Speech after repair of isolated cleft palate and cleft lip and palate

Transcription

1 British Journal of Plastic Surgery (2001), 54, The British Association of Plastic Surgeons doi: Yojps BRITISH JOURNAL OF [ ~ J PLASTIC SURGERY Speech after repair of isolated cleft palate and cleft lip and palate M. J. Timmons, R. A. Wyatt* and T. Murphy* Department of Plastic Surgery, Bradford Royal Infirmary, Bradford; and *Department of Oral Health and Development, Turner Dental School, University of Manchester, Manchester, UK SUMMARY The speech of children with isolated cleft palate (CP) repaired by one surgeon has been compared with the speech of children with some form of unilateral cleft lip and palate (CLP) repaired by the same surgeon. All palate repairs included an intravelar veloplasty. We identified 57 children (5-12 years old) with cleft palates repaired in infancy, of which three patients with other medical problems were excluded. Of the 54 patients, 44 (81%) attended for review (27 CP, 17 CLP). Video recordings were analysed by two speech and language therapists, using the Cleft Audit Protocol for Speech. The CP patients had no evidence of permanent fistulas. Final speech outcomes were similar for CP and CLP patients. Intelligibility was normal in 10 (37%) CP and nine (53%) CLP patients. Mild consistent hypernasality was present in five (18.5%) CP and four (23.5%) CLP patients. No patients had moderate or severe hypernasality or nasal emission. Mild consistent hyponasality was present in five (18.5%) CP and five (29%) CLP patients. Moderate consistent hyponasality was present in one (4%) CP patient. Dysphonia was present in eight (30%) CP and seven (41%) CLP patients. Cleft-type characteristics were noted in 11 (41%) CP and nine (53%) CLP patients. No CLP patients but 10 (37%) CP patients had required a pharyngoplasty (P = 0.004, Fisher's exact test). Possible reasons for this (age, cleft type, surgeon and surgery) are discussed The British Association of Plastic Surgeons Keywords: cleft palate, speech, Cleft Audit Protocol for Speech, fistula. An isolated cleft palate is the commonest type of cleft lip and palate deformity seen in many cleft units. Outcome studies such as the Eurocleft six-centre international study I and the Clinical Standards Advisory Group (CSAG) study in the UK 2 have assessed the treatment of complete unilateral cleft lip and palate. The conclusions of these studies do not necessarily apply to isolated cleft palate patients, who do not have the additional variables of a cleft lip and a cleft alveolus. Outcome studies of surgery for isolated cleft palate are equally important. Since this surgery is primarily undertaken to achieve normal speech, the speech of a group of isolated cleft palate patients has been investigated by two speech and language therapists from another unit and compared with the speech of patients who had some form of unilateral cleft lip and palate. Patients and methods From a register of all patients with primary cleft lip or palate repaired by the first author (MJT), all the children from the beginning of his consultant practice who had an isolated cleft palate or some form of unilateral cleft lip and palate repaired in infancy, and who were aged 5 years or more at the beginning of the study, were identified (n = 57). Of these 57 patients, three were excluded because they had other medical problems (severe mental handicap and other abnormalities; XXXXY syndrome; and foetal alcohol syndrome). Of the remaining 54 patients, 33 (61%) had cleft palate only and 21 (39%) had cleft lip and palate. Patients with submucous cleft palates were not included. In the isolated cleft palate group, 14 patients (42%) had been recorded as having the Pierre Robin sequence. This meant that they were considered to have a significantly small mandible, which might cause anaesthetic problems and difficulties with surgical access; they did not necessarily have the other features of the 'classical' Pierre Robin sequence. Ten of the 54 patients were not brought for review or had moved away; 44 patients (81%) were assessed. Of these 44 patients, 27 (61%) had cleft palate only and 17 (39%) had cleft lip and palate. In the isolated cleft palate group, 13 (48%) had the Pierre Robin sequence. Of the 17 cleft lip and palate patients, 11 had a complete unilateral cleft lip and palate, two had an incomplete cleft lip and a complete cleft palate, and four had a complete cleft lip and an incomplete cleft palate. The 44 patients had a wide range of backgrounds and abilities. Some of the children appeared to have above-average intelligence, while others had special educational needs because of below-average intelligence. Of the 44 patients, 14 (32%) were from Asian families (i.e. with Pakistani or Indian names), a similar proportion to that of Asian patients (n = 17, 31%) in the group of 54 patients. Surgery The unilateral cleft lips were repaired at 2-6 months of age (mean: 3.2 months). The palates were repaired at similar ages in the two groups: isolated cleft palate at 5-14 months (mean: 9.6 months), cleft lip and palate at 6-12 months (mean: 8.9 months). All the soft palates were repaired using an intravelar veloplasty, with transposition of the muscles and closure of the soft palate in three layers. The isolated cleft palates 377

2 378 British Journal of Plastic Surgery were repaired in one operation. At the beginning of the series, the hard palate was repaired using Veau flaps. In subsequent patients the hard palate was closed using bipedicled Langenbeck flaps whenever possible. In some cases, one of the Langenbeck flaps was converted to a Veau flap to achieve closure, a technique also used by /~byholm 3 and Millard et al.4 For the isolated cleft palate patients, the repairs were Veau (9), Langenbeck (11), medial Langenbeck (1), Veau/Langenbeck (4) and midline (2). For the cleft lip and palate patients, the repairs were Veau (5), Langenbeck (9) and Veau/Langenbeck (3). I z~ate:,,,, ) ~t No: Noralal I)iffeleni fi, onl olt~ ehildrlm'~ ~il~el~ hot noi ellolllll io ~ilae exlnlmellt m'~'on~s~son~ I n m m :~ mm I c~c~ I U I I ~ I m'uimm" '"II" Normal I II- l)vsljbonic i eo~-~t: IHNI,--- ~, ALVEO/.~IR [] Pal~l~ / ~lml 9 ~..,o.,~... I I I I m 8aO,~ to uv,,ttar I I Ph~rv,.j,al I I -.~ior Ond CTCi - Po~tedcr Oral C'l~s - Noc~atl CTCI Pud~ CTCI - D~,elopm~al ~ro~z Hi PGSI'~LAR t d, Y 'I d3 -- II m ~,~..~.~..~ m N.~ ~ I)\I~IlIN t Figure 1---Cleft Audit Protocol for Speech. Chart for patients' speech results. Available from the Department of Speech and Language Therapy, St Andrew's Centre for Plastic Surgery, Broomfield Hospital, Chelmsford, UK. (Reproduced with permission of A. Harding, K. Harland and R. Razzell). All patients had their palate repair assessed postoperatively at the time of discharge from hospital, in nearly all cases by the surgeon, and at their first postoperative outpatient clinic, in all cases by the surgeon. The case notes were reviewed to check for any record of a fistula due to wound breakdown and to ascertain whether the patient had a Secondary pharyngoplasty. Fistulas Since the quality of the surgical repair and complications with healing could affect speech, the isolated cleft palate patients were assessed by two orthodontists, as part of an orthodontic audit. 5 All patients and their parents were asked whether there were any problems, e.g. escape of liquid through the nose, that might indicate a palatal fistula. All patients had an intraoral examination. Speech A videotape recording of each child was made by a research speech and language therapist (TM) from another unit. She was one of the two research speech and language therapists who had recorded and analysed speech for the CSAG study. The same video recording equipment was used for this study. The ages of the two groups at the time of the recordings were similar: isolated cleft palate 5-12 years (mean: 7.6 years); cleft lip and palate 5-11 years (mean: 7.8 years). The children were recorded speaking English. Samples of rote speech, repetition of sentences from the Cleft Audit Protocol for Speech (CAPS) and spontaneous conversation were recorded. The videotapes were independently reviewed by two research speech and language therapists (TM, RAW). The speech results were summarised on CAPS charts (Fig. 1). Cleft Audit Protocol for Speech CAPS analyses the main features of speech from intelligibility to consonant characteristics (Fig. 1). For intelligibility, nasality, nasal airflow and grimace, the presence or absence of any abnormality and its severity are recorded as a number (Table 1). The results are then colour-coded from dark green (normal) to red (major concern) (Fig. 1). Table 1 Cleft Audit Protocol for Speech: keys for speech ratings Nasality and nasal airflow Hypemasal resonance/hyponasal resonance/audible nasal emission/nasal turbulence 0 not present 0-1 mild and occasional 1 mild and consistent 2 moderate and consistent 3 severe and consistent Grimace 0 not present 1 nasal flare 2 nasal grimace 3 facial grimace Summary of speech pattern no CTCs (cleft-type characteristics) anterior oral CTCs posterior oral CTCs non-oral CTCs passive CTCs developmental errors other 1. lateralisation/lateral 3. backing to velar 5. pharyngeal 8. weak/nasal consonants 2. palatalisation/palatal 4. backing to uvular 6. glottal 9. nasal realisations 7. active nasal fricatives 10. absent pressure consonants

3 Speech after cleft palate repair 379 'Voice', the presence or absence of distorted phonalion (dysphonia), but not its type or severity, is recorded and colour-coded. The presence or absence of any cleft-type characteristics is recorded as a number (Table 1). The cleft-type characteristics are colour-coded from light green (indicating articulation deficits not related to palatal function) to pink or red (a sign of some degree of velopharyngeal incompetence that needs further investigation and may require secondary surgery). In general, green indicates a favourable outcome, yellow indicates that there is a problem requiring at least monitoring, and pink and red indicate a poor outcome requiring further investigation. The exact interpretation of the colours varies, depending on which speech parameter is being assessed. Statistics Inter-observer agreement between the two speech and language therapists was assessed using the kappa statistic (k). A value of k= 0 indicates no agreement better than chance, k=l indicates that agreement is perfect and k = indicates a moderate agreement. 6 The pharyngoplasty results were compared with Fisher's exact test. Results Primary healing and fistulas The case notes record that one isolated cleft palate patient had a temporary postoperative fistula at the site of wound breakdown at the junction of the hard palate and soft palate, which had healed by the time of review one month after surgery; another patient was found to have an asymptomatic small fistula at the same site when examined under anaesthesia for secondary surgery. The orthodontic review of the isolated cleft palate patients found no symptoms or signs of palatal fistulas. 5 One cleft lip and palate patient (the first in the series) had a symptomatic fistula in the alveolus and adjacent anterior palate at the site of wound breakdown of part of a vomerine flap repair. This fistula was later closed. One cleft lip and palate patient had an asymptomatic pinholesized fistula of the soft palate. Speech audit All 44 videotapes were reviewed by TM and 41 were reviewed by RAW. Agreement between the two therapists was fair to very good; where they disagreed, it was nearly always by just one grade. For intelligibility there was complete agreement in 29 cases (71%) (k--0.54). For hypernasality, comparing normal (0) with mild (0-1 and 1), there was complete agreement in 25 cases (61%) (k= 0.38). For hyponasality, comparing normal (0) with mild (0-1 and 1) and moderate (2), there was complete agreement in 31 cases (76%) (k=0.59). For voice, there was complete agreement in 35 cases (85%) (k = 0.7). For the consonant articulation categories of normal (0) and cleft-type characteristics (1--4), there was complete agreement in 39 cases (95%); because of the large number of permutations of cleft-type characteristics for this very high level of inter-observer agreement, a single kappa value was not calculated. TM was more critical of speech problems than RAW. In CAPS and other protocols using scores with numbered categories, the results of two observers who are in close agreement cannot be averaged to a single figure. All the results presented here are from TM's assessments. The speech profiles for each patient are shown in Figure 2. The majority of speech problems occurred in a minority of patients. For example, four patients were only just intelligible to strangers (intelligibility grade: pink/3) and three of them were in the group of four patients with mild and consistent nasal emission (nasal emission grade: yellow/l). There was complete agreement between the two speech and language therapists about the poor intelligibility of the four patients. The four patients with poor intelligibility were younger than average, three of the four had isolated cleft palate repairs and none had been identified as having an obvious syndrome by their paediatricians. One had moderate general developmental delay and special educational needs. One had delayed speech and language development. One had social problems and a very poor attendance record at various clinics. Three of the four patients spoke English as a second language. However, of the 44 patients only a minority of those with English as a second language had significant speech problems on CAPS assessment. For other patients, belowaverage intelligence with special educational needs and poor social circumstances increase the risk of speech problems. Even so, good speech is still possible in such circumstances, e.g. patient SA with cleft lip and palate (Fig. 2B). Intelligibility was slightly worse in the isolated cleft palate group (Fig. 3). Direct measures of velopharyngeal competency, e.g. resonance, were very similar in the two groups (Fig. 3). No patients had moderate or severe hypernasality, nasal emission or nasal turbulence. Dysphonia was heard in eight (30%) of the isolated cleft palate patients and seven (41%) of the cleft lip and palate patients. No cleft-type characteristics were heard in 16 (59%) of the isolated cleft palate patients and eight (47%) of the cleft lip and palate patients (Fig. 4). The difference between the two groups was most noticeable for the articulatory errors of lateralisation and palatalisation, anterior oral cleft-type characteristics which were more common in the cleft lip and palate patients. PharyngopIasties To achieve these speech results, 10 (37%) of the isolated cleft palate patients, but none of the cleft lip and palate patients, had required a pharyngoplasty (P=0.004, Fisher's exact test). Of the original group of 54 patients, none of those whodid not attend for review had a pharyngoplasty. Of the 27 isolated cleft palate patients, 13 (48%) had the Pierre Robin sequence, which did not increase the risk of having a pharyngoplasty: five of the 10 pharyngoplasty patients (50%) had the Pierre Robin sequence. Nine of the l0 pharyngoplasty patients had clefts of the hard and soft palates. One child had ENT surgery, including an adenoidectomy, in a peripheral hospital; before this he had had slight hypernasality and nasal emission, but after the adenoidectomy his speech became worse.

4 380 British Journal of Plastic Surgery A CLEFT LIP AND PALATE B Figure 2--CAPS results lot (A) isolated cleft palate patients and (B) cleft lip and palate patients (patient idenlifying codcs in the leit-hand column). Six patients had posterior pharyngeal flap pharyngoplasties 7 and tour patients had Hynes pharyngoplasties with lateral mucosa/muscle flaps transposed, overlapped and inset, usually at the inferior edge of the adenoids, s Pharyngoplasties were done at mean age of 5.4 years (range: 4-8 years). Five patients were between 5 and 7 months old when they had their isolated cleft palates repaired; two later required a pharyngoplasty. Five patients were between 12 and 14 months old when they had their isolated cleft palates repaired; one later required a pharyngoplasty. Three phm'yngoplasty patients had residual mild and consistent hypernasality and three others had mild and consistent hyponasality. One patient had mild and occasional hypernasality and hyponasality and two others had mild and occasional hyponasality. Out of the 44 patients, one had moderate and consistent hyponasality (Fig. 3A); she had an isolated cleft palate but had not had a pharyngoplasty. Ten of the 16 patients with some degree of hyponasality after cleft palate repair had not had a pharyngoplasty.

5 Speech after cleft palate repair 381 B 100% 9O% 8O% 7O% 10% 0% ISOLATED CLEFT PALATE SPEECH ANALYSIS Hylxmul N. Er,~AI~ N, Tulbulence eck.nl~ CLEFT LIP & PALATE SPEECH ANALYSIS Int e41igl~flly Hygetnmllal Hypocalltr N E n'~on N. Tut~l~,oe Gdm~l~ Voice Figure 3--Intelligibility, nasality, nasal airflow, grimace and voice. (A) Isolated cleft palate patients: intelligibility, normal n= 10 (37%); mild consistent features: hypernasality n = 5 (18.5%), hyponasality n = 5 (18.5%), nasal emission n=3 (11%). (B) Cleft lip and palate patients: intelligibility, normal n = 9 (53%); mild consistent features: hypernasality n =4 (23.5%), hyponasality n=5 (29%), nasal emission n= 1 (6%). A ] 21!,2 B o 16 NO CTC$ 8 8 CLEFT-TYPE CHARACTERISTICS (CTCs) IN ISOLATED CLEFT PALATE PATIENTS 9 ~letlor oral CTC! po~d~l~ oral CTCs Non.rid CTCe 4 3 L--q... Anti,.or ~al CTCu po[dbr~o~ ~al CTCt Non~tal CTCe P~shta CTCs CLEFT-TYPE CHARACTERISTICS (CTCs) IN CLEFT LIP & PALATE PATIENTS passive CTCs Figure 4--Cleft-type characteristics for (A) isolated cleft palate patients and (B) cleft lip and palate patients. Discussion The main reason for repairing an isolated cleft palate in a child is to enable the child to develop normal speech. Normal speech is a complex mixture of sounds. There is no universally agreed scheme for describing and measuring resonance, consonant production and other features of normal speech and abnormal speech, such as can occur with a cleft palate. 9 Dalston et al proposed that the minimum parameters of speech that should be reported in studies of cleft palate surgery are: nasality (hypernasality and hyponasality), nasal air escape, intelligibility and articulation. 1~ Even when these parameters are assessed, comparing studies is difficult because the parameters are often rated on different scales, for example hypemasality was measured on a four-point scale by Semb et al u and on a seven-point scale by Morris et al. 12 Sell et al proposed a protocol for use in British cleft units: GOS.SP.ASS (Great Ormond Street Speech Assessment). t3 A survey of its use revealed significant ambiguities in the protocol, which was therefore revised in 1998 (GOS.SP.ASS.'98). 14 GOS.SP.ASS was considered too detailed to be used nationally as a protocol for audit and so CAPS was developed. The two protocols share many features, for example the list of sentences for speech sample elicitation and the summary of cleft-type characteristics. CAPS has now been accepted as the basis for speech audit by many cleft units in the UK. Unlike the GOS.SP.ASS protocol and others such as those used by McWilliams et a115 and Morris et al, 12 CAPS includes an assessment of intelligibility, as recommended by Dalston et al. l~ Intelligibility is difficult to judge because it is influenced by many variables in addition to features such as hypernasality, for example accent and the rate of speaking, the listener's acuity and experience in assessing speech, and external factors such as the context and background noise. 9 However, although intelligibility alone should never be used to report speech results, the intelligibility rating does provide an overall assessment of the speech of a child and is useful for comparison with other children. CAPS records the presence or absence of dysphonia. Dysphonia, particularly a hoarse or weak breathy voice, has many possible causes including velopharyngeal inadequacy. In our study 15 of the 44 patients had dysphonia; in most cases there were other clear signs of velopharyngeal incompetence but in a few cases the other signs were mild or absent. The CAPS colour-coding system gives a clear display of the features of speech for each patient and for groups of patients. The significance of the colours depends on what features are being assessed. The colours indicate the severity and, to some extent, the frequency as well as the presence or absence of any nasality, but only indicate the presence or absence of cleft-type characteristics. Inter- and intra-observer reliability is essential for CAPS and other systems to assess speech, e.g. GOS.SP.ASS. lntra-observer reliability was not assessed in our study. Inter-observer reliability depends on the training and expertise of the assessors and the use of standard high-quality equipment to make and to listen to the speech recordings. Even the conditions for listening should be standardised. One factor which may have affected the difference between the two speech and language therapists in our study (TM and RAW) was that not only was TM more experienced in using the CAPS system but she also sometimes listened to the video recordings with headphones while RAW listened to the

6 382 British Journal of Plastic Surgery recordings with loudspeakers. Inter- and intra-observer reliability also depends on what is being assessed. As our study indicates, complete agreement is more likely with a severe speech problem or with the presence or absence of a cleft-type characteristic. Differences between assessors have also been found in other studies of speech. In assessments of 13 speech recordings using the original GOS.SRASS system, Sell and Harding agreed on the presence or absence of hypernasality in 84% of cases, hyponasality in 92% of cases and nasal air escape in 77% of casesj 3 They did not attempt to measure the severity or consistency of any of the speech parameters; if they had, the differences between them may have been more marked. Detailed assessment of speech after cleft palate repair is essential, but the potential for some difference between assessors using CAPS or other systems has to be taken into consideration when interpreting the results of audits in either a single unit or in different units. In our study using CAPS, the isolated cleft palate patients had a similar outcome to the unilateral cleft lip and palate patients. None of the patients had moderate or severe hypernasality, about 20% in each group had mild consistent hypernasality and a few had mild and consistent nasal escape. Intelligibility was worse in the isolated cleft palate patients, with eight out of 27 (30%) of them, but only two out of 17 (12%) of the cleft lip and palate patients, being different enough to provoke comment or being only just intelligible. The isolated cleft palate group tended to have fewer cleft-type characteristics, especially anterior oral lateralisation and palatalisation. At the time of our study, the only published work on the use of CAPS to assess outcomes after cleft palate surgery was by the CSAG. 2 They carried out an audit of 5-year-old and 12-year-old children who had a complete unilateral cleft lip and palate repaired in the UK. They used a modification of CAPS, and only reported intelligibility, consonants 'sufficiently serious to require therapy and/or surgery' and 'minor' errors, and hypernasality on a simplified scale (normal tone; mild/inconsistent hypernasality; appreciable hypernasality). Caution is needed when comparing percentages because of the different numbers of patients and the different age groups in the two studies. Also, for reasons that were not explained, the numbers in the CSAG study varied: for example, 453 children had intelligibility scores but only 397 children had nasality scores. Our intelligibility ratings of 37% normal in isolated cleft palate patients and 53% normal in cleft lip and palate patients (average ages 7.6 years and 7.8 years, respectively) compare favourably with the CSAG results. Intelligibility was normal in 19% of all 5-year-olds and 47% of 5-year-olds in the best unit of those that entered 10 or more 5-year-olds into the CSAG study. 2 Intelligibility was normal in 47% of all 12-year-olds but was not reported for the best unit. 2 In the CSAG study there was moderate or severe hypernasality in 17% of 5-year-olds and 15% of 12-yearolds. The nasality ratings in our patients are comparable with the best unit in the CSAG study in that none of the patients had moderate or severe hypernasality, but not so good in that 88% of 5-year-olds and 100% of 12-year-olds in the best unit in the CSAG study had normal resonancefl The final outcome for our patients, with 18.5% of isolated cleft palate and 23.5% of cleft lip and palate patients having mild consistent hypernasality, is comparable with the outcome for 98 children at the age of 5 years, 16% of whom had some degree of consistent hypernasality after a primary cleft Palate repair by Sommerlad in the years 1991 to a6 Pharyngoplasties The indications for pharyngoplasty vary between centres and pharyngoplasty rates only give a general impression of the results of cleft palate surgery. 15 The overall pharyngoplasty rate for our patients was 23%, compared with a 45% secondary pharyngeal flap rate after intravelar veloplasty reported by McWilliams et all5 and a 15.4% secondary surgery rate reported by Sommerlad for all patients in his first 5 years of consultant practice. 17 However, our study emphasises the importance of distinguishing between cleft types because all the pharyngoplasties were in the isolated cleft palate group, of whom 37% had pharyngoplasties. This is comparable to other series, with pharyngoplasty rates of 25% for 24 Veau II (hard-soft) cleft palates, % for 40 isolated cleft palates in patients aged between 11 and 25 years and 16.7% for 18 isolated cleft palates in patients aged between 5 and 10 years. 12 One of our patients required a pharyngoplasty because his speech had deteriorated after he had an adenoidectomy in a peripheral hospital. What were the reasons for the failure of primary surgery in the other isolated cleft palate patients? Age Age at the time of primary cleft palate repair was not the cause. The mean ages of the two groups at the time of primary surgery were similar. Pharyngoplasty rates were similar for repairs at age 5-7 months and at age months. The mean age at the time of pharyngoplasty was 5.4 years (range: 4-8 years). While in some units this may be considered late, in others it may be considered early. However, even if pharyngoplasties were advised at an earlier age and for more moderate problems than in other units, this does not explain the marked difference in the pharyngoplasty rates between cleft types in our study. The mean ages of the two groups were similar and so the indications for pharyngoplasty were the same during the follow-up of both groups. Primary healing and fistulas The need for a pharyngoplasty was not due to poor healing or a fistula. Primary healing after surgery was good and there were no permanent fistulas. These results are reassuring but not exceptional. In 1963, Jolleys and Savage reported 91% primary healing after 57 cleft palate repairs (with normal preoperative nose and throat swabs); it should be noted that nearly all their cases had modified Langenbeck repairs, closing the soft palate onlyj 9 Others have reported fistula rates due to wound breakdown of 6.5% 2o and 0%. 12

7 Speech after cleft palate repair 383 Isolated cleft palate patients Patients with isolated cleft palate may have other congenital anomalies, some as part of specific syndromes. The reported incidence of these syndromes varies between centres. 12 At the time of our study, children with cleft lip and palate were not routinely screened by a geneticist. Three children (5%) were excluded from our study because of coexisting medical problems. There were no recognised cases of velocardiofacial syndrome, Stickler syndrome or other syndromes in the patients whose speech was assessed. Some studies exclude patients with the Pierre Robin sequence. 18 The outcomes of cleft palate surgery are just as important for these and other excluded patients as for the patients included. We were as inclusive as possible and this probably contributed to the number of unsatisfactory speech results in our study. In our study, isolated cleft palate patients had poorer results after primary surgery than did cleft lip and palate patients. There have been many studies correlating speech results with the severity of the cleft palate (Veau classification: I (soft palate), II (hard and soft palates), III (unilateral cleft lip and palate), IV (bilateral cleft lip and palate)). 18'2~ Some have found no difference between cleft types and others have found that the more severe the cleft the worse the speech. Our results did not show a simple correlation between the Veau classification and speech. Instead they matched the results of studies which have shown that patients with Veau II cleft palates have worse speech after primary surgery than those with Veau III cleft palates. For example, Marrinan et al reported pharyngoplasty rates of 25% for Veau II and 11% for Veau III after either Langenbeck or Veau-Wardill-Kilner repairs. 18 Kirschner et al reported completely normal resonance in 87% of Veau III patients and 85.7% of Veau IV patients but in only 68.4% of Veau I and II patients after a modified Furlow repair. 21 7he surgeon and the surgery The patients in our series were treated by the surgeon at the beginning of his consultant practice over a period when he was a relatively low-volume operator. However, the trend for isolated cleft palate patients to have more speech problems after primary surgery than unilateral cleft lip and palate patients has continued, over a period when the annual number of cleft palate operations has increased. Treating large numbers of patients does not, of course, necessarily produce good results. What is interesting is that with the same surgeon doing the same soft palate operation the results were satisfactory for the cleft lip and palate patients but not for the isolated cleft palate patients. Isolated cleft palates are more sensitive to any imperfections in the operative technique, and to improve the speech results for isolated cleft palate patients the technique needs to be improved. This will not be achieved just by altering the hard palate surgery, which is what most of the debate about eponyms like Veau, Langenbeck, Delaire, Malek, etc. is concerned with. Speech depends much more on the muscles of the soft palate. If the cleft palate muscles are not repositioned or are repositioned with only a limited intravelar veloplasty, then only a minority of patients will achieve completely normal speech, 22 Better results are achieved with more complete transposition of the palatal muscles. Sommerlad et al showed that secondary muscle correction in previously repaired cleft palates reduced hypernasality and nasal escape, z3 Sommerlad has also developed a more radical muscle correction for primary cleft palate repairs and has used an operating microscope during the surgery since For 483 cleft palate repairs over three 5-year periods between 1978 and 1992 his secondary surgery rates were 15.4%, 7.2% and 2.8%, respectively, which he ascribed to his muscle correction becoming more radical and refined. 17 The Furlow technique, using flaps in the soft palate which allow the muscles to be easily seen, also results in a complete transposition of the soft palate muscles. Randall and La Rossa reported a significant improvement in speech outcomes after they changed from repairs including an intravelar veloplasty to Furlow repairs, ~5 and their hospital continues to achieve good speech results with a modified Furlow technique. In a study of these results, Kirschner et al noted the different resonance rates for Veau I-IV cleft palates mentioned above and the need for secondary pharyngeal flap surgery in 14% of Veau I and II clefts, 4.3% of Veau III clefts and 3.6% of Veau IV clefts. 21 They attributed the results to the Furlow technique rather than to the increasing skill of the two senior surgeons, since the results were equally good for experienced and relatively inexperienced surgeons in their hospital. In summary, it seems that improved visibility of the soft palate muscles, using the operating microscope, or Furlow flaps, or both, and precise dissection and reorientation of the muscles, with either Sommerlad's or Furlow's technique, will enable more children with isolated cleft palates to develop normal speech. Because of this, the operating microscope is now being used during primary cleft palate repairs and cleft palate re-repairs 23 by the first author (MJT). Conclusions There are different types of cleft palate, and studies of speech after cleft palate surgery should distinguish between the types of cleft palate being assessed. In our study, isolated cleft palate patients had more problems with speech after primary surgery than unilateral cleft lip and palate patients. Possible reasons for this have been discussed but some of the questions raised will be best answered by prospective studies of large numbers of patients treated in a few centres. 2 Pharyngoplasty rates alone are a crude indicator of speech outcomes. CAPS is a useful system for assessment of speech after cleft palate surgery. Its main drawback is that it does not indicate the severity or frequency of clefttype characteristics. CAPS is easy for speech and language therapists to use, all major parameters of speech are measured and the colour-coding makes it easier to report and assess the results. Routine prospective CAPS data collection will make it easy to carry out audits in each unit and to compare results in different units. This study is a retrospective audit of speech using CAPS and, based as it is on the results from the first

8 384 British Journal of Plastic Surgery 5 years of a surgeon's consultant practice, it can perhaps be considered to show the minimum standards for speech outcomes after the repair of isolated cleft palates. Acknowledgements We wish to thank L. Strachan, N. Hughes (speech and language therapists), A. Broughton (specialist health visitor) and A. Ryall (audit administrator) for help with organising this study; Simon Littlewood and Laura Mitchell, Department of Orthodontics, St Luke's Hospital, Bradford, for the orthodontic audit; colleagues in the Turner Dental School, Manchester, for the statistical analyses; G. Spyrou for the computer graphics; and the Department of Medical Illustration, Bradford Hospitals NHS Trust, for the illustrations. This study was partly funded by the Plastic Surgery and Burns Research Unit, University of Bradford. References 1. Shaw WC, Dalai E, Asher-McDade C, et al. A six-center international study of treatment outcome in patients with clefts of the lip and palate: Part 5. General discussion and conclusions. Cleft Palate Craniofac J 1992; 29: Clinical Standards Advisory Group. Cleft lip and/or palate. London: The Stationery Office, 1998, 3. Abyholm FE. Discussion at the Meeting of the British Association of Plastic Surgeons Cleft Group, Manchester, October Millard DR Jr, Flynn W, Rao MP. Combining the von Langenbeck and the Wardill-Kilner operations in certain clefts of the palate. Cleft Palate Craniofac J 1992; 29: Litflewood S J, Mitchell L, Timmons MJ. Orthodontic audit of isolated cleft palate patients after surgical repair. (Poster). The Sixth European Craniofacial Congress, Manchester, June Altman DG. Practical Statistics for Medical Research. London: Chapman & Hall, Owsley JQ Jr. Lawson LI, Miller ER, Blackfield HM. Experience with the high attached pharyngeal flap. Plast Reconstr Surg 1966; 38: Hynes W. The results of pharyngoplasty by muscle transplantation in "failed cleft palate" cases, with special reference to the influence of the pharynx on voice production. Ann R Coll Surg Engl 1953; 13: Wyatt R, Sell D, Russell J, Harding A, Harland K, Albery E. Cleft palate speech dissected: a review of current knowledge and analysis. Br J Plast Surg 1996; 49: Dalston RM, Marsh JL, Vig KW, Witzel MA, Bnmsted RM. Minimal standards for reporting the results of surgery on patients with cleft lip, cleft palate, or both: a proposal. Cleft Palate J 1988; 25: Semh G, Borchgrevink H, Saether I-L, Ramstad T. Multidisciplinary management of cleft lip and palate in Oslo, Norway. In Bardach J, Morris HL, eds. Multidisciphnary Management of Cleft Lip and Palate, Philadelphia: W.B. Saunders, 1990: Morris HL, Bardach J, Ardinger H, et al. Multidisciplinary treatment results for patients with isolated cleft palate. Plast Reconstr Surg 1993; 92: Sell D, Harding A, Grunwell P. A screening assessment of cleft palate speech (Great Ormond Street Speech Assessment). Eur J Disord Commun 1994; 29: Sell D, Harding A, Grunwell P. GOS.SP, ASS.'98: an assessment for speech disorders associated with cleft palate and/or velopharyngeal dysfunction (revised). Int J Lang Commun Disord 1999; 34: McWilliams B J, Randall P, LaRossa D, et al. Speech characteristics associated with the Furlow palatoplasty as compared with other surgical techniques. Plast Reconstr Surg 1996; 98: Mehendale FV, Hattee C, Sommerlad BC. Predicting the results of primary cleft palate repair. Summer Meeting of the British Association of Plastic Surgeons, Birmingham, July Sommerlad BC. Towards a more functional palate repair. Hunterian Lecture, Royal College of Surgeons of England, London, December Marrinan EM, LaBrie RA, Mulliken JB. Velopharyngeal function in nonsyndromic cleft palate: relevance of surgical technique, age at repair, and cleft type. Cleft Palate Craniofac J 1998; 35: Jolleys A, Savage JP. Healing defects in cleft palate surgery--the role of infection. Br J Plast Surg 1963; 16: Emory RE Jr, Clay RP, Bite U, Jackson IT. Fistula formation and repair after palatal closure: an institutional perspective. Plast Reconstr Surg 1997; 99: Kirschner RE, Wang P, Jawad AF, et al. Cleft-palate repair by modified Fuflow double-opposing Z-plasty: the Children's Hospital of Philadelphia experience. Plast Reconstr Surg 1999; 104: Marsh JL, Grames LM, Holtman B. Intravelar veloplasty: a prospective study. Cleft Palate J 1989; 26: Sommedad BC, Henley M, Birch M, Harland K, Moiemen N, Boorman JG. Cleft palate re-repair--a clinical and radiographic study of 32 consecutive cases. Br J Plast Surg 1994; 47: The Authors M. J. Timmons MA, MChir, FRCS, Consultant Plastic Surgeon Department of Plastic Surgery, Bradford Royal Infirmary, Bradford, West Yorkshire BD9 6R J, UK. R. A. Wyatt BSc, MRCSLT, Research Speech and Language Therapist T. Murphy BSc, MRCSLT, Research Speech and Language Therapist Department of Oral Health and Development, Turner Dental School, University of Manchester, Manchester, UK, Correspondence to Mr M. J. Timmons. Paper received 26 September Accepted 26 February 2001, after revision.