The Active Movement Scale: An
Evaluative Tool for Infants With
Obstetrical Brachial Plexus Palsy
Christine Curtis, MSc, Derek Stephens, MSc,
Howard M. Clarke, PhD, MD,
David Andrews, PhD, Toronto, Ontario, Canada
Newborns with peripheral nerve lesions involving the upper extremity are difficult to evaluate. The
reliability of the Active Movement Scale (AMS),a tool for assessing motor function in infants with
obstetrical brachial plexus palsy (OBPP),was examined in 2 complementary studies. Part A was
an interrater reliability study in which 63 infants younger than 1 year with OBPP were indepen-
dently evaluated by 2 physical therapists using the AMS. The scores were compared for reliability
and controlled for chance agreement by using � statistics. Overall � analysis of the 15 tested
movements showed a moderate strength of score agreement (� � 0.51). Quadratic-weighted �
) statistics showed that 8 of the 15 movements tested were in the highest strength of
agreement category (�
� 0.81�1.00). Five movements showed substantial agreement (�
0.61�0.80),and 2 movements had moderate agreement (�
� 0.41� 0.60). The overall �
was 0.89. Part B was a variability study designed to examine the dispersion of scores when infants
with OBPP were evaluated with the AMS by multiple raters. Ten pediatric physical therapists with
varying degrees of experience using the scale attended a 1
⁄2-hour instructional workshop on
administration of the tool for infants with OBPP. A chain-block study design was used to obtain 30
assessments of 10 infants by 10 raters. A 2-way analysis of variance indicated that the variability
of scores due to rater factors was low compared with the variability due to patient factors and that
variation in scores due to rater experience was minimal. The results of part A indicate that the AMS
is a reliable tool for the assessment of infants with OBPP when raters familiar with the scale are
compared. The results of part B suggest that,with minimal training,raters with a range of
experience using the AMS are able to reliably evaluate infants with upper-extremity paralysis. (J
Hand Surg 2002;27A:470-478. Copyright © 2002 by the American Society for Surgery of the
Key words: Obstetrical brachial plexus palsy,outcome measure,reliability,movement,scale.
The evaluation of upper-extremity motor power in
infants with obstetrical brachial plexus palsy (OBPP)
is a challenge to clinicians. The ability to accurately
document motor function and to measure change (or
lack of change) in movement over time has important
implications for determining the natural history of
From the Department of Rehabilitation Services and the Department of Biostatistics and Division of Plastic Surgery, The Hospital for Sick Children,
Toronto, Ontario, Canada; and the Department of Surgery and Department of Statistics, University of Toronto, Toronto, Ontario, Canada.
Supported by the Division of Plastic Surgery and the Research Institute at The Hospital for Sick Children.
Received for publication July 30, 2001; accepted in revised form January 29, 2002.
No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.
Reprint requests: Christine Curtis, MSc, The Hospital for Sick Children, Department of Rehabilitation Services, 555 University Ave, Toronto,
Ontario, Canada M5G 1X8.
Copyright © 2002 by the American Society for Surgery of the Hand
470 The Journal of Hand Surgery
OBPP. Determining appropriate treatment options
and measuring treatment outcomes are likewise chal-
lenging. Although a number of tools have been used
to quantify upper-extremity motor power and func-
tion in infants with upper-extremity weakness, there
are no reports on the reliability or validity of these
tools in this population.
Brachial plexus injuries in infants usually occur as
a consequence of difficult childbirth.
Erb’s palsy display weakness or paralysis of the
muscles served by C5, C6� C7 nerve roots. Patients
with total brachial plexus palsy (C5, C6, C7, C8 �
T1) are more severely affected and have involvement
of hand grasp function.
The degree of spontaneous motor recovery in chil-
dren with OBPP varies widely. Some studies suggest
that the majority of infants recover sufficiently and
do not require primary surgical intervention.
vances in microsurgical techniques and pediatric an-
esthesia, however, have led to a renewed interest in
the surgical management of this condition.
lack of a uniform grading system makes it difficult to
compare outcomes and determine the natural history
of brachial plexus lesions in infants.
Many methods have been used to describe or quan-
tify motor function in children with OBPP. The British
Medical Research Council (BMRC)
system of man-
ual muscle testing is the most recognized scale for the
evaluation of strength for patients with peripheral nerve
injuries. This test uses limb segment positioning with-
out and against gravity and manual resistance to grade
muscle strength on a 6-point scale (0 � no contraction,
5 � normal power). The BMRC scale as a measure of
strength for infants with OBPP has been reported by a
number of investigators.
Gilbert and Tassin
suggested a modified BMRC
scale for the evaluation of children with OBPP to
account for the difficulties encountered in examining
infants with manual resistance. The M0–M3 scale
has been used as an outcome measure in some stud-
This scale is limited in the ability to
distinguish improvements in motor recovery because
it has only 1 grade for classifying partial movement.
described a method of evaluating children
with OBPP based on the ability of the child to
perform functional positioning of the affected limb.
Although used as an outcome measure by a number
this system can be used
only with a older, cooperative child. This scale is not
suitable for use with infants and therefore cannot be
used to compare across treatments in infancy.
A recent report by Chuang et al
introduced a new
evaluation system for predicting the sequelae of late
OBPP. Intraoperative findings are matched with a
score of 10 (a grading system that awards points
based on functional ability and the feasibility of
reconstructing deficient joint movement) to predict
deformity and functional limitations in the older
child. The investigators noted, however, that this
system is not suitable for use with infants and is not
amenable to comparisons before and after recon-
Eng et al
used 2 impairment rating scales to
correlate functional outcome with conservative man-
agement. One scale was used to classify newborn
function on initial examination, and a second scale
was used as a comparison to evaluate late outcome.
The validity and reliability of these scales have not
A number of reports used descriptive accounts of
function, such as “good,” “fair,” and “poor,” to char-
acterize clinical and surgical outcomes.
cause of the descriptive nature of these studies, com-
parison of results is not possible.
To our knowledge there is no evaluative tool cur-
rently available that is practical, sensitive, and reli-
able for quantifying motor function in infants with
upper-extremity paralysis. For these reasons we de-
veloped in the Brachial Plexus Clinic at The Hospital
for Sick Children our own scale for evaluating active
movement in infants with OBPP. The Active Move-
ment Scale (AMS) (Table 1) is an ordinal 8-grade
scale designed to capture changes in arm movement.
We believe that this scale offers a number of advan-
It can be used to grade movement in entire
upper extremities of infants and young children, and
it does not require the child to perform tasks on
Table 1. The Active Movement Scale
Observation Muscle Grade
No contraction 0
Contraction, no motion 1
Motion � 1/2 range 2
Motion � 1/2 range 3
Full motion 4
Motion � 1/2 range 5
Motion � 1/2 range 6
Full motion 7
Reprinted with permission.
The Journal of Hand Surgery / Vol. 27A No. 3 May 2002 471
command. In contrast to individual muscle testing,
the AMS evaluates overall joint movements. Smaller
changes in movement can potentially be detected
through the use of an 8-point grading system. Fi-
nally, the AMS can be used for the entire life of a
child and as both a preoperative and postoperative
The purpose of the 2 complementary studies was
to examine aspects of reliability of the AMS for the
assessment of infants with OBPP. Specifically, the
objective of part A was to examine the interrater
reliability of 2 experienced raters using the AMS.
Part B was conducted to examine rater and patient
variability and to compare variations in scores asso-
ciated with multiple raters with different levels of
experience using the AMS.
Materials and Methods
Part A: Interrater Reliability Study
Study Design. A single trial design with multiple
patients was used to determine total reliability and
chance agreement of the AMS by 2 experienced
raters. Consecutive rather than simultaneous evalua-
tion of the patients was conducted due to the neces-
sary interaction between rater and infant that is re-
quired to properly apply the tool.
The sample size requirement was estimated ac-
cording to the recommendation of Donner and Eli-
With a type I error of 0.05 and power of
0.80, it was calculated that a minimum of 50 patients
was required to show an underlying true reliability of
the AMS �0.80.
Patients. Sixty-three infants younger than 1 year
with a unilateral brachial plexus lesion as determined
by the Brachial Plexus Team at The Hospital for Sick
Children were recruited over a 28-month period for
the first reliability study. Children with a concomi-
tant diagnosis such as a neurologic condition (eg,
birth asphyxia or cerebral palsy) or fracture (eg,
clavicle, humerus, or rib) were excluded. The fami-
lies were informed of the study in consecutive order
as they presented to the Brachial Plexus Clinic, and
consent to participate in the data collection was ob-
tained from a parent of each child.
Data Collection. Two pediatric physical therapists
from the Hospital for Sick Children, each experi-
enced in using the AMS (more than 20 assessments
each before starting the study), performed assess-
ments using the scale on the 63 infants. Using a
standardized form, the raters evaluated 15 functional
movements of the affected arm of each child. Raters
were randomly assigned to be first assessor by coin
toss and were blinded to the other’s findings. Each
rater’s evaluations took no longer than 15 minutes to
Guidelines for Use of the Active Movement
Scale. Consistency within the study was facilitated
through the use of the following operational instruc-
tions for the application of the AMS
1. A score of 4 must be achieved (full range of motion
with gravity eliminated) before a higher score can be
assigned. This clarifies scoring when limited move-
ment is present both with gravity eliminated and
2. Movement grades are assigned within the available
range of passive motion; for example, if a flexion
contracture is present at the elbow, full range of
extension is scored if the elbow can be extended to
the limits of the contracture.
3. Movement is assessed within the age-appropriate
range of motion, with the uninvolved contralateral
limb used as a control to estimate the extent of
available normal range.
4. Extension of the fingers is assessed at the metacar-
pophalangeal joints. Finger flexion is evaluated by
observing the distance at rest between the fingertips
and the palm and then observing the active motion as
a fraction of that distance, both with and without
5. Digital flexion or extension is given a single grade
by using the movement score of the best finger; for
example, if the index finger scores a grade of 7 for
flexion and the other digits score 2, then the finger
flexion score is 7.
The evaluation is performed with the upper body
and arms of the infant exposed. Ideally, the child is
placed on a flat, firm surface with room to move or
roll. A variety of toys should be available to stimu-
late movement. Rattles or toys that make sounds are
especially useful. Gravity-eliminated movements are
assessed first to determine whether scores higher
than 4 can be assigned. To grade shoulder flexion, for
example, the child is placed in the gravity-eliminated
position of side-lying on the unaffected arm with the
affected arm uppermost. A toy is placed within the
child’s view and moved in a way to attract attention.
Tactile stimulation of the arm using the toy followed
by movement of the toy in a forward direction draws
attention to the arm and encourages flexion of the
shoulder. If minimal active movement is seen, the
anterior deltoid region of the shoulder is palpated to
detect flickers of movement. If less than full range of
available passive movement is obtained compared
472 Curtis et al / The Active Movement Scale
with the contralateral normal side, then a score of 3
or lower is given. If full range of forward flexion is
obtained (yielding a score of 4), the child is placed in
a supported sitting position to view movement
against gravity. Again, the child is encouraged to
reach forward for an object. An against-gravity score
of 5 or higher is assigned depending on the greatest
range of motion observed.
All joint movements are scored after observation
in gravity-eliminated and against-gravity positions.
Parents may participate in encouraging movement if
a child is especially anxious with strangers. The
entire assessment can be performed by observing the
child at play in 3 positions: supine, side-lying, and
Statistical Analysis. Three tests of agreement were
used to analyze the data from part A of the study:
percentage agreement, �, and �
Percentage agreement is a reliability test for cate-
gorical variables, estimating the ability of raters to
agree on category ratings.
In this case it was a
measure of how often the raters agreed on the score
for each of the 15 tested movements. � is an appro-
priate measure of ordinal data that calculates the
average rate of agreement and controls for expected
agreement by chance.
It has a maximum of 1.00
when agreement is perfect, and a value of 0 indicates
no agreement better than chance. Landis and Koch
categorized the � statistic according to strength of
agreement with 0.00 to 0.20 representing slight
agreement, 0.21 to 0.40 fair agreement, 0.41 to 0.60
moderate agreement, 0.61 to 0.80, substantial agree-
ment, and 0.81 to 1.0 excellent agreement.
coefficients were also used to deter-
mine if disagreement between the rater’s scores was
minor (differing by only 1 or 2 grades) or extreme.
For this study we chose quadratic weights, which
base disagreement weights on the square of the
amount of discrepancy.
Part B: Variability Study
Study Design. A chain-block design using 10 pa-
tients and 10 raters was selected to examine the
variability of AMS scores. A chain block is a method
of collecting data that allows for the effect of 2 or
more variables (in this case raters and patients) to be
studied using a 2-way analysis of variance. This
design was used for 2 purposes: (1) to facilitate the
randomization of raters and patients and (2) to pro-
vide a practical means of obtaining a total of 30
assessments in a single test period. Thirty assess-
ments were chosen as a sample size of convenience
for a single day of data collection.
Patients. The patient sample was obtained from
the database of the Brachial Plexus Clinic at the
Hospital for Sick Children after approval of the study
was obtained from the institutions’ Research Ethics
Board. The inclusion criteria for patients were iden-
tical to those listed in part A, with the exception that
all children were required to live within a 2-hour
driving distance of the Hospital for Sick Children.
The 47 potential participants were randomized using
a table of random numbers, and the parent of each
child was approached in order of randomization by
the primary investigator. The study was thoroughly
explained to the parent(s) of each child and consent
was obtained. The patient sample consisted of the
first 10 infants for whom consent was obtained.
Raters. The raters were 10 volunteer licensed
physical therapists recruited from the Hospital for
Sick Children and the Bloorview-McMillan Chil-
dren’s Treatment Centre and had 2 or more years of
current, full-time, pediatric employment. Therapists
with a range of experience using the AMS were
sought to participate from the 2 centers.
AMS Teaching Workshop. A1
to teach the raters about the AMS and to standardize
its application for infants with obstetrical brachial
plexus injuries was conducted on the day before the
data collection. The Brachial Plexus Clinic physical
therapist and co-developer of the scale organized and
led the workshop. Didactic lectures, video presenta-
tion, patient demonstration, and practice with dolls
were used to illustrate the administration of the tool.
At the conclusion of the workshop all raters felt
confident that they could reasonably apply the scale.
Data Collection. All data for part B of the study
were collected on the day after the teaching work-
shop. Each child received a total of 3 assessments by
Each assessment involved using the AMS to evaluate
15 joint movements of the involved upper extremity.
The raters performed the assessments in accordance
with the guidelines outlined in part A. Evaluations were
scheduled a maximum of one
⁄2 hour apart, and each
assessment took no longer than 15 minutes to complete.
To avoid viewing or recording their own scores, the
raters verbally conveyed their results to an assistant
who recorded the scores on a data sheet. Raters were
blinded to the results of their previous assessments and
to the assessments of the other raters.
Statistical Analysis. A 2-step process was pro-
posed to analyze the data. First, a general linear,
The Journal of Hand Surgery / Vol. 27A No. 3 May 2002 473
2-way analysis of variance was used to test the
significance of variance in scores due to rater and
patient factors. A statistically significant difference
in scores between these factors would then allow the
construction of box plots that would be used to
compare the variability of factors between and within
raters and patients. Box plots are graphs that display
the entire set of data in visual form and contain the
25th and 75th centiles, interquartile range (IQR), and
median. The IQR is a measure of data dispersion. A
large amount of dispersion of the IQR indicates
Part A: Interrater Reliability Study
The descriptive statistics for patients in Part A of
the study are listed in Table 2. The boy-to-girl patient
ratio was almost evenly divided, and the ages within
these groups were evenly distributed. Many more
patients in the sample had right-sided involvement
than left, which matched the OBPP population at
Percentage agreement, �, and �
scores for the
15 tested movements are listed in Table 3. Forearm
supination represented the least accurate percentage
of agreement (49%) between the 2 raters. Eight of
the 15 movements had total agreement of 90% or
more. Analysis of the scores using the � statistic
indicated a lower level of agreement for all 15 move-
ments than was obtained using percentage agreement
alone. With the effects of chance eliminated, the
overall strength of agreement of the AMS was 0.51
with a 95% confidence interval of 0.46 to 0.56, a
score considered in the moderate range of agreement.
The overall �
coefficient of the AMS was 0.89
with a 95% confidence interval of 0.87 to 0.91,
indicating that the raters showed excellent agreement
using the scale as a whole and that any disagreements
in scores were small. �
statistics of 8 of the 15
movements were in the highest strength of agreement
� 0.81–1.0). Five movements showed
substantial agreement (�
� 0.61–0.80), and 2
movements (pronation and supination) had moderate
� 0.41–0.60) (Table 4).
Part B: Variability Study
Descriptive statistics for the 10 patients in part B
are given in Table 5. Review of the patient statistics
suggested a highly varied sample. The boy-to-girl
ratio was 6:4, and the ages were well distributed with
a range of 1 month to 1 year. Ninety percent of the
patients had right-sided involvement. Three of the
patients had total plexus involvement, and 2 patients
had undergone surgical brachial plexus reconstruc-
Table 3. Agreement Coefficients of Raters in Part A
Joint Movement Agreement (%) � (95% CI) �
Shoulder abduction 51 0.37 (0.23,0.51) 0.72 (0.64,0.80)
Shoulder adduction 92 0.64 (0.39,0.89) 0.75 (0.51,0.97)
Shoulder flexion 63 0.54 (0.39,0.68) 0.82 (0.74,0.89)
Shoulder external rotation 65 0.48 (0.33,0.63) 0.68 (0.55,0.82)
Shoulder internal rotation 95 0.55 (0.10,1.00) 0.62 (0.17,1.00)
Elbow flexion 62 0.51 (0.37,0.65) 0.93 (0.89,0.97)
Elbow extension 92 0.59 (0.32,0.86) 0.92 (0.82,1.00)
Forearm pronation 97 0.49 (0.00,1.00) 0.49 (0.00,1.00)
Forearm supination 49 0.33 (0.19,0.46) 0.56 (0.41,0.71)
Wrist flexion 90 0.48 (0.14,0.81) 0.83 (0.61,1.00)
Wrist extension 78 0.52 (0.36,0.69) 0.81 (0.72,0.90)
Finger flexion 97 0.65 (0.38,0.93) 0.89 (0.79,0.99)
Finger extension 98 0.88 (0.64,1.00) 0.98 (0.94,1.00)
Thumb flexion 98 0.85 (0.56,1.00) 0.96 (0.86,1.00)
Thumb extension 76 0.56 (0.38,0.74) 0.66 (0.49,0.83)
CI, confidence interval.
Table 2. Descriptive Statistics for Patients in Part A
(Range) Right Side Left Side
Boys 32 (50.8%) 5 (1–11) 18 14
Girls 31 (49.2%) 5 (1–11) 22 9
Total 63 5 (1–11) 40 (63.5%) 23 (36.5%)
474 Curtis et al / The Active Movement Scale
The raters in this study were all physical therapists
with pediatric experience ranging from 2 to 30 years
(mean, 12.5 years). Experience using the AMS
ranged from none (6 raters) to more than 50 evalu-
ations (2 raters).
The appendix shows the results for the 2-way
analysis of variance of the raters’ scores. (This ap-
pendix can be viewed at the Journal’s Web site,
www.jhandsurg.org.) The purpose of this appendix is
to determine if the correct study model was chosen
and if indeed factors associated with raters or pa-
tients account for the differences in the scores (the
underlying assumption is that there is no difference
in scores due to patient or rater factors). With p� .05
and a high Fscore, we rejected the hypothesis that
there was no difference in scores due to rater and
patient factors. The results of the analysis of variance
were then expressed through the creation of box plots
of the variance coefficients of the 2 groups (patients
and raters). Comparison of score dispersion by pa-
tient and rater factors can be seen in Figure 1. Ex-
amination of the plots reveals that dispersion of the
scores due to factors associated with the patients is
greater than the dispersion due to factors attributed to
the raters. The median of rater factors is a coefficient
greater than 90. The notable feature of this plot is
that despite the extreme range in variability of patient
factors, the raters were able to evaluate these patients
with a high degree of precision as evidenced by the
concentration of rater factors in the high coefficient
range. Figure 2 shows a comparison of rater variance
by experience, with 6 of the 10 raters having no
previous experience using the AMS before the teach-
ing workshop. By inspection it appears that little
difference in scoring can be attributed to raters’
previous experience using the AMS and that the
variability coefficients shown by each of these
groups is high (� 75).
Measurement is the underlying basis of scientific
investigation, and the precise measurement of com-
plex clinical phenomena is one of the challenges of
clinical and surgical research.
A consensus regard-
ing a standardized method of evaluating motor power
in infants with OBPP has yet to be established. To
understand the natural history of OBPP and design
Table 4. Strength of Agreement of 15 Tested Movements in Part A
Pronation Shoulder abduction Shoulder flexion
Supination Shoulder adduction Elbow flexion
Shoulder external rotation Elbow extension
Shoulder internal rotation Wrist flexion
Thumb extension Wrist extension
Table 5. Descriptive Statistics of Patients in Part B
A Boy 6 Right No Total
B Boy 1 Right No Upper
C Boy 7 Right No Total
D Boy 9 Right Yes Total
E Girl 12 Right No Upper
FGirl 3 Right No Upper
G Boy 11 Left Yes Upper
H Girl 8 Right No Upper
I Boy 3 Right No Upper
J Girl 11 Right No Upper
The Journal of Hand Surgery / Vol. 27A No. 3 May 2002 475
valid outcome studies, researchers must use a con-
sistent method of evaluation of upper-extremity
function that is independent of the patient’s verbal
comprehension and ability to cooperate voluntarily.
This evaluation method should uphold standards of
appraisal such as reliability and validity.
Various systems for quantifying motor function in
the upper extremities of infants with OBPP have
None of these systems
has been reported to have validity or reliability for
the assessment of infants with upper-limb paralysis.
The BMRC system of manual muscle testing, al-
though reliable for the examination of adults,
not suited for use with infants because it requires that
the patient understand the nature of the examination.
Simplification of the 6-point (0–5) MRC scale by
Gilbert and Tassin
to a 4-point grading system
(M0–M3), although practical in design, severely
limits the ability of the scale to discriminate motor
recovery. All partial movements are classified as M2.
This grade covers a wide range of incomplete move-
ments, from slight movement with gravity eliminated
to almost full range of movement against gravity.
Scores in this category will fail to differentiate im-
provements in partial movement over time and will
not distinguish between functional and nonfunctional
The discriminative ability of a scale is dependent
on the number of categories used. Nishisato and
have shown that for reliability coefficients in
the range normally encountered, from 0.4 to 0.9, the
reliability drops as fewer categories are used.
Streiner and Norman
suggested that the minimum
number of categories used by raters should be in the
range of 5 to 7 to maximize the precision of a scale.
In response to the difficulty of evaluating infants
with OBPP and the limitations of reported measure-
ment systems, we developed a new tool: the AMS.
The 8-point construct of this scale maximizes the
potential for precision and reliability of measure-
ment. Five scores that categorize less than full move-
ment (grades 2–6) allow for the discrimination of
changes in partial movement over time. Within the
grades of partial movement there is the possibility of
distinguishing movements that suggest poor recovery
(grades 2–5) or useful function (grade 6).
In the Brachial Plexus Clinic at the Hospital for
Sick Children, we have been using the AMS as an
evaluative tool for infants with OBPP for the last 10
years. To date, more than 5,000 assessments of 700
children have been recorded with this system in a
single database. The use of this scale for clinical and
scientific evaluation has been reported else-
The distinct advantages of this tool are
that it can be used on infants and children of any age, it
can be applied before and after surgery, and it allows
for direct comparison of paired data to facilitate statis-
With many factors affecting the overall reliability
of a tool, there is no single analytic approach that can
by itself define all aspects of error and variability in
a measure. Therefore, we conducted 2 analyses in an
attempt to source various components of measure-
ment error and variation associated with the AMS.
Figure 2. Box plot comparison of AMS scores by rater
Figure 1. Box plot comparison of AMS score dispersion
by patient and rater factors.
476 Curtis et al / The Active Movement Scale
The results of the interrater reliability study sug-
gest that the AMS is a reliable scale for the evalua-
tion of infants with OBPP when used by 2 experi-
enced physical therapist raters. Although the overall
� statistic of the scale was 0.51 when controlling for
agreement by chance alone, the overall �
cient of 0.89 indicated that the raters’ scores were
close when there was a discrepancy. The majority of
15 individually tested movements were in the excel-
lent category of agreement when analyzed with
. The advantage of using �
is that it is
identical to the intraclass correlation coefficient
(ICC) and therefore allows comparison with other �
and ICC results.
The results of our previous work
showed that 5 movements of the upper extremity
(elbow flexion and elbow, wrist, finger, and thumb
extension) combined into a test score are the best
predictors of final outcome of OBPP patients at 3
months of age. Four of these 5 movements had
excellent agreement, and 1 movement (thumb exten-
sion) had substantial agreement when examined in
this current analysis.
Some limitations of part A of this study were that
only 2 raters, both experienced in using the AMS,
collected the data, and the amount of variation within
the patient and rater samples were unknown. Part B
of this work was designed to address these limita-
tions by examining the effect of multiple raters with
differing amounts of experience using the AMS on
the dispersion of scores of a randomized sample of
In part B we found that patient factors clearly
accounted for more variability in AMS scores than
rater factors. The rater factors were narrowly dis-
persed with a median variance coefficient greater
than 90. From these results we can conclude that
rater precision was high when evaluating patients
with extremes in variability and that the AMS was a
The second study also suggested that raters’ pre-
vious experience with the AMS did not account for
the higher reliability in scores when compared with
inexperienced raters. Educational preparation of the
raters in the application of the AMS may have ac-
counted for the parity that was shown between ex-
perienced and inexperienced raters’ scoring ability.
The provision of detailed guidelines for the use of the
scale may also have contributed to the accuracy of
the raters’ results. A determination of rater precision
by using the AMS without the benefit of an educa-
tional workshop should serve to further define as-
pects of reliability of this tool.
We believe that 30 evaluations obtained from a
randomized sample of 10 patients were sufficient to
yield useful information about rater and patient vari-
ability. With a limited patient sample size, it was
impossible to further define the exact sources of
patient error. Sources of error associated with pa-
tients may have included fatigue (varying levels of
wakefulness between or within assessments), interest
(varying levels of attention to motivational cues or
objects), anxiety (differing reactions to strangers),
age (factors associated with development), and level
of lesion (degree of paralysis).
Part A of this work helped to identify upper-
extremity movements in infants that are less reliably
evaluated with the AMS, namely, forearm pronation
and supination. Future development of the tool
should be directed at improving the reliability of
measuring these movements. The addition of intrin-
sic evaluation may aid also in further defining levels
of nerve root lesions when using the AMS as an
The authors thank Wendy Barden, BScPT, MSc, for her assistance in
patient evaluation for our initial study.
1. Duchenne G. De l’E
lectrisation localise´e et de son appli-
cation a` la pathologie et a` la the´rapeutique par courants
induits et par courants galvaniques interrompus et conti-
nus. 3rd ed. Paris: Librairie J.B. Baillie`re et fils, 1872.
2. Sever J. Obstetric paralysis. Am J Dis Child 1916;12:541–
3. Tachdjian MO. Pediatric orthopedics. 2nd ed. Vol. 3.
Philadelphia: Saunders, 1990:2009–2082.
4. Greenwald AG, Schute PC, Shiveley JL. Brachial plexus
birth palsy: a 10-year report on the incidence and progno-
sis. J Pediatr Orthop 1984;4:689.
5. Jackson ST, Hoffer MM, Parrish N. Brachial-plexus palsy
in the newborn. J Bone Joint Surg 1988;70A:1217.
6. Michelow BJ, Clarke HM, Curtis CG, et al. The natural
history of obstetrical brachial plexus palsy. Plast Reconst
7. Alanen M, Ryo¨ppy S, Varho T. Twenty-six early opera-
tions in brachial birth palsy. Z Kinderchir 1990;45:136–
8. Boome RS. Indications and timing of surgery for obstetric
brachial plexus palsies. In: Vastama¨ki M, ed. Current
trends in hand surgery: Proceedings of the 6th Congress of
the International Federation of Societies for Surgery of the
Hand. Helsinki, Finland: Elsevier, 1995:275–277.
9. Capek L, Clarke HM, Curtis CG. Neuroma-in-continuity
resection: early outcome in obstetrical brachial plexus
palsy. Plast Reconstr Surg 1998;102:1555–1562.
10. Carlstedt T, Stro¨mbeck C. Surgical vs. conservative treat-
ment of obstetrical brachial plexus palsy: a preliminary
study. In: Vastama¨ki M, ed. Current trends in hand
The Journal of Hand Surgery / Vol. 27A No. 3 May 2002 477
surgery: Proceedings of the 6th International Congress of
the International Federation of Societies for Surgery of the
Hand (IFSSH). Helsinki, Finland: Elsevier, 1995:255–259.
11. Clarke HM, Al-Qattan MM, Curtis CG, Zuker RM. Ob-
stetrical brachial plexus palsy: results following neurolysis
of conducting neuromas-in-continuity. Plast Reconstr Surg
12. Gilbert A. Long-term evaluation of brachial plexus surgery
in obstetrical palsy. Hand Clin 1995;11:583–595.
13. Piatt JH Jr. Neurosurgical management of birth injuries of
the brachial plexus. Neurosurg Clin North Am 1991;2:175.
14. Slooff ACJ. Obstetric brachial plexus lesions and their
neurosurgical treatment. Microsurgery 1995;16:30–34.
15. Kawabata H, Kawai H, Masatomi T, Yasui N. Accessory
nerve neurotization in infants with brachial plexus birth
palsy. Microsurgery 1994;15:768–772.
16. Laurent JP, Lee R, Shenaq S, et al. Neurosurgical correc-
tion of upper brachial plexus birth injuries. J Neurosurg
17. Slooff BA, Blaauw G. Obstetric brachial plexus lesions in
children born in a breech presentation and their neurosur-
gical treatment. In: Vastama¨ki M, ed. Current trends in
hand surgery: Proceedings of the 6th Congress of the
International Federation of Societies for Surgery of the
Hand (IFSSH). Helsinki, Finland: Elsevier; 1995:269–273.
18. Medical Research Council: Nerve Injuries Committee.
Aids to the investigation of peripheral nerve injuries.
London: His Majesty’s Stationary Office, 1942: 48.
19. Sherburn EW, Kaplan SS, Kaufman BA, et al. Outcome of
surgically treated birth-related brachial plexus injuries in
twenty cases. Pediatr Neurosurg 1997;27:19–27.
20. Solonen KA, Telaranta T, Ryo¨ppy S. Early reconstruction
of birth injuries of the brachial plexus. J Pediatr Orthop
21. Shenaq SM, Berzin E, Lee R, et al. Brachial plexus birth
injuries and current management. Clin Plast Surg 1998;25:
22. Sjo¨berg I, Erichs K, Bjerre I. Cause and effect of obstetric
(neonatal) brachial plexus palsy. Acta Pædiatr Scand 1988;
23. Gilbert A, Tassin J-L. Obstetrical palsy: a clinical, patho-
logic, and surgical review. In: Terzis JK, ed. Microrecon-
struction of nerve injuries. Philadelphia: Saunders, 1987:
24. Hentz VR, Meyer RD. Brachial plexus microsurgery in
children. Microsurgery 1991;12:175–185.
25. Gilbert A, Razaboni R, Amar-Khodja S. Indications and
results of brachial plexus surgery in obstetrical palsy. Or-
thop Clin North Am 1988;19:91–105.
26. Mallet J. Paralysie obste´tricale du plexus brachial. Traite-
ment des se´quelles. Primaute´ du traitment de l’e´paule—
Me´thode d’expression des re´sultats. Rev Chir Orthop
Reparatrice Appar Mot 1972;58:166–168 (suppl 1).
27. Meyer RD. Treatment of adult and obstetrical brachial
plexus injuries. Orthopedics 1986;9:899–903.
28. Gilbert A, Whitaker I. Obstetrical brachial plexus lesions.
J Hand Surg 1991;16B:489–491.
29. Gilbert A, Brockman R, Carlioz H. Surgical treatment of
brachial plexus birth palsy. Clin Orthop 1991;264:39–47.
30. Duclos L, Gilbert A. Obstetrical palsy: early treatment and
secondary procedures. Ann Acad Med Singapore 1995;26:
31. Chuang DC-C, Ma H-S, Wei F-C. A new evaluation sys-
tem to predict the sequelae of late obstetric brachial plexus
palsy. Plast Reconstr Surg 1998;101:673–685.
32. Eng G, Binder H, Getson P, O’Donnell R. Obstetrical
brachial plexus palsy (OBPP) outcome with conservative
management. Muscle Nerve 1996;19:884–891.
33. Hunt D. Surgical management of brachial plexus birth
injuries. Dev Med Child Neurol 1988;30:823–828.
34. Jahnke AH, Bovill DF, McCarroll HR Jr, et al. Persistent
brachial plexus birth palsies. J Pediatr Orthop 1991;11:
35. Kwai H, Yamamoto K, Murase T, et al. Brachial birth
palsy—cervical myelography and early brachial plexus
surgery. In: Vastama¨ki M, ed. Current trends in hand
surgery: Proceedings of the 6th Congress of International
Federation of Societies for Surgery of the Hand. Helsinki,
Finland: Elsevier, 1995:261–267.
36. Kawabata H, Masada K, Tsuyuguchi Y, et al. Early mi-
crosurgical reconstruction in birth palsy. Clin Orthop 1987;
37. Brown KLB. Review of obstetrical palsies: nonoperative
treatment. Clin Plast Surg 1984;11:181–187.
38. Clarke HM, Curtis CG. An approach to obstetrical brachial
plexus injuries. Hand Clin 1995;11:563–581.
39. Donner A, Eliasziw M. Sample size requirements for reli-
ability studies. Stat Med 1987;6:441–448.
40. Portney LG, Watkins MP. Foundations of clinical research:
applications to practice. Stamford, CT: Appleton & Lange, 1993.
41. Cohen J. Coefficient of agreement for nominal scales. Educ
Psychol Meas 1960;20:37.
42. Landis JR, Koch GG. The measurement of observer agree-
ment for categorical data. Biometrics 1977;33:159–174.
43. Cohen J. Weighted kappa: nominal scale agreement with
provision for scaled disagreement or partial credit. Psychol
44. Norman GR, Streiner DL. Biostatistics, the bare essentials.
St. Louis: Mosby, 1994:166.
45. Altman DG. Practical statistics for medical research.
London: Chapman and Hall, 1991:33.
46. Wright JG, McLeod R, Lossing A, et al. Measurement in
surgical clinical research. Surgery 1996;119:241–244.
47. Iddings DM, Smith LK, Spencer WA. Muscle testing: part
2. Reliability in clinical use. Phys Ther Rev 1961;41:249–
48. Wadsworth CT, Krishnan R, Sear M, et al. Intrarater reli-
ability of manual muscle testing and hand-held dynametric
muscle testing. Phys Ther 1987;67:1342–1346.
49. Nishisato S, Torii Y. Effects of categorizing continuous
normal variables on product-moment correlation. Jap
Psych Res 1970;13:45–49.
50. Streiner DL, Norman GR. Health measurement scales: a
practical guide to their development and use. 2nd ed. New
York: Oxford University Press, 1995:35.
51. Al-Qattan MM. The first multi-disciplinary obstetrical bra-
chial plexus clinic in Saudi Arabia. J Hand Surg 1996;21B:
478 Curtis et al / The Active Movement Scale
Appendix 1. Analysis of Variance of Total Scores
Source DF Seq. SS Adj. SS Seq. MS F* p†
Rater 9 7,233.9 612.4 803.8 52.34 .000
Patient 9 11,087.1 11,087.1 1,231.9 80.22 .000
Error 11 168.9 15.4
Total 29 18,489.9
DF, degrees of freedom; Seq. SS, sequential sum of squares; Adj. SS, adjusted sum of squares; Seq. ms, sequential mean square.
*Significant F ratio.
†p � .05.
The Journal of Hand Surgery 1