NICE commissioned an external assessment centre (EAC) to review the evidence submitted by the company. This section summarises that review. Full details of all the evidence are in the project documents on the NICE website.
3.1 There were 4 peer-reviewed studies on myCOPD, including 3 randomised controlled trials (RCTs) and 1 observational study. All trials were done in the NHS with 6 weeks or 3 months of follow up. The sample sizes ranged from 41 to 90 people with mild, moderate or severe chronic obstructive pulmonary disease (COPD). Of the 3 RCTs, TROOPER used a non-inferiority design comparing myCOPD with a face-to-face pulmonary rehabilitation programme in people who were referred for rehabilitation (Bourne et al. 2017). The other 2 trials examined the use of myCOPD as a self-management tool: RESCUE, a feasibility trial, compared myCOPD with usual care including a written self-management plan in people who were discharged from hospital after an acute exacerbation (North et al. 2020). EARLY was a superiority RCT that compared myCOPD with standard care in people with mild to moderate COPD or recently diagnosed COPD (Crooks et al. 2020).
3.2 There was also real-world evidence on clinical benefits, health service use, patient experience and usage of the myCOPD app from 10 local evaluations. A service evaluation from Southend University Hospital explored using myCOPD to support a home-based pulmonary rehabilitation programme. Other local evaluations assessed the effect of myCOPD on self-managing COPD. Most evidence was from interim evaluations designed to inform commissioning decisions or service developments. The methodology, patient numbers or characteristics, clinical outcomes and follow-up periods were not fully reported in these evaluations. However, the EAC concluded that the real-world evidence reflected the use of myCOPD in clinical practice and the findings of these evaluations would be generalisable to local health services.
3.3 Evidence on using myCOPD for self-managing COPD showed improvements in clinical outcomes. There was evidence from trials and real-world evidence evaluating myCOPD for self-managing COPD in people with different COPD severity and exacerbation history. Results suggested improvements in clinical outcomes such as the COPD assessment test (CAT) score, 6‑minute walking test (6MWT) distances and inhaler techniques after using myCOPD. Of those who were admitted to hospital because of an acute exacerbation, RESCUE showed a significant CAT score mean difference of -4.49 (95% confidence interval [CI] -8.41 to -0.58, n=41) in all people in the study regardless of whether they completed the study. RESCUE also showed a non-significant CAT score mean difference of -2.94 (95% CI -6.92 to 1.05, n=35) with myCOPD in people who completed the study at 3‑month follow up (North et al. 2020). RESCUE showed a significant reduction in inhaler errors in people using myCOPD compared with people having standard care (relative risk, 0.38; 95% CI 0.18 to 0.80). Compared with people having standard care, people using myCOPD were less likely to have exacerbations (relative risk, 0.58; 95% CI 0.32 to 1.04).
3.4 TROOPER reported no statistically significant difference in CAT score and 6MWT between the intervention groups, indicating myCOPD was not inferior to face-to-face care for pulmonary rehabilitation (Bourne et al. 2017). Health-related quality of life was not worse in people using myCOPD for pulmonary rehabilitation compared with those having face-to-face rehabilitation. The EAC considered that the TROOPER trial was well designed, but the sample size was small. The company provided further information on the sample size calculation to show non-inferiority at consultation. The additional information did not address the EAC's methodological concerns, such as the choice of the non-inferiority limit, and the EAC considered a larger trial would probably be needed to confirm the results. Real-world evidence also reported improvements in CAT scores and 6‑minute walk test with myCOPD to support home-based pulmonary rehabilitation (NHS Southend clinical commissioning group [CCG] evaluations).
3.5 For self-managing COPD, usage data of myCOPD reported in the RESCUE and EARLY trials showed its use declined over time. TROOPER used myCOPD for pulmonary rehabilitation. It showed that although adherence declined over time it was comparable in both groups (Bourne et al. 2017). After 6 weeks, 66% of people using myCOPD had completed 2 or more sessions compared with 69% of people in the comparator group who attended 2 or more face-to-face sessions. Real-world evidence showed varying use of the app. There is no direct evidence on the relationship between adherence and clinical outcomes using myCOPD.
3.6 NICE's public involvement programme did a survey of people using myCOPD. In this, people reported that myCOPD was easy to use (n=297/359, 82.7%) and helped improve their understanding of the condition and manage symptoms. Three-quarters of people who responded (n=267/358, 74.6%) felt confident in managing COPD symptoms after using the app. Of those who used the app to control COPD symptoms, 66.1% (n=220/333) felt there had been a reduction in the number of exacerbations experienced after using the app. People thought that myCOPD was a helpful tool and provided useful information that improved their confidence in managing COPD.
3.7 The company submitted cost models in 2 different populations:
people discharged from hospital after an acute exacerbation of COPD who used myCOPD for self-managing their condition (the AECOPD model)
people eligible for pulmonary rehabilitation who used myCOPD for pulmonary rehabilitation (the PR model).
The company also presented results for 2 different pricing scenarios. For healthcare providers who purchased a population-based unlimited myCOPD licence package, the company's AECOPD model showed that myCOPD was cost saving by £204,641 per CCG over 1 year compared with standard care (not accounting for savings from pulmonary rehabilitation). The estimated savings in the PR model within the population licence package was £20,269 per CCG per year. Also, if the myCOPD licence package is purchased solely for pulmonary rehabilitation services, the estimated saving is £8,707 per pulmonary rehabilitation service per year.
The EAC's changes to the cost models more accurately reflect the uptake rate of myCOPD for self-management and pulmonary rehabilitation
3.8 The EAC considered the company's 2 model structures to be appropriate. But it did not agree with the 100% uptake rate in the company's models. For the AECOPD model, the EAC used an uptake rate of 46% based on data from the RESCUE study (North et al. 2020). This rate is the proportion of people who chose to use myCOPD in the RESCUE trial.
3.9 For the PR model, the EAC changed the decision point from when people were referred to a pulmonary rehabilitation service to the point at which people have opted in or shown they would be willing to use myCOPD.
3.10 In the AECOPD model, the EAC base-case results showed that myCOPD saved £86,297 per CCG per year. The saving was influenced by the myCOPD uptake rate and the 90‑day readmission rate. The EAC's threshold analysis suggested that using myCOPD becomes cost incurring if the uptake rate is below 26% or the 90‑day readmission rate is higher than 0.30 admissions per person. It judged that both these values were plausible. At consultation, the company noted that the national activation rate was 48%. The EAC did an additional analysis of the AECOPD model using this activation rate for the uptake, which resulted in a small increase in the cost savings.
3.11 For the PR model, the EAC's base case showed cost savings of £22,779 per CCG per year or £11,093 per pulmonary rehabilitation service per year depending on the licence package purchased. These savings were influenced by the number of people being referred and the uptake rate of pulmonary rehabilitation programmes. The EAC's threshold analysis suggested that using myCOPD becomes cost incurring if:
the referrals to the pulmonary rehabilitation service are below 249 people per year or
the uptake rate is below 9.8% for myCOPD alone or
the uptake rate for myCOPD alone is below 1.9% and the uptake rate for a hybrid option is below 12.2%.
The EAC considered that there was uncertainty around the values used to inform the clinical inputs in the model. (For further details about the threshold analysis, see the EAC assessment report in the supporting documentation for this guidance.)