Treating children with speech and language impairments
BMJ 2000; 321 doi: https://doi.org/10.1136/bmj.321.7266.908 (Published 14 October 2000) Cite this as: BMJ 2000;321:908
Six hours of therapy is not enough
- James Law, reader,
- Gina Conti-Ramsden, professor of child language and learning
- Language and Communication Sciences, City University, London EC1V 0HB
- School of Education, University of Manchester, Manchester M13 9PL
Papers p 923
About 5-8% of children under the age of 5 have developmental impairments of speech and language. This proportion is higher than that for any other neurodevelopmental condition occurring at that age.1 Parents are concerned about these impairments, and the number of children being referred to speech and language therapy services is increasing.2
These impairments are characterised by a low level of speech and language skills. Such difficulties may occur secondary to disabilities such as cerebral palsy, sensorineural hearing loss, or autism. Impairment may also be the main symptom in a constellation of comorbid difficulties, such as challenging behaviour or otitis media.3
Although spontaneous remission of symptoms in primary speech and language disorders sometimes occurs many children will experience long term effects from these disorders. Studies of samples of children from different communities show that children who are at the extreme ends of the distribution of speech and language impairment are at risk of developing problems that can persist into adulthood.4-6 The inability to communicate with peers can have a marked effect on wellbeing.
Given what we know about the stability of speech and language impairments across time, what role can intervention play? There is evidence to suggest that some interventions can modify intelligence,7 and the literature about the Head Start programmes in the United States has shown that preschool programmes have a long term impact in terms of social outcomes (for example, in reducing the incidence of teenage pregnancy or incarceration).8 Clinical experience suggests that speech (whether difficulties involve dyspraxic— that is, neuromotor—or phonological presentations) and vocabulary can be modified but that it is much more difficult to change elements of syntax and verbal comprehension.
At first glance the picture painted by Glogowska et al in this issue of the BMJ (p 923) is gloomy.9 Interventions for speech and language impairments do not seem to work. However, there are some features of this study that should be interpreted cautiously. On average the children spent just six hours with their speech and language therapist in 12 months. How long would it take most people to change their speech and language behaviours? More than six hours, we would argue, even if clients were highly motivated. It is particularly important to note that both groups of children in the study (those who were given therapy and those who were not) continued to have marked language difficulties.
This study also needs to be set against a recent systematic review of studies of speech and language impairments that identified effect sizes for randomised and quasi-experimental study designs on the order of one standard deviation.10 This corresponds to a shift from the 25th to the 5th centile: a good improvement by any standard. These studies all included children of comparable ages and levels of language impairment. The source of the difference provides a potential explanation for the findings of Glogowska and colleagues. All of the studies in the review offered more treatment. In many cases the studies were carried out in university clinics and could best be described as efficacy rather than effectiveness studies. On the other hand, Glogowska et al's project is a study of the routine clinical services that are currently available to children in the United Kingdom.
Taken together the data indicate that offering limited amounts of speech and language therapy is not a tenable solution to the problem. The six hours provided did not necessarily reflect the choice of the speech and language therapists in the study but rather a constraint imposed on them by the “package of care” model of service delivery. The data suggest that such a simplistic model is not helpful and that the practitioners and their managers should be able to offer a more flexible package of interventions. This is likely to require a reorganisation of speech and language therapy services, but this is the point of practising evidence based medicine: when you fill the evidence gap you need to act.