2-3 Page Paper Needed Asap Needed in ****6 hours*** Re: Aphasias

Effect of choir activity in the rehabilitation of aphasia: a blind, randomised, controlled pilot study Anna Zumbansena,c,d,e, Isabelle Peretzb,c,d, Carole Angladea, Josée Bilodeaue, Suzanne Généreuxe, Michelyne Huberte and Sylvie Héberta,c,d,e

aSchool of Speech Pathology and Audiology, Faculty of Medicine, Université de Montréal, Quebec, Canada; bDepartment of Psychology, Faculty of arts and science, Université de Montréal, Quebec, Canada; cCRBLM, Centre for Research on Brain, Language and Music, McGill University, Quebec, Canada; dBRAMS, International Laboratory for Research on Brain, Music, and Sound, Université de Montréal, Quebec, Canada; eIUGM, Institut Universitaire de Gériatrie de Montréal, Quebec, Canada

ABSTRACT Background: Aphasia challenges the functional communication abilities that people use for everyday social interaction. People with non-fluent aphasia struggle to express themselves when speaking, yet they have been shown to better pronounce words when singing familiar songs or novel songs in a choral context. Choral singing and familiar songs are at the core of choir practice, a recreational activity that has long been offered in aphasic asso- ciations. Beneficial effects of this activity have been suggested for aphasia rehabilitation but have never been tested in a randomised controlled trial (RCT). Aim: Our study addressed whether the process of such a study is feasible and aimed to provide the first controlled trial of the effects of choir practice on aphasia rehabilitation. Methods: We piloted a three armed-prospective, randomised, par- allel-group, open-label, blinded end point (PROBE) pragmatic study with a 1:1:1 allocation ratio and stratification for aphasia severity. The primary outcome was improvement in functional communication. Secondary outcomes included various speech and language skills, mood and quality of life. Participants with chronic aphasia were recruited from five rehabilitation centres and three associations in the greater Montreal area from November 2011 to November 2012. Assessments were carried out by speech and language pathologists before and after a 6- month intervention period where participants had to attend either weekly choir sessions (experimental condition), drama classes (control condition) or neither of these (waiting list). The process feasibility of the design was explored based on rates of recruit- ment, exclusion, refusal, compliance and completion. Results: Twenty-two participants were recruited and randomised, corresponding to a recruitment rate of 1.8 participant/month. The following rates were measured: Exclusion 40%; Refusal 11%; Compliance 86%; Completion 77%. Changes from pre- to post- activity did not differ significantly between groups in any outcome measure although individual analyses showed various significant changes in different participants. Overall, a significant positive correlation was found between attendance to any social activity and functional communication improvements.

ARTICLE HISTORY Received 8 April 2016 Accepted 16 August 2016

KEYWORDS Aphasia; singing; functional communication; treatment; social activity

CONTACT Anna Zumbansen [email protected]

APHASIOLOGY, 2017 VOL. 31, NO. 8, 879–900 https://doi.org/10.1080/02687038.2016.1227424

© 2016 Informa UK Limited, trading as Taylor & Francis Group

Conclusion: A RCT to investigate the effectiveness of choral singing in people with aphasia is feasible. A large sample of people with aphasia is most probably necessary in order to test specific effects of social activities with sufficient statistical power and could be recruited in a large multisite trial. However, prior studies addres- sing more specifically the scientific (estimating treatment effect and its variance) resources and management feasibility are needed. For now, our results should encourage people with apha- sia to participate in social activities in general.

Introduction

Aphasia is an impairment of language function caused by brain damage and takes multiple forms depending on the severity of expression and comprehension deficits (Benson & Ardila, 1996). People with chronic aphasia have a lower quality of life compared with non-brain-injured adults (Ross & Wertz, 2003). Aphasia has a strong negative impact on language use in everyday functional communication, which leads to significant reduction in social participation (Dalemans, De Witte, Beurskens, Van Den Heuvel, & Wade, 2010; Darrigrand et al., 2011; Davidson, Howe, Worrall, Hickson, & Togher, 2008; Le Dorze & Brassard, 1995). Therefore, the primary aim of aphasia therapy is to improve functional communication (Brady, Kelly, Godwin, & Enderby, 2012).

Anecdotal reports of stroke patients with severe expressive aphasia state that these patients can sing words of familiar songs even if they cannot speak properly (Jacome, 1984; Yamadori, Osumi, Masuhar, & Okubo, 1977). Singing with lyrics can be regarded as an interesting bridge from relatively preserved musical production to impaired language expression in aphasia. Moreover, singing is a common activity, even in people with no musical training. Thus, a number of studies have investigated if singing could be successfully used in aphasia rehabilitation. One recent case study showed that the number of correctly reproduced words during singing was greater when compared with speaking, but this superiority effect was only found for familiar lyrics (Straube, Schulz, Geipel, Mentzel, & Miltner, 2008). When songs are novel to the patient, the singing advantage vanishes (Hébert, Racette, Gagnon, & Peretz, 2003; Peretz, Gagnon, Hébert, & Macoir, 2004; Racette, Bard, & Peretz, 2006; Straube et al., 2008). However, singing in synchrony with an auditory model (i.e., choral singing) helps improve the number of correct words produced by non-fluent aphasics (Racette et al., 2006). In fact, the beneficial effect of choral singing on speech production has been used as a facilitation technique in a range of speech and language music therapies for aphasia (Hurkmans et al., 2012; Zumbansen, Peretz, & Hébert, 2014b) such as the Melodic intonation therapy (MIT, Albert, Sparks, & Helm, 1973; Sparks, Helm, & Albert, 1974). This speech and language therapy approach has accumulated evidence of effectiveness in generalising the facilitation effect of choral singing to natural speech recovery (Zumbansen, Peretz, & Hébert, 2014a; Zumbansen et al., 2014b).

Speech and language therapy is traditionally given in patient–therapist dyadic ses- sions and recommended treatment intensity are well above what health care systems can offer, in particular in the chronic stage of post-stroke aphasia (Bhogal, Teasell, &

880 A. ZUMBANSEN ET AL.

Speechley, 2003; Code & Petheram, 2011). One cost-effective avenue to complement individual speech and language treatments is group interventions (Elman, 2007). When delivered by speech and language therapists, group interventions appear to be as effective as individual therapy on functional communication measures (Brady et al., 2012; Pulvermüller et al., 2001; Wertz et al., 1981; Yao, Xue, & Li, 2005). Other, non- therapeutically driven group activities that are commonly organised in associations of people with aphasia have been far less studied but might also have a therapeutic effect because they offer opportunities for social stimulation and support. Such activities are usually labelled as social support and stimulation activities and comprise various forms of support groups and recreational activities (Brady et al., 2012). Choir practice falls into this category and has been organised in associations of people with aphasia for many years (e.g., http://www.aphasie.ca; http://strokeachord.com/; the Cal State East Bay’s Aphasia Tones Choir). Choir activity not only offers opportunities for social stimulation and support but also has at its core the production of familiar and novel songs in speech- facilitating choral singing. Furthermore, listening to music has been shown more effec- tive than audiobooks in reducing negative mood in stroke patients (Särkämö et al., 2008). Therefore, choir activity might have greater effects in the rehabilitation process of people with aphasia than other social, non-musical activities.

Few studies have investigated the beneficial effects of choir practice in aphasia. Cohen (1992) explored the effect of singing familiar songs in group sessions of 30 min, three times a week for 3 weeks in eight participants with a variety of neurogenic expressive speech and language acquired disorders. She reported improvements in speech articulation and intelligibility but did not assess language or communication skills. In another study (Tamplin, Baker, Jones, Way, & Lee, 2013), quantitative and qualitative measures were collected from 13 people with aphasia before and after 20 weeks of a choir activity conducted 2 h per week by a music therapist. Quantitative analysis showed a pre/post trend towards a reduction in psychological distress while thematic analysis of semi-structured interviews with three participants and five caregivers highlighted the importance of peer support and revealed better con- fidence within social interactions, mood and motivation. Some interviewees perceived only speech improvements in singing with no carry-over effect to normal speech, while others reported generalised improvement of verbal output. Since these two studies (Cohen, 1992; Tamplin et al., 2013) did not include a control group, it is not possible to know if improvements were related to choral singing rather than social stimulation. Therefore, there is a need for research on the effects of the practice of choral singing for rehabilitation of aphasia using controlled studies.

Conducting randomised controlled trials (RCT) is particularly challenging and costly. As demonstrated by the small sample size of most aphasia treatment RCTs, recruitment of participants is not easy and high attrition has been highlighted as an issue in these studies (Brady et al., 2012). Moreover, the heterogeneity of clinical profiles in aphasia introduces considerable variability in the response to treatment (Howard, 1986). Because the effects of choir activity on aphasia rehabilitation have never been tested in a RCT, our study primarily addressed whether such a trial is feasible. In addition, we aimed to provide the first controlled data of the effects of choir practice on aphasia rehabilitation. Thus, we measured changes in functional communication but also in various speech and language skills, mood and quality of life.

APHASIOLOGY 881

Methods

Trial design

We piloted a three armed-prospective, randomised, parallel-group, open-label, blinded end point (PROBE) pragmatic study with a 1:1:1 allocation ratio and stratification for aphasia severity.

Study settings

Participants were recruited from five rehabilitation centres and three associations in the greater Montreal area from November 2011 to November 2012. All group activities took place in the premises of the three main associations. Six experienced speech-language pathologists assessed the study outcomes before and after the intervention phase, conducting testing sessions either in the premises of the participating associations, in their professional office or at participant’s home, at their convenience. To ensure better test–retest reliability, the same clinician did the pre- and post-intervention assessment for a given participant. Participants were assessed and started the 6-month intervention phase as and when they were enrolled. However, in order to avoid the 2-months summer break in the associations, participants who were recruited between February and August 2012 underwent their baseline assessment close to- and started their intervention phase in September 2012.

Participants’ eligibility criteria

Participants had to be francophone and have chronic aphasia (>1 year). The following exclusion criteria were applied: current participation in a speech and language treat- ment programme, participation in a choir or drama activity in the past 2 years, dementia or severe non-verbal cognitive deficit, uncorrected vision or severe hearing deficits preventing normal participation in social activities. Given the high frequency of hearing deficits with age, volunteers wearing hearing aids were asked to provide the fitting measurements of their hearing aids and a certified audiologist was consulted to assess participants’ hearing thresholds. We followed his clinical opinion on their ability to participate in the social activities of the study.

Interventions

Participants attended either choir sessions (experimental condition, Choir group), drama classes (control condition, Drama group) or neither (Waiting list) during 6-months (26 weeks). In the Choir group, participants were not allowed to also take drama classes and had to attend the weekly 2-hour long choir session. The same experienced choir leader directed the sessions. Participants assigned to the Drama group were not allowed to also attend choir sessions and had to follow the weekly 2-hour long drama session. Sessions were directed by two experienced drama teachers. Four participants of the Drama group attended sessions directed by the same teacher, while two (P15 and P17) were directed by another one. All activity directors had over 10 years of experience in

882 A. ZUMBANSEN ET AL.

leading groups of people with aphasia in their respective specialty. Finally, participants assigned to the waiting list could not be precluded from participating in other activities offered by the associations for ethical reasons. Attendance to any activity was recorded between the two evaluation times.

Trial outcomes and testing material

Primary outcome—functional communication The Test Lillois de Communication (TLC, Rousseaux, Delacourt, Wyrzykowski, & Lefeuvre, 2003) is a detailed questionnaire to be filled by speech and language pathologists based on their observation of patients’ verbal and non-verbal communication, as well as their attention to communication during the whole clinical examination and in a commu- nication task inspired from the therapy PACE (Promoting Aphasic Communicative Effectiveness, Davis & Wilcox, 1981; Wilcox & Davis, 1978). We used the total score of the Verbal communication scale (maximum score: 30 points), the total score of the Non- verbal communication scale (maximum score: 30 points) and the overall functional communication score, which is a weighted computation of the different subscales (maximum score: 100).

Secondary outcomes—speech and language skills Motor-speech agility—The total score of the Diadochokinetic rate subtest of the Apraxia battery for adults (ABA2, Dabul, 2000) was used to monitor changes in motor-speech agility. The task consists of rapid repetitions of syllable series increasing in complexity and the score reflects the number of correct repetitions in 3 s for the two-syllable series and in 5 s for the three-syllable series.

Automatised series—The production of the five automatised series of the MT86 (Nespoulous et al., 1992) was assessed by giving 3 points if the participant succeeded without help, 2 points if the participant succeeded with oral cueing, 1 point if the participant succeeded in unison with the examiner, and 0 if the participant failed. The maximum score was 15.

Repetition—The repetition score was based on the participant’s first attempt when repeating words and sentences in the repetition tasks of the MT86 and the ABA2 (subtests 2 and 5). We combined the stimuli from different tests in an attempt to improve the sensibility to change of the measure. Each of the 106 items was scored according to the rules of subtest 2 of the ABA2 by giving 2 points if the participant succeeded without hesitation, effort or articulatory error, 1 points if he self-corrected, delayed significantly or committed one or more articulatory errors but maintained the correct number of syllables and general conformation of the word, and 0 points if the participant produced no response, gave the wrong number of syllables or misarticulated to the extent that the word was no longer recognisable. The maximum score was 212 points.

Naming—The naming score was computed from the naming tasks of the MT86 and ABA2 (subtest 4), in an attempt to improve the sensibility to change of the measure with the number of items. Each of the 41 items was scored as specified by the MT86 manual by giving 1 point if the expected word was produced within 5 s (with or without phonetic or phonemic errors) and 0 points otherwise. The maximum score was 41 points.

APHASIOLOGY 883

Connected speech informativeness—The efficiency in conveying and transmitting correct information to the listener in connected speech was assessed with a test requiring the description of 15 complex line drawing pictures of several characters acting in daily situations (Zumbansen & Hébert, 2011; Zumbansen et al., 2014a). We divided the number of Correct Information Units (CIU) in a speech sample by the number of words in the sample. CIU are words that are intelligible in context and that accurately convey information related to the eliciting stimulus (Nicholas & Brookshire, 1993). Informativeness was scored with the help of the software Cordial Analyseur (Synapse-développement, 2010) for words counts. The connected speech informativeness score was the average of the scores on the 15 items (converted to percentage).

Comprehension—The language comprehension score was derived from the three auditory comprehension tasks of the MT86 (i.e., Compréhension orale de mots, Compréhension orale de phrases, Manipulation d’objets sur consignes verbales). As outlined in the MT86 manual, each of the 55 items was scored by rating the first response with 1 point if it was correct and 0 points otherwise. The maximum score was 55 points.

Secondary outcomes—mood Participants’ mood was assessed with the Visual Analogue Mood Scales (VAMS, Stern, 1997). On each scale, drawings of two faces are connected with a 10 cm vertical line. One face has a neutral expression while the other represents a mood state (afraid, confused, sad, angry, energetic, tired, happy or tense). Participants have to place a mark across the line at the point that describes how they feel. To avoid negative numbers, the negative mood T-scores were multiplied by −1 and we then added 1000 to all mood T-scores. The final mood score was the average of the eight subscales.

Secondary outcomes—quality of life Quality of life was measured with the short version of the Sickness Impact Profile (SIP), which has been validated for the francophone aphasic population (SIP65, Bénaim et al., 2003). It is organised in 11 subscales: fatigue and sleep, mood and mental state, physical state, activities at home and in the garden, activities outside the house, relationships with close ones, thinking, memory and concentration, ability to communicate, leisure activities, and diet. We summed points across all subscales and calculated a global percentage score of quality of life.

Sample size

The sample size was limited by the number of eligible participants that could be recruited over 1 year (November 2011–2012) in the greater Montreal area.

Randomisation

The first author enrolled participants and assigned them to one of the three study arms in a 1:1:1 ratio according to a computer-generated randomisation sequence in permuted blocks of six within aphasia severity strata (Mild–Moderate; Moderate; Severe) that were based on connected speech ratings of the MT86 (Table 3).

884 A. ZUMBANSEN ET AL.

Blinding

Clinicians who did the pre- and post-intervention assessment were blind to participant’s allocation. The first author, who randomised the participants, did not assess the study outcomes. To ensure assessors remained blinded to group allocation, all post-therapy assessment sessions were video recorded. These sessions were monitored for any disclosure of group allocation. Additionally, assessors were asked to declare if they thought they had been unblinded to group allocation during the outcome assessments.

Statistical methods

Baseline and group comparison data were analysed on an intention to treat basis. We did not exclude available data from participants even if they attended fewer activity sessions than others. Missing data from participants lost to follow-up were modelled using the method of the last observation carried forward. We performed non-parame- trical tests because of the small number of participants per group and the heterogeneity of their clinical profiles.

Preliminary analyses were run on outcome scores at baseline to ensure that groups were comparable. Individual changes for each outcome variable (post- minus pre- activity scores) were then calculated and Kruskal–Wallis H-tests were used to assess differential changes between groups in each outcome measure.

Additional analyses

Data were secondarily analysed as a multiple case study, hence individual data are presented. Non-parametrical tests were run to determine if changes were significant for each outcome variable in each participant (MacNemar or Wilcoxon test, depending on the nature of the variable). Whenever possible, the available test norms were used and a change above 1 standard deviation was considered significant (i.e., for the Functional communication and the Motor-speech ability scores).

In order to examine whether attendance to all social activities (including Choir, Drama and other activities) had an impact on the changes in functional communication, we also ran a Spearman’s rank–order correlation between these variables. Missing data were excluded for this analysis.

Feasibility analysis

We examined the recruitment, exclusion and refusal rates and the recruitment issues in the context of the study setting (involving five rehabilitation centres and three associa- tions in the greater Montreal area from November 2011 to November 2012). In addition, we measured the compliance and completion rates.

Ethical aspects

Participants gave written informed consent and the study was approved by the Ethics committees of the Centre for interdisciplinary research in rehabilitation of Greater

APHASIOLOGY 885

Montreal (CRIR), the rehabilitation hospital Villa-Medica and the Institut universitaire de gériatrie de Montréal (IUGM).

Results

Feasibility results

Participant flow, losses and exclusions Figure 1 displays the flow diagram of the study phases according to the Consolidated Standards of Reporting Trials (CONSORT, Schulz, Altman, & Moher, 2010). Of 37 volun- teers assessed for eligibility, 15 were excluded either because they did not meet the inclusion criteria (n = 10), they planned to move (n = 2) or they declined to participate (n = 2). Thus the exclusion rate was 40% (15/37) and the refusal rate 11% (3/27). Twenty- two eligible participants gave their consent and were randomly assigned to the three study arms in 1 year, which corresponds to a recruitment rate of 1.8 participants per month. The completion rate was 77% (17/22). Five participants were lost to follow-up: one participant from the Waiting list declined the post-activity testing. Two participants from the Drama group did not like the activity and two participants from the Choir group could not attend the sessions because the association’s new offices and choir room were not accessible to wheel chairs for a few months.

Recruitment process The recruitment was conducted in the greater Montreal area from November 2011 to November 2012. The first author presented the project to the speech and language therapists and clinical research coordinators (when they had one) in each of the five participating centres and organised conferences in the three associations of people with aphasia. Thirty-seven volunteers were assessed for eligibility during this year. Of these, seven were approached via the rehabilitation centres and 30 via the associations.

The low recruitment via the rehabilitation centres was partly due to the lack of, or inadequate resources dedicated to clinical research in three of the five centres. As a

Figure 1. Flow diagram of the study phases according to the consolidated standards of reporting trials (CONSORT, Schulz et al., 2010).

886 A. ZUMBANSEN ET AL.

result, no participant was recruited from these centres. The best collaborating centre screened their archives for francophone patients with chronic aphasia, sent 25 recruit- ment letters which were followed up with phone calls. Seven patients could not be reached, 1 patient turned out to be ineligible, 11 patients refused and 4 patients accepted to be contacted by the first author to be assessed for eligibility.

Compliance and attendance to social activities The compliance to the randomised allocation was 86% (19/22). One participant (on the Waiting list) did not strictly follow the randomised allocation and attended three choir sessions, and two participants (in the Drama group) dropped the activity because they did not enjoy it.

Participants in the Choir group attended a mean of 61% of their planned choir sessions. As indicated in Figure 1, two participants had to discontinue the choir activity because they could not access the choir room anymore. In the remaining participants, the compliance was high (P03: 20/22; P13: 14/23; P08: 20/22; P04: 18/23; P10: 21/24). Following the instructions for the study, none attended drama sessions. However, all but one also participated in other social activities, so that the choir sessions represented a mean of 48% of their overall social activities in the participating associations (P03: 38%; P13: 61%; P08: 50%; P04: 47%; P10: 34%; P12: 100%; P06: 4%).

Participants in the Drama group attended amean of 57% of their planned drama sessions (P22: 12/20; P21: 18/24; P15: 22/24; P01: 21/24; P17: 9/23; P02: 20/24; P14: 3/24; P11: 1/24). All but two (P22 and P17) also participated in other social activities, so that the drama sessions represented a mean of 49% of their overall social activities in the participating associations (P22: 100%; P21: 53%; P15: 29%; P01: 40%; P17: 100%; P02: 50%; P14: 14%; P11: 4%).

There was no statistically significant difference in overall attendance to all social activities (including choir, drama and other social activities) between the different groups (Table 6). Overall attendance to social activities was low (0–19) for six participants (P12, P22, P17, P18, P19, P07, P16, P09), medium (20–39) for five participants (P13, P04, P06, P21, P14, P11, P20, P05), high (40–59) for four partici- pants (P03, P08, P01, P02) and very high (60–79) for two participants (P10, P15). Only three participants of the Waiting list attended no social activity in the participating associations.

Trial results

Baseline data Characteristics of the 22 randomised participants are displayed in Tables 1–4. All were right-handed. Table 1 presents their demographic information. Seven participants (5 females) were in the Choir group, eight (3 females) in the Drama group and seven (5 females) on the Waiting list. There were no significant differences between the three study groups with regard to age, education, previous music or drama experience and post-aphasia onset time. All but one participant acquired aphasia following a stroke. P22, the youngest participant had a brain tumour surgically removed six and a half years before the beginning of the study.

As shown in Table 2, no participant scored below the cut-off at the Functional association of object drawings subtest of the visual agnosia battery (PEGV, Agniel,

APHASIOLOGY 887

oanette, Doyon, & Duchein, 1992). Other subtests of this battery, when failed, did not show a severe deficit. Thus, it was assumed that participants could be evaluated for other skills using visual material. They had various levels of abilities in planning accord- ing to the Total Move Score of the Tower of London-DX (Culbertson & Zillmer, 2001), and in short-term memory according to the Spatial Span of the WMS-III (Wechsler, 1997b). However, they all scored within the norms of their respective reference popula- tion on the Matrix reasoning test (Wechsler, 1997a) and were considered eligible for the study.

Aphasia was assessed following the diagnosis guidelines of Helm-Estabrooks and Albert (2004), with the francophone language evaluation battery for aphasia (MT86, Nespoulous et al., 1992) and a verbal fluency task (Cardebat, Doyon, Puel, Goulet, & Joanette, 1990). There was a wide range of aphasic clinical profiles, with various fluent and non-fluent aphasias, and mild to severe impairments (Table 3).

Table 4 displays participants’ musical abilities according to the Montreal Battery of Evaluation of Musical Abilities (MBEMA, Peretz et al., 2013). This version was chosen to limit the testing duration, although to date norms are only available for adults much younger than the participants of the present study. Thirteen participants scored below

Table 1. Participant’s demographics.

Group Participant Sex Age Education in years

Music experience in years

Drama experience in years

Time post- onset in years

Lesion etiology

Choir P03 F 73 7 0 0 1.3 Stroke P13 F 54 11 8 3 14.9 Stroke P08 M 65 9 0 0 4.7 Stroke P04 F 69 7 0 0 2.5 Stroke P10 M 59 11 0 0 2.9 Stroke P12 F 55 7 0 0 2.6 Stroke P06 F 69 12 0 6 9.5 Stroke Mean 63.4 9.1 1.1 1.3 5.5 SD 7.5 2.2 3.0 2.4 5.0

Drama P22 M 28 11 0 0 6.5 Brain tumour P21 F 35 9 0 0 4.9 Stroke P15 M 50 17 10 0 1.9 Stroke P01 M 82 6 0 0 1.3 Stroke P17 F 47 11 0 0 3.3 Stroke P02 F 82 9 0 0 12.5 Stroke P14 M 50 13 1 0 1.0 Stroke P11 M 58 11 0 0 5.4 Stroke Mean 54.0 10.9 1.4 0 4.6 SD 19.6 3.2 3.5 0 3.8

Waiting list P20 F 39 10 10 0 2.3 Stroke P18 F 46 11 0 2 22.0 Stroke P19 F 46 13 0 0 1.5 Stroke P05 M 69 10 0 0 1.1 Stroke P07 F 67 23 6 0 9.0 Stroke P16 M 50 19 0 0 18.3 Stroke P09 F 61 12 0 0 35.3 Stroke Mean 54.0 14.0 2.3 0.3 12.8 SD 11.6 5.0 4.1 0.8 13.0

Comparison between groups (p-values of the Kruskal–Wallis Test)

.222 .102 .804 .261 .647

888 A. ZUMBANSEN ET AL.

Ta bl e 2.

Pa rt ic ip an ts ’n

on -v er ba ls ki lls .

PE G V

W AI S- III

W M S- III

To w er

of Lo nd

on -D X

G ro up

Pa rt ic ip an t

Id en tic al ge om

et ric al

pa tt er ns /1 0

O ve rla pp

in g

fi gu

re s/ 36

Fu nc tio

na la

ss oc ia tio

n of

ob je ct

dr aw

in gs /1 0

Ca te go

ric al

as so ci at io n of

ob je ct

dr aw

in gs /1 0

M at rix

re as on

in g/ 26

Sp at ia l

sp an /3 2

To ta lm

ov e

sc or e

Ch oi r

P0 3

9 32

10 10

5 11

70 P1 3

10 35

10 8[ 9]

11 12

31 P0 8

10 32

10 7

7 12

65 P0 4

10 31

9 8[ 9]

7 13

80 (− 2. 5)

P1 0

7[ 8]

26 [3 0]

10 8[ 9]

6 10 [1 1]

64 P1 2

9 28 [3 0]

10 10

4 9[ 11 ]

99 (− 3. 8)

P0 6

10 35

10 10

14 15

64 M ea n

9. 29

31 .2 9

9. 86

8. 71

7. 71

11 .7 1

67 .5 7

SD 1. 11

3. 35

0. 38

1. 25

3. 56

1. 98

20 .4 9

D ra m a

P2 2

10 36

10 10

24 20

21 P2 1

10 36

10 10

17 16

21 P1 5

8 31

10 10

19 11

44 P0 1

6[ 8]

28 [3 0]

9 10

4 7

40 P1 7

10 31

10 9

12 8[ 11 ]

80 (− 2. 8)

P0 2

6[ 8]

30 10

9 4

11 83

(− 2. 7)

P1 4

10 33

10 9

11 15

32 P1 1

10 36

10 9

12 14

44 M ea n

8. 75

32 .6 2

9. 88

9. 50

12 .8 8

12 .7 5

45 .6 3

SD 1. 83

3. 11

0. 35

0. 54

6. 98

4. 33

23 .9 5

W ai tin

g lis t

P2 0

10 36

10 10

13 9[ 12 ]

93 (− 3. 9)

P1 8

10 34

10 9

12 8[ 11 ]

36 P1 9

10 36

10 10

18 10 [1 1]

52 P0 5

10 33

10 10

4 8[ 10 ]

79 (− 2. 5)

P0 7

7[ 8]

29 [3 0]

10 8[ 9]

6 9[ 10 ]

52 P1 6

10 35

10 10

19 15

11 P0 9

9 30

10 10

5 9[ 11 ]

29 M ea n

9. 43

33 .2 9

10 .0

9. 57

11 .0

9. 71

50 .2 9

SD 1. 13

2. 81

0. 00

0. 79

6. 16

2. 43

28 .4 5

Co m pa ris on

be tw ee n gr ou

ps (p -v al ue s of

th e

Kr us ka l– W al lis

Te st )

.8 06

.4 96

.6 08

.2 94

.4 44

.1 70

.2 94

Vi su al ag no

si a w as

as se ss ed

w ith

th e Pr ot oc ol e d’ év al ua tio

n de s gn

os ie s vi su el le s (P EG

V, Ag

ni el et

al ., 19 92 ), re as on

in g w ith

th e M at rix

re as on

in g su bt es t of

th e W ec hs le r ad ul t in te lli ge nc e

sc al e—

th ird

ed iti on

(W AI S- III ,W

ec hs le r, 19 97 a) ,s ho

rt -t er m

m em

or y w ith

th e Sp at ia ls pa n su bt es t of

th e W ec hs le r m em

or y sc al e—

th ird

ed iti on

(W M S- III ,W

ec hs le r, 19 97 b) ,a nd

pl an ni ng

sk ill s w ith

th e To w er

of Lo nd

on (T O L- D X,

Cu lb er ts on

& Zi llm

er ,2 00 1) .I f sc or es

ar e be lo w th e no

rm s (i. e. ,b

el ow

− 2S D or

be lo w cu t- off

sc or e) ,d

is ta nc e in

SD to

th e m ea n of

th e re fe re nc e

po pu

la tio

n (c or re sp on

di ng

to pa rt ic ip an ts ’ de m og

ra ph

ic s) is pr ov id ed

in pa re nt he se s ne xt

to th e pa rt ic ip an t’s

sc or e.

W he n on

ly cu t- off

sc or es

w er e av ai la bl e in

pu bl is he d m at er ia l, w e

in di ca te d th ei r va lu e in

sq ua re

br ac ke ts .

APHASIOLOGY 889

Ta bl e 3.

Pa rt ic ip an ts ’a ph

as ia pr ofi le s.

M T- 86

ap ha si a ba tt er y su bt es ts

M T- 86

Ra tin

gs of

co nn

ec te d sp ee ch

G ro up

Pa rt ic ip an t

Ve rb al

fl ue nc y te st

(a ni m al s in

90 s)

N am

in g/

31 Re pe tit io n/

30 Co

m pr eh en si on

/ 47

W or ds /

9 Se nt en ce s/

38

G lo ba l

re du

ct io n

of fl ue nc y

Ag ra m m at is m

Sy nt ac tic

de vi at io ns

W or d

fi nd

in g

di ffi cu lty

Ph on

et ic

de vi at io ns

Ph on

em ic

de vi at io ns

(a nd

/o r

ja rg on

) Ve rb al

de vi at io ns

M ot or -

sp ee ch

di ag no

si s

Ap ha si a

la be l

O ve ra ll

se ve rit y

Ch oi r

P o 3

14 (– 2. 2)

25 26

34 (– 2. 7)

8 26

+ N

+ +

+ +

N + +

+ +

N Tr an sc .

Se ns or y

M od

.

P 1 3

6( –3 .4 )

14 (– 16 .8 )

28 31 (– 7. 1)

9 22

+ +

+ +

+ + +

+ + +

+ +

+ N

M ix .

M od

. P 0

8 2( –4 .8 )

28 22 (– 3. 2)

44 9

25 + + +

+ + + +

+ +

+ + +

+ + +

+ D ys ar th ria

Tr an sc .

m ot or

Se ve re

P 0 4

8( –3 .3 )

28 28

35 (– 4. 9)

9 26

N N

+ + +

N + +

+ +

N Tr an sc .

Se ns or y

M od

.

P 1 0

1( –4 .5 )

1( –3 0. 2)

14 (– 13 .4 )

28 (– 8. 6)

9 19

+ + +

+ +

+ + +

+ + +

+ +

+ +

+ AO

S M ix .

Se ve re

P 1 2

4( –3 .0 )

14 (– 11 .9 )

27 32 (– 6. 4)

9 23

+ +

+ + +

+ + +

N +

+ N

M ix .

M od

. P 0

6 15 (– 2. 2)

26 (– 4. 4)

32 41 (– 2. 1)

9 32

N N

+ + +

N N

+ N

Tr an sc .

Se ns or y

M ild – M od

. D ra m a

P 2 2

15 25 (– 2. 7)

29 43

8 35

N N

+ + +

N N

+ N

M ix .

M ild – M od

. P 2

1 9( –2 .3 )

9( –1 7. 4)

14 (– 20 .3 )

33 (– 5. 3)

9 24

+ +

+ + +

+ + +

+ + +

+ +

+ +

AO S

M ix .

M od

. P 1

5 8( –3 .4 )

10 (– 20 .9 )

20 (– 8. 0)

43 8

35 + +

+ + +

+ + +

+ + +

+ + +

N AO

S Br oc a

M od

. P 0

1 16

25 18 (– 4. 6)

36 8

28 +

N N

+ +

+ + +

+ N

Co nd

uc tio

n M od

. P 1

7 1( –2 .5 )

3( –1 5. 9)

8( –3 2. 6)

25 (– 11 .1 )

8 17

+ + +

n/ a

n/ a

+ + +

+ + +

+ +

+ + +

AO S

G lo ba l

Se ve re

P 0 2

10 (– 3. 0)

23 (– 2. 5)

28 40

8 32

+ +

+ +

+ +

N +

+ +

N An

om ic

M od

. P 1

4 3( –4 .2 )

8( –2 3. 0)

9( –1 7. 9)

33 (– 6. 1)

8 25

+ + +

+ + +

+ + +

+ + +

+ + +

+ +

AO S

M ix .

Se ve re

P 1 1

8( –3 .4 )

28 (– 2. 3)

29 42

9 33

+ +

+ + +

+ +

N N

+ N

Tr an sc .

M ot or

M od

.

W ai tin

g lis t

P 2 0

5( –2 .2 )

11 (– 11 .1 )

20 (– 14 .5 )

35 (– 5. 6)

9 26

+ N

+ + +

+ + +

N + +

+ +

N W er ni ck e

M od

.

P 1 8

12 21 (– 5. 1)

28 (– 2. 3)

33 (– 6. 7)

8 25

+ +

+ + +

+ +

+ + +

+ +

AO S

M ix .

M od

. P 1

9 10

25 (– 2. 7)

17 (– 19 .0 )

39 (– 5. 3)

9 30

+ +

+ + +

+ + +

+ +

+ +

+ +

+ AO

S M ix .

M od

. P 0

5 18

29 26 (– 2. 6)

45 8

37 +

N +

+ N

+ N

N Co

nd uc tio

n M ild – M od

. P 0

7 4( –3 .6 )

12 (– 18 .8 )

10 (– 17 .0 )

33 (– 6. 1)

8 25

+ + +

+ + +

+ + +

+ + +

+ + +

+ +

AO S

M ix .

Se ve re

P 1 6

13 (– 2. 7)

13 (– 17 .8 )

23 (– 5. 3)

35 (– 5. 1)

8 27

+ +

+ + +

+ + +

+ + +

+ + +

+ +

AO S

M ix .

M od

.– se ve re

P 0 9

8( –3 .1 )

23 (– 7. 5)

26 (– 2. 6)

42 9

33 +

N +

+ N

+ +

N N

Co nd

uc tio

n M ild – M od

.

If sc or es

ar e be lo w

th e no

rm s (i. e. , be lo w

− 2S D

or be lo w

cu t- off

sc or e) , di st an ce

in SD

to th e m ea n of

th e re fe re nc e po

pu la tio

n (c or re sp on

di ng

to pa rt ic ip an ts ’ de m og

ra ph

ic s)

is pr ov id ed

in pa re nt he se s ne xt

to th e pa rt ic ip an t’s

sc or e. Ap

ha si a la be ls w er e de te rm

in ed

fo llo w in g th e di ag no

si s gu

id el in es

of H el m -E st ab ro ok s an d Al be rt (2 00 4) ,b

as ed

on th e pr es en te d da ta .N

:n or m al ;+

+ + :

se ve re ;A

O S: Ap

ra xi a of

sp ee ch ;M

ix .: M ix ed ;M

od .: M od

er at e; Tr an sc .: Tr an sc or tic al .

890 A. ZUMBANSEN ET AL.

the norms of this reference population. The Choir group tended towards lower musical abilities, but there was no significant difference between the groups according to the Kruskal–Wallis test (Table 4).

Table 5 displays statistics of baseline measurements of the trial outcomes. There was no statistically significant difference between groups at baseline for the primary outcome measure and most of the secondary outcome measures. However, there was a statistically significant difference between groups for Mood and Quality of life scores. Pairwise comparison tests (using the Dunn–Bonferroni approach) showed that the Mood score was significantly higher in the Waiting list than in the Choir group, and the Quality of life score was significantly higher in the Drama group than in the Choir group.

Trial outcomes Following the intention to treat principle, all 22 randomised participants were included in the between-group comparison of changes from pre- to post-activity. As shown in Table 6, changes between pre- to post-activity did not differ significantly between groups in any outcome measure.

Table 4. Participants’ musical abilities. Group Participant Pitch/20 Rhythm/20 Memory/20

Choir P03 12 (−2.6) 12 (−4.1) 9 (−6.4) P13 14 18 16 P08 11 (−3.1) 9 (−6.1) 12 (−4.3) P04 11 (−3.1) 11 (−4.7) 11 (−5.0) P10 13 (−2.1) 15 (−2.1) 11 (−5.0) P12 10 (−3.5) 15 (−2.1) 15 (−2.1) P06 16 20 15 (−2.1) Mean 12.4 14.3 12.7 SD 2.1 3.9 2.6

Drama P22 16 17 17 P21 18 16 20 P15 18 16 15 (−2.1) P01 9 (−4.1) 10 (−5.4) 9 (−6.4) P17 11 (−3.1) 8 (−6.7) 11 (−5.0) P02 11 (−3.1) 18 13 (−3.6) P14 9 (−4.1) 16 13 (−3.6) P11 13 (−2.1) 18 15 (−2.1) Mean 13.1 14.9 14.1 SD 3.8 3.8 3.4

Waiting list P20 15 14 (−2.7) 18 P18 11 (−3.1) 17 15 (−2.1) P19 14 15 (−2.1) 19 P05 10 (−3.6) 16 13 (−3.6) P07 14 16 18 P16 17 16 (−1.4) 11 (−5.0) P09 13 (−2.1) 15 (−2.1) 13 (−3.6) Mean 13.4 15.6 15.3 SD 2.4 1.0 3.1

Comparison between groups (p-values of the Kruskal–Wallis Test)

.786 .708 .345

Musical abilities were assessed with the Montreal battery of evaluation of musical abilities (MBEMA, Peretz et al., 2013). Norms of the abbreviated MBEMA are available for young adults (18–20 years). If scores are below the norms (i.e., below −2SD), distance in SD to the mean of the reference population is provided in parentheses next to the participant’s score.

APHASIOLOGY 891

Additional analyses Additional analyses were performed based on the 17 participants who completed the study (i.e, we removed the 5 participants lost to follow-up).

Individual analyses revealed variable significant changes in different participants (Table 6). Two participants (P03 and P20) improved in four of the outcome measures, notably in Functional communication and Connected speech informativeness. They had moderate receptive and fluent aphasias. P03 was in the Choir group and P20 was on the Waiting list. Both attended various social activities with medium to high frequency. Two other participants (P15 and P10) attended over 60 sessions of social activities in the 6-month activity period. They had expressive, non-fluent aphasias that were moderate in P15 and severe in P10. P15, who attended 77 group sessions, significantly improved in Quality of life. He did not improve in language skills but did so in non-verbal communication. P10, on the other hand, improved in speech and language expression skills as well as in non-verbal communication. Participants with the lowest attendance (P16 and P18) either did not improve significantly on any measure (P18), or showed significant negative changes (P16). Changes in two parti- cular cases (P08 and P22) do not seem to be related to attendance to social activities. Indeed, P08 attended 40 group sessions but showed a decrease in functional com- munication, while P22 attended only 12 sessions and had a significant improvement in Quality of life.

We found a statistically significant, positive correlation between Functional commu- nication changes and attendance to social activities (Figure 2, rs(15) = .51, p = .04).

Table 5. Baseline measurements of primary and secondary outcomes.

Choir group Drama group

Waiting List

Outcome variable M SD M SD M SD

Comparison between groups (p-values of the Kruskal–Wallis Test)

Functional communication

Global score 71. 9 8.8 78.1 13.6 77.9 6.1 .393 Verbal score 21.1 3.9 21.5 5.7 21.1 3.5 .668 Non-verbal score 19.4 6.0 23.6 6.0 23.1 2.0 .145

Motor-speech skills

Motor-speech ability 14.0 7.9 9.6 7.1 12.0 7.6 .511 Language expression

Automatised series 12.4 1.7 10.4 2.6 10.6 3.0 .243 Repetition 161.7 26.1 138.2 49.9 138.7 27.6 .411 Naming 26.9 12.9 22.4 13.0 27.1 8.9 .618 Connected speech

informativeness 28.3 9.9 30.6 15.0 32.3 11.0 .950

Language comprehension

Comprehension of words and sentences

39.9 6.7 42.0 7.5 42.7 6.8 .684

Mood Positive mood score 970.9 6.9 977.0 6.7 980.2 3.7 .032

Quality of life Positive quality of life score 49.4 7.5 68.3 14.3 63.8 9.5 .016

892 A. ZUMBANSEN ET AL.

Ta bl e 6.

Re su lts

of th e 22

ra nd

om is ed

pa rt ic ip an ts . Pr im ar y ou

tc om

e Se co nd

ar y ou

tc om

es

At te nd

an ce s

Fu nc tio

na lc om

m un

ic at io n

M ot or -

sp ee ch

sk ill s

La ng

ua ge

ex pr es si on

La ng

ua ge

co m pr eh en si on

M oo d

Q ua lit y of

lif e

G ro up

Pa rt ic ip an t

Ch oi r

se ss io ns

D ra m a

se ss io ns

Al ls oc ia l

ac tiv iti es

G lo ba ls co re

Ve rb al sc or e

N on

- ve rb al

sc or e

M ot or -

sp ee ch

ab ili ty

Au to m at is ed

se rie s

Re pi tit io n

N am

in g

Co nn

ec te d

sp ee ch

in fo rm

at iv en es s

Co m pr eh en si on

of w or ds

an d

se nt en ce s

Po si tiv e m oo d

sc or e

Po si tiv e

qu al ity

of lif e sc or e

Ch oi r

P o 3

20 0

53 58 –8 4*

19 –2 4*

12 –2 5*

20 –1 9

14 –1 2

15 5– 15 6

33 –3 7

28 –3 5*

38 –3 6

97 1– 96 5

58 –6 2

P 1 3

14 0

23 70 –8 0*

14 –1 9*

24 –2 8*

19 –1 2

12 –1 2

16 6– 17 4

22 –2 7

27 –3 3

36 –3 6

97 1– 96 5

53 –5 5

P 0 8

20 0

40 87 –7 9* (– )

23 –2 0* (– )

29 –2 8

6– 7

11 –1 3

13 0– 13 5

36 –3 6

35 –3 5

50 –4 7

96 8– 94 1

57 –4 9

P 0 4

18 0

38 69 –8 3*

25 –2 3* (– )

14 –2 5*

16 –1 5

15 –1 4

19 0– 18 3

36 –3 4

36 –3 6

40 –4 0

95 8– 95 7

45 –3 5

P 1 0

21 0

61 75 –7 8

20 –2 0

22 –2 4*

4– 13 *

11 –1 2

13 2– 13 1

1– 8*

8– 5

32 –2 7

97 5– 96 8

51 –5 5

P 1 2

4 0

4 68 –6 8

22 –2 2

16 –1 6

9– 9

11 –1 1

16 0– 16 0

24 –2 4

28 –2 8

35 –3 5

98 0– 98 0

43 –4 3

P 0 6

1 0

28 77 –7 7

25 –2 5

19 –1 9

24 –2 4

13 –1 3

19 8– 19 8

36 –3 5

37 –3 7

48 –4 8

97 4– 97 4

39 –3 9

M ea n

14 .0

0. 0

35 .3

6. 6

0. 7

4. 1

0. 1

0. 0

0. 7

2. 0

1. 4

–1 .4

–6 .7

–1 .1

SD 8. 2

0. 0

19 .1

11 .4

3. 2

5. 6

4. 7

1. 3

4. 8

3. 3

3. 6

2. 0

9. 4

5. 3

D ra m a

P 2 2

0 12

12 95 –9 6

26 –2 7

30 –3 0

23 –2 4

14 –1 3

19 8– 20 2

33 –3 7

49 –5 5

50 –5 0

97 7– 98 2*

86 –9 4*

P 2 1

0 18

34 82 –8 3

21 –2 3*

20 –2 5

9– 11

12 –1 1

12 3– 12 1

16 –2 0

30 –2 7

38 –3 3

98 5– 97 1

83 –8 6

P 1 5

0 22

77 86 –9 0

24 –2 4

26 –

28 .5 *

6– 8

10 –1 3

14 8– 14 9

16 –1 3

22 –2 2

46 –4 0

96 7– 96 4

62 –7 4*

P 0 1

0 21

52 58 –6 7*

25 –2 5

10 –1 6*

4– 0

8– 13

13 4– 12 2

33 –3 5

24 –2 2

43 –4 4

97 8– 97 4

63 –6 6

P 1 7

0 9

9 59 –5 8

9– 8

21 –2 1

7– 8

8– 6

68 –7 0

3– 5

11 –2 4*

28 –2 4

96 7– 95 9

40 –4 5

P 0 2

0 20

40 83 –8 3

23 –2 3

25 –2 5

18 –1 6

11 –1 4

18 2– 18 8

31 –3 2

18 –2 0

46 –4 3

98 2– 98 5*

75 –7 2

P 1 4

0 3

22 74 –7 4

18 –1 8

25 –2 5

3– 3

7– 7

70 –7 0

9– 9

40 –4 0

36 –3 6

98 0– 98 0

69 –6 9

P 1 1

0 1

24 89 –8 9

26 –2 6

26 –2 6

7– 7

13 –1 3

18 3– 18 3

38 –3 8

52 –5 2

49 –4 9

98 1– 98 1

68 –6 8

M ea n

0. 0

13 .3

33 .8

1. 8

0. 3

0. 9

0. 0

0. 9

–0 .1

1. 3

2. 2

–2 .1

–2 .5

3. 5

SD 0. 0

8. 3

22 .6

3. 4

0. 9

2. 3

2. 1

2. 5

5. 4

2. 3

5. 2

2. 7

6. 1

4. 9

W ai tin

g lis t

P 2 0

0 0

32 80 –8 6*

23 –2 4

23 –2 6*

23 –1 6

11 –9

12 2– 12 9

15 –1 7

19 –2 6*

39 –4 7*

97 7– 97 7

72 –7 2

P 1 8

0 0

0 85 –8 5

23 –2 3

26 –2 6

12 –6

10 –1 2

17 6– 17 1

29 –3 3

34 –3 9

36 –3 8

98 4– 98 2

69 –6 2

P 1 9

3 0

15 83 –8 2

24 –2 3

24 –2 4

2– 1

6– 9

10 9– 12 5*

35 –3 8

45 –4 6

47 –4 5

98 1– 98 3

58 –6 9

P 0 5

0 0

33 82 –8 4

23 –2 6*

24 –2 3

14 –2 2

15 –1 2

16 7– 16 6

39 –3 5

48 –4 8

53 –5 3

98 0– 98 1

74 –7 4

P 0 7

0 0

17 68 –7 0

16 –1 5

21 –2 3*

2– 2

8– 10

10 6– 10 1

20 –2 5

24 –2 0

36 –3 6

97 5– 97 9

52 –4 6

P 1 6

0 0

0 73 –6 8* (– )

16 –1 5

24 –2 2* (– )

16 –2 2

13 –1 2

14 1– 14 2

20 –2 2

29 –2 7

39 –4 0

98 1– 97 2* (– )

69 –6 9

P 0 9

0 0

0 76 –7 6

23 –2 3

20 –2 0

15 –1 5

11 –1 1

15 0– 15 0

32 –3 2

27 .2 7

49 –4 9

98 5– 98 5

52 –5 2

M ea n

0. 4

0. 0

13 .9

0. 6

0. 1

0. 3

0. 0

0. 1

1. 9

1. 7

1. 1

1. 3

–0 .5

–0 .4

SD 1. 1

0. 0

14 .7

3. 2

1. 5

1. 7

5. 6

2. 3

7. 5

3. 0

3. 9

3. 2

4. 1

5. 9

Co m pa ris on

be tw ee n

gr ou

ps (p -v al ue s of

th e Kr us ka l– W al lis

Te st )

< .0 01

< .0 01

.0 67

.5 01

.8 61

.3 32

.8 23

.8 23

.9 93

.8 82

.8 57

.0 82

.1 90

.2 11

Pr e-

an d po

st -o ut co m e m ea su re s ar e di sp la ye d se pa ra te d by

a hy ph

en .* in di ca te s a si gn

ifi ca nt

ch an ge

fr om

pr e-

to po

st -in

te rv en tio

n. (− )i nd

ic at es

w he n a si gn

ifi ca nt

ch an ge

w as

ne ga tiv e. M ea n

an d st an da rd

de vi at io n of

th e ch an ge s ar e pr ov id ed

fo r ea ch

gr ou

p.

APHASIOLOGY 893

Discussion

This study primarily addressed the feasibility of a RCT testing the effect of choir activity in aphasia rehabilitation. The intervention was feasible and the rates of recruitment (1.8 partici- pants/month in a metropolitan area such as Montreal), compliance (86%) and completion (77%) found in this study give elements that may be used to plan such a trial. The recruitment rate is not unusual in aphasia clinical trial and could havebeenhigherwith better collaboration and resources in the participating rehabilitation centres (only two of the five centres provided the study with participants). The completion rate (77%) could be improved by ensuring accessibility of the activity rooms to wheel chairs (2/5 participants lost to follow-up had to discontinue the activity for this reason). This study also provides the first controlled trial to investigate the effects of choir practice on aphasia rehabilitation. We found no specific effects at the group level on Functional communication improvements, our primary outcome, when comparing 6months of choir activity to drama classes and to aWaiting list. Similar resultswere found on secondary outcomes including speech and language skills, mood and quality of life. Individual analyses revealed variable significant changes in different participants. Thus, the benefits of choir practicemay not be specific to group singing butwe discuss below a number of reasons why the choir practice effect may not have emerged in this pilot study.

Limitations

Sample size It is very likely that this study was underpowered to detect any statistical significance with this small sample size. First, if social activities do help people with aphasia in their

Figure 2. Significant correlations between attendance to all social activities and improvement on the functional communication measure. Participants of the Choir, Drama and Waiting group are indicated with circles, triangles and diamonds, respectively.

894 A. ZUMBANSEN ET AL.

rehabilitation process, these effects are likely small. To our knowledge, only one RCT compared group language therapy to social activities in participants with chronic post- stroke aphasia (Elman & Bernstein-Ellis, 1999). The speech and language treatment included specific goals such as improvement of functional communication, confidence in communicative situations, self-awareness of personal goals and recognition of the progress made. In contrast, social activities were proposed to control for social stimulation. They were comprised of movement classes, creative/performance arts groups, church activities and support groups. The group communication treatment led to greater improvements on measures of language and communication compared with the control group. Although some of the participants improved while attending the social activities, no significant changes were found at the group level. This suggests that non-therapeutically driven group activities may have less therapeutic effect than a purposefully designed intervention.

In addition, multiple sources of variability may lower the effect size in our group compar- ison. As is often the case in group studies with aphasic populations, participants had a wide range of aphasia profiles, both in terms of aphasia types and severity. We stratified the randomisation for severity but not for aphasia types. It is still possible that people with certain forms of aphasia could benefit more from group singing than other social activities. The immediate facilitation of verbal output when singing familiar songs or in choral singing conditions has been shown in participants with non-fluent aphasias (Racette et al., 2006; Straube et al., 2008). Moreover, singing in Melodic Intonation Therapy has been shown to promote connected speech improvements in participants with Broca’s aphasia (Schlaug, Marchina, & Norton, 2008; Zumbansen et al., 2014a). Thus, group singing might have a specific beneficial effect for people with non-fluent aphasia. In our three groups, significant improvements in various outcome measures were found at the individual level, but there were not enough participants per group to identify which type of aphasia would benefit more from choir activity than other social activities. This could be examined in future studies with a greater number of participants randomised with a stratification for the aphasia type.

Finally, large sample size would also limit the influence of other sources of variability such as non-verbal cognitive skills (Helm-Estabrooks, 2002; Hinckley, Carr, & Patterson, 2001). For example, Hinckley et al. (2001) highlighted that executive functions are impor- tant to treatment response in aphasia therapy. They found that the lower the scores on executive function tests (Raven’s matrix reasoning and Wisconsin card sort test), the longer it takes patients to learn strategies that improve functional communication. Our partici- pants had a wide range of levels in non-verbal cognitive skills (Table 2), including executive functions (as tested by the Matrix reasoning and the Tower of London tests), that may influence the way they benefited from the social activities.

Study design We examined the feasibility of implementing a RCT on the effects of choir practice in the rehabilitation of people with aphasia. The following elements of the study design could be adapted in order to optimise the detection of pre–post effects of this social activity.

Our interventions may not have been provided at sufficient intensity to create the required change. It has been shown that even individual speech and language therapy must be intensive to achieve desired improvements (Bhogal, Teasell, Foley, & Speechley, 2003). Thus, it is possible that the intensity of the examined activities was not sufficient to elicit differential group effects. We tested the effect of group singing compared with

APHASIOLOGY 895

drama classes as usually practiced in associations of people with aphasia, that is, at a frequency of 2 h per week. Moreover, we included data of participants even if they did not attend all the planned sessions. In a study with non-musician, healthy subjects, Unwin, Kenny and Davis (2002) reported significant improvement in mood immediately after one single choir session. This improvement remained only partly 1 week later. Since we did not collect data at the beginning and the end of single activity sessions, our study does not show whether choir practice has an immediate effect on our outcome measures. Moreover, even if changes occurred at the single session level, they might not have lasted long enough until the next weekly practice, preventing a cumulative effect to operate. Future studies investigating the immediate and lasting effects of choir practice in people with aphasia could shed some light on the importance of frequency of this social activity.

One could also argue that the measures used in our study as well as the RCT by Elman and Bernstein-Ellis (1999) were not sufficiently sensitive to demonstrate change in performance. To our knowledge, there was no standardised assessment tool for franco- phone people with aphasia that had been tested for its ability to monitor changes at the time of our study. Such a tool is currently in development (Zumbansen, Frachon, Quiquempois, Hébert, & Thiel, 2016). In the present study, we attempted to improve the sensibility of the outcome measures by combining validated stimuli from different standardised diagnostic tests.

Data analysis We used intention to treat analysis for the main outcomes. This method is recom- mended to avoid overoptimistic estimates of the efficacy of an intervention but also goes with susceptibility to Type II error because non-compliers and participants lost to follow-up are not removed from the analysis (Gupta, 2011).

One participant from the Waiting list (P19) did not strictly follow the randomised allocation and attended three choir sessions. We do not think that this particular case of protocol violation could have significantly changed the results of the study since there was a statistically significant difference between groups with regard to choir attendance (Table 6).

In contrast, we believe that variability in attendance to allocated activities may have impacted the results. Subgroup analyses with participants reaching a specific compli- ance rate were not possible with the number of participants in our study. This approach may be justified as additional analyses in a larger trial given the probable importance of the intervention intensity.

The missing data of the five participants lost to follow-up were modelled with the method of the last observation carried forward. Because the last and only observation was done at baseline, this led to consider that 23% of our cohort (5/22) did not change in any outcome measure. We agree that it might have impacted our study results too. If such a percentage of participants lost to follow-up is to be expected in a RCT, the study design needs to be planned to strengthen the modelling of missing data, for example, with multiple measurements of the primary outcome during the intervention period (Streiner & Geddes, 2001).

896 A. ZUMBANSEN ET AL.

Generalisability and interpretation

The recruitment and completion rate obtained in our pilot study indicate that a RCT with a similar design is feasible with regards to its process but would require numerous study sites in distant geographical areas to reach an adequate sample size. Before engaging costs in such a multisite trial, other studies are necessary to estimate the treatment effect, its variance (scientific feasibility) and to assess resources and management issues. We believe that our feasibility results and considerations are generalisable to the study of other social activities and can help the continuing research on social interventions in people with aphasia.

Our additional analyses showed a significant correlation between the improve- ments in functional communication and attendance to a variety of social activities offered by the associations. This does not imply a causal relationship between the frequency of social activities and communication improvements but at least two, non- mutually exclusive interpretations can be discussed. First, it is likely that the more people with aphasia attend social activities, the more they stimulate their functional communication skills, thereby training and improving these abilities. Conversely, it is possible that the people who frequently participate in social groups are those who already have a good potential for improvement. However, taking into account other existing data on the benefits of social participation, it is reasonable to consider social activities as a valuable complement to speech and language therapy for aphasia rehabilitation (Cruice, Worrall, & Hickson, 2006; Vickers, 2010). Finally, two participants from the Drama group discontinued the intervention because they did not like this particular social activity. If there is actually no specific effect in any social activity compared with others, then any preferred social activity should be encouraged in order to foster frequent participation and thereby improvements in functional communication.

Registration

Due to its pilot nature, this study was not registered as a RCT.

Acknowledgements

Funding for this project was provided by the Centre for Research on Brain, Language and Music (CRBLM) and the Research Centre of the Montreal University Geriatric Institute (CRIUGM). This work was also supported by scholarships to Anna Zumbansen from the Collaborative Research and Training Experience (CREATE) Program in Auditory Cognitive Neuroscience from the Natural Sciences and Engineering Research Council of Canada (NSERC) and the Faculty of Graduate Studies of Université de Montréal.

We thank Christine Cantin (choir leader), Richard Gaulin and Isabelle Côté (drama teachers), Brigitte Damien and Marie-Andrée Laberge (speech-language pathologists), and Philippe Fournier and Charles-Édouard Basile (audiologists). We especially thank Alice Perdereau and Cynthia Keurentjes for their help in analysing the data and the associations Aphasie Rive-Sud, Association Québécoise des Personnes Aphasiques (AQPA) and Théâtre Aphasique for their long-lasting collaboration.

APHASIOLOGY 897

Disclosure statement

No potential conflict of interest was reported by the authors.

Funding

This work was supported by the Centre for Research on Brain, Language and Music [R0020225]; CRIUGM—Centre de Recherche de l’Institut Universitaire de Gériatrie de Montréal [CAREC pilote 2010-11].

References

Agniel, A., Joanette, Y., Doyon, B., & Duchein, C. (1992). Protocole d’évaluation des gnosies visuelles (PEGV). Isbergues: L’Ortho Édition.

Albert, M. L., Sparks, R. W., & Helm, N. A. (1973). Melodic intonation therapy for aphasia. Archives of Neurology, 29, 130–131. doi:10.1001/archneur.1973.00490260074018

Bénaim, C., Pélissier, J., Petiot, S., Bareil, M., Ferrat, E., Royer, E., . . . Hérisson, C. (2003). Un outil francophone de mesure de la qualité de vie de l’aphasique: Le SIP-65: A French questionnaire to assess quality of life of the aphasic patient: the SIP-65. Annales de réadaptation et de médecine physique, 46, 2–11. doi:10.1016/S0168-6054(02)00306-9

Benson, D. F., & Ardila, A. (1996). Aphasia: A clinical perspective. New York, NY: Oxford University Press on Demand.

Bhogal, S. K., Teasell, R., & Speechley, M. (2003). Intensity of aphasia therapy, impact on recovery. Stroke, 34, 987–993. doi:10.1161/01.STR.0000062343.64383.D0

Bhogal, S. K., Teasell, R. W., Foley, N. C., & Speechley, M. R. (2003). Rehabilitation of aphasia: More is better. Topics in Stroke Rehabilitation, 10, 66–76. doi:10.1310/RCM8-5TUL-NC5D-BX58

Brady, M. C., Kelly, H., Godwin, J., & Enderby, P. (2012). Speech and language therapy for aphasia following stroke. Cochrane Database of Systematic Reviews, 5, CD000425. doi:10.1002/14651858. CD000425.pub3

Cardebat, D., Doyon, B., Puel, M., Goulet, P., & Joanette, Y. (1990). Evocation lexicale formelle et sémantique chez des sujets normaux. Performances et dynamiques de production en fonction du sexe, de l’âge et du niveau d’étude. Acta Neurologica Belgica, 90, 207–217.

Code, C., & Petheram, B. (2011). Delivering for aphasia. International Journal of Speech-Language Pathology, 13, 3–10. doi:10.3109/17549507.2010.520090

Cohen, N. S. (1992). The effect of singing instruction on the speech production of neurologically impaired persons. Journal of Music Therapy, 29, 87–102. doi:10.1093/jmt/29.2.102

Cruice, M., Worrall, L., & Hickson, L. (2006). Quantifying aphasic people’s social lives in the context of non-aphasic peers. Aphasiology, 20, 1210–1225. doi:10.1080/02687030600790136

Culbertson, C. W., & Zillmer, E. A. (2001). Tower of London Drexel University (TOL DX): Technical manual. Toronto: Multi-Health Systems Incorporated (MHS).

Dabul, B. L. (2000). Apraxia battery for adults (2nd ed.). ABA2. Austin, TX: Pro-ed. Dalemans, R. J. P., De Witte, L. P., Beurskens, A. J. H. M., Van Den Heuvel, W. J. A., & Wade, D. T.

(2010). An investigation into the social participation of stroke survivors with aphasia. Disability & Rehabilitation, 32, 1678–1685. doi:10.3109/09638281003649938

Darrigrand, B., Dutheil, S., Michelet, V., Rereau, S., Rousseaux, M., & Mazaux, J.-M. (2011). Communication impairment and activity limitation in stroke patients with severe aphasia. Disability and Rehabilitation, 33, 1169–1178. doi:10.3109/09638288.2010.524271

Davidson, B., Howe, T., Worrall, L., Hickson, L., & Togher, L. (2008). Social participation for older people with aphasia: The impact of communication disability on friendships. Topics in Stroke Rehabilitation, 15, 325–340. doi:10.1310/tsr1504-325

898 A. ZUMBANSEN ET AL.

Davis, G. A., & Wilcox, M. J. (1981). Incorporating parameters of natural conversation in aphasia treatment. In R. Chapey (Ed.), Language intervention strategies in adult aphasia (pp. 169–193). Baltimore: William & Wilkins.

Elman, R. J. (2007). Group treatment of neurogenic communication disorders: The expert clinician’s approach. San Diego, CA: Plural Pub.

Elman, R. J., & Bernstein-Ellis, E. (1999). The efficacy of group communication treatment in adults with chronic aphasia. Journal of Speech, Language, and Hearing Research, 42, 411–419. doi:10.1044/jslhr.4202.411

Gupta, S. K. (2011). Intention-to-treat concept: A review. Perspectives in Clinical Research, 2, 109. doi:10.4103/2229-3485.83221

Hébert, S., Racette, A., Gagnon, L., & Peretz, I. (2003). Revisiting the dissociation between singing and speaking in expressive aphasia. Brain, 126, 1838–1850. doi:10.1093/brain/awg186

Helm-Estabrooks, N. A. (2002). Cognition and aphasia: A discussion and a study. Journal of Communication Disorders, 35, 171–186. doi:10.1016/S0021-9924(02)00063-1

Helm-Estabrooks, N. A., & Albert, M. L. (2004). Manual of aphasia and aphasia therapy. Austin, TX: Pro-ed.

Hinckley, J. J., Carr, T. H., & Patterson, J. P. (2001). Relationships between cognitive abilities, treatment type and treatment time in aphasia. Paper presented at the 31st Annual Clinical Aphasiology Conference, Santa Fe, NM.

Howard, D. (1986). Forum: Evaluating intervention beyond randomised controlled trials: The case for effective case studies of the effects of treatment in aphasia. British Journal of Disorders of Communication, 21, 89–102. doi:10.3109/13682828609018546

Hurkmans, J., De Bruijn, M., Boonstra, A. M., Jonkers, R., Bastiaanse, R., Arendzen, H., & Reinders- Messelink, H. A. (2012). Music in the treatment of neurological language and speech disorders: A systematic review. Aphasiology, 26, 1–19. doi:10.1080/02687038.2011.602514

Jacome, D. E. (1984). Aphasia with elation, hypermusia, musicophilia and compulsive whistling. Journal of Neurology, Neurosurgery & Psychiatry, 47, 308–310. doi:10.1136/jnnp.47.3.308

Le Dorze, G., & Brassard, C. (1995). A description of the consequences of aphasia on aphasic persons and their relatives and friends, based on the WHO model of chronic diseases. Aphasiology, 9, 239–255. doi:10.1080/02687039508248198

Nespoulous, J.-L., Lecours, A. R., Lafond, D., Lemay, A., Puel, M., Cot, F., . . . Giroux, F. (1992). Protocole Montréal-Toulouse d’examen linguistique de l’aphasie MT-86 M1béta Module standard initial (2nd ed.). Isbergues: Ortho-Edition.

Nicholas, L. E., & Brookshire, R. H. (1993). A system for quantifying the informativeness and efficiency of the connected speech of adults with aphasia. Journal of Speech Language and Hearing Research, 36, 338–350. doi:10.1044/jshr.3602.338

Peretz, I., Gagnon, L., Hébert, S., & Macoir, J. (2004). Singing in the brain: Insights from cognitive neuropsychology. Music Perception, 21, 373–390. doi:10.1525/mp.2004.21.3.373

Peretz, I., Gosselin, N., Nan, Y., Caron-Caplette, E., Trehub, S. E., & Béland, R. (2013). A novel tool for evaluating children’s musical abilities across age and culture. Frontiers in Systems Neuroscience, 7. doi:10.3389/fnsys.2013.00030

Pulvermüller, F., Neininger, B., Elbert, T., Mohr, B., Rockstroh, B., Koebbel, P., & Taub, E. (2001). Constraint-induced therapy of chronic aphasia after stroke. Stroke, 32, 1621–1626. doi:10.1161/ 01.STR.32.7.1621

Racette, A., Bard, C., & Peretz, I. (2006). Making non-fluent aphasics speak: Sing along! Brain, 129, 2571–2584. doi:10.1093/brain/awl250

Ross, K. B., & Wertz, R. T. (2003). Quality of life with and without aphasia. Aphasiology, 17, 355–364. doi:10.1080/02687030244000716

Rousseaux, M., Delacourt, A., Wyrzykowski, N., & Lefeuvre, M. (2003). Test Lillois de Communication (TLC). Isbergues: L’Ortho-Édition.

Särkämö, T., Tervaniemi, M., Laitinen, S., Forsblom, A., Soinila, S., Mikkonen, M., . . . Hietanen, M. (2008). Music listening enhances cognitive recovery and mood after middle cerebral artery stroke. Brain, 131, 866–876. doi:10.1093/brain/awn013

APHASIOLOGY 899

Schlaug, G., Marchina, S., & Norton, A. (2008). From singing to speaking: Why singing may lead to recovery of expressive language function in patients with Broca’s aphasia. Music Perception, 25, 315–323. doi:10.1525/mp.2008.25.4.315

Schulz, K. F., Altman, D. G., & Moher, D. (2010). CONSORT 2010 statement: Updated guidelines for reporting parallel group randomised trials. BMJ, 340, c332. doi:10.1136/bmj.c332

Sparks, R. W., Helm, N. A., & Albert, M. L. (1974). Aphasia rehabilitation resulting from melodic intonation therapy. Cortex, 10, 303–316. doi:10.1016/S0010-9452(74)80024-9

Stern, R. A. (1997). Visual analog mood scales, VAMS. Odessa, FL: Psychological Assessment Resources.

Straube, T., Schulz, A., Geipel, K., Mentzel, H.-J., & Miltner, W. H. R.. (2008). Dissociation between singing and speaking in expressive aphasia: The role of song familiarity. Neuropsychologia, 46, 1505–1512. doi:10.1016/j.neuropsychologia.2008.01.008

Streiner, D., & Geddes, J. (2001). Intention to treat analysis in clinical trials when there are missing data. Evidence Based Mental Health, 4, 70–71. doi:10.1136/ebmh.4.3.70

Synapse-développement. (2010). Cordial Analyseur. Toulouse, France: Synapse développement. Tamplin, J., Baker, F. A., Jones, B., Way, A., & Lee, S. (2013). ‘Stroke a Chord’: The effect of singing in

a community choir on mood and social engagement for people living with aphasia following a stroke. NeuroRehabilitation, 32, 929–941. doi:10.3233/NRE-130916

Unwin, M. M., Kenny, D. T., & Davis, P. J. (2002). The effects of group singing on mood. Psychology of Music, 30, 175–185. doi:10.1177/0305735602302004

Vickers, C. P. (2010). Social networks after the onset of aphasia: The impact of aphasia group attendance. Aphasiology, 24, 902–913. doi:10.1080/02687030903438532

Wechsler, D. (1997a). Wechsler adult intelligence scale—Third edition. San Antonio, TX: Harcourt Assessment.

Wechsler, D. (1997b). Wechsler memory scale–Third edition. San Antonio, TX: Psychological Corporation.

Wertz, R. T., Collins, M. J., Weiss, D., Kurtzke, J. F., Friden, T., Brookshire, R. H., . . . Porch, B. E. (1981). Veterans administration cooperative study on aphasia: A comparison of individual and group treatment. Journal of Speech, Language, and Hearing Research, 24, 580–594. doi:10.1044/ jshr.2404.580

Wilcox, M. J., & Davis, G. A. (1978). Procedures for promoting communicative effectiveness in an aphasic adult. Paper presented at the Miniseminar presented at the Annual Convention of American Speech and Hearing Association, San Francisco.

Yamadori, A., Osumi, Y., Masuhar, S., & Okubo, M. (1977). Preservation of singing in Broca’s aphasia. Journal of Neurology, Neurosurgery & Psychiatry, 40, 221–224. doi:10.1136/jnnp.40.3.221

Yao, J., Xue, Y., & Li, F. (2005). Clinical application research on collective language strengthened training in rehabilitation nursing of cerebral apoplexy patients with aphasia. Chinese Nursing Research, 19, 482–484.

Zumbansen, A., Frachon, J., Quiquempois, D., Hébert, S., & Thiel, A. (2016). Abstract TP148: The Montreal assessment of connected speech offers good psychometric properties to monitor ecological language recovery in post-stroke aphasia. Stroke, 47, ATP148–ATP148.

Zumbansen, A., & Hébert, S. (2011). Quantifying natural speech in people with aphasia. Paper presented at the CRSNG-CREATE annual workshop on Auditory Cognitive Neuroscience, Montréal, QC.

Zumbansen, A., Peretz, I., & Hébert, S. (2014a). The combination of rhythm and pitch can account for the beneficial effect of Melodic intonation therapy on connected speech improvements in Broca’s aphasia. Frontiers in Human Neuroscience, 8. doi:10.3389/fnhum.2014.00592

Zumbansen, A., Peretz, I., & Hébert, S. (2014b). Melodic intonation therapy: Back to basics for future research. Frontiers in Neurology, 5. doi:10.3389/fneur.2014.00007

900 A. ZUMBANSEN ET AL.

Copyright of Aphasiology is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.