2019 Feb 20. doi: 10.1089/dia.2018.0328. [Epub ahead of print]

Cost-Effectiveness Analysis of the MiniMed 670G Hybrid Closed-Loop System Versus Continuous Subcutaneous Insulin Infusion for Treatment of Type 1 Diabetes.

Jendle J1Pöhlmann J2de Portu S3Smith-Palmer J2Roze S4.



Hybrid closed-loop (HCL) systems combine continuous glucose monitoring with continuous subcutaneous insulin infusion (CSII) to continuously self-adjust basal insulin delivery. Relative to CSII, HCL improves glycemic control and reduces the risk of hypoglycemia but has higher acquisition costs. The aim of this analysis was to assess the cost-effectiveness of the MiniMed™ 670G HCL system versus CSII in people with type 1 diabetes (T1D) in Sweden.


Cost-effectiveness analysis, from a societal perspective, was performed over patient lifetimes using the IQVIA CORE Diabetes Model. Clinical data were sourced from a study comparing the MiniMed 670G system with CSII in people with T1D. Cost data, expressed in 2018 Swedish krona (SEK), were obtained from Swedish reference prices and published literature.


The MiniMed 670G system was associated with a quality-adjusted life-year (QALY) gain of 1.90 but higher overall costs versus CSII, leading to an incremental cost-effectiveness ratio (ICER) of SEK 164,236 per QALY gained. Use of the HCL system resulted in a lower cumulative incidence of diabetes-related complications. Higher HCL system acquisition costs were partially offset by reduced complication costs and productivity losses. In people with T1D poorly controlled at baseline, the MiniMed 670G system was associated with 2.25 incremental QALYs versus CSII, yielding an ICER of SEK 15,830 per QALY gained.


The MiniMed 670G system was associated with clinical benefits and quality-of-life improvements in people with T1D relative to CSII. At a willingness-to-pay threshold of SEK 300,000 per QALY gained, this HCL system likely represents a cost-effective treatment option for people with T1D in Sweden.

Artikeln i sin helhet som pdf utabn lösenord




www red DiabetologNytt



In 2016, an estimated 42,500 people with type 1 diabetes (T1D) were treated in specialist clinics in Sweden, while ∼400 patients were newly diagnosed with T1D.1 The prevalence of diabetes, including T1D, is projected to increase substantially in Sweden over the coming decades.2 Currently, diabetes is still associated with excess mortality relative to the general population.2,3 A recent study using the Swedish National Diabetes Register data showed that a diagnosis of T1D before reaching 10 years of age was associated with a loss of 16 (95% confidence interval [CI] 14.0–18.1) life-years relative to matched, diabetes-free controls.4 Similarly, all-cause mortality over a median 10 years of follow-up was 1.9 (95% CI 1.71–2.11) deaths per 100,000 person-years in people with T1D, compared with 0.60 (95% CI 0.56–0.66) deaths per 100,000 person-years in diabetes-free controls.4

Good glycemic control is crucial to reduce the incidence of long-term diabetes-related complications.5,6 International diabetes treatment guidelines state that people with T1D should be offered intensive treatment to achieve a glycated hemoglobin (HbA1c) target below 7.0% (53 mmol/mol).7–9 Glycemic control, however, needs to be balanced against the risk of hypoglycemia, leading to recommendations of multiple daily self-monitoring of plasma glucose (SMPG) testing.7,10 In clinical practice in Sweden, only 44% of people with T1D were reported to perform at least four SMPG tests per day.11 More than 50% of patients stated that barriers to SMPG testing were lack of time or simply forgetting to perform the test.

Real-time continuous glucose monitoring (RT-CGM), coupled with continuous subcutaneous insulin infusion (CSII), is an established treatment approach that improves glycemic control, reduces the risk of hypoglycemia, and reduces the SMPG burden.12–14 According to Swedish guidelines, sensor-augmented pump (SAP) therapy, that is, the combination of RT-CGM and CSII, can be offered to people with T1D who suffer recurrent episodes of hyper- or hypoglycemia or who fail to reach glycemic goals with either CGM or CSII alone.7 While data on SAP use in Sweden are still lacking, 23% of people with T1D treated in specialist clinics in 2016 were already equipped with CSII, with frequent hypoglycemia, high HbA1c levels, and glucose variability as the indication for CSII in more than 70% of patients.1

Recent technological advances have allowed to go beyond SAP and develop sensor-driven devices that use glucose monitoring data to adjust basal insulin levels automatically.12,14 These systems, which combine RT-CGM, CSII, and software analyzing CGM data and both predicting and adjusting insulin levels, mimic the function of the pancreas.15 Recently, the first hybrid closed-loop (HCL) system, the MiniMed™ 670G system (Medtronic MiniMed, Northridge, CA), has become available in Europe. The MiniMed 670G system uses the novel SmartGuard™ Auto Mode feature, which self-adjusts basal insulin every 5 min based on real-time sensor glucose readings. Self-adjustment of insulin increases patient time spent in the euglycemic range while reducing the risk of hypoglycemia.16–18 The device is classified as “hybrid” closed-loop as user interaction is still required, for example, patients need to enter their carbohydrate intake to calculate meal-time boluses and must perform regular SMBG testing.

Use of this HCL system was associated with reductions in HbA1c, hypoglycemia and sensor glucose variability in a before/after trial of 30 adolescents and 94 adults with T1D.17,18 In a recent retrospective analysis covering real-world data of the first 3141 patients using the MiniMed 670G system, the system's Auto Mode was shown to be associated with more time spent in the target glucose range relative to baseline.19 The glycemic benefits of the MiniMed 670G system were similar regardless of patients' prior CGM use.18,20 A recent survey showed that diabetes management distress was reduced and diabetes technology perceived more positively in adult and adolescent patients using the system.21

The clinical benefits of the MiniMed 670G system are associated with an increased acquisition cost relative to CSII. It is therefore necessary to balance clinical gains versus incremental costs to establish if the MiniMed 670G system provides good value for money compared with CSII. The aim of the present study was to assess the cost-effectiveness of the MiniMed 670G system versus CSII in people with T1D in Sweden.


In the present study, the MiniMed 670G HCL system was shown to likely be cost-effective versus CSII in people with T1D in Sweden. Treatment with this HCL system, which combines CSII and RT-CGM with the ability to adjust and stop insulin infusion based on CGM data, was associated with gains in life expectancy and improvements in quality of life due to fewer and delayed diabetes-related complications. These clinical benefits came at increased acquisition costs, which were partially offset by lower complication and indirect costs, yielding an ICER of SEK 164,236 per QALY gained for the MiniMed 670G system versus CSII. At a willingness-to-pay threshold of SEK 300,000 per QALY gained, the HCL system was therefore considered to provide good value for money relative to CSII.

In sensitivity analyses, the benefits associated with the MiniMed 670G system were shown to be substantial in people with poorly controlled HbA1c at baseline. In this population with HbA1c ≥7.5% (58 mmol/mol) at baseline (mean baseline HbA1cwas 8.23% [66 mmol/mol]), HbA1c was estimated to be reduced by 0.97% (10.6 mmol/mol) with HCL relative to CSII, leading to a gain of 2.25 QALYs and an ICER of SEK 15,380 per QALY gained. As poorly controlled people with T1D are at high risk of developing long-term diabetes-related complications, improved glycemic control is associated with considerable improvements in clinical and economic benefits.5 In Sweden, population-level HbA1c targets in T1D still had not been met by 2015, indicating room for further improvements.1 Specifically, only 21.2% of people with T1D achieved the HbA1c target below 7.0% (52 mmol/mol) set by Swedish treatment guidelines.1,8 As people with poor glycemic control have an increased risk of all-cause mortality relative to people with good glycemic control and the general, diabetes-free population, these findings indicate that novel treatment options such as HCL, which improve glycemic control, may have beneficial clinical impacts in this population.51 While user input and commitment are currently still required, the MiniMed 670G system can be considered an important step toward reducing user input, which may also help to reduce the detrimental impact of infrequent SMPG monitoring, which is observed in Sweden and associated with worse glycemic control.11,52,53

The incidence and quality-of-life impact of SHEs were also identified as an important driver of cost-effectiveness outcomes. In the before/after study of the MiniMed 670G system, no SHEs were observed while patients were being treated with HCL.17,18 However, even if only a 50% reduction in SHE incidence relative to CSII was assumed, the MiniMed 670G system would be considered cost-effective at a willingness-to-pay threshold of SEK 300,000 per QALY gained. Similarly, even a small utility gain of 0.0184 per annum, due to reduced FoH, was associated with an ICER of SEK 275,591 per QALY gained, that is, below the willingness-to-pay threshold of SEK 300,000 per QALY gained. While further evidence is required on the link between HCL and FoH, it is plausible that patients benefit from increased confidence in their ability to avoid hypoglycemia, thereby reducing FoH, which is associated with reduced quality of life and poor glycemic control.54,55

The present cost-effectiveness analysis has some limitations. The clinical study from which cohort data and treatment effects were derived was a single-arm, nonrandomized, before/after study conducted in the United States.17,18 In addition, the run-in phase (2 weeks) was considerably shorter than the study period (12 weeks), indicating an imbalance in the available data for time periods before and after HCL treatment, which was acknowledged by the study authors.17 As with all clinical studies, investigator/patient interaction might limit the generalizability of results. However, this study of 124 people with T1D was considered the best available evidence on the MiniMed 670G system as the study had more participants and longer follow-up than previous studies of HCL. Notably, clinical trials using other HCL systems have recently become available.56,57 While these trials could not be used for the present cost-effectiveness analyses as they did not investigate the MiniMed 670G system, both studies demonstrated that the use of HCL was associated with improved glycemic control and a lower risk of hypoglycemia. A further potential shortcoming of the analysis is the exclusion of training time from the analysis. Another limitation, which is common to most health economic analyses, was the use of short-term clinical data to project outcomes over patient lifetimes. Long-term projections of clinical and cost outcomes therefore partly relied on assumptions, for example, regarding the impact of HCL treatment on FoH, and were associated with uncertainty. These issues were addressed by using an established validated model of diabetes and by exploring different scenarios in sensitivity analyses. Again, in the absence of long-term, real-world data on the only recently developed HCL system, the clinical study was considered to be the best source for long-term projections using the CDM.22–24

The findings of the present study were aligned with those from a recent report published by the Swedish Dental and Pharmaceutical Benefits Agency, which suggested that the benefits of the MiniMed 670G system relative to the MiniMed 640G device were likely to justify a replacement with the MiniMed 670G system despite slightly higher acquisition costs.58The results also mirrored previous findings, which indicated the benefits of technological advances in insulin delivery. Recently, the DIAMOND and GOLD trials in adult patients with T1D treated with multiple daily injections (MDI) of insulin reported that adding RT-CGM to MDI was associated with improved glycemic control in U.S. and Swedish patients, respectively.59,60 Similarly, switching from MDI to CSII in patients with T1D using RT-CGM improved time-in-range although at the expense of increased RT-CGM-measured hypoglycemia.61 In an observational national diabetes registry study of Swedish people with T1D, insulin pump therapy was found to be associated with lower cardiovascular mortality than MDI treatment over a follow-up of 6.8 years, while another observational study showed larger HbA1c reductions with CSII versus MDI.62,63 While CSII was superior to MDI, several technologies that built on CSII were found to be superior to CSII, in turn. In the INTERPRET study, a 12-month observational study in people with T1D, SAP was shown to be associated with a higher proportion of people with good glycemic control and lower SHE incidence than CSII.48 Similar findings were reported by a systematic review and meta-analysis.50

In cost-effectiveness analyses based on these findings, SAP was found to be cost-effective relative to CSII in a wide range of settings, including France, Denmark, and Sweden.64–66 Taken together, these findings can be interpreted to suggest that technological advances that increase the effectiveness and convenience of insulin delivery, for example, by adding RT-CGM to CSII, generally, although not always, improve clinical and economic outcomes in people with T1D. Evidence on this issue is, for example, available from cost-effectiveness analyses conducted in the U.S. setting although results from this setting may not be directly transferable to Sweden due to differences in health systems and diabetes care.67,68 In a recent within-trial and long-term cost-effectiveness analysis based on the DIAMOND trial, pump use was found to be associated with higher societal costs and reduced life expectancy and QALE in the United States.69 In contrast, adding RT-CGM to MDI in adults with poorly controlled T1D was associated with an ICER of USD 98,108 per QALY gained in the United States, with RT-CGM improving glycemic control and reduced hypoglycemia incidence.70 As noted earlier, the beneficial effect of the MiniMed 670G system on glycemic control was found to be independent of prior RT-CGM treatment as reductions in HbA1cwere observed for people with and without prior RT-CGM.20 The present cost-effectiveness analysis showed that the MiniMed 670G system, by improving glycemic control and reducing hypoglycemia incidence, likely represents good value for money for people with T1D in Sweden.