6 Implementation considerations

The following considerations around implementing the evidence generation process have been identified through working with system partners.

6.1 General

  • Developers should provide training for staff to support use of the technologies

  • The evidence generation process is most likely to succeed with dedicated research staff to reduce the burden on NHS staff, and by using suitable real-world data to collect information when possible.

  • Evidence generation should be overseen by a steering group including researchers, commissioners, practitioners and representatives with lived experience.

  • Contributing services or centres should be chosen to maximise the generalisability of evidence generated, or to improve data collection for important subgroups.

  • Careful planning of approaches to information governance is vital.

6.2 Topic-specific

  • Head-to-head comparison between the technologies is limited because they may be designed for different groups of children and young people. For example, several developers have indicated that their technology is not designed for use by children and young people with low mood only. This may limit the availability of evidence for this subgroup

  • There is variation in the services and settings where the technologies may be used. Treatment processes and outcome performance vary. This may create issues with the generalisability of any evidence generated to the wider NHS, reducing its utility for NICE decision making. Developers should provide clear descriptions of the services and settings in which the study is done, and the characteristics of the included children and young people. Generalisability may be improved by using a range of sites, or sites considered more representative of national norms

  • Using school settings for data collection may support follow up, but these settings are narrower in scope than the NICE early value assessment

  • There is a wide spectrum of neurodivergent children and young people, and their support and needs vary considerably. Complete evidence across this spectrum will be difficult to generate

  • Information collected about engagement and drop out may not necessarily correlate with outcomes. For example, children and young people may choose to drop out early if they feel better. Developers should consider whether it is possible to collect information on any further interactions with health services after drop out

  • Using CHU-9D requires a licence, with an associated cost.

ISBN: 978-1-4731-5365-3