INT J LANG COMMUN DISORD, NOVEMBER–DECEMBER VOL.

2013,

48, NO. 6, 640–650

Research Report Effect of maxillary osteotomy on speech in cleft lip and palate: perceptual outcomes of velopharyngeal function Valerie J. Pereira†, Debbie Sell†‡§ and Jyrki Tuomainen¶ †Great Ormond Street Hospital for Children NHS Foundation Trust, London, UK ‡North Thames Regional Cleft Service, London, UK §Centre for Nursing and AHP Research, London, UK ¶Speech, Hearing and Phonetic Sciences, University College London, London, UK

(Received October 2012; accepted May 2013) Abstract Background: Abnormal facial growth is a well-known sequelae of cleft lip and palate (CLP) resulting in maxillary retrusion and a class III malocclusion. In 10–50% of cases, surgical correction involving advancement of the maxilla typically by osteotomy methods is required and normally undertaken in adolescence when facial growth is complete. Current evidence for the impact of the surgery on velopharyngeal function is weak and mixed. Aims: The first objective of the study was to investigate the nature of the effect of maxillary osteotomy on the perceptual outcomes of velopharyngeal function in CLP. The second objective was to establish if speech changes seen early at 3 months post-operation persisted for a year after/following surgery’, when it is considered that the maxilla is relatively stable. Methods & Procedures: Twenty consecutive patients with CLP undergoing maxillary osteotomy by a single surgeon were seen pre-operatively (T1), 3 months (T2) and 12 months (T3) post-operation. A non-cleft control group (NonCLP) undergoing surgery was also recruited. Speech data were collected using the Cleft Audit Protocol for Speech—Augmented (CAPS-A). A velopharyngeal composite score-summary (VPC-SUM) was derived from specific CAPS-A-rated parameters. An external CAPS-A-trained therapist, blinded to the study, rated the randomized samples and inter-rater reliability was established. Outcomes & Results: For the CLP group, hypernasality and nasal turbulence increased significantly post-operation. Planned comparisons were significant for T1–T2 only with a medium effect size. For hypernasality, the CLP group differed statistically from the NonCLP group at T2 and T3. For nasal turbulence, the CLP group differed statistically from the NonCLP group at T2. For VPC-SUM, there were statistically significant changes postoperatively between T1–T2 and T1–T3 only with medium effect sizes for the CLP group only. Conclusions & Implications: This study provides evidence that maxillary osteotomy affects patients with and without CLP differently. In patients with CLP, surgery may impact negatively on velopharyngeal function for speech and changes seen early on at 3 months post-operatively appear to persist at 12 months postoperatively. The findings in this study have implications for the speech care pathway of patients with CLP undergoing maxillary osteotomy in terms of assessment, review and management. Keywords: maxillary osteotomy, cleft lip and palate, velopharyngeal function.

What this paper adds? What is already known on this subject? The current evidence base for the impact of maxillary osteotomy on velopharyngeal function in cleft lip and palate (CLP) is weak due to variable and sometimes poor study methodology. This affects the patient’s informed consent process to surgical intervention as potential risks for speech are currently ill-defined. What this study adds? The study found that maxillary osteotomy impacts negatively on velopharyngeal function with increases in nasality and nasal turbulence in patients with CLP, and that speech changes seen 3 months post-operatively are stable. These Address correspondence to: Valerie J. Pereira, Department of Chinese and Bilingual Studies, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong; e-mail: [email protected] International Journal of Language & Communication Disorders C 2013 Royal College of Speech and Language Therapists ISSN 1368-2822 print/ISSN 1460-6984 online  DOI: 10.1111/1460-6984.12036

Effect of maxillary osteotomy on velopharyngeal function

641

findings have implications for the type of information provided to patients during the patient informed consent process to maxillary osteotomy, as well as the speech care pathway in terms of elements of assessment, review and management. The evidence suggests that management of acquired velopharyngeal inadequacy from maxillary osteotomy can be undertaken at an earlier point than the currently practised 12 months’ post-operative stage.

Introduction A well-known sequelae of cleft lip and palate (CLP) is abnormal facial growth. The facial profile of the individual with CLP is often described as ‘concave’ due to the ‘vertical deficiency in the mid-face and vertical excess in the lower face’ (Ross 1987: 21). This maxillary retrusion becomes increasingly evident during the pubertal growth spurt (e.g. Lello 2005). The need for surgical correction of maxillary retrusion in individuals with CLP ranges from 10% to 50% (e.g. Friede et al. 2011). Maxillary retrusion also occurs in individuals without CLP. In a US National Health and Nutrition Estimates Survey (NHANES-III), an estimated 5% of the population presented with this dentofacial deformity requiring surgical intervention (Proffit and White 2003). Surgical correction of maxillary retrusion involves advancement of the maxilla, which is sometimes undertaken with a simultaneous setback of the mandible. Maxillary advancement surgery is typically carried out using conventional osteotomy methods (Lello 2005). Currently, the most commonly used surgical technique is the Le Fort I osteotomy with maxillary advancement, undertaken in almost 84% of maxillary osteotomy cases (Cheung and Chua 2006). Two of the most commonly reported post-operative complications of maxillary osteotomy are skeletal relapse of the maxilla and velopharyngeal insufficiency. The stability of the maxilla following surgery may be affected by tethering of an existing pharyngeal flap and/or scar retraction. In the same review of maxillary advancement cases by Cheung and Chua (2006), the authors reported that horizontal relapse of the maxilla was most frequently reported at 1-year post-operatively (57% of reported cases) and this can have implications for speech. At present, the exact nature of the relationship between skeletal relapse of the maxilla and speech changes remains unknown. Nevertheless, by 12 months the maxilla is considered to be relatively stable and therefore a time frame which can be justified for measuring outcomes after orthognathic surgery (Eurocran 2003). Maxillary osteotomy and its effect on velopharyngeal function Maxillary advancement surgery is often multidimensional in nature, resulting in physical alterations that affect the size of the velopharyngeal orifice, dimensions

of the palate and nasopharyngeal area (Schendel et al. 1979, Okushi et al. 2011). Okushi et al. (2011) reported increases in the antero-posterior and left-right diameter of the velopharyngeal orifice. These physical alterations can lead to velopharyngeal dysfunction (VPD), which occurs when the velopharyngeal sphincter does not close consistently or completely for production of oral sounds (e.g. Kummer 2011). VPD is characterized by hypernasal resonance, abnormal nasal airflow, passive cleft type speech characteristics and compensatory articulation (e.g. Harding and Grunwell 1998). Where hypernasality is a supra-segmental feature reflecting excessive nasal resonance as a result of abnormal coupling of the oral and nasal cavities (Henningsson et al. 2008), nasal airflow reflects abnormal or inappropriate airflow through the nasal cavity on production of high-pressure consonants (John et al. 2006). Henningsson et al. (2008) differentiated between audible nasal emission described as non-turbulent frication, and nasal turbulence described as a turbulent ‘snorting’ noise. Passive cleft speech characteristics (CSCs) (Sell et al. 2009) include ‘weakened/nasalized consonants, nasal realizations of plosives or fricatives, absent pressure consonants, and gliding of fricatives/affricates’ (e.g. Sell et al. 1999, Grunwell and Sell 2005: 78), although it is recognized that gliding of fricatives can occur as part of normal speech development (Grunwell 1987: 223). Nasalization involves airflow through the oral and nasal cavities and often results in a reduction in intra-oral air pressure so that oral consonants are weak and/or nasalized (Sell et al. 1994). Nasal realizations of plosives (e.g. /b/ — [m]) is a classic CSC associated with lack of intra-oral pressure (Sell et al. 1999). Compensatory articulation, e.g. pharyngeal and glottal articulation (Trost 1981), may be a symptom of ongoing VPD but could also represent a persisting error of previous VPD, even after surgical or structural treatment (Lohmander et al. 2009). Whilst some studies have found that maxillary osteotomy results in VPD (e.g. Trindade et al. 2003), others have shown the reverse (e.g. Dalston and Vig 1984). Pereira et al. (2013) recently undertook a systematic review of the maxillary osteotomy literature using levels of evidence and power calculations. The results showed that for the speech parameter of velopharyngeal function, post-hoc power was low, ranging from 0.051 to 0.717 (where the generally accepted level is 0.800; Cohen 1992), and effect sizes were in the main

642

Valerie J. Pereira et al.

small, and at most medium (d = 0.024–0.696). The authors concluded that the evidence for the effects of maxillary osteotomy on velopharyngeal function in CLP is currently ill-defined and inconclusive. In many cases, this is related to study design issues such as small sample sizes and/or variable speech methodology providing potentially unreliable results. Perceptual assessment of speech in CLP: current approaches It is internationally accepted that perceptual assessment of speech is at the core of speech methodology in CLP (e.g. Lohmander and Olsson 2004, Sell 2005). One recent approach to speech analysis in CLP is the Cleft Audit Protocol for Speech-Augmented (CAPS-A; John et al. 2006). CAPS-A has undergone rigorous assessment of face, content and criterion validity, reliability, and acceptability, and is used in inter-centre CLP audit studies in the UK and Ireland (John et al. 2006). It includes the evaluation of articulation, resonance and nasal airflow. The use of a composite score to reflect velopharyngeal function based on specific speech parameters such as hypernasality and nasal airflow rated perceptually, has been previously advocated (e.g. Lohmander et al. 2009). With perceptual assessment of speech, several issues need to be considered carefully: listeners, recordings, playback and reliability. For example, in terms of listeners, Sell (2005: 103) advocated that ‘the gold standard methodological approach when reporting audit and research outcomes’ is the use of blind and independent analysis of speech by therapists specializing in the field. Additionally, listeners should be blinded to the aims of the study, the participants (for within-subject design studies), participant groups (for between-subject design studies) and datum points for intervention studies (Pereira 2012). To undertake unbiased ratings and analyses of speech and reliability studies, the data must be audio and/or video recorded (e.g. Lohmander and Olsson 2004, Sell 2005). Live ratings have inherent biases including clinician knowledge of the patient and whether the patient is pre- or post-surgical. Recent recommendations in CLP have been the use of both audio and video recordings (CAPS-A; John et al. 2006; and Cleft Palate International Speech Issues CLISPI; CLISPI n.d.). Consideration also needs to be given to the speech sample elicited and recorded. Several authors have advocated the use of sentences and sentence repetition as they provide a controlled way of data sampling (Sell and Grunwell 2005). In assessing hypernasality, although many authors recommend a speech sample devoid of nasal consonants (e.g. Henningsson et al. 2008), Sweeney and Sell (2008: 3–4) proposed an alternative view suggesting that ‘a speech sample without nasal con-

sonants is not representative of normal speech and may well represent a biased sample for assessment of hypernasality’ and, furthermore, ‘a speech sample, which represents the content of spontaneous speech, may have greater face validity’. The aims of the current study were to examine the nature of the impact of maxillary osteotomy on velopharyngeal function in patients with CLP and to explore if changes seen immediately (3 months post-operatively) persist a year post-operatively. Perceptual speech analyses that adhered to current standards articulated in the literature were used in the study. Method Participants There were two participant groups. All participants were native speakers of English. No participant reported or presented with significant hearing and/or learning difficulties and none had any known syndrome diagnosis:

r Group 1 (N = 20): consisted of a consecutive series of adolescents and adults with CLP (CLPGRP) undergoing maxillary osteotomy (with or without a mandibular setback) by a single surgeon, within the North Thames Regional Cleft Service in London. There were 16 males and four females. The mean age was 20;2 years (range 18;1– 30 years, SD = 2;6 years). r Group 2 (N = 10): consisted of adolescents and adults with no CLP (NonCLP) undergoing maxillary osteotomy (with or without a mandibular setback). Participants in this group were recruited from, and seen at, one of two NHS sites: Great Ormond Street Hospital (GOSH) or the Royal London Hospital (Barts and the London NHS Trust). There were six males and four females. Mean age was 25.7 years (range = 17;0–50;1 years, SD = 10;8 years). One participant had a diagnosis of high-functioning autism spectrum disorder which did not affect his ability to participate in the study. Baseline measurement and datum points Participants in Groups 1 and 2 were seen on three separate occasions: 0–3 months pre-operatively (T1), 3 months post-operatively (T2) and 12 months postoperatively (T3). These post-operative datum points were aimed at capturing early and late speech changes following advancement surgery, as recognized as the typical post-operative data points in the osteotomy field (e.g. Eurocran 2003, Trindade et al. 2003).

Effect of maxillary osteotomy on velopharyngeal function Table 1. The adapted CAPS-A listening protocol Speech sample Spontaneous speech Rote speech: counting from 1 to 20, days of the week, months of the year Sentences

Action Rate intelligibilitya Rate hypernasality, hyponasality, nasal emission and nasal turbulence Transcribe consonant production

Note: In the adapted protocol, only the audio component was provided to the raters. a Intelligibility was rated but not included in the analyses as the focus of this study was the effect of maxillary osteotomy on velopharyngeal function.

Materials and procedure All speech data were collected using a standardized speech protocol based on the CAPS-A (John et al. 2006) by the principal investigator (PI). The speech sample consisted of a brief conversation, rote speech (counting from 1 to 20, 60 to 70, days of the week, and months of the year), and the repetition of a set of standardized sentences. All speech data were recorded using a Panasonic digital video camera (NV-GS70) and with a Rode NT3 hypercardiod condenser microphone on a Samson boom microphone floor stand. The recommendation of using a plain background and the framing of a subject were followed (CLISPI n.d., John et al. 2006). All data were edited by the University College London (UCL) Medical Photography and Video Production team based at GOSH. Further editing of the digitized samples was undertaken by the PI to ensure that there was no remaining information about the participants, particularly in relation to speech and/or surgical status. For the purposes of this study, the CAPS-A listening protocol was adapted (table 1). Only the audio component was utilized as including the visual component would have shown if the participant was pre- or postoperative, thereby introducing bias. The audio component was extracted from the digitized samples using RER Audio Converter 3.7.5.0412 (A-S) and converted to .wav format (uncompressed CD audio quality). All audio samples were subsequently randomized according to participant group and datum point. There were six missing samples of a possible 110 across the two participant groups. The missing samples for the CLPGRP reflect two participants who failed to attend their 3-month post-operative evaluation. The four missing samples for the NonCLP are represented by two participants who did not attend their 3-month post-operative evaluation, one participant who could not be traced for his final 12-month post-operative appointment, and one other participant for whom sound was not recorded at his 12-month appointment. Ratings study A speech and language therapist with more than 25 years’ experience in the field of CLP served as the pri-

643 Table 2. Analyses and coding of perceptual speech data from the CAPS-A Parameter

Description of each scalar point

Hypernasality

Absent Borderline–minimal Mild Moderate Severe 0: Absent on pressure consonants 1: Seldom heard 2: Prevalent/often heard 0: Absent on pressure consonants 1: Seldom heard 2: Prevalent/often heard

Nasal emission Nasal turbulence

Note: Scalar points and descriptions for hypernasality are from the CAPS-A (John et al. 2006), and those for nasal emission and turbulence are from Sell et al. (2011). Source: Adapted from John et al. (2006).

Table 3. The CAPS-A cleft speech categories (CSCs)a Non-oral CSCs Pharyngeal articulation Glottal articulation Active nasal fricatives Double articulation Passive CSCs Weak or nasalized consonants Nasal realization of plosives, and/or suspected passive nasal fricativeGliding of fricatives/affricates Note: a The other two cleft speech categories in the CAPS-A are anterior oral CSCs and posterior oral CSCs. These have been excluded in the analyses as they are not usually associated with velopharyngeal dysfunction. Source: Adapted from John et al. (2006).

mary rater (Rater 1) with established inter- and intrarater reliability as part of the CAPS-A training programme. Rater 1 rated a total of 104 samples and was blinded to the aims of the study, participant groups and datum points. All ratings were undertaken independently using studio professional headphones (Beyerdynamic DT100), particularly designed to seal out ambient noise. Repetition of the samples was permitted, allowing the rater to check and confirm ratings and transcriptions. Coding and analyses Articulation, hypernasality, nasal emission, and nasal turbulence were rated and coded according to the CAPSA guidelines, on separate ordinal scales with differing numbers of scalar points (table 2). In terms of articulation, only the non-oral CSCs and passive CSCs (which are characteristics or symptoms of VPD) were coded and analysed for this analysis (table 3). For each articulation characteristic within each speech category, a score of ‘0’ was assigned if no consonants were affected, ‘1’ if less than or equal to two consonants were affected, and ‘2’ if three or more consonants were affected.

644

Valerie J. Pereira et al.

Table 4. Velopharyngeal composite scoring using the CAPS-A Speech parameter

CAPS-A ratings

Hypernasality

If 0–1 If 2–4 Nasal emission (NE) If NE = 0 and NT = 0 or Nasal turbulence (NT) If NE = 1 and NT = 0 or If NE = 0 and NT = 1 If NE = 1 and NT = 1 or If NE = 2 and NT = 0 or 1 or If NT = 2 and NE = 0 or 1 If absent or if one or two Non-oral CSCsa , e.g. pharyngeal articulation consonants are affected If three or more consonants are affected Passive CSCsa , e.g. weak If absent or if one or two or nasalized consonants consonants are affected If three or more consonants are affected

Scoring 0 1 0 1 0 1 0

Table 5. Inter-rater reliability for the ratings study Parameters/CSCs

Correlation (rs ) Po (%)a Po – 1 (%)b

Hypernasality Nasal emission Nasal turbulence Non-oral CSCs Pharyngeal articulation Glottal articulation Active nasal fricatives Double articulation Passive CSCs Weak or nasalized consonants Nasal realization of plosives

0.742∗ 0.503∗ 0.671∗

76.7 68.3 88.3

95.0 93.3 100

– 1.000∗∗ – –

98.3 98.3 91.7 93.3

98.3 100 95.0 100

95.0

100

95.0

98.3

∗

0.997 0.543

∗

Notes: Po = perfect agreement. b Po – 1 = precision of –1 to +1 agreement. –, Reliability was not computed as there was no variance in the data set for this parameter. ∗ p < 0.01, ∗ ∗ p < 0.001. a

1

Note: a CSCs, cleft speech characteristics (Sell et al. 2009) (see table 3). Source: Adapted from Lohmander et al. (2009) and Pereira (2012).

Reliability study A second CAPS-A trained rater with more than 10 years’ experience in CLP served as Rater 2 for the inter-rater reliability study. Ten participants from the CLPGRP and five from the NonCLP were identified and subsequently randomized according to participant group and datum point. Five normal speech samples collected by Pereira (2012) were also included in the reliability study and all samples were subsequently randomized. Fifty samples were rated in the inter-rater reliability study, constituting 48% of the total number of speech samples rated by Rater 1. The same studio headphones were used for the reliability study and all ratings were undertaken independently. As the speech parameters were rated on ordinal type scales, reliability was evaluated using Spearman’s correlation (rs ). In addition, per cent agreement between raters was also calculated using two statistics: perfect agreement (Po), and whether raters agreed to a precision of –1 to +1 scores (Po – 1). Perfect agreement was deemed acceptable when per cent agreement was equal to or more than 80%; and Po – 1 agreement was deemed acceptable when per cent agreement was equal to or more than 90% (Lord and Corsello 2005: 738, Zarcone et al. 1991). Velopharyngeal composite score-summary CAPS-A (VPC-SUM CAPS-A) Velopharyngeal composite scores were derived from the CAPS-A data and ratings study. An adapted version of the system by Lohmander et al. (2009) was used (table 4). Velopharyngeal composite scores were then summed to obtain a velopharyngeal composite score-summary (VPC-SUM) with a minimum of ‘0’ and a maximum of ‘4’. Clinical interpretation for each summary score

Source: Adapted from Pereira (2012).

was based on Lohmander et al. (2009): 0–1 = sufficient velopharyngeal function; 2 = borderline deficit; and 3–4 = insufficient velopharyngeal function. Statistical analyses All statistical analyses were undertaken using SPSS version 15.0 and effect sizes were calculated using G∗ Power 3.1.2 (Faul et al. 2007), where d = 0.2 is small, d = 0.5 is medium, and d = 0.8 is large (Cohen 1992). As the CAPS-A data were ordinal, a Friedman’s test (using a Chi Square distribution) was used to look at within-group differences over time (i.e. T1–T2–T3) for the CLPGRP and NonCLP separately. Planned comparisons across pairs of datum points (i.e. T1–T2, T1–T3, T2–T3) were undertaken using the Wilcoxon Signed Rank test. Between-group comparisons (CLPGRP and NonCLP) at each datum point were undertaken using the Mann– Whitney U-test. The significance criterion was set at α = 0.05 unless otherwise indicated. Results Inter-rater reliability for CAPS-A data Inter-rater reliability for the ratings study ranged from rs = 0.503 to 1.000 (all p < 0.01). Per cent agreement ranged from 76.7% to 98.3% for perfect agreement and from 93.3% to 100% for precision of +1 to –1 agreement (table 5). Non-oral CSCs, passive CSCs, resonance and nasal airflow Within-group differences over time For the NonCLP group, there was no statistically significant change following maxillary osteotomy for any of

Effect of maxillary osteotomy on velopharyngeal function

645

Table 6. CAPS-A within-group comparisons across pairs of datum points for the CLP group Parameter Hypernasality Nasal turbulence Nasal emission Non-oral CSCs Pharyngeal articulation Glottal articulation Active nasal fricatives Double articulation Passive CSCs Weak or nasalized consonants Nasal realization of plosives

Datum points Significance Effect size (d) a

T1–T2 T1–T3 T2–T3 T1–T2 T1–T3 T2–T3 T1–T2 T1–T3 T2–T3

0.026 0.053 0.380 0.041a 0.083 0.608 0.157 0.194 1.000

0.516 0.300 0.201 0.768 0.474 0.213 0.459 0.357 0.049

T1–T2 T1–T3 T2–T3 T1–T2 T1–T3 T2–T3 T1–T2 T1–T3 T2–T3 T1–T2 T1–T3 T2–T3

1.000 1.000 1.000 0.655 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000

0.000 0.000 0.000 0.147 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

T1–T2

0.317

0.134

T1–T3 T2–T3 T1–T2

1.000 0.317 1.000

0.000 0.134 0.000

T1–T3 T2–T3

0.317 0.317

0.257 0.257

Note: a Statistically significant difference across pairs of datum points.

the parameters evaluated. No participant in this group was rated as having any hypernasality, nasal emission, nasal turbulence or the presence of any CSCs pre- or post-operation. All effect sizes were, therefore, ‘0’. For the CLPGRP, however, hypernasality χ 2 (2) = 6.643, p = 0.036, and nasal turbulence χ 2 (2) = 6.059, p = 0.0489, changed significantly following surgery. For hypernasality, planned comparisons showed a statistically significant difference between T1 and T2, and an almost significant difference between T1 and T3. For nasal turbulence, there was also a statistically significant difference between T1 and T2. For both parameters, the difference between T2 and T3 was not statistically significant. Effect sizes for T1–T2 were medium for hypernasality and approaching large for nasal turbulence (table 6). At a group level, there was a minor reduction in hypernasality and nasal turbulence after the surgery, which yielded a smaller difference between T2 and T3 and also T1 and T3 compared with T1–T2. This resulted in a small to medium effect size for T1–T3 and a small effect size for T2–T3 for both parameters (table 6). The pattern suggests that maxillary osteotomy

Table 7. Individual ratings for hypernasality and nasal turbulence at each datum point Cases/ datum points 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Hypernasalitya

Nasal turbulenceb

T1

T2

T3

T1

T2

T3

3 3 0 0 0 0 0 0 0 0 0 0 2 0 0 0 1 0 0 0

3 3 2 2 0 0 0 0 0 1 0

3 4 1 0 0 0 0 0 0 2 0 0 3 0 0 2 0 0 0 1

0 0 0 0 0 1 0 0 0 0 0 0 0 – 0 0 1 0 1 0

0 0 1 2 0 1 0 1 1 0 0

0 0 2 0 0 0 1 1 0 0 0 0 2 0 0 2 2 0 0 0

a

3 0 0 3 2 a

0 0

a

2 0 1 0 2 a

1 0

Notes: a For hypernasality: 0 = absent hypernasality, 1 = borderline–minimal, 2 = mild, 3 = moderate and 4 = severe. b For nasal turbulence: 0 = absent on pressure consonants, 1 = seldom heard and 2 = prevalent/often heard. c These participants failed to attend their appointment at T2. –, Rater omitted rating for this participant at T1.

has a ‘true’ impact on hypernasality and nasal turbulence in CLP and that the changes seen at 3 months post-operatively are stable. In terms of hypernasality, 16 participants had oral resonance pre-operatively (table 7). Of these 16, five acquired hypernasality post-operatively, four at T2 and one at T3. Of the four participants who had hypernasality pre-operatively, three showed further deterioration post-operatively, although this resolved in one of the three cases at T3. In terms of nasal turbulence, 17 participants presented with no nasal turbulence preoperatively (table 7). Of these 17, eight cases acquired nasal turbulence post-operatively, two at T3 only. Only one of the three cases with pre-operative nasal turbulence deteriorated further post-operatively. There were no statistically significant differences over time for nasal emission, non-oral CSCs or for passive CSCs. Between-group differences There were statistically significant differences between the CLPGRP and the NonCLP at T2 and T3 for hypernasality, and at T2 for nasal turbulence. There were no statistically significant differences between the two groups for any of the other speech parameters. Table 8 shows that effect sizes for hypernasality increased from medium at T1 to large at T2 and T3. For nasal

646

Valerie J. Pereira et al.

Table 8. CAPS-A between-group comparisons at each datum point CLP versus NonCLP Parameter Hypernasality Nasal turbulence Nasal emission Non-oral CSCs Pharyngeal articulation Glottal articulation Active nasal fricatives Double articulation Passive CSCs Weak or nasalized consonants Nasal realization of plosives

Table 9. Individual VPC-SUM scores at each datum point VPC-SUM Cases/datum points

Datum point Significance Effect size (d) T1 T2 T3 T1 T2 T3 T1 T2 T3

0.157 0.029a 0.048a 0.215 0.017a 0.072 0.491 0.156 0.229

0.634 1.145 0.881 0.583 1.225 0.849 0.283 0.678 0.547

T1 T2 T3 T1 T2 T3 T1 T2 T3 T1 T2 T3

1.000 1.000 1.000 0.502 0.505 0.502 1.000 1.000 1.000 1.000 1.000 1.000

0.000 0.000 0.000 0.288 0.323 0.288 0.000 0.000 0.000 0.000 0.000 0.000

T1

0.334

0.425

T2 T3 T1

0.229 0.334 1.000

0.559 0.425 0.000

T2 T3

1.000 0.502

0.000 0.288

Note: a Statistically significant difference between groups.

turbulence a similar trend was found. These results suggest that maxillary osteotomy has a negative impact on hypernasality and nasal turbulence only in participants with CLP. Velopharyngeal composite score-summary: VPC-SUM All participants in the NonCLP had a summary score of ‘0’ at all datum points indicating velopharyngeal closure, suggesting that maxillary osteotomy does not impact negatively on velopharyngeal function in this clinical group. For the CLPGRP, however, there was a statistically significant difference in VPC-SUM scores over time χ 2 (2) = 9.769, p = 0.008. Planned comparisons across pairs of datum points showed statistically significant differences between T1 and T2 (p = 0.014) and between T1 and T3 (p = 0.024) but not between T2 and T3 (p = 1.000). Effect sizes were medium between T1 and T2 (d = 0.59) and between T1 and T3 (d = 0.50) but negligible between T2 and T3 (d = 0.068).

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

T1

T2

T3

1 2 0 0 0 0 0 0 0 0 0 0 1 0 0 0 1 0 0 0

2 3 1 2 0 0 0 0 0 0 0

3 3 1 0 0 0 0 0 0 1 0 0 2 0 0 2 1 0 0 0

a

2 0 0 1 2 a

0 0

Note: a These participants failed to attend their appointment at T2.

At a group level, mean rating increased from 0.22 at T1 to 0.72 at T2 and T3. On an individual level, 16 (of a total of N = 20) participants had a rating of ‘0’ at T1 indicating velopharyngeal closure (table 9). Of these 16, four cases received ratings of at least ‘1’ post-operatively. For all four cases, there were no perceptual symptoms of VPD at T1, i.e. none had hypernasality, nasal airflow or presented with passive or non-oral CSCs. This would suggest that even participants judged as having normal speech pre-operatively are potentially at risk of acquiring VPD following maxillary osteotomy. Discussion We investigated the nature of the impact of maxillary osteotomy on velopharyngeal function in patients with CLP and also observed if speech changes that are acquired immediately after surgery persist a year postoperatively. The findings of this study indicate that maxillary osteotomy results in a negative impact on velopharyngeal function only in individuals with CLP. The specific speech parameters affected are hypernasal resonance and nasal turbulence. There were also no statistically significant differences between the 3- and 12month post-operative datum points for both parameters, indicating that the profile seem immediately postoperatively is stable. These findings are similar to those reported by Chua et al. (2010), one of the seven studies identified and described in the systematic review by Pereira et al. (2013). Where the acquired hypernasality rate in the current

Effect of maxillary osteotomy on velopharyngeal function study was 31% (5/16), Chua et al. (2010) reported a similar rate of 36%. In their study, the authors also found that this deterioration in resonance, which was seen at 4 months post-operation, was maintained at 17 months post-operation. At present there is no robust literature on the impact of maxillary osteotomy on nasal turbulence. The same study by Chua et al. (2010) only made reference to individual changes in nasal emission postoperatively, which may have included nasal turbulence. Nasal turbulence usually reflects a small velopharyngeal gap (Kummer et al. 2003). This can usually be ascertained through the use of instrumental evaluation such as nasendoscopy, which provides a three-dimensional view of the velopharyngeal portal. Despite the non-significant changes over time for the cleft group in terms of non-oral and passive CSCs, velopharyngeal composite-summary scores increased significantly post-operatively, indicating deterioration in overall velopharyngeal function. Statistically significant changes between the pre- and post-operative datum points but not between the 3- and 12-month postoperative datum points provide evidence that maxillary osteotomy has a negative impact on velopharyngeal function in individuals with a cleft condition and that the deterioration seen at 3 months is stable. Effect sizes decreased from medium between pre- and both postoperative datum points to negligible between the 3- and 12-month post-operative datum points, providing further substantiating evidence. Most of the published literature on the effect of maxillary osteotomy on velopharyngeal function is based on instrumental evaluation (Pereira 2012). In the review by Pereira et al. (2013), the authors identified and described three studies: Chua et al. (2010)—nasendoscopy; Kummer et al. (1989)—videofluoroscopy, both of which involve direct observation of the velopharyngeal mechanism; and Dalston and Vig (1984)—pressure-flow methods. Whilst two of the studies (Chua et al. 2010, Kummer et al. 1989) reported deterioration in velopharyngeal function following surgery, Dalston and Vig (1984) found no statistical change over time. This body of evidence is conflicting, with none of the studies achieving a calculated post-hoc power of at least 0.8 and effect sizes also ranging widely from d = 0.064 to 0.696 (Pereira 2012, Pereira et al. 2013). Velopharyngeal composite scores based on perceptually rated speech parameters have, however, been reported in other maxillary osteotomy studies (e.g. Heli¨ovaara et al. 2004, Janulewicz et al. 2004). Heli¨ovaara et al. (2004) included parameters such as hypernasality, audible nasal emission, weakness of plosives and compensatory articulation. The degree of VPD was subsequently based on the severity of the ‘symptoms’. Janulewicz et al. (2004) used the University of Pittsburgh-weighted values for speech symptoms asso-

647 ciated with velopharyngeal incompetence scale which included nasal emission, facial grimace, nasality, phonation and articulation characteristics such as reduced intra-oral pressure for plosives or fricatives. Using these velopharyngeal composite scores, both Heli¨ovaara et al. (2004) and Janulewicz et al. (2004) reported velopharyngeal deterioration following maxillary osteotomy in individuals with a cleft condition. However, it is important to note that both studies were retrospective in nature, thereby susceptible to biases and incomplete datasets (Shaw and Semb 2005). The current findings, which were based on a prospective study, also add to the work of Witzel (1989) that pre-operative borderline velopharyngeal status is a potential risk factor in the context of maxillary osteotomy. In the current study, all four participants who acquired velopharyngeal insufficiency following surgery presented with no perceptual symptoms pre-operatively. All four cases had oral resonance, no nasal airflow errors and presented with no passive or non-oral CSCs. The perceptual evidence, therefore, suggests that even cases judged as having normal velopharyngeal function preoperatively are susceptible to acquiring VPD as a result of maxillary osteotomy. Witzel (1989: 200), who described pre-operative borderline velopharyngeal status as a risk factor, defined borderline as ‘normal nasal resonance or clinically insignificant hypernasality, inaudible nasal air emission (via mirror test), and borderline or marginal velopharyngeal closure during speech’. Witzel (1989: 200) added that such individuals ‘usually have small pinhole gaps in the velopharyngeal valve through which bubbles or barium or mucus are observed during videofluoroscopy, nasendoscopy, or both’. Witzel’s definition seems to imply the necessity of instrumental measurement of velopharyngeal function and status and/or perhaps the inadequacy of relying solely on perceptual outcomes of velopharyngeal function. However, the definition of what constitutes borderline velopharyngeal status differs across studies, and the variation in the type of instrumentation used to evaluate velopharyngeal function has resulted in conflicting evidence (e.g. Trindade et al. 2003, Chua et al. 2010). There are also reliability and validity issues related to the varying instrumental measurement methods (Gilleard et al. 2009, Sell and Pereira 2011). For example, Gilleard et al. (2009: 179) found that uncalibrated nasendoscopy edits were unreliable as ‘even relatively small differences in object-lens distance and variation in object position within the field of view cause significant changes in magnification and distortion’. The problem is exacerbated by the nature of the scales used in the visual perceptual ratings of recorded images based on videofluoroscopic and/or nasendoscopic images, as such rating scales may not be sufficiently sensitive for small or subtle changes in velopharyngeal function to be identified (Pereira 2012).

648 A plausible explanation as to why velopharyngeal deterioration is found only in individuals with a cleft condition following maxillary osteotomy is the lack of ‘adaptability’ of the velopharyngeal muscles to achieve velopharyngeal closure due to scarring from previous palate surgery (Witzel 1989), or inadequate ‘velar stretch’ (e.g. Schendel et al. 1979). Maxillary advancement surgery increases the antero-posterior and left– right diameters of the velopharyngeal area (Okushi et al. 2011), requiring the soft palate to stretch adequately to achieve velopharyngeal closure during speech. Schendel et al. (1979) found that following advancement of the maxilla, soft palate length increased by only 1.8 mm for individuals with a cleft condition, in contrast to an increase of 2.5 mm in those without a cleft. This provided evidence for the intrinsic hypoplastic nature of the velum and associated tissues in individuals with a cleft. Witzel (1989: 200) added that ‘previous scarring in the palate [i.e. from the primary palate repair] may have a tethering effect that impedes muscular stretch and movement’. More recently, Tian and Redett (2009) stated that the direction and orientation of the collagen fibres of the palatal aponeurosis may be disrupted following palatal surgery and post-surgical scarring, thereby affecting posterior velar stretch. This would explain why individuals without a cleft and no previous palate repair and subsequent surgical scarring are less likely to acquire velopharyngeal deterioration following maxillary advancement surgery. Further prospective osteotomy research with valid and reliable measurement of ‘velar length’ pre- and post-operatively, for example, is necessary to determine the full extent of the relationship between ‘velar stretch’ and velopharyngeal function. Although perceptual assessment of speech remains at the core of speech methodology in CLP (e.g. Sell 2005, Sweeney and Sell 2008), Dalston et al. (1988) recommended that dynamic assessment of the velopharyngeal mechanism during speech should include at least one instrumental method. Direct visualization of the velopharyngeal portal using measures such as nasendoscopy and videofluoroscopy not only ‘provide[s] information about structure, movement and extent of closure’, but also is useful in determining management of VPD (Sell and Pereira 2011: 147). Although not reported here, all participants in the CLPGRP in the current study underwent instrumental evaluation of velopharyngeal function using nasendoscopy and lateral videofluoroscopy at all datum points. Multiple regression analyses identified specific velar parameters, which were rated and measured on lateral videofluoroscopic images as valid predictors of acquired VPD following maxillary osteotomy. This has subsequently led to the recommendation of the inclusion of lateral videofluoroscopy as a key assessment in the speech care pathway of future patients with CLP undergoing maxillary osteotomy (Pereira 2012).

Valerie J. Pereira et al. Implications The current study adds to the currently existing limited evidence base in the field of maxillary osteotomy and speech research in CLP. The clinical significance of the study is in the informing of the speech care pathway for future patients with CLP undergoing maxillary osteotomy, and in the improvement of the patient informed consent process to surgical intervention. Future patients should be counselled accordingly and be told of the potential risk of the surgical procedure on velopharyngeal function and the subsequent need for velopharyngeal surgery. In terms of the speech care pathway, the current findings have implications for the assessment and review processes. Pre- and post-operative perceptual speech assessment should focus on the parameters of hypernasality and nasal turbulence as these have been identified as the significant clinical features of acquired VPD following maxillary osteotomy. This should be undertaken in conjunction with lateral videofluoroscopy as already alluded to above. In terms of the timing of post-operative review, the current evidence suggests that an earlier datum point of 3 months post-operatively is sufficient compared with the currently practised one year following surgery. For many of these patients, orthognathic surgery represents the final stage in their cleft care pathway. As Bradbury (2005: 368) succinctly stated: However, when treatment such as an osteotomy goes well and the result is successful for the adolescent, the rewards can be very great. The young person often describes the feeling of having lost the stigma of the ‘cleft appearance’ and can now relate to others with more social confidence. There is also a sense of relief that the final treatment hurdle has been.

Conclusions This study has adhered to current approaches and international standards of speech and study methodology in the field of CLP with the inclusion of a non-cleft group undergoing maxillary osteotomy, a speech reliability study with two post-operative datum points, including an adequate follow-up at 12 months. Using effect sizes in addition to statistical analyses, the study provides strong evidence for the negative impact of maxillary osteotomy on velopharyngeal function in individuals with a cleft condition. In addition, the study has demonstrated that velopharyngeal changes seen immediately post-operatively are stable and permanent. The findings have clear implications for the speech care pathway of future patients with CLP undergoing maxillary osteotomy, encompassing elements of assessment, review and management.

Effect of maxillary osteotomy on velopharyngeal function Acknowledgements Funding by Sport Aiding Research for Kids (SPARKS UK, Grant no. is 2DTK) was awarded to Dr D. Sell. Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

References BRADBURY, E., 2005, Growing up with a cleft: the impact on the child. In A. C. H. Watson, D. Sell and P. Grunwell (eds), Management of Cleft Lip and Palate (London: Whurr), pp. 365–385. CHEUNG, L. K. and CHUA, H. D., 2006, A meta-analysis of cleft maxillary osteotomy and distraction osteogenesis. International Journal of Oral and Maxillofacial Surgery, 35(1), 14–24. CHUA, H. D., WHITEHILL, T. L., SAMMAN, N. and CHEUNG, L. K., 2010, Maxillary distraction versus orthognathic surgery in cleft lip and palate patients: effects on speech and velopharyngeal function. International Journal of Oral and Maxillofacial Surgery, 39(7), 633–640. CLEFT PALATE INTERNATIONAL SPEECH ISSUES (CLISPI), n.d. (available at: https://www.clispi.org) (accessed on 12 July 2012). COHEN, J., 1992, A power primer. Psychological Bulletin, 112(1), 155–159. DALSTON, R. M., MARSH, J. L., VIG, K. W., WITZEL, M. A. and BUMSTED, R. M., 1988, Minimal standards for reporting the results of surgery on patients with cleft lip, cleft palate, or both: a proposal. Cleft Palate Journal, 25(1), 3–7. DALSTON, R. M. and VIG, P. S., 1984, Effects of orthognathic surgery on speech: a prospective study. American Journal of Orthodontics, 86(4), 291–298. EUROCRAN, 2003, Eurocran: Timing of Minimal Records (available at: http://www.eurocran.org/content.asp?contentID=779) (accessed on 12 July 2012). FAUL, F., ERDFELDER, E., LANG, A. G. and BUCHNER, A., 2007, G∗ Power 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior Research Methods, 39(2), 175–191. FRIEDE, H., LILJA, J. and LOHMANDER, A., 2011, Long-term, longitudinal follow-up of individuals with UCLP after the Gothenburg primary early veloplasty and delayed hard palate closure protocol: maxillofacial growth outcome. Cleft Palate Craniofacial Journal, 49(6), 649–656. GILLEARD, O., SOMMERLAD, B., SELL, D., GHANEM, A. and BIRCH, M., 2009, Nasendoscopic image analysis: an analysis of measurement uncertainties. In C. E. Raposo do Amaral and Ana Beatriz Albino de Almeida (eds), 11th International Congress on Cleft Lip and Palate and Related Craniofacial Anomalies (Italy: Medimond), pp. 179–184. GRUNWELL, P., 1987, Clinical Phonology, 2nd edn (London: Chapman & Hall). GRUNWELL, P. and SELL, D., 2005, Speech and cleft palate/velopharyngeal anomalies. In A. C. H. Watson, D. Sell and P. Grunwell (eds), Management of Cleft Lip and Palate (London: Whurr), pp. 68–86. HARDING, A. and GRUNWELL, P., 1998, Active versus passive clefttype speech characteristics. International Journal of Language and Communication Disorders, 33(3), 329–352. ¨ HELIOVAARA , A., HUKKI, J., RANTA, R. and HAAPANEN, M. L., 2004, Cephalometric pharyngeal changes after Le Fort I osteotomy in different types of clefts. Scandinavian Journal Plastic and Reconstructive Surgery and Hand Surgery, 38(1), 5–10. HENNINGSSON, G., KUEHN, D. P., SELL, D., SWEENEY, T., TROSTCARDAMONE, J. E. and WHITEHILL, T. L., 2008, Universal

649 parameters for reporting speech outcomes in individuals with cleft palate. Cleft Palate Craniofacial Journal, 45(1), 1–17. JANULEWICZ, J., COSTELLO, B. J., BUCKLEY, M. J., FORD, M. D., CLOSE, J. and GASSNER, R., 2004, The effects of Le Fort I osteotomies on velopharyngeal and speech functions in cleft patients. Journal of Oral and Maxillofacial Surgery, 62(3), 308–314. JOHN, A., SELL, D., SWEENEY, T., HARDING-BELL, A. and WILLIAMS, A., 2006, The cleft audit protocol for speech-augmented: a validated and reliable measure for auditing cleft speech. Cleft Palate Craniofacial Journal, 43(3), 272–288. KUMMER, A. W. 2011. Type and causes of velopharyngeal dysfunction. Seminars in Speech and Language, 32(2), 150–158. KUMMER, A. W., BRIGGS, M. and LEE, L., 2003, The relationship between the characteristics of speech and velopharyngeal gap size. Cleft Palate Craniofacial Journal, 40(6), 590–596. KUMMER, A. W., STRIFE, J. L., GRAU, W. H., CREAGHEAD, N. A. and LEE, L., 1989, The effects of Le Fort I osteotomy with maxillary movement on articulation, resonance, and velopharyngeal function. Cleft Palate Journal, 26(3), 193–199. LELLO, G. E., 2005, Orthognathic surgery. In A. C. H. Watson, D. A. Sell and P. Grunwell (eds), Management of Cleft Lip and Palate (London: Whurr), pp. 338–351. LOHMANDER, A. and OLSSON, M., 2004, Methodology for perceptual assessment of speech in patients with cleft palate: a critical review of the literature. Cleft Palate Craniofacial Journal, 41(1), 64–70. LOHMANDER, A., WILLADSEN, E., PERSSON, C., HENNINGSSON, G., BOWDEN, M. and HUTTERS, B., 2009, Methodology for speech assessment in the Scandcleft project—an international randomized clinical trial on palatal surgery: experiences from a pilot study. Cleft Palate Craniofacial Journal, 46(4), 347– 362. LORD, C. and CORSELLO, C., 2005, Diagnostic instruments in autistic spectrum disorders. In F. R. Volkmar, R. Paul, A. Klin and D. J. Cohen (eds), Handbook of Autism and Pervasive Developmental Disorders (New York, NY: Wiley), pp. 730–771. OKUSHI, T., TONOGI, M., ARISAKA, T., KOBAYASHI, S., TSUKAMOTO, Y., MORISHITA, H., SATO, K., SANO, C., CHIBA, S., YAMANE, G. Y. and NAKAJIMA, T., 2011, Effect of maxillomandibular advancement on morphology of velopharyngeal space. Journal of Oral and Maxillofacial Surgery, 69(3), 877–884. PEREIRA, V., 2012, The effect of maxillary advancement on speech, nasality and velopharyngeal function in cleft lip and palate. PhD thesis, UCL Institute of Child Health, London. PEREIRA, V. J., SELL, D. and TUOMAINEN, J., 2013, The impact of maxillary osteotomy on speech outcomes in cleft lip and palate: an evidence-based approach to evaluating the literature. Cleft Palate Craniofacial Journal, 50(1), 25–39. PROFFIT, W. and WHITE, W. R., 2003, Dentofacial problems: prevalence and treatment need. In W. Proffit and D. Sarver (eds), Contemporary Treatment of Dentofacial Deformities (St. Louis, MI: Mosby), pp. 2–28. ROSS, R. B., 1987, Treatment variables affecting facial growth in complete unilateral cleft lip and palate. Cleft Palate Journal, 24(1), 5–77. SCHENDEL, S. A., OESCHLAEGER, M., WOLFORD, L. M. and EPKER, B. N., 1979, Velopharyngeal anatomy and maxillary advancement. Journal of Maxillofacial Surgery, 7(2), 116–124. SELL, D., 2005, Issues in perceptual speech analysis in cleft palate and related disorders: a review. International Journal of Language and Communication Disorders, 40(2), 103–121. SELL, D. and GRUNWELL, P., 2005, Speech assessment and therapy. In A. C. H. Watson, D. Sell and P. Grunwell (eds), Management of Cleft Lip and Palate (London: Whurr), pp. 227–257.

650 SELL, D., HARDING, A. and GRUNWELL, P., 1994, A screening assessment of cleft palate speech (Great Ormond Street Speech Assessment). European Journal of Disorders of Communication, 29, 1–15. SELL, D., HARDING, A. and GRUNWELL, P., 1999, GOS.SP.ASS.’98: an assessment for speech disorders associated with cleft palate and/or velopharyngeal dysfunction (revised). International Journal of Language and Communication Disorders, 34(1), 17– 33. SELL, D., JOHN, A., HARDING-BELL, A., SWEENEY, T., HEGARTY, F. and FREEMAN, J., 2009, Cleft audit protocol for speech (CAPS-A): a comprehensive training package for speech analysis. International Journal of Language and Communication Disorders, 44(4), 529–548. SELL, D. and PEREIRA, V., 2011, Instrumentation in the analysis of the structure and function of the velopharyngeal mechanism. In A. Lohmander and S. Howard (eds), Cleft Palate Speech (Chichester: Wiley), pp. 145–166. SELL, D., SWEENEY, T., HARDING-BELL, A. and JOHN, A., 2011, An update on cleft audit protocol for speech (CAPS-A). Paper presented at the Craniofacial Society of Great Britain and Ireland Annual Conference, Pre-Conference Special Interest Group Meeting, Ireland, 2011. SHAW, W. C. and SEMB, G., 2005, Choosing the best treatment for the child with a cleft. In A. C. H. Watson, D. Sell and

Valerie J. Pereira et al. P. Grunwell (eds), Management of Cleft Lip and Palate (London: Whurr), pp. 386–403. SWEENEY, T. and SELL, D., 2008, Relationship between perceptual ratings of nasality and nasometry in children/adolescents with cleft palate and/or velopharyngeal dysfunction. International Journal of Language and Communication Disorders, 43(3), 265–282. TIAN, W. and REDETT, R. J., 2009, New velopharyngeal measurements at rest and during speech: implications and applications. Journal of Craniofacial Surgery, 20(2), 532– 539. TRINDADE, I. E., YAMASHITA, R. P., SUGUIMOTO, R. M., MAZZOTTINI, R. and TRINDADE, A. S., JR, 2003, Effects of orthognathic surgery on speech and breathing of subjects with cleft lip and palate: acoustic and aerodynamic assessment. Cleft Palate Craniofacial Journal, 40(1), 54–64. TROST, J. E., 1981, Articulatory additions to the classical description of the speech of persons with cleft palate. Cleft Palate Journal, 18(3), 68–70. WITZEL, M. A., 1989, Commentary. Cleft Palate Journal, 26(3), 199–200. ZARCONE, J. R., RODGERS, T. A., IWATA, B. A., ROURKE, D. A. and DORSEY, M. F., 1991, Reliability analysis of the motivation assessment scale: a failure to replicate. Research in Developmental Disabilities, 12(4), 349–360.