Marilyn Langevin and Deborah Kully
Institute for Stuttering Treatment & Research (ISTAR), The University of Alberta, 3rd floor; 8220 - 114 Street, Edmonton, Alberta, Canada, T5J 2P4
This report focuses on the process of maintenance for four subjects over a 5 year longitudinal period. Findings indicated that: fluctuations in fluency occurred but were not excessive, slower rates of speech can sound relatively natural, the expectancy to stutter remained salient in spite of good maintenance of speech gains, there were general trends for either maintenance of improved post-treatment confidence in approaching a variety of speech situations or improvement over the follow-up period, and either post treatment normalization of communication attitudes was maintained or normalization was achieved at 5 years. All subjects were maintaining a substantial amount of their post- treatment speech and emotional/attitudinal gains at 5 years follow-up.
- The clinical program
S1, 29 years of age, with a grade 12 education, presented with mild-moderate stuttering and a family history of stuttering. He had previously received approximately 4 hours of fluency shaping therapy, 1 hour of which occurred in the year immediately prior to the intensive clinic. S3, 37 years of age, with post-secondary education presented with moderate stuttering and a family history of stuttering. He had no previous therapy. S10, aged 15 years, was in high school. His stuttering was mild and he had no family history of stuttering. He had a long history of therapy which was either unsuccessful or of short term benefit. He had not received therapy for one year prior to the intensive clinic. S16, aged 42, with post-secondary education, presented with very severe stuttering. He had no family history of stuttering and had a long history of non-avoidance and stuttering management therapy with no durable improvements in fluency but had not received therapy for 4 years prior to the intensive clinic. “
Outcome evaluation measures
The dependent speech measures are percentage of syllables stuttered (%SS) and syllables spoken per minute (SPM) obtained in audio-taped telephone calls consisting of 2 minutes of subject talk time. At pre- and post-treatment calls were made within the clinic to businesses and at follow- up surprise calls were made to clients at their home or work by unfamiliar research assistants. Naturalness (NAT) ratings of 5”â year speech samples are also reported.
Non-speech dependent measures include the (a) Revised Communication Attitude Inventory (S-24) (Andrews & Cutler, 1974) for which the mean for nonstutterers is 9.14 (S.D. = 5.38), (b) Perceptions of Stuttering Inventory (PS1) (Woolf, 1967), (C) the approach scale of the Self-Efficacy Scaling by Adult Stutterers (SESAS-Approach) (Ornstein & Manning, 1985), and (d) the Speech Performance Questionnaire (SPQ) adapted from Perkins (1981). These questionnaires are routinely administered to inform treatment and evaluate outcome (Langevin & Kully, in press). In addition subjects completed reports on maintenance activities and therapy accessed during the follow-up period. During follow-up, questionnaires were mailed to the subjects’ homes for completion.
Reliability of speech and naturalness measures.
Seven, or 26%, of the speech samples in this investigation were included in the samples randomly chosen to evaluate inter-rater agreement in the CSP 5 year Outcome Project. Four research assistants who weretrained to count stuttered and fluent syllables using Kully’s (1986) counting guidelines and were independent of the investigation served as raters. See Boberg & Kully (1994) for a more detailed description of the training program for raters. Rater 1, who was not able to participate in rating samples from all measurement periods, rated a subset of samples from pre, post and 1 year follow-up. Raters 2, 3 and 4 rated samples from all measurement periods.
Inter-rater agreements for %SS and SPM are displayed Figure 1. Pearson correlations for %SS for Rater 1 with each of Raters 2, 3 and 4 were .97, .99 and .98 (p=.Ol, two tailed) respectively. For Rater 2, Correlations with Raters 3 and 4 were .93, .92 respectively, and the correlation for Rater 3 with 4 was .93 (all significant at p=.01,). For SPM, correlations for Rater 1 with Raters 2, 3 and 4 were .95, .99, .99 respectively. Correlations for Rater 2 with R3 and R4 were .93 and .86 respectively, and the correlation for Rater 3 with 4 was .96 (all significant at p=.01).
Figure 1. Inter-rater agreement for %SS and SPM.
Naturalness ratings (NAT) in the CSP 5 year Outcome Project were made for all subjects and matched controls by 30 independent listeners using Martin, Haroldson and Triden’s (1984) nine point scale (9 = highly unnatural speech; 1: highly natural speech). At the time of this writing only intra-rater agreements were available. Twenty of the 123 samples were re-rated, 89.3% of which were within 1 scale value of the original rating.
Figure 2. Speech and non-speech data for SI and S3.
Figure 3. Speech and non-speech data for S10 and S16.
S3’s %SS decreased from 10 % at pre-treatment to 0% at post-treatment. It remained unchanged at 1 year, rose to 2% at 3 years, and then decreased to .59% at 5 years. His 5th year %SS reflected a 94.1% improvement relative to pre-treatment. During follow-up his SPM fluctuated between a low of 140.8 at 5 years and a high of 182.9 at 1 year. His NAT rating for the 5 year sample was 3.5, slightly lower than the highest mean for normally fluent speakers referred to above. His PSI total score fluctuated between a low of 5 at 1 year and a high of 25 at 4 years but dropped again at 5 years. Interestingly his Expectancy subscale score did not improve until year 1 and was consistently higher (more negative) than the Struggle and Avoidance subscales. In spite of the higher levels of expectancy to stutter, which one would expect might be associated with less desire to approach speech situations, S3’s follow-up SESAS scores improved slightly from post-treatment and remained stable to 5 years. His S24 scores remained normal and relatively stable through the follow- up period as well. S3’s maintenance record revealed that he belonged to a self-help group throughout the 5 year period, that he attended “casual” refresher clinics generally 2 times per year (for 4 of the 5 years) and that, remarkably, he maintained home practice at 3 to 4 times per week throughout the 5 year period. S3’s SPQ responses indicated that he almost always had to pay attention to his speech to be fluent, that his fluency skills almost always worked, that he almost always had the skills to sound normal when controlling speech, that his overall level of fluency was generally good, that he seldom preferred to stutter than use fluency skills, and that improvements in speech were attributed to therapy in the CSP.
Sl0’s %SS dropped from 3.7% at pre-treatment to 1.6% at post-treatment. This reflected a 56.8% improvement relative to 3.7 %SS at pre-treatment. Over the 5 year follow-up period his rate fluctuated by no more than 1 %SS and at 5 years he was maintaining 27.3% improvement relative to pre-treatment. His SPM fluctuated between a low of a 156.1 at 1 year and a high of 183.1 at 5 years. His 5 year sample was given a mean NAT rating of 2. S10’s PSI total scores fluctuated during the follow-up period and again it was notable that the Expectancy subscale scores were consistently higher than other subscale scores at all follow-up periods. His SESAS-Approach scores showed nearly continuous improvement during follow-up. Interestingly Sl0’s S24 scores did not approach the mean of 9.14 for nonstutterers (Andrews & Cutler, 1974) until 2 years post-treatment but thereafter remained normal at 3 and 5 years follow-up. S10 did not belong to a self-help group and he did not access follow-up speech therapy in the reported years. His maintenance report indicated that he began with monthly home practice in the first six months following therapy, after which home practice fluctuated between weekly, monthly, and no practice in the remainder of the reported years. Responses to the SPQ indicated that he sometimes had to pay attention to his speech to be fluent, that his fluency skills worked most of the time, that he almost always had the skills to sound normal when controlling speech, that his overall level of fluency was generally good, that he sometimes preferred to stutter than use fluency skills, and that improvements in speech were attributed to therapy in the CSP.
Unfortunately S16, who lived overseas, could not be contacted by telephone in the first two years of follow-up, however, questionnaires were returned for the first 4 follow-up periods. Despite regression in %SS at 3 and 4 years, his” improvement relative to pre-treatment was 82%, 72% and 97% at 3, 4, and 5 years respectively. The improvement to 1.12 %SS at 5 years from 36.8% pre- treatment was associated with a rate of 120.3 SPM which had improved from 44.7 SPM at pre- treatment. His speech rate remained relatively stable over the reported years. The NAT rating given to his 5 year sample was 4.6, slightly higher than the mean of 4.5 (S.D. =l.9) for experimental subjects in O’Brian et al. (2003) and within the range of that given to matched controls in that study. This rating may have been influenced by S1 6’s accent (Mackey et al., 1997), however, it does give clinically useful information about the degree of naturalness that can be achieved at slower rates of speech. Sl6’s PSI scores remained relatively stable in the reported years. His SESAS-Approach scale score returned to pre-treatment levels at 1 year but improved consistently from the 2”â to the 4”â year. An almost mirror image of this data trend is reflected in the S24 scores which were showing regression at 1 year and then improved consistently to normalize at 4 years. S16 did not belong to a self help group. He accessed follow-up therapy in the 3"‘ year, 2 times a month for 6 months.
Home practice progressed from daily in the first year to not all in the majority of the succeeding reported years, but he indicated that he was making many phone calls in the 4”â year in his work. SPQ responses indicated that to be fluent S 16 had to pay attention to his speech most of the time, that his fluency skills worked some of the time and were dependent on the situation and people, that he had the skills to sound normal when controlling speech most of the time, that his overall level of fluency was generally good, that he would rather stutter than use fluency skills some of the time, and that improvements in speech were attributed to therapy in the CSP and other therapy.
- Discussion and conclusion
In terms of the non-speech measures, it is notable that the PSI Expectancy subscale scores were almost always higher (more negative) than the Struggle or Avoidance subscale scores from post-treatment to 5 years. The implications of this are unknown. Indeed, it seems preferable that Expectancy scores would be the same as Struggle and Avoidance scores, however, it may not be logical given the subjects’ long histories of stuttering and the variability in stuttering that is characteristic of the condition. For S3, it is possible that the higher Expectancy scores contributed to his diligence in carrying out home practice throughout the 5 year period.
Over the 5 year measurement period SESAS - Approach scores showed little fluctuation for two subjects (S1 and S3), almost consistent improvement for one subject (S10), and a return to post-treatment levels at 1 year with continuous improvement thereafter for the 4”â subject (S16). S24 scores normalized for three subjects post-treatment and remained so for two (S1 and S3) over the 5 years and returned to the normal for two (S10 and S16) after relapse or regression at 1 year. A visual inspection of the non-speech graphs revealed some relationships among the non-speech measures over the 5 year period though the patterns were not consistent. For example, Sl0’s more negative communication attitudes (S24) at 1 year were associated with an expected increase in negativity in perceptions (PSI; particularly the Avoidance and Expectancy subscales), however, there was no associated drop in the SESAS-Approach score that would also have been expected. A stronger relationship occurred in Sl6’s data: less positive communication attitudes at 1 year (S24) were associated with less confidence in approaching speech situations (SESAS) and more negative perceptions (PSI). Visual comparison of the non-speech graphs with the %SS graphs revealed some 202 Theory, research and therapy in fluency disorders corresponding increases in %SS with increases in negativity of perceptions and attitudes and loss of confidence (e. g., S1 at 3 years), however, these associations were not consistent (e.g., for S16 an increase in %SS at 4 years was instead associated with an improvement in attitudes at 4 years).
Reported home practice activities for 3 subjects revealed a decline in practice over the years, which is consistent with usual verbal reports of clients. Given the favourable %SS data, it appears that continued home practice was not needed to maintain fluency gains over the 5 year period for 3 of the 4 subjects. Whether or not S3 needed to maintain his rigorous schedule of practice is unknown. Certainly he determined that he did so.
SPQ responses indicated that as a group these subjects perceived that they sounded normal either most of the time or almost always when controlling speech. Only one subject reported having to almost always pay attention to speech. Interestingly two subjects (S1, S3) seldom preferred to stutter than use fluency skills and two (S10, S16) did so some of the time. All subjects rated their fluency as generally good fluency. These self-perceptions together with their corroborating %SS data indicate that these subjects were maintaining speech gains achieved in treatment. These subjects are representative of the 76 to 80% of subjects who in previous studies (Boberg & Kully, 1994; Langevin & Boberg ,l993), successfully maintained treatment gains at 1 and 2 years. However, the degree to which they are representative of the 36 subjects in the 5 Year Outcome Project is still to be determined. It will be remembered that these subjects were selected because they had the most complete non-speech data over the 5 year period.
In conclusion, all subjects were maintaining a substantial amount of their post-treatment speech and non-speech gains. The data show expected but not excessive year to year fluctuations in %SS. NAT ratings indicate that substantially slower speech rates needed by some clients to achieve more fluent speech can sound relatively natural. The non-speech data indicate that (a) the expectancy to stutter remained more salient than perceptions of struggle and avoidance in spite of good fluency levels, (b) there were general trends for either maintenance of improved post-treatment SESAS-Approach scale scores or improvement over the follow-up period, and (c) there were two trends in the S24 scores: normalized post-treatment attitudes were either maintained over the five year period or achieved at 5 years after 1 year relapse or regression. Maintenance reports generally confirm clinical experience which suggests that there is a gradual decline in home practice in the months and years following therapy and that, as is generally known and encouraged, some follow- up therapy was accessed to assist with maintenance in the post-treatment period. The relationship between treatment and outcome was evident in the substantial decrease in post-treatment %SS
and improvements in non-speech data. Three subjects attributed speech improvements solely to the CSP and one (S16), attributed speech improvements to the CSP and other therapy. It will be remembered that this subject brought a long history of non-avoidance and stuttering management therapy which had not resulted in durable gains but no doubt supported these components of the CSP. These findings are encouraging and give preliminary indications of the process of maintenance experienced by some clients.
Andrews, G., & Cutler, J. (1974). Stuttering therapy: The relations between changes in symptom level and attitudes. Journal of Speech and Hearing Disorders, 39, 312-319. Bloodstein, O. (1995). A handbook on stuttering (5”' ed. ). San Diego, CA: Singular Publishing
Boberg, E. (Ed.). (1981). Maintenance of fluency. New York, NY: Elsevier North Holland, Inc.
Boberg, E., & Kully, D. (1985). Comprehensive stuttering program. San Diego, CA: College-Hill Press.
Boberg, E., & Kully, D. (1994). Long-term results of an intensive treatment program for adults and adolescents who stutter. Journal of Speech and Hearing Research, 3 7, 1050-1059. Section 4. Outcomes and Efficacy of Interventions 203
Ingham, R. J. (1993). Current status of stuttering and behavior modification - II. Principal issues and practices in stuttering therapy. Journal of Fluency Disorders, 18, 57-79.
Kully, D. (1986). Counting guidelines. Unpublished document, Institute for Stuttering Treatment & Research. Edmonton, AB. ~
Kully, D., & Langevin, M. (1999). Intensive treatment for stuttering adolescents. In R.F. Curlee (Ed.), Stuttering and related disorders of fluency (pp.139-159). New York, NY: Thieme Medical Publishers, Inc.
Langevin, M., & Boberg, E. (1993). Results of an intensive stuttering therapy program. Journal of Speech-Language Pathology and Audiology, 17, 158-166.
Langevin, M., & Boberg, E. (1996). Results of an intensive stuttering therapy with adults who clutter and stutter. Journal of Fluency Disorders, 21, 315-327.
Langevin, M., & Kully, D. (2003). Evidence-based treatment of stuttering: III. Evidence-based practice in a clinical setting. Journal of Fluency Disorders, 28, 219-236.
Mackey, L. S., Finn, P., Ingham, -R.,J. (1997). Effect of speech dialect on speech naturalness ratings: A systematic replication of Martin, Haroldson, and Triden (1984). Journal of Speech, Language, and Hearing Research, 40, 349-350.
Martin, R. R., & Haroldson, S. K. (1992). Stuttering and speech naturalness: audio and audiovisual judgements. Journal of Speech and Hearing Research, 35,521-528.
Martin, R.R., Haroldson, S.K., & Triden, K.A. (1984). Stuttering and speech naturalness. Journal of Speech and Hearing Disorders, 49, 53-58.
O’Brian, S., Onslow, M., Cream, A., & Packman, A. (2003). The Camperdown program: Outcomes of a new pro1onged-speech treatment model. Journal of Speech, Language and Hearing Research, 46, 933-946.
Ornstein, A., & Manning, W. (1985). Self-efficacy scaling by adult stutterers. Journal of Communication Disorders, 18, 313-320.
Perkins, W. H. (1981). Measurement and maintenance of fluency. In E. Boberg (Ed.), Maintenance of fluency (pp. 147-178). New York, NY: Elsevier North Holland.
Runyan, C. M., Bell, J. N., & Prosek, R. A. (1990). Speech naturalness ratings of treated stutterers. Journal of Speech and Hearing Disorders, 55, 434-438.
Woolf, G. (1967). The assessment of stuttering as struggle, avoidance and expectancy. British Journal of Disorders of Communication, 2, 158-171.
The authors thank Shelli Wright for her work in collecting naturalness ratings as part of her masters degree project submitted to the Department of Speech Language Pathology at the University of Alberta.