Received: 13 April 2022 Revised: 14 January 2023 Accepted: 2 June 2023

DOI: 10.1002/pmrj.13039

ORIGINAL ARTICLE

Effects of mobility dose on discharge disposition

in critically ill stroke patients

Nicole Mazwi MD 1 | India Lissak MS 2 | Karuna Wongtangman MD 3,4 |

Katharina Platzbecker MD 3,5 | Lea Albrecht 6 | Bijan Teja MD, MBA 3,7 |
Xinling Xu PhD 3 | Nicole M. Morteo DPT 8 | Tawnee Sparling MD 9 |
10 10
Nicola Latronico MD | Silvia Barbieri MD | Manfred Blobner MD, PhD 6,11 |
Stefan J. Schaller MD 6,12 | Matthias Eikermann MD, PhD 3,13
1
Department of Rehabilitation Medicine, University of Washington, Seattle, Washington, USA
2
Department of Neurology, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts, USA
3
Department of Anesthesia, Critical Care and Pain Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Massachusetts, USA
4
Department of Anesthesia, Faculty of Medicine, Siriaj Hospital, Mahidol University, Bangkok, Thailand
5
Department of Clinical Epidemiology, Leibniz Institute for Prevention Research and Epidemiology—BIPS, Bremen, Germany
6
Department of Anesthesiology and Intensive Care, Technical University of Munich, School of Medicine, Munich, Germany
7
Departments of Anesthesiology and Critical Care Medicine, St. Michael’s Hospital, Toronto, Ontario, Canada
8
Department of Physical Therapy, Massachusetts General Hospital, Boston, Massachusetts, USA
9
Department of Physical Medicine and Rehabilitation, Spaulding Rehabilitation Hospital, Massachusetts General Hospital, Harvard Medical School, Boston,
Massachusetts, USA
10
Department of Anesthesia, Critical Care and Emergency, Spedali Civili University Hospital, University of Brescia, Brescia, Italy
11
University of Ulm, Faculty of Medicine, Department of Anesthesiology and Intensive Care, Ulm, Germany
12
Charité—Universitätsmedizin Berlin, corporate member of Freie Universität Berlin and Humboldt-Universität zu Berlin, Department of Anesthesiology and
Operative Intensive Care, Berlin, Germany
13
Montefiore Medical Center, Department of Anesthesiology, Albert Einstein College of Medicine, Bronx, New York, USA

Correspondence Abstract
Nicole Mazwi, MD, University of Washington,
Department of Rehabilitation Medicine,
Background: Mobilization in the intensive care unit (ICU) has the potential to
325 9 Avenue, Box 359612, Seattle, improve patient outcomes following acute stroke. The optimal duration and
WA 98104, USA. intensity of mobilization for patients with hemorrhagic or ischemic stroke in the
Email: nmazwi@uw.edu
ICU remain unclear.
Objective: To assess the effect of mobilization dose in the ICU on adverse dis-
Funding information
MGH ECOR, Grant/Award Number: 230898 charge disposition in patients after stroke.
Design: This is an international, prospective, observational cohort study of
critically ill stroke patients (November 2017–September 2019). Duration and
intensity of mobilization was quantified daily by the mobilization quantification
score (MQS).
Setting: Patients requiring ICU-level care were enrolled within 48 hours of
admission at four separate academic medical centers (two in Europe, two in
the United States).
Participants: Participants included individuals (>18 years old) admitted to an
ICU within 48 hours of ischemic or hemorrhagic stroke onset who were func-
tionally independent at baseline.
Interventions: Not applicable.

Nicole Mazwi and India Lissak are co-first authors with equal contribution.

PM&R. 2023;1–10. http://www.pmrjournal.org © 2023 American Academy of Physical Medicine and Rehabilitation. 1
 19341563, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/pmrj.13039 by Institut Pasteur Full PN, Wiley Online Library on [24/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
2 EARLY MOBILITY IN CRITICALLY ILL STROKE PATIENTS

Main Outcome Measure: The primary outcome was adverse discharge

disposition.
Results: Of the patients screened, 163 were eligible for inclusion in the study.
One patient was subsequently excluded due to insufficient data collection
(n = 162). The dose of mobilization varied greatly between centers and patients,
which could not be explained by patients’ comorbidities or disease severity. High
dose of mobilization (mean MQS > 7.3) was associated with a lower likelihood of
adverse discharge (adjusted odds ratio, [aOR]: 0.14; 95% confidence interval [CI]:
0.06–0.31; p < .01).
Conclusion: The increased use of mobilization acutely in the ICU setting may
improve patient outcomes.

INTRODUCTION care facility, or in-hospital mortality. We used the

mobilization quantification score (MQS), a mobilization
Stroke is a leading cause of long-term physical disabil- assessment instrument that incorporates duration and
ity globally and creates significant challenges for intensity, to quantify the daily dose of mobilization.15
patients and caregivers.1 Survival among critically ill
stroke patients has improved in recent years.2 As a
result, there is a growing patient population with long- METHODS
lasting physical and neurocognitive sequelae of stroke.
These sequalae significantly affect functional indepen- Study design
dence and quality of life.3
Traditionally, rehabilitative efforts aimed at mini- This was an international, multicenter, observational
mizing disability have been primarily undertaken in study of patients with acute stroke in ICUs conducted
the postacute setting. More recently, mobilization in at four institutions in North America and Europe.
critically ill patients has been shown to be safe,4–8 Patients were enrolled in Germany, Italy, and two cen-
although studies of efficacy have shown mixed ters in the United States. The research protocol was
results. A Very Early Rehabilitation Trial after stroke approved by the institutional review boards (IRBs) of all
(AVERT), the largest randomized controlled trial of participating centers.
early mobilization after acute stroke in stroke units,
found that patients who received very early mobiliza-
tion (defined as mobilization initiated within 24 hours Participants
of hospital admission) had worse functional out-
comes at 3 months compared with those who did We enrolled critically ill patients with ischemic or hemor-
not.9 Notably, this study did not include patients in an rhagic stroke meeting the following inclusion criteria:
intensive care unit (ICU) setting. However, a post hoc (1) 18 years or older, (2) new onset anterior circulation
analysis of the same study cohort found that patients stroke, (3) symptom onset <48 hours prior to enroll-
who were mobilized in shorter, more frequent ses- ment, (4) expected ICU length of stay (LOS) ≥ 48 hours
sions had improved functional outcomes.10 In addi- from the time of screening, and (5) functionally indepen-
tion, a prospective single center study suggested that dent at least 2 weeks prior to symptom onset (defined
mobilization is feasible and safe in patients after by a Barthel Index score ≥ 70).16 Patients all received
acute aneurysmal subarachnoid hemorrhage and guideline standard of care interventions within their local
may improve long-term patient outcomes.11,12 There- environments. Although we aimed to be broadly inclu-
fore, we suspect that the varying success in neurocri- sive, we excluded patients who had been transferred
tical care patients may be partly due to unique from another facility with a stay exceeding 48 hours,
medical considerations in this population as well as a patients for whom a goals of care discussion was ongo-
lack of consensus on the definition, optimal duration ing and patients with lower extremity amputations.
and intensity of mobilization.13,14 Patients with posterior circulation strokes or traumatic
In this study, we sought to determine whether intracranial hemorrhage were also excluded because of
mobilization dose in the ICU could predict adverse the variability of these patients’ presentations and
discharge disposition in patients with severe ische- courses. Written informed consent was obtained either
mic or hemorrhagic stroke. Adverse discharge is directly from the patient or through an authorized repre-
defined as discharge to a long-term care facility, sentative in accordance with local IRB standards.
skilled nursing facility, swing bed provider (eg, small Stroke severity was calculated using the clinical stroke
hospitals that provide skilled nursing facility care), severity score, a single scale to compare patients with
hospice at the patient’s home, hospice in a health both ischemic and hemorrhagic stroke.17,18
 19341563, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/pmrj.13039 by Institut Pasteur Full PN, Wiley Online Library on [24/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
MAZWI ET AL. 3

Power analysis Primary exposure

Although there has been some research conducted on The primary exposure was mean daily MQS. Daily
early mobilization of neurocritical care patients, none of MQS from nursing and physical therapy were summed
these studies provided data compatible to base power throughout the ICU stay and subsequently divided by
calculations for our primary aim. Therefore, for the number of daily MQS measurements to arrive at the
power calculations, we relied on a similar study in surgi- mean daily MQS. Nonlinearity between the mean daily
cal ICU patients,19 while acknowledging the caveat that dose of mobilization and log odds of adverse discharge
the two groups of patients (neurocritical vs. surgical disposition was detected; thus, the mean MQS was
ICU) may not reflect the same cohort. We estimated a dichotomized to arrive at the exposure variable high
correlation of 0.25 between mobilization dose and dis- versus low mean daily mobilization. The median served
charge disposition. Using a two-tailed alpha error of as a cutoff for the binary variable high versus low
0.05, we calculated that a sample size of 160 patients mobilization.
provides a power of >0.8 for the primary outcome.

Statistical analysis
Mobility data
A multivariate logistic regression model was used as
The health care team documented the type and dura- the primary model to assess whether mobilization dose
tion of daily mobilization for each patient. The defini- affects discharge disposition independent of other
tions of mobilization therapy are provided in Table 1. known confounders: Barthel Index,16 clinical stroke
At all centers, mobilization was prioritized as soon severity score (calculated as a percent max of the
as patients were deemed hemodynamically stable National Institutes of Health Stroke Scale/Functional
(no significant orthostatic blood pressure or pulse vari- Outcome in Patients With Primary Intracerebral Hemor-
ability when mobilized). Mobility was either physical rhage scores), Acute Physiology and Chronic Health
therapy (PT) or nursing led. PT-directed mobilization Evaluation II (APACHE II),25 and Charleson Comorbid-
was documented separately by physical therapists in ity Index (CCI). The clinical stroke severity score17,18
the medical record. Nursing-directed mobility was col- was used to compare stroke severity.
lected daily either from documentation in the medical To evaluate the performance of the primary model
record or by an interview with the nurse conducted analyzing the effect of high versus low mean mobiliza-
by the research team. The data were then used to tion on adverse discharge disposition, we conducted a
calculate the MQS, a measure adopted from the Hosmer–Lemeshow test as well as calculated the sen-
existing ICU mobility scale20,21 to incorporate both sitivity, specificity, positive predictive value, and nega-
time spent mobilized as well as mobilization level tive predictive value of the model. To further explore
(Table 1). A description and supplemental calcula- the relationship between mobilization dose (as a contin-
tions for the MQS are found in the Data S1. uous variable) and outcomes, we used the Youden
Index to investigate the optimal mean mobilization
dose. We also examined the dose–response relation-
Outcomes ship between mean MQS and adverse discharge dispo-
sition using a generalized linear model. Mean MQS
The primary outcome was adverse discharge disposi- was assessed over the duration of hospitalization to
tion, defined as discharge to a long-term care facility, account for LOS.
skilled nursing facility, hospice at the patient’s home,
hospice in a health care facility, or in-hospital mortality.
We additionally collected data on secondary outcomes RESULTS
including ICU and hospital LOS, transfer and ambulation
subdomains of the mini-modified Functional Indepen- Patients admitted to an ICU across four institutions
dence Measure at ICU and hospital discharge,22,23 as were screened from November 2017 to September
well as the Glasgow Outcome Scale—Extended24 and 2019. Of these, 273 patients were eligible for inclusion,
mortality assessed at 90 days following stroke onset and 163 consented to participate (Figure 1). One patient
(see Data S1). was excluded from the analysis due to insufficient data
Data indicative of harm that could possibly be linked collection. The median value of the composite mean
to mobilization were also collected from nursing and phy- daily MQS across the study cohort was 7.3. Patients
sician notes. These adverse events were ascertained with a mean MQS < 7.3 were characterized as having a
from the medical record and included worsened neuro- low MQS (n = 81), and patients with mean MQS ≥ 7.3
logical deficits, falls, angina, myocardial infarction, deep were considered to have a high MQS (n = 81). A greater
vein thrombosis, dizziness, and pulmonary embolism. percentage of patients with ischemic stroke (69%) had
 19341563, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/pmrj.13039 by Institut Pasteur Full PN, Wiley Online Library on [24/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
4 EARLY MOBILITY IN CRITICALLY ILL STROKE PATIENTS

TABLE 1 Calculating the mobilization quantification scale (MQS).

Modified
ICU mobility Calculation (level of
SOMS scale Mobilization level Mobilization description Unit definition activity x units)

0 1 Passive range of Passively rolled or passively 60 min = 1 unit 1 x units

motion exercised by health care
providers (eg, passive cycling),
but not actively moving.
1 1 Active sit/exercise in Any activity in bed, including rolling, 15 min = 1 unit 1 x units
bed or exercise in bridging, active exercises, cycle
chair ergometry and active assisted
exercises; not moving out of bed
or over the edge of the bed.
2 2 Passively to/in chair Hoist, passive lift, or slide transfer 30 min = 1 unit 2 x units
to the chair.
2 3 Sitting on edge of bed May be assisted by staff, but 5 min = 1 unit 3 x units
involves actively sitting over the
side of the bed with some trunk
control.
3 4 Standing of any kind Weight bearing through the feet in 5 min = 1 unit 4 x units
the standing position, with or
without assistance. This may
include use of a standing lifter
device, tilt table or body-weight
supported gait training.
3 5 Active stand-step/ Able to step or shuffle through 5 min = 1 unit 5 x units
shuffle transfer to standing to the chair. This
chair involves actively transferring
weight from one leg to another
to move to the chair. If the
patient has been stood with the
assistance of a medical device,
they must step to the chair (not
included if the patient is wheeled
in a standing lifter device).
3 6 Step in place >/=4 x or Able to walk on the spot or less 5 min = 1 unit 6 x units
walk <15 ft (5 m) than 15 ft (5 m) by lifting
alternate feed (must be able to
step at least 4 times, twice on
each foot), with or without
assistance.
4 7 Walk w/2+ assist Walking away from the bed/chair by 5 min = 1 unit 7 x units
>/= 15 ft (5 m) at least 15 ft (5 m) assisted by
two or more people.
4 8 Walk w/1 assist Walking away from the bed/chair by 5 min = 1 unit 8 x units
>/= 15 ft (5 m) at least 15 ft (5 m) assisted by
one person.
4 9 Walk independently w/ Walking away from the bed/chair by 5 min = 1 unit 9 x units
device >/= 15 ft at least 15 ft (5 m) with a gait
(5 m) aid, but no assistance from
another person. In a wheelchair-
bound person, this activity level
including wheeling the chair
independently 15 ft (5 m) away
from the bed/chair.
4 10 Walk independently w/o Walking away from the bed/chair by 5 min = 1 unit 10 x units
device >/= 15 ft at least 15 ft (5 m) without a gait
(5 m) aid or assistance from another
person.
Abbreviation: ICU, intensive care unit; SOMS, surgical intensive care unit optimal mobilisation score.
 19341563, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/pmrj.13039 by Institut Pasteur Full PN, Wiley Online Library on [24/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
MAZWI ET AL. 5

FIGURE 1 CONSORT diagram. CONSORT, Consolidated Standards of Reporting Trials; ICU, intensive care unit; LOS, length of stay.

TABLE 2 Patient characteristics.

Low mobilization dose High mobilization dose
Baseline characteristics (mean MQS < 7.3) N = 81 (mean MQS ≥ 7.3) N = 81 p value

Gender, female, n (%) 44 (54%) 39 (48%) .43

Age, years, mean ± SD 68.3 ± 13.2 66.1 ± 15.5 .33
Body mass index, kg/m2, mean ± SD 28.1 ± 7.1 27.4 ± 7.2 .50
Type of stroke, n (%) .01
Ischemic 40 (49%) 56 (69%)
Hemorrhagic 41 (51%) 25 (31%)
Glasgow coma score at admission, median (IQR) 9 (7,11) 13 (10,14) <.01
Acute physiology and chronic health evaluation II 18 (15,21) 16 (12,20) .04
score, median (IQR)
Barthel Index, mean ± SD 98.2 ± 4.9 98.7 ± 4.1 .49
% maximum of stroke severity score, mean ± SD 51.3 ± 19.4 44.2 ± 24.4 .04
Charlson comorbidity index, median (IQR) 3 (2,5) 3 (2,5) .25
Frailty phenotype modified, median (IQR) 0 (0,1) 0 (0,1) .65
Abbreviations: IQR, interquartile range; MQS, mobilization quantification score.

high MQS compared to patients with hemorrhagic stroke Primary analysis

(31%; p = .01), but there was no significant difference in
adverse discharge disposition between the two groups Within the study cohort, 72 (44.4%) patients had
(p = .39). There were significant differences between adverse discharge disposition. Among patients with
patients with low versus high MQS with respect to a low MQS, 55 (67.9%) had adverse discharge dispo-
admission Glasgow Coma Scale score (p < .01), sition compared to 17 (20.1%) with high MQS
APACHE II score (p = .04), and stroke severity score (Figure 2). High MQS was associated with lower odds
(p = .04), which were subsequently controlled for of adverse discharge disposition (odds ratio [OR]:
through multivariable analysis. Among the study popula- 0.13; 95% confidence interval [CI]: 0.06–0.26;
tion, 112 (69%) had hypertension, 36 (22%) had diabe- p < .01). When adjusting for APACHE II, CCI, and
tes, 34 (21%) had atrial fibrillation, 24 (15%) were stroke severity score, the association of high MQS
current smokers, 42 (32%) had a history of smoking, with lower likelihood of adverse discharge disposition
and 47 (29%) had dyslipidemia. Additional baseline remained robust (adjusted OR [aOR]: 0.14; 95% CI:
characteristics on admission are summarized in Table 2. 0.06–0.31; p < .01).
 19341563, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/pmrj.13039 by Institut Pasteur Full PN, Wiley Online Library on [24/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
6 EARLY MOBILITY IN CRITICALLY ILL STROKE PATIENTS

F I G U R E 2 Violin plots of mean mobilization quantification score (MQS) during hospitalization for patients with and without adverse
discharge disposition. Data points are jittered for improved visualization.

For the outcome adverse discharge disposition, the mobilization (mean duration ≤41 minutes per day). This
sensitivity/specificity of high MQS was 71% and 76% relationship was maintained when controlling for
respectively, the positive predictive value was 79%, APACHE II, CCI, and stroke severity (aOR: 0.11, 95%
the negative predictive value was 68%, and the area CI: 0.05–0.25; p < .01). Patients who achieved a level
under the receiver operating curve was 0.74 (95% CI: of ≥ 5 or greater on the modified ICU mobility scale
0.66–0.81). The Hosmer—Lemeshow test yielded a (active stand or step-shuffle to the chair at minimum) at
chi-square of 4.42 (p = .82), indicating that the model did any point during their ICU admission had lower odds of
not demonstrate lack of goodness of fit. Youden Index adverse discharge disposition compared to patients
yielded an optimal mean MQS threshold of 5.9 during criti- whose maximum mobilization level was < 5 during the
cal illness as the optimal minimum dose for patients to course of their ICU stay (OR: 0.14, 95% CI: 0.07–0.29;
avoid adverse discharge disposition. A generalized linear p < .01). This finding remained robust when controlling
model confirmed that there was a dose–response relation- for APACHE II, CCI, and stroke severity (aOR: 0.18,
ship between mean MQS and discharge disposition 95% CI: 0.08–0.40; p < .01).
(accuracy 0.72; 95% CI: 0.64–0.79; p < .01; Figure 3).
The association between MQS and adverse discharge
disposition was unchanged when using the optimal mean Adverse event rates
MQS threshold of 5.9, as determined by the Youden
Index analysis (aOR: 0.11; 95% CI: 0.05–0.25; p < .01). Among patients included in our cohort, two (1.2%) had
In an analysis examining the effect of mean duration falls (one with low mean MQS versus one with high
of mobilization on adverse discharge disposition, high mean MQS), three (1.9%) experienced angina (one
mean duration of mobilization (mean duration with low mean MQS versus two with high mean MQS),
>41 minutes per day) was associated with lower odds three (1.9%) had a myocardial infarct (three with low
of adverse discharge disposition (OR: 0.11, 95% CI: mean MQS versus none with high mean MQS), four
0.05–0.23; p < .01) compared to low mean duration of (2.5%) had deep venous thrombosis (three with low
 19341563, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/pmrj.13039 by Institut Pasteur Full PN, Wiley Online Library on [24/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
MAZWI ET AL. 7

F I G U R E 3 Dose response relationship between mean mobilization quantification score (MQS) during hospitalization and probability of
adverse discharge disposition.

mean MQS versus one with high mean MQS), and nine Mobilization represents a promising therapy to reduce
(5.6%) experienced pulmonary embolism (six with low ICU-acquired muscle weakness.26,27 ICU-acquired mus-
mean MQS vs. three with high mean MQS). cle weakness has been shown to increase mortality and
prolong illness.6,28–30 However, studies have found
heterogeneous results when assessing the impact of
DISCUSSION mobilization on outcomes. The AVERT study9 raised the
concern for possible harm in patients who were mobilized
In this international, multicenter observational trial, we very early (within 24 hours); however, a post hoc analysis
found that mobilization dose in the ICU in patients after revealed that patients who were mobilized in shorter,
stroke varies when adjusting for stroke severity, comor- more frequent sessions had more favorable outcomes.10
bidities, and other predictors of receiving a high dose of There was no mobilization-induced harm detected in
mobilization therapy. High mobilization dose was asso- our study sample of ICU patients with stroke, as
ciated with a lower risk of losing the ability to live inde- assessed by reported adverse events. Given the poten-
pendently after hospital discharge. This relationship tial harm identified in overmobilizing patients in the non
was maintained when adjusting for stroke severity, ICU-setting, identifying an optimal mobilization dose is
comorbidities and study center. Benefits were seen in critical.9 The authors of the AVERT trial noted that the
both short-term (ICU LOS, hospital LOS, and discharge harmful effects of early mobilization after stroke seen in
destination) and long-term outcomes (90-day functional their study may have been due to changes in blood
status and mortality). Importantly, we assessed the opti- pressure, inhibition of penumbral tissue reperfusion, or
mal dose of mobilization and found that a mean MQS of increased rebleeding in the case of patients with hemor-
5.9, which is equal to walking in place for approximately rhagic stroke. Ensuring patients are hemodynamically
5 minutes daily, or sitting stationary in a chair for stable (no significant orthostatic blood pressure or pulse
1.5 hours daily, was the optimal minimum dose for variability when mobilized) prior to the initiation of mobili-
patients to avoid adverse discharge disposition. zation may help to prevent these complications.
 19341563, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/pmrj.13039 by Institut Pasteur Full PN, Wiley Online Library on [24/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
8 EARLY MOBILITY IN CRITICALLY ILL STROKE PATIENTS

Two important factors that differ across studies of Study limitations

mobilization are the initiation time point and the
nature of the mobilization interventions; the definition Our study was subject to the challenges inherent in a
of early mobilization has yet to be standardized. In prospective observational study. We used a combined
previously published literature, early mobilization has MQS that included nursing-led and PT-directed mobili-
been used to define mobilization within a specific zation. Although this approach enabled us to analyze
threshold of hours or days post injury or to quantify composite mobilization, there may be benefit in looking
all ICU-specific mobilization in comparison to mobili- at these exposures independently. The MQS has sev-
zation in the outpatient or rehabilitation setting.9 eral benefits over other scales, including the combina-
Furthermore, the specific types of mobilization prac- tion of duration and intensity in a single score. We
tices vary substantially. Some studies require early intentionally included patients with both ischemic and
mobilization to be protocol driven and goal directed, hemorrhagic stroke to increase generalizability of our
whereas others consider mobilization to be any study; however, there may be subgroups within these
facilitated passive or active movement.13,19,31,32 We populations that benefit from different doses more than
chose to approach this issue broadly, including all others. We also did not analyze our patient population
mobilization that encompassed both lower-level by therapeutic intervention and so cannot make com-
activity (ie, passive range of motion in bed) and ments about whether various therapies prior to ICU
higher-level activity (eg, ambulating) that was part of admission affect these findings. Despite these limita-
the nursing and PT-directed treatment plan. tions, our study has several notable strengths. Our pro-
The precise mechanism of action by which mobiliza- spective study was clinically pragmatic and included
tion is linked to improved outcomes is unclear among multiple centers across three countries capturing highly
neurocritical care patients; however, it is believed that detailed mobilization data. We present granular data for
these benefits are mediated by both a reduction in mobilization measures and outcomes and collect total
immobility-related complications and the promotion of mobilization from a range of providers, allowing for a
neuroplasticity during a critical recovery window. ICU- complete calculation of mobilization dose encompass-
acquired immobility has been linked to increased intra- ing both mobilization type and duration.
muscular fat, decreased cardiovascular reserve, and
muscle atrophy.33 Although additional studies are needed
to determine whether mobilization can prevent these CONCLUSION
immobility-related complications, previous research sug-
gests that early mobilization has the potential to do This observational study identified the potential ben-
so.19,34–38 The critical sensitive period represents a win- efits of early mobilization in an inclusive population
dow of heightened neuroplasticity after stroke,39 during of patients in the ICU with ischemic and hemor-
which mobilization may improve outcomes by means of rhagic stroke. We found that both the duration and
enhanced neuroplasticity.40 intensity of mobilization may be important contribu-
In adjusted analyses controlling for APACHE II, tors to outcome, lending support to the increased
CCI, and stroke severity, high mobilization dose use of mobilization in the ICU setting. Although this
reduced adverse discharge disposition rates and practice is generally considered safe, additional
improved 90-day outcomes. Although the dose of mobi- studies are needed to evaluate whether there are
lization applied varied across centers, the association specific subgroups that may benefit most from mobi-
between high mobilization dose and lower risk of lization. Mobilization in the ICU offers a promising
adverse discharge disposition was independent intervention to improve both short and long-term out-
of center geography and local standards of care. The comes in critically ill patients after stroke.
mean MQS was higher when the propensity score,
which accounted for several clinical variables as well AC KN OW LED GME NT S
as the individual enrollment centers, was high. This We are grateful to Judith A. Fong and Mark F. Wheeler,
study demonstrates that routine mobilization therapy MD for their generosity. Thank you to Hillary Kelly, Can
led by nursing and PT is feasible in patients with acute Ozan Tan, and Karen Waak for their assistance in
stroke across hospitals in the United States and executing the study.
Europe. Prior studies have demonstrated that both the
dose and timing of mobilization are important to opti- DIS CL OS UR ES
mize the rehabilitative benefit.19,41 This study builds None.
upon prior research by employing a novel metric of
quantifying dose and timing of mobilization into a com- OR CI D
posite metric as well as further demonstrating an asso- Tawnee Sparling https://orcid.org/0000-0001-5237-
ciation between mobilization dose and outcomes. 0984
 19341563, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/pmrj.13039 by Institut Pasteur Full PN, Wiley Online Library on [24/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
MAZWI ET AL. 9

Stefan J. Schaller https://orcid.org/0000-0002-6683- 19. Schaller SJ, Anstey M, Blobner M, et al. Early, goal-directed
9584 mobilisation in the surgical intensive care unit: a randomised
controlled trial. Lancet. 2016;388(10052):1377-1388.
20. Hodgson C, Needham D, Haines K, et al. Feasibility and inter-
R E F E REN CE S rater reliability of the ICU mobility scale. Heart Lung. 2014;43(1):
1. Avan A, Digaleh H, Di Napoli M, et al. Socioeconomic status and 19-24.
stroke incidence, prevalence, mortality, and worldwide burden: 21. Tipping CJ, Bailey MJ, Bellomo R, et al. The ICU mobility scale
an ecological analysis from the Global Burden of Disease study has construct and predictive validity and is responsive. A multicen-
2017. BMC Med. 2019;17(1):191. ter observational study. Ann Am Thorac Soc. 2016;13(6):887-893.
2. Ovbiagele B, Goldstein LB, Higashida RT, et al. Forecasting the 22. Wagner AK, Fabio T, Zafonte RD, Goldberg G, Marion DW,
future of stroke in the United States: a policy statement from Peitzman AB. Physical medicine and rehabilitation consultation:
the American Heart Association and American Stroke Associa- relationships with acute functional outcome, length of stay, and
tion. Stroke. 2013;44(8):2361-2375. discharge planning after traumatic brain injury. Am J Phys Med
3. Dhamoon MS, Moon YP, Paik MC, et al. Quality of life declines Rehabil. 2003;82(7):526-536.
after first ischemic stroke the Northern manhattan study. Neurol- 23. Gennarelli TA, Champion HR, Copes WS, Sacco WJ. Compari-
ogy. 2010;75(4):328-334. son of mortality, morbidity, and severity of 59,713 head injured
4. Dirkes SM, Kozlowski C. Early mobility in the intensive care unit: patients with 114,447 patients with extracranial injuries.
evidence, barriers, and future directions. Crit Care Nurse. 2019; J Trauma. 1994;37(6):962-968.
39(3):33-42. 24. NIH/NINDS. CDE Detailed Report; Stroke; Activities of Daily
5. Hodgson CL, Capell E, Tipping CJ. Early mobilization of patients Living/Performance; Glasgow Outcome Scale Extended (GOSE).
in intensive care: organization, communication and safety fac- 25. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE
tors that influence translation into clinical practice. Crit Care. II: a severity of disease classification system. Crit Care Med.
2018;22(1):77. 1985;13(10):818-829.
6. Morris PE, Goad A, Thompson C, et al. Early intensive care unit 26. Tipping CJ, Harrold M, Holland A, Romero L, Nisbet T,
mobility therapy in the treatment of acute respiratory failure. Crit Hodgson CL. The effects of active mobilisation and rehabilitation
Care Med. 2008;36(8):2238-2243. in ICU on mortality and function: a systematic review. Intensive
7. Nydahl P, Sricharoenchai T, Chandra S, et al. Safety of patient Care Med. 2017;43(2):171-183.
mobilization and rehabilitation in the intensive care unit. System- 27. Patel BK, Pohlman AS, Hall JB, Kress JP. Impact of early mobili-
atic review with meta-analysis. Ann Am Thorac Soc. 2017;14(5): zation on glycemic control and ICU-acquired weakness in criti-
766-777. cally ill patients who are mechanically ventilated. Chest. 2014;
8. Mayer KP, Joseph-Isang E, Robinson LE, Parry SM, Morris PE, 146(3):583-589.
Neyra JA. Safety and feasibility of physical rehabilitation and 28. Mueller N, Murthy S, Tainter CR, et al. Can sarcopenia quanti-
active mobilization in patients requiring continuous renal fied by ultrasound of the rectus Femoris muscle predict adverse
replacement therapy: a systematic review. Crit Care Med. 2020; outcome of surgical intensive care unit patients as well as frailty?
48(11):e1112-e1120. A prospective observational cohort study. Ann Surg. 2016;
9. group ATC. Efficacy and safety of very early mobilisation within 264(6):1116-1124.
24 h of stroke onset (AVERT): a randomised controlled trial. 29. Needham DM, Korupolu R, Zanni JM, et al. Early physical medicine
Lancet. 2015;386(9988):46-55. and rehabilitation for patients with acute respiratory failure: a quality
10. Bernhardt J, Churilov L, Ellery F, et al. Prespecified dose- improvement project. Arch Phys Med Rehabil. 2010;91(4):536-542.
response analysis for A Very Early Rehabilitation Trial (AVERT). 30. Van Aerde N, Meersseman P, Debaveye Y, et al. Five-year
Neurology. 2016;86(23):2138-2145. impact of ICU-acquired neuromuscular complications: a pro-
11. Karic T, Roe C, Nordenmark TH, Becker F, Sorteberg W, spective, observational study. Intensive Care Med. 2020;46(6):
Sorteberg A. Effect of early mobilization and rehabilitation on 1184-1193.
complications in aneurysmal subarachnoid hemorrhage. 31. Hodgson CL, Bailey M, Bellomo R, et al. A binational multicenter
J Neurosurg. 2017;126(2):518-526. pilot feasibility randomized controlled trial of early goal-directed
12. Karic T, Roe C, Nordenmark TH, Becker F, Sorteberg A. Impact mobilization in the ICU. Crit Care Med. 2016;44(6):1145-1152.
of early mobilization and rehabilitation on global functional out- 32. Wollersheim T, Grunow JJ, Carbon NM, et al. Muscle wasting
come one year after aneurysmal subarachnoid hemorrhage. and function after muscle activation and early protocol-based
J Rehabil Med. 2016;48(8):676-682. physiotherapy: an explorative trial. J Cachexia Sarcopenia Mus-
13. Fuest K, Schaller SJ. Recent evidence on early mobilization in cle. 2019;10(4):734-747.
critical-ill patients. Curr Opin Anaesthesiol. 2018;31(2):144-150. 33. Lipshutz AK, Gropper MA. Acquired neuromuscular weakness
14. Hernandez S, Kittelty K, Hodgson CL. Rehabilitating the neuro- and early mobilization in the intensive care unit. Anesthesiology.
logical patient in the ICU: what is important? Curr Opin Crit Care. 2013;118(1):202-215.
2021;27(2):120-130. 34. Askim T, Bernhardt J, Loge AD, Indredavik B. Stroke patients do
15. Scheffenbichler FT, Teja B, Wongtangman K, et al. Effects of not need to be inactive in the first two-weeks after stroke: results
the level and duration of mobilization therapy in the surgical ICU from a stroke unit focused on early rehabilitation. Int J Stroke.
on the loss of the ability to live independently: an international 2012;7(1):25-31.
prospective cohort study. Crit Care Med. 2021;49(3):e247-e257. 35. Cumming TB, Thrift AG, Collier JM, et al. Very early mobilization
16. Mahoney FI, Barthel DW. Functional evaluation: the Barthel after stroke fast-tracks return to walking: further results from the
index. Md State Med J. 1965;14:61-65. phase II AVERT randomized controlled trial. Stroke. 2011;42(1):
17. Racosta JM, Di Guglielmo F, Klein FR, et al. Stroke severity 153-158.
score based on six signs and symptoms the 6S score: a simple 36. Govan L, Langhorne P, Weir CJ, Stroke Unit Trialists C. Does
tool for assessing stroke severity and In-hospital mortality. the prevention of complications explain the survival benefit of
J Stroke. 2014;16(3):178-183. organized inpatient (stroke unit) care?: further analysis of a sys-
18. Andersen KK, Olsen TS. One-month to 10-year survival in the tematic review. Stroke. 2007;38(9):2536-2540.
Copenhagen stroke study: interactions between stroke severity 37. Schaller SJ, Scheffenbichler FT, Bose S, et al. Influence of the
and other prognostic indicators. J Stroke Cerebrovasc Dis. initial level of consciousness on early, goal-directed mobilization:
2011;20(2):117-123. a post hoc analysis. Intensive Care Med. 2019;45(2):201-210.
 19341563, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/pmrj.13039 by Institut Pasteur Full PN, Wiley Online Library on [24/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
10 EARLY MOBILITY IN CRITICALLY ILL STROKE PATIENTS

38. Schweickert WD, Pohlman MC, Pohlman AS, et al. Early physi- S UP P O RT IN G I NF O R M AT IO N
cal and occupational therapy in mechanically ventilated, critically Additional supporting information can be found online
ill patients: a randomised controlled trial. Lancet. 2009;
in the Supporting Information section at the end of this
373(9678):1874-1882.
39. Krakauer JW, Carmichael ST, Corbett D, Wittenberg GF. Getting article.
neurorehabilitation right: what can be learned from animal
models? Neurorehabil Neural Repair. 2012;26(8):923-931.
40. Kelly C, Foxe JJ, Garavan H. Patterns of normal human brain How to cite this article: Mazwi N, Lissak I,
plasticity after practice and their implications for neurorehabilita- Wongtangman K, et al. Effects of mobility dose
tion. Arch Phys Med Rehabil. 2006;87(12 suppl 2):S20-S29. on discharge disposition in critically ill stroke
41. Paton M, Lane R, Paul E, Cuthburtson GA, Hodgson CL. Mobili-
patients. PM&R. 2023;1‐10. doi:10.1002/pmrj.
zation during critical illness: a higher level of mobilization
improves health status at 6 months, a secondary analysis of a 13039
prospective cohort study. Crit Care Med. 2021;49:e860-e869.