/clinical/,/clinical/cckm-tools/,/clinical/cckm-tools/content/,/clinical/cckm-tools/content/questionnaires/,/clinical/cckm-tools/content/questionnaires/related/,

/clinical/cckm-tools/content/questionnaires/related/name-97124-en.cckm

201606168

page

100

UWHC,UWMF,

Tools,

Clinical Hub,UW Health Clinical Tool Search,UW Health Clinical Tool Search,Questionnaires,Related

Assessing Activity in Joint Replacement Patients

Assessing Activity in Joint Replacement Patients - Clinical Hub, UW Health Clinical Tool Search, UW Health Clinical Tool Search, Questionnaires, Related


The Journal of Arthroplasty Vol. 13 No. 8 1998
Assessing Activity in Joint
Replacement Patients
Christopher A. Zahiri, MD,* Thomas P. Schmalzried, MD,* t
Edward S. Szuszczewicz, MD,* and Harlan C. Amstutz, MDt
Abstract: Outcome evaluations of lower extremity joint reconstructions should
include an assessment of patient activity. In vivo wear assessments of total joint
prostheses should be based on a measure of use, not time in situ or a proxy such as age
or gender; however, clinicians lack a simple method to reliably assess the activity of
patients with joint replacement. The modern pedometer can be a satisfactory means
of quantifying the use of lower extremity joints. The pedometer, however, requires
special effort on the part of the physician or evaluator and the patient. Therefore, we
compared the quantitative assessment of walking activity of 100 total joint replace-
ment patients, as measured with a pedometer, to the UCLA activity score and a simple
visual analog scale that can easily be employed during a routine office evaluation.
Both the UCLA activity rating (P = .002) and the visual analog scale rating of the
investigator (P = .00001) had a strong correlation with the average steps per day as
recorded by the pedometer. There was, however, up to a 15-fold difference in the
average steps per day for individual patients with the same UCLA score. The visual
analog scale as rated by the patients of their own activity did not have as strong a
correlation with the pedometer data (P = .08) as did patient age (P = .049). For
practical reasons, the pedometer is probably best reserved for the evaluation of
extreme cases of activity (or inactivity). This study indicates that both the UCLA
activity rating and the investigator visual analog scale are valid for routine activity
assessment in a clinical setting. Adjustments of the UCLA activity score for the
frequency and intensity of activity, as can be done with the investigator visual analog
scale, increase the accuracy of the activity rating. Key words: total joint replacement,
activity, pedometer, wear.
For more than 30 years of joint arthroplasty, patient
activity has been a recognized variable in the clinical
outcome equation \[1-6\]. Commonly used outcome
evaluations \[5,7-10\], however, assess only what a
patient is capable of doing and do not assess what
the patient actually does. As total joint replacement
technology has been applied to a wide range of
From the *Harbor-UCLA Medical Center, Division of Orthopaedic
Surgery, Torrance; and t Joint Replacement Institute at Orthopaedic
Hospital, Los Angeles, California.
Supported by the Los Angeles Orthopaedic Foundation, CA.
Reprint requests: Thomas P. Schmalzried, MD, Joint Replace-
ment Institute at Orthopaedic Hospital, 2400 South Flower
Street, Los Angeles, CA 90007.
Copyright © 1998 by Churchill Livingstone ®
0883-5403/1308-000853.00/0
patients with highly variable activities, assessment
of patient activity becomes imperative.
Walking is the most important single physical
activity that affects the wear of knee and hip
prostheses \[3\]. Assessment of patient activity
using an electronic digital pedometer has been
reported \[11\]. The accuracy of the pedometer used
was determined to be >92% for a variety of ambu-
latory activities. The i 11 patients in the study, who
were all independent community ambulators, aver-
aged about 0.9 million cycles per year for each lower
extremity joint. The most important finding was the
substantial variability in individual activity, ranging
from 395 steps per day to 17,718 steps per day (a
45-fold difference). Patients under 60 years of age
890

were about 30% more active on average than
patients 60 years of age or older (P = .023). Men
were about 30% more active on average than
women (P = .037).
The modern pedometer can be a satisfactory
means of quantifying the use of lower extremity
joints \[12\]. The pedometer requires special effort,
however, on the part of the physician or evaluator
and the patient. The evaluator needs to instruct and
oversee the correct use of the device, and the patient
has to comply with proper use of the device.
Therefore, routine use of a pedometer may be
impractical in many clinical settings. The goal of the
present study was to compare the quantitative
assessment of walking activity of total joint replace-
ment patients, as measured with a pedometer, with
3 other methods of assessing patient activity that
can easily be performed during a routine office
evaluation.
Materials and Methods
From our study of quantitative assessment of
walking activity in l lI patients who had joint
replacement \[11\], complete sets of data on all 3
activity rating scales were available for 100 of these
patients. Criteria for inclusion of patients were as
follows \[1 i\]: i) interested patients, ii) patients who
were at least 6 months postoperative from lower
extremity joint replacement with independent com-
munity ambulatory ability, and iii) patients with a
body habitus and gait conducive to recording their
ambulatory activity with the pedometer. The aver-
age age for these 100 patients was 58.6 years (range,
23-82 years). There were 48 men (average age, 60.7
years; range, 36-80 years) and 52 women (average
age, 56.6 years; range, 23-82 years).
Activity Rating Scales
In addition to the pedometer, the activity of each
patient was assessed with 3 methods:
1. Each patient was rated by the investigator
using the 10-point University of California Los
Angeles (UCLA) activity-level rating \[13\] (Table 1).
The evaluation has 10 descriptive activity levels
ranging from wholly inactive and dependent on
others (level 1), to moderate activities such as
unlimited housework and shopping (level 6), to
regular participation in impact sports such as jog-
ging or tennis (level 10). At the time of the evalua-
tion, the investigator queried the patient about
Assessing Patient Activity • Zahiri et al. 891
Table 1. UCLA Activity-Level Rating
Level Activity
10 Regularly participate in impact sports such as jogging,
tennis, skiing, acrobatics, ballet, heavy labor, or
backpacking
Sometimes participate in impact sports
Regularly participate in very active events, such as
bowling or golf
Regularly partidpate in active events, such as bicycling
Regularly participate in moderate activities, such as
swimming and unlimited housework or shopping
Sometimes participate in moderate activities
Regularly participate in mild activities, such as walking,
limited housework, and limited shopping
Sometimes participate in mild activities
Mostly inactive: restricted to minimal activities of daily
living Wholly inactive: dependent on others; cannot leave
residence
participation in the various activities described in
the UCLA activity level rating. The UCLA score is
based on participation in the highest-rated activity,
regardless of the frequency or intensity of participa-
tion.
2. After the interview to determine the UCLA
activity level, the investigator rated the patient's
activity relative to other joint replacement patients
using a simple visual analog scale. The scale, a
horizontal line 10 cm in length, had least active
indicated on the left-hand margin, average in the
middle, and most active on the right-hand margin
(Fig. 1). The investigator made a mark along this
continuum that indicated his or her perception of
the patient's activity relative to other joint replace-
ment patients. This activity rating not only consid-
ered the type(s) of activities, but also considered the
frequency and intensity. Both of these evaluations
were performed by 2 physicians with experience in
total joint replacement (H.C.A., 49; T.P.S., 51) with-
out knowledge of the patient's pedometer data.
3. The patient independently (without knowl-
edge of the investigator's assessment) rated his or
her own activity relative to other people on another
copy of the same visual analog scale. Scoring of the
visual analog scale was done by placing a 10-cm
ruler along the horizontal line and measuring the
distance to the mark that indicated the patient's
Least active Average Most active
Fig. 1. The 10-cm visual analog scale used to assess
patient activity.

892 The Journal of Arthroplasty Vol. 13 No. 8 December 1998
activity level. The distance, measured to the nearest
0.1 cm, was the activity rating.
Linear regression analysis was used to determine
the correlations between the various methods of
assessing activity. Analysis of variance was done to
determine the difference in activity between the
patient subgroups.
Results
Table 2. Walking Activity of Joint
Replacement Patients
Steps Cycles P
Patients n per Day per Year Value
All 100 5,078 926,735
<60 yr 43 5,933 1,082,773 .018
~60 yr 57 4,434 809,205
Men 48 5,853 1,068,173 .018 Women 52 4,363 796,248
Men<60 yr 17 7,187 1,311,628 .002
Patients averaged 5,078 steps per day by the
pedometer, which extrapolates to 0.93 million cycles
per year for each lower extremity joint. The most
active patient averaged 17, 718 steps per day, which
was about 3.2 million cycles per year for each lower
extremity joint. This patient was a 70-year-old man.
The least active patient in this study, a 71-year-old
woman, although an independent community am-
bulator, averaged only 395 steps per day, or only
73,000 cycles per year. The difference between the
most active and the least active patient represents a
45-fold range in activity levels. Regression analysis
was done on the average steps per day with age as
an independent variable. Age was correlated with
patient activity (P = .049) but with a high degree of
variability (SD = 3,156 steps per day) (Fig. 2).
Comparison of various patient subgroups is pre-
sented in Table 2. Patients <60 years old (n = 43)
averaged 5,933 steps per day; patients ->60 years
(n = 57) averaged 4,434 steps per day. This 34%
increase in average walking activity of joint replace-
ment patients <60 years compared to patients >60
years was statistically significant (P = .018). Men
20,000 -
15,000"
"~ 10,000 - . • •
g • ! .- .
• ~ s,000
• i
20 30 40 50 60 70 80 90
Age in Years
Fig. 2. Scatterplot of patient age versus quantitative activity.
A high degree of variability (SD = 3,156 steps per day) is
demonstrated. Linear regression analysis indicates that walk-
ing activity is significantly correlated with age (P = .049).
(n = 48) averaged 5,853 steps per day; women
(n = 52) averaged 4,363 steps per day. There was
also a 34% increase in average walking activity of
the men with a total joint prosthesis as compared
with the women (P = .018). Men <60 years
(n = 17) averaged 7,187 steps per day (about 1.3
million cycles per year), which is 55% higher than
the average activity of the other patients (n -- 83)
who averaged 4,646 steps per day (about 848,000
cycles per year) (P = .002).
Activity Rating Scales
The average UCLA patient activity score was 6.3
(range, 2-10; SD--1.4) (Table 3). Patients <60
years old had an average UCLA score of 6.4, and
patients >60 years had an average UCLA score of
6.2. The average UCLA score for men was 6.7
(range, 2-10; SD +_ 1.6), whereas the average UCLA
score for women was 5.9 (range, 3-9; SD _ 1.1).
Using the visual analog scale as rated by the
investigator, the average patient activity score was 6.1
(range, 2-10; SD-+ 1.6) (Table 3). Patients <60
years old had an average score of 6.3 (range, 3-9;
SD + 1.4), whereas patients ->60 years had an
average score of 5.9 (range, 2-10; SD + 1.8). The
average score for men was 6.7 (range, 5-10;
SD _+ 1.4), whereas the average score for women
was 5.4 (range, 2-9; SD + 1.6).
Using the visual analog scale as rated by the patient,
the average patient activity score was 6.6 (range, 1-10;
SD, + 1.9) (Table 3). Patients <60 years had an average
score of 6.4 (range, 3-10; SD -+ 1.7), whereas pa-
tients :>60 years had an average score of 6.7 (range,
1-10; SD _+ 2.0). The average score for men was 7.0
(range, 3-10; SD _ 1.7), whereas the average score
for women was 6.2 (range, 1-10; SD _+ 1.9).
Both the UCLA activity rating (P = .002) (Fig. 3)
and the visual analog scale rating of the investigator
(P = .00001) (Fig. 4) had a strong correlation with
the average steps per day as recorded by the pedom-
eter. Further, the UCLA activity rating and the
investigator visual analog scale were highly corre-

Assessing Patient Activity • Zahiri et al. 893
Table 3. Activity Rating Scores
UCLA Activity Score
Average SD Range
Investigator Visual Analog Score Patient Visual Analog Score
Average SD Range Average SD Range
All patients 6.3 1.4 2-10
<60 yr 6.4 1.1 4-9
-----60 yr 6.2 1.5 2-10
Men 6.7 1.6 2-10
<60 yr 6.8 1.3 4-9
-->60 yr 6.6 1.7 2-10
Women 5.9 1.1 3-9
<60 yr 6.1 1.0 4-9
->60 yr 5.7 1.1 3-8
6.1 1.6 2-10 6.6 1.9 1-10
6.3 1.4 3-9 6.4 1.7 3-10
5.9 1.8 2-10 6.7 2.0 1-10
6.7 1.4 5-10 7 1.7 3-10
7.1 1.2 6-9 6.8 1.6 5-10
6.6 1.5 5-10 7 1.8 3-10
5.4 1.6 2-9 6.2 1.9 1-10
5.7 1.3 3-9 6.1 1.8 3-9
5.1 1.8 2-9 6.4 2.1 1-10
lated (P< .00001) (Fig. 5). Despite the overall
correlation of UCLA score with the pedometer data,
there was substantial variability in the average steps
per day for individual patients with the same UCLA
score, in whom the difference in the average steps
per day could vary by as much as a factor of 15. For
example, with a UCLA score of 8, an 80-year-old
man averaged only 1,189 steps per day, whereas a
55-year-old man (also with a UCLA score of 8)
averaged 17,680 steps per day, about a 15-fold
difference.
The visual analog scale as rated by the patients of
their own activity did not have as strong a correla-
tion with the pedometer data (P = .08) (Fig. 6) as
did patient age (P = .049). It appears that, in gen-
eral, patients tend to perceive themselves to be
more active than they are, especially women and
patients ---60 years. Further, there was greater vari-
ability (higher SDs) in the visual analog scale ratings
by the patients compared with the visual analog
scale ratings by the investigator (Table 3).
The equation for the slope of the linear regression
line for the plot of the average steps per day as
recorded by the pedometer versus the investigator
visual analog scale (Fig. 4) is:
y = 821.1x + 110.7
where y represents the average number of steps per
20,000 -
15,000.
~. 10,000,
s,ooo
<
20,000 -
15,000
• • • • • ~. |0,000
i "
i .~.....-~-"~1 : • '~ ,,ooo-
• • I o
UCLA Activity Rating Scale
Fig. 3. Scatterplot of UCLA activity rating versus quanti-
tative activity. On a population basis, there is a good
correlation between the UCLA score and the pedometer
data (P = .002). Note the categorical nature of the UCLA
activity rating system and the substantial variability in the
average steps per day for individual patients with the
same UCLA score. For example, the difference between
the least active and the most active patient for the UCLA
score of 8 was about 15-fold.
• o• • •
°•i ~i • ~•R • •
I I I I ll0
2 4 6 8
Investigator V-A Scale
Fig. 4. Scatterplot of investigator visual analog scale
versus quantitative activity. There is less variability in the
average steps per day as compared to the UCLA activity
rating. Linear regression analysis indicates the best corre-
lation with the pedometer data (P = .00001 ) as compared
with the other methods. The equation for the slope of the
line, y = 821.1x + 110.7, can be used to convert the
activity score (represented by x) to an estimate of the
average steps per day (represented by y).

894 The Journal of Arthroplasty Vol. 13 No. 8 December 1998
10' •
~ t • , 6-
" i:: "
2" •
0 i i I l
2 ,4 6 8 It0
UCLA Activity Rating
Fig. 5. Scatterplot of UCLA activity rating versus investi-
gator visual analog scale. Linear regression analysis indi-
cates a significant correlation between the two
(P < .00001).
day and x represents the investigator visual analog
score. This equation can be used to convert the
activity level as determined by the investigator
visual analog scale to an estimate of the average
steps per day for a patient with that activity score.
For example, a patient with an investigator visual
analog activity score of 7.5 would, based on the
experience with these 100 patients, average 6,269
steps per day, or about 1.14 million cycles per year.
Another application of these data can be obtained
by determining the average steps per day of patients
20,000 -
15,000"
10,000.
r.~
~ 5,000.
<
•e
o• • • 00 •
e• | • • oeo
O • • •
i t i i /
2 4 6 8 1 0
Patient V-A Scale
Fig. 6. Scatterplot of patient visual analog scale versus
quantitative activity. Linear regression analysis indicates
that patients' rating of their own activity levels was not
correlated to the pedometer data at the 95% confidence
level (P = .08).
within a defined range of the investigator visual
analog activity score (Table 4). Four activity sub-
groups were defined: scores between 0 to 4 (low
activity), 4.1 to 6 (moderately low activity), 6.1 to 8
(moderately high activity), and 8.1 to 10 (high
activity). The average number of steps progressively
increases with each subgroup. Patients with scores
between 0 and 4 averaged about 3,100 steps per day,
(about 570,000 cycles per year), patients with scores
between 4.1 and 6 averaged about 4,300 steps/day
(about 785,000 cycles per year), patients with scores
between 6.1 and 8 averaged about 5,700 steps/day
(just over 1 million cycles per year), and patients
who scored between 8.1 and 10 averaged about
7,800 steps/day (about 1.4 million cycles per year).
The difference in the activity between the sub-
groups is significant (P = .0007).
Discussion
In the evaluation and reporting of the outcome of
total joint replacement, age and gender have com-
monly been used as proxies for activity \[4,5\]. Al-
though age and gender are correlated to activity on
a cohort basis, quantitative assessments of patient
activity using an electronic digital pedometer has
demonstrated a high degree of individual variability.
Thus, on an individual basis, activity assessments
cannot be based on age and gender \[11\]. Although
quantitative activity assessment by using a pedom-
eter may be the most accurate and sensitive method
currently available, issues related to use of the
pedometer limit its utility in the clinical setting.
Wear has become the focus of total joint replace-
ment. Wear assessments should be based on a
measure of use, not time in situ or a proxy such as
age or gender. Clinicians need a simple method to
assess reliably the activity of patients with joint
replacement. Commonly used outcome evalua-
tions \[5,7-10\] assess what a patient is capable of
doing (function), not what a patient actually does
(activity). The UCLA hip evaluation \[13\] includes
an assessment of patient activity. The UCLA hip
evaluation is not widely used. Further, this activity
Table 4. Average Amount of Activity Based on the
Investigator Visual Analog Scale
Investigator Visual Average Steps Cycles
Analog Score n per Day per Year
0-4.0 11 3,124 570,130
4.1-6.0 40 4, 301 784,933
6.1-8.0 38 5,681 1,036,783
8.1-10 11 7,776 1,419,120

Assessing Patient Activity • Zahiri et al. 895
rating scale has not been validated. In this study, we
found a strong correlation between the pedometer
data and UCLA activity rating (P = .002), bringing
validity to the UCLA scale. There was substantial
variability (up to 15-fold) in the average steps per
day for individual patients with the same UCLA
activity score. A limitation of the UCLA activity
rating is the categorical nature of the descriptions
for the 10 activity levels, making the UCLA activity
rating scale insensitive to the frequency and inten-
sity of an activity.
The investigator visual analog rating demon-
strated the best correlation with the pedometer data
(P = .00001). This simple scale allows the investiga-
tor to integrate frequency and intensity as well as
type(s) of activity into the assessment. The basis for
the investigator visual analog activity rating is the
interview with the patient to determine the UCLA
activity rating. Once the types of activities have
been identified, the investigator can then adjust the
visual analog rating for the frequency and intensity
of activity. It is therefore not surprising that the
UCLA activity score and the investigator visual
analog score are highly correlated (P < .00001).
The visual analog scale rating by the patients of
their own activity did not have as strong a correla-
tion with the pedometer data (P - .08) as the other
rating scales. In fact, patient self-assessment of
activity had a poorer correlation to the pedometer
data than patient age (P = .049). The assessment of
patient activity by the physician-investigators in this
study was more accurate than the patients' assess-
ment. This is probably because patients have little
perspective of other patients' activity levels, whereas
a physician, or another individual with experience
in the evaluation of patients with total joint replace-
ment, has a broader perspective for this assessment.
Outcome evaluations of lower extremity joint
reconstructions should include an assessment of
patient activity. Because of practical considerations,
use of a pedometer is probably best reserved for the
evaluation of extreme cases of activity (or inactiv-
ity). This study indicates that both the UCLA activity
score and the investigator visual analog scale are
valid for routine activity assessment in the clinical
setting. Adjustments of the UCLA activity score for
the frequency and intensity of activity, as can be
done with the investigator visual analog scale,
increase the accuracy of the activity rating.
References
1. Kabo JM, Gebhard JS, Loren G, Amstutz HC: In vivo
wear of polyethylene acetabular components. J Bone
Joint Surg Br 75:254, 1993
2. Seedhom BB, Dowson D, Wright V: Wear of solid
phase formed high density polyethylene in relation to
the life of artificial hips and knees. Wear 24:35, 1973
3. Seedhom BB, Wallbridge NC: Walking activities and
wear of prostheses. Ann Rheum Dis 44:838, 1985
4. Chandler HP, Reineck FT, Wixson RL, McCarthy JC:
Total hip replacement in patients younger than thirty
years old. J Bone Joint Surg Am 63:1426, 1981
5. Charnley J: The low friction principle: low friction
arthroplasty of the hip: theory and practice. Springer-
Verlag, New York, 1979
6. Dorr LD, Takei GK, Conaty JP: Total hip arthroplasty
in patients less than forty-five years old. J Bone Joint
Surg Am 65:474, 1983
7. American Academy of Orthopaedic Surgeons: Coun-
cil of Musculoskeletal Specialty Societies Outcomes
Data Collection Package. American Academy of Ortho-
paedic Surgeons, Rosemont, IL, 1996
8. Harris WH: Traumatic arthritis of the hip after disloca-
tion and acetabular fractures: treatment by mold
arthroplasty: an end result study using a new method
of result evaluation. J Bone Joint Surg Am 51:737,
1969
9. Insall JN, Dorr LD, Scott RD, Scott WN: Rationale of
the knee society clinical rating system. Clin Orthop
248:13, 1989
10. Johnston RC, Fitzgerald RH, Harris WH, et al: Clinical
and radiographic evaluation of total hip replacement:
a standard system of terminology for reporting re-
sults. J Bone Joint Surg Am 72:161, 1990
11. Schmalzried TP, Szuszczewicz ES, Northfield MR, et
al: Quantitative assessment of walking activity in
joint replacement patients. J Bone Joint Surg Am
80:54-59, 1998
12. Bassey E J, Dallosso HM, Fentem PH, et al: Validation
of a simple mechanical accelerometer (pedometer)
for the estimation of walking activity. Eur J Appl
Physiol 56:323, 1987
13. Amstutz HC, Thomas B J, Jinnah R, et al: Treatment of
primary osteoarthritis of the hip: a comparison of
total joint and surface replacement arthroplasty. J
Bone Joint Surg Am 66:228, 1984