Research Watch Report: "Intensive" Evidence-Based Aphasia Interventions

According to the National Institute on Deafness and Other Communication Disorders (NIDCD, 2015), about 1 million people in the United States are currently living with aphasia. ASHA’s (2017) data suggests that aphasia was the 2nd most commonly treated diagnosis by SLPs who work with adults from 2009 - 2015, after dysphagia, based on the percentage of time spent providing services. Aphasia dropped to 3rd place in 2017, after dysphagia (1^st) and dementia (2^nd), however, dementia only outranked aphasia by 1%. In any case, aphasia represents a significant portion of SLPs caseloads. 

Not only is aphasia common among our adult clients, but it is also often debilitating for these individuals because of its pervasive and chronic effects. Studies have documented the diminished quality of life reported by people with aphasia (PWA), which stems primarily from communication difficulties, isolation, decreased participation, and environmental barriers (see Hillari., et al., 2012). Given its prevalence as well as its significant influence on PWA and their caregivers, the disorder is one that is frequently addressed in the literature. The articles selected for discussion in this research brief focus on some of the latest evidence on some aphasia interventions that most SLPs are likely somewhat familiar with, including a modified version of constraint-induced language therapy and two investigations of telepractice and technology applications for individuals with aphasia.

Article 1: Mozeiko, J., Myers, E., & Coelho, C. (2018). Treatment response to a double administration of constraint-induced language therapy in chronic aphasia. Journal of Speech, Language, and Hearing Research, 61, 1664-1690.

Background: Why was this study conducted? Although current outpatient rehabilitation models do not typically provide a mechanism for this type of service provision, evidence suggests that aphasia outcomes are improved by intensive services (see Brady et al., 2016). Mozeiko and colleagues explain that this finding is not as straightforward as it might seem, as recovery is multifactorial, and current studies have not necessarily accounted for complex issues including aphasia severity and different treatment approaches. They also point to the inconsistencies in the literature on the definition of “intensive” services, which range from two hours to just under nine hours per week, depending on the study. 

One specific treatment type that is, by definition, delivered on an intensive basis is constraint-induced language therapy (CILT). CILT is a treatment approach in which the client must respond in the oral-verbal modality. CILT incorporates the principles of intensity (i.e., long treatment sessions) and massed practice over a brief period of time, usually a few weeks. The authors cite evidence suggesting that CILT is effective, with positive language outcomes at the completion of the intervention period, and even additional language improvements afterward. However, when it comes to CILT’s effectiveness based on aphasia severity, the results are less clear. Given that CILT appears to offer promising outcomes, including continued improvement after the intervention period has ended, Mozeiko and colleagues set out to answer three questions, which have been consolidated here for simplicity: (1) What, if anything, is the effect of aphasia severity on language outcomes following CILT? (2) Is there an additional benefit to providing two consecutive CILT interventions?

Method: Who participated in the study and what did they do? (Note: The methods of this study were quite complex and the following has been written to preserve the most relevant information in order to facilitate understanding of the study. Readers are directed to the source article if more detailed information about the experiment is needed.)

Participants included two individuals with mild fluent aphasia and two individuals with severe aphasia (one nonfluent, one fluent) based on the Western Aphasia Battery-Revised (WAB-R) Aphasia Quotient (AQ). The authors selected a multiple baseline design across participants. A multiple baseline design is a single-subject design in which baselines are established for several different behaviors. The intervention is provided, and the behaviors of interest are probed at specific times during the treatment. 

The treatment stimuli were cards depicting objects in full color, including high frequency objects and low frequency objects. The treatment protocol focused on verbal expression and included eight levels of complexity. Level 1 required participants to name a high frequency object, while Level 2 required them to name a low frequency object. Subsequent levels increased the complexity of the target, adding additional objects, adjectives, prepositions, and finally, to having the participants produce a full sentence description of the pictured object (Level 8). 

Prior to beginning the treatment protocol, baseline probes were conducted in order to establish the starting point (i.e., level) for each participant. Baseline probes included naming objects (using the color picture cards) and discourse production. Discourse was elicited by showing the participants Norman Rockwell illustrations and asking them to describe what was happening in the drawing. The authors included the productive discourse measure in order to test generalization from the treatment, which focused on discrete productions, to a more functional language context. The discourse probes were scored for productivity (quantity of information), efficiency (speed of production), and informativeness (proportion of the amount of relevant information compared to the total amount of information conveyed).

The CILT intervention was implemented using the full color picture cards in a game context similar to Go Fish. Participants were grouped into tryads (groups of three). Each participant received five to seven cards and asked the other participants in their group if they had a particular card. They were instructed to use complete verbal responses. The clinician first modeled the procedures of the game until the participants demonstrated competence. In keeping with the principles of CILT, requests and responses were made in the oral-verbal modality, with no written or gestural communication accepted as correct. The clinician, however, could provide phonemic or semantic cues when needed to facilitate the oral-verbal production.

The CILT intervention protocol was administered twice. Each treatment period lasted two weeks. Participants received services five days per week, for a three-hour session each day, resulting in a total of 30 hours of treatment per treatment period (60 hours over the two treatment periods). Following the completion of Treatment Period I, participants had a break from treatment for five weeks, then Treatment Period II commenced. Throughout the treatment period, probes were conducted to measure participants’ performance on treated and untreated stimuli (full color picture cards) and discourse production (Norman Rockwell descriptions). In addition, both the WAB-R and the Boston Naming Test (BNT) were administered four times throughout the treatment period.

Results: What were the outcomes of the experimental measures? Since the authors used a single-subject design, they provided results for each individual participant as well as summative data across the participants. They also reported effect sizes to illustrate differences in scores between various points in time (i.e., baseline probes to post-treatment probes). However, they stressed that effect sizes must be interpreted in conjunction with the visual, graphed data in order to fully appreciate the results. Effect sizes, as interpreted by the authors of the study, are reported here.

Treated and untreated stimuli. The results indicated variability in performance for the participants with mild aphasia (M1 and M2) compared to those with severe aphasia (S1 and S2).  M1 and M2 started at Level 4 and progressed to Level 8 during the course of the treatment. Medium to large gains were revealed for M1 and M2 for both treated and untreated stimuli, suggesting generalization. For S1 and S2, who both started at Level 1 and progressed to Level 2, small to medium gains were achieved, but only for treated stimuli, suggesting a lack of generalization. These effects were maintained by all participants at the maintenance period follow-up probe, four weeks after completion of Treatment Period II.

Discourse production. For measures of discourse production, M1 and M2 were already at a functional level at baseline. S1’s discourse analysis measures suggested significant gains. However, the authors noted that S1’s description relied heavily on overlearned or stereotypical utterances, so the gains are somewhat artificial. S2’s discourse production measures revealed more moderate gains.

Standardized measures. M1 and M2 made only small gains on the standardized measures (WAB-R and BNT), while S1 and S2 both made large gains on these measures after Treatment Period I and Treatment Period II.

One treatment period versus two. The authors reported that, as expected, all participants achieved clinical gains on all measures during Treatment Period I. However, results varied by aphasia severity in terms of the effectiveness of Treatment Perod II. The participants with mild aphasia (M1 and M2) benefited from the second treatment period for the treated and untreated stimuli only, and the participants with severe aphasia (S1 and S2) demonstrated additional gains, beyond Treatment Period I, on the BNT only. 

Clinical applications: What are the take-home messages for me as an SLP? The findings of this study are consistent with a number of previous studies suggesting that CILT can lead to significant clinical gains. Although some additional improvements could be attributed to the second treatment period, the authors report that Treatment Period II was “arguably of lesser value than the first treatment period” (p. 1683). In addition, the authors noted that they observed improvements in the participants that were not necessarily indexed by their measures. For example, they noted that S1 and S2 improved their ability to respond to cues, which is a functionally significant skill, but it was not specifically measured in this study.

In addition, the authors concluded that the individuals with mild aphasia demonstrated greater improvement on the specific naming task, including generalization to untrained stimuli. On the other hand, the participants with severe aphasia demonstrated significant improvement on the standardized tests, but did not realize significant gains on the naming task. Finally, the authors found that all of the participants maintained their gains when tested again four weeks following the completion of Treatment Period II.

Although the authors of this study acknowledged its limitations and lack of generalizability, they also determined that these results are “promising” (p. 1685). Practicing clinicians may take the results of this study, along with the results of a number of previous studies, to provide further evidence of the efficacy of the CILT approach, which includes the principles of responses constrained to the oral-verbal modality, massed practice, and intensive services. The authors also credit the use of small groups and the interactive game context as crucial to the success of the approach. Though current models of service provision do not typically allow for the intensive services and massed practice aspect in the clinical setting, clinicians can encourage clients and their caregivers to engage in more intensive practice sessions as part of their home program, strictly using the oral-verbal modality for responses. In addition, aphasia groups and interactive games can be fairly easily incorporated in a number of clinical settings.

Article 2: Simic, T., Leonard, C., Laird, L., Cupit, J., Hobler, F., & Rochon, E. (2016). A usability study of internet-based therapy for naming deficits in aphasia. American Journal of Speech-Language Pathology, 25, 642-653.

Background: Why was this study conducted?  With much of our population aging, the incidence of stroke in the United States is on the rise, which means that the number of individuals living with chronic post-stroke aphasia is also set to increase. Telerehabilitation is being considered in order to meet the growing needs of this clinical population, particularly those who live in outlying areas and/or are homebound. The authors cite a number of previous studies suggesting that telerehabilitation for individuals with aphasia appears to be a viable alternative to face-to-face speech-language pathology services. However, while some studies suggest that aphasia telepractice is feasible, at least one study (Galliers et al., 2011) found that individuals with aphasia have difficulty using technology because of their communication disorders. This study examined the usability of an online application for providing a technology-based analogue of a specific naming intervention approach, Phonological Components Analysis (PCA; Leonard et al., 2008).

Method: Who participated in the study and what did they do?  Participants included six individuals with mild to moderate aphasia resulting from a single left-hemisphere stroke. Participants had normal/corrected vision and visual perception skills as well as no known history of substance abuse, mental illness, or other neurological illness. The online PCA treatment protocol was called PhonoCom, and had been previously developed as a technology-based version of face-to-face PCA intervention. Earlier studies established that PCA was effective for increasing naming skills among PWA. For detailed information on the PCA treatment protocol, readers are directed to Leonard et al. (2008). 

Prior to initiation of the treatment protocol, a one-hour, face-to-face initial training session was conducted for participants and clinicians. The goal of the session was to orient everyone involved to the online PhonoCom application. In addition, the participants were given an aphasia-friendly user guide and cheat sheet to assist them with accessing the PhonoCom website.

Treatment occurred in real time through a video conferencing application in order to provide both visual and auditory information simultaneously, for a total of six hours over four visits during the study. The participants interacted with clinicians via PC laptops from their own homes or other locations remote from the clinician. During each treatment session, the clinician first guided the participant through the PCA protocol, a chart which was visible on the participant’s screen along with a picture of the target. The chart included prompts to encourage the participants to analyze phonological components of the target item to facilitate their naming of the item. The prompts included: first sound (“what does it start with?”); first sound associate (“what other words start with the same sound”); rhyme (“what does this rhyme with?”), and number of syllables (“how many syllables does this word have?”). Participants were allowed to use verbal responses or choose from a list of options with a mouse click. 

Researchers measured effectiveness, or the ability to complete the tasks successfully (i.e., opening the web browser, navigating to the application, logging in, etc.), and efficiency, or the effort required to use the application. Effectiveness was judged based on a scale of 0 (someone other than the participant had to complete the task) to 3 (the task was completed by the participant independently). Efficiency was indexed by the number of errors made by participants during each session on a scale of 0 (no errors) to 4 (six or more errors). One of the study’s authors observed each participant’s sessions to assess these variables.

In addition, following the six hours of treatment, all participants and clinicians completed a scale to indicate their opinions about the usability of the system. Semistructured interviews were also conducted in order to further assess participants’ and clinicians’ thoughts about the system. For both of these measures, assessment tools were modified to be aphasia-friendly as needed.

Results: What were the outcomes of the experimental measures?  Analyses revealed that all of the participants exhibited high levels of effectiveness across the four sessions, with most of them needing some assistance (effectiveness score of 2) or no assistance (effectiveness score of 3) to complete the tasks. Participants improved on their efficiency over the four sessions, making more errors initially and significantly fewer errors by the last session. The usability survey showed that clinicians found the system less usable (based on subjective report) than the participants did. Comments during the semistructured interviews were generally positive for all of the areas identified: convenience, usefulness, interest, interaction with clinician/participant, rapport, remoteness, audio/video quality, layout/design, and interaction with technology. Negative comments mostly included issues with sound/picture quality and dexterity issues (e.g., difficulty using the mouse), and some indicated that participants and/or clinicians felt the remote experience left something to be desired in the way of rapport building.

Clinical applications: What are the take-home messages for me as an SLP? While some PWA might have discomfort or difficulty accessing various technologies due to their communication disorders, the participants in this study were able to learn how to utilize the system initially and also to improve this ability over four sessions. In other words, we cannot assume that our clients with aphasia cannot or will not want to take advantage of developing technologies for increasing their communication skills. This particular study did not measure or report on treatment effectiveness. However, the next study reviewed here will provide some insight into that aspect of technology-based treatment.

Article 3: Kurland, J., Liu, A., & Stokes, P. (2018). Effects of a tablet-based home practice program with telepractice on treatment outcomes in chronic aphasia. Journal of Speech, Language, and Hearing Research, 61, 1140-1156. 

Background: Why was this study conducted? As noted previously in this research brief, the authors of this study also emphasize that the typical model of outpatient-based aphasia treatment does not align with current best evidence, which suggests that PWA benefit most from intensive services. Evidence suggests that providing intervention in larger blocks of time (i.e., several hours daily) over a shorter period (i.e., 2 weeks) is advantageous, but outpatient treatment is usually provided for smaller time blocks over a longer time frame. The mismatch between the evidence-based model and the model that is most prevalent in practice was the impetus for this study. The authors proposed that the use of home practice (HP) computer software applications might be one way to address the gap between scientific evidence and clinical practice in this area. The authors cite studies of iPad-based HP applications that suggest these programs can improve language and cognitive skills as well as aid in maintenance of progress made in therapy. The authors of this study sought to combine an intensive aphasia treatment program, use of iPad-based HP apps, and weekly teletherapy sessions in order to determine whether these methods would assist with maintenance and to examine the feasibility of using technology for aphasia treatment. 

Method: Who participated in the study and what did they do? Participants included 21 individuals (eight women, 13 men) with chronic aphasia secondary to a single MCA stroke. Assessments included the Western Aphasia Battery-Revised (WAB-R), the Boston Diagnostic Aphasia Examination (BDAE), an object naming task, and the Apraxia Battery for Adults-Second Edition. Aphasia severity was determined based on the WAB-R AQ and the BDAE Aphasia Severity Rating Scale. Following completion of a two-week intensive face-to-face intervention program, participants were assigned to a six-month iPad-based HP program and weekly teletherapy with an SLP. 

Participants, who were reported to have variable baseline proficiency with technological platforms, received training as needed in order to use iPads and the HP and video conferencing applications. Most participants required two to three 45-minute sessions to reach basic proficiency for independent use. Researchers verified that all participants had functioning wireless internet connections in their homes.

Participants’ HP programs included both stimuli that had been treated during their face-to-face intervention and untreated stimuli. The stimuli were black and white line drawings of common objects and actions that could be clearly represented in pictures (i.e., “bending” showed an image of a person bending over at the waist). The researchers created two individualized HP books for each participant’s iPad using iBooks Author. One iBook included 20 chapters of pictures of objects from the stimulus set and the other iBook included 20 chapters of picturable actions from the stimulus set. Each chapter consisted of four interactive pages. For example, the first page showed a picture and a series of prompts, from least support to most support, to assist the participant with naming the pictured object or action. The first prompt was “What is it?” and the final prompt was a video of an individual slowly saying and/or performing (in the case of the action words) the word. The other interactive pages included a picture-to-word matching task and an odd-man-out category task. The fourth interactive page allowed participants to check the accuracy of their responses on the tasks.

Participants were instructed to practice 5-6 days per week on their own, for at least 20 minutes each day. They were also asked to log their practice times. Weekly teletherapy sessions with a clinician lasted 30 minutes and were focused on casual conversation and questions from the clinician about HP compliance, technical issues, and additional feedback about using the device.

Dependent measures (i.e., participant assessments) were collected at four different times: at the beginning of the HP program, each month during the HP program, immediately following completion of the HP program, and during the first four months after completion of the HP program. These measures included subtests of the BDAE, the Boston Naming Test, and percent accuracy of naming for both treated and untreated stimuli.

Results: What were the outcomes of the experimental measures? Analyses revealed that the participants with severe aphasia made smaller gains on the naming task than did their counterparts with moderate-severe or mild-moderate aphasia. Performance was slightly better for stimuli that pictured objects than for action words, in general, but especially so for those with severe and moderate-severe aphasia. Participants performed better with treated stimuli than untreated stimuli. Gains on treated stimuli were maintained after the completion of the HP program, but were more pronounced and longer-lasting among those with mild-moderate aphasia only. In addition, most of the participants were able to use the HP program successfully, though those with more severe aphasia had more difficulty. Greater compliance with the HP program was associated with greater gains in the naming measure.

Clinical applications: What are the take-home messages for me as an SLP? The results of the study suggest that, as expected, aphasia severity plays a significant role in recovery and particularly maintenance of treatment gains. The authors of the study concluded that severity also likely prevented some of the participants from developing the ability to independently use the iPad and/or HP applications, despite fairly extensive training, due to their language deficits.

This study, when considered along with Articles 1 and 2, present some compelling evidence for the inclusion of technological applications as an adjunct to face to face sessions, with some considerations. These will be addressed in the next section.

Final Thoughts

The three studies summarized in this research brief focused on some questions that are central to the treatment of aphasia: (1) Does more intense therapy lead to stronger gains? (2) Can PWA utilize technology successfully? (3) Can technology help increase and/or extend the gains PWA make in face to face treatment sessions with SLPs? The answers to all of these questions seem to be “yes”, with some caveats.

The findings of these studies suggest that severity plays a role in the participants’ language gains and access to software applications. Participants with both milder and more severe forms showed improvements following both Treatment Periods I and II in Article 1. However, the measures they improved upon varied such that participants with mild aphasia improved on treated items and on generalization of naming skills, while those with severe aphasia improved only their scores on the Boston Naming Test. Of course, increased BNT scores are positive, but might not translate to functional communication skills as generalization to untreated stimuli could. Notably, aphasia severity did not affect maintenance, at least in the short-term – both those with milder and more severe aphasia had maintained their progress at post-treatment testing, four weeks after the end of Treatment Period II.

Articles 2 and 3 found that individuals with aphasia can use and benefit from technology and software applications. However, the results were conditional, such that those with more severe aphasia had more difficulty using technology and had a less robust maintenance effect than did participants with milder aphasia.

Taken together, the evidence reviewed here supports the use of the following approaches and strategies in aphasia intervention: interactive small groups with structured game-type activities; intensive services (defined in a number of different ways in the literature); constraining responses to the oral-verbal modality; and, therapy software applications for self-directed and/or prescribed home practice, counseling clients that home program compliance has been shown to lead to better outcomes. However, SLPs must also keep in mind that individual differences in PWA, such as severity, could affect the ability to successfully access technology, and therefore, language gains associated with intensive face to face services and/or home practice applications.