Differential Diagnosis and Treatment of Children with Speech Disorder

Barbara Dodd

• Publisher: Wiley
• Year: 2013
• ISBN:
  9781118713334
• Edition: 2
• Topic: Medicine
• Subtopic: Audiology & Speech Pathology

Paediatric speech and language therapists are challenged by diminished resources and increasingly complex caseloads. The new edition addresses their concerns. Norms for speech development are given, differentiating between the emergence of the ability to produce speech sounds (articulation) and typical developmental error patterns (phonology). The incidence of speech disorders is described for one UK service providing crucial information for service management.

The efficacy of service provision is evaluated to show that differential diagnosis and treatment is effective for children with disordered speech. Exploration of that data provides implications for prioritising case loads. The relationship between speech and language disorders is examined in the context of clinical decisions about what to target in therapy.

New chapters provide detailed intervention programmes for subgroups of speech disorder: delayed development, use of atypical error patterns, inconsistent errors and development verbal dyspraxia.

The final section of the book deals with special populations: children with cognitive impairment, hearing and auditory processing difficulties. The needs of clinicians working with bilingual populations are discussed and ways of intervention described. The final chapter examines the relationship between spoken and written disorders of phonology.

PART I

UNDERSTANDING SPEECH-DISORDERED CHILDREN

CHAPTER 1

Children with speech disorder: defining the problem

BARBARA DODD

Communication disability can be defined as an impaired ability to use spoken and written language to express thought or to understand others’ language. Most children who are referred for clinical assessment of a communication difficulty have a speech disorder. Their speech is difficult or impossible to understand because it is characterized by many mispronunciations of words. However, these children are far from being a homogeneous group. They differ in terms of the severity of their difficulty, the underlying cause of the disorder, the characteristics of their speech errors, the degree to which other aspects of their language, such as syntax, semantics and pragmatics, are involved, and their response to treatment. They also differ in terms of their response to their impaired ability to communicate: some seem unaware of their lack of intelligibility; others withdraw socially or become overtly frustrated by their difficulty in making themselves understood.

Broomfield and Dodd’s (2004a) incidence survey in the UK found that 6.4% of otherwise normal children had a speech disorder in the absence of any other sensory, cognitive or physical difficulty. Incidence is the number of new cases referred in a given population during a specified time (Enderby and Phillip, 1986). Prevalence figures (‘the total number of people with [a disorder] at any one time in a given population’, Enderby and Phillip, 1986, p. 152) for speech disorder range from 2% to 25% of the normal preschool/school population (e.g. Kirkpatrick and Ward, 1984; Enderby and Phillip, 1986; Shriberg et al., 1999; Law et al., 2000). To these children, whose disorder is specific to speech and/or language, must be added those whose speech difficulty is part of a more general handicap, such as hearing impairment or physical or intellectual disability. Chapter 5 describes the epidemiology of speech disorder, examining factors that might place children at risk: gender, socio-economic status, family history of communication impairment and family size.

There are, however, difficulties with most of the epidemiological data available. Different research groups use different criteria for the identification of speech disorder (Broomfield and Dodd, 2004b). In addition, the term ‘speech disorder’ encompasses a heterogeneous population. It includes, among others, children who have a lisp (i.e. misarticulation of /s/) but whose speech is intelligible, those whose speech is unintelligible due to omissions and substitutions of speech sounds in words but who can articulate all sounds perfectly in isolation, those born with an anatomical anomaly, such as cleft palate, who develop disordered speech despite surgical repair, children who have had earlier periods of impaired hearing but who currently have no hearing loss, children with motor speech disorders, children who have suffered emotional trauma and children from impoverished language-learning environments. Shriberg (2003, p. 502) argues that ‘accurate differential diagnosis of a patient’s disorder, including information on both original and maintaining causes, is necessary to determine the optimum form and content of treatment’. Despite consensus about the need for a classification system for developmental speech disorders (e.g. US National Institute of Health’s 2003 call for research on classification), as yet there is no agreed approach to classification that would allow better clinical management.

Approaches to the classification of speech-disordered children

Age of acquisition

In some cases, it is obvious at birth that a child is at serious risk for a later speech disorder (e.g. children with intellectual disability such as Down syndrome, those with hearing impairment, anatomical anomalies and physical disabilities, such as cerebral palsy). These are congenital disorders. Other children’s disorders emerge during the first years of life, when they fail to develop speech at the appropriate age, their errors are atypical of normal development or their rate of development is so slow that their phonology becomes delayed in comparison to that of their peers. Most children are referred for assessment of a speech disorder during their third or fourth year (see Chapter 5). This group is categorized as having a developmental disorder. In some cases, children whose speech and language development has followed a normal path acquire a speech disorder due to accident (e.g. head injury) or illness (e.g. meningitis leading to hearing loss). Thus, one major classificatory division is between congenital, developmental and acquired disorders.

Severity

A study of 20 speech-disordered children showed that the percentage of consonants produced correctly on a standard assessment ranged between 21% and 98% (Garrett and Moran, 1992). It seems obvious that one simple way of categorizing children with speech disorder is in terms of the severity of their disorder: mild, moderate and severe (Shriberg et al., 1997b). However, in most clinical reports the severity rating assigned is subjective and dependent on the clinician’s experience. They have yet to agree criteria for labelling severity of a particular type of language sample (e.g. imitation vs. conversation vs. picture naming) in terms of the three major categories and their hybrids (mild-moderate and moderate-severe).

Some procedures provide arbitrary cut-off points between categories in terms of the number of consonants in error. Shriberg et al. (1997a) report that the percentage consonants correct (PCC) metric, calculated from a 5–10-minute conversational speech sample, is psychometrically robust. The results are categorized such that a PCC of > 90% indicates a mild classification, 65–85% indicates mild-moderate impairment, 50–65% suggests a moderate-severe impairment and < 50% indicates a severe speech disorder. Shriberg et al. (1997a) list concerns about the reliability and validity of the PCC metric. The speech sample obtained may be inadequate and, if conversational, certain highly frequent sounds (e.g. /s/) will be more heavily weighted. Omissions and substitutions will be weighted equally with distortion of speech sounds, although the three error types effect on intelligibility differs. Vowel sounds are not included; and there is a need for standardization data that take age and gender into account. While Shriberg et al. (1997a) provide data that alleviate some of these concerns, the PCC metric, although important in describing the level of difficulty, provides little useful information for differential diagnosis of subgroups of speech disorders.

The process of differential diagnosis was described by Peterson and Marquardt (1990) as the integration of information from the result of measurement of speech behaviour with contextual information (e.g. from case history and other professionals’ reports) to identify the causal and maintenance factors specific to an individual’s disorder. Identification of a cause–effect relationship allows the distinction between speech disorders that have similar surface characteristics, but differ in terms of prognosis (i.e. the need for and outcome of intervention), and the type of intervention that is appropriate. There seems to be no evidence that severity measures discriminate between subgroups of children with speech disorder in terms of the type of intervention indicated, or outcome.

Aetiology

The application of the medical model to the classification of communication disorders has a long tradition in speech and language pathology. It is important to identify the aetiology of a child’s speech difficulty, if that is possible. A major diagnostic distinction is between those children whose disorder is caused by organic factors and those for whom no organic aetiology can be identified. This latter group is often described as having a ‘functional disorder’. It is the role of physicians (and other professionals such as audiologists and clinical psychologists) to diagnose the disease states, neurological lesions and anatomical anomalies that disable the speech production mechanism (Perkins, 1977). In some cases the cause of the disordered speech is relatively easy to identify: hearing loss, anatomical anomalies (e.g. inadequate velo-pharyngeal closure leading to nasal emission), intellectual disability (e.g. Down syndrome) and neurological lesions leading to motor speech disorders (e.g. the dysarthria associated with cerebral palsy), or aphasia with phonological involvement. However, the proportion of speech-disordered children for whom a clear-cut organic cause can be identified is relatively small. Most children are eventually assigned to the ‘functional’ category (Gierut, 1998).

Shriberg (1982) argues that ‘functional’ is a default classification for children showing no significant deficits in structural, cognitive or psychological systems and that classification systems must be developed that include all children. He proposes that speech-disordered children should be diagnostically categorized in aetiological ‘families’:

• Speech mechanism (i.e. including subtypes where causality is associated with hearing, motor speech or craniofacial involvement).
• Cognitive-linguistic factors (i.e. including subtypes where causality is associated with general intellectual ability and receptive and expressive linguistic ability).
• Psychosocial factors (i.e. including both caregiver and school input, plus child-specific factors such as aggression and maturity).

The major difficulty associated with aetiological classification systems is that it is rarely possible to establish a single causal factor. Fox et al. (2002) attempted to categorize 66 speech-disordered children according to Shriberg’s (1982) system. Around half the children were unable to be classified under any one of the causal factors listed. For example, in one case study of a phonologically disordered child (Leahy and Dodd, 1987), three possible causal factors were apparent:

1. The child had had a series of middle-ear infections during early childhood, and although several audiological assessments showed no significant loss, it ...