Language and social/emotional problems identified at a universal developmental assessment at 30 months

Behavioural problems identified in the preschool years often persist [1, 2] and their association with adverse physical, mental health and forensic outcomes in adulthood is now generally accepted [2‐6]_.

In 2010 the Scottish Government’s Health Department mandated a new universal child health contact between 24-30 months to identify children who might benefit from further support. There had been no universal child health surveillance contact beyond 16 weeks prior to this policy recommendation, though families may have received routine contacts from their health visitor depending on their assessed need and customary practices within their region of the health service. The universal surveillance contact was removed in part because of a lack of evidence demonstrating effectiveness [7].

Language development is closely related to broader social development, and there is a high incidence of language/communication difficulties in children with emotional and behavioural problems [8‐12]. The strength of the association between language delays and behaviour problems is however still subject to debate [13].

The majority of late talkers identified in the preschool period tend to move into the normal range by school age [14‐18] but continue to have significantly weaker language skills than peers with similar backgrounds. As with behavioural problems, language difficulties identified at this age can persist in some form into later childhood [19‐21]_, adolescence [22, 23] and adult life [24‐26]. Although some studies have sought to conduct longitudinal follow-up, much of the research in this area uses retrospective data with older participants from clinical samples and is therefore less representative of the whole population and less informative of the relationship between early language and behavioural development.

There is an increasingly strong case for identification of language and behavioural [27] problems in early childhood given the range of effective interventions for early neurodevelopmental and communication problems [28, 29]. Language delay in preschool children is common [30], and many children are not identified until they begin formal education [31]. In the United Kingdom, parents are generally given advice about normal language development and are then expected to identify children with problems and notify their general practitioner or health visitor [32]. The effectiveness of this approach in identifying children with speech and language difficulties is unclear and parental concerns about toddlers predominantly emphasise eating, sleeping and toileting problems [33]. Screening for language delay using standardised measures could be a more reliable method of identifying language and communication problems in the pre-school years [34, 35]_.

A pilot evaluation of a universal 30 month contact [36] showed that community child health nurses identified a substantial number of children with language delay and behavioural problems who were previously considered at low risk of developmental difficulties and who would normally not have received any preventative health service input until school entry.

This aims of this study are to a: assess the uptake of a new routine 30 month health visitor contact and b: to quantify the prevalence of language delay and social/emotional/behavioural difficulties and their potential overlap.

From a potential 543 children in the eligible population for this contact, data were obtained from 486 (90%) children who either received the 30 month contact or had attempted contacts. Full data were obtained on 421 children, 78% of those eligible.

Of the sample of 486 children, 64.8% were assigned to the Core HPI (lowest risk status) before the contact, with 17.5% Additional HPI (intermediate risk) and 6% Intensive HPI (children requiring multi-agency involvement). 10.7% of children had missing HPI information and 1% had not yet had HPI allocated.

Table 1 demonstrates that a large proportion (48.3%) of participants in this sample were, at the time of data collection, living within the most deprived areas of Greater Glasgow.

We sought to assess the uptake of a new routine 30 month health visitor contact and to quantify the prevalence of language delay and social/emotional difficulties.

There was a 90% return rate of contact forms from the health visitors, and assessments were completed on 78% of eligible children. Health visitors were provided with additional local administrative support to allow for time to be allocated to these 30 month contacts. Almost half of the families eligible for assessment had homes in the most deprived quintile of Scottish areas of residence. Contact completion rates did not vary significantly by socio-economic status, which contrasts with earlier Scottish data demonstrating a decline in coverage with increasing age of pre-school children and that this decline was steeper in the most deprived quintile [50].

A large proportion of 30 months old children give indications of having at least some social/emotional or language difficulty. Of the sample of 425 children for whom we have SDQ data; 8.8% were above the threshold for the Total Difficulties Score of the SDQ, while a higher proportion (13.6%) had scores suggesting likely emotional and conduct problems (following further examination of the individual subscales). Similar data have been reported from the Growing up in Scotland cohort at age 34 months [51]. No socio-economic patterning of language or social/emotional/behavioural problems was evident in our sample. This may reflect our modest sample size, but it is equally possible that social gradients demonstrated elsewhere [52] may emerge later in the preschool years.

Based on the simple language screening questions; 3-8% of children in the sample had some indication of language delay.

There was substantial overlap between language delay and social/emotional difficulties as judged by abnormal scores on the SDQ. The association between abnormal SDQ scores and failure to produce 50 words was significant (P < 0.05) for each subscale of the SDQ although the use of multiple comparisons should be noted in interpreting the strength of these associations. Only around one third (12/33) of children with language delay had no abnormal SDQ subscale scores, and children with abnormal SDQ subscale scores are about three times as likely to have language delay than those with no abnormal subscale scores. Our analysis appears to show that it is more common for children who have language delay to also have co-occurring socio-emotional difficulties (60.6% - 63.6%) than for children who have socio-emotional difficulties to also have language delay (13.4% - 4.9%). These findings support other reports that language delay without associated neurodevelopmental problems is relatively uncommon in early childhood [53].

The 30 month universal nurse contact reported here appears to identify significant developmental needs which may otherwise have gone unmet. We have demonstrated that the assignment of children to the 3 HPI groups in the first 4 months of life based on the anticipated degree of support required did not usefully predict which children would have possible problems at age 30 months. The (English) National Service framework has proposed that decisions about category assignment should be deferred, often to the end of the first year. This view has been was supported by research in health surveillance [54, 55] and endorsed in the 2006 revision of Health for All Children.

The pathways from assessment to intervention are currently being optimised and judgement of the overall cost and utility of the 30-month surveillance visit will need to take into account the wider impact on primary care and specialist services. Because screening for communication problems and behavioural difficulties could result in a substantial increase in secondary care workload, it is important to refine any universal screening measures as much as possible in order to identify only those children whose problems are likely to be persistent, to have a serious impact on education and life chances, and be remediable. Further research work is ongoing to ascertain the screening performance of the approach described here, based on parental reporting of difficulties through detailed psychometric assessment of children with and without identified problems. The development of effective interventions to improve outcomes in children with confirmed developmental difficulties at this age remains a high priority.