Statement of principal findings
Following comprehensive searches of the current literature, two studies comparing ORBIT with psychoeducation (one each in the UK34 and Sweden38) and one UK study comparing active stimulation with sham stimulation36 were included in the review of clinical effectiveness. All were assessed as being at low risk of bias. However, there were important differences between study size and follow-up duration across the comparisons. ORBIT studies had larger sample sizes and longer follow-up than for Neupulse. Additionally, all ORBIT data used in economic modelling were available in the public domain and published in peer-reviewed articles, whereas the economic modelling for Neupulse relied on company-provided data through personal communication, with a shorter follow-up duration for transition probability model inputs.
The majority of intermediate, clinical and patient-reported outcomes specified in the scope were reported across the three studies, but the actual measures utilised and time points they were administered were not consistent across studies. In general, tic severity scores measured using the YGTSS-TTSS (primary outcome) were lower in the respective intervention groups at follow-up periods ranging from 4 weeks (Neupulse)36 to 12 months (the two ORBIT studies).34,38 However, no improvements were observed in all three studies concerning the YGTSS-Impairment score. Responses to secondary outcome measures were mixed, and participants did not consistently show a better response to the intervention compared with the control at all assessed time points and across studies. In general, the level of engagement, adherence to treatment and dropouts were similar between intervention groups. The Neupulse study lacked reporting of AEs.
Economic modelling shows that there remains substantial residual uncertainty in the assessment of cost-effectiveness, particularly for the Neupulse intervention. Key areas of remaining uncertainty for Neupulse include:
- non-publicly available Neupulse pricing information
- short follow-up of only 4 weeks for Neupulse
- a lack of evidence on longer-term effectiveness and safety for economic modelling.
For the economic modelling of ORBIT, there were some uncertainties regarding long-term intervention costs for ORBIT that would be required to deliver sustained trial observed benefits over the longer term.
For model population more generally, there was uncertainty about the most appropriate link between tic severity score and health state costs and utilities, in particular whether small improvements in YGTSS scores have a direct impact on generic QoL measures such as EQ-5D or CHU-9D.
Despite longer-term follow-up available for ORBIT than Neupulse, economic modelling results over a lifetime horizon were uncertain because of a lack of information about the maintenance of long-term effectiveness beyond the trial follow-up periods to inform economic modelling
Strengths and limitations of the assessment and uncertainties
This assessment was conducted using thorough and robust methods. The main limitation of the systematic review of clinical effectiveness was the paucity of evidence for the technologies of interest. The limited amount of evidence and inconsistencies in the outcomes measured and the timing of these measurements also hampered efforts to conduct further meaningful analyses.
The identified studies did not involve a comparison with face-to-face behavioural therapy, which is the current standard of care. Therefore, for the ORBIT studies, it was not possible to separate the effects of online delivery from those of ERP. Moreover, both ORBIT studies did not include a non-active control treatment (e.g. waitlist) to evaluate the natural course of the disease over time, especially in young children. For the ORBIT studies in particular, the comparator, psychoeducation, may be a more active, beneficial comparator than is routinely available in UK clinical practice. Therefore, results of cost-effectiveness modelling might under-estimate cost-effectiveness compared to usual NHS care.
While a reduction of tic severity measured by the YGTSS-TTSS score was greater among participants receiving the intervention (digitally enabled ERP in the ORBIT studies and home-administered MNS in the Neupulse study), the fact that there were no improvements in the YGTSS-Impairment score casts some doubt on whether a reduction in tic severity translates into an improvement in daily life, including self-esteem, social interactions and school/work performance. Similarly, the inconsistent improvements in quality-of-life measures (i.e. C&A-GTS-QOL) over follow-up times and across studies reinforce this doubt.
It is also unclear why only the YGTSS-TTSS score was selected as the primary outcome in the included studies, but not the YGTSS-Impairment score.
Both ORBIT studies included participants with moderate to severe tic disorders, who had a low rate of common comorbidities and modest use of medications for tics. Therefore, the reported findings cannot be generalisable to more severe tic disorder populations. Similarly, most participants in at least one of the ORBIT studies (the UK study) were White, which limits the generalisation of findings to other ethnic groups.
The longest follow-up period in the UK ORBIT study and the Swedish ORBIT study was 18 and 12 months, respectively. The current data for Neupulse, which has been designed to control tic symptoms on demand, refer to stimulation delivered for a maximum of 4 weeks. There are no data on longer durations of stimulation or longer-term outcomes to inform economic modelling.
The assessment of cost-effectiveness relied on several major, but highly uncertain assumptions about longer-term intervention costs that might be required to maintain the intervention’s effectiveness as well as the most appropriate long-term extrapolation assumptions. Scenario and probabilistic sensitivity analyses show that there is substantial residual uncertainty, making it difficult to define the most plausible ICERs for either comparison.
Key areas for future research
- As the included studies are the first published adequately powered randomised studies to assess the technologies under investigation, future replication is needed to confirm the observed results.
- Future studies of longer duration should compare digitally enabled therapy for tics versus face-to-face behavioural therapy and should also consider including a non-active intervention (e.g. waitlist) to monitor the natural course of the disease over time.
- In addition to tic severity, future studies should measure the impact of digitally enabled therapy on participants’ daily lives as their primary outcome.
- To assess differences in treatment response among participants, future studies should plan appropriate subgroup analyses according to the participants’ sex distribution (males vs. females) and common comorbidities.
- Future studies should include economic evaluations and should consider collection of longitudinal data to improve long-term modelling of treatment effectiveness with a focus on determining the impact of changes in clinical outcomes on QoL and costs.
Patient and public involvement
We did not involve patient and public involvement partners in the preparation of our report. However, we shared our findings with NICE, whose appraisal committees include representatives from patient organisations and networks, ensuring that the patient perspective is considered in the decision-making process.
Equality, diversity and inclusion considerations
This project involves an evidence synthesis and economic evaluation. The scope was defined by the NICE, which is committed to promoting equality of opportunity, eliminating unlawful discrimination and fostering good relations between people with particular protected characteristics and others.
- Tic disorders are more common in boys than girls at a ratio of 4 : 1.
- Patient-facing digital health technologies may be unsuitable for people with cognitive impairment, problems with manual dexterity, learning disabilities, and visual and/or hearing impairments. Carer or advocate assistance may be required to utilise the digital platform, and consideration of this should be made by the referring practitioner when considering appropriate intervention for the child or young person.
- Children and young people, or their families/carers, with English as a second language may have difficulties navigating digital technologies provided in English.
- People’s ethnic, religious and cultural background may affect their views of digital health interventions. HCPs should discuss the language and cultural content of digital health interventions with users before provision.
- Specific groups may particularly benefit from improved access to online behavioural therapy, for example:
- Those living in areas not currently served by specialist clinical centres might have difficulty travelling to face-to-face appointments if public transport is unreliable, costly and if parents are unable to drive them.
- Adolescents may have an increased engagement with this format of intervention.
- Children and young people with more chaotic home lives may lack the family support required to ensure that they attend face-to-face sessions. These families may also be less likely to seek help in the first place and or be less able to navigate the healthcare system.
- However, accessibility would not be improved for those who are unable to engage with a digital service due to a lack of equipment, unavailability of internet connection or lack of experience with computers or lack the privacy needed to complete this intervention. Additional support and resources may be needed for these individuals.
Chronic tic disorders and TS can significantly affect people’s daily lives. Under the Equality Act 2010, a person has a disability if they have a physical or mental impairment that has a substantial and long-term effect on their ability to do typical day-to-day activities. Age, sex, disability, race and religion are protected characteristics under the Equality Act (2010).