree of muscle weakness (n ¼ 1); weaning failure and ‘timed inspiratory effort index’ (n ¼ 1); those dependent on mechanical ventilation
with tracheostomy (n ¼ 1); if patient is too weak, use ventilator and never IMT (n ¼ 1); those on noninvasive ventilation and those who are at the weaning stage (n ¼ 1);
impossible to judge and depends on each patient and each pathology (n ¼ 1); dependent on device as some devices are ineffective for very weak patients (n ¼ 1); patients
ventilated for more than 7 days (n ¼ 1); not specified (n ¼ 1).
d
Participant perception of the importance of IMT in ICU patients in comparison with ‘whole-body’ rehabilitation.

available evidence were less likely to use IMT in the ICU (Table 3). in the ICU, they were 4.1 times more likely to use IMT with these
Participants who believed that patients in the ICU with MIP of patients. The primary reason cited for not using IMT in patients in
30e60 cmH2O were most likely to benefit from IMT were more the ICU was a lack of equipment. Sedation, delirium, and access to
likely to use IMT (Table 3). Participants who perceived that IMT was equipment were reported as the most common barriers to training
just as important as whole-body rehabilitation were more likely to in patients deemed suitable.
implement IMT, while those who perceived it was ‘not as important Compared with the one previous study investigating IMT in
as whole-body rehabilitation’ were less likely to use IMT. patients in the ICU, physiotherapists appear more likely to use IMT
in their practice (63% vs 36%);19 however, as response rates in the
3.9. Considerations for future training current study cannot be described, this relative improvement
should be interpreted with caution. Techniques used for IMT varied
Almost half (41% [n ¼ 123]) of the participants stated an online in both studies; however, the use of a threshold device was greater
learning module was the preferred method for future training in in the present study (69% vs 5%).19 The rate of MIP assessment in
IMT for patients in the ICU. Face-to-face training was preferred by our study was four times that reported in the French study (64% vs
35% (n ¼ 106) of participants. Reading journal articles was the 16%);19 however, techniques applied to assess MIP varied between
preferred option for 14% (n ¼ 43) of participants, and 5% (n ¼ 14) of the two studies. Electronic IMT devices have only more recently
participants would rather complete a postgraduate certificate in been studied and were likely not available to clinicians at the time
IMT in patients in the ICU. the French survey was conducted.11,19 Other factors accounting for
the difference in IMT application between our study and the French
4. Discussion study include evolution of sedation practices in the ICU,21 publi-
cation of more evidence,12 and availability of clinical guidelines.9,10
The main finding of this study was that 63% (n ¼ 225) of ICU In the present study, participants cited a lack of equipment as a
physiotherapists responding to this survey used IMT in patients in reason for not using IMT more commonly than in the French survey
the ICU. Given the risk of response bias, this rate is likely to be an (65% vs 33%).19 Lack of equipment and resources to provide IMT
overestimate of the current international practice. IMT techniques should be challenged in the context of the known benefits of IMT
used varied, but two-thirds (69% [n ¼ 130]) of participants used a for these patients (including enhancing weaning success and
threshold device. Other techniques such as deep breathing, improving QOL).12,13 Spring-loaded threshold devices are evidence
controlled diaphragmatic breathing, and incentive spirometry were based,10,12 inexpensive, and readily available.22 It is also likely that
used as methods of IMT despite a lack of supporting all barriers have not been captured in this study, given the high risk
evidence.10,12 The majority (64% [n ¼ 145]) of those therapists who of response bias, particularly in regions with very low responses.
used IMT also assessed MIP in their patients, most commonly using Future studies should explore the barriers in these countries as
the negative inspiratory force function on the ventilator. This study local barriers, and enablers may be different.
also found participants were 4.8 times more likely to use IMT in Participant knowledge of the available evidence varied. One-
patients if they did not consider equipment a barrier to the use of third (36% [n ¼ 108]) of participants correctly stated that patients
IMT. If participants were aware of the evidence for IMT in patients with moderate inspiratory muscle weakness (MIP 30e60 cmH2O)
 E. Hearn et al. / Australian Critical Care 35 (2022) 527e534 533

are most likely to benefit from threshold-based IMT.18 Being ‘un- CRediT authorship contribution statement
aware’ of the evidence for IMT was also significantly associated
with a reduced likelihood of using IMT in patients in the ICU Ellie Hearn: Project administration, Conceptualisation, Meth-
(OR ¼ 4.1, 95% CI ¼ 1.69e10.1). As 27% (n ¼ 35) of participants were odology, Visualisation, Writing; Rik Gosselink: Conceptualisation,
unaware of the utility of IMT in patients recovering from prolonged Methodology, Writing; Nicole Freene: Formal Analysis, Visual-
mechanical ventilation, education about this technique could be isation, Writing; Ianthe Boden: Conceptualisation, Methodology,
enhanced both as part of university training and for experienced Writing; Margot Green: Conceptualisation, Methodology, Writing;
clinicians. This survey found participants would prefer online Bernie Bissett: Conceptualisation, Methodology, Writing, Editing,
learning modules as the method of future training for IMT. To our Supervision.
knowledge, an online learning training package, specific to IMT in
the ICU, is not yet available. Development of such a resource would Conflict of interest
be an important next step in translating evidence to practice more
widely. While respondents to this survey preferred online learning, The authors do not have any conflicts of interest to declare in
preferences of nonresponders may be important in the design of relation to this publication.
effective training in IMT, and medical and nursing staff should be
included owing to the inherent multidisciplinary approach needed
in an ICU context.10 Appendix A. Supplementary data
The relatively high rate of adverse events noted in this study
(34% [n ¼ 50]) was somewhat surprising and in contrast to the low Supplementary data to this article can be found online at
rate of adverse events reported in other studies of rehabilitation https://doi.org/10.1016/j.aucc.2021.08.002.
practice in the ICU in Australia (e.g., 1%)23,24 including inpatients on
vasopressor therapy (7%).25 The contrast between the adverse References
event rate in Oceania (12% [n ¼ 5]) and other regions (45% [n ¼ 45])
is intriguing, but it is not clear whether this is a true difference, a [1] Dres M, Dube  B-P, Mayaux J, et al. Coexistence and impact of limb muscle and
diaphragm weakness at time of liberation from mechanical ventilation in
function of recall bias, or related to differences in defining an medical intensive care unit patients. Am J Respir Crit Care Med 2016;195(1):
‘adverse event’. It could be argued that issues such as transient 57e66.
dyspnoea or anxiety-related tachycardia should not be classified as [2] Levine S, Nguyen T, Taylor N, et al. Rapid disuse atrophy of diaphragm fibers in
mechanically ventilated humans. N Engl J Med 2008;358(13):1327e35.
‘adverse events’ because these are unlikely to result in any long-
[3] Demoule A, Jung B, Prodanovic H, et al. Diaphragm dysfunction on admission
term problems for the patient and could be in fact a common to the intensive care unit. Prevalence, risk factors, and prognostic impact-A
experience of high-intensity strength training. Only 3% (n ¼ 3) of prospective study. Am J Respir Crit Care Med 2013;188(2):213e9.
participants reported adverse events necessitating medical inter- [4] Goligher EC, Fan E, Herridge MS, et al. Evolution of diaphragm thickness
during mechanical ventilation. Impact of inspiratory Effort. Am J Respir Crit
vention (e.g., removal of lines, bronchoscopy), and this is much Care Med 2015;192(9):1080e8.
closer to the adverse event rate described in other rehabilitation [5] Dres M, Dube BP, Mayaux J, et al. Coexistence and impact of limb muscle and
activities in the ICU.26 diaphragm weakness at time of liberation from mechanical ventilation in
medical intensive care unit patients. Am J Respir Crit Care Med 2017;195(1):
This was the first attempt to describe IMT use by ICU physio- 57e66.
therapists around the world. The major limitation of this study [6] Dres M, Jung B, Molinari N, et al. Respective contribution of intensive care
was the limited response rates from some countries and the sus- unit-acquired limb muscle and severe diaphragm weakness on weaning
outcome and mortality: a post hoc analysis of two cohorts. Crit Care
ceptibility to response bias. Physiotherapists with a specific in- 2019;23(1):370.
terest in IMT may have had greater motivation to participate, [7] Goligher EC, Brochard LJ, Reid WD, et al. Diaphragmatic myotrauma: a
potentially leading to selection bias and overreporting of IMT use. mediator of prolonged ventilation and poor patient outcomes in acute res-
piratory failure. Lancet Respir Med 2019;7(1):90e8.
Selection bias was minimised using a standardised email and
[8] Goligher EC, Dres M, Fan E, et al. Mechanical ventilation-induced diaphragm
Twitter post for survey distribution. Furthermore, there may have atrophy strongly impacts clinical outcomes. Am J Respir Crit Care Med
been a low rate of participation from North American therapists 2018;197(2):204e13.
[9] Bissett B, Leditschke IA, Green M, et al. Inspiratory muscle training for
who do not identify with the title ‘physiotherapist’. Future studies
intensive care patients: a multidisciplinary practical guide for clinicians. Aust
should consider broadening the term to include ‘respiratory Crit Care 2018;32(3):249e55.
therapists’ to better capture practices around the world and [10] Bissett B, Gosselink R, van Haren FMP. Respiratory muscle rehabilitation in
should explore the understanding of medical and nursing staff patients with prolonged mechanical ventilation: a targeted approach. Crit
Care 2020;24(1):103.
with respect to facilitating IMT in patients recovering from me- [11] Tonella RM, Ratti LDSR, Delazari LEB, et al. Inspiratory muscle training in the
chanical ventilation. intensive care unit: a new perspective. J Clin Med Res 2017;9(11):929e34.
[12] Vorona S, Sabatini U, Al-Maqbali S, et al. Inspiratory muscle rehabilitation in
critically ill adults. A systematic review and meta-analysis. Ann Am Thorac Soc
2018;15(6):735e44.
5. Conclusion [13] Bissett BM, Leditschke IA, Neeman T, et al. Inspiratory muscle training to
enhance recovery from mechanical ventilation: a randomised trial. Thorax
2016;71(9):812e9.
In this survey of ICU physiotherapists, 63% (n ¼ 225) used IMT in [14] Martin AD, Smith BK, Davenport PD, et al. Inspiratory muscle strength training
patients in the ICU and the majority of participants applied improves weaning outcome in failure to wean patients: a randomized trial.
evidence-based techniques and had at least some knowledge of the Crit Care 2011;15(2):R84.
[15] Cader SA, Vale RGdS, Castro JC, et al. Inspiratory muscle training improves
available evidence. These results should be interpreted with maximal inspiratory pressure and may assist weaning in older intubated
caution owing to response bias and are likely to represent an patients: a randomised trial. J Physiother 2010;56(3):171e7.
overestimate of the current international practice. This study [16] Elkins M, Dentice R. Inspiratory muscle training facilitates weaning from
mechanical ventilation among patients in the intensive care unit: a systematic
identified the need to improve access to equipment and the need to
review. J Physiother 2015;61(3):125e34.
minimise sedation and delirium in the ICU. Overcoming these [17] Bissett B, Leditschke IA, Green M. Specific inspiratory muscle training is safe in
barriers is a multidisciplinary challenge, but close collaboration selected patients who are ventilator-dependent: a case series. Intensive Crit
between physiotherapists, medical staff, and nursing staff to opti- Care Nurs 2012;28(2):98e104.
[18] Bissett B, Wang J, Neeman T, et al. Which ICU patients benefit most from
mise training conditions could improve the recovery of critically ill inspiratory muscle training? Retrospective analysis of a randomized trial.
patients. Physiother Theory Pract 2019;9:1e6.
534 E. Hearn et al. / Australian Critical Care 35 (2022) 527e534

[19] Bonnevie T, Villiot-Danger J, Gravier F, et al. Inspiratory muscle training is [23] Leditschke IA, Green M, Irvine J, et al. What are the barriers to mobilizing
used in some intensive care units, but many training methods have uncertain intensive care patients? Cardiopulm Phys Ther J 2012;23(1):26e9.
efficacy: a survey of French physiotherapists. J Physiother 2015;61(4):204e9. [24] Berney SC, Harrold M, Webb SA, et al. Intensive care unit mobility practices in
[20] Frey B. Snowball sampling. The SAGE Encyclopedia of Educational Research, Australia and New Zealand: a point prevalence study. Crit Care Resusc
Measurement, and Evaluation; 2018. p. 1e4. 2013;15(4):260e5.
[21] Shehabi Y, Bellomo R, Mehta S, et al. Intensive care sedation: the past, present [25] Rebel A, Marzano V, Green M, et al. Mobilisation is feasible in intensive care
and the future. Crit Care 2013;17(3):322. patients receiving vasoactive therapy: an observational study. Aust Crit Care
[22] Phillips. Threshold IMT: Koninklijke Phillips. 2020. Available from: https:// 2019;32(2):139e46.
www.philips.com.au/healthcare/product/HCHS730010/treshold-inspiratory- [26] Nydahl P, Sricharoenchai T, Chandra S, et al. Safety of patient mobilization and
muscle-trainer [Accessed 12 August 2020]. rehabilitation in the intensive care unit. Systematic review with meta-anal-
ysis. Ann Am Thorac Soc 2017;14(5):766e77.