Theoretical accounts and clinical management of pediatric speech sound disorders (SSD) are limited by previous research. Participants' speech difficulties have been inadequately described, reflecting the lack of clarity in existing diagnostic guidelines. Performance measures have primarily focused on the articulation of consonants in single words rather than phonological competence and the cognitivelinguistic abilities underlying speech development. Evidence-based practice reviews conflate studies of heterogeneous speech-disordered populations who received widely differing amounts of therapy using different intervention approaches. More recently, however, researchers have begun to explore assessment measures that allow differential diagnosis of subgroups of SSD in terms of qualitative analyses of speech errors and underlying cognitive linguistic abilities. These measures have allowed long-term follow-up to better predict which children will have future literacy difficulties and to identify specific underlying deficits that inform intervention.

Based on the Middlesborough data, the estimated national incidence rate of referrals who attend for assessment and who have speech and language disability is 85 000-90 000 children per year (14.6% of births). While findings from only one Primary Care Trust must be treated cautiously, they provide paediatric speech and language therapy services managers with information that might guide service planning.

Features indicative of best practice are highlighted and tentative suggestions made that would enable services to address the challenges of serving bilingual paediatric caseloads.

Longitudinal case studies of the successive phonological acquisition of two Cantonese–English bilingual children, aged 2;3 to 3;1 years and 2;9 to 3;5 years, are presented. The children were assessed at 4-week intervals. The first assessment of their phonology occurred when they had been exposed to English for three months. Phoneme acquisition and phonological process data revealed that both children had separate phonological systems for the two languages. The two phonological systems for each child developed in similar ways to monolingual children acquiring Cantonese and English. However, a number of error patterns, indicative of disorder in monolingual children, were evident in the children's phonological systems in English and in Cantonese. These patterns have been documented as normal error patterns for successive bilingual Cantonese–English speaking children. The difference between normal successive bilingual phonological development and normal monolingual development is addressed.

The DAPPLE (Dynamic Assessment of Preschoolers’ Proficiency in Learning English) is currently being developed in response to a clinical need. Children exposed to English as an additional language may be referred to speech and language therapy because their proficiency in English is not the same as their monolingual peers. Some, but not all, of these children are likely to have a core language learning difficulty. Clinicians need to be able to distinguish disorder from difference due to a child’s language learning context. The assessment used a test–teach–test format to examine children’s ability to learn vocabulary, sentence structure and phonology. The assessment, which takes less than 60 minutes to administer, was given to 26 children who were bilingual: 12 currently on a speech and language therapy caseload and 14 children matched for age and socio-economic status who had never been referred to speech and language therapy. The DAPPLE data clearly discriminated the two groups. The caseload group required a greater amount of prompting to identify targeted words in the receptive vocabulary assessment and performed less well in the post-teaching expressive component. For sentence structure, the caseload group required more cues to acquire the targeted clause elements in the teaching phase. The caseload group made more phoneme errors at the initial and final assessments than the controls, and the type of errors made differed. Teaching resulted in greater positive change in percent phonemes correct for the caseload participants. Qualitative analyses of individual children’s performance on the DAPPLE suggested that it has the potential to discriminate core language deficits from difference due to a bilingual language learning context. Future directions for development of the test are considered.