Brief Communication

Comparing Interventions for Selective Mutism:

A Pilot Study

Katharina Manassis, MD, FRCPC1; Rosemary Tannock, PhD2

Objective: To examine the outcome within 6 to 8 months of medical and nonmedical

intervention for children with severe selective mutism (SM).
Method: Children with SM (n = 17) and their mothers, seen in a previous study, attended
follow-up appointments with a clinician. Obtained by maternal report were: treatment
received, current diagnosis (based on semi-structured interview), speech in various
environments, and global improvement. An independent clinician also rated global
functioning.
Results: The diagnosis of SM persisted in 16 children, but significant symptomatic
improvement was evident in the sample. All children had received school consultations.
Children who had been treated with selective serotonin reuptake inhibitors (SSRI) (n = 10)
showed greater global improvement, improvement in functioning, and improvement in
speech outside the family than children who were unmedicated (n = 7). No differences
were evident for children receiving and not receiving additional nonmedical intervention.
Conclusions: The findings suggest the potential benefit of SSRI treatment in severe SM,
but randomized comparative treatment studies are indicated.
Can J Psychiatry 2008;53(10):700–703

Clinical Implications

· A persistent condition, SM merits intervention as significant symptomatic

improvement can occur.
· For children with severe SM, SSRIs may be beneficial.
· School consultation may also be helpful in reducing symptoms of SM in affected
children.

Limitations

· As the study was not randomized, there might have been economic, educational, or
severity-related selection biases among families electing to have various
interventions.
· The small sample size and lack of a nonintervention control group limit our ability to
draw definite conclusions.
· Improvement ratings by teacher report (in addition to mothers and clinicians) would
have strengthened the findings.

Key Words: selective mutism, anxiety disorder, child psychiatry, pharmacotherapy,

selective serotonin reuptake inhibitor

700 W La Revue canadienne de psychiatrie, vol 53, no 10, octobre 2008


espite using normal speech (usually at home), children
D with SM fail to speak in selective environments (often at
school).1 Treatment of this potentially debilitating condition
has received limited research attention, partly owing to its rar-
ity (prevalence is estimated at 0.7%2). Findings suggest that
SM may be related to social phobia,3 with anxiolytic medica-
tions and psychotherapy having been examined. Medication
studies included a randomized controlled trial of fluoxetine4
with some open trials of other serotonergic medications, and
numerous case reports. Follow-up intervals were relatively
brief (9 to 12 weeks), and it is unclear whether samples in
these trials were representative of children with SM in the
general population. Psychotherapy studies consist mostly of
case reports, with only one small controlled trial of behav-
ioural intervention5 and no gold standard therapy in the field.
No studies have contrasted outcomes for different treatments.
Given the high persistence rates of SM and associated mor-
bidity,6 comparative trials are needed to guide clinicians’
interventions and improve the chances of returning such chil-
dren to a normative developmental course. This pilot study is a
first step toward such trials.
Methods
In our study, 20 children with SM and their parents, partici-
pants in a study of cognitive and linguistic abilities in SM at a
children’s hospital, were invited to return 6 to 8 months after
their initial research assessment.7 When assessed initially, all
children had been mute for at least one full school year. All
procedures were approved by the Research Ethics Board of
the Hospital for Sick Children in Toronto, and parents pro-
vided written informed consent. Verbal assent was obtained
from children whenever possible.
At follow-up, treatment received was determined and out-
comes were measured using brief parent-report question-
naires (SMQ8 and CGI9) and a semi-structured parent
interview (Anxiety Disorders Interview Schedule10) adminis-
tered by a trained child psychiatrist blind to treatment status.
This psychiatrist also completed the CGAS11 based on the
interview and clinical notes on current functioning for each
child. The SMQ ascertains quantity of speech in 3 environ-
ments: family, school, and other social situations, and has
been used in numerous research studies of SM. In this study,
Abbreviations used in this article
CGAS Children’s Global Assessment Scale
CGI Clinical Global Improvement Rating
SM selective mutism
SMQ Selective Mutism Questionnaire
SSRI selective serotonin reuptake inhibitor
The Canadian Journal of Psychiatry, Vol 53, No 10, October 2008 W
Comparing Interventions for Selective Mutism: A Pilot Study
both the total score and the sum of the 2 nonfamilial environ-
ment scores were related to treatment, as children with SM
often speak normally at home. Treatment was determined
largely by parental preference, with no attempt to assign chil-
dren to particular treatments apart from providing school
consultations to all participants. Monitoring of treatment was
left to clinicians’ judgment, as would be the case in commu-
nity settings.
Results
Among the original 20 subjects seen, 17 returned for a 6- to
8-month follow-up appointment (85%) with at least one par-
ent. All mothers attended, so results are by maternal report.
Nonparticipants did not differ from participants in demo-
graphic characteristics or initial CGAS or SMQ scores, sug-
gesting the follow-up sample was representative of the
original. Similarly, demographic characteristics, initial
CGAS, and initial SMQ did not differ for medicated, com-
pared with unmedicated children, nor for those who did,
compared with those who did not receive psychotherapy. The
17 participants had a mean age of 7.83 (SD 1.28) years at ini-
tial assessment (that is, 8.33 years at follow-up). Twelve were
female and 5 were male, showing a trend towards female pre-
ponderance (÷2 = 2.88, P = 0.09). All but one initially met cri-
teria for both SM and social phobia on parental interview
(Anxiety Disorders Interview Schedule). The remaining
child only met criteria for SM.
Diagnoses remained stable over 6 to 8 months, with the
exception of one child who no longer met criteria for SM at
follow-up.
Significant changes in CGAS or SMQ over 6 to 8 months
were examined for all 17 children using paired sample t tests
(Table 1). Significant improvements were found in the
CGAS (t = 3.49, P = 0.003), the total SMQ (t = 3.45, P =
0.004), and the SMQ school and other scales (t = 3.40, P =
0.004). Significance was maintained when applying the
Bonferroni correction for multiple t tests. Global improve-
ment ratings by mothers averaged 3 for the sample, corre-
sponding to improved.
All children’s schools received monthly telephone consulta-
tions from an experienced member of the clinical research
team. All children were initially medication-free, but were
offered treatment with an SSRI at time of assessment. Ten of
the 17 families agreed to such treatment. Eight children
received a liquid preparation of fluoxetine, and 2 received
sertraline. Dosage was titrated upwards gradually depending
on response, with final dosages of 10 to 25 mg of fluoxetine
and 25 to 50 mg sertraline. Three families admitted providing
the medication inconsistently to their child (stopping it with-
out the physician’s knowledge in one case), though no formal
measures of compliance were done. Insomnia was reported
701
Brief Communication

Table 1 Mean and SDs in relation to intervention

Samplea Medicated Unmedicated Any therapy No therapy
Measure (n = 17) (n = 10) (n = 7) (n = 10) (n = 7)

CGAS (initial) 47.94 (5.40) 46.20 (4.94) 50.43 (5.38) 49.20 (6.23) 46.14 (3.62)
CGAS (6 months) 57.82(8.38)b 62.00 (8.14)b 51.86 (4.18) 54.80 (7.25) 62.14 (8.45)
Total SMQ (initial) 6.25 (1.24) 5.95 (1.36) 6.74 (0.89) 6.65 (1.09) 5.57 (1.27)
Total SMQ (6 to 8 months) 7.43 (1.51)b 7.67 (1.70) 7.09 (1.25) 7.57 (1.51) 7.24 (1.65)
SMQ school and other 3.08 (0.81) 2.80 (0.74) 3.60 (0.71) 3.43 (0.82) 2.47 (0.19)
(initial)
SMQ school and other 4.06 (1.13)b 4.24 (1.25)c 3.79 (0.97) 4.17 (1.22) 3.90 (1.08)
(6 to 8 months)
CGI 2.94 (1.08) 2.50 (0.97)c 3.57 (0.98) 3.10 (0.99) 2.71 (1.25)
a
Significance indicated for whole-sample changes from baseline to 6 to 8 months.
b
P < 0.01; cP < 0.05

in one child on fluoxetine, but responded to dosage reduction. Discussion

No other medication side effects were reported.
Psychotherapy or speech therapy was not offered at assess-
ment owing to resource constraints, but 10 families obtained
such therapy privately for their children (5 with a speech ther-
apist; 5 with a psychologist). The average number of sessions
was 9.20 (SD 2.93). Overall, there were 4 children who
received both interventions, 6 children who received medica-
tion only, 6 children who received therapy only, and one child
who received neither. A medication ´ therapy chi-square was
not significant (that is, children were not more or less likely to
receive one modality based on receiving the other).
Repeated measures ANOVAs were done to determine any 6-
to 8-month outcome differences in children receiving medical
or nonmedical intervention (Table 1). Owing to sample size
constraints, specific medications and specific types of therapy
could not be examined separately. As mentioned, there were
significant improvements with time in the CGAS, total SMQ,
and sum of the SMQ school and other subscales. There were
also significant medication ´ time interactions for the CGAS
(F = 9.58; df = 1,14; P = 0.007) and the SMQ school and other
scales (F = 6.24; df = 1,14; P = 0.028), with medicated chil-
dren showing greater improvements than unmedicated chil-
dren. There was no such interaction for the total SMQ. There
were no significant therapy ´ time interactions, though there
was a trend toward greater CGAS improvement for children
that did not receive therapy (F = 3.85; df = 1,14; P = 0.07).
Maternal Global Improvement ratings were compared by
intervention modality using t tests. Medicated children
showed significantly greater improvement than unmedicated
children (t = 2.23, P = 0.04). There was no significant differ-
ence related to nonmedical intervention.
In this clinical sample of children with severe SM, improve-
ments were evident by clinician and maternal report after 6 to
8 months, but most children still met criteria for SM. This is
consistent with previous reports of high persistence rates in
clinical samples.6
Choice of interventions might have been influenced by selec-
tion biases including financial circumstances, parental edu-
cation, and child severity. Nevertheless, by maternal report,
using serotonergic medications increased the child’s degree
of overall improvement and reduced mutism outside the
home. Clinicians also rated medically treated children as
showing higher functional gains than nonmedicated chil-
dren. Previous efficacy trials of fluoxetine in SM showed
similar outcomes.4,12 Interestingly, medication-related gains
occurred despite noncompliance in some cases and a lack of
the rigorous medication monitoring typical of efficacy trials.
This finding suggests potential effectiveness for SSRIs in
treating SM in real-world environments. The follow-up inter-
val (6 to 8 months) is also more typical in community use of
SSRIs than in short-term efficacy studies. Further medication
effectiveness studies are indicated, though, especially as chil-
dren did not all receive the same SSRI, and SM may respond
differently to different SSRIs. Because all children in this
sample received school consultations regarding the manage-
ment of SM, we also cannot conclude that improvement was
due to medication alone. The possible benefits of such
consultations, alone or in combination with medication,
warrant further study.
Psychotherapy and speech therapy did not appear to affect
outcomes, except for a trend toward greater CGAS improve-
ments in the absence of therapy. The heterogeneity of thera-
pies received by the children makes this finding difficult to

702 W La Revue canadienne de psychiatrie, vol 53, no 10, octobre 2008

 Comparing Interventions for Selective Mutism: A Pilot Study

interpret. For example, it is possible that highly trained thera- References

pists using a manualized treatment might have obtained more
encouraging results. The diversity of nonmedical treatment
approaches received, however, speaks to the lack of a gold
standard therapy for SM and the need to further develop and
evaluate potentially useful therapies. The sample was too
small to determine if combining medication and therapy
would be more helpful than doing either alone.
Conclusions
Despite the significant limitations of this preliminary, nonran-
domized study and recent controversy surrounding the use of
SSRIs in children,13 the findings suggest a potential benefit to
using serotonergic medications in children with persistent
SM. Randomized effectiveness trials that include long-term
follow-up would further elucidate using SSRIs in this popula-
tion, alone or in combination with other interventions. Sup-
port was not found for non-medical interventions apart from
(possibly) school consultation, but standardized nonmedical
interventions or treatment packages that include medical and
nonmedical components await further development and
evaluation.
1. American Psychiatric Association. Diagnostic and statistical manual of mental
disorders. 4th ed. Washington (DC): APA; 1994.
2. Bergman RL, Piacentini J, McCracken JT. Prevalence and description of
selective mutism in a school-based sample. J Am Acad Child Adolesc
Psychiatry. 2002;41:938 –946.
3. Kristensen H. Selective mutism and comorbidity with developmental
disorder/delay, anxiety disorder, and elimination disorder. J Am Acad Child
Adolesc Psychiatry. 2000;39:249–256.
4. Black B, Uhde T. Treatment of elective mutism with fluoxetine: a double blind
placebo-controlled study. J Am Acad Child Adolesc Psychiatry.
1994;33:1000 –1006.
5. Calhoun J, Koenig KP. Classroom modification of elective mutism. Behav
Ther. 1973;4:700 –702.
6. Steinhausen HC, Juzi C. Elective mutism: an analysis of 100 cases. J Am Acad
Child Adolesc Psychiatry. 1996;35:606 –614.
7. Manassis K, Tannock R, McInnes A, et al. The sounds of silence: language,
cognition and anxiety in selective mutism. J Am Acad Child Adolesc
Psychiatry. 2007;46:1187–1195.
8. Bergman RL, Keller ML, Piacentini J, et al. The development and psychometric
properties of the selective mutism questionnaire. J Clin Child Adolesc Psychol.
2008;37:456-464.
9. Guy W. Global Improvement Scale (part of the Clinical Global Impressions and
Improvement Scales). Rockville (MD): National Institute of Mental Health;
1976.
10. Silverman WK, Albano AM. Anxiety Disorders Interview Schedule for
DSM-IV. New York (NY): Graywind Publications Inc; 1996.
11. Shaffer D, Gould MS, Brasic J, et al. A children’s global assessment scale
(CGAS). Arch Gen Psychiatry. 1983;40:1228–1231.
12. Dummit ES III, Klein R, Tancer NK, et al. Fluoxetine treatment of children
with selective mutism: an open trial. J Am Acad Child Adolesc Psychiatry.
1996;35:615–621.
13. Lam R, Kennedy S. Prescribing antidepressants for depression in 2005: recent
concerns and recommendations. A statement for the Canadian Psychiatric
Association. Can J Psychiatry. 2005;49:1–6.

Funding and Support

This work was supported by a Type A Grant from the Ontario
Mental Health Foundation.
Acknowledgements
We acknowledge the contributions of Mr David Avery (school
liaison coordinator), Ms Lisa Fiksenbaum (research coordinator),
and Dr Alison McInnes, Dr E Jane Garland, Dr Sandra Clark, and
Dr Klaus Minde (coinvestigators on original study).
Manuscript received November 2007, revised, and accepted January
2008.
1
Senior Associate Scientist, Hospital for Sick Children, Toronto, Ontario;
Associate Professor of Psychiatry, University of Toronto, Toronto,
Ontario.
2
Senior Scientist, Hospital for Sick Children, Toronto, Ontario; Professor,
University of Toronto, Toronto, Ontario.
Address for correspondence: Dr K Manassis, Department of Psychiatry,
Hospital for Sick Children,555 University Avenue, Toronto, ON
M5G 1X8; katharina.manassis@sickkids.ca

Résumé : Comparer les interventions pour le mutisme sélectif : une étude pilote
Objectif : Examiner le résultat en 6 à 8 mois d’une intervention médicale et non médicale auprès d’enfants souffrant de
mutisme sélectif (MS) grave.
Méthode : Des enfants souffrant de MS (n = 17) et leurs mères, vus dans une étude précédente, ont participé à des
rendez-vous de suivi avec un clinicien. Les rapports maternels présentaient : le traitement reçu, le diagnostic actuel
(selon une entrevue semi-structurée), le langage dans différents milieux, et l’amélioration générale. Un clinicien
indépendant cotait aussi le fonctionnement global.
Résultats : Le diagnostic de MS persistait chez 16 enfants, mais une amélioration symptomatique significative était
évidente dans l’échantillon. Tous les enfants avaient reçu des consultations à l’école. Les enfants qui avaient été traités
aux inhibiteurs spécifiques du recaptage de la sérotonine (ISRS) (n = 10) présentaient une plus grande amélioration
générale, une plus grande amélioration du fonctionnement, et une plus grande amélioration du langage en dehors de la
famille que les enfants qui n’étaient pas médicamentés (n = 7). Il n’y avait pas de différence manifeste pour les enfants
qui recevaient ou ne recevaient pas d’intervention non médicale additionnelle.
Conclusions : Les résultats suggèrent le bénéfice potentiel du traitement aux ISRS pour le MS grave, mais des études
aléatoires de traitements comparés sont indiquées.

The Canadian Journal of Psychiatry, Vol 53, No 10, October 2008 W 703