GMFM: Gross Motor Function
Measure, Part I



Kathy McKellar, “Knowledge Broker”
January 2007
Based on a presentation by Dianne Russell, CanChild
Centre for Childhood Disability Research, Knowledge
Broker project co-investigator
KB study




Looking at clinical knowledge and
appropriate use of:
GMFCS
GMFM
Motor Growth Curves (MCG’s):
prognosis, treatment planning
Interaction of Concepts
ICF 2001
Health Condition
(disorder/disease)
Body
function&structure
(Impairment)
Environmental
Factors
Activities
(Limitation)
Participation
(Restriction)
Personal
Factors
GMFM: Why was it developed?


To answer the question: “How do we
measure small but important changes in
motor function for children with CP?”
Development started in 1984
GMFM


Criterion-referenced test: evaluates
performance of motor skills on that day;
useful for comparison over time
Measures how much of a task the child
can accomplish, rather than how well
the task is completed (quantity, not
movement quality)
Who is the GMFM appropriate
for?



Children with CP: original validation
sample included kids 5 mo- 16 yrs
May be appropriate for children with
other diagnoses
GMFM is appropriate for children whose
motor skills are at or below those of a
typical 5 year old.
GMFM Formats



GMFM-88: 88 items
GMFM-66: 66 items
GMFM-88 with reported scores for kids
with Down Syndrome
Examiner Qualifications



For use by pediatric PT’s
Before testing children, PT’s should
familiarize themselves with the
scoresheet and the administration and
scoring guidelines
CD ROM training available
Time required



GMFM 88: approx. 45-60 minutes
GMFM 66: faster, allows for some
missing data (items that are not tested)
Can be completed in more than 1
session (ideally complete all items
within 1 week)
GMFM-88
88 items in 5 gross motor dimensions (for
ease of administration):
 lying and rolling
 crawling and kneeling
 Sitting
 Standing
 walking, running and jumping
GMFM-66

Same dimensions, but 22 items
eliminated (mostly in lying position)
Validation of the GMFM-88

Reliability
•
•
Test-retest (ICC = 0.99) ( dimensions ranged .92.99)
Inter-rater (ICC = 0.99) (dimensions ranged .87.99)
• Validity
• Gradient of change: pre-school children without
CP>children with ABI>children with CP
• Children with CP who were young & mild > older &
more severe
Validation of the GMFM-88

Change over 6 months as judged by parents,
therapists, and a “masked” video analysis was
correlated with change scores on the GMFM88
Further evidence of reliability &
validity



Reliability established by others outside the GMFM
team (Bjornson et al. 1994;1998, Nordmark et al.
1997)
Responsiveness (Bjornson et al. 1998; Kolobe et al.
1998
Discriminative validity (Palisano et al 2000)
Why use the GMFM?




Reliable, valid
Internationally accepted:
Translated into
several languages, including Dutch, French, German,
Icelandic, Japanese
Considered best practice
Used as an outcome measure
Used as an outcome measure



Surgery (rhizotomy, pallidal stimulation, muscle
tendon)
Drugs (botulinum toxin, intrathecal baclofen)
Physical therapy (including ambulatory aids &
orthoses)

Horseback riding

Strength training & physical fitness
Use of the GMFM in other
populations



Osteogenesis imperfecta (Ruck-Gibis et
al. 2001)
Lymphoblastic leukemia (Wright et al.
1998)
Down syndrome (Russell et al. 1998)
Validation for children with Down
syndrome


Compared the results using the standard
scoring method with an alternate method of
scoring using caregiver report “Reported
Score” (for items which the therapist couldn’t
entice the child to demonstrate)
Found stronger evidence of reliability, validity
& responsiveness with “reported score”
Equipment




GMFM kit
Need smooth floor, large firm exercise
mat, toys for motivation, large bench or
table for cruising
Five steps with railing
Wheeled stool
Environment



Room large enough to accommodate
the equipment, the child and the
examiner
Private area
Consistent environment for retesting
Clothing


Shorts and Tshirt are ideal
Testing is done without shoes
Preparing for Testing


Have manual, equipment, score sheet
ready.
Room booked, mat in place, as well as
other required furniture
Testing




Items may be tested in any order, but
be careful not to miss any! (esp. when
using the GMFM 88)
Verbal encouragement or demonstration
is permitted
Maximum 3 trials for each item
Spontaneous performance of any item
is acceptable
Non-compliance



Strategies such as “follow the leader” or role
playing can be used
Toys and incentives can be used as
motivators (eg. creep through a tunnel)
If a child refuses to attempt an item that you
think they can do, return to the item at the
end of the test, or try it again in in another
session. You can also circle “not tested”.
Scoring the GMFM






Scores 0-3 or NT
0- does not intitiate task
1- intitiates task (<10%)
2- partially completes task (10-99 %)
3- completes task (100%)
Sometimes generic scoring as above,
other times specific criteria for each
level
Scoring the GMFM, cont.




The score given is based on the best
performance out of the 3 trials
If undecided about what score to assign,
choose the lower of the 2 possible scores
Any item that has been omitted or that the
child is unable (or unwilling) to attempt must
be indicated as NT
In the GMFM 88, NT items are scored 0, but
in the GMFM 66, NT items are treated as
missing data
Item 36
O n the floor: A ttains sitting on
sm all bench
0 = does n ot initiate sitting
1 = initiates sitting
2 = partially attains sitting
3 = attains sitting
N T = N ot tested
G eneric S coring K e y
Initiates= com pletes less
than 10% of task
P artially c om pletes=
com pletes > 10% to less
than 100%
# 58: S tanding:lifts R foot, arm s free, 10 secs .
•
0= does n ot lift R foot, arm s free
•
1= lifts R foot, arm s free, < 3 secs.
•
2= lifts R foot, arm s free, 3 -9 secs.
•
3= lifts R foot, arm s free, 10 secs.
GMFM Part II… to follow






GMFM-88 vs. 66
Scoring
GMAE
Interpretation of results
Motor Growth Curves
GMFCS, GMFM, MCG’s: how do they
relate?
Knowledge Broker study
CanChild research project looking at



clinical knowledge and appropriate use
of:
GMFCS
GMFM
Motor Growth Curves (MCG’s)
GMFM Part II







Quick review
Scoring
GMAE
Interpretation of results
GMFM-88 vs. 66
Motor Growth Curves
GMFCS, GMFM, MCG’s: how do they relate?
GMFM


Criterion-referenced test: evaluates
performance of motor skills on that day;
useful for comparison over time
Measures how much of a task the child
can accomplish, rather than how well
the task is completed (quantity, not
movement quality)
Who is the GMFM appropriate
for?



Children with CP: original validation sample
included kids 5 mo- 16 yrs
May be appropriate for children with other
diagnoses: osteogenesis imperfecta,
lymphoblastic leukemia, Down syndrome
GMFM is appropriate for children whose
motor skills are at or below those of a typical
5 year old.
GMFM- 88 and 66
GMFM 88: 88 items in 5 gross motor
dimensions:
 lying and rolling
 crawling and kneeling
 Sitting
 Standing
 walking, running and jumping
GMFM-66: Same dimensions, but 22 items
eliminated (mostly in lying position)
Scoring of the GMFM 88/66



Math or no math
Graph or no graph
Computer or no computer
GMFM-88 score: math!



Sum the item scores within dimensions and
transfer to the summary score section on the
score sheet.
A percent score for each of the 5 dimensions
is calculated.
The total percent score for each dimension is
averaged to obtain the total score (round off
to the nearest whole number)
Scoring with aids/orthotics





Use GMFM-88 only
First complete the GMFM without the
aid/orthosis, then retest with aid/orthosis
For repeat testing at a later dater, apply the
same aid at the same item number
Aids/orthoses could have positive and
negative effects
Mark an “A” for the aided score on the score
sheet
GMFM-88 - scoring issues (i)




Scoring leads to an overall % score as well as
dimension % scores
Change scores: T2 - T1 = GMFM  score
Assumes that all % changes/unit of time have
the same meaning…
...but we don’t really know what a ‘unit’ of
change means clinically! (Some changes
might be easier to attain than others)
GMFM-88 - scoring issues (ii)




GMFM-88 scaling is ‘ordinal’ (ordered)
Cannot assume that a unit of change
has the same meaning across the scale
Really need ‘interval’ scaling, whereby a
‘unit’ of change has the same meaning
throughout the scale
Hence the need for Rasch (itemresponse) analysis
What is Rasch Analysis?

It is a way to analyse data to assess the
‘fit’, order and relative difficulty of items
that measure a construct (e.g., GMF)
RASCH SCALING OF THE GMFM




Identified items which did not “fit” the
unidimensional constructeliminated 22
items (GMFM-66)
Items are now arranged in order of difficulty
(empirical)
Response options within items are weighted
according to difficulty
Interval scale…so that a unit of change has
the same meaning across the scale (thus
improving the interpretability of scores)
GMFM-66



Only 66 items administered (asterixed
on score sheet)
Enter scores into the computer
program: Gross Motor Ability Estimator
(GMAE)
Not possible to calculate the score with
pencil and paper
Gross Motor Ability Estimator
(GMAE)


User-friendly program to analyze
GMFM-66 scores with a built-in tutorial
Allows entry of data in two formats:
1.
2.
Research - from ASCII files or text only
files (files entered into a statistical
package –SPSS)
Individual GMFM-66 item scores for one
or more children
Why use a computer program
to score?


Provides an estimate of a child’s score even when
not all items have been administered
Provides a database to keep child information and
track GMFM-66 scores over time- case summary
report


Produces item maps- arrange items by order of
difficulty
It’s easy! No math, but graphs!
What is the GMFM-66 score?
The GMFM-66 score is an interval-level measure of function where subjects are placed on
an ability continuum ranging from 0 (low motor ability) to 100 (high motor ability).
Interval level scoring makes comparisons of change over time more meaningful because a
difference of, for example, 10 points means the same whether the child is at the lower end or
the upper end of the scale.
Case Summary Report



Summarizes demographic data
Summarizes score, including error
(standard error and 95% confidence
interval)
Graphs scores over time
Item Maps

By item order or by difficulty order- by
difficulty order is the most useful
Lower Motor
GMFM-66 Score
with 95%
Confidence
Intervals
Higher Motor
Ability
Ability
Clinical Use of Item Maps
and Case Summaries
• understand/interpret change
• identify relatively easier and more difficult
‘next steps’ for a child
• discuss and communicate a child’s progress
• set appropriate goals and plan interventions
Interpretation of
GMAE print-outs
Group exercise
Questions for groups:




What is the child’s GMFM-66 score?
Are there any unexpected scores?
What would you expect the child to
accomplish next?
What activities might you work on in PT
with this child?
GMFM 88 and 66

Good reasons to choose one or the
other.
GMFM-88 & GMFM-66
General Issues



Items are administered and scored the same,
with the exception of a new category of ‘Not
Tested’ (NT) to differentiate a true “0” from
an item not attempted
If administer the GMFM-88 with NT, the data
can also be used to calculate score for GMFM66
Strengths of GMFM-88



Reliable and valid measure of change
over time in children with CP and
children with Down syndrome
Widely used in practice and research
GMFM is most responsive to change in
children with CP under age 5 years
Limitations of GMFM-88



Time to administer
- all items must be administered
Must give a score of “0” for items if the child
refuses or assessor fails to administer
Score based on number of items completed
regardless of difficulty
When should I use the GMFM-88?




For a more detailed description of skills
especially for children whose skills are
primarily in Lying and Rolling activities (e.g.,
infants, or children classified at GMFCS Level
V)
No access to a computer
Assessing effects of aides and orthoses
Assessing children with diagnosis other than
CP
Strengths of GMFM-66





Reliable and valid measure of change over
time in children with CP
Items are ordered by difficulty
A score can be derived with a less-thancomplete assessment
Item maps useful in understanding motor
function and in planning goals
Computer program allows tracking of
individual children’s scores over time
Possible Limitations of
GMFM-66



Requires use of a computer program for
scoring
May need some time to learn how to
interpret item maps
No longer able to calculate dimension
scores
When should I use the GMFM-66?



Assessing children with cerebral palsy where
the interval properties of the scale are
important (e.g. Research purposes, change
over time)
When you have limited time to administer all
items
Access to a computer and the GMAE scoring
program
Motor Growth Curves

More graphs!
Motor Growth Curves




Derived from a longitudinal study
657 children, >2600 GMFM
assessments
Children <6 years assessed every 6
mo., older children assessed every 9-12
mo.
Plotted GMFM-66 score against age
Longitudinal Motor Growth Curves for Children with Cerebral Palsy by
GMFCS Level Using GMFM-66 (N=2624 observations)
How can the Motor Growth
Curves be used?
• Describe patterns of gross motor function for
children with cerebral palsy over time
• Estimate a child’s future motor capabilities
(prognosis)
• Compare child’s GMFM-66 score with children
in the sample of a similar age and severity
GMFM-66 plateau


Does not mean therapy is not needed!
Work on quality, functional goals,
equipment needs, prevention of
secondary problems.
Putting the measures all
together…..
• Several different purposes :
discriminative (descriptive)
evaluative
prognostic (predictive)
• Can be used together to track and evaluate
change over time and determine how the
rate of change compares to children of
similar abilities and ages
Case Study
• Beth
Use of Motor Measures at QA


How could these measures work for us?
*** Most useful if used by all PT’s, in
both EIP and SAP
Descargar

GMFM: Gross Motor Function Classification Measure