n Feature Article
Orthopedic Inpatients’ Ability to Accurately
Reproduce Partial Weight Bearing Orders
Serena Yu, B Physiotherapy; Tony McDonald, BAppSc (Physio);
Christabel Jesudason, BAppSc (Physio), MHlthSc; Kathy Stiller, BAppSc (Physio), PhD;
Thomas Sullivan, BMa, CompSc (Hons), BSSc
abstract
Full article available online at Healio.com/Orthopedics. Search: 20131219-10
Partial weight bearing is often prescribed for patients with orthopedic injuries. Patients’
ability to accurately reproduce partial weight bearing orders is variable, and its impact on
clinical outcomes is unknown. This observational study measured patients’ ability to reproduce partial weight bearing orders, factors influencing this, patients’ and physiotherapists’
ability to gauge partial weight bearing accuracy, and the effect of partial weight bearing accuracy on long-term clinical outcomes. Fifty-one orthopedic inpatients prescribed partial
weight bearing were included. All received standard medical/nursing/physiotherapy care.
Physiotherapists instructed patients in partial weight bearing using the hand-under-foot,
bathroom scales, and/or verbal methods of instruction. Weight bearing was measured on
up to 3 occasions during hospitalization using a force-sensitive insole. Factors that had the
potential to influence partial weight bearing accuracy were recorded. Patients and their
physiotherapists rated their perception of partial weight bearing accuracy. Three-month
clinical follow-up data were retrieved from medical records. The majority of patients (72%
or more) exceeded their target load, with mean peak weight bearing as high as 19.3 kg over
target load (285% of target load). Weight bearing significantly increased over the 3 measurement occasions (P<.001) and was significantly associated with greater body weight
(P=.04). Patients and physiotherapists were unable to accurately gauge partial weight bearing accuracy. The incidence of clinically important complications at 3 months was 9% and
not significantly associated with partial weight bearing accuracy during hospitalization
(P≥.45). Patients are unable to accurately reproduce partial weight bearing orders when
trained with the hand-under-foot, bathroom scales, or verbal methods of instruction.
The authors are from the Physiotherapy Department (SY, TM, CJ, KS), Royal Adelaide Hospital,
North Terrace; and the Data Management and Analysis Centre (TS), Discipline of Public Health,
University of Adelaide, Adelaide, Australia.
The authors have no relevant financial relationships to disclose.
A Royal Adelaide Hospital Allied Health Research Grant was used to purchase the SmartStep equipment
and software. The authors thank the patients who participated in this study; Ms Naomi Haensel (Director,
Physiotherapy); and Mr George Potter (Head of Clinical Services, Orthopaedic and Trauma Service).
Correspondence should be addressed to: Tony McDonald, BAppSc (Physio), Physiotherapy Department,
Royal Adelaide Hospital, North Terrace, Adelaide, Australia 5000 (tony.mcdonald@health.sa.gov.au).
Received: July 21, 2013; Accepted: July 29, 2013; Posted: January 15, 2014.
doi: 10.3928/01477447-20131219-10
e10
ORTHOPEDICS | Healio.com/Orthopedics
n Feature Article
R
estricted or partial weight bearing
is often prescribed during the rehabilitation of patients with pelvic
or lower limb fractures following trauma
or elective orthopedic surgery, with the
aim of protecting the healing bone and/or
surgical construct while providing a stimulus for bone growth. Bone remodeling is
promoted by dynamic (as opposed to static) loading, and loading in discrete bouts
separated by recovery periods is more effective than continuous cyclic loading.1
Bony or implant failure can occur from
a single excessively high load or from repetitive loading above a tolerance point.2
In clinical practice, the maximum amount
of weight bearing, or target load, is usually prescribed by an orthopedic surgeon,
taught to the patient by a physiotherapist,
and achieved by the use of crutches or a
walking frame. Methods for instructing
patients in partial weight bearing include
simple techniques such as the hand-underfoot method and bathroom scales, or more
complex methods such as force platforms
and ambulatory devices that use a forcesensitive insole.
The ability to reproduce partial weight
bearing orders has been investigated in
both healthy subjects and patient samples.
In studies involving healthy subjects, the
ability to reproduce partial weight bearing
orders has been variable.3-13 For example,
Malviya et al10 found that 12 healthy subjects were able to reproduce partial weight
bearing orders immediately and 1 hour
after receiving instruction using bathroom
scales, whereas Dabke et al3 found that 6
healthy subjects were unable to reproduce
partial weight bearing orders when measured a few days after an initial instruction
period using bathroom scales, with 4 overshooting the prescribed limit by a mean of
27% of body weight and 2 undershooting
the limit by a mean of 12% of body weight.
The method of instruction in partial weight
bearing appears to influence partial weight
bearing accuracy in healthy subjects, with
lower levels of accuracy found for partial
weight bearing training using the hand-
JANUARY 2014 | Volume 37 • Number 1
under-foot or bathroom scales techniques
and improved accuracy when biofeedback
devices are used.5-7 No factors associated
with reduced partial weight bearing accuracy have been consistently identified
in healthy subjects, but single studies
published after the commencement of the
current study have reported reduced accuracy being associated with the male sex,6
increased body weight,7 and greater body
mass index.6 The effect of the time that has
elapsed since training on partial weight
bearing accuracy is unclear: Hustedt et
al8 found that partial weight bearing accuracy had not significantly changed 24
hours after training for 12 healthy subjects,
whereas all 6 healthy subjects in the study
by Dabke et al3 were unable to reproduce
partial weight bearing orders a few days
after the initial period of instruction.
In studies involving patient samples,
the ability to reproduce partial weight
bearing orders has also been shown to be
variable.11,14-21 As an example, Vasarhelyi
et al11 found that maximum ground forces
were almost 3 times the target load in elderly patients (older than 60 years) and
over 3 times the target load in younger patients. As with healthy subjects, improved
partial weight bearing accuracy was demonstrated when biofeedback devices were
used for training.15,18,19 Factors that were
significantly associated with weight bearing above the target load in patient samples were variable between studies and
included female sex,17 being elderly,11,20
increased body weight,20 certain psychomotor skills,20 greater total walking time,
less pain, and greater anxiety.17 The effect
of time on patients’ ability to reproduce
partial weight bearing orders was similar to that seen in healthy subjects, with
reduced accuracy and increased weight
bearing documented over time, particularly after discharge from hospital.16,18,19 An
additional finding, reported by Hurkmans
et al,22 was that physiotherapists were unable to accurately visually estimate patients’ partial weight bearing accuracy,
irrespective of their level of experience.
Although it seems that the ability of
both patients and healthy subjects to accurately reproduce partial weight bearing
orders is variable, the clinical importance
of this finding is unclear. No studies were
identified that investigated whether an
inability to accurately reproduce partial
weight bearing orders adversely affected
clinical outcomes in a patient sample.
The aims of the current observational
study were to measure whether orthopedic inpatients were able to accurately
reproduce partial weight bearing orders,
to investigate factors that might predict
patients’ accuracy at reproducing partial weight bearing orders, to investigate
patients’ and physiotherapists’ ability to
gauge partial weight bearing accuracy,
and to investigate whether the ability to
accurately reproduce partial weight bearing orders affected long-term clinical outcomes.
Materials and Methods
This observational study was approved by the Royal Adelaide Hospital
Research Ethics Committee, registered
with the Australian New Zealand Clinical
Trials Registry, and supported by a Royal
Adelaide Hospital Allied Health Research
Grant.
Setting and Patients
The study was conducted within the
Orthopaedic and Trauma Service at the
Royal Adelaide Hospital, a 650-bed, tertiary-care, urban public hospital in Australia,
between February 2012 and September
2012. The Orthopaedic and Trauma Service
has 82 designated beds with approximately
4300 admissions per year.
Potential participants were inpatients
admitted to the Orthopaedic and Trauma
Service with an orthopedic injury/condition where the management included the
prescription of partial weight bearing. The
need for partial weight bearing and target
load was determined by each patient’s orthopedic surgeon and documented in the
medical records. Patients were excluded
e11
n Feature Article
from participation if they refused consent.
Patients were withdrawn from the study
if their weight bearing status changed to
non- or full weight bearing, if there was
equipment failure, or if staff were unable
to record data due to a lack of time.
Study Protocol
After satisfying the selection criteria,
potential participants were approached by
one of the investigators regarding participation. The project aims and format were
explained, a written information sheet
provided, and informed written consent
obtained from those patients willing to
participate. Consent for patients who were
cognitively impaired or younger than 18
years was obtained from an appropriate
family member.
All patients received usual medical,
nursing, and physiotherapy care during
and after their hospital admission. From
a physiotherapy perspective, all patients
were assessed by a physiotherapist soon
after admission to hospital and/or surgery
and received interventions based on assessment findings, as per usual practice.
These interventions included exercise
prescription, mobilization, provision of
mobility aids, education, advice, and organization of appropriate community
equipment/resources. Instruction in partial weight bearing during mobilization
was carried out by physiotherapists. On
the first occasion of mobilization, this
involved the hand-under-foot, bathroom
scales, and/or verbal methods of instruction, at the discretion of the treating physiotherapist, until the patient and physiotherapist were confident that the patient
was reproducing the partial weight bearing orders accurately. Further instruction
in partial weight bearing on subsequent
episodes of mobilization was undertaken
as deemed necessary by the physiotherapist. Over the duration of the study, the
physiotherapy care to patients was provided by 4 physiotherapists working in the
Orthopaedic and Trauma Service whose
experience ranged from recent graduates
e12
to senior physiotherapists (more than 15
years of experience).
Outcomes
The primary outcome was patients’
ability to reproduce partial weight bearing orders when mobilizing with physiotherapists during the early period of
hospitalization. This was recorded using
a force-sensitive insole (SmartStep monitor; Andante Medical Devices, Inc, White
Plains, New York). Measurements were
recorded on days 1 and 2 of mobilization
and again between days 4 and 6 of mobilization. On each measurement occasion, data were recorded during the entire
period of mobilization. For the purposes
of this study, the outcomes reported on
each occasion of mobilization were peak
weight bearing and mean weight bearing.
These data were expressed in absolute figures (ie, kg), absolute figures compared
with the target load (ie, kg above or below
target load), and as a percentage of the target load. The SmartStep device can be set
to provide simultaneous audio and visual
feedback; for the purposes of this study,
no feedback was provided to the patients
or the treating physiotherapists. The data
generated by the force-sensitive insole
were received and stored temporarily in
a portable compact wireless control unit
worn around the ankle. Data were relayed
from the control unit to a computer running the SmartStep software.
To investigate factors that may predict patients’ ability to reproduce partial
weight bearing orders, data including sex,
age, weight, body mass index, ability to
understand English, preexisting mobility,
level of assistance required during mobilization, and level of pain were recorded.
The patient’s ability to understand English
was rated by the treating physiotherapist
based on his or her perception gained
during assessment and treatment. Each
patient’s level of assistance required during mobilization was rated by the treating
physiotherapist using the Iowa Level of
Assistance Scale.23 Each patient rated his
or her level of pain using a verbal analog
scale (0=no pain to 10=worst pain imaginable). In addition, each patient and his or
her treating physiotherapist were asked to
independently rate their perceptions of the
patient’s accuracy at reproducing partial
weight bearing orders during each occasion of mobilization using predetermined
descriptive terms (ie, very accurate, reasonably accurate, unsure, somewhat inaccurate, very inaccurate, variable).
To investigate the long-term clinical
impact of patients’ ability to accurately
reproduce partial weight bearing orders,
broad clinical outcomes at 3 months were
retrieved from data routinely recorded in
the medical records by orthopedic surgeons/registrars when patients attend an
outpatient clinic. These outcomes included information about major complications
and radiographic findings.
Sample Size
Vasarhelyi et al11 found that the SD
for the peak force difference from the
preset load to the actual load was 170.4
N (17.4 kg). Using this SD, the authors
determined that a total sample size of 47
patients would allow them to estimate the
mean peak weight difference with a precision of ±5 kg (where precision is defined
as the half-width of a 95% confidence interval).
Data Analyses
Data were checked for accuracy and
cleaned prior to analyses. Frequency
histograms for the weight bearing outcomes were visually inspected and data
were found to be approximately normally
distributed. Changes over time in partial
weight bearing accuracy were assessed
using linear mixed-effects models, with
time included as a categorical fixed effect and patient included as a random effect. Predictors of partial weight bearing
accuracy were also assessed using linear
mixed-effects models, again with patient
included as a random effect and with adjustment for time. All analyses were un-
ORTHOPEDICS | Healio.com/Orthopedics
n Feature Article
dertaken using SAS version 9.3 statistical
software (SAS Institute Inc, Cary, North
Carolina). Statistical significance was assessed at a 2-tailed P value less than .05.
Results
Patients
A total of 60 patients were screened
for eligibility. Of these, 57 (95%) met
the inclusion criteria (3 refused consent).
Six patients were subsequently withdrawn from the study because of insufficient staff time to undertake the required
measurements (n=4), equipment failure
(n=1), or a change in status from partial
weight bearing to nonweight bearing
(n=1). Therefore, data from 51 patients
were analyzed. Baseline demographic and
clinical characteristics of the 51 patients
are shown in Table 1. The sample comprised a nearly even spread of men and
women, with a mean age of 56 years. The
majority of patients were able to comprehend English (n=49 [96%]) and walked
independently without a mobility aid prior to admission (n=34 [67%]). A fall was
the most frequent reason for admission
(n=20 [39%]), a lower limb fracture was
the most frequent lower limb injury (n=26
[51%]), and virtually all patients required
surgery (n=49 [96%]). Intrahospital postoperative complications occurred in 5
(10%) patients (Table 1).
Mobility Data
Descriptive mobility data from the
3 measurement occasions are shown in
Table 2. The majority of patients were
prescribed partial weight bearing with a
target load of 10 kg, mobilized using a
standard walking frame, and trained in
partial weight bearing using bathroom
scales. Most patients required 2 staff
members to assist during mobilization:
moderate to maximal assistance on the
first measurement occasion and minimal
to moderate assistance by the third measurement occasion. Pain levels were variable between patients, with pain most
frequently confined to the affected lower
JANUARY 2014 | Volume 37 • Number 1
limb. The rating most frequently given by
both the patients and the treating physiotherapists to describe their perceptions
of partial weight bearing accuracy was
“reasonably accurate.” The treating physiotherapists rated partial weight bearing
accuracy as “somewhat inaccurate” or
“very inaccurate” more frequently than
the patients (Table 2).
Data from the force-sensitive insole
are summarized in Table 3. Mean values
for peak weight bearing over the 3 measurement occasions ranged from 21.0 to
32.8 kg. On all 3 measurement occasions,
mean values for peak weight bearing exceeded the target load. The highest values
were obtained on the third measurement
occasion, where the peak weight exceeded the target load by a mean of 19.3
kg, representing 285% of the target load.
Peak weight bearing exceeded the target
load for the majority of patients on all 3
measurement occasions. On an individual
basis, peak weight bearing exceeded the
target load by as much as 50 kg, representing 600% of the target load.
Mean weight bearing values over each
occasion of mobilization were considerably lower than the peak weight bearing
data, with means ranging from 8.2 to 15.0
kg. Mean weight bearing exceeded the
target load on the third measurement occasion only, by 2.1 kg, representing 134%
of the target load. Mean weight bearing
exceeded the target load for a minority
of patients on the first 2 measurement occasions, with this figure rising to 50% on
the third measurement occasion. On an
individual basis, mean weight bearing
exceeded the target load by up to 31 kg,
representing 412% of the target load.
The authors examined whether patients
whose peak weight bearing exceeded the
target load also recorded mean weight
bearing in excess of the target load: from
a total of 108 sets of data obtained across
the 3 measurement occasions, peak weight
bearing exceeded the target load 86 times,
with mean weight bearing also exceeding
the target weight on 35 (41%) of these 86
times. The authors also examined whether
the patients who overshot the target load
on the first measurement occasion continued to do so on subsequent occasions:
36 patients overshot the target load on the
first occasion; 24 of these were tested on
the second measurement occasion and 14
on the third measurement occasion, and in
all instances peak weight bearing in excess of the target load was documented.
Factors Predicting Partial Weight Bearing
Accuracy
A significant change over time was
found for peak weight bearing (expressed as a percentage of the target load)
(P=.0003), with partial weight bearing
significantly higher on the third measurement occasion (mean, 285%) compared with both the first (mean, 185%)
and second (mean, 205%) measurement
occasions (P<.001 and P=.005, respectively). Similarly, mean weight bearing
(expressed as a percentage of the target
load) significantly changed over time
(P<.0001), with significantly higher values on the third measurement occasion
(mean, 134%) compared with both the
first (mean, 71%) and second (mean,
83%) measurement occasions (P<.0001
and P=.0006, respectively).
Patients’ weight was significantly associated with partial weight bearing accuracy in terms of peak weight bearing
(expressed as a percentage of the target
load), with every 1-kg increase in weight
associated with a 2% increase in peak
weight bearing (P=.04). For mean weight
bearing, every 1-kg increase in weight
was associated with a 0.8% increase in
mean weight bearing, which approached
statistical significance (P=.053).
No other significant predictors of partial weight bearing accuracy for either
peak or mean weight bearing were found.
Specifically, age (P≥.82), sex (P≥.30),
body mass index (P=.08), preexisting mobility (P≥.72), method of partial
weight bearing training (P≥.50), level of
assistance required during mobilization
e13
n Feature Article
Table 1
Patient Demographic and Clinical Characteristics
Characteristic
26 (51)
Mean age, y
56±24
No. (%)
Type/location of lower limb injury
No. (%)
Female
Mean body mass index, kg/m2
Characteristic
27±6
Fracture
26 (51)
Pelvic
6 (12)
Femur
8 (16)
Tibia
10 (20)
Multiple
Occupation
Retired
23 (45)
Tradesperson/related worker
8 (16)
Clerical, sales, and service
5 (10)
Professional
4 (8)
Pension
3 (6)
Unemployed
3 (6)
2 (4)
Total hip arthroplasty
13 (25)
Primary
2 (4)
Revision
5 (10)
Periprosthetic fracture
4 (8)
Infection
2 (4)
Total knee arthroplasty
5 (10)
2 (4)
Student
3 (6)
Periprosthetic fracture
Laborer/related worker
2 (4)
Infection
2 (4)
Revision
1 (2)
Language
English first language
English comprehension
Congenital condition
48 (94)
49 (96)
Past medical history
Musculoskeletal condition
5 (10)
Hip dysplasia
4 (8)
Hip avascular necrosis
1 (2)
Other
35 (69)
2 (4)
3 (6)
Removal of femoral screws
1 (2)
Cognitive impairment
1 (2)
Tibial sarcoma
1 (2)
Psychiatric condition
5 (10)
Surgery
49 (96)
12 (24)
Intrahospital postoperative complications
5 (10)
Neurological condition
Obesity
Other
17 (33)
Preadmission mobility
Independent without walking aid
34 (67)
Independent with walking aid
15 (29)
Walking frame
10 (20)
Stick
5 (10)
Independent with wheelchair
1 (2)
Dependent with walking frame
1 (2)
20 (39)
Elective surgery
14 (27)
Infection
4 (8)
Motor bike accident
4 (8)
Motor vehicle accident
1 (2)
Pedestrian vs motor vehicle accident
2 (4)
Bicycle accident
1 (2)
Crush injury
2 (4)
Assault
2 (4)
Tumor
1 (2)
e14
1 (2)
Hemarthrosis
1 (2)
Postural hypotension
2 (4)
Foot drop
1 (2)
Postoperative complications at 3-mo follow-up
Primary reason for admission
Fall
Deep vein thrombosis, pulmonary embolus
4 (9)a
Loosening of internal fixation requiring
surgery
2 (4)
Nonunion of fracture requiring surgery
1 (2)
Chronic regional pain syndrome requiring
medication
1 (2)
a
Data missing from 6 patients, so n=45.
(P≥.22), and level
of pain (P≥.11) did
not
significantly
predict
partial
weight bearing accuracy. The effect
of the ability to
understand English
on partial weight bearing accuracy was
unable to be tested because of the low
number of patients unable to understand
English (n=2).
Data pertaining to patients’ and physiotherapists’ perceptions of partial weight
bearing accuracy were dichotomized (as
accurate or inaccurate) due to the small
ORTHOPEDICS | Healio.com/Orthopedics
n Feature Article
No. (%)
Table 2
Descriptive Mobility Data From 3 Measurement
Occasions
First
Occasion
(n=51)
Second
Occasion
(n=38)
Third
Occasion
(n=20)
15 (29)
8 (21)
3 (16)
1 (2)
3 (8)
1 (5)
Unsure
10 (20)
6 (16)
1 (5)
Variable
2 (4)
1 (3)
0 (0)b
Characteristic
Somewhat inaccurate
No. (%)
Very inaccurate
First
Occasion
(n=51)
Second
Occasion
(n=38)
Third
Occasion
(n=20)
10
41 (80)
31 (82)
16 (80)
Very accurate
6 (12)
3 (8)
2 (11)
15
1 (2)
0 (0)
0 (0)
Reasonably accurate
14 (28)
13 (35)
6 (33)
20
4 (8)
4 (11)
1 (5)
Somewhat inaccurate
14 (28)
10 (27)
4 (22)
25
3 (6)
2 (5)
2 (10)
Very inaccurate
10 (20)
6 (16)
4 (22)
30
2 (4)
1 (3)
1 (5)
Unsure
3 (6)
1 (3)
0 (0)
Variable
4 (8)
4 (11)c
2 (11)d
Characteristic
PWB target, kg
Physiotherapists’ rating of PWB accuracy
Walking aid
Standard walking frame
Gutter walking frame
Axillary crutches
35 (69)
26 (68)
13 (65)
4 (8)
3 (8)
1 (5)
12 (24)
7 (18)
4 (20)
Elbow crutches
0 (0)
1 (3)
1 (5)
Gutter crutches
0 (0)
1 (3)
1 (5)
Hand-under-foot
14 (27)
12 (32)
8 (40)
Scales
36 (71)
20 (53)
5 (25)
Verbal
10 (20)
10 (26)
10 (50)
Abbreviation: PWB, partial weight bearing.
a
Some patients had more than 1 method of instruction, hence
n>sample size.
b
Data missing for 1 patient, hence n=19.
c
Data missing for 1 patient, hence n=37.
d
Data missing for 2 patients, hence n=18.
Main training methodsa
No. of staff assisting
1
2 (4)
7 (18)
5 (25)
2
49 (96)
31 (82)
15 (75)
Independent
3 (6)
2 (5)
2 (10)
Standby (nearby
supervision)
2 (4)
2 (5)
3 (15)
Minimal (1 point of
contact)
11 (22)
7 (18)
5 (25)
Moderate (2 points of
contact)
17 (33)
18 (47)
7 (35)
Maximal (≥3 points of
contact)
17 (33)
8 (21)
1 (5)
1 (2)
1 (3)
2 (10)
4.9±3.0
(0-10)
4.7±2.8
(0-9)
4.0±3.0
(0-8)
42 (82)
32 (84)
15 (75)
4 (8)
2 (5)
1 (5)
2 (4)
1 (3)
1 (5)
21 (41)
19 (50)
13 (68)
Iowa Level of Assistance Scale
Failed
Mean pain severity (range)
Location
Lower limb only
Upper and lower limb
Patients’ rating of PWB accuracy
Very accurate
Reasonably accurate
JANUARY 2014 | Volume 37 • Number 1
sample size. No
significant differences (P≥.30) were
found between patients who thought
they were accurate
compared
with
those who thought
they were inaccurate for either peak
or mean weight
bearing (expressed
as a percentage of
the target load).
Similarly, no significant difference
(P=.93) was found
between patients
whom the physiotherapists thought
were accurate and
those whom the
physiotherapists
thought were inaccurate for peak
weight
bearing
(expressed as a
percentage of the target load). However,
for mean weight bearing, patients whom
the physiotherapists thought were accurate recorded a mean of 94% of the target load compared with 120% for those
whom the physiotherapists thought were
inaccurate, which approached statistical
significance (P=.053).
Clinical Outcomes at 3 Months
Three-month follow-up clinical data
were available for 45 (88%) of the 51
patients, with 6 patients lost to follow-up
(Table 1). Four (9%) of the 45 patients had
complications that could have potentially
resulted from an inability to accurately reproduce partial weight bearing orders: 2
had loosening of fracture internal fixation
requiring further surgery, 1 had nonunion
at a fracture site requiring further surgery,
and 1 developed chronic regional pain
syndrome. Peak weight bearing for these
4 patients exceeded their target load on
all 3 measurement occasions (n=3 on the
second and third measurement occasions),
with values ranging from 120% to 500%
of the target load recorded. Mean weight
e15
n Feature Article
bearing for these 4 patients was over the
target load for 1 patient only on each of
the 3 measurement occasions, with values
ranging from 62% to 192% of the target
load documented. Linear mixed-effects
modeling revealed no significant association between the presence of a clinically important complication at 3-month
follow-up and partial weight bearing accuracy during hospitalization (P≥.45).
Discussion
This observational study investigated
the ability of orthopedic inpatients to accurately reproduce partial weight bearing
orders, factors that may predict the ability
to reproduce partial weight bearing orders, patients’ and physiotherapists’ ability to gauge partial weight bearing accuracy, and whether the ability to accurately
reproduce partial weight bearing orders
affected long-term clinical outcomes. The
authors found that the majority of patients
were unable to accurately reproduce partial weight bearing orders and exceeded
their target load, with mean values for
peak weight bearing being as high as
19.3 kg over the target load (representing
285% of target load) and individual values as much as 50 kg over the target load
(600% of target load). Partial weight bearing accuracy changed significantly over
the 3 measurement occasions, with higher
values for peak and mean weight bearing
found on the third measurement occasion. Although partial weight bearing accuracy was not able to be predicted based
on age, sex, body mass index, preexisting
mobility, method of partial weight bearing training, level of assistance required
during mobilization, or level of pain, patients’ weight affected partial weight bearing accuracy, with greater weight significantly associated with higher peak weight
bearing. The patients and treating physiotherapists were unable to accurately gauge
partial weight bearing accuracy. Despite
patients’ inability to accurately reproduce
partial weight bearing orders during the
period of hospitalization, the incidence of
e16
Table 3
Data From the Force-sensitive Insole for 3 Measurement Occasions
Mean (95% CI)
First
Occasion
(n=50)a
Second
Occasion
(n=38)
Third
Occasion
(n=20)
Absolute, kg
21.0
(16.9 to 25.1)
23.2
(18.9 to 27.4)
32.8
(27.3 to 38.3)
Compared with target load, kg
+8.4
(4.1 to 12.7)
+10.8
(6.4 to 15.2)
+19.3
(13.6 to 25.0)
Compared with target load, %
185
(145 to 225)
205
(162 to 249)
285
(225 to 346)
36 (72)
30 (79)
20 (100)
Absolute, kg
8.2
(6.4 to 10.0)
9.4
(7.7 to 11.2)
15.0
(10.9 to 19.2)
Compared with target load, kg
-4.4
(-6.5 to -2.2)
-2.9
(-5.2 to -0.6)
+2.1
(-3.1 to 7.3)
Compared with target load, %
71
(55 to 87)
83
(65 to 101)
134
(90 to 178)
11 (22)
14 (37)
10 (50)
Characteristic
Peak weight bearing
No. (%) of patients over target load
Mean weight bearing
No. (%) of patients over target load
Abbreviation: CI, confidence interval.
a
Data were lost for 1 patient, hence n=50.
long-term complications that could have
potentially arisen from excessive weight
bearing was less than 10% and not significantly associated with partial weight bearing accuracy during hospitalization.
This study’s results regarding patients’ inability to accurately reproduce
partial weight bearing orders concur with
previous studies.11,14,16,18,20,21 Also similar to previous research, the current authors found no learning effect over time,
instead finding that weight bearing above
the target load progressively and significantly increased over the 3 measurement
occasions.11,16,18,19,24 In terms of factors
that influenced partial weight bearing accuracy, this study’s finding that increased
body weight is associated with higher
weight bearing concurs with the recent
results of Hustedt et al7 and Ruckstuhl
et al20 and is particularly important given the increasing prevalence of obesity
worldwide. However, unlike previous
studies, the current authors did not find
that partial weight bearing accuracy was
significantly affected by sex,6,17 age,11,20
or body mass index.6 Neither patients nor
physiotherapists were able to accurately
gauge partial weight bearing accuracy,
which confirms and extends the findings
of Hurkmans et al.22
The current study was the first to investigate whether an inability to accurately
reproduce partial weight bearing orders
was associated with poorer long-term
clinical outcomes. No such association
was found. However, these results should
be interpreted with caution. It is possible
that there are subgroups of patients (eg,
with a specific profile or specific lower
limb diagnosis [eg, anatomical site, fracture pattern, fracture fixation]) where accurate partial weight bearing is vital in
terms of long-term clinical outcomes.
One of the major limitations of the
current study is that partial weight bear-
ORTHOPEDICS | Healio.com/Orthopedics
n Feature Article
ing accuracy was only measured on 3 occasions at most. Although more frequent
data collection during the period of hospitalization may have been desirable, this
was impractical because of the amount of
time required to collect the data from the
force-sensitive insole. The authors chose
to measure partial weight bearing accuracy on the first and second days of mobilization to provide an indication of initial
partial weight bearing accuracy and a few
days later to see if partial weight bearing
accuracy changed over time. Although
there was a substantial loss to follow-up
from the first to third measurement occasions, this most often resulted from patients being discharged from the hospital
and was thus unavoidable. Measurement
of partial weight bearing accuracy following discharge from the hospital would
have been desirable but was beyond the
scope of the current study.
It is acknowledged that the 3-month
follow-up data collected provided, at best,
only a broad indication of clinical outcome. Nevertheless, the authors believe
these data should have revealed patients
who experienced a clinically important
loss of fracture reduction, nonunion, and/
or implant loosening, which may have resulted from excessive loading. However,
it is important to acknowledge that the
authors did not control or measure partial
weight bearing accuracy following discharge from hospital. Despite these limitations, the authors believe the 3-month
follow-up data provide important preliminary data evaluating the potential effect of
an inability to accurately reproduce partial
weight bearing orders on long-term clinical outcome.
The authors acknowledge that the
sample was heterogeneous with respect
to factors such as preadmission mobility
and target load. However, they elected to
include a wide range of patients because
this is reflective of their usual clinical
workload. Future research involving more
homogeneous samples may provide further insight into partial weight bearing
JANUARY 2014 | Volume 37 • Number 1
accuracy, factors predicting partial weight
bearing accuracy, and effect on long-term
clinical outcomes.
This study’s findings support previous data demonstrating that patients
are unable to accurately reproduce partial weight bearing orders, at least when
trained using the hand-under-foot, bathroom scales, and/or verbal methods of instruction, and it is difficult to predict who
will be inaccurate at partial weight bearing. Furthermore, neither the patient nor
the treating physiotherapist appear able to
accurately gauge partial weight bearing
accuracy.
Based on these findings, the authors’
clinical recommendation is that for patients in whom accurate partial weight
bearing is deemed critical, the handunder-foot, bathroom scales, and verbal
methods of training should not be used.
Although previous research shows biofeedback devices offer advantages in
terms of short-term partial weight bearing accuracy, they are expensive and time
consuming to use, and retention of partial
weight bearing accuracy after an initial
period of training appears to deteriorate.
Hence, more intensive and prolonged
training using biofeedback devices, or
even continual use of biofeedback devices
for the duration of partial weight bearing,
may be required. Alternatively, in patients
in whom accurate partial weight bearing
is deemed critical, modification of partial
weight bearing orders to lessen the target
load may be necessary to avoid excessive
loading, particularly in heavier patients.
Research should be undertaken to
explore whether patients’ inability to reproduce partial weight bearing orders
improves or deteriorates further following discharge from hospital and to investigate whether an inability to accurately
reproduce partial weight bearing orders
adversely affects long-term clinical outcomes. The effect of inaccurate partial
weight bearing on outcomes such as fracture fragment movement (eg, using techniques such as radiostereometric analysis)
in specific diagnostic categories would be
of interest; however, it is acknowledged
that such studies may quickly become irrelevant as surgical practices continue to
evolve.
Conclusion
This study’s findings suggest that
patients are unable to accurately reproduce partial weight bearing orders when
trained with the hand-under-foot, bathroom scales, or verbal methods of instruction. Partial weight bearing increased over
time, and greater body weight was associated with partial weight bearing inaccuracy. Patients and physiotherapists were
unable to accurately gauge partial weight
bearing accuracy. The effect of partial
weight bearing inaccuracy on long-term
clinical outcomes is unclear.
References
1.Lee CS, Szczesny SE, Soslowsky LJ.
Remodeling and repair of orthopedic tissue:
role of mechanical loading and biologics:
Part II. Cartilage and bone. Am J Orthop.
2011; 40(3):122-128.
2. Hustedt JW, Blizzard DJ, Baumgaertner MR,
Leslie MP, Grauer JN. Current advances in
training orthopaedic patients to comply with
partial weight-bearing instructions. Yale J
Biol Med. 2012; 85(1):119-125.
3. Dabke HV, Gupta SK, Holt CA, O’Callaghan
P, Dent CM. How accurate is partial weightbearing? Clin Orthop Relat Res. 2004;
(421):282-286.
4. Eisele R, Weickert E, Eren A, Kinzl L. The
effect of partial and full weight-bearing on
venous return in the lower limb. J Bone Joint
Surg Br. 2001; 83(7):1037-1040.
5. Gray FB, Gray C, McClanahan JW. Assessing
the accuracy of partial weight-bearing instruction. Am J Orthop. 1998; 27(8):558-560.
6. Hustedt JW, Blizzard DJ, Baumgaertner MR,
Leslie MP, Grauer JN. Effect of age on partial
weight-bearing training. Orthopedics. 2012;
35(7):e1061-e1067.
7. Hustedt JW, Blizzard DJ, Baumgaertner MR,
Leslie MP, Grauer JN. Is it possible to train
patients to limit weight bearing on a lower
extremity? Orthopedics. 2012; 35(1):e31e37.
8.Hustedt JW, Blizzard DJ, Baumgaertner
MR, Leslie MP, Grauer JN. Lower-extremity
weight-bearing compliance is maintained
over time after biofeedback training.
Orthopedics. 2012; 35(11):e1644-e1648.
e17
n Feature Article
9. Krause D, Wünnemann M, Erlmann A, et al.
Biodynamic feedback training to assure learning partial load bearing on forearm crutches.
Arch Phys Med Rehabil. 2007; 88(7):901-906.
15. Hershko E, Tauber C, Carmeli E. Biofeedback
versus physiotherapy in patients with partial weight-bearing. Am J Orthop. 2008;
37(5):E92-E96.
10. Malviya A, Richards J, Jones RK, Udwadia
A, Doyle J. Reproducibility of partial weight
bearing. Injury. 2005; 36(4):556-559.
11.Vasarhelyi A, Baumert T, Fritsch C,
Hopfenmuller W, Gradl G, Mittlmeier T.
Partial weight bearing after surgery for fractures of the lower extremity: is it achievable?
Gait Posture. 2006; 23(1):99-105.
16.Hurkmans HL, Bussmann JB, Selles RW,
Benda E, Stam HJ, Verhaar JA. The difference between actual and prescribed weight
bearing of total hip patients with a trochanteric osteotomy: long-term vertical force
measurements inside and outside the hospital. Arch Phys Med Rehabil. 2007; 88(2):200206.
12. Winstein CJ, Pohl PS, Cardinale C, Green
A, Scholtz L, Waters CS. Learning a partial-weight-bearing skill: effectiveness of
two forms of feedback. Phys Ther. 1996;
76(9):985-993.
17.Hurkmans HL, Bussmann JB, Benda E,
Haisma JA, Verhaar JA, Stam HJ. Predictors
of partial weight bearing performance after
total hip arthroplasty. J Rehabil Med. 2010;
42(1):42-48.
13.Youdas JW, Kotajarvi BJ, Padgett DJ,
Kaufman KR. Partial weight-bearing gait using conventional assistive devices. Arch Phys
Med Rehabil. 2005; 86(3):394-398.
18.Hurkmans HL, Bussmann JB, Benda E,
Verhaar JA, Stam HJ. Effectiveness of audio feedback for partial weight-bearing in
and outside the hospital: a randomized controlled trial. Arch Phys Med Rehabil. 2012;
93(4):565-570.
14.Ebert JR, Ackland TR, Lloyd DG, Wood
DJ. Accuracy of partial weight bearing after
autologous chondrocyte implantation. Arch
Phys Med Rehabil. 2008; 89(8):1528-1534.
e18
19. Pataky Z, De León Rodriguez D, Golay A,
Assal M, Assal JP, Hauert CA. Biofeedback
training for partial weight bearing in patients
after total hip arthroplasty. Arch Phys Med
Rehabil. 2009; 90(8):1435-1438.
20.Ruckstuhl T, Osterhoff G, Zuffellato M,
Favre P, Werner CM. Correlation of psychomotor findings and the ability to partially
weight bear. Sports Med Arthrosc Rehabil
Ther Technol. 2012; 4:6.
21. Tveit M, Karrholm J. Low effectiveness of
prescribed partial weight bearing. J Rehabil
Med. 2001; 33(1):42-46.
22.Hurkmans HL, Bussmann JB, Benda E.
Validity and interobserver reliability of visual
observation to assess partial weight-bearing.
Arch Phys Med Rehabil. 2009; 90(2):309313.
23. Shields RK, Enloe LJ, Evans RE, Smith KB,
Steckel SD. Reliability, validity, and responsiveness of functional tests in patients with
total joint replacement. Phys Ther. 1995;
75(3):169-179.
24. Aranzulla PJ, Muckle DS, Cunningham JL.
A portable monitoring system for measuring
weight-bearing during tibial fracture healing.
Med Eng Phys. 1998; 20(7):543-548.
ORTHOPEDICS | Healio.com/Orthopedics