Accepted Manuscript Title: The oxygen cost of rehabilitation interventions in mechanically ventilated patients Author: Claire Black Micheal Grocott Mervyn Singer PII: DOI: Reference:
S0031-9406(18)30301-8 https://doi.org/doi:10.1016/j.physio.2019.06.008 PHYST 1117
To appear in:
Physiotherapy
Received date: Revised date: Accepted date:
22 October 2018 11 March 2019 21 June 2019
Please cite this article as: Claire Black, Micheal Grocott, Mervyn Singer, The oxygen cost of rehabilitation interventions in mechanically ventilated patients, (2019), https://doi.org/10.1016/j.physio.2019.06.008 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
cr
ip t
The oxygen cost of rehabilitation interventions in mechanically ventilated patientsI
us
Abstract
Objective: Early rehabilitation is assumed to be a crucial intervention to
an
facilitate weaning from mechanical ventilation in critically ill patients and to limit their long-term functional dependence. However, little is
M
known about the physiological load imposed on patients during such interventions. Without the ability to quantify the exercise intensity of re-
d
habilitation interventions it is impossible to establish a clear separation
te
between usual care and intervention groups in randomised controlled trials. This may explain the lack of definitive benefit of rehabilitation
Ac ce p
in published trials. We sought to characterise the physiological load, ˙ 2 ), of the physical activities carmeasured as oxygen consumption (VO
ried out during rehabilitation interventions in mechanically ventilated participants.
Design: Observational Study Setting: Single centre medical-surgical university hospital ICU.
Participants: 26 mechanically ventilated participants ventilated >7 days, able to participate in a rehabilitation program. Intervention: Oxygen consumption (measured by the Medgraphics Ultima Preprint submitted to Physiotherapy
July 1, 2019
Page 1 of 19
breath-by-breath gas exchange analysis system) and heart rate were
ip t
measured continuously pre-, during and post-standard rehabilitation sessions.
cr
Results: 52 sessions were recorded in 26 participants. There was considerable variation in the oxygen cost of the physical activities between
us
participants. The recovery time for 1 in 4 rehabilitation sessions was longer than the rehabilitation activity time.
an
Conclusions: Absolute exercise intensity in mechanically ventilated ICU
M
participants, as measured by oxygen consumption, is not activity-dependent. Contribution of paper
We present novel data regarding the oxygen cost of rehabilitation inter-
te
following:
d
ventions in mechanically ventilated participants in ICU and demonstrate the
Ac ce p
• the oxygen cost of rehabilitation activities varies between participants, • oxygen consumed during rehabilitation interventions may be influenced by factors such as the participants active contribution to the activity.
Keywords: Rehabilitation, Exercise, Oxygen consumption, Mechanical ventilation, Critical illness,
1
1. Introduction
2
Patients who survive critical illness have a severely reduced exercise ca-
3
pacity [1, 2]. However, the metabolic cost of individual rehabilitation ac-
4
tivities in mechanically ventilated patients is, as yet unknown. The current 2
Page 2 of 19
assumption is that the metabolic cost of an individual rehabilitation activity,
6
such as siting over the edge of the bed (SOEB) is the same for each patient
7
and similar to that of a healthy individual.
ip t
5
cr
8
A failure to appropriately tailor individualised exercise programs may
10
lead to under-training of some patients, delaying their recovery by not al-
11
lowing them to reach their full rehabilitation potential, while over-training
12
others, potentially placing them under considerable physiological stress.
an
us
9
13
The available evidence suggests that active rehabilitation interventions
15
[3],[4], increase a patients energy expenditure beyond their resting metabolic
16
rate. However the intensity at which an intervention becomes a training load
17
will be dictated by the patients duration of exposure to both bed rest and
18
sepsis, along with their incumbent level of physiological fitness before they
19
arrive in ICU. Without the ability to quantify the exercise intensity of re-
20
habilitation interventions [5] it is impossible to establish a clear separation
26
Ac ce p
te
d
M
14
27
tima (MGU) (St Paul, Minneapolis, MN, USA) device in mechanically venti-
28
lated (MV) ICU patients [10] and demonstrated the feasibility of measuring
29
˙ 2 during rehabilitation interventions [11]. Here we report a pilot study to VO
21
22
23
24
25
between usual care and intervention groups in randomised controlled trials. This may be why we have yet to see evidence of a definitive benefit of rehabilitation in published trials.[6, 7, 8, 9]
The gold standard for measuring energy expenditure is breath-by-breath
gas exchange analysis (BBGEA). We recently validated the Medgraphics Ul-
3
Page 3 of 19
investigate the oxygen cost of rehabilitation activities in mechanically venti-
31
lated patients recovering from critical illness.
ip t
30
32
The primary objective of this study was to describe rehabilitation ac-
34
tivities in mechanically ventilated ICU patients in terms of duration and
35
˙ 2. intensity, as given by VO
36
2. Methods and materials
an
us
cr
33
Ethical approval was granted for the study by the UK National Research
38
Ethics Service (REC reference number 11/LO/1646). This single-centre ob-
39
servational study was conducted between August 2012 and March 2015 in
40
a 26-bed medical surgical ICU in London, UK. Participants were included
41
if they were aged 18 years or over, had been receiving mechanical ventila-
42
tion for over >7 days, had a tracheostomy, and they or their representative
43
gave informed consent/surrogate approval. Exclusion criteria were an in-
44
ability to exercise due to pre-existing conditions or current morbidity e.g.
46
47
48
49
d
te
Ac ce p
45
M
37
severe dementia, motor neurone disease, severe stroke, severe critical illness neuromyopathy, moderate-to- severe stenotic valvular heart disease, primary pulmonary hypertension, hypertrophic cardiomyopathy, or unstable angina. 2.1. Participant methodology Before each session, participants were assessed for suitability to exercise
50
by the ICU physiotherapist, as per normal practice at University College
51
Hospital (Supplementary material: Screening participants prior to exercise).
52
Rehabilitation sessions consisted of sitting the participant over the edge of
53
the bed (SOEB), SOEB and balance activities, then progressing them, as 4
Page 4 of 19
able, to standing, transferring and walking. Early termination of exercise
55
was based upon criteria described in the supplementary material (Indication
56
for termination of rehabilitation sessions). The participant was encouraged
57
to achieve their maximum functional level of activity during each rehabilita-
58
tion session.
cr
ip t
54
us
59
The MGU flow sensor was calibrated using the manufacturers instruc-
61
tions and placed in the participants ventilator circuit 30 minutes prior to
62
the rehabilitation session. Individual participant data were recorded during
63
rehabilitation sessions up to 3 times a week until they were liberated from
64
mechanical ventilation. The variables, recorded by the Breeze Suite soft-
65
˙ 2 ware supplied with the MGU, were: time from the beginning of the VO
66
˙ 2 (mL.min-1 ), carbon dioxide production (mL.min-1 ), resrecording (sec), VO
67
piratory rate (bpm), minute ventilation (l.min-1 ) and respiratory exchange
68
ratio (RER). Additionally, we recorded the time at which each rehabilita-
69
tion activity occurred to allow retrospective analysis of individual activities.
71
72
73
74
M
d
te
Ac ce p
70
an
60
Blood pressure was recorded from either an indwelling arterial cannula or a non-invasive cuff, and ECG monitoring continued throughout. Resting (HRrest ) and peak (HRpeak ) heart rate during exercise were recorded from the participants bedside monitor. 2.2. Breath by breath gas exchange analysis
75
The Medgraphics Ultima (MGU) BBGEA device provides a direct mea-
76
surement of oxygen consumption during exercise. It measures both inspira-
77
tory and expiratory flow through a bi-directional, patented flow-sensor. The
78
MGU device is widely used for exercise testing in spontaneously breathing 5
Page 5 of 19
individuals and, more recently, in mechanically ventilated participants [10].
80
2.3. Assessments of functional status
ip t
79
The ICU Functional Status Score (ICU-FSS) was measured in all partic-
82
ipants at each testing session [12]. The FSS-ICU examines the participant’s
83
ability to perform the five functional tasks: rolling, transfer from supine to
84
sit, SOEB, transfer from sit to stand, and walking. Each task is evaluated
85
using an eight-point ordinal scale ranging from 0 (unable to perform) to
86
7 (complete independence). The total score ranges from 0-35, with higher
87
scores indicating better physical functioning. The minimally minimum clin-
88
ically important difference is reported to be 2 - 5 [13].
M
an
us
cr
81
89
Pre-admission activity levels were estimated from a General Practice
91
Physical Activity Questionnaire (GPPAQ) score [14]. This information was
92
obtained from either the participant and/or relatives. The GPPAQ is a val-
93
idated screening tool for use in primary care to assess adult (16 - 74 years)
95
96
97
98
99
100
te
Ac ce p
94
d
90
physical activity levels. It provides a simple, 4-level Physical Activity Index categorising participants as active, moderately active, moderately inactive, or inactive. Participants were asked to report their perceived exertion at the beginning and end of the rehabilitation sessions using the Borg rate of perceived exertion scale, a self-reported tool with a category scale ranging from 6 to 20 points [15]. 3. Analysis
101
The total session VO2 (mL.kg-1 ) was estimated by calculating the area
102
˙ 2 (mL.kg.-1 min-1 ) curves bound by the start of the under the individual VO 6
Page 6 of 19
rehabilitation session, defined as the time the participant initially began to
104
˙ 2 returned to within 10% of the baseline resting value SOEB and the time VO
105
before exercise (Figure 1). Total session VO2 (mL.kg-1 ) was thus calculated
106
˙ 2 as the product of the duration of the session (minutes) and the mean VO
107
(mL.min-1 ) of the session.
cr
ip t
103
us
108
˙ 2 (mL.min-1 ) was calculated as the mean VO ˙ 2 (mL.min-1 ) durResting VO
110
˙ 2 at rest ing 10 consecutive minutes where there was a <10% variation in VO
111
within the 30 minutes prior to the rehabilitation session. The VO2 (mL.kg-1 )
112
attributable to the rehabilitation session itself, i.e. the VO2 (mL.kg-1 ) con-
113
sumed above resting VO2 , was calculated by subtracting the equivalent VO2
114
(mL.kg-1 ) at rest from the total session VO2 (mL.kg-1 ) (Figure 1). The VO2
115
values attributable to the rehabilitation session are presented as a percent-
116
age of the resting VO2 . To draw comparison between the percentage change
117
in VO2 between different rehabilitation activities, the activities were cate-
118
gorised as either sitting or standing.
124
Ac ce p
te
d
M
an
109
125
testing, increase in pressure support during the session (PSinc ), Haemoglobin
126
(Hb), Sequential Organ Failure Score (SOFA), GPPAQ, weight, gender, age.
127
A step down approach was used to remove non-significant variables from the
119
120
121
122
123
For hypothesis generating purposes, multilevel univariate regression [16]
was performed with percentage change in VO2 as the dependent variable. A beyond optimal model was created. The explanatory variables used were; rehabilitation activity; categorised as either sitting (SOEB and SOEB ± balance exercises) or standing (STS, MOS or BCT), ICU-FSS on the day of
7
Page 7 of 19
model and a restricted maximum likelihood estimation used to estimated
129
the variance of the model. The interaction between ICU-FSS and rehabilita-
130
tion activity was assessed using ANOVA of the log-likelihoods of the relevant
131
models. This was to test if the ICU-FSS was influencing the rehabilitation
132
activity i.e the higher the ICU FSS the more likely it would be for the par-
133
ticipant to carry out a higher level of activity. The residuals of the final
134
model were then plotted and examined for heterogeneity. This method of
135
regression was used to control for the within-participant nature of the reha-
136
bilitation sessions, i.e allowing for correlation between multiple observations
137
from the same participant.
138
4. Results
139
4.1. Recruitment
d
M
an
us
cr
ip t
128
Forty-four participants were recruited between August 2012 and March
141
2015, performing 125 tests in total. No data were collected for two partici-
142
pants as one weaned from mechanical ventilation before the first test, while
144
145
146
147
Ac ce p
143
te
140
the other became cardiovascularly unstable before the first test and remained so thereafter. Figure 2 shows the participant recruitment diagram and Table 1 participant characteristics. These participants had been mechanically ventilated for a median 22.5 (range 4 -103) days before their first rehabilitation session and mechanically ventilated for a median of 30 days (range
148
11 -103) before their first evaluated rehabilitation session within the study.
149
The median highest SOFA score was 10 (range 2 -14). Using the GPPAQ
150
classification of participant activity prior to ICU admission, obtained either
151
from the participant or a surrogate, 11 participants were classed as inactive, 8
Page 8 of 19
Median
Min
Max
Age (years)
71
31
85
Weight (kg)
70
MV prior to recruitment
27
cr
n = 26
35
114
11
103
us
Variable
ip t
17 as relatively inactive, 12 relatively active and 4 active.
Highest SOFA
2
14
2
1
4
an
GPPAQ
9
MV = mechanical ventilation
GPPAQ = General practice physical assessment questionnaire
M
SOFA = Sepsis Organ Failure Score
Table 1: Characteristics of the 26 participants
d
152
Table 2 shows the characteristics of the participants during the 52 recorded
154
rehabilitation sessions with respect to body temperature, haemoglobin and
155
ICU-FSS. During six tests, three participants were receiving total parenteral
157
158
159
160
Ac ce p
156
te
153
nutrition with the remainder fed enterally via a nasogastric tube. One participant was receiving renal replacement therapy during 4 tests. 4.2. Rehabilitation sessions A total of 104 rehabilitation sessions involving rehabilitation activities to
at least SOEB were assessed in 42 participants. Predominantly related to
161
technical issues (listed in Figure 2), the percentage change in VO2 from rest
162
could only be calculated in 52 of the 104 tests. These comprised 15 SOEB
163
episodes, 11 SOEB and balance work, 7 single sit-to-stand, 10 sit-to-stand
164
>1, and 9 bed-chair transfers. The minimum rehabilitation session duration 9
Page 9 of 19
Min
Max
C Reactive Protein (g.L-1 )
53
2
319
Haemoglobin (mg.L-1 )
8
7
11
SOFA
3
1
10
Temperature (C)
37
35.7
38
White Blood Cell Count (109 L-1 )
11
4
26
3
28
ICU-FSS total
12
an
SOFA = Sepsis Organ Failure Score
ip t
Median
cr
n = 52
us
Variable
ICU-FSS total = ICU Functional Status Score total
M
Table 2: Characteristics of the participants for the 52 recorded rehabilitation sessions.
was 5 minutes and 35 seconds to transfer from bed to chair. The maximum
166
duration was 39 minutes and 21 seconds to SOEB and to stand one or more
167
times (Table 3). Session duration was the total time from when the partici-
168
pant started to move, to when VO2 had recovered to within 10% of baseline.
169
Rehabilitation activity duration was the total time from when the partic-
171
172
173
174
175
te
Ac ce p
170
d
165
ipant started to move to returning to supine. Recovery proportion is the proportion of the whole rehabilitation session that was taken with recovery to within 10% of baseline VO2 . For ten rehabilitation sessions participant recovery time was greater than the actual rehabilitation activity duration.
Box plots of the rehabilitation categories (sitting or standing) and the
176
˙ 2 (mL.kg.-1 min-1 ) of the session, the total VO2 of the session (mL.kg-1 ), mean VO
177
and the percentage change in VO2 of the session are shown in Figure 3. The
178
mean (SD) percentage increase in VO2 from rest to SOEB ± balance activi10
Page 10 of 19
SOEB
Session Duration
Rehabilitation Duration Recovery Proportion
Min Median Max Min Median 15 06:11 15:00 36:36 04:34 08:46
Max
Min Median
Max
23:28
0.20
0.51
0.69
ip t
Activity
n
cr
Rehabilitation
17:38
0.18
0.34
0.59
STS x1
7 10:55 18:32 26:19 05:06 10:17
16:11
0.29
0.37
0.72
STS > 1
10 07:44 22:04 39:21 05:53 15:13
25:10
0.22
0.31
0.47
MOS or BCT
9 05:35 12:07 28:04 02:52 07:43
28:04
0.00
0.35
0.49
an
us
SOEB & balance 11 13:19 21:04 31:45 06:51 13:38
SOEB = sit over the edge of bed, STS = sit to stand, MOS = march on spot, BCT = bed chair transfer.
Table 3: Session duration = time from initiation of movement to VO2 recovering to within
M
10% of baseline, Rehabilitation duration = time from initiation of movement to returning to supine and Recovery proportion = the proportion of the whole rehabilitation session
d
taken with recovery.
ties was 23.3 (11.2). The mean percentage increase in VO2 from rest to stand-
180
ing and/or transferring from bed to chair was 34.8 (13.3). The interaction
181
i.e the way ICU-FSS influenced rehabilitation category was non-significant
183
184
185
186
187
Ac ce p
182
te
179
(p = 0.31) and not included in the final model (Supplementary Table 1). When residuals were plotted (Supplementary Figure 1), one participant was identified as a significant outlier. ˙ 2 (and VO2 ) The final selected model of the percentage change in VO
included ICU-FSS, the pressure support increment and the rehabilitation ac-
188
tivity. Therefore, a participant with an ICU-FSS of 0 who could SOEB with
189
a pressure support increment of 0 would expect an 12% (95% CI: 4-20) in-
190
˙ 2 . For every point increase in ICU-FSS, a 1% (95% CI: 1-2) crease in their VO 11
Page 11 of 19
˙ 2 could be expected. For each cm H2 O increase in pressure increase in VO
192
˙ 2 could be expected. support, a 1.5% (95% CI: 0-3) increase in VO
ip t
191
193
The calculated pseudo R2 for this model was 0.63. This suggests that
195
activity, functional status and pressure support increment explain up to 63%
196
˙ 2 during a rehabilitation session in the ICU. of the change seen in VO
197
5. Discussion
an
us
cr
194
This study aimed to describe rehabilitation activities in mechanically ven-
199
˙ 2 ) and rehabilitation activity tilated patients in terms of the intensity (VO
200
duration. We present novel data regarding the oxygen cost and duration of
201
rehabilitation interventions in mechanically ventilated participants in ICU
202
and demonstrate the following: (i) the oxygen cost of rehabilitation varies
203
between participants; (ii) the oxygen cost of the rehabilitation activity was
204
associated with the level the participant actively contributed the activity;
205
(iii) the recovery time was greater than the duration of the rehabilitation
207
208
209
210
211
d
te
Ac ce p
206
M
198
activity in one in four rehabilitation sessions.
It is important to acknowledge that the population studied were both
persistently mechanically ventilated and critically ill. Therefore it is not possible to extrapolate the findings of this study to all ICU patients receiving rehabilitation. The large standard deviation of the sample data reflects in
212
part the relatively small sample size but also the heterogeneity of the partici-
213
pants recruited and the multiplicity of factors influencing energy expenditure
214
˙ 2 meaduring a rehabilitation session. Technical issues often prevented VO 12
Page 12 of 19
surement during rehabilitation sessions in our mechanically ventilated partic-
216
ipants indicating that the technique is feasible although challenging. Where
217
measurements could be performed, we noted the main factor that influenced
218
˙ 2 was the participants’ physical function status, the percentage change in VO
219
rather than the actual rehabilitation activity. The regression model showed
220
that an increase in ICU-FSS was associated with an increase in the percent-
221
˙ 2 . The relationship between ICU-FSS and change in VO ˙ 2 may relate age VO
222
to reduced O2 consumption in individuals who are less able to contribute
223
actively to the rehabilitation activity. Reasons for being less able to partici-
224
pate actively may include general fatigue, less muscle to recruit as a result of
225
ICU-acquired sarcopenia, and/or decreased oxygen utilisation due to bioen-
226
ergetic dysfunction or changes in fibre type composition. While ICU-FSS
227
˙ 2 , this does not assist in estimating in isolation may be associated with VO
228
an individual patients’ exercise capacity, nor the intensity at which they are
229
working. We found that ICU-FSS was highly correlated with the rehabil-
230
itation activity, presumably because the improvement in physical function
236
Ac ce p
te
d
M
an
us
cr
ip t
215
237
perception of breathlessness will enable them to achieve more in a rehabilita-
238
tion session. We noted a trend towards an increase in rehabilitation activity
239
duration in those participants in whom pressure support was increased (Sup-
231
232
233
234
235
(measured by ICU-FSS) meant the participant was more likely to engage in a higher level of activity (Supplementary Figure 3). ˙ 2 . This Incrementing ventilatory support could influence the change in VO
is commonly done during a rehabilitation session on the assumption that offloading the participants mechanical respiratory load and/or reducing their
13
Page 13 of 19
plementary Figure 2). This may suggest that the perception of breathlessness
241
or mechanical respiratory load are potentially important factors that can be
242
manipulated in mechanically ventilated participants. However, we cannot
243
exclude the possibility that therapists consider this enables them to push
244
participants harder. Further study is warranted, but also challenging. A
245
participant would need to perform the same rehabilitation activity twice,
246
with or without an increase in pressure support. We found many partici-
247
pants were fatigued after just one activity and could perform no more.
an
us
cr
ip t
240
248
There are several limitations to the current study. It was not possible
250
to test repeatability as participants were often too fatigued at the end of a
251
rehabilitation session to repeat the same activity. The few comparisons that
252
were possible between different days were not necessarily valid due to changes
253
in the participants physiological status. There were insufficient repeated
254
episodes to test the sensitivity of the measurement technique to measure
255
change over time. The regression analysis did not detect any alteration in
257
258
259
260
261
d
te
Ac ce p
256
M
249
˙ 2 over time. This is likely due to insufficient the percentage change in VO numbers of participants repeating the same activity over time. Given the nature of rehabilitation in the ICU, participants were rehabilitated to their maximum functional level on each occasion. Reasons for ceasing exercise were unfortunately not recorded; these may have provided useful insights into the limitations to exercise experienced.
14
Page 14 of 19
262
6. Conclusion Breath-by-breath gas exchange analysis remains a challenging technique
264
to perform consistently. However, it can provide valuable information regard-
265
ing the participants oxygen consumption during rehabilitation. We found
266
˙ 2 , is not activity-dependent. that absolute exercise intensity, as measured by VO
267
Therefore the assumption that a rehabilitation activity such as SOEB has the
268
same absolute exercise intensity for each patient is not valid. This highlights
269
the need to monitor each individual patients’ workload during rehabilitation
270
in real time at the point of care.
271
List of changes
272
7. Ethical Approval
275
276
277
278
cr
us
an
M d
8. Funding
te
274
UK National Research Ethics Service (REC reference number 11/LO/1646).
Ac ce p
273
ip t
263
This work was funded by a National Institute for Health Research (NIHR)
Clinical Academic Training Fellowship. 9. Conflict of Interest None.
15
Page 15 of 19
279
10. Disclaimer This report presents independent research funded by the NIHR. The views
281
expressed are those of the author(s) and not necessarily those of the NHS,
282
the NIHR, or the Department of Health.
283
11. Figure Legends
us
cr
ip t
280
Figure 284 1: Oxygen consumption calculation for rehabilitation sessions. a = VO2 attributable to exercise session, b = Total VO2 at rest, a + b = Total
286
session VO2
an
285
M
Figure 287 2: Recruitment of participants to the study
Figure 288 3: Box plots of a. percentage change in session VO2 , b. session VO2
291
292
293
294
295
296
297
d
12. References
te
290
(mL.kg.-1 min-1 ) and c. session VO2 (mL.kg.-1 ). ** p <0.01
Ac ce p
289
[1] M. S. Herridge, Legacy of intensive care unit-acquired weakness, Crit Care Med. 37 (10 Suppl) (2009) S457–S461. URL PM:20046135
[2] S. Benington, D. McWilliams, J. Eddleston, D. Atkinson, Exercise testing in survivors of intensive care-is there a role for cardiopulmonary exercise testing?, J Crit Care 27 (1) (2012) 89–94. URL http://www.ncbi.nlm.nih.gov/pubmed/21958985
298
[3] N. Collings, R. Cusack, A repeated measures, randomised cross-over
299
trial, comparing the acute exercise response between passive and active 16
Page 16 of 19
sitting in critically ill patients, BMC Anesthesiol 15 (1) (2015) 1. doi:
301
10.1186/1471-2253-15-1.
302
URL http://www.biomedcentral.com/1471-2253/15/1
ip t
300
[4] J. Sommers, E. Klooster, S. B. Zoethout, H. L. A. van den Oever, F. Nol-
304
let, R. Tepaske, J. Horn, R. H. H. Engelbert, M. van der Schaaf, Feasi-
305
bility of exercise testing in patients who are critically ill: A prospective,
306
observational multicenter study., Arch Phys Med Rehabil 100 (2) (2019)
307
239–246. doi:10.1016/j.apmr.2018.07.430.
an
us
cr
303
[5] S. C. Slade, C. E. Dionne, M. Underwood, R. Buchbinder, B. Beck,
309
K. Bennell, L. Brosseau, L. Costa, F. Cramp, E. Cup, L. Feehan,
310
M. Ferreira, S. Forbes, P. Glasziou, B. Habets, S. Harris, J. Hay-Smith,
311
S. Hillier, R. Hinman, A. Holland, M. Hondras, G. Kelly, P. Kent, G.-J.
312
Lauret, A. Long, C. Maher, L. Morso, N. Osteras, T. Peterson, R. Quin-
313
livan, K. Rees, J.-P. Regnaux, M. Rietberg, D. Saunders, N. Skoetz,
314
K. Sogaard, T. Takken, M. van Tulder, N. Voet, L. Ward, C. White,
316
317
318
319
320
d
te
Ac ce p
315
M
308
Consensus on exercise reporting template (cert): Modified delphi study, Physical Therapy 96 (10) (2016) 1514–1524.
[6] A. Parker, K. M. Tehranchi, D. M. Needham, Critical care rehabilitation trials: the importance of ‘usual care, Crit Care 17 (5) (2013) 183. doi: 10.1186/cc12884.
URL http://ccforum.com/content/17/5/183
321
[7] P. E. Morris, M. J. Berry, D. C. Files, J. C. Thompson, J. Hauser,
322
L. Flores, S. Dhar, E. Chmelo, J. Lovato, L. D. Case, R. N. Bakhru,
323
A. Sarwal, S. M. Parry, P. Campbell, A. Mote, C. Winkelman, R. D. 17
Page 17 of 19
Hite, B. Nicklas, A. Chatterjee, M. P. Young, Standardized rehabilita-
325
tion and hospital length of stay among patients with acute respiratory
326
failure: A randomized clinical trial., JAMA 315 (24) (2016) 2694–2702.
327
doi:10.1001/jama.2016.7201.
cr
ip t
324
[8] L. Denehy, E. H. Skinner, L. Edbrooke, K. Haines, S. Warrillow,
329
G. Hawthorne, K. Gough, S. V. Hoorn, M. E. Morris, S. Berney, Exer-
330
cise rehabilitation for patients with critical illness: a randomized con-
331
trolled trial with 12 months of follow-up, Crit Care 17 (4) (2013) R156.
332
doi:10.1186/cc12835.
333
URL http://ccforum.com/content/17/4/R156
M
an
us
328
[9] N. E. Brummel, T. D. Girard, E. W. Ely, P. P. Pandharipande,
335
A. Morandi, C. G. Hughes, A. J. Graves, A. Shintani, E. Murphy,
336
B. Work, B. T. Pun, L. Boehm, T. M. Gill, R. S. Dittus, J. C. Jackson,
337
Feasibility and safety of early combined cognitive and physical therapy
338
for critically ill medical and surgical patients: the activity and cognitive
340
341
342
343
te
Ac ce p
339
d
334
therapy in icu (act-icu) trial, Intensive Care Med 40 (3) (2014) 370–9. URL http://www.ncbi.nlm.nih.gov/pubmed/24257969
[10] J. Black, C, P. Grocott, M, M. Singer, Metabolic monitoring in the intensive care unit: a comparison of the medgraphics ultima, deltatrac ii, and douglas bag collection methods, Br J Anaesth 114 (2) (2015)
344
661–8.
345
URL http://www.ncbi.nlm.nih.gov/pubmed/25354946
346
[11] G. M. Black CJ, S. M., The oxygen cost of rehabilitation in mechanically
347
ventialted paitents., Am J Resp Crit Care Med 195 (2017) A2742. 18
Page 18 of 19
[12] J. M. Zanni, R. Korupolu, E. Fan, P. Pradhan, K. Janjua, J. B. Palmer,
349
R. G. Brower, D. M. Needham, Rehabilitation therapy and outcomes
350
in acute respiratory failure: an observational pilot project., J Crit Care
351
25 (2) (2010) 254–262. doi:10.1016/j.jcrc.2009.10.010.
cr
ip t
348
[13] M. Huang, K. S. Chan, J. M. Zanni, S. M. Parry, S.-C. G. B. Neto,
353
J. A. A. Neto, V. Z. M. da Silva, M. E. Kho, D. M. Needham, Functional
354
status score for the icu: An international clinimetric analysis of valid-
355
ity, responsiveness, and minimal important difference., Crit Care Med
356
44 (12) (2016) e1155–e1164. doi:10.1097/CCM.0000000000001949.
359
an
M
358
[14] H. I. D. Physical Activity Policy, General practise physical activity questionnaire.
[15] G. A. Borg, Psychophysical bases of perceived exertion., Med Sci Sports
d
357
us
352
Exerc 14 (5) (1982) 377–381.
361
URL
362
fcgi?dbfrom=pubmed&id=7154893&retmode=ref&cmd=
364
365
http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.
Ac ce p
363
te
360
prlinks
[16] D. Bates, M. M¨achler, B. Bolker, S. Walker, Fitting linear mixed-effects models using lme4, Journal of Statistical Software 67 (1) (2015) 48.
19
Page 19 of 19