SINGING FROM THE SAME HYMN SHEET – The im importance of adopting a population approach to in intervention for child ild wit ith la language le learning dif iffic icult ltie ies James Law Professor of Speech and Language Science University of Limerick – Practitioner Conference May 2017
Is child language a public health problem?
What is a public health problem? For a health problem to be considered a public health issue, three criteria must be met:- • It must place a large burden on society, a burden that appears to be getting larger; • The burden must be distributed unfairly (i.e., certain segments of the population are unequally affected); • There must be evidence that upstream preventive strategies could substantially reduce the burden of the condition
The burden - Is it getting larger? • Prevalence estimates suggest DLD is as prevalent as childhood obesity, reported to be 7% in Australia.Approximately 5% to 8% of children may have DLD. • In population-based samples, which use broader criteria and include less severe DLD, estimates are higher. Population-based studies report 14% to 20% of 4-5 year old children may be affected by DLD, with similar levels also reported at age 7 years. • In Australia, there has been a major increase in the number of speech pathology service claims made to Medicare, Australia’s publicly funded universal health care system. The speech pathology Medicare service items reported went from 3,051 in 2004-05, to 115,167 in 2012-13, with majority of services for 0-14 year olds.
and in the US.. • Two nationally representative surveys that include measures of speech and language disorders in children at multiple points in time are the National Survey of Children’s Health and the National Survey of Children with Special Health Care Needs. • The National Survey of Children’s Health showed an increase in prevalence of speech and language disorders from 3.8 percent (n = 2,697) in 2007 to 4.8 percent (n = 3,916) in 2011, a 26 percent increase. • The National Survey of Children with Special Health Care Needs showed an increase in prevalence from 3.2 percent (n = 8,435) in 2005-2006 to 5.0 percent (n = 11,936) in 2009- 2010, an increase of 56 percent • “ The best available evidence shows an increase in the prevalence of speech and language disorders over the past decade in the U.S. child population. Trends in annual Supplemental Security Income initial allowances parallel this overall increase. “
Is the burden distributed unfairly? • Social gradient
Social gradie ient in in oral l la language skil kills amongst 5-6 6 year old ld child ildren on the Australian Early Development Census (AEDC) in in 2015 and the UK’s Mille illenium Cohort Study in in 2005 16 14 % Developmenetally vulnerable language skills 12 10 8 6 4 2 0 Most Quintile 2 Quintile 3 Quintile 4 Least disadvantaged disadvantaged
Access • Access to services is not equitably distributed – i.e., it is not easy for all children to access relevant services. • It is often those families who are most in need of services, who access them the least. • Barriers to accessing health services reported by vulnerable families include cost, as well as availability and accessibility of health services. • Not only are there financial barriers to access, but more socially advantaged parents are more likely to have the skills and knowledge based on their education and experience to be resourceful and access the services they need.
What can you do about it?
The three pillars of evidence based practice..+1
Best research evidence needs to be:- Accessible – is it easy to find? Readily interpretable – what does it mean? Meaningful in a practice context – can I use it? Divisible into its components - what are the key elements that make an intervention work? Realistic – what sort of effects should we be expecting? Translatable - across cultures and languages Meaningful to policy makers and those mediating whether an intervention is introduced and supported 11
And the “What works” (WW) for children with speech and language needs and the Communication Trust WW interactive website:- www.thecommunicationtrust.org.uk/schools/what-works
“Foundations for Life” Review of the evidence for early developmental Section Header slide interventions – in many cases not even including Subtitle language development 13
14
Source http://onlinelibrary.wiley.com/doi/10.1002/14651858.CD004110/pdf
Intervention evidence • Lots of intervention evidence available from What Works for SLCN, SpeechBite etc plus 15 or so systematic reviews • Need to separate out into:- • Universal • Targeted selective • Targeted indicated • Specialist • Not always an easy task (partly because people are not clear about their sampling)
How robust is the evidence?
Findings from the EEF report.. • 44 intervention studies (quasi-experimental and randomised) • Specific outcomes • Phonological awareness ( n= 8 ): an understanding of the sound structure of the spoken language. • Vocabulary: expressive and receptive (n=20) – the ability to use or understand words. • Expressive language (n=7) : children’s ability to use language in an accurate and coherent manner. • Receptive language (n=9): children’s ability to understand complex language forms including grammar, inferential use of language etc. • Sample sizes varied considerably from 12 (Tsybina and Eriks-Brophy, 2010) to 2250 (Apthorp et al.2012) and the studies varied considerably in terms of whether they were clear Universal (7), or targeted – selective (20) or targeted – indicated/specialist (17) • Equally balanced between programmes and practices
What did we recommend? 1. There is a need to explore the potential role of parent child interaction interventions with young children as a means of promoting children’s language abilities and ensuring that children are ready for learning when they get to nursery at 2-3 years. • Care needs to be taken to identify parent/child dyads where there is some concern about the interaction AND there is an identified language difficulty. • The outcomes for such a study would be improved interaction, vocabulary and potential early word combinations. • The comparison intervention here would most likely be with routine care – from health visitors and other community services. 2. There is a need for an efficacy trial of training teachers (professional development) to deliver interventions within the classroom drawing on the work of Piasta, Dockrell and The Hanen Centre’s Learning Language and Loving It. • The outcome for such a study should be vocabulary (receptive and expressive), narrative skills and pre- reading skills. • The comparison here should be with routine care in comparable early years settings AND with targeted (indicated) interventions provided by specialist staff such as speech and language therapists.
Some thoughts on the randomisation question • Randomisation reduces bias by washing out the contextual factors – but are those factors of interest • Randomisation per se tells you nothing about how an intervention works – just whether • Internal and external validity – especially the latter means the results are generalizable – you can take what you are doping in one place and apply it to another • Serious concerns that this may not be feasible in highly contexualised, complex interventions such as intervention for children with language learning difficulties. Cf the example of classroom size (Cartwright and Hardie 2012) • Comes back to both your population and to the way that the intervention is administered. First and second generation interventions commonly vary widely in their impact Remember the best external evidence is only one of the pillars of EBP • We also need well informed patients and the practitioners perspective plus, of course, a better understanding of the active ingredients of interventions
But what about active ingredients – the application of behaviour change techniques • The Medical Research Council (Craig et al. 2008) - only by understanding how an intervention works can we design, apply and replicate effective interventions; • Based on behaviour change theory drawing from smoking cessation and increasing healthy eating and physical exercise (Michie et al. 2011); • BCT replicable and aims to alter the causal processes
Active ingredients - Logic models – a public health example
Logic models and theories of change.. Deliver Monitor impact intervention – Parent and child Monitor primary Identify children – wellbeing, meeting respond to outcome – with SLCN QOL, intervention speech/language predetermined independence criteria
What are the ingredients of speech and language therapy interventions? • The Behaviour Change Technique Taxonomy (BCTT) comprises 93 Behaviour Change Techniques (BCTs) agreed by a panel of international experts (Michie et al, 2013); • SLT interventions are complex due to many interacting elements within them, difficulties with replication and to identify the causal mechanisms, or active ingredients (Michie et al., 2011a; • Developed by team at Newcastle precisely for children with speech and language disorders focussing on phonological awareness (Stringer and Toft 2016 plus student input from Atkinson and Spalding).
Recommend
More recommend
