Caregivers’, teachers’, and assistants’ use and learning of partner strategies in communication using high-tech speech-generating devices with children with severe cerebral palsy

ABSTRACT

Communication with speech generating devices (SGDs) with children with severe physical, communicative and cognitive impairments, such as children with cerebral palsy (CP), can be difficult. Use of partner strategies facilitates the communication and instructional approaches such as feedback and role play facilitate communication partners’ learning in how to use partner strategies. To describe communication partners’ use and learning about partner strategies in SGD-mediated communication with children with severe CP. Questionnaires (n = 65) were sent to caregivers (n = 30), teachers (n = 17), and teaching or personal assistants (n = 18) of children with severe CP. Response rate was 80%. To ask open-ended questions was the most frequently used partner strategy and aided augmented input the least frequently used partner strategy. Most commonly, participants learned partner strategies from speech and language pathologists (SLPs) who used verbal instructions when teaching partner strategies but seldom or never feedback, role play or video examples. Communication partners’ learning about partner strategies in SGD-mediated communication is inadequate and needs to be improved. SLPs, who are the main prescribers of SGDs and responsible for training and support in using them, should consider using instructional approaches when teaching communication partners about partner strategies in communication with an SGD.

KEYWORDS:

• augmentative and alternative communication
• cognitive impairment
• communication

Introduction

Children with severe physical, communicative, and cognitive impairments can benefit from unaided and aided augmentative and alternative communication (AAC). AAC has positive effects on functional communication, challenging behaviors, and receptive and expressive language skills (Drager, Light, & McNaughton, 2010). Examples of unaided AAC are facial expressions, gestures, vocalizations, and movements. Aided AAC include external equipment, for example communication boards with pictures or graphic symbols, or a speech generating device (SGD). SGDs range from mid-tech battery-operated SGDs to high-tech computerized SGDs with synthetic or digitalized speech (Myrden, Schudlo, Weyand, Zeyl, & Chau, 2014). SGDs give individuals with severe communication impairments a voice and enable their participation in education, in face-to-face communication, and in communication over a distance (Caron & Light, 2016; Ganz et al., 2017; Holmqvist, Thunberg, & Peny Dahlstrand, 2017). SGDs can be directly accessed with selection aids such as pointing, trackballs, joysticks, and contacts, or by eye gaze technology. In recent decades, SGDs accessed using eye gaze technology have become more readily available, which provides individuals with severe physical and communicative impairments with enriched communicative opportunities (Borgestig, Sandqvist, Ahlsten, Falkmer, & Hemmingsson, 2016; Townend et al., 2016).

Despite its benefits, SGD-mediated communication (i.e. communication with an SGD) can be difficult to achieve (Baxter, Enderby, Evans, & Judge, 2012). It requires a lot of skill and effort from the individuals involved as it is a co-construction between the nonspeaking child (i.e. child with severe speech motor disorder) and the typically speaking person (Hörmeyer & Renner, 2013; Norén, Svensson, & Telford, 2013). By using the SGD with the communication partner, the child may develop linguistic, strategic, social, and operational competencies (Light & McNaughton, 2014), which are also valuable competencies in communication with an SGD. The child may successively understand the content of the symbols and how to use the navigational principles for communicative functions such as to request, ask, tell something, or answer a question. Moreover, children in need of eye gaze technology may need additional training because the technology requires a combination of visual, social, cognitive, and motor skills (Sargent, Clarke, Price, Griffiths, & Swettenham, 2013). Borgestig et al. (2016) found that children with physical, communicative, and cognitive impairments continued to improve their eye gaze performance over a period of 12–15 months. Furthermore, communication with an SGD can be facilitated if the communication partners use partner strategies (Gevarter & Zamora, 2018; Soto & Clarke, 2017). Current research on high-tech AAC interventions report positive effects on the child’s comprehension and expression when partner strategies, such as responsive strategy, environmental arrangements, open-ended question, aided augmented input, and behavior chain interruption strategy or least-to-most, are used (Ganz et al., 2017; Gevarter & Zamora, 2018; Morin et al., 2018).

Using responsive strategy, the communication partner awaits, attends to, and responds to the child’s intentional and non-intentional verbal and nonverbal communications (Fey et al., 2006). A detailed analysis of communication with an SGD shows that communication is facilitated when the communication partner awaits the prolonged SGD production process and attends to the nonspeaking person’s unaided AAC (Hörnmeyer & Renner, 2013; Engelke, & Higginbotham, 2013). Furthermore, the communication partner can arrange communicative appealing situations (i.e. use environmental arrangements) (Kaiser & Roberts, 2013). Environmental arrangements are an effective strategy for teaching individuals with severe physical, communicative, and cognitive impairments to request items and to initiate social interactions (Roche, Sigafoos, Lancioni, O’Reilly, & Green, 2015). The communication partner can ask open-ended questions (i.e. ask questions that require more than yes or no as an answer) – a strategy that can provide the nonspeaking person with additional opportunities for more linguistically complex answers with an SGD (Soto & Clarke, 2017). In addition, the communication partner can use aided augmented input, (i.e. point to and name symbols on the SGD during ongoing conversation) to model language. This strategy has been found to be highly effective across diagnoses and ages as it supports both comprehension and expression (Allen, Schlosser, Brock, & Shane, 2017; O’Neill, Light, & Pope, 2018). Moreover, the communication partner can use the behaviour chain interruption strategy or the least-to-most strategy. These strategies, which use stepwise interruptions in familiar routines, have been found to be effective in teaching individuals with cognitive impairments to request things (Carter & Grunsell, 2001; Finke et al., 2017; Kent-Walsh, Binger, & Malani, 2010).

Due to the specific communicative competencies required by communication partners, professionals need to consider how to provide training that targets communication partners. Communication partners typically need opportunities for reflection, time for problem-solving, their own practice and feedback (Dunst & Trivette, 2009). Kent-Walsh and McNaughton (2005) and Kent-Walsh et al. (2010) have proposed models of instructional approaches to be used when training communication partners in the use of partner strategies. These models include (1) description and modeling of the partner strategy by the instructor, and (2) the communication partner’s own practice of the partner strategy in role play and in real life with feedback from the instructor (Kent-Walsh, Murza, Malani, & Binger, 2015). Without adequate support and training, there is a risk of low use or even abandonment of SGDs (Anderson, Balandin, & Stancliffe, 2014; Bailey, Parette, Stoner, Angell, & Carroll, 2006a; McMillian, 2008; Stadskleiv, 2017; van Niekerk & Tonsing, 2015). Tegler, Pless, Blom Johansson, and Sonnander (2018) found inequalities in training targeting communication partners in SGD-mediated communication. Teachers and assistants received more training than caregivers and typically only one caregiver in the family received training. Previous research suggests that communication demands differ between home and school. Caregivers often understand the child’s unaided AAC and may not have the same motivation to use aided AAC as communication partners unfamiliar with the child may have (Bailey et al., 2006a). Teachers, on the other hand, may need aided AAC to ensure that the child is learning and is able to demonstrate the knowledge acquired (Holmqvist et al., 2017).

The most common reason for severe physical impairments in childhood is cerebral palsy (CP). The prevalence of CP is 2 out of every 1,000 children (Oskoui, Coutinho, Dykeman, Jette, & Pringsheim, 2013). CP is a permanent, nonprogressive disorder that occurs in the developing fetal or infant brain that affects motor activity (Rosenbaum et al., 2007). CP is often associated with impairments of speech, language, cognition, vision, attention, and executive functions (Rosenbaum et al., 2007). The prevalence of severe speech motor disorders and communication impairments in children with CP ranges between 46% and 55% (Hustad, Oakes, McFadd, & Allison, 2016; Zhang, Oskoui, & Shevell, 2015). Furthermore, there is a high comorbidity between severe speech motor disorder, communication impairment, and cognitive impairment: 94% of children with severe physical impairment GMFCS IV-V (gross motor functional classification system) (Compagnone et al., 2014) have communication impairments and 70.1% have severe speech motor disorder (Zhang et al., 2015). Accordingly, children with severe CP can benefit from AAC.

Previous research supports that children with CP can communicate with high tech SGDs (Clarke & Wilkinson, 2008; Hörmeyer & Renner, 2013; Soto & Clarke, 2017). However, this research is limited. There are a few studies that used conversation analysis which shows how communication partners use open-ended questions, environmental arrangements, and responsive strategy in SGD-mediated communication with individuals with CP (Clarke & Wilkinson, 2008; Hörmeyer & Renner, 2013). There are also some studies that used interviews and questionnaires to acquire data on the perceptions and practices of communication partners and individuals with CP (Caron & Light, 2016; Holmqvist et al., 2017). According to these studies, the SGD can provide additional communicative opportunities in social and learning activities. Finally, there are a few intervention studies that show positive effects when communication partners use environmental arrangements, open-ended questions, and linguistic modeling (Soto & Clarke, 2017; Ballin, Balandin, Stancliffe, 2012).

In conclusion, it has been shown that children with CP can communicate with high-tech SGDs (Borgestig et al., 2016; Clarke & Wilkinson, 2008; Ganz et al., 2017; Gevarter & Zamora, 2018; Hörmeyer & Renner, 2013; Roche et al., 2015); partner strategies are effective in SGD-mediated communication (Ballin, Balandin, & Stancliffe, 2012; Gevarter & Zamora, 2018, 2018; Morin et al., 2018; Soto & Clarke, 2017); and instructional approaches are effective when teaching communication partners how to use partner strategies (Kent-Walsh & McNaughton, 2015; Dunst & Trivette, 2009; Kent-Walsh et al., 2010). However, communication partners report inadequate professional support, which can affect their ability to scaffold communication using the SGD (Anderson et al., 2014; Bailey et al., 2006a; McMillian, 2008; Stadskleiv, 2017; Tegler et al., 2018; van Niekerk & Tonsing, 2015). The specific aim of the present study is to investigate important communication partners’ (caregivers, teachers, and assistants) use and learning of partner strategies in communication with children with severe physical, communicative, and cognitive impairments using high-tech SGDs. Specifically, this study asks:

1. Do communication partners use partner strategies when interacting with children who use high-tech SGDs?

2. If so, which strategies (among those presented) do communication partners use?

3. How have communication partners learned about partner strategies in SGD-mediated communication?

4. Do differences exist in the use and learning of partner strategies in SGD-mediated communication between the home and school contexts?

Method

A cross-sectional descriptive and comparative design was used with questionnaires developed specifically for the study.

Setting

In Sweden, where this study took place, SGDs are funded by tax revenues and are free of charge for users. Multidisciplinary SGD teams including SLPs, occupational therapists, physiotherapists, and technicians provide assessment, training and support to children with CP and their communication partners. There are guidelines for the prescription of assistive devices (Blomquist & Jacobson, 2011) which are regulated by the Hälso- och sjukvårdslagen (2017:30) (Sweden’s Health and Medical Service Act). According to the guidelines, the prescriber is responsible for the assessment of needs prior to prescription, the individualization of the device, and informing and training the child and his/her communication partners. Most commonly, the prescriber of an SGD is an SLP.

Participants

Caregivers, teachers, teaching assistants, and personal assistants of children for whom high-tech SGDs had been prescribed were recruited throughout Sweden from October 2014 to June 2015. To recruit the participants, the children first needed to be identified.

The study-specific inclusion criteria for the children were: (a) prescribed a high-tech SGD (e.g. Tobii, Rolltalk, Tellus, and Grid Pad Go), (b) 7–18 years of age, (c) severe CP, and (d) intellectual disability. Severe CP was defined as children performing at Level IV or V of the GMFCS and Level III−V of the Manual Ability Classification System (MACS) (Compagnone et al., 2014). Intellectual disability was established by the child’s participation in the 9-year special school system (Swedish National Agency for Education, 2011). In Sweden, participation in the special school system postulates intellectual disability diagnosed by a psychologist. Accordingly, the children fulfilling the study-specific inclusion criteria did not have oral speech, they used a wheelchair, had difficulties using their hands and had an intellectual disability. The only exclusion criterion was children also diagnosed with autism spectrum disorder.

At the time of the data collection there were 711 Swedish children aged 7–18 years (born 1997–2008) with severe CP performing at Level IV−V of the GMFCS according to the Swedish follow-up surveillance program and quality registry for individuals with CP (CPUP, e-mail correspondence, 2014-11-17). It was not possible to determine from the register how many of these children had an SGD and intellectual disability but not autism.

The recruitment of participants was therefore carried out step-wise starting with identifying the potential participants. Managers of technical aid centers in all Swedish county councils or regions received written information about the study and a request to assign a contact who could identify potential participants. Seventeen out of 20 county councils responded. The assigned contacts identified 39 children, and the caregivers of 30 of these children were informed about the study. The caregivers of the remaining nine children were not informed because the contacts hesitated to intrude on the caregivers, or the children were hospitalized, or there was no SLP involved (however, access to an SLP was not an inclusion criterion). The contacts obtained written informed consent from the caregivers of 16 children and forwarded those to the first author. The first author then telephoned the caregivers to get the contact details of the child’s SLP, teachers, and assistants. Caregivers who were not fluent in spoken and written Swedish were contacted by the child’s SLP, who forwarded the contact details to the first author. Translators authorized by the Swedish Legal, Financial and Administrative Services Agency (Kammarkollegiet) translated information about the study and written consent into these caregivers’ native languages. Finally, the first author telephoned the teachers and assistants to obtain their verbal consent. Detailed information about each child’s age and GMFCS and MACS level was obtained from the child’s SLP.

In total, 65 communication partners, caregivers (n = 30), teachers (n = 17), and assistants (n = 18), received the questionnaire. The response rate was 80.0% (n = 52). Twenty-three caregivers, 14 teachers, and 15 assistants filled in and returned the questionnaires. There were thus 42 female and 10 male participants in the study from the South (n = 30), the Middle (n = 19), and the North (n = 3) of Sweden. Thirty-four participants (65.4%) provided answers in relation to SGDs accessed with eye gaze technology and 18 participants (34.6 %) in relation to other access methods. There were 10 caregivers (19.2%) who were not fluent in Swedish but spoke Arabic, Albanian, or Tigrinya. Eight teachers and six assistants had previous experience of SGDs.

The participants answered questions in relation to 16 children identified for the study (8 girls and 8 boys) aged 7–18 years (M = 12.6). There were 1–4 respondents (participants) per child. Ten children used a Tobii, three used a Rolltalk, two used a computer/iPad, and one used a Grid Pad Go. The number of symbols and the languages of the SGDs were not known. The SGD was used for different purposes (Table 1). Ten children used eye gaze technology, three children used contacts (e.g. trackballs), and three children used direct pointing as their access method. Nine children had used their SGD for at least 2 years and five children for 3 years or more. For two children, no information about the period of use of their SGD was provided.

Table 1. Daily use of the SGD according to caregivers (n = 23), teachers (n = 14), and assistants (n = 15), (n_tot = 52)

	Caregivers weekday n %	Caregivers weekends n %	Teachers n %	Assistents n %
Choose activity	6 (26,9)	9 (40,1)^a	8 (66,7)^b	8 (61,5)^b
Answer questions	7 (33,3)^b	9 (39,1)	7 (58,3)^b	6 (50,0)^c
Ask questions	7 (30,4)	8 (34,8)	5 (41,7)^b	4 (33,3)^c
Explain that something is wrong	3 (13,6)^a	5 (21,7)	6 (50,0)^b	5 (45,5)^d
Narrate	5 (22,7)^a	7 (30,0)	5 (41,7)^b	7 (58,3)^c
Express social phrases	3 (13,6)^a	5 (21,7)	5 (45,5)^c	3 (25,0)^c
Play games	10 (43,5)	14 (60,9)	7 (58,3)^b	6 (46,2)^b

^amissing value (n = 1), ^bmissing value (n = 2), ^cmissing value (n = 3), ^dmissing value (n = 4)

Questionnaires

Two study-specific questionnaires about the use of and learning about partner strategies in SGD communication targeting either caregivers or teachers and assistants were developed (see Appendix). The questionnaires were based on previous research and the clinical experience of the first author and covered five domains: demographic and background data, hardware and software of the SGD, use of the SGD, use of partner strategies, and learning about partner strategies. There were 15 closed-ended comprehensive questions with subitems directed to caregivers and 16 closed-ended comprehensive questions with subitems directed to teachers and assistants with the option to add comments to each question. The questions were answered with fixed-alternative responses or on scales.

Questions about the use of partner strategies were responded to on an 8–9 level ordinal scale; 6 times/day or more, 4–5 times/day, 2–3 times/day, 1 time/day, every second day, every third day, more seldom, and never. In the 9-level scale, seldom was divided into more seldom because other AAC is more effective and more seldom because of lack of vocabulary. Questions about from whom caregivers, teachers, and assistants had learned about partner strategies and what instructional approaches the SLPs had used when teaching the partner strategies in SGD-mediated communication were responded to on a 7-point Likert scale with response alternatives ranging from 6 (to a high degree) to 0 (not at all).

The questionnaires were piloted for relevance and feasibility by caregivers of children meeting the study-specific criteria (n = 2), their teachers (n = 2), assistants (n = 2), and a director of research and development of the Riksförbundet Rörelsehindrade Barn och Ungdomar (Sweden’s National Association for Physically Disabled Children and Youth) (n = 1). Minor adjustments were made according to comments from the pilot respondents. None of these respondents participated in the study.

Procedure

The caregivers, teachers, and assistants were given three choices in how to respond to the questionnaire: online, on paper, or by telephone by the first author telephoning the participants. Those caregivers who were not fluent in Swedish could either respond to the questionnaire translated by an authorized translator or use an authorized interpreter (Kammarkollegiet). Telephone interviews with caregivers not fluent in Swedish were conducted by the first author together with an interpreter using a speakerphone. The telephone interviews were then transcribed while the interview was going on. The first author and the interpreter later checked the transcripts to ensure accuracy. Two reminders were provided. The time to respond to the questionnaire varied from 1 to 6 months depending on when the participants had been contacted during the data collection period.

Data analysis

Statistical Package for the Social Sciences (SPSS (version 22.0), 2015) was used for descriptive statistics. The answers given on the 8- or 9-level scale ranging from 6 times/day or more often to never were collapsed into three groups; ‘once a day or more often,’ ‘every second or third day,’ and ‘once a week or less often’. The answers given on the scale ranging from 6 to 0 were also collapsed into three groups; often (6–4), seldom (3–1), and never (0).

Conventional content analysis was used to analyze the comments on the questions about the use of and learning about partner strategies (Hsieh & Shannon, 2005). Credibility was assured by a thorough description of the data collection process and how the data were analyzed. The analyzing process comprised four steps. Firstly, all comments relating to the research questions were carefully read through. Phrases and sentences that described an attitude, a belief or an experience of the question were highlighted and meaning units were identified. Secondly, the meaning units were condensed, and thirdly, coded. Lastly, the codes were categorized according to their meaning. One comment was excluded because it was too short and vague. The analyzing process continued until consensus was reached among the four authors. In all, there were 35 codes and 5 categories. Citations are used to provide examples of the categories.

Ethical considerations

The ethical regulations and guidelines were followed according to Swedish Law 2003:460 (CODEX, 0000). The study was approved by the Regional Ethical Review Board in Uppsala, Sweden (Reg.no. 2014/200). The information letters sent to participants contained information about the study, voluntariness, anonymity, confidentiality, and the option to withdraw without explanation.

Results

The questionnaires were answered online (n = 37), on paper (n = 8), and by telephone (n = 7). Missing values per questionnaire domain were: (1) demographic and background data 4.2%, (2) hardware and software of the SGD 3.1%, (3) use of the SGD 8.9%, (4) use of partner strategies 6.0%, and (5) learning about partner strategies 5.3%.

Use of partner strategies

As can be seen in Table 2, open-ended questions were the most commonly used partner strategy among the four alternatives responsive strategy, environmental arrangements, open-ended questions, and aided augmented input. Caregivers _WD (i.e. weekdays) (47.6%), caregivers _WE (i.e. weekends) (59.1%), teachers (83.3%), and assistants (76.9%) used open-ended questions once a day or more often in communication with the SGD. The least used strategies were aided augmented input followed by environmental arrangements. Caregivers _WD (27.3%), caregivers _WE (36.4%), teachers (53.8%), and assistants (61.5%) used aided augmented input once a day or more often. Caregivers _WD (22.7%), caregivers _WE (28.6%), teachers (69.2%), and assistants (66.7%) used environmental arrangements once a day or more often. Responsive strategy was use to a medium. Caregivers _WD (36.3%), caregivers _WE (45.4%), teachers (61.5%), and assistants (76.9%) used responsive strategy once a day or more often. Not shown in Table 2, caregivers (39.1%), teachers (42.9%), and assistants (46.7%) used the behaviour chain interruption strategy to teach the child SGD communication.

Table 2. The numbers and percent of caregivers’ (n = 23), teachers’ (n = 14) and assistants’ (n = 15) use of partner strategies in SGD-mediated communication (n_tot = 52)

		Once a day or more n (%)	Every second or third day n (%)	Once a week or less n (%)
Responsive strategy
	Caregivers WD^a	8 (36.3)	3 (13.6)	11 (50.0)
	Caregivers WE^a	10 (45.4)	5 (22.7)	7 (31.8)
	Teachers^a	8 (61.5)	3 (23.1)	2 (15.3)
	Assistants^b	10 (76.9)	1 (7.7)	2 (15.3)
Environmental arrangements
	Caregivers WD^a	5 (22.7)	4 (18.2)	13 (59.1)
	Caregivers WE^b	6 (28.6)	4 (19.0)	11 (52.4)
	Teachers^a	9 (69.2)	1 (7.7)	3 (23.1)
	Assistants^c	8 (66.7)	0 (0.0)	4 (33.3)
Open-ended questions
	Caregivers WD^b	10 (47.6)	5 (23.8)	6 (28.6)
	Caregivers WE^a	13 (59.1)	3 (13.6)	6 (27.3)
	Teachers^b	10 (83.3)	0 (0.0)	2 (16.7)
	Assistants^b	10 (76.9)	0 (0.0)	3 (23.1)
Aided augmented input
	Caregivers WD ^a	6 (27.3)	3 (13.6)	13 (59.1)
	Caregivers WE^a	8 (36.4)	4 (18.2)	10 (45.4)
	Teachers^a	7 (53.8)	1 (7.7)	5 (38.5)
	Assistants^b	8 (61.5)	1 (7.7)	4 (30.8)

WD = Weekdays, WE = Weekends

^amissing value (n = 1), ^bmissing value (n = 2)

In general, caregivers used the four partner strategies less often than teachers and assistants. In addition, caregivers used the strategies more often on weekends than on weekdays. The biggest difference among the participants was the use of environmental arrangements. Caregivers _WD (22.7%), caregivers _WE (28.6%), teachers (69.2%), and assistants (66.7%) used this partner strategy once a day or more often.

Learning about partner strategies in SGD-mediated communication

The participants most commonly learned about partner strategies from SLPs; caregivers (34.7%), teachers (35.7%), and assistants (28.5%) often learned from SLPs (Table 3). In addition, participants learned in different ways. Caregivers (30.4%) often learned from other professionals (e.g. occupation therapists, special education teachers) compared to teachers (14.2%) and assistants (14.2%). Teachers (28.5%) often learned from the literature compared to caregivers (8.6%) and assistants (0%).

Table 3. The numbers and percent of caregivers (n = 23), teachers (n = 14) and assistants (n = 15) who had learned to use partner strategies in SGD-mediated communication often, seldom, or never from different persons, literature or websites (n_tot = 52)

		Caregivers n (%)	Teachers n (%)	Assistants n (%)
The child’s speech and language pathologist
	Often	8 (34,7)	5 (35,7)	4 (28,5)^a
	Seldom	9 (40,9)	4 (28,5)	5 (35,7)^a
	Never	6 (26,0)	5 (35,7)	5 (35,7)^a
Other professionals in the SGD team e.g. occupational therapists, teachers, physiotherapists
	Often	7 (30,4)	2 (14,2)	2 (14,2)^a
	Seldom	7 (30,4)	4 (28,5)	3 (21,4)^a
	Never	9 (39,1)	8 (57,1)	9 (64,2)^a
Teacher
	Often	5 (21,7)	1 (7,6)^a	2 (14,2)^a
	Seldom	2 (8,6)	4 (30,7)^a	1 (7,1)^a
	Never	16 (69,5)	8 (61,5)^a	11 (78,5)^a
Assistants
	Often	3 (13,0)	2 (14,2)	1 (7,6)^a
	Seldom	1 (4,3)	3 (21,4)	5 (35,7)^a
	Never	19 (82,6)	9 (64,2)	8 (57,1)^a
Caregivers
	Often	3 (13,0)	−	2 (14,2)^a
	Seldom	3 (13,0)	6 (42,8)	7 (50,0)^a
	Never	17 (73,9)	9 (64,2)	5 (35,7)^a
Other student at school
	Often	4 (17,3)	1 (7,6)	2 (14,2)^a
	Seldom	2 (8,6)	5 (35,7)	2 (14,2)^a
	Never	17 (73,9)	8 (57,1)	10 (71,4)^a
Internet, web sites
	Often	3 (13,0)	3 (23,0)^a	−^a
	Seldom	3 (13,0)	3 (23,0)^a	3 (21,4)^a
	Never	17 (73,9)	7 (53,8)^a	11 (78,5)^a
Literature
	Often	2 (8,6)	4 (28,5)	−^a
	Seldom	2 (8,6)	5 (35,7)	2 (14,2)^a
	Never	19 (82,6)	5 (35,7)	12 (85,7)^a

^amissing value (n = 1)

SLPs most commonly used verbal instructions and modeling to teach partner strategies with the SGD (Table 4). A high percentage of SLPs never used feedback when practicing a strategy to teach caregivers (68.1%), teachers (46.1%), and assistants (69.2%); feedback on videotaped material to teach caregivers (86.3%), teachers (84.6%), and assistants (85.7%); or role play to teach caregivers (86.3%), teachers (84.6%), and assistants (78.5%) communication with the SGD. There were differences between participant groups. SLPs often used verbal descriptions to teach teachers (33.3%) compared to caregivers (18.1%) and assistants (9.0%); and SLPs often used modeling to teach assistants (30.7%) compared to caregivers (18.1%) and teachers (23.0%).

Table 4. The numbers and percent of caregivers (n = 23), teachers (n = 14), and assistants (n = 15) who reported which instructional approaches the SLP use to teach them to use partner strategies in SGD-mediated communication during the last 12 months (n_tot = 52)

		Caregivers n %	Teachers n %	Assistents n %
Verbal description of strategy
	Often	4 (18,1)^a	4 (33,3)^b	1 (9,0) ^c
	Seldom	10 (45,4)^a	4 (33,3)^b	6 (54,5)^c
	Never	8 (36,3)^a	4 (33,3)^b	4 (36,3)^c
Written description of strategy
	Often	3 (13,6)^a	1 (7,6)^a	−1
	Seldom	7 (31,8)^a	5 (38,4)^a	3 (21,4)^a
	Never	12 (54,5)^a	7 (53,8)^a	11 (78,5)^a
Modeling strategy
	Often	4 (18,1)^a	3 (23,0)^a	4 (30,7)^b
	Seldom	6 (27,2)^a	3 (23,0)^a	3 (23,0)^b
	Never	12 (54,5)^a	7 (53,8)^a	6 (46,1)^b
Feedback when practicing strategy
	Often	3 (13,6)^a	2 (15,3)^a	2 (15,3)^a
	Seldom	4 (18,1)^a	5 (38,4)^a	3 (23,0)^a
	Never	15 (68,1)^a	6 (46,1)^a	8 (69,2)^a
Feedback on videotaped material
	Often	2 (9,0)^a	−1	−1
	Seldom	1 (4,5)^a	2 (15,3)^a	2 (14,2)^a
	Never	19 (86,3)^a	11 (84,6)^a	12 (85,7)^a
Role play
	Often	2 (9,0)^a	−1	2 (14,2)^a
	Seldom	1 (4,5)^a	2 (15,3)^a	1 (7,1)^a
	Never	19 (86,3)^a	11 (84,6)^a	11 (78,5)^a
Videoclip of others using the strategy
	Often	1 (4,5)^a	1 (7,6)^a	−2
	Seldom	2 (9,0)^a	2 (15,3)^a	1 (7,6)^b
	Never	19 (86,3)^a	10 (76,9)^a	12 (92,3)^b

^a(missing (n=1), ^bmissing value (n=2), ^cmissing value (n=4).

Results of the qualitative analysis

The comments were grouped into five categories: communication partners’ feelings, barriers to communication with an SGD, facilitators to communication with an SGD, prerequisites for communication with an SGD, and use of partner strategies in SGD-mediated communication.

Caregivers described being grateful for all the possibilities that the SGD provided the child with, but they were also distressed, frustrated, and lacked energy.

We are so grateful to get access to a Tobii [SGD], but at the same time distressed at not receiving more help… Is it supposed to be like this, SHOULD I, the mother, have the knowledge, energy and time for everything? (Caregiver 4)

Barriers to communication with an SGD were: lack of support from SLPs and other professionals; lack of own knowledge and practice and/or lack of time, energy and drive. In addition, high turnover among assistants hindered communication with the SGD, and the difficulties in the SLP’s ability to instruct bilingual caregivers was a barrier.

We have not received any help from the SLP. He [the child] would be able to communicate even more and [his communication would be] more comprehensible if we had received more help (Caregiver 3)

On weekday mornings it is impossible to have time to use the Tobii [SGD]. We use paperboard communication, a calendar and the schedule [low tech AAC] to communicate about toilet, clothing and breakfast (Caregiver 4).

We have often felt a little bit lost concerning this [behaviour chain interruption strategy] and [we] need more education (Teacher 14)

It is hard to understand. I have received a lot of information from the speech and language pathologist, but it was hard to understand (Caregiver 10)

Effective support from SLPs facilitated communication with the SGD.

Our speech and language pathologist has been working really hard with the student…most often he finds the symbols better than we do (Assistant 1)

Prerequisites for communication with an SGD were knowledge about partner strategies and the possibility to learn within the family.

[I] don’t have enough knowledge about the Tobii to do this analysis [BCIS] (Caregiver 6)

I have received instruction from the child’s mom (Caregiver 20)

The use of partner strategies in SGD-mediated communication was limited. Communication partners commented that they did not use partner strategies or that there was no need to use partner strategies because the child communicated well anyway. Partner strategies were used with low tech AAC to model language.

[I] don’t use an augmented input strategy with the SGD because the student can use it himself. [I] use an augmented input strategy with the blissboard [low tech AAC] to model sentences. (Teacher 4)

I have taken a course in how the computer [SGD] works and encouraged the student to use it and do it in a fun way. I have not analysed anything [BCIS] (Assistant 2)

Discussion

The findings of this study were based on reports from 23 caregivers, 14 teachers, and 15 assistants who communicated regularly with 16 children with severe CP and intellectual disability using high-tech SGDs. In summary, open-ended questions was the partner strategy most frequently used to support SGD-mediated communication and aided augmented input the least frequently used. The participants mostly learned about SGD communication from SLPs, who often used verbal descriptions of partner strategies and seldom or never used feedback or role play when teaching partner strategies.

Use of partner strategies and use of SGD

A high use of open-ended questions is promising because this strategy contributes to vocabulary expansion and grammatical improvement and it provides individuals with severe physical and communicative impairments with the opportunity to choose the topic (Hörmeyer & Renner, 2013; Soto & Clarke, 2017). A lower rate of use of aided augmented input could be problematic because this strategy has been found to contribute to extending vocabulary and the use of multi-symbol utterances (Allen et al., 2017; Soto & Clarke, 2017; Binger et al., 2011). The participants in this study commented that aided augmented input was not needed because the child used the SGD unimpeded. Instead, the strategy was used on low-tech AAC to model language. This finding mirrors research by Holmqvist et al. (2017) who found that communication partners did not use aided augmented input on SGDs accessed with eye gaze technology but on the child’s communication board. In the current study, 34 participants (65.4%) provided answers in relation to SGDs accessed with eye gaze technology. Research on aided augmented input in relation to eye gaze technology is limited. Using aided augmented input in relation with eye gaze technology requires special arrangements because of the infrared light. The infrared light needs to be closed off or covered to ensure that it does not interfere with the strategy. There is a lack of knowledge about if and how to use aided augmented input in relation to eye gaze technology (Allen et al., 2017) – a knowledge gap that needs to be investigated further.

Differences between the participants’ use of partner strategies were discovered. Teachers and assistants used partner strategies more often than caregivers; the SGD was used more often at school than at home: and caregivers commented on high stress and time limitations and they used the strategies more often on weekends than on weekdays. The reasons for these differences are not known, but might reflect a few aspects. Firstly, the results may indicate that there are different needs for SGD-mediated communication at school and at home. Teachers want to implement SGDs in educational settings (McMillian, 2008). Unlike caregivers, they have an educational role; they need to ensure that all children, including children with severe physical, communicative and cognitive impairments, can participate in the education system and that they can evaluate the child’s knowledge to plan their further education. Secondly, the caregivers in the study commented that they suffered high stress and lack of energy in relation to SGD-mediated communication. According to Raina et al. (2005), caregivers to children with severe disabilities often experience higher levels of physical and emotional stress compared to caregivers to typically developed children. The comments on stress and fatigue in combination with lower use of partner strategies should be taken seriously because caregivers’ participation in SGD interventions is important (Bailey et al., 2006a; Bailey, Stone, Parette, & Angell, 2006b). The caregivers in this study commented that they might have provided more SGD-mediated communication if they had received more support from the SLP. Previous research confirms that SLP support is often inadequate (Tegler et al., 2018).

Learning about partner strategies in SGD-mediated communication

Most commonly, the participants in this study learned about partner strategies in SGD-mediated communication from their SLPs. Despite this, they commented on their need for more support from the SLPs, which is borne out by previous research on limited professional support (Anderson et al., 2014; Bailey et al., 2006a; Crisp & Ellett, 2014; Stadskleiv, 2017). In the present study, the reported lack of support from SLPs might reflect the SLPs’ low use of instructional approaches. A considerable number of the participants reported that SLPs never used feedback and role play, which is problematic because communication partners need to practice the use of partner strategies and they need positive feedback to implement the strategy (Kent-Walsh & McNaughton, 2015; Dunst & Trivette, 2009).

Finally, caregivers not fluent in Swedish commented on their difficulties understanding the SLP. Research and regulations demand that AAC interventions targeting these caregivers are to be adjusted linguistically and culturally and that interpreters should be used (Pickl, 2011; Soto & Yu, 2014). This might also be an area in need of development.

Methodological considerations

Even though the results are based on a relatively small number of participants, they represent communication partners to 16 out of 39 (41%) of the potential number of children identified for the study throughout Sweden. The process of recruitment was complex. The contacts did not inform the caregivers of all the children potentially identified for the study about the study and did not ask them to participate. This is problematic since research ethics recommend that all potential participants should be informed about the study and given the choice to participate. However, if the number of children included in the study and their communication partners had been known in advance, another data collection method such as interview or observation would have been considered. The use of self-reporting to investigate participants’ own behaviors has limitations and the results should therefore be interpreted with caution. Self-reporting presumes advanced awareness and there is a risk of recollection bias. Nonetheless, the response rate to the questionnaire was 80%, which is considered high. To assure the reliability of the survey, professional terms such as aided augmented input and environmental arrangements were not used in the questionnaire. Instead, simple explanations were produced to illustrate each partner strategy. Face validity was examined in a pilot study; and the first author went through the questionnaire with the interpreter before data collection via telephone with non-Swedish-speaking participants. It is believed that, despite these limitations, this study adds valuable information about caregivers’, teachers’, and assistants’ use and learning of partner strategies in SGD-mediated communication with children with severe physical, communicative and cognitive impairments due to CP.

Conclusion and clinical implications

Communication partners’ learning about partner strategies in SGD-mediated communication with children with severe physical, communicative and cognitive impairments due to CP is inadequate and needs to be improved. SLPs, who are the major prescribers of SGDs and responsible for training and support in their use, should consider using more and different teaching approaches, such as feedback and role play, when teaching communication partners how to use partner strategies in communication when using an SGD.

Acknowledgments

We wish to thank the participating caregivers, teachers, and assistants for responding to the questionnaire and the SLPs for providing contact details and demographic data.

Additional information

Funding

The study was financed by the Sävstaholm Foundation and Health and Habilitation Services, Uppsala County Council, Norrbacka-Eugenia Foundation, the Foundation Promobilia, the Foundation Folke Bernadotte, the Gillberg Foundation and Uppsala County Council Research and Developmental funds.