Clinical Rehabilitation 2009; 23: 352–361
Writing SMART rehabilitation goals and achieving
goal attainment scaling: a practical guide
Thamar JH Bovend’Eerdt Oxford Brookes University and Oxford Centre for Enablement, Nuffield Orthopaedic Centre,
Oxford, Rachel E Botell Oxford Centre for Enablement, Nuffield Orthopaedic Centre, Oxford and St Mary’s Hospital, Leeds and
Derick T Wade Oxford Centre for Enablement, Nuffield Orthopaedic Centre, Oxford, UK
Received 7th November 2008; manuscript accepted 7th November 2008.
Objective: To describe a practical method of setting personalized but specific goals in
rehabilitation that also facilitates the use of goal attainment scaling.
Background: Rehabilitation is a complex intervention requiring coordinated actions
by a team, a process that depends upon setting interdisciplinary goals that are
specific, clear and personal to the patient. Goal setting can take much time and still
be vague. A practical and standardized method is needed for being specific.
Method: A novel approach to writing specific, measurable, achievable, realistic/
relevant and timed (SMART) goals is developed here. Each goal can be built up
by using up to four parts: the target activity, the support needed, quantification of
performance and the time period to achieve the desired state. This method can be
employed as part of goal attainment scaling and the other levels can be easily and
quickly formulated by adding, deleting and/or changing one or more of the (sub)parts.
Discussion: The success of goal setting and goal attainment scaling depends on the
formulation of the goals. The method described here is a useful tool to standardize
the writing of goals in rehabilitation. It saves time and simplifies the construction
of goals that are sufficiently specific to be measurable.
Introduction team where standard, single-treatment packages
are rarely, if ever, appropriate.1 In this context a
Many patients attending rehabilitation services goal-planning process should be used to ensure
have multifactorial, complex problems that often that all the people involved, especially the patient,
require several or many different interventions to agree on the goals of rehabilitation, on the meth-
be given by different people, frequently in a spe- ods to be used to achieve these goals, and on each
cific sequence. Rehabilitation is the archetypical person’s role in this process.2
‘complex intervention’, comprising a multitude of It is also well recognized that goal setting is an
complicated activities and actions. It is a problem- effective way of achieving behavioural change in
solving process delivered by a multiprofessional people.3,4 Some of the characteristics of goals that
effectively alter behaviour are that the goals:
should be relevant to the person concerned,
should be challenging but realistic and achievable,
Address for correspondence: Thamar JH Bovend’Eerdt,
Oxford Centre for Enablement, Windmill Road, Headington, and should be specific (in order to measure them).5
Oxford OX3 7LD, UK. e-mail: t.bovendeerdt@brookes.ac.uk There is some evidence concerning the benefits of
ß SAGE Publications 2009
Los Angeles, London, New Delhi and Singapore 10.1177/0269215508101741
Writing SMART rehabilitation goals 353
goal setting in rehabilitation, particularly around service is increasingly expected to show that treat-
the use of goal attainment scaling as an outcome ments are having the desired effect.
measure.6 Thus the researcher (TB) and the clinical
Thus goal setting is an essential part, and indeed service (RB) set out to achieve a process that sets
the central part of the interdisciplinary rehabilita- goals that:
tion process.
Nonetheless, there is relatively little research on are individualized to a particular patient;
the best way of setting goals in rehabilitation and can be written without too much effort, time or
many questions on the best method still remain. specific training;
For example, does the patient (and family) need allow accurate, unambiguous determination of
to be present at the goal setting meeting, or is it goal achievement;
sufficient to establish their wishes and expectations are flexible enough to cover most situations.
beforehand and to check afterwards that the goals
set are acceptable? What is an appropriate number This novel method for writing SMART goals
of goals? What is an appropriate time frame? can be used as a method simply to write better
One particular question is ‘how should one write goals, but it can be expanded to allow the goal
(specify) a goal?’. It is generally agreed that a good attainment scaling method to be used at little
goal is specific, measurable, achievable, realistic/ extra cost.
relevant and timed (SMART)7 but defining the
characteristics of a SMART goal is less easy.
Moreover writing SMART goals in rehabilitation Background assumptions
is often perceived as time-consuming and difficult.
Well-defined goals are particularly needed for
This article is based on four assumptions. First,
goal attainment scaling which is sometimes used it will consider rehabilitation as taking place
in rehabilitation as way of measuring success. within the pre-eminent (biopsychosocial) model
Goal attainment scaling is a method for evaluating of illness used in rehabilitation, namely an
the attainment of goals. Originally goal attainment expanded version of the World Health Organiza-
scaling was developed simply as an outcome mea- tion’s International Classification of Functioning,
sure but the process may also be in itself a ther- Disability and Health (ICF) model.9
apeutic intervention and a useful tool in case Consequently, it then assumes that rehabilitation
management.8 goals will usually be set around observed beha-
Goal attainment scaling is particularly depen- viours at the WHO ICF levels of activities and par-
dent on defining goals that are measurable,7 ticipation. This does not deny the importance of
which is not always easy because each goal other goals concerning the patient’s personal
requires several different levels to be defined. Yip experiences or the patient’s context (personal, phy-
and colleagues8 developed standardized goal sical or social). Indeed it should be possible to use
attainment scaling menus to address the difficul- or adapt this method for goals in those realms.
ties associated with writing multiple goals. However, the method described here focuses on
However, these menus may be at the cost of activities because they are most easily defined,
some of the advantages of goal attainment scaling, and they probably are of most concern both to
such as its client centred and individual approach. the patient and to those who pay for health care.
Locally the rehabilitation service has developed Third, the description assumes that preliminary
and undertaken goal planning for many years but work with the patient (and relevant other parties)
there has been a long-standing unease within the has already established necessary background
local service about the lack of specificity in some information: the patient’s wishes and expectations,
goals set. A current randomized trial of a rehabi- and all the additional information needed. Goals
litation intervention (motor imagery) needed to must always be set in realms that are of interest to
use individualized specific goals as an outcome the patient. Additionally it may be important to
measure. Finally, and at the same time, the clinical investigate the wishes and expectations of other
354 TJH Bovend’Eerdt et al.
parties such as family members, friends and work The process of goal attainment scaling includes
colleagues, whoever is paying for the service, and five steps11 and our method will focus on steps
(occasionally) team members. It is also important 1–3, illustrated in Figure 1. It must be emphasized
to know sufficient other information to ensure again that before starting step 1 it is essential to
that the goal is potentially achievable, and to iden- know what the patient’s wishes and expectations
tify the actions needed to achieve the goal. In other and goals are and to know enough about
words, this method is only a part of the complete the patient’s situation (disease, impairments, con-
goal-setting process. text, etc.) to allow the team to set valued and
Thus, finally, this method assumes that the team achievable goals.
will only set goals that are attainable and realistic
for the patient to achieve.
Step 1: Defining the expected goals
Goal attainment scaling – introduction The key innovation described in this article is a
structured approach to specifying a goal, and
Goal attainment scaling is the term used to describe this is the important first step in goal attainment
a simple method of scoring (quantifying) the achie- scaling. Even if goal attainment scaling is not used,
vement of goals. Rather than simply stating that a this method allows one to write a SMART goal.
goal has or has not been achieved, attainment scal- The method involves ‘building up’ an expected
ing recognizes that sometimes achievement exceeds goal using four parts:
expectation, whereas at other times achievement is
less than expected but nonetheless there is some pro- specifying the target activity (a behaviour);
gress towards the goal, and (rarely) there may be no specifying the support needed;
progress towards goals set, or even deterioration. quantifying the performance; and
Goal attainment scaling is a structured specifying the time period to achieve the desired
approach to recording goal achievement and was state.
first introduced in the 1960s by Kiresuk and
Sherman10 within a mental health service. The
approach is based on predicting the expected
goal to be achieved, accompanied by two states Part I. Specify the target activity
above the expected outcome and two states Rehabilitation is, ultimately, concerned with
below, one of which is usually (but not inevitably) altering behaviour whether that behaviour is (a)
the current (or ‘baseline’) state. observed activities or participation in social activ-
The process of goal attainment scaling was ities, such as dressing or working, (b) the reporting
chosen by us because it is already reasonably by a person of their internal experiences (such as
well researched with evidence that it is at least as pain), or (c) the report of a person about their
sensitive as a measure of change as other standar- interpretation of activities and experiences (such
dized scales,6 and moreover it may in itself as their own assessment of quality of life, or
improve outcome. Furthermore the scoring satisfaction, or social role performance).
system can be adapted to take into account vari- In the context of setting specific and measurable
ables such as the difficulty of achieving a goal and goals it is easiest to focus upon target behaviours
the patient’s priority, and the scoring system can concerned with activity and participation.
encompass more than one goal but still give a Common examples include mobility and the
single outcome value. many activities of daily living (personal, domestic,
It is not necessarily easy to write a goal specifi- community, vocational, etc.). The method
cally, but the additional challenge when using goal described here can extend to the reporting of
attainment scaling in particular is to write a series experience and perception, but this article will
of five well-defined potential states for each goal, not consider these aspects in any detail; there is
and to do so quickly and easily. some discussion later.
Writing SMART rehabilitation goals 355
Identify patient’s goals and Identify relevant contextual factors
expectations (Environment, resources, etc.)
Specify target activity
(behaviour) Activity
Specify support needed
People
S
Physical aids
T
E
P Cognitive, language or
1 other aids
Quantify performance
Timing
Distance/amount
Frequency
Specify time period to
achieve goal Time period
1 = a little important
Weight the goals 2 = moderately important
S Importance 3 = very important
T
E
P 1 = a little difficult
2 = moderately difficult
Difficulty
2 3 = very difficult
Define other levels by adding, –2 much less than expected
deleting or changing one or
more of ‘support needed’
S and/or ‘quantify performance’ –1 less than expected level
T
E 0 goal (expected level)
P
3 1 better than expected
2 much better than expected
Figure 1 Flowchart for writing goals in goal attainment scaling.
356 TJH Bovend’Eerdt et al.
This first part has the largest number of The third subpart of support concerns the way
possibilities and identifies the functional purpose that items in the environment can be set up to
of the goal. provide informational support encoded or present
The behaviour should be specified as clearly and within the environment; it is the meaning or invo-
explicitly as possible: ‘walking indoors’ rather than luntary consequence associated with the object that
‘mobilizing’, and ‘cooking a three-course meal’ is important. Examples include lists to prompt the
rather than ‘preparing food’. Phrases such as person to sequence actions, sign posting for orien-
‘using left hand in functional tasks’ are too tation, and barriers that remind the person not to
vague and need more detail such as ‘brushing go somewhere.
teeth using left hand’.
In rehabilitation some activities are commonly
targeted, and one might use a list such as the
Rehabilitation Activities Profile12 or the ICF Part III. Quantify performance
core sets for stroke13 as a checklist both to Activities can be described both qualitatively,
ensure that all relevant activities have been consid- using judgement, and quantitatively in terms
ered when setting goals and to standardize the of some measurable aspect of the behaviour.
behavioural descriptions used, to an extent. The patient’s perception of quality (and, to a
lesser extent, the judgement of other people) is
of importance but it is not easily standardized.
Thus qualitative descriptions have been left out in
Part II. Specify specific support this method although an assessment of quality
Behaviour is a (goal-directed) interaction with could be used as an option if quantification is
the environment, whether objects or other not possible.
people. In rehabilitation it is often necessary to Performance can be quantified in three ways:
modify or provide additional environmental
factors for the behaviour to succeed. There are by the time taken to achieve a set quantity of
several environmental supports to consider, and the activity, and/or
thus this part is divided into three subparts. by the quantity of a continuous activity per-
The first subpart concerns support given by formed (e.g. distance) in a set time, and/or
people in the environment: by the quantity of a discrete activity occurring
in a period of time (e.g. its frequency).
hands-on, practical or physical assistance (such
as assisting in a transfer, cutting food, doing up Any activity that has a reasonably clear start
shoe laces); or and finish can be timed, and timing allows a rea-
emotional and stand-by support to increase sonably accurate and sensitive (to change) method
self-confidence; or of quantification that, incidentally, will often also
cognitive, structural support such as prompting be associated with the quality of performance.
and reminding. Timing should be widely used. Examples include
time to walk to the post office, time taken to get
The second subpart concerns specific objects in up and dressed, and time to complete a shopping
the environment – extra aids, or particular adapta- trip successfully. Generally (but not inevitably)
tions to objects – that need to be present. It covers time will be shortened as performance improves.
the field of physical equipment, for example: Distance or amount is commonly used to quan-
tify activities, for example the distance walked in 2
specific items that can be moved around (such minutes, or the number of words typed in 5 min-
as a walking stick, wheelchair, or hoist); or utes. It could also be the distance walked before
adaptation to personal items (such as clothing being stopped by pain, or the amount of time elap-
or cutlery); or sing before fatigue supervenes.
an adapted fixed environment (such as a ramp, Any activity that occurs repeatedly can also be
or a stair rail). counted. If the activity is a desired activity then an
Writing SMART rehabilitation goals 357
increase will usually be specified (e.g. number of Step 3: Scaling the goal
letters filed successfully) but counting can also
apply to unwanted activities (such as falling, In the goal attainment scaling process, once the
swearing, forgetting, needing prompts or dropping initial goal has been set in terms of the perfor-
objects) when a decrease will usually be the desired mance level expected at a specified time (which is
change. defined as the level scoring ‘0’), four more perfor-
mance levels need to be specified: two that are
better than and two that are worse than the goal.
The particular advantage of the structured
Part IV. Specify time period to achieve the desired approach to defining a goal outlined above (step 1)
state is that it allows easy definition of better than expected
The last step is to specify the time period over and worse than expected states. These states are
which (or date when) the target state is to be achieved by adding, deleting and/or varying one
achieved. In practice many services review pro- or more of the parts or subparts from step 1.
gress at set intervals varying from weekly, through Thus, states that indicate exceeding the goal will
every 4–6 weeks, up to every 3–6 months. This involve one or more of:
time will vary depending on the rehabilitation set-
ting (post-acute or longer term) and the goal set
succeeding with less support from people;
(most commonly short- or medium-term goals).
succeeding with a less supportive physical
It is important to remember that rehabilitation
environment;
concerns changing behaviours, which depends
succeeding with a less supportive ‘cognitive’
upon learning by the person or people concerned.
environment;
Behavioural change takes time. Consequently, in
being faster (usually);
complex cases it is rarely appropriate to set a
an increase in quantity (e.g. distance); and/or
review point at less than four weeks away.
doing the activity more or less frequently.
Moreover, the process described here is probably
too ‘expensive’ in terms of staff time to warrant its States that indicate underachievement will be
use for shorter term goals. The principles may be the reverse.
used by individual therapists, but setting complex The goal that was set in step 1 is level ‘0’; it is the
multiprofessional team goals simply for one or level that the team believes can be achieved by the
two weeks may best be done less formally. specified time. Two states that reflect a better out-
come than expected (þ1, þ2) and two states that
reflect a worse outcome than expected (1, 2)
need to be specified. Level 1 is somewhat less
Step 2: Weighting the goal than the expected level and level 2 is much less
than the expected level. Levels þ1 and þ2 are when
Traditionally in goal attainment scaling, each goal the patient performs somewhat better than expected
is weighted for importance and difficulty. and much better than expected, respectively.
However, it is possible not to score importance It is possible for one of these levels to be the
and difficulty and simply assign a weight of 1 to current level of performance (see discussion
the goal. If wanted, each goal can be weighted for later), but it will still need accurate specification
importance and/or difficulty. The importance is using this system.
determined by the patient, and the difficulty by
the clinician. Both importance and difficulty are
ranked on a 3-point scale, ranging from 1 (a little
importance/difficult) to 3 (very important/diffi- Step 4: Evaluating goal achievement
cult). If weighting is used, it needs to be used con-
sistently and uniformly for all goals and in all At the appointed review date the level achieved is
patients if any comparison is being undertaken. determined by the patient and the team.
358 TJH Bovend’Eerdt et al.
Step 5: Scoring goal achievement with Mr R and will have to supply the long-
handled sponge. The physiotherapist will have to
practise activity-related balance and the nursing
The score is calculated by applying the formula14, 15:
staff will have to implement the techniques in his
10ðWi Xi Þ daily routine.
GAS ¼ 50 þ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ð1 ÞW2i þ ðW2i Þ
where Wi is the weight (importance difficulty) Step 2: Weighting the goals
assigned to the i-th goal; Xi is the numerical value A weight for importance and difficulty is
achieved for the i-th goal; and is the expected assigned to the goal. For Mr R the goal is very
correlation of the goal scales (normally 0.3). important (score 3) and it is moderately
Calculating the score is discussed in more detail difficult (score 2). The weight for this goal is
elsewhere.11 importance difficulty; 3 2 ¼ 6.
Example 1. An illustration employing steps 1–3 Step 3: Defining other levels
Mr R, 73 years old, had a stroke two months ago. The goal is: To wash in the shower with verbal
He used to live independently in an apartment with prompting using a long-handled sponge in 15 min-
an adapted shower. The stroke has left him with utes on a daily basis using a checklist within four
slightly reduced balance and some apraxia. weeks (level 0). The other levels are defined by
He has expressed the desire to be able to wash him- adding, removing or changing one or more of
self in the shower on his own. It is anticipated that the (sub)-part (from II and III) that are specific
at the time of discharge Mr R will need a small care for Mr R.
package to provide some help at home. The inter-
disciplinary team will need to write a SMART goal Level 1 is the current level: To wash in the
for Mr R to work towards washing himself in the shower with physical assistance of one person
shower on his own. on a shower chair within four weeks.
Level 2 is less than current: To wash in the
shower with physical assistance of one person
Step 1: Defining the goal on a shower wheelchair within four weeks.
By selecting possibilities from each (sub)-part Level 1 is somewhat better than expected: To
a SMART goal is created. The occupational thera- wash in the shower with a long-handled sponge
pist suggests that Mr R needs verbal prompting in 15 minutes on a daily basis within four weeks.
(support by people) to perform this activity and Level 2 is much better than expected: To inde-
would be safe doing this if he had a long-handled pendently wash in the shower in 15 minutes on
sponge (support by objects). The psychologist sug- a daily basis within four weeks.
gests using a checklist (cognitive, structural, com-
munication support) to increase his independence.
He should be able to do it within 15 minutes
(quantifying by timing) on a daily basis (quanti-
fying by frequency) within four weeks (time period
Practical application – some points
to achieve state). So the result is the SMART goal:
To wash in the shower with verbal prompting We have learned four lessons from our experience
using a long-handled sponge in 15 minutes on a with goal setting in rehabilitation in general and
daily basis using a checklist within four weeks. with the method described here:
This goal is clear for Mr R as well as for the
interdisciplinary team. The psychologist will have Getting the team to work together as a team.
to teach him the strategy of checklists. The occu- When setting goals, aim for them to be
pational therapist will have to practise the activity interdisciplinary (i.e. to require collaborative
Writing SMART rehabilitation goals 359
working by two or more team members). Discussion
Interdisciplinary rehabilitation is effective,16
and creating interdisciplinary goals improves We describe a new, structured method for writing
the collaboration of the various disciplines goals that are specific and measurable without too
and creates clear aims for the patient and the much effort. Goal attainment scaling is a techni-
disciplines (see example 2). que that is increasingly used in rehabilitation, but
Making goals relevant and important. This its success depends upon formulating unambigu-
should follow on from discussions with the ous goals and the method described here is a useful
patient, but we find it is much more likely if tool to achieve this: it is flexible enough to cover
goals are based on activities (or social partici- most situations; it is patient-specific; it saves time
pation) (see example 2). and effort; and it can easily be taught and used by
Scoring goal attainment (a). When scoring the whole team. In this article we have focused on
goal attainment (step 4), it is possible that the construction of the target goals and levels of
none of the predefined levels precisely repre- achievement (steps 1–3). More detailed informa-
sents the patient’s level. However, it is our tion on goal attainment scaling in general is avail-
experience that the team can score the appro- able elsewhere.6,10,14,17 Some particular additional
priate level without significant difficulty. points that we have considered are discussed here.
In example 1, the attainment of the patient Weighting the importance and difficulty of
after four weeks is actually: able to wash in goals seems intuitively good, and different meth-
the shower with verbal prompts in 10 minutes ods for weighting are available. In the accompany-
on a daily basis using a checklist. This level is ing article by Turner-Stokes11 a 4-point (0–3)
not one of the predefined levels. However, it is weighting scale is suggested. However, this
fairly obvious that the appropriate level is level means that items rated ‘0’ score ‘0’. This may be
0. We suggest that this difference is noted on appropriate in that unimportant goals or goals
the score sheet. that can easily be achieved should not be set and
Scoring goal attainment (b). When scoring goal should not score at all. However in our view, it is
attainment (step 4), occasionally the level inappropriate to waste time setting goals that are
achieved is in between two predefined levels. of no importance and/or are very easily achieved.
We suggest always choosing the lower Consequently we have restricted the scale to 1, 2 or
(less good) level in this case and making a 3 (see Figure 1).
note of the actual level on the score sheet. Furthermore, the additional value (information
content) of weighting goals is unknown, and it is
quite possible that it is an unnecessary complica-
tion. Whether importance and/or difficulty are
scored or whether goals are always assigned a
Example 2. Illustration of an interdisciplinary weight of 1, it is vital that a consistent approach
goal at the activity level is used for all goals given to an individual patient
The physiotherapist has identified weak hip and for all patients where aggregation of data is
extensor power and poor stability around the contemplated. Comparing patients where impor-
hips. An obvious goal for the physiotherapist tance and difficulty are scored with patients
would be: To bridge with verbal prompting to where only importance is scored or without any
clear the bed by 10 cm and hold for 5 seconds weighting at all is completely invalid.
within four weeks. This goal complies reasonably The score attributed to the current state when
well with the SMART criteria: specific, measur- goals are set is also subject to debate. Some
able, achievable, relevant and timed. However, a authors set ‘2’ as the current state, but this
better goal that would be interdisciplinary, prob- approach risks missing a deterioration in the
ably more relevant and at the activity level, would patient’s state15 (i.e. there would be a floor
be: To pull up the trousers independently, using effect). One suggested remedy is to add a further
bridging, within four weeks. level, ‘3,’ to indicate deterioration from the
360 TJH Bovend’Eerdt et al.
current level (set as ‘2’).18 Another suggested state such as pain, mood and quality of life because
remedy is to set the current state as ‘1,’ but behavioural correlates usually exist. For example a
although this allows for deterioration it reduces patient’s self-report is in fact a behaviour and, more
sensitivity to improvement by removing the importantly, these subjective states usually have
option of ‘has made some progress, but not as externally observed behavioural sequelae such as
much as expected’. taking symptomatic treatments (e.g. analgesic
Our suggestion is to set the current state at the drugs), reducing or altering other activities
extreme (‘2’), and to score (record) any dete- (e.g. sleeping less long, not going to work).
rioration as ‘3’ in the patient’s record but to Two final points must be emphasized. The
score it as ‘2’ when scoring, acknowledging method described here has not itself been evalu-
that this overestimates their state. Unexpected ated against other techniques for defining different
deterioration is sufficiently rare to make this a outcome states (such as using a predefined menu).
minor problem. However, it is currently being used in clinical prac-
Goal setting in general, and goal attainment tice and as an outcome measure in a randomized
scaling in particular, has generally been applied controlled trial, and it is proving to be a user
when improvement (recovery) is the expected friendly, practical and quick tool in both clinical
direction of change. However, goals can also be and research practice, without compromising the
set in situations where deterioration is the expecta- patient-centred and individualized approach.
tion (e.g. in motor neurone disease); under these Second, in this discussion we have outlined sev-
circumstances the goal of treatment is to reduce eral variations on the theme of scoring the achieve-
the extent or consequences of disease progression. ment of outcome (concerning weighting, score
The same general approach to scoring should be attributed to the current state, handling unexpected
used here: ‘þ2’ would represent an outcome state change, etc.). A clinical team or researcher may
much better than expected and ‘2’ a state much choose whichever approach seems most appropri-
worse than expected. As above, it is probably best
ate in their circumstances, but it is imperative that
to set the current state as ‘þ2’, with ‘þ3’ being
only one method is used with all patients to be
used to record (but not score) a completely unex-
analysed in a group, and that the exact methods
pected improvement. In patients who deteriorate,
used in any report or analysis are specified.
the level ‘0’ (‘expected state’) is the state antici-
pated as a result of the intervention with ‘1’
and ‘2’ being worse states.
The approach we have put forward depends Clinical messages
upon defining different states at a fixed time. In
principle it would be possible to fix a state and to Goals in rehabilitation can be constructed
vary the time taken to reach that state as an alter- using four parts: the target activity, the sup-
native means of scoring. For example the state port needed, quantification of performance
might be ‘washing up breakfast dishes without and the time for achievement.
being reminded and without breakage’ and one This method can scale outcome by changing
could aim to achieve this by six weeks, with (sub)parts to give five levels.
achievement by five weeks being level ‘þ1,’ four The method is easy and quick, patient-speci-
weeks ‘þ2,’ and seven weeks ‘1’ and not achiev- fic, and applies to most situations.
ing it by 10 weeks being ‘2’. We are unaware of
this method being used for goal attainment scal-
ing, but recording the time to achieve a state
(such as recurrence of a cancer) is a common ana- Acknowledgements
lytic technique. We would like to thank all the staff at the
The method has been described here primarily in Oxford Centre for Enablement and Claire Guy
relation to activities. These are most easily for their help and effort in developing this
described. The technique should, however, be method. We would also like to thank Joan
applicable to most outcomes, including subjective Warren for her financial support.
Writing SMART rehabilitation goals 361
References community mental health programs. Community
Ment Health J 1968; 4: 443–53.
1 Shiell A, Hawe P, Gold L. Complex interventions 11 Turner-Stokes L. Goal attainment scaling (GAS) in
or complex systems? Implications for health rehabilitation: a practical guide. Clin Rehabil 2009;
economic evaluation. Br Med J (Clin Res Ed) 23: 362–70.
2008; 336: 1281–83. 12 Jelles F, Van Bennekom CA, Lankhorst GJ,
2 Wade DT. Goal planning in stroke rehabilitation: Sibbel CJ, Bouter LM. Inter- and intra-rater
How? Topics Stroke Rehabil 1999; 6: 16–36. agreement of the Rehabilitation Activities Profile.
3 Locke EA, Bryan JF. The effects of goal-setting, J Clin Epidemiol 1995; 48: 407–16.
rule-learning, and knowledge of score on 13 Geyh S, Cieza A, Schouten J, Dickson H,
performance. Am J Psychol 1966; 79: 451–57. Frommelt P, Omar Z et al. ICF core sets for stroke.
4 Locke EA, Latham GP. Building a practically J Rehabil Med 2004; 44(suppl): 135–41.
useful theory of goal setting and task motivation. 14 Tennant A. Goal attainment scaling: current
A 35-year odyssey. Am Psychol 2002; 57: methodological challenges. Disabil Rehabil 2007;
705–17. 29: 1583–88.
5 Schmidt RA, Wrisberg WA. Motor learning and 15 Law LSH, Dai MOS, Siu A. Applicability of goal
performance: a situation-based learning approach, attainment scaling in the evaluation of gross motor
fourth edition. Human Kinetics, 2007. changes in children with cerebral palsy. Hong Kong
6 Hurn J, Kneebone I, Cropley M. Goal setting as an Physiother J 2004; 22: 22–28.
outcome measure: a systematic review. Clin Rehabil 16 How do stroke units improve patient outcomes? A
2006; 20: 756–72. collaborative systematic review of the randomized
7 Schut HA, Stam HJ. Goals in rehabilitation trials. Stroke Unit Trialists Collaboration. Stroke
teamwork. Disabil Rehabil 1994; 16: 223–26. 1997; 28(11): 2139–44.
8 Yip AM, Gorman MC, Stadnyk K, Mills WG, 17 Rockwood K, Howlett S, Stadnyk K, Carver D,
MacPherson KM, Rockwood K. Powell C, Stolee P. Responsiveness of goal
A standardized menu for Goal Attainment attainment scaling in a randomized controlled
Scaling in the care of frail elders. The Gerontologist trial of comprehensive geriatric assessment.
1998; 38: 735–42. J Clin Epidemiol 2003; 56: 736–43.
9 World Health Organization. International 18 Steenbeek D, Meester-Delver A, Becher JG,
classification of functioning, disability and health. Lankhorst GJ. The effect of botulinum toxin type
Accessed 23 September 2008, from: www.who.int/ A treatment of the lower extremity on the level of
classifications/icf/en/ functional abilities in children with cerebral palsy:
10 Kiresuk TJ, Sherman RE. Goal attainment scaling: evaluation with goal attainment scaling.
a general method for evaluating comprehensive Clin Rehabil 2005; 19: 274–82.
