6 Considerations

Current practice

6.1 The Committee discussed the current care pathway for people with type 1 diabetes in the NHS. It heard from clinical specialists that care for people with type 1 diabetes is usually offered in stages, depending on whether target HbA1c is maintained or if they have had disabling hypoglycaemia. The Committee heard that standard practice is for people to self‑monitor their blood glucose levels with a glucose meter (capillary blood glucose testing) and to have insulin therapy with multiple insulin injections throughout the day (multiple daily insulin injections). It was aware that people who have difficulty maintaining target HbA1c or who have disabling hypoglycaemia with multiple daily insulin injections are eligible for continuous subcutaneous insulin infusion (insulin pump therapy) in accordance with NICE's technology appraisal guidance on continuous subcutaneous insulin infusion for the treatment of diabetes mellitus. The Committee noted that people who continue to have difficulty despite using continuous subcutaneous insulin infusion may sometimes be offered continuous glucose monitoring through individual funding requests, but if they have extreme difficulty they may be considered for an allogeneic pancreatic islet cell transplantation. The Committee concluded that based on current clinical practice, the integrated sensor‑augmented pump therapy systems were most likely to have benefit in people who are being considered for continuous glucose monitoring.

6.2 The Committee discussed the impact of type 1 diabetes and the perceived benefits of the integrated sensor‑augmented pump therapy systems with patient and carer experts. The Committee heard that one of the greatest fears for people with type 1 diabetes and their carers is severe hypoglycaemia. The Committee heard from clinical specialists that around 30% of people with type 1 diabetes have problematic hypoglycaemia, which can affect many aspects of daily life and result in substantial anxiety. The Committee heard that using the integrated sensor‑augmented pump therapy systems could provide greater independence for children and young adults with type 1 diabetes by enabling participation in sports and giving reassurance when they stay away from home. It also acknowledged that the technologies could offer similar advantages to adults. The Committee noted that using the MiniMed Paradigm Veo system with the low‑glucose suspend function could reduce anxiety substantially, particularly in carers of children and young adults who may experience disrupted sleep for many years because of fear of the risk of nocturnal hypoglycaemia. The Committee recognised that the integrated sensor‑augmented pump therapy systems have the potential to offer substantial benefits to people with type 1 diabetes and their carers, particularly for those who have had disabling hypoglycaemia.

Clinical evidence

6.3 The Committee reviewed the evidence available on the clinical effectiveness of the MiniMed Paradigm Veo system, and the Vibe and G4 PLATINUM CGM system. It noted that there was substantial heterogeneity in the characteristics of the included populations, particularly for baseline HbA1c levels, exclusions relating to severe hypoglycaemia, age of patients and previous use of continuous subcutaneous insulin infusion. It also noted that there was heterogeneity in the primary clinical endpoints and length of follow‑up. The Committee concluded that the substantial heterogeneity in the studies limited the number of comparisons that could be drawn from the available data.

6.4 The Committee considered the validity of HbA1c as a primary endpoint in the studies included in the meta‑analyses. The Committee heard from clinical specialists that HbA1c was an established and accepted surrogate outcome measure that has been shown to be associated with clinical outcomes. The Committee concluded that using HbA1c in the analyses was reasonable and provided a link to long‑term outcomes.

6.5 The Committee noted that no published studies reported comparative data for the Vibe and G4 PLATINUM CGM system and so the External Assessment Group had included data from an integrated sensor‑augmented pump therapy system without low‑glucose suspend as a clinical proxy. The Committee acknowledged that there were no data to suggest that the effectiveness of the Vibe and G4 PLATINUM CGM system was equivalent to the proxy system and concluded that in the absence of comparative data, the clinical effectiveness of the Vibe and G4 PLATINUM CGM system was unknown.

6.6 The Committee considered the validity of the network meta‑analyses. It noted that 3 sets of network meta‑analyses were available: 1 for adults, 1 for children and 1 for the economic modelling, which included estimates derived from mixed populations over multiple follow‑up time points. The Committee considered that most of the effect estimates were based on indirect comparisons that were drawn from studies with less than 12 months follow‑up. It also noted that no clinical data were available for 2 of the comparators included in the final scope: continuous glucose monitoring with continuous subcutaneous insulin infusion (non‑integrated) and continuous glucose monitoring with multiple daily insulin injections. The Committee concluded that the effect estimates derived from the network meta‑analyses, particularly those constructed for the economic model, were limited because they were most likely confounded by indirect comparisons, small sample sizes and heterogeneity in both the included populations and duration of follow‑up.

6.7 The Committee considered the results of the network meta‑analyses for adults. It noted that direct evidence from the ASPIRE in‑home study showed that at 3 months follow‑up, the MiniMed Paradigm Veo system significantly reduced both daytime and nocturnal hypoglycaemic event rates compared with an integrated subcutaneous insulin infusion and continuous glucose monitoring system. It also noted that no impact on HbA1c was seen in the ASPIRE in‑home study, although the study was not powered to detect changes in HbA1c. Also, no impact on HbA1c was seen in indirect comparisons with capillary blood testing and continuous subcutaneous insulin infusion and capillary blood testing with multiple daily injections. The Committee concluded that it was plausible that the MiniMed Paradigm Veo system may have clinical benefit for adults with type 1 diabetes who experience frequent episodes of severe hypoglycaemia.

6.8 In addition to the meta‑analyses for adults, the Committee also considered data from 1 randomised cross‑over study (the SWITCH study) and 2 observational studies (Choudhary et al. 2011; Choudhary et al. 2013) that were included in a narrative analysis of supplementary evidence. The Committee noted that although the SWITCH study included a mixed population, it supported using continuous glucose monitoring sensors to decrease the amount of time spent in hypoglycaemia. In addition, the Committee considered that although the 2 observational studies contained small numbers of patients, they provided evidence that in highly selected populations it was plausible that using integrated sensor‑augmented pump therapy systems could have benefit for people with frequent episodes of hypoglycaemia. Further, the Committee noted that Choudhary et al. (2011) provided proof‑of‑concept data for the low‑glucose suspend function of the MiniMed Paradigm Veo system, which appeared to reduce the duration of nocturnal hypoglycaemia in a high‑risk group, and heard from clinical specialists that the study population appeared to be representative of the population in which the MiniMed Paradigm Veo system would be used in practice. The Committee concluded that the supplementary data from these 3 studies supported the conclusions drawn from the network meta‑analyses.

6.9 The Committee considered the results of the network meta‑analyses for children. It noted that direct evidence from Ly et al. (2013) showed that the MiniMed Paradigm Veo system significantly reduced the rate of moderate and severe hypoglycaemic events when compared with capillary blood testing and continuous subcutaneous insulin infusion at 6‑month follow‑up. It also noted that the MiniMed Paradigm Veo system did not have a significant impact on HbA1c at 6 months. The Committee considered that although Ly et al. (2013) included a mixed population, most of the patients (≈70%) were aged less than 18 years and it was the only evidence available to provide an indication of the likely clinical impact of the system in children. The Committee concluded that it was plausible that the MiniMed Paradigm Veo system may have clinical benefit for children with type 1 diabetes who experience frequent episodes of severe hypoglycaemia.

6.10 The Committee discussed the overall strength of the evidence base for the integrated sensor‑augmented pump therapy systems. It considered that, despite limited evidence to suggest that the MiniMed Paradigm Veo system may have benefit in reducing rates of severe hypoglycaemia, the overall evidence base to support using the integrated sensor‑augmented pump therapy systems was weak. The Committee heard from clinical specialists and patient experts that companies often provide online software for people to upload data from their sensor‑augmented pump therapy system to a company‑maintained database. It was also suggested that that these data are rarely analysed and published. The Committee concluded that these technologies and their successive versions could offer substantial benefits to patients and that robust data need to be generated to support the claimed benefits of these technologies and their reimbursement value.

Cost effectiveness

6.11 The Committee considered the cost‑effectiveness analyses for the MiniMed Paradigm Veo system. It noted that 3 base cases were available, but that none included a comparison with continuous glucose monitoring and multiple daily injections because of a lack of clinical‑effectiveness data. The Committee noted that the cost‑effectiveness analyses were done using the IMS CORE Diabetes Model. It discussed the advantages and disadvantages of the model, and heard from experts that although the model was well validated its structure and underlying clinical data were likely to favour interventions that are aimed at reducing HbA1c and associated long‑term complications. The Committee heard from the External Assessment Group that the impacts of short‑term outcomes associated with hypoglycaemic events are more difficult to capture in the model because, unlike change in HbA1c, the rate of hypoglycaemia cannot be changed over time. The Committee also heard from clinical specialists that the risk equations included in the model were dated and may underestimate the risk of long‑term complications for people with type 1 diabetes. The Committee concluded that despite its extensive validation, the IMS CORE Diabetes Model may not be suitable for investigating the impact of interventions on short‑term clinical outcomes, and that it was likely that this impact could have been undervalued in the analyses.

6.12 The Committee considered the treatment‑effect estimates that had been used in the economic modelling. It noted that these treatment effects were limited by the heterogeneity in the meta‑analyses constructed for the economic model. It heard from clinical specialists that in the mixed population base case, it was implausible that HbA1c increases from baseline with continuous subcutaneous insulin infusion because this is an intervention that is intended to reduce HbA1c. Further, it also heard that for the population who experience hypoglycaemia, the assumption that the hypoglycaemic event rate is equivalent for capillary blood testing with continuous subcutaneous insulin infusion and continuous glucose monitoring with continuous subcutaneous insulin infusion was likely to be implausible because the addition of continuous glucose monitoring is intended to reduce the occurrence of hypoglycaemia. The Committee concluded that, in general, the cost‑effectiveness analyses could not be considered robust because the insufficient evidence base for clinical effectiveness leads to a large amount of uncertainty in the incremental clinical‑effect estimates.

6.13 The Committee heard from clinical specialists that the analysis most relevant to using the MiniMed Paradigm Veo system in clinical practice is the base case that includes people who are assumed to have frequent hypoglycaemic events. However, the Committee heard that there may also be a small but clinically significant group of people who have both frequent hypoglycaemia and uncontrolled HbA1c, which may be as a result of brittle diabetes. The Committee also acknowledged that people who have frequent hypoglycaemic events may have increased HbA1c because they try to avoid hypoglycaemia and considered that in practice it was likely that there would be an overlap in the populations modelled for this assessment. However, the Committee concluded that despite the limitations of the clinical‑evidence base, the analysis for people who have frequent hypoglycaemic events should be considered further because of the plausible clinical benefit in this subgroup.

6.14 The Committee considered the cost‑effectiveness analyses for the MiniMed Paradigm Veo system in the severe hypoglycaemia population. The Committee noted that the technology was not cost effective when compared with capillary blood testing with multiple daily injections (£188,100 per quality‑adjusted life year [QALY] gained) and capillary blood testing with continuous subcutaneous insulin infusion (£189,300 per QALY gained), but considered that the incremental cost‑effectiveness ratios (ICERs) for these comparisons were likely to reduce to between £57,900 to £85,300 per QALY gained when treatment effects for HbA1c and an additional utility increment for fear of hypoglycaemia were included. The Committee also noted that the MiniMed Paradigm Veo system dominated when compared with continuous glucose monitoring and continuous subcutaneous insulin infusion (non‑integrated), but acknowledged that this comparison was mainly driven by cost. The Committee concluded that at face value, the analyses suggested that the MiniMed Paradigm Veo system could not be considered cost effective when compared with capillary blood testing with multiple daily injections and capillary blood testing with continuous subcutaneous insulin infusion because of the high incremental cost of the technology.

6.15 The Committee considered the impact of hypoglycaemia and heard from patient and carer experts that fear of severe hypoglycaemic events was associated with a substantial impact on quality of life. Fear of hypoglycaemia may lead to people restricting their daily activities and can cause significant anxiety for carers, particularly parents who may have to wake several times a night to check on their child. The Committee also heard that severe hypoglycaemic events could have a substantial impact on quality of life that extends beyond the perspective of the NHS and personal social services. For example, children may miss school because of a hypoglycaemic event, and adults may need to be absent from work to recover. The Committee concluded that severe hypoglycaemia has a sustained impact on quality of life and can limit usual daily activities for people with type 1 diabetes and their carers.

6.16 The Committee questioned the extent to which the ICERs produced for the severe hypoglycaemia population reflected the benefits associated with a reduction in hypoglycaemic events. The Committee heard from the External Assessment Group that although additional utility increments for fear of hypoglycaemia had been included in scenario analyses, it is possible that the studies from which the utility values were derived had not captured the persistent fear of catastrophic events. Additionally, the Committee heard from clinical specialists that severe hypoglycaemic events may have a substantial impact on ambulance services. The Committee agreed that it was likely that the full impact of hypoglycaemia had not been captured in the model. Further, it considered that persistent fear of hypoglycaemia could be fully captured in the model if published evidence was available. The Committee concluded that although the analysis for people with frequent hypoglycaemia was highly uncertain, it was likely that the benefits of the technologies were not fully captured in the ICER and that it was plausible that the MiniMed Paradigm Veo system with low‑glucose suspend would have benefit in a highly selected population.

6.17 The Committee questioned the extent to which the cost‑effectiveness analyses are applicable to children and young people. The Committee heard from the External Assessment Group that the CORE Diabetes Model could not be adapted for use in children because of the underlying risk equations that were incorporated in the model. The Committee considered that, in general, the ICERs would most likely be lower for children than those available for adults because of the increased risk of adverse outcomes over a longer duration of disease. The Committee also heard from clinical specialists and patient and carer experts that the low‑glucose suspend function of the MiniMed Paradigm Veo system could be of greater benefit in younger children who may have an unpredictable response to insulin and rapid onset of hypoglycaemia. The Committee concluded that it was plausible that the ICERs for the MiniMed Paradigm Veo system would be lower if they were extrapolated to children, particularly those who experience frequent hypoglycaemic events.

6.18 The Committee discussed Roze et al. (2015), which was submitted to the Committee as a pre‑publication manuscript. The Committee noted that the ICERs included in the manuscript were substantially different to those produced by the External Assessment Group's analyses. The Committee heard from the External Assessment Group that the treatment effects used in the different analyses were the primary driver of the differences in the ICERs. The Committee concluded that the more conservative combination of HbA1c changes and hypoglycaemic event rates applied in the analyses by the External Assessment Group provided a fairer assessment of the incremental benefits associated with using integrated sensor‑augmented pump therapy.

Additional considerations

6.19 The Committee heard from a clinical specialist that young children are often considered to have greater sensitivity to insulin than adults, which can make the management of type 1 diabetes more unpredictable in this population. The Committee noted that the use of the MiniMed Paradigm Veo system should be considered for younger children, particularly those of pre‑school age who may be at risk of rapid onset of severe hypoglycaemia.

6.20 The Committee noted that no clinical‑effectiveness data were available for using integrated sensor‑augmented pump therapy systems in pregnant women. The Committee agreed that this was an important subgroup and wished to encourage further research in this area.

6.21 The Committee considered the rapid pace of development of new technologies designed to help with monitoring blood glucose levels in people with type 1 diabetes, and noted the importance of the progress in developing closed‑loop systems and the artificial pancreas. The Committee acknowledged that new versions of the integrated sensor‑augmented pump therapy systems were becoming available to the NHS and wished to encourage evidence generation on the clinical effectiveness of these new technologies.

  • National Institute for Health and Care Excellence (NICE)