Research Watch Report: Evidence-Based Treatment Approaches for Acquired Apraxia of Speech

Introduction

Acquired apraxia of speech (AOS) is a motor speech disorder that results in decreased intelligibility and impaired prosody. Specifically, AOS results from damage to the motor planning and programming areas of the brain, typically in the region of the Sylvian fissure. AOS is characterized by: slow rate of speech; speech sound errors that are predominantly distortions; pauses between syllables; and, dysprosody (McNeil, Robin, & Schmidt, 1997; 2009). Dysarthria is another motor speech disorder characterized by impaired speech production, but the cause of the impairment is muscle weakness. AOS may be differentiated from dysarthria by the absence of weakness in muscles responsible for speech production.

According to the ASHA AOS Practice Portal (n.d.), incidence and prevalence data for AOS are not readily available. This is attributed to the frequent co-occurrence of AOS, or AOS-like symptoms, in individuals with nonfluent aphasia. Indeed, AOS is most often exhibited in the presence of another communication disorder, presenting as the primary communication diagnosis in less than 7% of patients on the speech-language pathology (SLP) caseload at one prominent hospital in the United States (Mayo Clinic; ASHA, n.d.).

Although AOS is more often a secondary diagnosis, co-occurring with aphasia and/or dysarthria, the disorder represents an often difficult practice area for SLPs. Possibly because it is more likely to occur within the context of other communication disorders, AOS is often a complex problem, and can be difficult to treat when the ultimate goal is oral communication, particularly in more severe cases. In some cases, especially among individuals who present with concomitant cognitive-linguistic deficits, the most realistic goal is augmentative and alternative communication (AAC) system (ASHA, n.d.).

A number of AOS treatment approaches are supported by scientific evidence. These different treatments are generally categorized as articulatory-kinematic approaches, sensory cueing approaches, and rhythm and rate control approaches (ASHA, n.d.). Articulatory-kinematic treatments are those that emphasize remediation of the motor planning impairments. They typically employ principles of motor learning (PML). PML are assumptions about the ways in which individuals use practice and feedback to establish motor patterns. (For more detailed information on PML, the reader is referred to Maas et al., 2008.) Sensory cueing approaches focus on the use of multimodality stimulation, such as visual, auditory and tactile input to assist clients with sound production. Rate and rhythm control approaches, or prosodic facilitation approaches as they are sometimes called, are specifically geared toward remediating the prosodic impairments associated with AOS, but have also been found to improve speech sound production in that population (Wambaugh, Nessler, Cameron, & Mauszycki, 2012).

Some specific treatments incorporate aspects of more than one of these categories. For instance, Sound Production Treatment (SPT), a treatment method developed by Julie Wambaugh and colleagues, is primarily an articulatory-kinematic approach, but it also includes sensory cues. This is an important point to remember when considering treatment efficacy as reported in various studies. For example, Ballard, Wambaugh, Duffy, Layfield, Maas, Mauszycki, & McNeil (2015) conducted a systematic review of 26 articles published between 2004 and 2012. They categorized 24 of the studies as being articulatory-kinetic approaches and 2 as rate and/or rhythm control approaches. SPT studies were included in the articulatory-kinetic category. Those authors found strong evidence of efficacy for both articulatory-kinetic and rate and/or rhythm control approaches, but they did not compare specific treatments within those categories.

As an aside, AOS represents a small proportion of the current literature base in speech-language pathology.  A search of ASHAWire (which includes ASHA journals only) for treatment of acquired AOS yielded a total of 19 research articles between January 2016 and March 2020. Compare this to a search for treatment for aphasia, which yielded 52 research articles for the same timeframe.

The articles summarized in this research brief examine current research of two specific AOS or AOS/aphasia treatments, SPT and Combined Aphasia and Apraxia of Speech Treatment (CAAST). The table below provides details about each of these treatments. Another important note is that some of these studies report their results in terms of effect size. Effect sizes, in the studies reviewed here, are statistical calculations of the difference between a group’s pre-treatment and post-treatment performance. In essence, we interpret larger effect sizes to mean that the degree of change resulting from the treatment was clinically meaningful. Smaller effect sizes, on the other hand, indicate that any change was less meaningful or not meaningful at all.

Effect size is different from measures of statistical significance. Statistical significance is an indicator of whether a result is reliable or whether it could have been obtained by chance. In research articles, authors report p (probability) values, and results are typically considered statistically significant if the p value is <.05. This means that there is less than a 5% possibility that the result was obtained due to chance (and a 95% possibility that the result is reliable.). The relationship between statistical significance and clinical significance is not always predictable. For example, a result may be statistically significant, meaning that the result is reliable, but also have a small effect size, meaning that it is not clinically significant. The opposite may also be true (large effect size and not statistically significant).

Table. Details about Two Specific Treatment Approaches

Article 1. Wambaugh, J., Nessler, C., Wright, S., Mauszycki, S., DeLong, C., Berggren, K., & Bailey, D. (2017). Effects of blocked and random practice schedule on outcomes of sound production treatment for acquired apraxia of speech: results of a group investigation. Journal of Speech-Language-Hearing Research, 60, 1739-1751.

Background: What was the rationale and/or clinical question guiding this study? Given the evidence supporting PML in limb rehabilitation, these researchers set out to determine whether the PML of blocked vs. random practice also applies to motor speech rehabilitation, specifically AOS treatment using SPT. Blocked practice for AOS refers to a treatment schedule in which one sound is addressed in consecutive trials before moving on to another sound. Random practice, on the other hand, refers to a schedule in which target sounds are practiced in a random order. The authors cite evidence from the limb rehabilitation literature suggesting that blocked practice leads to more rapid improvements, but that random practice leads to increased skill retention and generalization. They predicted that they would find a similar pattern for AOS in this study.

Method: Who participated in the study and what did they do? Seven women and 13 men participated in the study (a total of 20 participants), ranging in age from 29 to 83 years. The participants’ months post-onset of a CVA or focal brain injury ranged from 6 to 259 months. All were diagnosed with mild, moderate, or severe AOS and aphasia. Based on the Western Aphasia Battery-Revised (WAB-R; Kertesz, 2007) and the Porch Index of Communicative Ability (PICA; Porch, 2001), 17 participants had Broca’s aphasia, two had anomic aphasia, and one had global aphasia. All participants demonstrated the ability to produce some monosyllabic words.

The researchers designed two experimental conditions, SPT-R (random) and SPT-B (blocked). All participants received both conditions. One group received SPT-R as their first treatment condition and SPT-B as their second treatment condition, while the other group received SPT-B first, then SPT-R. Participant groups were matched as much as possible for aphasia type, aphasia severity, and AOS severity.

Stimuli were developed to address six sounds that were in error for each participant based on his or her specific productions. One set of stimuli was used during treatment and the other set, untreated stimuli, was used to conduct probes. Treatment sessions lasted approximately an hour and the number of treatment sessions was equivalent across participants. The dependent measure was percent articulatory accuracy for treated sounds. Participants’ baseline percent articulatory accuracy for targeted sounds was recorded prior to the initiation of treatment. Probes for articulatory accuracy for the targeted sounds were conducted two weeks following the first treatment condition, two weeks following the second treatment condition, and again 10 weeks following the second treatment condition.

Results: What were the outcomes of the experimental measures? Effect sizes were calculated for each participant’s articulatory accuracy for treated phonemes. Mean effect sizes for both treatment conditions (random and blocked) were small to medium for treated stimuli and small for untreated stimuli. Percent change scores were also calculated and analyzed for statistical significance. These analyses revealed that participants performed significantly better on stimuli that were treated during the random schedule during treatment and at all probes (2-week follow-ups and 10-week follow-up). For the untreated stimuli, there was no significant difference between the random schedule and the blocked schedule.

Clinical Application: What are the take-home messages for me as an SLP? In general, these results partially replicate what has been found previously in limb rehabilitation research: random practice results in more robust gains during treatment. However, where limb rehabilitation evidence suggests that blocked practice is more beneficial for retention and generalization, this study found that there was no discernable difference between the two schedules for untreated items. The authors of the study also reported that they observed individual variability among their participants that is not obvious when considering these group outcomes. Specifically, they found that participants with mild-moderate AOS demonstrated responded more favorably to treatment than did those with severe AOS. Taken together, the results of this study suggest that the differences between random and blocked practice schedules may be statistically significant, but perhaps not clinically meaningful, particularly where generalization is concerned.

Article 2. Wambaugh, J., Wright, S., Boss, E., Mauszycki, S., DeLong, C., Hula, W., & Doyle, P. (2018). Effects of treatment intensity on outcomes in acquired apraxia of speech. American Journal of Speech-Language Pathology, 27, 306-322.

Background: What was the rationale and/or clinical question guiding this study? In this follow up to their 2017 publication on blocked vs. random practice schedules, Wambaugh and colleagues set out to examine the effects of another PML, intensity, on improvement in AOS. They referenced Kleim and Jones’ (2008) discussion of principles of neuroplasticity, in which those authors state that “intensity matters”, for promoting brain change. However, some studies suggest that a less intense schedule, called distributed practice (fewer treatment sessions over a longer timeframe), leads to superior learning when compared to a more intense schedule, or massed practice (more treatment sessions over a shorter period of time). The researchers set out to determine if SPT delivered in both more intense and less intense schedules resulted in differential effects on participants’ improvement on treated and untreated speech sounds.

Method: Who participated in the study and what did they do? Participants included 5 men, ages 44 to 64 years, with chronic AOS according to the McNeil et al. (1997, 2009) diagnostic criteria. Two participants presented with moderate-severe AOS, two with moderate AOS, and one with severe AOS. All participants also had aphasia according to the WAB-R (Kertesz, 2007), and overall intelligibility among the participants ranged from 34% to 86% based on the Assessment of Intelligibility of Dysarthric Speech (ASSIDS; Yorkston and Beukelman, 1981).

Participants were assigned to one of two treatment conditions: Intense (SPT-I) or Traditional (SPT-T). The SPT-I group members participated in nine intervention sessions, one hour per session, each week, for three weeks. The SPT-T group members participated in three intervention sessions, one hour per session, each week, for nine weeks. Both treatment conditions resulted in a total of 27 one-hour sessions. Each participant received both treatment conditions over the course of the experiment. Treatment stimuli were created for each individual participant and included a total of six target sounds that were relatively consistent error productions for each participant. Three sounds were addressed during the first treatment condition and the other three sounds were addressed during the second treatment condition. Treatment stimuli were practiced in a blocked manner, such that all items for a specific phoneme were presented consecutively before moving on to another sound.

A single-subject design was utilized so that each participant’s change during treatment was compared with his own baseline performance. This is different from group designs, which typically have a control group and an experimental group, and the performance of the groups is compared. Each participant received his first treatment condition (SPT-I or SPT-T), then after an intervening period of two or three weeks, received the other treatment condition. Probes for both treated and untreated stimuli were conducted before the initial treatment phase, at the conclusion of the initial treatment phase, at the conclusion of the second treatment phase, two weeks after the completion of the second phase, and again eight weeks after the completion of the second phase. 

Results: What were the outcomes of the experimental measures? The researchers measured two dependent variables, articulatory accuracy for treated phonemes and intelligibility based on the ASSIDS (Yorkston et al., 1981). With regard to the ASSIDS results, four participants demonstrated increased overall intelligibility following treatment. The fifth participant’s overall intelligibility was essentially unchanged.

Effect sizes were calculated to compare pre- and post-treatment articulatory accuracy. For three of the five participants, both treatment conditions produced similarly favorable outcomes for the treated stimuli. In other words, effect sizes for these participants were essentially equivalent for both treatment conditions for the treated items. For the other two participants, the traditional treatment condition led to larger effect sizes for the treated stimuli. The traditional treatment condition also led to greater effect sizes for the untreated stimuli across participants.

Clinical Application: What are the take-home messages for me as an SLP? Given that the findings for treated stimuli across conditions were equivalent for three participants and the traditional condition proved more effective for two participants, as well as the finding that the traditional condition also led to improved generalization to untreated stimuli, this evidence suggests that distributed practice (traditional model of treatment) may be superior to massed practice (intense treatment) for individuals with AOS. This finding appears to provide support for the more prevalent model of speech-language service provision in the U.S., which favors one or two sessions per week for a number of weeks or months. This model may also be more favorable for our clients and their families as multiple sessions per week, especially in the outpatient setting, can cause hardships in terms of scheduling, transportation, and the like.

Article 3 Wambaugh, J., Wright, S., Mauszycki, S., Nessler, C., & Bailey, D. (2018). Combined aphasia and apraxia of speech treatment (CAAST): Systematic replications in the development of a novel treatment. International Journal of Speech-Language Pathology, 20, 247-261.

Background: What was the rationale and/or clinical question guiding this study? CAAST was developed based on an evidence-based Response Elaboration Training (RET; Kearns, 1985) and SPT. CAAST represents a combination of RET and SPT and addresses both expressive language deficits and motor speech deficits among those with chronic expressive aphasia and AOS. One critical distinction between CAAST and SPT, however, is that SPT utilizes treatment stimuli developed by the clinician, whereas the client generates the targets (either spontaneously or when given a model utterance) in CAAST. An earlier study of the efficacy of CAAST was published in 2014 (Wambaugh, Wright, Nessler, & Mauszycki, 2014). The purpose of this 2018 study was to test a modification of the initial CAAST protocol, which included an additional emphasis on experimentally testing the SPT component.

Method: Who participated in the study and what did they do? Two men and two women, ages 39-69 years, served as participants for the study. All participants presented with chronic Broca’s aphasia and AOS. The time post-onset ranged from 26 to 103 months. Aphasia type and severity were confirmed according to the WAB-R (Kertesz, 2007) and the PICA (Porch, 2001). Overall intelligibility among the participants ranged from 76% to 98% based on the Assessment of Intelligibility of Dysarthric Speech (ASSIDS; Yorkston and Beukelman, 1981). AOS was diagnosed according to the McNeil et al. (2009) diagnostic criteria. Two participants presented with moderate-severe AOS, and two presented with mild-moderate AOS, based on both overall intelligibility and prevalence of speech sound errors.

The study, like the one discussed in Article 2,  followed a single-subject design. Therefore, each participant’s performance was compared to his or her baseline performance, which was measured prior to the beginning of intervention. Additional probes were conducted during treatment as well as two weeks after the completion of treatment and six weeks after the completion of treatment. Treatment sessions took place three times per week for 60 to 75 minutes per session with most participants receiving approximately 40 sessions.

Treatment followed the CAAST protocol (see Table for a description of the protocol). A treatment activity called the narrative discourse elicitation task utilized 24 black and white line drawings depicting common actions as stimuli. These stimuli were divided into three sets of eight cards each, two sets of which were to be addressed in treatment. The third set of cards was used for probing generalization of treatment effects to untreated items. In addition to the narrative discourse elicitation task, researchers also devised a speech elicitation task, in which participants were asked to produce a short sentence with simple syntax following the clinician’s model.

The dependent measures in the study were correct information units (CIUs), percent consonants correct (PCC) and word intelligibility. CIUs are essentially a way to quantify meaningful verbal expression in terms of its level of informativeness, and were derived from the narrative discourse task. PCC was calculated based on the sentence elicitation task, and word intelligibility was measured by the ASSIDS.

Results: What were the outcomes of the experimental measures? Effect sizes were reported for each participant's CIU outcomes. Three of the four participants’ effect sizes for treated stimuli were large, and large effect sizes were also found for two participants for the untreated stimuli. In addition, for all participants, number of CIUs showed an increase when measured following treatment. Three participants exhibited increased PCC at the six-week follow-up probe. Likewise, three participants’ increased word intelligibility as measured by the ASSIDS; the fourth participant was at 98% prior to treatment and after treatment.

Clinical Application: What are the take-home messages for me as an SLP? Based on the findings of this study, combining treatment for expressive aphasia and for AOS to address both disorders in a systematic way can be effective. The participants in this study, all of whom were more than two years post onset, improved the amount of meaningful information produced as well as speech production accuracy and intelligibility. In addition, the results of this study suggest that this treatment approach may be effective for those with across a range of severity levels.

Final Thoughts

The three studies summarized here provide support for the use of SPT in a traditional intervention model (distributed practice) as well as for CAAST for clients who present with both AOS and aphasia, which applies to the vast majority of clients with AOS. In addition, the evidence suggests that whether we target one sound at a time or randomize stimuli to address multiple sounds at once, individuals with AOS are able to generalize speech improvements gained using SPT.

These studies do have caveats, however, chief among them the small numbers of participants for Articles 2 and 3. These small numbers of participants necessitate the use of methods other than group or cohort designs, and they are not considered as one of the higher levels of evidence, although ASHA no longer uses a hierarchy to denote levels of evidence when engaging in evidence-based practice (see www.asha.org/Research/EBP/Assess-the-Evidence/). On the other hand, the use of single-subject design provides a treatment model that is analogous to clinical sessions in that participants (clients) are compared to themselves in order to measure change over time. They also allow for a more heterogeneous group of participants, because participants don’t have to be grouped or matched. This heterogeneity can demonstrate the efficacy of a treatment among different age levels, severity levels, and months post-onset, which is also helpful for clinicians.              

References

American Speech-Language-Hearing Association. (ASHA; n.d.). Acquired Apraxia of Speech. (Practice Portal). Retrieved March 6, 2020, from www.asha.org/Practice-Portal/Clinical-Topics/Acquired Apraxia of Speech/.

American Speech-Language-Hearing Association (ASHA; n.d.). Step 3: Assess the Evidence. Retrieved April 1, 2020, from https://www.asha.org/Research/EBP/Assess-the-Evidence/.

Ballard, K. J., Wambaugh, J. L., Duffy, J. R., Layfield, C., Maas, E., Mauszycki, S., & McNeil, M. R. (2015). Treatment for acquired apraxia of speech: A systematic review of intervention research between 2004 and 2012. American Journal of Speech-Language Pathology, 24(2), 316-337.

Kearns, K. P. (1985). Response elaboration training for patient-initiated utterances. Clinical Aphasiology, 15, 196-204.

Kertesz, A. (2007). WAB-R: Western aphasia battery-revised. PsychCorp.

Kleim, J., & Jones, T. (2008). Principles of experience-dependent plasticity: implications for rehabilitation after brain damage. Journal of Speech, Language, and Hearing Research,

51, S225-239.

Maas, E., Robin, D. A., Hula, S. N. A., Freedman, S. E., Wulf, G., Ballard, K. J., & Schmidt, R. A. (2008). Principles of motor learning in treatment of motor speech disorders. American Journal of Speech-Language Pathology, 17, 277-298.

McNeil, M. R., Robin, D. A, & Schmidt, RA. (1997). Apraxia of speech: Definition, differentiation, and treatment. In M. R. McNeil (Ed.), Clinical management of sensorimotor speech disorders, 311-344.

McNeil, M. R., Robin, D. A., & Schmidt, R. A. (2009). Apraxia of speech: Definition and differential diagnosis. Clinical management of sensorimotor speech disorders, 2nd

edition, 249-267.

Porch, B. E. (2001). Porch Index of Communicative Ability-Revised (PICA-R). Alberquerque (NM): Pica Programs.

Wambaugh, J. L., Kalinyak-Fliszar, M. M., West, J. E., & Doyle, P. J. (1998). Effects of treatment for sound errors in apraxia of speech and aphasia. Journal of Speech, Language, and Hearing Research, 41(4), 725-743.

Wambaugh, J., & Martinez, A. (2000) Effects of modified response elaboration training with apraxic and aphasic speakers, Aphasiology, 14, 603-617.

Wambaugh, J. L., Nessler, C., Cameron, R., & Mauszycki, S. C. (2012). Acquired apraxia of speech: The effects of repeated practice and rate/rhythm control treatments on sound production accuracy. American Journal of Speech-Language Pathology, 21, S5-27.

Wambaugh, J. L., Wright, S., Nessler, C., & Mauszycki, S. C. (2014). Combined aphasia and apraxia of speech treatment (CAAST): Effects of a novel therapy. Journal of Speech, Language, and Hearing Research, 57(6), 2191-2207.

Yorkston, K. M., & Beukelman, D. R. (1981). Assessment of intelligibility of dysarthric speech. Austin, TX: Pro-Ed.