Research Report

Physical Therapist Treatment of

Patients in the Neurological Intensive
Care Unit: Description of Practice

Downloaded from https://academic.oup.com/ptj/article/95/7/1006/2686465 by guest on 05 December 2020

Peter D. Sottile, Amy Nordon-Craft, Daniel Malone, Darcie M. Luby,
Margaret Schenkman, Marc Moss
P.D. Sottile, MD, University of Col-
orado, Anschutz Medical Campus,
Research Building 2, 9th Floor,
Background. Although studies have established the safety and feasibility of phys-
12700 E 19th Ave, Aurora, CO ical therapy in the critical care setting, minimal information about physical therapist
80045 (USA). Address all corre- practice in the neurological intensive care unit (NICU) is available.
spondence to Dr Sottile at:
peter.sottile@ucdenver.edu. Objective. This study describes physical therapists’ treatment of people admitted
A. Nordon-Craft, PT, DSc, School of to a NICU.
Medicine, University of Colorado.

D. Malone, PT, PhD, CCS, School of Design. People admitted to the NICU with a diagnosis of subarachnoid hemor-
Medicine, University of Colorado. rhage, subdural hematoma, intracranial hemorrhage, or trauma were retrospectively
studied.
D.M. Luby, PT, DPT, School of
Medicine, University of Colorado.
Methods. Data on patient demographics, use of mechanical ventilation, and
M. Schenkman, PT, PhD, FAPTA, intracranial pressure (ICP) monitoring were collected. For each physical therapy
School of Medicine, University of
Colorado.
session, the length of the session, the location (NICU or post-NICU setting), and the
presence of mechanical ventilation or ICP monitoring were recorded. Data on safety
M. Moss, MD, School of Medicine, parameters, including vital sign response, falls, and dislodgement of lines, were
University of Colorado.
collected.
[Sottile PD, Nordon-Craft A,
Malone D, et al. Physical therapist Results. Over 1 year, 180 people were admitted to the NICU; 86 were evaluated
treatment of patients in the neu-
by a physical therapist, for a total of 293 physical therapy sessions in the NICU
rological intensive care unit:
description of practice. Phys Ther. (n⫽132) or post-NICU setting (n⫽161). Only one session (0.3%) was stopped,
2015;95:1006 –1014.] secondary to an increase in ICP. The first physical therapy session occurred on NICU
day 3.0 (25%–75% interquartile range⫽2.0 – 6.0). Patients received a median of 3.4
© 2015 American Physical Therapy
Association sessions per week (25%–75% interquartile range⫽1.8 –5.9). Patients with mechanical
ventilation received less frequent physical therapy sessions than those without
Published Ahead of Print:
mechanical ventilation. Patients with ICP monitoring received less frequent sessions
February 5, 2015
Accepted: January 28, 2015 than those without ICP monitoring. However, after multivariate analysis, only the
Submitted: March 13, 2014 admission Glasgow Coma Score was independently associated with physical therapy
frequency in the NICU. Patients were more likely to stand, transfer, and walk in the
post-NICU setting than in the NICU.

Limitations. The results are limited by the retrospective, single-center nature of

the study. There is inherent bias of evaluating only those patients who had physical
therapy, and therapists were unable to completely adjust for the severity of illness of
a given patient.

Conclusions. Physical therapy was performed safely in the NICU. Patients who
Post a Rapid Response to required invasive support received less frequent physical therapy.
this article at:
ptjournal.apta.org

1006 f Physical Therapy Volume 95 Number 7 July 2015

 Physical Therapy in Neurological Intensive Care

G
rowing evidence supports motor control, which can have Method
early mobilization and physi- immediate and profound effects on Design
cal therapy to improve both balance, mobility, and the ability to We conducted a retrospective
short and long-term physical func- perform skilled movements. Altera- cohort study to describe the current
tion in patients who are critically ill. tions in muscle tone may impair physical therapist interventions re-
Muscle strength decreases 1% to 2% range of motion (ROM) more rapidly ceived by people admitted to a NICU
daily in patients who are critically ill in people with central nervous sys- and through their acute care hospi-
and confined to bed, and early inter- tem impairments. Moreover, such talization. Adults initially admitted to
ventions appear to limit the long- people frequently have specific and the NICU at a university hospital
term effects of neurological inten- debilitating impairments of percep- from January 1, 2012, until Decem-
sive care unit (NICU)–associated tion and cognition that may affect ber 31, 2012, with a primary diagno-

Downloaded from https://academic.oup.com/ptj/article/95/7/1006/2686465 by guest on 05 December 2020

weakness.1,2 In a recent systematic the implementation of early physical sis of SAH, SDH, ICH, or trauma were
review, early mobilization of patients therapy.4 identified from hospital admission
who required mechanical ventilation records. Physical therapy depart-
improved muscle strength, increased The focus of neurocritical care is to ment administrative records identify-
ventilator-free days, and decreased attempt to reduce the chronic dys- ing people who received physical
length of hospital stay.3 Patients who function and debilitation of acute therapy orders during their hospital-
received physical therapy while crit- neurological diseases, and some evi- ization were cross-referenced with
ically ill had improved long-term out- dence supports the use of intensive NICU admission records. From this
comes, including the ability to per- physical therapy in the neurocritical group of people, those with active
form independent activities of daily care setting.5 For example, a ran- physical therapist consultations
living after hospital discharge.2 domized controlled trial of early were included for further review.
Importantly, physical therapy can be physical therapy within 24 hours of People younger than 18 years of age
performed safely for most patients admission in people who experi- were excluded.
who are critically ill.1,3 enced a stroke suggested improved
functional recovery. However, that The NICU consists of 10 beds primar-
Despite this growing body of evi- study excluded people who required ily used for patients who need neu-
dence supporting early physical ther- admission to the NICU.6,7 In a care- rosurgical and neurological care. It is
apy for patients who are critically ill, fully selected cohort of people with staffed by one physical therapist
the efficacy and safety of early phys- SAH and normal ICP, early physical Monday through Friday, with week-
ical therapy remain unclear with therapy was performed safely and end coverage determined by the pri-
respect to people who have acute did not significantly change the ICP mary physical therapist. Additional
nervous system injury. People are with ROM or limb exercises.8,9 physical therapists can be added to
often admitted to the NICU with the Despite these studies, data about the the unit as needed.
diagnosis of acute ischemic stroke, implementation, effects, and safety
subdural hematoma (SDH), sub- of physical therapy for patients in Data Collection
arachnoid hemorrhage (SAH), intra- the NICU are scarce. Administrative data, physician notes,
cranial hemorrhage (ICH), hydro- physical therapist notes, nursing
cephalus, or traumatic brain injury. The purpose of this report is to flow sheets, and respiratory therapist
Patients who require neurocritical describe the current physical thera- flow sheets were extracted from
care differ from patients who have pist interventions for people with electronic medical records. Data
other critical illnesses in specific SAH, SDH, ICH, or trauma in a NICU were extracted by a single person
and important ways. For example, in a large university teaching hospi- (P.D.S.). We did not perform a pro-
patients in the neurocritical care set- tal. We hypothesized that people spective audit of data entry to evalu-
ting often require specialized and admitted to the NICU with invasive ate the accuracy of the retrospective
invasive monitoring, such as intracra- monitoring or mechanical ventila- data. Basic characteristics—includ-
nial pressure (ICP) monitoring. Such tion would receive physical therapy ing age, sex, race, admission diagno-
monitoring can limit physical ther- safely but might receive less frequent sis, admission Glasgow Coma Score
apy interventions. Indeed, physical and less intensive physical therapy (GCS), tracheostomy, craniotomy,
therapy may increase the ICP and be than people without such invasive date of physical therapist consulta-
contraindicated for some patients in support. tion, post–acute care discharge loca-
the neurocritical care setting. Addi- tion (home, skilled nursing facility,
tionally, people with acute nervous long-term acute care setting, inpa-
system injury often have altered tient rehabilitation setting, or

July 2015 Volume 95 Number 7 Physical Therapy f 1007

Physical Therapy in Neurological Intensive Care

Table 1.
Characteristics of Patients by Diagnosisa

Characteristic Total SAH SDH ICH Trauma

Patients 86 42 (49) 17 (20) 19 (22) 8 (9)

Age, y, X (SD) 60 (17) 56 (18) 67 (15) 64 (13) 54 (21)

Men 45 (52) 17 (40) 13 (76) 11 (58) 5 (63)

White 68 (79) 33 (79) 15 (88) 14 (74) 6 (75)

NICU length of stay, d, median (25%–75% IQR) 9.0 (3.0–14.0) 12.0 (5.8–17.0) 3.0 (2.0–9.0) 6.0 (2.0–14.0) 12.0 (3.0–49.0)

Ventilated 50 (58) 31 (74) 4 (24) 10 (53) 5 (63)

b
Ventilator days, median (25%–75% IQR) 3.0 (1.0–12.0) 1.0 (1.0–11.0) 6.5 (1.5–11.5) 5.5 (1.0–12.5) 22.0 (5.5–55.0)

Downloaded from https://academic.oup.com/ptj/article/95/7/1006/2686465 by guest on 05 December 2020

Tracheostomy 9 (10) 4 (10) 0 (0) 2 (11) 3 (38)

ICP monitoring 35 (41) 16 (38) 11 (30) 4 (11) 4 (11)

ICP days,b median (25%–75% IQR) 12.0 (3.0–19.0) 16.0 (12.5–21.8) 2.0 (1.0–3.0) 10.5 (4.0–23.0) 23.0 (6.5–37.3)

Hospital length of stay, d, median (25%–75% IQR) 12.0 (6.0–20.3) 13.5 (9.8–21.5) 6.0 (3.0–12.0) 8.0 (4.0–20.0) 21.5 (6.8–51.5)

Mortality 8 (9) 2 (5) 4 (24) 1 (5) 1 (13)

Discharged to home 46 (53) 26 (62) 10 (59) 8 (42) 2 (25)

a
Data are numbers (percentages) of patients unless otherwise indicated. SAH⫽subarachnoid hemorrhage, SDH⫽subdural hematoma, ICH⫽intracranial
hemorrhage, NICU⫽neurological intensive care unit, IQR⫽interquartile range, ICP⫽intracranial pressure.
b
Ventilator days and ICP days were calculated for patients who needed mechanical ventilation or ICP monitoring only.

another hospital), and mortality— ties, or bed weights), transferring the NICU with that during the hos-
were extracted. The duration of (sitting or standing), standing, and pitalization. Similarly, we performed
mechanical ventilation, the type and ambulating. logistic regression analysis, while
duration of ICP monitoring, and the adjusting for age, craniotomy, pres-
lengths of NICU and hospital stays Data Analysis ence of mechanical ventilation, and
also were collected. Ventilator and Characteristics are reported as mean presence of ICP monitoring, to deter-
NICU days were calculated. and standard deviation or as median mine the odds of particular physical
and 25% to 75% interquartile range therapist interventions occurring in
For each physical therapy session, (IQR). Outcome variables included the NICU. Data were analyzed with
data on safety parameters and their the following: time to first physical JMP 10.0 (SAS Institute Inc, Cary,
effects on the physical therapy ses- therapy session, frequency of physi- North Carolina).
sion were collected; these data cal therapy sessions per week (in the
included changes in vital signs, NICU and over the course of the hos- Role of the Funding Source
arrhythmia, changes in the ICP, falls, pitalization), median duration of ses- Funding was provided by National
and dislodgement of lines. Physical sions, and types of interventions at Institutes of Health grants R01
therapists and bedside nurses gener- each session. Data were stratified NR011051 and K24 HL089223.
ally decided the safety of initiating a between NICU sessions and post-
given therapy session. Additionally, NICU sessions, sessions with ventila- Results
date and location of therapy (NICU tion and sessions without ventilation Characteristics of Patients
or post-NICU setting [ie, after NICU in the NICU, and sessions with ICP Over the course of 1 year, 180 peo-
discharge to the floor or ward but monitoring and sessions without ICP ple were admitted to the NICU with
before discharge from the acute care monitoring in the NICU. All out- SAH, SDH, ICH, or trauma; 87 (48%)
setting]), mean GCS on the day of comes were compared with Wil- received a physical therapist consul-
each session, duration of each ses- coxon rank sum tests. To adjust for tation, and 86 were formally evalu-
sion, and the presence of mechanical potential confounders, we per- ated by a physical therapist. The pri-
ventilation or ICP monitoring were formed linear regression analysis, mary diagnoses in these 86 patients
obtained from physical therapist while adjusting for age, admission were SAH (n⫽42), SDH (n⫽17), ICH
notes. The types of therapy per- GCS, craniotomy, presence of (n⫽19), and trauma (n⫽8) (Tab. 1).
formed were categorized as ROM, mechanical ventilation, and pres- The average age was 60 years
bed-based interventions (sitting at ence of ICP monitoring, to compare (SD⫽17), 45 patients (52%) were
the edge of the bed, posture activi- the frequency of physical therapy in men, and 68 patients (79%) were

1008 f Physical Therapy Volume 95 Number 7 July 2015

 Physical Therapy in Neurological Intensive Care

white. The admission GCS was 14.1 (ICP monitor and arterial line). One Ventilated Versus Not Ventilated
(25%–75% IQR⫽9.7–15) and did not additional session was limited by the Of the 50 patients who required
vary significantly by age (P⫽.48) or patient’s discomfort. Consequently, mechanical ventilation, 20 (40%)
sex (P⫽.20). Overall, 50 patients 8 sessions were noted as being received physical therapy while still
(58%) required mechanical ventila- limited. ventilated, for a total of 36 physical
tion for a median of 3.0 days (25%– therapy sessions. The median time to
75% IQR⫽1.0 –12.0), and 35 patients NICU Versus Post-NICU Setting ordering of a physical therapist con-
(41%) required ICP monitoring or Of the 86 patients who were evalu- sultation was similar for patients
treatment. Of these, 21 had external ated by a physical therapist, 75 had requiring mechanical ventilation in
ventricular drains, 4 had ICP moni- physical therapist consultations and the NICU and those who did not
toring devices, 9 had subdural evac- 60 received physical therapy while require mechanical ventilation

Downloaded from https://academic.oup.com/ptj/article/95/7/1006/2686465 by guest on 05 December 2020

uating port systems, and 1 had a lum- in the NICU. The remaining patients in the NICU (1.0 day [25%–75%
bar drain. Twenty-three patients had physical therapist consultations IQR⫽0.0 – 4.0] versus 1.0 day [25%–
(27%) required a craniotomy. The and evaluations after transfer to the 75% IQR⫽0.0 –2.8], P⫽.35), but
median NICU length of stay was 9.0 post-NICU setting but before hospi- patients needing mechanical ventila-
days (25%–75% IQR⫽3.0 –14.0), and tal discharge. The median time to tion started physical therapy later
the median hospital length of stay ordering of a physical therapist con- (5.0 days [25%–75% IQR⫽2.0 –13.5]
was 12.0 days (25%–75% IQR⫽6.0 – sultation was 1.0 day (25%–75% versus 2.0 days [25%–75% IQR⫽1.0 –
20.3). Eight patients (9%) died, 46 IQR⫽0.0 –3.0) after NICU admission, 3.8], P⫽.0001). Patients who
(53%) were discharged to home, 7 and physical therapy sessions started received mechanical ventilation and
(8%) were discharged to a skilled at hospital day 3.0 (25%–75% physical therapy had a lower median
nursing facility, 12 (14%) were dis- IQR⫽2.0 – 6.0). Patients with a GCS GCS at each physical therapy session
charged to a long-term acute care of ⱕ8 started physical therapy later than those who did not receive
setting, 11 (13%) were discharged to than those whose GCS was greater mechanical ventilation (9.3 [25%–
an acute care rehabilitation setting, than 8 (10 days [25%–75% IQR⫽3– 75% IQR⫽6.9 –10.3] versus 14.8
and 3 (3%) were discharged to 19.5] versus 3 days [25%–75% [25%–75% IQR⫽14 –15], P⬍.001).
another acute care hospital. IQR⫽1–5], P⬍.001). In total, there Physical therapy sessions for patients
were 293 physical therapy sessions; needing mechanical ventilation were
Safety 132 occurred in the NICU and 161 not statistically shorter than those
Eighty-six patients received 293 occurred in the post-NICU setting. for patients not needing mechanical
physical therapy sessions. In only 5 Patients received a median of 3.4 ses- ventilation in the NICU (20.0 min-
of these sessions (2.0%) was a sions per week (25%–75% IQR⫽1.8 – utes [25%–75% IQR⫽15.0 –25.0] ver-
change in the patient’s condition or 5.9) once physical therapy was initi- sus 25.0 minutes [25%–75%
vital signs noted. Only 1 of these 5 ated. Patients in the NICU received IQR⫽15.0 –30.0], P⫽.06). Once
sessions (0.3% of total sessions) was less frequent physical therapy ses- physical therapy was initiated,
stopped because of a change in the sions per week than did those in the patients who needed mechanical
patient’s status—in this case, an ele- post-NICU setting (2.1 [25%–75% ventilation received less frequent
vation of the ICP in the NICU. In 3 of IQR⫽1.2–5.1] versus 5.3 [25%–75% physical therapy sessions per week
these 5 sessions (1.0% of total ses- IQR⫽3.5–7.0], P⬍.0001). Overall, in the NICU (1.9 [25%–75%
sions), the patient became hypoxic physical therapy sessions lasted a IQR⫽1.2–3.5] versus 4.7 [25%–75%
and needed increased oxygen deliv- median of 25.0 minutes (25%–75% IQR⫽1.4 –7.0], P⫽.01) and over the
ery via a nasal cannula; 1 of these IQR⫽15.5–30.0) and did not differ entire course of the hospitalization
episodes occurred in the NICU. In 1 significantly in length between the (2.8 [25%–75% IQR⫽1.4 – 4.5] ver-
of these 5 sessions (0.3% of total ses- NICU and the post-NICU setting sus 4.4 [25%–75% IQR⫽2.2–7.0],
sions), the patient became hyperten- (P⫽.12). Physical therapists in the P⫽.01) than those who never
sive in a non-NICU setting; the NICU were more likely to perform required mechanical ventilation
hypertension was treated with oral ROM interventions than those in the (Tab. 2). Physical therapy sessions in
medications. No invasive monitors, post-NICU setting (29% versus 9% of the NICU for patients requiring
airways, or central venous access sessions, P⬍.0001). Standing, trans- mechanical ventilation were more
lines were dislodged during physical ferring, and gait training were less likely to include ROM interventions
therapy. No falls occurred during likely to be performed in the NICU than physical therapy sessions for
physical therapy. In 2 additional ses- than in the post-NICU setting patients who did not need mechani-
sions (0.7%), the patient’s ability to (Fig. 1). cal ventilation (68% versus 13% of
ambulate was limited by equipment sessions, P⬍.0001). All other inter-

July 2015 Volume 95 Number 7 Physical Therapy f 1009

Physical Therapy in Neurological Intensive Care

Downloaded from https://academic.oup.com/ptj/article/95/7/1006/2686465 by guest on 05 December 2020

Figure 1.
Percentage of physical therapy sessions with specific interventions in the neurological intensive care unit (NICU) and in the
post–neurological intensive care unit (Post-NICU) setting. ROM⫽range of motion.

ventions, including bed-based inter- was still in place, for a total of 44 NICU admission, P⫽.04). Patients
ventions, transferring, standing, and physical therapy sessions in the who required ICP monitoring started
ambulating, were more likely to NICU and 1 session with a lumbar physical therapy later in their NICU
occur with patients who did not re- drain in the post-NICU setting. The course than those who did not need
quire mechanical ventilation (Fig. 2). median time to ordering of a physical ICP monitoring in the NICU (5.0
therapist consultation was delayed in days [25%–75% IQR⫽2.0 –15.0] ver-
ICP Monitoring Versus No ICP patients receiving ICP monitoring sus 3.0 days [25%–75% IQR⫽1.0 –
Monitoring relative to those not receiving ICP 5.0] after NICU admission, P⫽.007).
Of the 35 patients who required ICP monitoring in the NICU (1.0 day Individuals who required ICP moni-
monitoring, 23 (66%) received phys- [25%–75% IQR⫽0.0 – 8.0] versus 0.0 toring during physical therapy ses-
ical therapy while the ICP monitor day [25%–75% IQR⫽0.0 –2.0] after sions had a lower median GCS for

Table 2.
Characteristics of Patients in the NICU Stratified by Ventilator Status and ICP Monitoringa

Not Ventilated Ventilated No ICP Monitoring ICP Monitoring

Characteristic (nⴝ36) (nⴝ50) P (nⴝ51) (nⴝ35) P

Initial GCS 15.0 (14.3–15.0) 11.8 (6.9–14.6) ⬍.001 15.0 (14.0–15.0) 10.2 (6.3–14.0) ⬍.001

Mortality, no. (%) of 2 (2.3) 6 (7.0) .30 0 8 (9.3) ⬍.001

86 patients

Discharged to home, 23 (27) 23 (27) .10 35 (40.7) 11 (12.8) ⬍.001

no. (%) of 86 patients

Time to physical therapist 1.0 (0.0–2.8) 1.0 (0.0–4.0) .35 0.0 (0.0–2.0) 1.0 (0.0–8.0) .04
consultation, d

Time to first physical 2.0 (1.0–3.8) 5.0 (2.0–13.5) ⬍.001 3.0 (1.0–5.0) 5.0 (2.0–15.0) .01
therapy session, d

GCS at first physical 14.7 (14.1–15) 9.2 (6.6–10) ⬍.001 14.8 (14.1–15) 9.4 (6.7–12.7) ⬍.001
therapy session

Length of physical therapy 25.0 (15.0–30.0), n⫽96 20.0 (15.0–25.0), n⫽36 .06 25.0 (15.0–30.0), n⫽88 25.0 (15.0–30.0), n⫽44 .41
sessions in NICU, min

Physical therapy sessions/wk 4.4 (2.2–7.0) 2.8 (1.4–4.5) .01 4.0 (1.9–7.0) 2.3 (1.4–4.5) .02
over hospitalization

Physical therapy sessions/wk 4.7 (1.4–7.0), n⫽21 1.9 (1.2–3.5), n⫽39 .01 2.6 (1.2–7.0), n⫽29 1.9 (1.2–3.5), n⫽31 .27
in NICU, n⫽60

Physical therapy sessions/wk 7.0 (3.5–7.0), n⫽28 5.0 (3.5–7.0), n⫽31 .69 7.0 (3.6–7.0), n⫽40 5.0 (1.9–7.0), n⫽19 .26
in post-NICU setting, n⫽59
a
Data are reported as median (25%–75% interquartile range) unless otherwise indicated. NICU⫽neurological intensive care unit, ICP⫽intracranial pressure,
GCS⫽Glasgow Coma Score.

1010 f Physical Therapy Volume 95 Number 7 July 2015

 Physical Therapy in Neurological Intensive Care

Downloaded from https://academic.oup.com/ptj/article/95/7/1006/2686465 by guest on 05 December 2020

Figure 2.
Percentage of physical therapy sessions with specific interventions in the neurological intensive care unit by mechanical ventilation.
ROM⫽range of motion.

each physical therapy session than therapy both in the NICU and ferring interventions (odds ratio⫽
those who did not require ICP mon- over the entire course of the hospi- 0.02; 95% confidence inter-
itoring (12.7 [25%–75% IQR⫽8.4 – talization, while controlling for age, val⫽0.001, 0.16; P⬍.001). For
14.9] versus 14.7 [25%–75% admission GCS, craniotomy, use of patients requiring mechanical venti-
IQR⫽13.2–15.0], P ⬍.001). The dura- mechanical ventilation, use of ICP lation, no physical therapy sessions
tions of physical therapy sessions monitoring, and ability of the patient included ambulation (Tab. 3). Intra-
were similar for patients who to follow instructions in the NICU. cranial pressure monitoring was
required ICP monitoring and those Only the admission GCS remained associated with less ambulation
who did not. Once physical therapy independently associated with less (odds ratio⫽0.18; 95% confidence
was initiated, patients who received frequent physical therapy in the interval⫽0.05, 0.51; P⫽.01).
ICP monitoring had rates of physical NICU (0.24 [25%–75% IQR⫽0.03–
therapy in the NICU that were simi- 0.45], P⫽.02). None of the variables Discussion
lar to those for patients who did not remained independently associated We analyzed physical therapist treat-
receive monitoring (1.9 [25%–75% with overall physical therapy fre- ment of 86 patients admitted to a
IQR⫽1.2–3.5] versus 2.6 [25%–75% quency over the course of the university NICU with SAH, SDH,
IQR⫽1.2–7.0], P⫽.27) but had less hospitalization. ICH, or trauma with regard to safety,
frequent physical therapy over the NICU or post-NICU status, and inva-
course of the hospitalization (2.3 For the physical therapy sessions in sive support and monitoring. First,
[25%–75% IQR⫽1.4 – 4.5] versus 4.0 the NICU, we performed a similar we found that physical therapy can
[25%–75% IQR⫽1.9 –7.0], P⫽.02) analysis, while controlling for age, be safely performed in the NICU.
(Tab. 2). Physical therapy sessions craniotomy, use of mechanical ven- Only a single treatment session (out
with ICP monitoring were more tilation, and use of ICP monitoring, of 293 reviewed) was discontinued,
likely to include ROM interventions to analyze the odds of various inter- secondary to an increase in the ICP.
than those without ICP monitoring ventions being performed. Only the There were no reports of adverse
(43% versus 22% of sessions, 102 physical therapy sessions in events associated with a physical
P⫽.0135). Bed-based interventions which patients could follow instruc- therapy session. Second, physical
were equally likely in physical ther- tions in the NICU were included therapy was performed less fre-
apy sessions with ICP monitoring because patients who could not fol- quently and with a lower intensity in
and those without ICP monitoring. low instructions received only pas- the NICU than in the post-NICU set-
Other interventions, including trans- sive ROM and bed-based interven- ting. Finally, patients requiring
ferring, standing, and walking, were tions. In this analysis, use of mechanical ventilation or ICP moni-
less likely in physical therapy ses- mechanical ventilation remained toring received less frequent and less
sions with ICP monitoring than in independently associated with more intensive physical therapy than
those without ICP monitoring ROM interventions (odds ratio⫽ those who did not require mechani-
(Fig. 3). 20.85; 95% confidence interval⫽ cal ventilation or ICP monitoring.
3.89, 144.50; P⫽.01) and fewer
Multivariate Analysis standing interventions (odds Growing evidence supports early
We performed multivariate analysis ratio⫽0.008; 95% confidence inter- physical therapy in other popula-
to assess the frequency of physical val⫽0.0003, 0.07; P⬍.001) or trans- tions of patients in the NICU, most

July 2015 Volume 95 Number 7 Physical Therapy f 1011

Physical Therapy in Neurological Intensive Care

Downloaded from https://academic.oup.com/ptj/article/95/7/1006/2686465 by guest on 05 December 2020

Figure 3.
Percentage of physical therapy sessions with specific interventions in the neurological intensive care unit by intracranial pressure (ICP)
monitoring. ROM⫽range of motion.

notably in medical and surgical ICUs. able to follow instructions in the hospital day 3. In contrast to
Multiple retrospective and prospec- NICU. Olkowski et al,8 who developed an
tive studies have demonstrated the early mobility program and prospec-
feasibility and safety of early, inten- However, data about physical ther- tively examined safety and feasibility
sive physical therapy in this setting. apy in the NICU are scarce. The pres- for patients who had SAH, we retro-
A recent systemic review of 17 pre- ent, descriptive study, in which the spectively examined physical thera-
vious studies concluded that early characteristics of physical therapist pist practice for a more diverse
physical therapy in this setting could treatment and safety were examined, population of patients (with SAH,
be performed safely and likely furthers the understanding of the SDH, and ICH), who required
resulted in improved outcomes.4,10 current practice in a large medical mechanical ventilation, in the NICU.
To this end, authors have argued for center NICU. The present study adds Additionally, we describe the effects
daily physical therapy in the NICU to the work of Brimioulle et al,9 who of mechanical ventilation and ICP
for patients who are critically ill.11 demonstrated, for 65 people with monitoring on physical therapist
Currently, 2 randomized controlled and without normal ICP, that most practice in the NICU. In general,
trials are actively enrolling patients physical therapy exercises— exclud- early physical therapy in the NICU is
in medical NICUs to assess the ben- ing isometric hip flexion— could associated with adverse events in 1%
efits of early physical therapy.9 The be performed without significant to 16% of physical therapy ses-
participants in these trials are ran- changes in the ICP. Similarly, sions.12 In the present study, only a
domized to receive daily physical Olkowski et al8 demonstrated, for 25 single session of physical therapy
therapy focusing on breathing exer- people with SAH and a low risk for was associated with a change in the
cises, strength exercises, mobility ischemic complications, that physi- ICP, and, in 4 other sessions, patients
activities, and ROM once they are cal therapy could safely be started at

Table 3.
Logistic Regression of Physical Therapy Interventions in the NICU at a Given Sessiona

Odds Ratio (95% CI)

Intervention Craniotomy Mechanical Ventilation ICP Monitoring Overall P

ROM 1.28 (0.31, 5.25) 20.85 (3.89, 144.50) 1.58 (0.40, 5.92) .01

Bed-based interventions 1.90 (0.47, 9.89) 0.63 (0.11, 5.11) 1.02 (0.25, 5.32) .36

Standing 2.73 (0.61, 19.30) 0.008 (0.0003, 0.07) 1.12 (0.28, 5.15) ⬍.001

Transferring 3.25 (0.94, 15.16) 0.02 (0.001, 0.16) 0.58 (0.19, 1.81) ⬍.001

Walking 0.52 (0.21, 1.29) No ambulation for patients who were ventilated 0.18 (0.05, 0.51) .01
(term not included)
a
Age was not significantly associated with any intervention. NICU⫽neurological intensive care unit, CI⫽confidence interval, ICP⫽intracranial pressure,
ROM⫽range of motion.

1012 f Physical Therapy Volume 95 Number 7 July 2015

 Physical Therapy in Neurological Intensive Care

had easily corrected changes in vital not be stable enough for early phys- ical therapy for those patients.
signs. ical therapy. Additionally, although Second, our study was a retrospec-
physical therapy can be performed tive chart review; therefore, we can
Importantly, despite experiences on a patient with a low GCS, we can report only association and not cau-
with early and intensive physical hypothesize that it is performed less sation. Importantly, our data were
therapy in surgical and medical frequently because of the increased limited by the information recorded
NICUs, the present study demon- nurse and physical therapist time in the medical records. Specifically,
strated that earlier and potentially needed to perform physical therapy physical therapist interventions
more intensive physical therapy is with such a patient and insufficient were broadly described in the medi-
not being performed for patients in staff to spend that time. cal records, limiting our ability to
the NICU. We found that physical describe completely the differences

Downloaded from https://academic.oup.com/ptj/article/95/7/1006/2686465 by guest on 05 December 2020

therapy started at a median of hospi- The present study demonstrated that in interventions between groups of
tal day 3 and significantly later in most of the physical therapy per- patients. Third, we could not com-
patients requiring mechanical venti- formed in the NICU consisted of pletely adjust for the impact of the
lation or ICP monitoring. In compar- ROM or bed-based interventions. severity of illness on the ability to
ison, in the Very Early Rehabilitation Functional activities, including gait initiate physical therapy. The physi-
Trial (AVERT) for stroke study, training, were performed less fre- cal therapists did not consistently
patients with stroke were random- quently in the NICU than in the post- document specific reasons for deem-
ized to receive mobilization within NICU setting. Our results stand in ing patients too medically unstable
24 hours of admission rather than contrast to the recent trend toward to participate in physical therapy.
the standard of care.6,7 Similarly, in early mobilization for patients who Presumably, patients needing ICP
studies of early physical therapy in are critically ill; it has been shown monitoring or mechanical ventila-
the medical NICU with positive out- that physical therapy—specifically, tion had more severe illness and may
comes, physical therapy was started ambulation— can be more intensive have been less ideal candidates
as early as 24 to 48 hours after intu- in the NICU than in the post-NICU for physical therapy. Instead, we
bation and focused on respiratory setting.15 Moreover, the present adjusted for age, craniotomy, and
muscle strength, ROM, functional study demonstrated that gait training GCS as indicators of the severity of
mobility training, and exercise and functional activities occurred illness. However, there still may have
training.13,14 less frequently in patients requiring been additional, unknown con-
invasive support in the NICU, includ- founders in our multivariate analysis.
Interestingly, although an admission ing mechanical ventilation or ICP Finally, our data described physical
GCS of 8 or less was associated with monitoring. Although these findings therapist practice at a single aca-
starting physical therapy later in the were likely due, in part, to the sever- demic institution and may not be
hospital course, the admission GCS ity of illness of a given patient in the applicable to other institutions.
and the GCS at each physical therapy NICU, we were unable to evaluate
session were lower in patients need- the specific reasoning behind an We described the current physical
ing mechanical ventilation or ICP individual therapist’s decision to per- therapist practice in the NICU at a
monitoring, suggesting that GCS form certain interventions. Thus, single academic center and demon-
alone was not a barrier to perform- patients in the NICU with, perhaps, strated that physical therapy can be
ing physical therapy. However, the the greatest risk for debilitation safely performed in the NICU. Addi-
admission GCS was independently received less intensive physical ther- tionally, we demonstrated that phys-
associated with less frequent physi- apy interventions than their ical therapy was performed less fre-
cal therapy in the NICU. Together, counterparts. quently and with less intensity in the
these data potentially indicate that NICU than in the post-NICU setting
although the GCS is not a barrier to Our study had multiple limitations. and was performed less frequently
initiating an individual physical ther- First, an inherent bias was intro- and with less intensity in patients
apy session, it does influence the tim- duced by the inclusion in our analy- requiring mechanical ventilation or
ing of physical therapy and the fre- sis of only 86 of 180 patients who ICP monitoring in the NICU than in
quency at which physical therapy received physical therapy instead of those who did not require such sup-
can be performed. A low GCS, all those admitted to the NICU. Data port. Despite these findings, a better
together with requirements for were not obtained for patients who understanding about the safety of
mechanical ventilation and ICP mon- did not have a physical therapist con- starting physical therapy earlier in
itoring, is a marker of disease sever- sultation, and we are unable to spec- the NICU is necessary. Although we
ity and suggests that a patient may ulate on the appropriateness of phys- demonstrated that the current inten-

July 2015 Volume 95 Number 7 Physical Therapy f 1013

Physical Therapy in Neurological Intensive Care

sity of physical therapy was safe and International Conference; May 16 –21, 8 Olkowski BF, Devine MA, Slotnick LE,
2014; San Diego, California. et al. Safety and feasibility of an early mobi-
feasible, further study is needed to lization program for patients with aneurys-
investigate whether the current Funding was provided by National Institutes mal subarachnoid hemorrhage. Phys Ther.
2013;93:208 –215.
intensity of therapy is sufficient to of Health grants R01 NR011051 and K24
HL089223. 9 Brimioulle S, Moraine JJ, Norrenberg D,
improve outcomes. Similarly, the Kahn RJ. Effects of positioning and exer-
necessary elements of physical ther- DOI: 10.2522/ptj.20140112 cise on intracranial pressure in a neurosur-
gical intensive care unit. Phys Ther. 1997;
apist interventions have not been 77:1682–1689.
established. A better understanding 10 Perme C, Chandrashekar R. Early mobility
References
of cognitive and perceptual deficits 1 Morris PE, Herridge MS. Early intensive
and walking program for patients in inten-
sive care units: creating a standard of care.
as well as the therapies for treating care unit mobility: future directions. Crit Am J Crit Care. 2009;18:212–221.
these deficits is needed. A random- Care Clin. 2007;23:97–110.
11 Schweickert WD, Pohlman MC, Pohlman

Downloaded from https://academic.oup.com/ptj/article/95/7/1006/2686465 by guest on 05 December 2020

ized controlled trial is needed to 2 Garzon-Serrano J, Ryan C, Waak K, et al. AS, et al. Early physical and occupational
Early mobilization in critically ill patients: therapy in mechanically ventilated, criti-
evaluate these issues. patients’ mobilization level depends on cally ill patients: a randomised controlled
health care provider’s profession. PM&R. trial. Lancet. 2009;373:1874 –1882.
2011;3:307–313.
12 Nordon-Craft A, Moss M, Quan D, Schen-
Dr Sottile, Dr Nordon-Craft, Dr Malone, Dr 3 Li Z, Peng X, Zhu B, et al. Active mobiliza- kman M. Intensive care unit–acquired
Schenkman, and Dr Moss provided concept/ tion for mechanically ventilated patients: a weakness: implications for physical thera-
systematic review. Arch Phys Med Reha- pist management. Phys Ther. 2012;92:
idea/research design. Dr Sottile, Dr Nordon- bil. 2013;94:551–561. 1494 –1506.
Craft, Dr Schenkman, and Dr Moss provided
4 van Gijn J, Kerr RS, Rinkel GJ. Subarach- 13 Morris PE, Goad A, Thompson C, et al.
writing. Dr Sottile and Dr Moss provided noid haemorrhage. Lancet. 2007;369: Early intensive care unit mobility therapy
data collection and analysis. Dr Moss pro- 306 –318. in the treatment of acute respiratory fail-
vided project management, fund procure- 5 Kitchener N, Hashem S, Wahba M, et al. ure. Crit Care Med. 2008;36:2238 –2243.
ment, facilities/equipment, institutional liai- Critical Care in Neurology. Flying Pub- 14 Pohlman MC, Schweickert WD, Pohlman
sons, and administrative support. Dr lisher & Kamps; 2012. AS, et al. Feasibility of physical and occu-
Malone, Dr Luby, and Dr Moss provided 6 Cumming TB, Thrift AG, Collier JM, et al. pational therapy beginning from initiation
consultation (including review of manuscript Very early mobilization after stroke fast- of mechanical ventilation. Crit Care Med.
tracks return to walking: further results 2010;38:2089 –2094.
before submission).
from the phase II AVERT randomized con- 15 Hopkins RO, Miller RR, Rodriguez L, et al.
This study was approved by the Colorado trolled trial. Stroke. 2011;42:153–158. Physical therapy on the wards after early
Multiple Institutional Review Board 7 Bernhardt J, Dewey H, Thrift A, et al. A physical activity and mobility in the inten-
very early rehabilitation trial for stroke sive care unit. Phys Ther. 2012;92:1518 –
(COMIRB).
(AVERT): phase II safety and feasibility. 1523.
A poster discussion of this research was pre- Stroke. 2008;39:390 –396.
sented at the American Thoracic Society

1014 f Physical Therapy Volume 95 Number 7 July 2015