Closed for comments This consultation ended on at Request commenting lead permission
The diagnostics advisory committee considered evidence on digital self-help for eating disorders from several sources. This included evidence submitted by Credo Therapies, Five Areas Ltd and stem4, a review of clinical and cost evidence by the external assessment group (EAG), and responses from stakeholders. Full details are available in the project documents for this guidance.
It is estimated that at least 1.25 million people in the UK have an eating disorder (Beat's data on eating disorder prevalence in the UK). Eating disorders are described as mental health conditions in which controlling food is used to cope with feelings and situations.
Having a binge eating disorder means eating very large quantities of food without feeling in control of it. This includes eating much faster than normal until feeling uncomfortably full, eating large amounts of food when not physically hungry or eating alone through embarrassment at the amount being eaten. It can also include feelings of disgust, shame or guilt during or after the binge. People with bulimia nervosa cycle between bingeing and trying to compensate for the overeating (purging) by vomiting, taking laxatives or diuretics, fasting, or exercising excessively. When symptoms are similar to an eating disorder but they do not exactly fit the typical symptoms for the condition, the condition may be diagnosed as other specified feeding or eating disorder (OSFED). Disordered eating refers to food- and diet-related behaviours that do not meet diagnostic criteria for recognised eating disorders but may still negatively affect physical and mental health.
Signs of eating disorders can be noticed in many settings, such as school, university, work, home or social care. Often the first healthcare contact who will do an initial assessment is a GP. After the initial assessment, people with a suspected eating disorder are usually referred to a community-based eating disorder service for further assessment or treatment.
While people wait for further assessment or treatment in specialist care, usual waiting list care may include:
further appointments at the GP practice
appointments with the eating disorder service while on the waiting list
signposting to voluntary, community and social enterprise organisations, for example, eating disorder charities
books or online resources (including the books used in eating-disorder-focused cognitive behavioural therapy [CBT-ED]-based guided self-help)
local groups or telephone helplines for additional support.
The incidence of eating problems and eating disorders are increasing. People often wait a long time for psychological treatment to start. More referrals to specialist care mean that the services cannot meet the increasing need for psychological treatment with the healthcare professional capacity available.
Earlier treatment could help prevent the condition from becoming more severe. There is a need for a treatment option that could start as soon as possible once eating problems are identified. This could be in primary care or straight after an eating disorder is diagnosed in specialist eating disorder service.
Using digital self-help does not depend on healthcare professional capacity to provide support for using the therapy. It could also offer people with signs and symptoms of eating disorders faster access to eating disorder therapy.
The EAG identified 3 randomised controlled trials (RCTs), 3 cohort studies and 3 qualitative studies on Overcoming Bulimia Online. The RCTs showed reductions in binge eating episode frequency and eating disorder symptom severity compared with usual waiting list care. The cohort studies also reported improvements in clinical outcomes during the study.
In only 1 of the RCTs on Overcoming Bulimia Online, people participating in the study used the technology as an unguided intervention. The committee noted that digital self-help may be more effective when it is guided than unguided. So the improvements in outcomes seen in the studies may have been smaller if no support was provided. The committee recalled that people often wait a long time to access guided self-help because it needs healthcare professional capacity. Unguided digital self-help could provide earlier access to psychological treatment. To confirm the effectiveness of Overcoming Bulimia Online, more data on the unguided use of the technology is needed. Despite the limited evidence on unguided therapy, the committee concluded that Overcoming Bulimia Online is likely to be clinically effective.
There were 3 cohort studies on Digital CBTe. In 2 of the 3 studies, people used Digital CBTe as an unguided intervention. All 3 studies showed reductions in binge eating episode frequency and eating disorder symptom severity during the study. There was 1 small cohort study on Worth Warrior. People participating the study used the technology as an unguided intervention. Clinical outcomes for some people in the study showed improvement.
The studies on Digital CBTe and Worth Warrior did not have comparator groups. The committee noted that this meant that it was possible that people's eating disorder symptoms improved for reasons other than the technologies. The study on Worth Warrior was also very small. So it was difficult to know if the improvements in the study happened by chance and if they could happen in a larger group of people. The committee concluded that it is uncertain whether Digital CBTe and Worth Warrior are likely to be clinically effective. To better understand this, comparative data on the technologies is needed.
None of the studies included a long-term follow up. Clinical experts noted long-term evidence on book-based self-help for context (it was not included in the assessment). It suggests that people who start with self-help are likely to have better outcomes in the long term compared with people who start with therapist-led sessions. This is because they are more actively involved in their own therapy. To better understand the effects of unguided digital self-help, longer-term data after its initial use, compared with no initial use, is needed.
In many studies, the proportion of people who did not complete the digital self-help treatment was high. Not much information was available on the people who did not complete the treatment or the reasons why. The Digital CBTe company representatives explained that their studies were done either in community or NHS settings. The attrition (non-completion) rates in these studies were higher than in tightly controlled clinical trials. But, the rates are typical and common in real-world evaluations of digital mental health interventions, particularly when self-guided or minimally supported. The committee concluded that to better understand the potential benefits of the technologies, more information on people who did not complete the digital self-help treatment, and the reasons why, is needed.
The committee noted that some people may particularly benefit from having access to unguided digital self-help. For example, people with less severe eating disorders who may otherwise wait longer for treatment and people who live in areas with lower specialist eating disorder service capacity.
Most study participants in the key studies were white women. Not all studies reported information on the participants' ethnicities or whether they had conditions that may make it more difficult to use or complete digital self-help. This could include whether participants were neurodivergent, had learning disabilities, visual, hearing or cognitive impairment or problems with manual dexterity, or were less used to using digital technologies in general. The patient and carer experts highlighted the importance of inclusive technologies. If digital self-help programmes are designed only with neurotypical women from white ethnic groups in mind, others may find it harder to engage with the therapy. The committee agreed that future studies should collect information about the characteristics of study participants. This should include participants' ethnicity and whether people have conditions that may make it more difficult to use the technology. Studies should aim to include a diverse group of people and an equality impact assessment.
The EAG adapted the model from NICE's guideline on eating disorders to estimate short-term resource use and costs for digital self-help technologies in primary care and specialist eating disorders services. The base-case model assumptions were conservative. This was because of the uncertainties in the evidence base. The model assumed that only people who completed the digital self-help treatment had an increased probability of no longer having eating disorder episodes (remission). Partially completing the treatment had no benefits. The model did not include potential improvements in health-related quality of life, avoided deaths or potential reductions in longer-term resource use and costs associated with comorbidity (such as obesity in binge eating disorder or other mental health conditions such as depression and anxiety). The EAG advised that in a future assessment, a longer-term model is needed to more fully capture the benefits and costs of the technologies.
The key clinical inputs to the short-term model were the probabilities of remission relapse (eating disorder symptoms returning after the condition being in remission). The probability of remission was taken from the Sánchez-Ortiz et al. (2011) study. This was because this was the only study that reported on remission using the current online format of Overcoming Bulimia Online. There was no evidence on how digital self-help affects relapse. So relapse probability in the base case was based on clinical expert estimates and assumed to be the same for people having usual waiting list care and people using Overcoming Bulimia Online. The EAG advised that mortality would be a key clinical input to a longer-term model, but it was not included in the short-term model. None of the studies reported on mortality. The committee concluded that more information on remission, relapse and mortality is needed.
Resource use in the model included healthcare use during a 1-year follow-up period. There was no evidence on the effect of using digital self-help on resource use and so it was based on clinical expert estimates from 2 group interviews. The experts noted that it was very difficult to give definitive estimates. The model assumed that people whose eating disorder was in remission after the initial digital self-help treatment or usual waiting list care, did not need further assessment or treatment. The clinical experts also noted that if digital self-help is helpful for people, it could reduce the need for further treatment as well as treatment length or intensity later in the care pathway. If there was evidence for this, it could be captured in a longer-term model. The committee concluded that to reduce uncertainty in the model, more short- and long-term data is needed on the effects of digital self-help on resource use and the NHS care pathway.
The conservative base-case analysis estimated that, compared with usual care in bulimia, using Overcoming Bulimia Online would save £5.52 in primary care and £39.86 in specialist eating disorder services. The EAG analysed several plausible alternative scenarios where Overcoming Bulimia Online was more effective or where higher resource use and costs were avoided. In some of these alternative scenarios the cost savings were considerably higher. The committee concluded that using Overcoming Bulimia Online was likely to be cost-effective use of NHS resources.
Because Digital CBTe and Worth Warrior did not have comparative clinical effectiveness evidence, the EAG did 2-way sensitivity analyses. These were to show how much more effective a hypothetical digital self-help technology at a given per-person cost would need to be than usual care to be potentially cost saving (using the EAG's conservative base-case assumptions). Based on these analyses, Digital CBTe and Worth Warrior would likely only be cost saving if:
some of the most conservative assumptions in the model were relaxed and
the effect sizes seen with Overcoming Bulimia Online were replicated for both technologies.
The committee concluded that it is uncertain whether using Digital CBTe and Worth Warrior is likely to be a cost-effective use of NHS resources.
User feedback from the studies on Overcoming Bulimia Online and Digital CBTe was mainly positive. The users who completed all or most of the treatment appreciated the technologies' usability, effect on their eating disorder, privacy and flexibility. The preliminary user feedback from the study on Worth Warrior included some positive views on the content and interactivity of the technology. In the NICE survey on views on using digital self-help, most people who had used digital self-help were likely to recommend it to others with eating disorders. Two studies on Digital CBTe also included the views of NHS staff. The staff felt that the technologies could be helpful for people with eating disorders and therefore reduce the intensity of further treatment or need for further support from the service. The committee concluded that it was likely that the technologies would be acceptable to people with eating disorders and healthcare professionals.
The committee discussed whether there was potential for harm if digital self-help is used in the NHS alongside usual waiting list care while further evidence is generated. Most studies did not report on adverse events but the committee noted that:
digital self-help may not be suitable for some people
if people do not complete the digital self-help course, they could feel demotivated and their eating disorder symptoms could get worse
people with severe eating disorders are at risk of crisis.
The clinical experts noted that self-help-type therapy in general is not suitable for people with severe eating disorders or a high mortality risk. This is because of the potential physical health impact of the eating disorder or other mental or physical health conditions. The technologies in the assessment are not intended for this population. The clinical experts explained that when unguided digital self-help is used alongside usual waiting list care, such as regular check-ins and routine physical monitoring, it is not expected to cause harm to people with eating disorders. The committee concluded that it was important that the technologies should be used after an initial eating disorder assessment in primary care or further assessment by specialist eating disorder services. It also concluded that the technologies should be used alongside usual waiting list care, such as regular check-ins and routine physical monitoring.
How are you taking part in this consultation?
You will not be able to change how you comment later.
You must be signed in to answer questions