Six studies provided evidence relevant to the decision problem in the scope, including 2 published randomised controlled trials, 1 unpublished randomised controlled trial and 3 non-comparative observational studies. The studies included between 12 and 197 people. Only 1 study used the current model of the device, the others used older models. One study was done in the UK.
The 3 randomised controlled trials showed a statistically significant improvement in patient-reported anxiety scores with Alpha‑Stim AID compared with drugs alone, a sham device or no treatment in people with anxiety disorders. The benefit of Alpha‑Stim AID in relieving anxiety symptoms was also reported consistently in the observational studies.
The studies were of short duration (usually 5 to 6 weeks) with only 1 observational study reporting longer‑term outcomes at 24 weeks.
Only 2 of the studies reported improvements in quality of life (Morriss et al. 2019, Lu and Hu 2014) using health questionnaire (EQ‑5D‑5L and WHOQOL BREF) scores. An improvement in quality of life was also reported with Alpha‑Stim AID in a patient survey run by NICE's Public Involvement Programme. For details, see the assessment report overview in the supporting documents for this guidance.
Adverse events reported with Alpha‑Stim AID in 2 studies included mild headache, dizziness, nausea and feeling strange. Similar symptoms were reported by people who used Alpha‑Stim AID in the patient survey that was done as part of this assessment. Additional symptoms reported in this survey included ear discomfort and worsening of anxiety symptoms. All reported adverse events were mild. The instructions for use note that prolonged electrotherapy stimulation treatment at currents higher than necessary may cause dizziness or nausea that can last for hours to days. The clinical experts did not identify any specific safety concerns with Alpha‑Stim AID.
In response to consultation comments about the mechanism of action of Alpha‑Stim AID, the external assessment centre (EAC) reviewed additional studies that were highlighted by the consultees. It concluded that 2 published studies (Kennerly 2004; Lande and Gragnani 2018), and 1 PhD thesis (Kennerly 2006) which reported electroencephalogram (EEG) parameters of brain activities after using the Alpha‑Stim AID device, were potentially relevant. The EAC noted, however, that none of these studies examined the association between changes in brain wave activity and changes in anxiety symptoms. The EAC also reviewed a conference abstract and an unpublished single case report, which reported an increase in brain alpha waves and a reduction in perceived anxiety symptoms after using the Alpha‑Stim AID device. But the EAC emphasised that these results should be treated with caution because of the small sample size and uncertainties around study details and the generalisability of the results.
Three clinical experts, including 1 expert with extensive expertise in interpreting EEG recordings, also reviewed the additional evidence and agreed with the EAC's conclusions. The expert with EEG expertise considered that further investigation, preferably in studies that include a sham control, is justified to improve the understanding of Alpha‑Stim AID's mechanism of action.
The company identified 1 UK study (Morriss et al. 2019). This reported the cost impact of Alpha‑Stim AID as a treatment option for people with anxiety disorders who were waiting for individual cognitive behavioural therapy (CBT) delivered by Improving Access to Psychological Therapies (IAPT) services. No additional economic analyses were identified by the EAC.
The company developed a decision tree model with a time horizon of 6 months. The model compared the cost of using Alpha‑Stim AID as an option for people waiting for individual CBT in IAPT services with that of individual CBT alone. The results showed that Alpha‑Stim AID was cost saving by £817.68 per person.
The EAC agreed with many of the assumptions in the company's IAPT model but found some limitations. In the UK observational study, a significant proportion of people who were offered Alpha‑Stim AID chose not to use it and preferred to wait for individual CBT (Morriss et al. 2019). The EAC therefore revised the model to reflect the reduced uptake of Alpha‑Stim AID. The EAC also modified the structure of the model to better reflect the current care pathway as outlined by clinical experts. The EAC's base case included drug treatment as an option at the start of the pathway and for people whose anxiety symptoms did not respond to Alpha‑Stim AID or individual CBT. The EAC excluded the second course of individual CBT for people whose anxiety symptoms did not respond to an initial course of treatment.
The EAC's base case showed that using Alpha‑Stim AID saved £80.79 per person compared with individual CBT. This was based on a 47.2% response rate with Alpha‑Stim AID (Morriss et al. 2019) and a 54.2% response rate with individual CBT (Gyani et al. 2013). The reported response rate for Alpha‑Stim AID included everyone who used it in the Morriss et al. study. However, many people may also have had individual CBT alongside or after Alpha‑Stim AID. The reported response rates for treatment combinations were more uncertain and varied. For instance, the response rate was 65.0% in people using Alpha‑Stim AID alone and 13.0% in people having Alpha‑Stim AID followed by individual CBT (Morriss et al. 2019). The EAC explored the effect of different response rates and treatment regimens. For details, see the addendum to the EAC's assessment report in the supporting documents for this guidance.
In response to committee, clinical expert and consultation comments, the EAC revised the IAPT cost model to more closely reflect the current IAPT pathway. Results from this model suggested that introducing Alpha‑Stim AID to the current pathway could be cost saving and result in fewer people being discharged from IAPT to primary care for further treatment. However, because of the limited data available and uncertainties in the modelling, the EAC concluded that the cost impact of using Alpha‑Stim AID in the current IAPT care pathway remained uncertain.
During consultation, the company submitted a new model, which included Alpha‑Stim AID in primary care as a first-line treatment for anxiety. The model was based on an unpublished study in the NHS and the results showed a cost saving of £285.00 per person when Alpha‑Stim AID was included in the care pathway. However, the EAC considered that the proposed model did not reflect clinical practice because it did not capture medication use, possible variations in treatment options and repeat visits to a GP or nurse.