Research

[A dissertation submitted in fulfilment of the requirements for the degree of Doctor of Philosophy in the School of Clinical Speech and Language Studies, University of Dublin, Trinity College, and funded by The Child Health Foundation, Temple Street.]
Date: July 2000

Summary:
This study aimed to develop a reliable and valid perceptual profile for the assessment of nasality and nasal airflow errors in speech. To date, clinicians rely on simple categorical or numerical scales of nasality and nasal airflow errors to assess speech. Such scales provide limited information on the type or degree of the presenting speech problems. This new scale aimed to describe the nature and the degree of these characteristics, thus improving reliability and validity. The second aim of the study was to assess the validity of the Perceptual Profile using instrumental assessment for investigations of nasality and nasal airflow errors and thus develop a protocol for assessment of nasality and nasal airflow errors in speech.

In order to develop the Perceptual Profile, working definitions of the terms used to describe nasality and nasal airflow errors were developed. The definitions formed the basis of the Perceptual Profile and qualitative descriptions of error categories were devised. The resulting scale was then tested rigorously for reliability, using percentage of agreement and kappa scores. Intra-rater reliability was assessed by rating the speech of twelve children presenting with nasality and/or nasal airflow errors twice from audio tape recordings. Results indicated good to excellent intra-rater reliability of the Perceptual Profile. Inter-rater reliability was assessed by comparing ratings of three Speech and Language Therapists (one specialist, one experienced and one inexperienced in the area). The speech samples of 20 children were analysed; percentage agreement and kappa scores were calculated for each pair of raters. Results indicated good inter-rater agreement for perceptual ratings of nasality and nasal airflow errors. However, kappa results were variable, ranging from good to poor. Agreements of ratings in the present study compared well with previous studies. The use of kappa analysis in the evaluation of reliability of nasality and nasal airflow assessment is new and comparison of the kappa analysis of the perceptual profile with other scales is not possible. The variable findings of reliability of the perceptual assessment demonstrates the need to supplement perceptual assessments with instrumental measurements.

Perceptual ratings of nasality were compared to instrumental measurements of nasality using the Nasometer, while perceptual ratings of nasal airflow errors were compared to instrumental measurements of flow and pressure using the PERCI SARS system. Normative data from instrumental measurements is required before instrumentation can be used to assess pathological speech. Normal pressure/flow measurements from 152 normal speaking children were obtained by Zajac (1998, personal communication). Previous studies have indicated the need to obtain local normative data for the Nasometer. In this study, seventy normal English-speaking Irish children were assessed on the Nasometer using a novel speech stimulus. The speech stimuli available to date have been found to be difficult for children (Watterson, Hinton & McFarlane, 1996) and the present study found that the American stimuli were culturally biased. The speech stimulus used for the present study allowed for nasometric analysis of sentences, which were categorised according to consonant type. The normative data obtained in the study will now provide baseline normal data for use of the Nasometer in Ireland.

Fifty children presenting with nasality and/or nasal airflow errors were assessed using the perceptual scale, the Nasometer and the pressure/flow system. Results were compared using correlational analysis, test sensitivity, specificity and overall efficiency ratings. Results indicated a strong relationship between perceptual ratings of nasality and nasalance scores on specific speech stimuli. There was a strong relationship between perceptual ratings of nasal emission and pressure/flow measurements, particularly nasal flow and velopharyngeal port area measurements. A weak relationship was found between pressure/flow measurements and perceptual ratings of the following nasal airflow errors: nasal turbulence, nasal fricatives and velopharyngeal fricatives.

Reliability and validity results identified sections of the Perceptual Profile that required revision. Overall results indicate that the Perceptual Profile has variable reliability, with acceptable agreement and, is a valid tool for the assessment of nasality and nasal airflow errors in the speech of children with velopharyngeal dysfunction. Results also indicate that the perceptual and instrumental assessment protocol is a valid and reliable assessment, which could be used in specialist centres for management of children with cleft palate and velopharyngeal dysfunction.

[Original Abstract]
Otitis media with effusion (OME) is common among children with cleft palate, and may lead to such long-term consequences as hearing loss, tympanic memberane retraction, and chronic otitis media (COM). In total, 104 children with cleft lip and/or palate treated for OME at our institution* were reiewed. Mean duration of follow-up was 6.9 years, and mean age at latest follow-up was 9.6 years. The incidence of COM was 19% and the incidence of cholesteatome was 1.9%. Ears showing such long term sequelae of OME as hearing loss, tympanic membrane retraction, and chronic otitis media, were noted to have undergone a significantly greater number of ventilation tube insertions than ears not showing these sequelae. Our findings would suggest that a conservative approach to the management of OME in children with cleft palate is more likely to be beneficial in the long term.

[Article Conclusion]
Otitis media with effusion is common in young children with cleft palate, however, as children get older, their otological status frequently improves. A higher incidence of tympanic membrane retraction, chronic otitis media , and hearing loss may be expected in ears treated with a greater number of ventilation tube insertions among children with cleft palate. However, a conservative approach to the treatment of otitis media with effusion in children is safe and without adverse long-term effects. Furthermore, aural rehabilitation in the form of hearing aids provides a feasible alternative to surgery. Our data would support a conservative approach to the treatment of OME in children with cleft palate. ventilation tubes should only be inserted when there is objective evidence of hearing loss, and of persistence of the effusion for more than three months, and only after the reasons for surgery and the non-surgical alternatives have been discussed with the child’s parents.

*The Children’s Hospital, Temple Street, Dublin, Ireland

The project consisted of 2 main parts. In the first part, the medical charts of all children with cleft lip and/or cleft palate undergoing plastic surgery at Temple St. between 1990 and 1995 were reviewed. 104 of these children also attended the ENT department at Temple St. because of otitis media with effusion (glue ear), and were followed up continuously by the ENT department for many years. The charts of these children were selected for detailed examination. The second part of the project consisted of a postal survey of the parents of all children who were registered on the cleft register at Temple St. in April 2002, and who had an address in Ireland. This register included children with cleft lip, cleft palate, and congenital VPI. 584 questionnaires were posted out, and 397 fully completed questionnaires were returned. The medical charts of all respondents were then reviewed.

Middle ear disease was found to be very common in children with cleft palate. Altogether, 71% of parents of children with cleft palate reported their child to have suffered from ear problems at some stage. This compares with 16% of parents of children with cleft lip only. Children with congenital VPI were also reported to have a high incidence of ear problems (63%).

The peak age of onset of ear problems in cleft palate was in the first year of life. However, only 31% of parents of children with cleft palate aged 0-1 years old reported ear problems. Ear problems were most prevalent in the 4-6 year old age group, with 56% reported to have suffered from ear problems in the preceding year, and 40% reported to have below normal hearing. Ear problems persisted at a high level right through to the 10-12 year old age group (46% ear problems in preceding year, and 46% below normal hearing). Ear problems tended to settle in older age groups, however, there was still a substantial incidence of ongoing problems (21-26%) and below normal hearing (24%).

58% of all children with cleft palate underwent grommet insertion at some stage. This percentage was higher in older age groups. Two-thirds of those children undergoing grommet insertion required to undergo this procedure on more than one occasion, and one-third required three or more sets of grommets. At long-term follow-up, those children who had received more sets of grommets were found to have poorer hearing and a higher incidence of chronic ear disease than those children who never underwent grommet insertion, or who underwent this procedure on only one occasion. The reason for this is not because grommets caused hearing loss or chronic ear disease; rather it reflects an increased severity of ear disease in some children with cleft palate, which makes them more likely to require to undergo grommet insertion on several occasions, and also makes them more likely to have persistent long-term problems.

Chronic ear disease (meaning the presence of permanent ear problems) developed in roughly one fifth of children with cleft palate. 8% of children with cleft palate ultimately required major ear surgery (mastoidectomy or repair of perforation), compared to only 2% of children with cleft lip alone. Children with congenital VPI also had a high incidence of requiring major ear surgery (18%).

The important observations from this project are as follows:

1. Not every child with cleft palate develops ear problems, although the vast majority do (71%).
2. Children with congenital VPI suffer from an incidence of ear problems which is similar to that seen in cleft palate. This risk of ear problems in VPI is not related to the presence of visible stigmata, such as bifid uvula or submucous cleft palate.
3. Although nearly all infants (children < 1 year old) with cleft palate are known to have fluid in their ears, most parents of such children in our survey reported no clinically apparent problems or hearing loss. The peak age incidence of ear problems was in the 4-6 year old age group, which is similar to that seen in children without clefts, however, unlike the case for children without clefts, whose ear problems usually resolve after this age, ear problems in children with cleft palate persisted right up to the age of 12 years.
4. Ear problems do settle with age for the majority of cleft palate children, and most ultimately end up with normal hearing, however, around one-quarter have persistent problems right up to adulthood.
5. Grommet insertion is very beneficial in children with glue ear in the short-term. They work by providing a communication between the middle ear and the outside atmosphere, thus allowing fluid in the middle ear to resolve, so improving hearing. However, grommets are designed to extrude (fall out) after 9-12 months. In children with more severe ear disease, the fluid often re-accumulates after this. These children may require further grommet insertion, but the grommet insertion does not alter the natural history of the ear disease, i.e. grommet insertion will not prevent long-term ear or hearing problems in the more severely affected children. Thus it is important for parents to realize that the objective of grommets in glue ear is to “buy some time”, while waiting for the child to hopefully “grow out” of the ear problems, and not to try to cure the underlying problem.

My thanks to my co-authors, Mr. Alexander Blayney, Mr. Michael Earley, Ian Miller, Anne McGillavery, and Jerome Sheahan (statistics),. My thanks also to all the parents who replied to the questionnaire, without whom this research would not have been possible. (Patrick Sheahan)

[Article furnished to the Cleft Lip & Palate Association of Ireland, March 2004]

During the period covered, 398 affected babies were born. This represents an overall occurrence of 1.28 per 1000 live births, or an incidence of 1:781. The incidence varies from 1:614 in 1985 to 1:938 in 1986. The report does not offer an explanation as to the wide variation from year to year.

The relative frequency of different types of clefts was reported as follows; primary palate only -16%, secondary palate only – 53%, complete cleft of both primary and secondary palate – 26%, and unconnected clefts of both primary and secondary palate – 5%. Although there were variations in the relative incidence of cleft types from year to year, the one consistent was that clefts of the secondary palate only were always in the majority.

The distribution of the type of cleft related to sex showed that significantly more males (56%) than females (44%) were affected. Only in clefts of the secondary palate were females in the majority (53% versus 47%). This difference was not noted as significantly different, however. Separate clefts of the lip and palate occurred exclusively in males with one exception (20 versus 1).

The report showed that of unilateral clefts, a significantly higher number (64%) affected the left side. It further showed that complete clefting of primary and secondary palate was significantly more left-sided in males and right-sided in females.

The report also compares the findings with those of other studies, concluding that the incidence of clefting in Northern Ireland is within the limits reported in the literature from the UK and elsewhere. Those limits, however, as reported and reiterated in this report vary from 1:89 per 1000 live births in Denmark (602 births in the period 1976-1981), to 1:30 in a Birmingham study covering 285 births over the period 1940-1950. The report with the largest number of births studied (Trent region – 930 births in the period 1973-1982) indicated an incidence rate of 1:82 per 1000 live births.