Constraint induced language therapy

Evidence-Based Systematic Review:
Effects of Intensity of Treatment
and Constraint-Induced Language
Therapy for Individuals With
Stroke-Induced Aphasia
Purpose: This systematic review summarizes evidence for intensity of treatment and
constraint-induced language therapy (CILT) on measures of language impairment and
communication activity/participation in individuals with stroke-induced aphasia.
Method: A systematic search of the aphasia literature using 15 electronic databases
(e.g., PubMed, CINAHL) identified 10 studies meeting inclusion/exclusion criteria. A
review panel evaluated studies for methodological quality. Studies were characterized
by research stage (i.e., discovery, efficacy, effectiveness, cost–benefit/public policy
research), and effect sizes (ESs) were calculated wherever possible.
Results: In chronic aphasia, studies provided modest evidence for more intensive
treatment and the positive effects of CILT. In acute aphasia, 1 study evaluated
high-intensity treatment positively; no studies examined CILT. Four studies reported
discovery research, with quality scores ranging from 3 to 6 of 8 possible markers. Five
treatment efficacy studies had quality scores ranging from 5 to 7 of 9 possible markers.
One study of treatment effectiveness received a score of 4 of 8 possible markers.
Conclusion: Although modest evidence exists for more intensive treatment and CILT for
individuals with stroke-induced aphasia, the results of this review should be considered
preliminary and, when making treatment decisions, should be used in conjunction
with clinical expertise and the client’s individual values.
KEY WORDS: aphasia, intensity, constraint, rehabilitation
Constraint-induced language therapy (CILT), a relatively new apha-sia
treatment, (Pulvermuller et al., 2001), has garnered consid-erable
interest from consumer groups and clients with aphasia.
The major components of CILT involve both forced use of verbal language
and massed practice (i.e., high intensity of treatment). Both are of con-siderable
importance in light of recent work in neuroscience demonstrat-ing
that the neuroplasticity of the adult brain can be impacted by several
experience-dependent principles, including intensity of training and forced
use of cognitive capacities (Kleim & Jones, 2008). Because intensity of
treatment is an important variable that affects neuroplasticity, any dis-cussion
of CILT must consider the impact not only of constraint but also
of treatment intensity on outcomes. Therefore, the aim of this evidence-based
systematic review (EBSR) was to examine the current state of the
evidence that used one or both principles of CILT for individuals with
stroke-induced aphasia: (a) constraining treatment responses to the verbal
Leora R. Cherney
Rehabilitation Institute of Chicago, IL
Janet P. Patterson
California State University,
East Bay, Hayward, CA
Anastasia Raymer
Old Dominion University, Norfolk, VA
Tobi Frymark
Tracy Schooling
American Speech-Language-Hearing
Association, Rockville, MD
Journal of Speech, Language, and Hearing Research • Vol. 51 • 1282–1299 • October 2008 • D American Speech-Language-Hearing Association
1092-4388/08/5105-1282
1282

modality only and (b) providing treatment on an inten-sive
delivery schedule.
The neurobiological principles and rehabilitation
techniques surrounding CILT were derived from animal
studies (Taub, 1977) as well as from a series of investi-gations
of limbmovements in patients with chronic stroke
and hemiplegia (Taub,Miller, Novack,&Cook, 1993; Taub
&Wolf, 1997). These studies found that motor behavior of
an impaired limb was modifiable with intensive practice
over a short period of time. The techniques collectively
known as constraint-induced movement therapy (CIMT)
involve restraining the use of the less affected arm for
the majority of waking hours to counteract the presumed
learned nonuse of the affected arm. Repetitive training
and shaping of movements of the affected arm take place
in intensive treatment (6 hr per day) over 2 consecutive
weeks.
In the past 20 years, a large body of evidence, ap-plying
either the original CIMT technique or a variant of
this approach, has accumulated in support of the efficacy
of constraint-induced (CI) therapy for rehabilitating
hemiparetic arm use in individuals with chronic stroke
(Hakkennes & Keating, 2005; Taub, Uswatte, & Piclikiti,
1999). To date, there are more than 120 published stud-ies,
including a single-site, placebo-controlled trial (Taub
et al., 2006) and a large multisite randomized trial (Wolf
et al., 2006). Although the magnitude of the treatment
effects have varied throughout the CIMT literature, the
generally positive outcomes have encouraged investigators
to examine the effects of CI and their potential benefits in
other domains of stroke rehabilitation, including aphasia.
Since being introduced to the field of aphasiology by
Pulvermuller and colleagues (2001), CILT has been met
with great interest. Pulvermuller et al. argued that pa-tients
with aphasia often use the communication chan-nel
that is accessible to them with the least amount of
effort, such as gestures and drawing, rather than spoken
language. A constrained approach for aphasia is incor-porated
in the therapeutic setting by forcing the patient
to communicate only through verbal channels while lim-iting
the use of all other communication channels. The
implementation of CILT in patients with aphasia con-sisted
of intensive practice through a series of language
games requiring verbal production of phrases and sen-tences
of increasing length that were shaped into succes-sively
better approximations.Results comparing language
improvements after 2 weeks of CILT to approximately
4 weeks of traditional therapy indicated significant im-provements
in the CI group on several standard clinical
tests, self-ratings, and blinded-observer ratings of the
patients’ communicative effectiveness in everyday life.
Two principles of CILT need to be considered when
interpreting the findings. The first principle, forced use
of verbal language, requires that all responses provided
during language treatment activities be constrained or
restricted to the spoken modality. Nonverbal communi-cation
attempts such as gesturing, drawing, or writing
are not allowed and are not considered as communica-tion
intents because participants are on opposite sides of
a barrier. The second principle, massed practice, involves
a high-intensity treatment schedule consisting of 3–4 hr
of treatment per day for 2 weeks. Participants in the
Pulvermuller et al. (2001) study were randomized to a
CILT group and a comparison group that received con-ventional
speech treatment provided on a distributed
schedule over 3–5 weeks. That is, the groups differed in
two dimensions of treatment: the type of treatment ad-ministered
and the intensity of the treatment provided.
Thus, it is difficult to determine whether the positive
treatment results reported for the CILT group as com-pared
to the conventional treatment group emanate from
the constrained forced language use, the intensity of the
treatment schedule, or a combination of these two factors.
Intensity of treatment has been a topic of interest for
some time in aphasia treatment studies (Poeck, Huber,
& Willmes, 1989). Findings from Robey’s (1998) meta-analysis
of the aphasia literature reported large effect
sizes (ESs) associated with treatment provided for 2 or
more hr per week. In a review focused on intensity and
outcomes, Bhogal, Teasell, and Speechley (2003) reported
better treatment outcomes in studies that provided in-tensive
treatment schedules. On average, the more in-tensive
treatment schedules equaled 8.8 hr per week for
11 weeks, compared with the less intensive schedules of
2 hr per week for 23 weeks of treatment. Although both
reviews examined the intensity literature, the schedule
of treatment provided in the cited studies is far less than
the intensity of services given in CILT.
This review examines the current state of the evi-dence
employing the dual principles of CILT for individ-uals
with stroke-induced aphasia. In addition to analyzing
and synthesizing the scientific research in this area, the
authors piloted the use of a newly created levels-of-evidence
scheme developed by the American Speech-
Language-Hearing Association (ASHA) National Center
for Evidence-Based Practice in Communication Disor-ders
(N-CEP), along with its advisory committee. This
scheme classifies treatment studies according to several
quality indicators and the phase of research (Robey,
2004). We piloted this system to determine its applica-bility
to treatment studies in the area of communication
sciences and disorders and provided feedback to N-CEP
and its advisory committee prior to the widespread
adoption of this system.
Clinical Questions
Ten clinical questions were identified by the authors
for review (see Table 1). In constructing these questions,
Cherney et al.: EBSR: Effects of Intensity of Treatment and CILT 1283

the nature of outcome measurements and stage of re-covery
from aphasia were considered. In keeping with
current ASHA policy documents, such as the Scope of
Practice in Speech-Language Pathology (ASHA, 2001),
the World Health Organization’s (2001) International
Classification of Functioning, Disability and Health: ICF
was selected to categorize the various outcome measures
representing language impairment (e.g., standard apha-sia
tests of phonology, syntax, and semantics) or commu-nication
activity/participation (e.g., functional real-life
use of language in connected speech measures or com-munication
ratings scales). Because the stage of recovery
from stroke-induced aphasia—acute or chronic—influences
recovery patterns (Robey, 1998) and studies typically
include participants in only one of these stages, the clin-ical
questions also included participants with chronic
and acute aphasia (distinguished by mean time post
onset [TPO]≤ 3months). Therefore, eight clinical ques-tions
reflect combinations of the following parameters:
(a) CILT principles—response constraint and treatment
intensity; (b) aphasia chronicity—acute and chronic; and
(c) outcome measures identified from the ICF—language
impairment and communication activity/participation.
Two additional questions were constructed to address
maintenance of treatment behavior in chronic aphasia
for a total of 10 clinical questions (see Table 1).
Method
A systematic search of the aphasia literature was
conducted to identify studies that directly investigated
CILT as the specific intervention for aphasia or directly
compared conditions of higher and lower intensity treat-ment
for aphasia. Fifteen electronic databases were
searched: PubMed, CINAHL, PsycINFO, PsycArticles,
CSA Linguistics and Language Behaviour Abstracts,
Combined Health Information Database, Health Source:
Nursing, Science Citation Index, ScienceDirect, NeLH,
REHABDATA, Social ScienceCitation Index, SUMSearch,
TRIP Database, and Cochrane Database of Systematic
Reviews. Additional searches were also performed on all
ASHA journals, the University of Pittsburgh Clinical
AphasiologyConference Proceedings,National Institutes
of Health (NIH) abstracts, and Google Scholar. In addi-tion,
references from all relevant articles were examined
to identify other applicable studies. A total of 25 ex-panded
search terms included those related to stroke-induced
aphasia, amount and intensity of treatment, and
CILT. Studies were initially considered for the review if
theywere published in a peer-reviewed journal from1990
to 20061, were written in English, and contained original
data addressing 1 ormore of the 10 clinical questions. For
studies addressing a clinical question related to intensity,
the keyword intensity or amount of treatment had to be
included in the abstract or the title of the article. In ad-dition,
studies were limited to those that included adults
18 years of age or older with stroke-induced aphasia.
Studies that included individuals with underlying cog-nitive
deficits or any other primary medical diagnoses
were excluded. Studies that used pharmacological inter-vention
for aphasia as the comparison treatment were
also excluded.
The initial search schematicized in Figure 1 yielded
441 citations. Two authors (TF and TS), blind to one an-other’s
results, reviewed each abstract and identified
36 abstracts as meeting the initial inclusion criteria with
91% agreement. Of those preliminarily accepted, 26 were
Table 1. Clinical questions.
Question No. Clinical Question—Intensity
1. For stroke-induced chronic aphasia, what is the influence of treatment intensity on measures of language impairment?
2. For stroke-induced chronic aphasia, what is the influence of treatment intensity on measures of communication activity/participation?
3. For stroke-induced acute aphasia, what is the influence of treatment intensity on measures of language impairment?
4. For stroke-induced acute aphasia, what is the influence of treatment intensity on measures of communication activity/participation?
5. For stroke-induced chronic aphasia, what treatment outcomes are maintained following intensive language treatment?
Clinical Question—CILT
6. For stroke-induced chronic aphasia, what is the influence of constraint-induced language therapy on measures of language impairment?
7. For stroke-induced chronic aphasia, what is the influence of constraint-induced language therapy on measures of communication
activity/participation?
8. For stroke-induced acute aphasia, what is the influence of constraint-induced language therapy on measures of language impairment?
9. For stroke-induced acute aphasia, what is the influence of constraint-induced language therapy on measures of communication
activity/participation?
10. For stroke-induced chronic aphasia, what treatment outcomes are maintained following constraint-induced language therapy?
1The original purpose of the systematic review was to examine the effects
of CILT. It became apparent that it would be impossible to adequately
complete this review independent of addressing treatment intensity, which
is a key principle of CILT. The search date was determined based on the
emergence of CI literature in the early 1990’s.
1284 Journal of Speech, Language, and Hearing Research • Vol. 51 • 1282–1299 • October 2008

rejected upon review of the full text by the author panel,
as they did not report original data or include direct evi-dence
of a contrast between conditions (e.g., a more in-tensive
condition vs. a less intensive condition) pertaining
to one or more of the clinical questions. Ten studies met
the final inclusion.
Each of these studies was independently reviewed
and evaluated for methodological quality by two of three
authors (LC, JP, AR). One study written by an author
(AR) was reviewed by two other panel authors (LC, JP).
The authors, blind to one another’s results, appraised
each study on the basis of quality indicators described in
Table 2. For each indicator, quality markers are ordered
from highest to lowest level of quality. A study received
1 point for each quality indicator if the highest level of
quality was incorporated. All nine indicators were rel-evant
to studies incorporating controlled trials, leading
to amaximum quality score of 9. For other study designs,
where intent-to-treat analysis was not relevant, the
maximum quality score was 8. Credit for study design
varied by stage of research (see Table 3). For example,
efficacy studies using controlled trial designs received
1 point. Discovery studies received 1 point for research
design if reporting retrospective case controls or single-participant
studies. Reliability of scores was assessed,
and any discrepancies in scoring were sent back to the
full author panel for resolution. All disagreementswere
resolved via consensus.
In addition to assessing methodological quality, the
phase of clinical research was determined using the
decision tree depicted in Figure 2. The four stages of
research—discovery, efficacy, effectiveness, and cost–
benefit/public policy research—are defined in Table 3.
A final synthesis of the body of scientific literature was
compiled into an evidence table on the basis of the study
qualitymarker score and corresponding stage of research.
ESs were calculated for outcomemeasureswhenever
possible. For group studies, Cohen’s d (Cohen, 1988) was
calculated (corrected for bias following Hedges & Olkin,
1985). ESs were calculated from group means and stan-dard
deviations or were estimated from results of anal-yses
of variance (ANOVAs) or t tests. We used Cohen’s
benchmarks for small, medium, and large ESs as .2, .5,
and .8, respectively. Whether these benchmarks are ap-plicable
to aphasia treatment studies has yet to be de-termined.
In the one single-participant experiment
(Raymer, Kohen,&Saffell, 2006), the authors calculated
theweighted ESs using the methods described by Beeson
and Robey (2006). Although Busk and Serlin (1992)
reported that large ESs in single-participant studies sur-pass
2.5, Beeson and Robey noted benchmarks for small,
medium, and large ESs in single-participant designs as
2.6, 3.9, and 5.8.
Results
Of the 10 studies that met inclusion criteria for the
systematic review, 5 studies investigated treatment in-tensity
(Basso & Caporali, 2001; Denes, Perazzolo, Piani,
& Piccione, 1996; Hinckley & Carr, 2005; Hinckley &
Craig, 1998; Raymer et al., 2006), 4 studies evaluated
Figure 1. Process for identification of included studies.

CILT (Maher et al, 2006; Meinzer et al, 2004; Meinzer,
Djundja,Barthel,Elbert,&Rockstroth, 2005; Pulvermuller,
Hauk, Zohsel, Neininger, & Mohr, 2005), and 1 study ex-amined
both (Pulvermuller et al., 2001). These studies
allowed us to address 7 of the 10 clinical questions (see
Table 1): Questions 1, 2, 5, 6, 7, and 10, which are related
to language impairment and communication activity/
participation in individuals with chronic aphasia, and
Question 3, which is related to the effects of treatment
intensity on measures of language impairment for in-dividuals
with acute aphasia. No study investigated
CILT or intensive treatment effects for measures of
Table 2. Quality indicators in the ASHA levels-of-evidence scheme.
Indicator Description Quality marker
Study design The type of design used in the study. Controlled trial (Efficacy research)*
Retrospective case control (Effectiveness or
cost–benefit research)*
Single participant study (Discovery research)*
Case series
Case study
Blinding The practice of keeping investigators or participants ignorant to
the group to which participants are assigned. For the purposes
of the critical appraisal, blinding refers to assessors only.
Assessors blinded*
Assessors not blinded or not stated
Sampling The method(s) used to choose and assign participants to the
experimental conditions in the study.
Random sample adequately described*
Random sample inadequately described
Convenience sample adequately described
Convenience sample inadequately described or
hand-picked sample or not stated
Group/participant
comparability
How similar the participants/groups were at the start of the
study or how adequately they were described.
Groups/participants comparable at baseline
on important factors (between-subject design)
or participant(s) adequately described
(within-subject design)*
Groups/participants not comparable at baseline
or comparability not reported or participant(s)
not adequately described
Treatment fidelity The procedure used to ensure that the treatment
was delivered as intended.
Evidence of treatment fidelity*
No evidence of treatment fidelity
Outcomes The measure(s) used in the study to quantify improvement. At least one primary outcome measure is
valid and reliable*
Validity unknown, but appears reasonable;
measure is reliable
Invalid and/or unreliable
Significance The likelihood that the study findings occurred by chance. p value reported or calculable*
p value neither reported nor calculable
Precision The size or magnitude of any difference found between the
treatment under investigation and the control condition.
Effect size and confidence interval
reported or calculable*
Effect size or confidence interval, but not
both, reported or calculable
Neither effect size nor confidence
interval reported or calculable
Intent-to-treat
(controlled trials only)
Participants are analyzed according to the group to which they were
initially assigned, regardless of whether or not they dropped out,
fully complied with the treatment, or crossed over and received
the other treatment. If number of participants at pre-treatment =
number at post-treatment, study received credit.
Analyzed by intent-to-treat*
Not analyzed by intent-to-treat*
*Highest level of quality needed to receive 1 point.

Table 3. Continuum of research stages.
Stage Description
Discovery Treatment approaches are developed and assessed in the context of whether
communication activity/participation in acute aphasia
(covered in Questions 4, 8, and 9).
Participants
Table 4 provides a detailed description of the 141 par-ticipants
in the 10 studies (e.g., age, education level,
gender, handedness, TPO, aphasia type, and severity).
The total sample size (N) of participants reported in the
studies does not account for the possible overlap of par-ticipants
across four studies: Pulvermuller et al. (2001
and 2005) and Meinzer et al. (2004 and 2005).
For the intensity studies, type of aphasia was pro-vided
in five studies; 81% (55 of 68) of the participants
were nonfluent. Severity of aphasia was provided in four
studies; 40% (23 of 57) of participants tended to have
more severe aphasia, with 30% (19 of 63) of participants
being described as having global aphasia. The mean TPO
was considered acute in one study (3.1 months in Denes
et al., 1996); the other studies included participants with
chronic aphasia, ranging from 11.2 to 67.6 months postonset.
Similarly for the CILT studies, 60% of participants
(42 of 70) had nonfluent forms of aphasia, typically de-scribed
as Broca’s aphasia. Most had moderate aphasia
(48%, 30 of 62), whereas 19 participants (31%) had mild
aphasia and 12 participants (19%) had severe aphasia.
Mean TPO ranged from 36.8 months to 90 months, all in
the chronic range.
Four of the 10 studies were conducted in Germany,
4 in the United States (U.S.), and 2 in Italy. Two German
studies (Pulvermuller et al., 2001, 2005) included mono-lingual
native speakers of German. The other two Ger-man
studies (Meinzer et al., 2004, 2005) did not mention
linguality of the participants. Two U.S. studies (Hinckley
& Carr, 2005; Hinckley & Craig, 1998) noted that all par-ticipants
were native speakers of English, whereas two
Figure 2. Stage of research.
they show promise of being efficacious.
Efficacy Promising interventions are tested in a rigorous way under ideal, highly
controlled conditions to determine the outcome that results.
Effectiveness The intervention is tested in a “real-world” clinical setting. This phase
is often conducted if the intervention demonstrates positive outcomes in a
highly controlled setting of a clinical trial.
Cost–benefit/public policy A study is conducted of the political and economic environment in which the
intervention is best delivered. This phase is often conducted once it has
been shown that an intervention is both efficacious and effective.

Table 4. Patient characteristics.
Study N Age Education Gender Handedness TPO Etiology Aphasia type
Aphasia
severity
Aphasia
severity
rating
Basso & Caporali,
2001 (Intensity)
6 35–48 yrs 8–17 yrs 4 M R 2–22 months 2 hemorrhagic 2 mixed nonfluent NR Test scores
M = 41 yrs M = 12.5 yrs 2 F M = 11.2 months 4 ischemic 2 global
2 nonfluent agrammatic
Denes et al.,
1996 (Intensity)
17 60.2 yrs 7.2 yrs 8 M R 2.4–4.5 months 4 hemorrhagic global Severe Test scores
9 F M = 3.1 months 13 ischemic
Hinckley & Craig,
1998, Study 3
(Intensity)
10 25–78 yrs 12+ yrs 6 M 9 R 7–47 months Thromboembolic 8 nonfluent 1 severe Subjective
M = 48 yrs 4 F 1 L M = 20 months 2 fluent 7 moderate
2 mild
Hinckley & Carr,
2005 (Intensity)
13 19–72 yrs NR NR R 6–99 months L CVA 12 Broca BDAE severity
score 1–3
Test scores
M = 50 yrs M = 27 months intensive 1 transcortical motor
M = 40 months nonintensive
Maher et al., 2006 (CILT) 9 41–73 yrs 12–16+ yrs 6 M 8 R 14–72 months L CVA NR 4 severe Test scores
M = 58 yrs 3 F 1 R+ M = 36.8 months 1 moderate
3 mild
1 NS
Meinzer et al., 2004 (CILT) 28 35–80 yrs NR 14 M R 12–156 months 20 ischemic 13 Broca NR NR
M = 54.6 yrs 14 F M = 43.8 months 8 hemorrhagic 4 Wernicke
3 global
6 not classified
2 amnesic
Meinzer et al., 2005 (CILT) 27 18–80 yrs NR 16 M NR 13–116 months 11 hemorrhagic 10 Broca 2 severe Test scores
M = 51.5 yrs 11 F M = 43.1 months 16 ischemic 8 Wernicke 15 moderate
3 amnesic 10 mild
1 global
5 not classified
Pulvermuller et al., 2001
(CILT and Intensity)
17 42–72 yrs 9–13 years 12 M 14 R 2–223 months L CVA 10 Broca 4 severe Test scores
M = 54.9 yrs M = 11 yrs 5 F 3 R+ M = 67.6 months 4 Wernicke 9 moderate
1 amnesic 4 mild
1 transcortical
1 conduction
Pulvermuller et al.,
2005 (CILT)
9 39–72 yrs 10–13 yrs 6 M 7 R 16–223 months L CVA 5 Broca 2 severe Test scores
M = 54.4 yrs M = 11.3 yrs 3 F 2 R+ M = 90 months 2 Wernicke 2 moderate
1 amnesic 2 mild
1 transcortical
Raymer et al.,
2006 (Intensity)
5 51–82 yrs 8–14 yrs 2 M 5 R 4–42 months L CVA 2 Broca NR Test scores
M = 70.8 yrs M = 11.4 yrs 5 F M = 18.4 months 2 conduction
1 mixed transcortical
Note. TPO = time postonset; yrs = years; NR = not reported; R+ = ambidextrous; L CVA = left cerebrovascular accident; NS = not severe; BDAE = Boston Diagnostic Aphasia Examination (Goodglass &
Kaplan, 1983).

other U.S. studies (Maher et al., 2006; Raymer et al.,
2006) did not address linguality.One Italian study (Denes
et al., 1996) included only native speakers of Italian. The
second Italian study (Basso & Caporali, 2001) included
5 participants who were native speakers of Italian and 1
who was a native of Sweden and completely bilingual.
The educational level of the participants was reported
in only 6 of the 10 studies. The mean years of educa-tion
varied from 10.9 years (Pulvermuller et al., 2001) to
15 years (Maher et al., 2006). One study (Hinckley &
Craig, 1998) reported that all participants completed
high school or above, and the Basso and Caporali (2001)
study reported educational attainment on only 4 of 6 par-ticipants.
Only one study (Hinckley&Carr, 2005) reported
mean data on socioeconomic status. Those participants
had a mean score of 2.3 on the Hollingshead Four Factor
Index of Social Status (Hollingshead, 1975).
Study Characteristics and Quality
Table 5 summarizes information regarding the in-tervention,
including schedule, amount and duration of
treatment, and outcomemeasures.Most studies provided
24–30 hr of treatment; several studies reported more
than 100 hr of treatment (e.g., Hinckley & Craig, 1998).
Treatment schedules varied across studies, as did the
nature of treatment provided.
Table 6 displays quality marker ratings obtained for
the studies. All 10 studies described participants suf-ficiently
and reported data in a manner in which statis-tical
significance was available. Studies were lacking in
several areas, however, including use of a clearly described
randomized assignment scheme for participants in group
designs, provided in only one study; blinding of assessors
to treatment conditions, indicated in only two studies; and
providing evidence of treatment fidelity, also noted in
only two studies.
Four studies were judged as discovery phase re-search
and five studies as efficacy research. Only one
study fell into the effectiveness stage of research, and
none addressed cost–benefit or public policy.
Intensity Results
Table 7 displays stage of research, quality scores,
and ESs for the six studies examining treatment inten-sity.
These studies addressed Questions 1, 2, 3, and 5
(see Table 1). No data were available to answer Ques-tion
4. Five of these studies contained sufficient data
for calculation of treatment ESs. The effect of intensity
in Denes et al. (1996), Pulvermuller et al. (2001), and
Hinckley andCarr (2005) was derived from between-group
comparisons for groups receiving intensive and non-intensive
treatment. The effect of intensity in Study 3 of
Hinckley andCraig (1998)was derived fromwithin-group
comparisons of the pre- and post difference scores from
each intensive 6-week training session compared with the
nonintensive 6-week training session. In Raymer et al.
(2006), the effects came from within-subject comparisons
across the individual participants.
Clinical Question 1: For stroke-induced chronic apha-sia,
what is the influence of treatment intensity on mea-sures
of language impairment? Four group studies used
impairment outcome measures for which eight ESs were
calculable, including seven large ESs, all in favor of more
intensive treatment. In the single-participant design of
Raymer et al. (2006), ESs were larger in the more inten-sive
condition for picture-naming acquisition and larger
in the less intensive condition for word/picture verifica-tion.
ESs could not be calculated for Basso and Caporali
(2001), who described case studies of three pairs of indi-viduals.
Individuals receiving more intensive treatment
showed greater gains on language impairment tasks than
did the comparison individualswho received a less intensive
schedule. Thus, the language impairment outcome mea-sures
favored more intensive treatment for all language
measures except one in participants with chronic aphasia.
what is the influence of treatment intensity on mea-sures
of communication activity/participation? Three
group studies used communication measures for which
nine ESs were calculable, including content unit (CU)
analysis, oral and written responses on a catalogue order-ing
task administered in quiet and concurrent noise con-ditions,
and the CommunicationActivities of Daily Living
(CADL–2; Holland,Frattali,&Fromm, 1999). ESs ranged
from –1.15 to 3.78,with five favoringmore intensive treat-ment
(four large effects) and four favoring less intensive
treatment (two large effects).
Basso and Caporali (2001) also contributed data to
this clinical question, as they described greater changes
in participants’ conversational abilities in all persons re-ceiving
more intensive treatment as compared with the
counterpart receiving less intensive treatment. Overall,
findings for outcome measures of communication activity/
participation were mixed, with some favoring less in-tensive
treatment and others favoring more intensive
treatment.
Clinical Question 3: For stroke-induced acute aphasia,
what is the influence of treatment intensity on measures of
language impairment? In one study of 17 individuals,
(Denes et al., 1996), the ESs for the Aachen Aphasia Test
(AAT; Huber, Poeck, & Williams, 1984) ranged from 0.39
(repetition tasks) to 1.20 (written language tasks), all fa-voring
more intensive treatment.
what treatment outcomes are maintained following
intensive language treatment? In the single-participant
design of Raymer et al. (2006), ESs for picture-naming

Table 5. Intervention variables.
Study Treatment program Treatment schedule Total amount Duration
Impairment
measure
Activity/participation
measure
Basso & Caporali,
2001 (Intensity)
All participants—individual treatment; conventional
aphasia treatment; varied techniques.
More intensive
treatment
NR All participants TT Picture description
(no analysis)
Goal areas = auditory comprehension, repetition,
naming, and conversation.
3 participants 5 hr/wk;
Home program
14–40 months Raven’s
Matrices
Treatment program—more intensive = clinic + home
program; less intensive = clinic.
2–3 hr/day
Less intensive
treatment
3 participants;
5 hr/wk
Denes et al., 1996
(Intensity)
All participants—individual treatment; ecological,
based on stimulation approach to restore efficient use
of language mainly in conversation.
All participants Intensive treatment
group
All participants AAT profile None
Goal areas = auditory comprehension and verbal
production.
45- to 60-min sessions
M = 130 sessions,
SD = 36,
range = 94–160
M = 6 months,
range = 5.2–7.0
TT
Technique = picture identification and conversation context.
Intensive treatment
group
Regular treatment
group
Repetition
6–7 sessions/wk
M = 60 sessions,
SD = 14,
range = 56–70
Written Language
Regular treatment
group
Naming
About 3 sessions/wk
Comprehension
Hinckley & Craig,
1998, Study 3
(Intensity)
All participants—individual, group and computer treatment;
functionalist/pragmatic approach. Standard functional
treatments (i.e., PACE and cueing hierarchy).
Intensive treatment
phase
All participants All participants BNT CU analysis from
picture description
23 hr/wk
294–306 hrs 18 wks
Nonintensive
treatment phase
6 wks intensive
phase
3–5 hr/wk
6 wks nonintensive
phase
6 wks intensive
phase
Hinckley & Carr,
2005 (Intensity)
All participants—context-based treatment that is based
on whole-task training and ecological validity.
Intensive treatment
group
Total number of
sessions per
participant-NR
NR PALPA CADL
Tasks = problem-solving catalogue ordering task
to develop compensatory strategies.
20 hrs individual
treatment/wk; 5 hr
group treatment/wk
All participants
Oral Naming Catalogue ordering
task
Techniques = role play, strategy development,
and context-specific cues. Nonintensive
treatment group
Minutes to reach
criterion on
catalogue ordering task
Written Naming
Treatment group—intensive treatment = individual + group;
nonintensive treatment = individual. 4 hr individual
treatment/wk
M = 233,
range = 29–597
(1–10 sessions)
(Continued on the following page)

Table 5 Continued. Intervention variables.
Impairment
measure
measure
Maher et al.,
2006 (CILT)
All participants—task = primarily dual card task. All participants All participants All participants WAB Story retelling
Clinician ratings
of post-treatment
narrative
Technique = shaping and cueing. 3-hr sessions 24 hr 2 wks BNT
Treatment group—CILT = verbal response required. 4 sessions/wk ANT
PACE = all communication modalities permitted.
Meinzer et al.,
2004 (CILT)
Treatment group—CIAT Verbal response required. Dual
card task (object drawings) 2–3 participants per group.
All participants All participants All participants AAT None
Technique = shaping and cueing. Model-based aphasia
treatment; deficit-specific approach.
3 hr/day for 10 days 30 hr 2 wks TT
NR = individual or group treatment.
Meinzer et al.,
2005 (CILT)
Treatment group—CIAT = verbal response required; dual
card task (object drawings); 2–3 participants per group.
Technique = shaping. All participants All participants All participants AAT CAL
CIAT plus = CIAT as above, plus dual card task (object 3 hr/day for 10 days 30 hr 2 wks CETI
drawings, written words, and photographs); daily
home program; daily communication practice with
family members.
2001 (CILT and
Intensity)
Treatment group CIAT group CIAT group CIAT group TT CAL
CIAT = verbal response required; dual card task
3 hr/day for 10 days M = 31.5 hr,
2 weeks AAT
(object drawings); 2–3 participants per group.
range = 23–33
Technique = Shaping and reinforcement.
Conventional aphasia treatment—syndrome-specific
standard approach; naming, repetition,
sentence completion, and conversation tasks.
Conventional
treatment group Conventional
treatment group
Conventional
treatment group
Session length NR 3–5 wks
M = 33.9 hr,
range = 20–54
3–5 wks
(Continued on the following page)

Table 5 Continued. Intervention variables.
Impairment
measure
measure
2005 (CILT)
All participants—CIAT; verbal response required; dual card
task (object drawings); 3 participants per group.
All participants All participants All participants Lexical decision
Reaction time
None
Technique = shaping and reinforcement.
3 hr/day for 10 days M = 31.3 hr, SD = 3.5,
range = 23–33
2 wks
AAT
TT
Repetition
Naming
Comprehension
Raymer et al.,
2006 (Intensity)
All participants—Moss-talk multimode matching exercises;
spoken and written word/picture matching.
All participants All participants NR WAB None
Alternating treatment
schedules
3–5 baseline sessions BNT
3–4 sessions/wk
12 training sessions Picture naming
1–2 sessions/wk
1-month break Word/picture
12 training sessions verification
4 participants = 24 hr
1 participant = 22 hr
Note. NR = not reported; AAT = Aachen Aphasia Test (Huber, Poeck, & Weniger, 1984); ANT = Action Naming Test (Nicholas, Obler, Albert, & Goodglass, 1985); BNT = Boston Naming Test (Kaplan,
Goodglass, & Weintraub, 2001); CADL = Communication Activities of Daily Living (Holland, Frattali, & Fromm, 1999); CAL = Communication Activity Log (Pulvermuller et al., 2001); CETI = Communicative
Effectiveness Index (Lomas et al., 1989); CU = content unit; PACE = Promoting Aphasics’ Communicative Effectiveness (Davis & Wilcox, 1985); PALPA = Psycholinguistic Assessment of Language Processing in
Aphasia (Kay, Lesser, & Coltheart, 1992); TT = Token Test (De Renzi & Vignolo, 1962); WAB = Western Aphasia Battery (Kertesz, 1982).

maintenance were 7.45 in the more intensive condi-tion
and 4.85 in the less intensive condition. ESs for a
word/picture verification task were 1.75 in the more
intensive condition and 2.14 in the less intensive con-dition.
Thus, maintenance effects of treatment inten-sity
were mixed in this study.
Stage of research and quality. Table 7 also shows the
number of quality markers for each study and the stage
of research. Of the two studies in the discovery phase of
research, one received a quality marker score of 5 out of
8 (Raymer et al., 2006) and the other received a quality
marker score of 3 out of 8 (Basso&Caporali, 2001). Three
studieswere efficacy studies (Denes et al., 1996;Hinckley
& Carr, 2005; Pulvermuller et al., 2001), each of which
attained a quality marker score of 6 or 7 out of 9. One
study (Hinckley & Craig, 1998) fell into the effectiveness
stage of research, achieving a quality score of 4 out of 8.
CILT Results
CILT was examined in five studies as shown in
Table 8. These studies addressed Questions 6, 7, and 10.
No studies of CILT in acute aphasia were available, so
Questions 8 and 9 could not be addressed. Cohen’s d val-ues
could be calculated for all five CILTstudies. The ESs
for Meinzer et al. (2004, 2005), and Pulvermuller et al.
(2005) represent within-subject effects, whereas those in
Maher et al. (2006) and Pulvermuller et al. (2001) come
frombetween-group comparisons of CILTwith a contrast-ing
treatment.
what is the influence of constraint-induced language
therapy on measures of language impairment? Across the
five group studies, 16 ESs were calculable, including 9
large ESs in favor of CILT. In individuals with chronic
aphasia, CILT had positive effects for overall aphasia bat-tery
scores (Maher et al., 2006; Meinzer et al., 2004) and
subtests of auditory comprehension, word retrieval, repe-tition,
and lexical decision (Maher et al., 2006; Meinzer
et al., 2004; Pulvermuller et al., 2005).
Meinzer et al. (2005) examined a modification of
CILT (described as CIAT) in which additional home ac-tivities
were included. There was no difference between
CILT and CIAT plus on outcome scores immediately
Table 6. Study quality marker variables.
Study Design
Assessor
blinding
Random
sampling
described
Participants
comparable/
described
Treatment
fidelity reported
Valid outcome
measure Significance
Intent-to-
treat Precision
Basso & Caporali,
2001 (Intensity)
Case study No No Yes No Yes Yes N/A No
Denes et al., 1996
(Intensity)
Controlled
trial
Yes No Yes No Yes Yes Yes Yes
Hinckley & Craig,
1998, Study 3
(Intensity)
Case series No No Yes No Yes Yes N/A Yes
Hinckley & Carr,
2005 (Intensity)
Controlled
trial
No No Yes Yes Yes Yes Yes Yes
Maher et al.,
2006 (CILT)
Controlled
trial
No No Yes Yes Yes Yes No Yes
Meinzer et al.,
2004 (CILT)
Meinzer et al.,
2005 (CILT)
Controlled
trial
No No Yes No Yes Yes Yes No
2001 (CILT
and Intensity)
Controlled
trial
Yes Yes Yes No Yes Yes No No
2005 (CILT)
Raymer et al.,
2006 (Intensity)
Single subject No No Yes No Yes Yes N/A Yes
Note. N/A = not applicable.

following training. Both groups reported significant im-provements.
For language impairment measures, CILT
consistently led to positive outcomes.
Clinical Question 7: For stroke-induced chronic
aphasia, what is the influence of constraint-induced lan-guage
therapy on measures of communication activity/
participation? Three of the 5 CILT studies included mea-sures
of communication activity/participation; ESs could
be calculated for 11 outcome measures incorporated in the
studies (e.g.,Communicative Activity Log, story retelling,
Communicative Effectiveness Index; Lomas et al., 1989).
ESs ranged from –.82 to 3.77, with eight favoringCILT (six
large effects) and three favoring a comparison treatment
(one large effect). Thus, CILT had positive outcomes for
severalmeasures of communication activity/participation,
although some favored a comparison condition.
ClinicalQuestion 10: For stroke-induced chronic apha-sia,
what treatment outcomes are maintained following
Table 7. Intensity studies, effect sizes, and methodologic quality.
Study Outcome measure(s) ICF Intensity effect size
Research
stage
Quality
score
Clinical
question
Basso & Caporali, 2001 AAT TT I Not calculable Discovery 3 of 8 1, 2
Raven’s matricesa I
Picture description I
Denes et al., 1996 AAT: TT I 0.60 Efficacy 7 of 9 3
Repetition I 0.39
Written Language I 1.20
Naming I 0.73
Comprehension I 0.91
Profile Level I 0.83
Hinckley & Craig, 1998, Study 3 Intensive Treatment I vs. Nonintensive Effectiveness 4 of 8 1, 2
BNT I 1.12
Content Unit Analysis A/P 0.53
Intensive Treatment II vs. Nonintensive
BNT I 0.95
Content Unit Analysis A/P 0.74
Hinckley & Carr, 2005 Catalogue ordering Efficacy 7 of 9 1,2
Oral (quiet) A/P –0.81
Oral (concurrent) A/P –0.05
Written (quiet) A/P –0.54
Written (concurrent) A/P 0.18
CADL-2 A/P –1.15
PALPA Oral Naming I 0.16
PALPA Written Naming I 1.48
Pulvermuller et al., 2001 AAT: Overall Test Scores I 2.18 Efficacy 6 of 9 1, 2
TT I 0.92
Naming I 1.12
Language Comprehension I 1.12
Repetition I Not calculable
Communicative Activity Log: Patients A/P 3.78
SLPs A/P 2.64
Raymer et al., 2006 WAB Aphasia Quotient I Not calculable Discovery 5 of 8 1, 5
BNT I Not calculable
Picture Naming Acquisition I 4.35 (lw), 11.37 (hw)
Picture Naming Maintenance I 4.85 (lw), 7.45 (hw)
Word/Picture Verification Acquisition I 2.72 (lw), 2.14 (hw)
Word/Picture Verification Maintenance I 2.14 (lw), 1.75 (hw)
Note. lw = low weighted intensity; hw = high weighted intensity; ICF = International Classification System; I = impairment; A/P = activity/participation;
SLP = speech-language pathologist.
aRaven’s Coloured Progressive Matrices (Raven, Court, & Raven, 1986)

constraint-induced language therapy? Two studies ad-dressed
Question 10. Meinzer et al. (2005) reported that
effects of CILT on measures of language impairment
and communication activity/participation were main-tained
at 6months compared with baseline performance.
Family members of the group receiving CIATplus with
the modified home program reported further gains in
communication effectiveness over the 6-month period.
Maher et al. (2006) reported follow-up testing at 1 month
post-CILT completion. Three of 4 CILT participants and
1 of 3 participants in Promoting Aphasics’ Communica-tive
Effectiveness (PACE) demonstrated continued in-creases
on aphasia tests over the 1-month period.
Stage of research and quality. Table 8 indicates that
two studies that fell into the discovery phase of re-search
received quality scores of 4 out of 8. Three studies
represented efficacy research, two of which received a
qualitymarker score of 6 out of 9 and one ofwhich received
a score of 5 out of 9.
Discussion
The purpose of this EBSR was to assess the in-fluence
of intensity of language treatment and CILT on
language and communication outcomes of individuals
with stroke-induced aphasia. A systematic search of the
literature from 1990 to 2006 yielded only 10 studies that
met predetermined inclusion criteria, with 5 studies ad-dressing
treatment intensity, 4 studies addressing CILT,
and 1 study addressing both. Although few studies were
expected to be found for CILT—a technique that was
Table 8. CILT studies, effect sizes, and methodologic quality.
Study Outcome measure(s) ICF CILT effect size Research stage Quality score Clinical question
Maher et al., 2006 Aphasia tests Efficacy 6 of 9 6, 7, 10
WAB Aphasia Quotient I 1.01
BNT I –0.16
Action Naming Test I 0.14
Linguistic measures in story retelling
Number of words A/P –0.72
Number of utterances A/P –0.82
Number of sentences A/P –0.19
Mean length of utterance A/P 0.33
Clinician ratings A/P Not calculable
Meinzer et al., 2004 AAT: Profile I 0.34 Discovery 4 of 8 6
Token Test I 0.81
Meinzer et al., 2005 AAT: Profile I 1.63 Efficacy 5 of 9 6, 7, 10
Subtests I Not calculable
CETI: Overall – relatives A/P 1.86
Communication Activity Log
Quantity: Patients A/P 1.99
Relatives A/P 2.35
Comprehension: Patients A/P .47
Relatives A/P 1.13
Pulvermuller et al., 2001 AAT: Overall Test Scores I 2.18 Efficacy 6 of 9 6, 7
TT I 0.92
Naming I 1.12
Language Comprehension I 1.12
Repetition I Not calculable
Communicative Activity Log: Patients A/P 3.77
SLPs A/P 2.64
Pulvermuller et al., 2005 AAT: TT I 0.25 Discovery 4 of 8 6
Repetition I 0.11
Naming I 0.25
Comprehension I 0.46
Lexical decision reaction times
Response time I 2.39
Pseudowords x time I 3.32
Words I Not calculable

introduced only in 2001—the paucity of studies directly
addressing treatment intensity was surprising in view of
the overall large number of studies examining language
treatment for stroke-induced aphasia. Given the small
number of studies, the conclusions of the present sys-tematic
reviewmust be considered preliminary. Further-more,
interpretation of ESs and their applicability to the
aphasia literature is not certain at this time. Neverthe-less,
certain trends in the literature from 1990 to 2006
may usefully inform both clinical practice and future re-search
in this area.
Increased treatment intensity was associated with
positive changes in outcome measures of language im-pairment
in 68 persons with chronic and acute aphasia.
All ESs calculated for group comparisons on language
impairment measures favored more intensive treatment
over less intensive treatment, including seven large ef-fect
sizes. In the single-participant study (Raymer et al.,
2006), ESs for one measure favored more intensive treat-ment
and one favored less intensive. The data for Clin-ical
Question 2 addressing change in communication
activity/participation were mixed, with some favoring
more intensive treatment and some favoring less inten-sive
treatment for persons with chronic aphasia. These
observations suggest that there can be complex interac-tions
among intensity of treatment schedule, type of
treatment, and type of outcome measure.
Maintenance of treatment intensity effects were
reported in one study of persons with chronic aphasia
(Raymer et al., 2006). Findings also were equivocal, fa-voring
more intensive treatment for one outcome mea-sure
and less intensive treatment for the other. However,
there was only one data point for each maintenance ob-servation.
Beeson and Robey (2006) suggest a minimum
of two data points for calculation of an ES. Therefore, this
result should be taken with caution.
Overall, five studies involving 90 participants
reported that CILT resulted in positive changes on mea-sures
of language impairment and communication activity/
participation in individuals with chronic aphasia, in-cluding
large ESs for 9 of 16 impairment measures and
6 of 11 activity/participation measures. Meinzer et al.
(2005) reported that effects weremaintained for as long
as 6 months following completion of CILT in 27 patients.
No data addressed the effect of CILT in patients with
acute aphasia, however. Finally, it is important to note
that Maher et al. (2006) also found significant improve-ments
on language measures for participants in their
control group, who participated in an intensive version
of PACE in which use of compensatory modalities is al-lowed.
This finding implies that intensity is a key com-ponent
of CILT.
Interpretation of the results of this EBSR and its
implications for clinical practice must be considered in
conjunction with the information about characteristics
of the participants. For the intensity studies, the major-ity
of participants had nonfluent forms of aphasia. When
severity of aphasia was provided, the majority of partic-ipants
tended to have more severe aphasia, with many
described as having a global aphasia. Therefore, conclu-sions
regarding intensity of treatment are most applica-ble
to individuals with severe nonfluent aphasia. Future
research needs to address, in particular, the impact of
treatment intensity for participants presenting with a
milder aphasia as well as participants who are catego-rized
as fluent. Similarly for the CILT studies, the ma-jority
of participants were nonfluent and moderately
impaired, thereby limiting the generalizability of the re-sults
in individuals with fluent aphasia and individuals
with mild and severe aphasia. Chronicity of aphasia is
another participant characteristic that requires atten-tion
in future studies. No study of CILT included in-dividuals
with acute aphasia, and only one treatment
intensity study addressed the impact of intensity on
acute aphasia.
One of the difficulties encountered in this system-atic
review was that of comparing results across studies
when there were differences in the outcome measures
used. Therefore, we attempted to find commonalities
by categorizing the types of outcome measures accord-ing
to the WHO (2001) ICF levels of body function (lan-guage
reception and production impairment) and activity/
participation. Overall, all of the studies included at least
one outcome measure at the language impairment level,
and most of these measures were standardized, valid as-sessment
instruments. Although all of the CILT studies
and four of the six intensity studies included at least
onemeasure at the communication activity/participation
level, there were fewer of these measures, they were typ-ically
individualized, and information on their validity
and reliability was often lacking. Future studies of CILT
and clinical practice should consider inclusion of more
measures at the communication activity/participation
level as well as studies that address quality of life in
patients with aphasia.
The variety of primary and secondary outcome mea-suresmade
comparisons across studies difficult and com-plicated
the determination of the validity and reliability
as they contribute to the quality of the study. Therefore,
our critical appraisal used a quality marker related to
outcome measures that represented only a minimal level
of quality in these areas. All studies achieved this min-imal
level of quality that related to use of a valid and
reliable primary outcome measure.
A number of methodological weaknesses in the 10
studies included in the EBSR were identified during the
critical appraisal of the quality of each study. The short-comings
across the studies related to primarily three

quality indicators: random assignment of subjects, as-sessor
blinding, and evidence of treatment fidelity. In
most of the studies, regardless of stage of research and
research design, participants were selected as a conve-nience
sample. Randomization is critical for future re-search,
as many evidence-based review organizations
(e.g., The Cochrane Collaboration) include primarily ran-domized
trials in their systematic reviews of health care
interventions. It is also essential that investigators doc-ument
procedures to ensure fidelity of the actual treat-ment
techniques and that some measure of procedural
reliability is provided. Clear documentation of the treat-ment
technique will facilitate the transference of the tech-nique
to the clinical setting.
Examination of the studies in relation to stage of re-search
clearly shows where the strengths of the current
research lie and the direction of future research. In-cluded
studies in the present EBSR were three efficacy
studies evaluating the impact of intensity on stroke-related
chronic aphasia and three efficacy studies eval-uating
the impact of CILT on chronic aphasia. Only one
intensity study evaluated effectiveness, and no study
evaluated effectiveness of CILT. No studies in either cat-egory
assessed policy change or cost effectiveness. There-fore,
as evidence for efficacy of more intensive treatment
and CILT continues to grow, future research must ad-dress
issues of effectiveness and cost effectiveness with
studies that are designed according to the criteria in-cluded
in the ASHA levels-of-evidence scheme to ensure
that the studies are of high quality.
Regardless of type of treatment, our general con-clusions
indicate that more treatment is better over a re-stricted
time interval. No studies have addressed what
might be the optimum amount and intensity of treat-ment;
however, it is possible that no single answer exists
and that optimum dosage of high-intensity treatment
varies depending on the type of treatment and the char-acteristics
of the participant’s aphasia, including chro-nicity.
Future studies, beginning at the discovery phase,
are needed to address the many confounds inherent in
resolving the issue of optimum intensity of treatment for
individuals with stroke-induced aphasia.
Although this systematic review has highlighted the
potential efficacy of CILT on the basis of its dual prin-ciples,
future studies need to be designed with the aim of
teasing out the impact of constraint and intensity on out-come.
This task is particularly difficult because high in-tensity
is such an integral part of CILT. However Maher
et al. (2006) have demonstrated that it is possible to ad-dress
this issue methodologically by providing an iden-tical
schedule of CILTand its control treatment. Although
results suggest that efficacy is related to the nature of the
treatment (i.e., constraint) rather than to intensity alone,
given the small number of participants, results are not
yet conclusive. More definitive evidence will be derived
only from (a) conducting randomized studies comparing
CILT with other treatment procedures given at a simi-lar
intensity schedule using participants of similar type
and severity of aphasia and (b) measuring outcomes at
the levels of language impairment and communication
activity/participation. Additionally, the optimum inten-sity
of treatment for CILT has yet to be determined. Fur-ther
studies comparing constraints at different levels of
intensity are needed.
Although the early evidence surrounding the efficacy
of CILT for individuals with aphasia shows somepromise,
the effects are similar to other intensive treatments that
do not employ the use of constraint. Therefore, these re-sults
should be taken under advisement. Clear and con-vincing
evidence supporting the effectiveness of CILT
over other aphasia treatments has not yet been estab-lished,
and further research comparing aphasia treat-ments
is warranted. Clinicians should interpret the
findings from this EBSR conservatively and always in
conjunction with their patients’ unique characteristics,
circumstances, and preferences when considering the
best course of treatment for and with their clients.
Finally, some observations need to be made following
this pilot implementation of ASHA’s levels-of-evidence
scheme. The indicators used to judge the quality of re-search
studies tend to focus on issues that are most per-tinent
to group study designs. Therefore, the usefulness
of the levels-of-evidence scheme in evaluating single-participant
research designs, which are used commonly
in the speech-language pathology literature, is in need of
further development. Although eight of the nine indica-tors
were applicable to single-participant design research,
other quality indicators related to the documentation of
experimental control and threats to external validity (e.g.,
stability of baselinemeasures) that are unique to this type
of research were not included. These additional issues of
design quality that arise in single-participant research
will need to be studied.
Another notable aspect of the current systematic re-view
is that all studies addressing the clinical questions,
including discovery studies, are included. Although
some systematic reviews include only group designs or
randomized clinical trials, we elected to include all stud-ies,
regardless of design. Ultimately, we need to give
greater weight to well-designed randomized trials. How-ever,
given the relative paucity of those studies in our
current literature base, the inclusion of all possible stud-ies
gives a broader perspective of the state of knowledge.
It is our hope that this systematic review will incite ad-ditional
rigorous clinical trials addressing the questions
of intensity of aphasia treatment and CILT, as well as
other areas of interest to the fields of speech-language
pathology and audiology.

Acknowledgments
This EBSR was supported by ASHA’s National Center
for Evidence-Based Practice in Communication Disorders
(N-CEP), the Advisory Committee on Evidence-Based Practice
in Communication Disorders, and ASHA’s Special Interest
Division 2: Neurophysiology and Neurogenic Speech and
Language Disorders. We thank the following individuals who
contributed to the preparation of this article: Rob Mullen,
N-CEP Director; BeverlyWang, N-CEP Information Manager;
and Floyd Roye, N-CEP Project Administrator. All members
of this evidence-based review panel agreed to declare no
competing interests in relation to this article. No author had
any paid consultancy or any other conflict of interest with this
document. The first three authors contributed equally to the
preparation of this article, and the order does not reflect any
differences in contribution.
References
ASHA. (2001). Scope of practice in speech-language pathology
[Scope of Practice]. Available from www.asha.org /policy.
Basso, A. A., & Caporali, A. (2001). Aphasia therapy or the
importance of being earnest. Aphasiology, 15, 307–332.
Beeson, P., & Robey, R. (2006). Evaluating single-subject
treatment research: Lessons learned from the aphasia
literature. Neuropsychological Review, 16, 161–169.
Bhogal, S.K., Teasell, R.,&Speechley, M. (2003). Intensity
of aphasia therapy, impact on recovery. Stroke, 34, 987–993.
Busk, P. L., & Serlin, R. (1992). Meta-analysis for single case
research. In T. R. Kratochwill & J. R. Levin (Eds.), Single-case
research design and analysis: New directions for psy-chology
and education. Hillsdale, NJ: Erlbaum.
Cohen, J. (1988). Statistical power analysis for the behavioral
sciences (2nd ed.). Hillsdale, NJ: Erlbaum.
Davis, G. A., & Wilcox, M. (1985). Adult aphasia rehabilita-tion:
Applied pragmatics. San Diego, CA: College Hill Press.
Denes, G., Perazzolo, C., Piani, A., & Piccione, F. (1996).
Intensive versus regular speech therapy in global aphasia:
A controlled study. Aphasiology, 10, 385–394.
De Renzi, E., & Vignolo, L. (1962). The Token Test: A
sensitive test to detect receptive disturbances in aphasics.
Brain, 85, 665–678.
Goodglass, H., & Kaplan, E. (1983). The Boston Diagnostic
Aphasia Examination. Philadelphia: Lea & Febiger.
Hakkennes, S., & Keating, J. (2005). Constraint-induced
movement therapy following stroke: A systematic review of
randomized controlled trials. Australian Journal of Physio-therapy,
51, 221–231.
Hedges, L.V.,&Olkin, I. (1985). Statistical methods for meta-analysis.
Orlando, FL: Academic Press.
Hinckley, J. J., & Craig, H. K. (1998). Influence of rate of
treatment on the naming abilities of adults with chronic
aphasia. Aphasiology, 12, 989–1006.
Hinckley, J. J.,& Carr, T. (2005). Comparing the outcomes of
intensive and non-intensive context-based aphasia treat-ment.
Aphasiology, 19, 965–974.
Holland, A., Frattali, C., & Fromm, D. (1999). Communica-tion
activities of daily living (2nd ed.). Austin, TX: Pro-Ed.
Hollingshead, A. B. (1975). Four Factor Index of Social
Status. New Haven, CT: Yale University.
Huber, H. P., Poeck, K., & Williams, K. (1984). The Aachen
Aphasia Test. In F. C. Rose (Ed.), Progress in aphasiology
(pp. 291–303). New York: Raven Press.
Huber, W., Poeck, K., & Weniger, D. (1984). The Aachen
AphasiaTest. InF. C. Rose (Ed.), Advances in neurology,Vol. 42:
Progress in aphasiology. New York: Raven Press.
Kaplan, E., Goodglass, H., & Weintraub, S. (2001). The
Boston Naming Test. Philadelphia: Lea & Febiger.
Kay, J., Lesser, R., & Coltheart, M. (1992). Psycholinguistic
assessments of language processing in aphasia. Hove, United
Kingdom: Psychology Press.
Kertesz, A. (1982). Western Aphasia Battery. Orlando, FL:
Grune & Stratton.
Kleim, J. A., & Jones, T. A. (2008). Principles of experience-dependent
neural plasticity: Implications for rehabilitation
after brain damage. Journal of Speech-Language-Hearing
Research, 51(Suppl.), S225–S239.
Lomas, J., Pickard, L., Bester, S., Elbard, H., Finlayson,
A., & Zoghaib, C. (1989). The Communicative Effective-ness
Index: Development and psychometric evaluation of
a functional communication measure for adult aphasia.
Journal of Speech Hearing Disorders, 54, 113–124.
Maher, L., Kendall, D., Swearengin, J., Rodriguez, A.,
Leon, S., Pingel, K., Holland, A., & Rothi, L. (2006).
A pilot study of use-dependent learning in the context of
constraint induced language therapy. Journal of the Inter-national
Neuropsychological Society, 12, 843–852.
Meinzer, M., Djundja, D., Barthel, G., Elbert, T., &
Rockstroth, B. (2005). Long-term stability of improved
language functions in chronic aphasia after constraint-induced
aphasia therapy. Stroke, 36, 1462–1466.
Meinzer, M., Elbert, C., Wienbruch, C., Djundja, D.,
Barthel, G., & Rockstroh, B. (2004). Intensive language
training enhances brain plasticity in chronic aphasia. BMC
Biology, 2, 1–9.
Nicholas, M., Obler, L., Albert, M., & Goodglass, H. (1985).
Lexical retrieval in healthy aging. Cortex, 21, 595–606.
Poek, K., Huber, W., & Willmes, K. (1989). Outcomes of
intensive language treatment in aphasia. Journal of Speech
and Hearing Disorders, 54, 471–479.
Pulvermuller, F. B., Neininger, B., Elbert, T., Mohr, B.,
Rockstroh, B., Koebbel, P.,&Taub, E. (2001). Constraint-induced
therapy of chronic aphasia after stroke. Stroke, 32,
1621–1626.
Pulvermuller, F., Hauk, O., Zohsel, K., Neininger, B., &
Mohr, B. (2005). Therapy-related reorganization of lan-guage
in both hemispheres of patients with chronic aphasia.
Neuroimage, 28, 481–489.
Raymer, A., Kohen, F., & Saffell, D. (2006). Computerised
training for impairments for word comprehension and
retrieval in aphasia. Aphasiology, 20, 257–268.
Raven, J. C., Court, J. H., & Raven, J. (1986). Raven’s
Coloured Progressive Matrices. London: H. K. Lewis & Co.
Robey, R. R. (1998). A meta-analysis of clinical outcomes in
the treatment of aphasia. Journal of Speech, Language, and
Hearing Research, 41, 172–187.
Robey, R. R. (2004). A five-phase model for clinical-outcomes
research. Journal of Communication Disorders, 5, 401–411.

Taub, E. (1977). Movement in nonhuman primates deprived
of somatosensory feedback. Exercise and Sports Sciences
Review, 4, 335–374.
Taub, E., Miller, N., Novack, T., & Cook, E. (1993). Tech-nique
to improve chronic motor deficit after stroke. Archives
of Physical Medicine and Rehabilitation, 74, 347–354.
Taub, E., Uswatte, G., King, D. K., Morris, D. M., Crago,
J. E., & Chatterjee, A. (2006). A placebo controlled trial of
constraint-induced movement therapy for upper extremity
after stroke. Stroke, 37, 1045–1049.
Taub, E., Uswatte, G., & Piclikiti, R. (1999). Constraint-induced
movement therapy: A new family of techniques with
broad application to physical rehabilitation—A clinical
review. Journal of Rehabilitation Research and Develop-ment,
36, 237–251.
Taub, E., & Wolf, S. (1997). Constraint-induced movement
techniques to facilitate upper extremity use in stroke
patients. Topics in Stroke Rehabilitation, 3, 38–61.
Wolf, S., Winstein, C. J., Miller, J., Taub, E., Uswatte, G.,
Morris, D., et al. (2006). Effect of constraint-induced move-ment
therapy on upper extremity function 3 to 9 months
after stroke: The EXCITE randomized clinical trial. Jour-nal
of the American Medical Association, 296, 2095–3104.
World Health Organization. (2001). International Classifi-cation
of Functioning, Disability and Health: ICF. Geneva,
Switzerland: Author.
Received September 4, 2007
Accepted February 24, 2008
DOI: 10.1044/1092-4388(2008/07-0206)
Contact author: Tobi Frymark, National Center for Evidence-
Based Practice in Communication Disorders, American
Speech-Language-Hearing Association, 2200 Research
Boulevard, Rockville, MD 20850. E-mail: tfrymark@asha.org.

Constraint induced language therapy

Constraint induced language therapy

More Related Content

What's hot (20)

Similar to Constraint induced language therapy (20)

Recently uploaded (20)

Constraint induced language therapy