Examining speech of very‐low‐birthweight children during everyday activities

It has been more than 40 years since the introduction of neonatal intensive care which resulted in improved outcomes for very-low-birthweight (VLBW) children (birthweight <1500g and ⁄ or gestational age <32wks). Longitudinal studies following infants into adulthood document that although 90% of VLBW infants do not have overt neurosensory disorders, they continue to be in greater risk of lower intelligence, poorer academic performance, and social isolation. As a result of these disorders, VLBW children who were born in the US in the late 1970s were 74% less likely to complete high school by the age of 19 than their normal birthweight siblings. However, since the 1970s, medical advances in neonatal intensive care have resulted in a more than 50% reduction in hospital mortality rates of VLBW children and a likely decrease in the incidence and severity of cognitive and learning deficits. In their study, Van der Spek et al. examined early speech and language development in VLBW children. Their findings add to the substantial evidence that VLBW children continue to be at risk of intellectual, motor, and speech disorders. The positive news from this study of recently born children is that although the 2-year-old VLBW children had a greater number of phonological errors than their typically developing peers, there was no increase in unusual phonological error patterns, and the VLBW children’s language development, both comprehension and vocabulary, did not differ from their typically developing peers. If this early trend continues through their preschool years, it may be an indication that VLBW children born in the 21st century have a better educational prognosis than the previous generation. Longitudinal studies have shown that preschool-aged children with isolated expressive phonological disorders appear to be clinically distinct from their peers with co-occurring phonological and language disorders and typically have better educational outcomes. Children with isolated phonological disorders are primarily at risk for spelling deficits, while children with cooccurring phonological and language disorders are additionally at risk of developing reading comprehension and decoding deficits. Accurate determination of specific speech and language deficits in very young children is challenging due their limited speech production, attention, cooperation, and reticence to communicate with unfamiliar examiners. Therefore, valid assessments need to include the interactions of young children in everyday activities with familiar caregivers. Conversational speech analysis, as used by Van der Speck et al., allows for the examination of specific speech and language parameters including vocabulary acquisition, word length, and phonological error analysis. One limitation of conversational speech analysis with very young children is that a 20-minute recorded interaction may not be a representative sample of the child’s communication ability. Analyses of longer conversational speech samples have been limited by the portability, durability, and memory capacity of the recorders and the hours required of the examiner to analyse the samples. Recent advances in voice recognition and signal processing technology may address these issues. Small recording devices that can be worn by a child have recently become commercially available and are capable of recording up to 16 hours of child and adult vocalizations. These extended-time speech samples can be uploaded from the recording devices to a computer for analysis. To date, automated analysis from these small recording devices is limited to analysing turn-taking and the number of vocalizations. As technology expands to enable further analysis of extended-time speech samples of infants and young children, more accurate examinations of specific risk factors for learning, reading, and speech disorders will be feasible and will facilitate treatment strategies targeted to improve chances for the educational success of VLBW children.