Intensive Versus Distributed Aphasia Therapy
A Nonrandomized, Parallel-Group, Dosage-Controlled Study
Background and Purpose—Most studies comparing different levels of aphasia treatment intensity have not controlled the dosage of therapy provided. Consequently, the true effect of treatment intensity in aphasia rehabilitation remains unknown. Aphasia Language Impairment and Functioning Therapy is an intensive, comprehensive aphasia program. We investigated the efficacy of a dosage-controlled trial of Aphasia Language Impairment and Functioning Therapy, when delivered in an intensive versus distributed therapy schedule, on communication outcomes in participants with chronic aphasia.
Methods—Thirty-four adults with chronic, poststroke aphasia were recruited to participate in an intensive (n=16; 16 hours per week; 3 weeks) versus distributed (n=18; 6 hours per week; 8 weeks) therapy program. Treatment included 48 hours of impairment, functional, computer, and group-based aphasia therapy.
Results—Distributed therapy resulted in significantly greater improvements on the Boston Naming Test when compared with intensive therapy immediately post therapy (P=0.04) and at 1-month follow-up (P=0.002). We found comparable gains on measures of participants’ communicative effectiveness, communication confidence, and communication-related quality of life for the intensive and distributed treatment conditions at post-therapy and 1-month follow-up.
Conclusions—Aphasia Language Impairment and Functioning Therapy resulted in superior clinical outcomes on measures of language impairment when delivered in a distributed versus intensive schedule. The therapy progam had a positive effect on participants’ functional communication and communication-related quality of life, regardless of treatment intensity. These findings contribute to our understanding of the effect of treatment intensity in aphasia rehabilitation and have important clinical implications for service delivery models.
Treatment intensity is an important consideration for the delivery of effective and efficient aphasia rehabilitation services. However, the optimal treatment intensity for aphasia rehabilitation is unknown.1 Neuroscience research suggests that intensive training is required to optimize neurological and functional recovery post stroke.2 Within the field of cognitive psychology, however, there is considerable evidence supporting the distributed practice effect, which suggests that optimal long-term learning in healthy humans is achieved with nonintensive or distributed training.3
Many meta-analyses and systematic reviews have been conducted in an attempt to synthesize the evidence for treatment intensity in aphasia rehabilitation.4–7 Bhogal et al8 found that studies that provided an increased dosage of therapy over a shorter duration (8.8 hours per week for 11.2 weeks) achieved positive therapeutic outcomes, whereas studies that provided a reduced dosage of therapy over a longer duration (2 hours per week for 22.9 weeks) did not. Consequently, this review provides support for the benefits of intensive therapy. However, a key limitation of this review is that the total dosage of aphasia therapy provided was not controlled. As such, it is difficult to differentiate between a dose–response relationship, whereby greater amounts of therapy result in greater therapeutic gains and the true effect of treatment intensity.
Additional reviews have reported more equivocal results for the benefits of intensive treatment. Cherney et al5 found no significant difference between the outcomes for intensive versus nonintensive treatment on impairment-based measures in the acute (time post onset, <3 months) or chronic stage (time post onset, ≥3 months) of stroke recovery. There was, however, a strong positive relationship between intensity and activity/participation-based therapy outcomes in the chronic stage. Furthermore, a Cochrane review conducted by Brady et al7 found that intensive aphasia therapy significantly reduced aphasia severity when compared with nonintensive aphasia therapy; however, significantly more individuals withdrew from intensive programs. As such, the evidence for intensive aphasia therapy remains mixed, and there is uncertainty as to whether intensive treatment is appropriate for all individuals with aphasia.
A small number of studies have systematically compared different levels of treatment intensity while controlling the dosage of aphasia therapy provided.9–12 Consistent with the distributed practice effect, Raymer et al11 and Sage et al9 provided evidence demonstrating that distributed aphasia therapy may be equally or more effective than intensive aphasia therapy when considering the long-term maintenance of treatment gains. However, the dosage-controlled studies conducted to date have provided a relatively limited dosage of impairment-based therapy,9,11,12 to a small sample of participants,9–12 using a repeated-measures, crossover design.9,11,12 In addition, Martins et al10 investigated the effect of treatment intensity in the subacute aphasia population, and as such, their findings may have been influenced by spontaneous recovery. Consequently, the comparative effect of intensive versus distributed aphasia therapy, when provided in a controlled dosage, warrants further investigation.
A preliminary study conducted at the University of Queensland evaluated the efficacy of the intensive, comprehensive aphasia program, Aphasia Language Impairment and Functioning Therapy (Aphasia LIFT), and found that the program positively affected participants’ language impairment and functional communication.13 This study aims to further investigate the effect of treatment intensity by comparing the efficacy of Aphasia LIFT when delivered in a dosage-controlled, intensive versus distributed treatment schedule on communication outcomes in adults with chronic aphasia.
This phase II study used a nonrandomized, parallel-group, pre–post-test design. Three intensive (LIFT) and 8 distributed (D-LIFT) trials of Aphasia LIFT were conducted at the University of Queensland and in rehabilitation centers in Brisbane and Sydney, Australia between November 2012 and August 2014. Trials consisted of groups of 2 to 6 participants, and the results of these trials were pooled for analysis. Participants were allocated to LIFT (n=16) and D-LIFT (n=18) based on their geographic location, the availability of a position within the research program, and personal factors (eg, participant availability, transport, and accommodation). This study was approved by the relevant institutional ethics committees, and written informed consent was obtained from participants before participation in study procedures.
Thirty-four adults with chronic aphasia resulting from unilateral, left hemisphere stroke(s) were recruited to participate in the study (Table 1; Table I in the online-only Data Supplement). All participants were >4 months time post onset, spoke fluent English before their stroke, and presented with residual word finding difficulties on the Boston Naming Test (BNT).14 Individuals with comorbid neurological conditions, severe apraxia of speech, or severe dysarthria were excluded from the study.
Assessment and Intervention
Participants completed a comprehensive speech and language assessment battery, administered by a qualified speech pathologist, before commencing therapy. Outcome measures were collected immediately post therapy and at 1-month follow-up. Where possible, assessments were administered by nontreating speech pathologists.
Treatment was based on the therapy principles outlined in Rodriguez et al.13 Participants each received 48 hours of aphasia therapy, comprised of 14 hours of impairment therapy, 14 hours of functional therapy, 14 hours of computer-based therapy, and 6 hours of group therapy. As anomia is a predominant feature of aphasia, impairment therapy primarily aimed to remediate word-retrieval deficits using a combined semantic feature analysis and phonological component analysis approach.15–17 Computer therapy also targeted word-retrieval impairments and included training with the software programs StepbyStep18 and Aphasia Scripts.19 Functional therapy was tailored to individuals’ communication goals and included a range of treatment approaches, for example, script training20 and communication partner training.21 Group therapy was based on the Aphasia Action Success Knowledge program (Grohn, Brown, Finch, Worrall, Simmons-Mackie, Thomas, unpublished data, 2012) and included education on stroke and aphasia, compensatory strategies for effective communication, and avenues to access further support.
A comprehensive Aphasia LIFT manual was developed to promote treatment fidelity. Therapy was provided by qualified speech pathologists who received training on the treatment approaches used in Aphasia LIFT. In some instances, computer therapy was facilitated by speech pathology students or a trained allied health assistant under the supervision of a qualified speech pathologist.
To evaluate the effect of treatment intensity, the total dosage of therapy, in number of therapy hours, remained constant and the frequency and duration of intervention varied between groups. Aphasia LIFT was delivered for 3 weeks (16 hours per week; total 48 hours), whereas D-LIFT was delivered for 8 weeks (6 hours per week; total 48 hours; Table 2). The cumulative treatment intensity for impairment therapy was measured according to the framework proposed by Warren et al22 (Table II in the online-only Data Supplement).
Outcome measures were selected based on the recommendations of the exploratory phase I/II study investigating the clinical efficacy of Aphasia LIFT.13 The BNT was administered as the primary outcome measure to assess participants’ word-retrieval abilities. Secondary outcome measures included a proxy-rated measure of participants’ functional communication (Communicative Effectiveness Index [CETI]23), and self-report measures of participants’ communication confidence (Communication Confidence Rating Scale for Aphasia)24 and communication-related quality of life (Assessment of Living with Aphasia25).
Two-tailed t tests and Fisher exact tests were used to compare the 2 cohorts, LIFT and D-LIFT, at baseline. To evaluate changes on the primary and secondary outcome measures, linear mixed models (LMM) were used. The use of LMM is preferable to general linear models (eg, regression, ANOVA, and ANCOVA) for modeling of longitudinal, repeated-measures data as it enables explicit modeling of correlation patterns and variance–covariance structures.26 To evaluate the effect of treatment by groups (LIFT and D-LIFT), separate models were fit for each outcome measure with time (pretherapy, post-therapy, and follow-up) and aphasia severity score from the Comprehensive Aphasia Test27 as fixed effects and participants as random effects. Furthermore, LMM were used to compare groups (LIFT and D-LIFT), covaried for aphasia severity score and pretherapy performance, on each outcome measure at post-therapy and follow-up. The BNT and CETI data were transformed before analysis using reflect and square root and square root transformations,26 respectively. Data approximated a normal distribution according to the Shapiro–Wilk test of normality28 (P>0.05).
The treatment groups were comparable for age, time post onset, sex, handedness, and measures of language impairment and functional communication at baseline (P>0.05). Thirty-two participants completed the Aphasia LIFT trial. Two D-LIFT participants (P29 and P31) withdrew from therapy because of acute-onset medical reasons and their data have been excluded from analyses. Another D-LIFT participant (P18) was not available for follow-up testing because of a change in personal circumstances. All 16 participants under the LIFT condition completed the therapy program. The mean therapy attendance rate was high (LIFT=47.7 hours; D-LIFT=47.9 hours), and the total dosage of therapy provided, in number of therapy hours, was comparable between groups (P=0.72). Furthermore, the cumulative treatment intensity for impairment-based therapy was comparable between groups (P=0.66; Table II in the online-only Data Supplement).
Primary Outcome Measures
Statistical analyses revealed a significant improvement in naming performance on the BNT at post-therapy compared with pretherapy for LIFT (F1,15=12.93; P=0.003) and D-LIFT (F1,15=29.92; P<0.001; Figure; Table III in the online-only Data Supplement). Likewise, there was a significant improvement in naming performance on the BNT at follow-up compared with pretherapy for LIFT (F1,15=6.50; P=0.02) and D-LIFT (F1,14.1=37.87; P<0.001). LMM, covaried for pretherapy BNT naming performance, revealed a significant difference between groups at post-therapy (F1,28=4.91; P=0.04) and follow-up (F1,27=11.85; P=0.002), with naming performance being significantly higher for D-LIFT compared with LIFT.
Secondary Outcome Measures
The CETI data are based on a sample of 28 participants (LIFT, n=15; D-LIFT, n=13), as 4 participants did not have a communication partner participate in the study. Participants’ functional communication, as measured by the CETI, was significantly higher for both groups at post-therapy, (LIFT: F1,9.9=31.57; P<0.001 and D-LIFT: F1,10.7=67.21; P<0.001) and follow-up (LIFT: F1,14=34.35; P<0.001 and D-LIFT: F1,12=71.97; P<0.001) compared with pretherapy. LMM, covaried for pretherapy CETI performance, revealed a trend favoring D-LIFT at post-therapy; however, this did not reach significance (P=0.05). There was no significant difference between groups on the CETI at follow-up (P=0.21).
Participants’ communication confidence, as measured by the Communication Confidence Rating Scale for Aphasia, was significantly higher for both groups at post-therapy (LIFT: F1,15=7.18; P=0.02 and D-LIFT: F1,15=16.56; P=0.001) and follow-up (LIFT: F1,15=6.08; P=0.03 and D-LIFT: F1,14.3=28.07; P<0.001) compared with pretherapy. There was no significant difference between groups on the Communication Confidence Rating Scale for Aphasia at post-therapy (P=0.79) or follow-up (P=0.48).
Finally, there was a significant improvement in participants’ communication-related quality of life, as measured by the Assessment of Living with Aphasia, for both groups at post-therapy (LIFT: F1,15=6.24; P=0.02 and D-LIFT: F1,15=10.81; P=0.005) and follow-up (LIFT: F1,15=9.64; P=0.007 and D-LIFT: F1,14.6=8.00; P=0.01). There was no significant difference between groups on the Assessment of Living with Aphasia at post-therapy (P=0.37) or follow-up (P=0.75).
This is the first dosage-controlled, parallel-group design study to compare the short and long-term therapeutic outcomes of an intensive versus distributed comprehensive aphasia program in participants with chronic aphasia. This study demonstrated that Aphasia LIFT had a positive and enduring effect on participants’ language impairment and functional communication. Interestingly, we found that aphasia therapy provided in a distributed schedule of 6 hours per week (8 weeks; total 48 hours) resulted in superior language gains on the primary outcome measure, the BNT, when compared with an intensive treatment regime of 16 hours per week (3 weeks; total 48 hours). This benefit of distributed training on word retrieval was maintained 1 month post therapy. Principles of experience-dependent neuroplasticity suggest that treatment intensity is a critical element driving neurological and functional recovery post stroke (ie, intensity matters).2 Our results indicate that when controlling the dosage of therapy provided, distributed therapy resulted in significantly greater impairment-based therapy gains compared with intensive therapy. The advantage for distributed training, with respect to the maintenance of treatment gains, is consistent with the results of Sage et al9 and provides support for the distributed practice effect. Sage et al9 used theories of learning and cognition to account for the benefit of distributed training on word retrieval in aphasia rehabilitation. As learning underpins the rehabilitation process, future consideration of these theories as they relate to the dosage, intensity, and duration of aphasia rehabilitation services is warranted.
With respect to measures of participants’ functional communication (CETI), communication confidence (Communication Confidence Rating Scale for Aphasia), and communication-related quality of life (Assessment of Living with Aphasia), we found comparable improvements for the intensive and distributed treatment conditions at post-therapy and 1-month follow-up. Although there was a trend favouring D-LIFT on the CETI at post-therapy, this did not reach significance. Importantly, these results indicate that both intensive and distributed treatment models had a positive and enduring effect on the real-life, functional communication of participants. Furthermore, a distributed therapy model did not reduce the efficacy of Aphasia LIFT, with respect to participants’ functional communication outcomes.
The results of previous dosage-controlled studies suggest that distributed therapy may result in equivalent or even superior long-term, clinical gains on impairment-based measures of word retrieval when compared with intensive therapy.9,11 However, because of design limitations and the use of small sample sizes, the generalizability of these results is limited. Our study sought to overcome previous methodological limitations by investigating the effect of treatment intensity using a parallel-group design in a larger sample of participants with chronic aphasia. Furthermore, in addition to controlling the total hours of therapy provided, our study measured the cumulative treatment intensity for impairment-based therapy, as per Warren et al,22 to ensure that the dosage of impairment therapy provided was consistent between groups. The findings of this phase II study build on the results of earlier research and provide increased support for the benefit of distributed training on impairment-based measures of word retrieval in adults with chronic aphasia. The next step in this systematic line of research is to conduct a large-scale randomized controlled trial to further investigate the effect of treatment intensity on short- and long-term therapy outcomes.
The outcomes of this study have significant implications for the scheduling and delivery of aphasia rehabilitation services. Highly intensive treatment protocols are an emerging service delivery model in aphasia rehabilitation29 and increasingly clinical guidelines in stroke management are advocating for the delivery of intensive services.30 However, our results provide support for a distributed model of aphasia therapy, with a significant advantage for distributed training on measures of language impairment and comparable gains on measures of functional communication and communication-related quality of life. In view of an ageing population and increasing demands for healthcare services, a distributed therapy model, such as that used in our study, presents an efficacious and potentially more feasible alternative model of care. Furthermore, for many individuals with aphasia, highly intensive treatment protocols may not be clinically appropriate, because of personal, medical, and logistical factors.32 Consistent with this argument, Brady et al7 found that significantly more individuals withdrew from intensive therapy than nonintensive therapy. We did not replicate this finding; however, it is acknowledged that our sample was comprised of individuals with chronic aphasia who volunteered to participate in Aphasia LIFT. As such, our sample may have been subject to selection bias. Further research into the clinical suitability and accessibility of intensive versus distributed service delivery models in aphasia rehabilitation is an important direction for future research.
Because of the complexity of behavioral interventions provided in intensive, comprehensive aphasia programs, it is difficult to determine which elements of therapy may respond to treatment intensity. Previous dosage-controlled research suggests that impairment-based therapy may be optimized with distributed training. However, it is possible that computer-based therapy or functional therapy targeting communication activity/participation may differentially respond to treatment intensity. Unfortunately, this research question cannot be resolved by investigating a comprehensive therapy program. Although this design may be viewed as a limitation of this study, it reflects the combination of therapy approaches that are used in clinical practice.
Outcome measurement for aphasia rehabilitation is complex, and therapy gains may be difficult to quantify with respect to the everyday relevance for individuals with aphasia. In view of the comprehensive nature of Aphasia LIFT, this study endeavored to measure outcomes across impairment and activity/participation domains. Although we found a significant advantage for D-LIFT on an impairment-based measure of word retrieval (BNT), it is important to note that both treatment conditions positively influenced the real-life effectiveness of participants’ communication at the activity/participation level, as measured by a validated assessment tool (CETI).
The definition of intensity in the aphasiology literature is ambiguous, ranging from 5 hours per week to >15 hours per week.5 This study aimed to compare the effect of 2 different levels of treatment intensity, provided at the same total dosage. Although the distributed schedule used (6 hours per week, including 2 hours of direct impairment therapy) is still less intensive than the 8.8 hours per week deemed necessary by Bhogal et al8 to achieve therapeutic gains, future research could evaluate the effect of an even less-intensive treatment model, such as 2 hours per week, which more closely approximates usual care.
A distributed model of Aphasia LIFT resulted in the superior acquisition and maintenance of language-impairment therapy gains on the primary outcome measure, the BNT, when compared with intensive therapy. Aphasia LIFT had a positive effect on participants’ functional communication, communication confidence, and communication-related quality of life, regardless of treatment intensity. This research contributes to our understanding of the effect of treatment intensity, independent of therapy dosage, on aphasia rehabilitation outcomes. Treatment intensity is integral to the provision of effective and efficient aphasia rehabilitation services. Furthermore, establishing optimal treatment intensity is an important research question, with implications extending beyond aphasia management to the multidisciplinary rehabilitation of stroke. Consequently, the outcomes of this research provide an important contribution to the field and have significant implications for clinicians, consumers, and service providers involved in stroke and aphasia rehabilitation.
The Communication Research Registry is acknowledged as a source of participant recruitment. We acknowledge the support provided by the speech pathologists at Prince of Wales Hospital (Randwick, New South Wales), St George Hospital (Kogarah, New South Wales), Royal Rehabilitation (Ryde, New South Wales), and The Royal Brisbane and Women’s Hospital (Herston, Queensland), and the people with aphasia and their family members.
Sources of Funding
This work was supported by the National Health and Medical Research Council Centre of Clinical Research Excellence in Aphasia Rehabilitation under grant number 569935, a Royal Brisbane & Women’s Hospital Foundation grant, and a Speech Pathology Australia postgraduate research grant. D. Copland was supported by an Australian Research Council Future Fellowship.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.115.009522/-/DC1.
- Received March 26, 2015.
- Revision received April 30, 2015.
- Accepted May 21, 2015.
- © 2015 American Heart Association, Inc.
- Cherney LR,
- Patterson JP,
- Raymer A,
- Frymark T,
- Schooling T.
- Bhogal SK,
- Teasell R,
- Speechley M.
- Kaplan E,
- Goodglass H,
- Weintraub S.
- 18.↵StepbyStep [computer program]. Steps Cottage, Littleton Drew, Wiltshire: Steps Consulting Limited; 2002.
- 19.↵AphasiaScripts [computer program]. Chicago: Rehabilitation Institute of Chicago; 2007.
- Kagan A,
- Simmons-Mackie N,
- Victor J,
- Carling-Rowland A,
- Hoch J,
- Huijbregts M,
- et al
- Tabachnick B,
- Fidell L.
- Swinburn K,
- Porter G,
- Howard D.
- Shapiro SS,
- Wilk MB.
- Legh-Smith JA,
- Denis R,
- Enderby PM,
- Wade DT,
- Langton-Hewer R.