The effect of closed-loop glucose control on C-peptide secretion in youth with newly diagnosed type 1 diabetes: the CLOuD RCT

Article history

The research reported in this issue of the journal was funded by the EME programme as award number 14/23/09. The contractual start date was in February 2016. The draft manuscript began editorial review in August 2022 and was accepted for publication in November 2023. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The EME editors and production house have tried to ensure the accuracy of the authors’ manuscript and would like to thank the reviewers for their constructive comments on the draft document. However, they do not accept liability for damages or losses arising from material published in this article.

Copyright statement

Copyright © 2024 Boughton et al. This work was produced by Boughton et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This is an Open Access publication distributed under the terms of the Creative Commons Attribution CC BY 4.0 licence, which permits unrestricted use, distribution, reproduction and adaptation in any medium and for any purpose provided that it is properly attributed. See: https://creativecommons.org/licenses/by/4.0/. For attribution the title, original author(s), the publication source – NIHR Journals Library, and the DOI of the publication must be cited.

2024 Boughton et al.

Preservation of C-peptide: evidence and approach

Type 1 diabetes is characterised by autoimmune destruction of pancreatic beta-cells. 9 Loss of beta-cells is gradual10 and at the clinical diagnosis of diabetes most people with type 1 diabetes have residual pancreatic beta-cells which can continue to secrete insulin for several additional years. Amelioration of hyperglycaemia after diagnosis allows partial recovery of beta-cell insulin secretory function, leading to a ‘honeymoon period’ with relatively low exogenous insulin requirements. 9 In the DCCT, 35% of participants with diabetes duration of 1–5 years had persistent islet cell function (meal-stimulated C-peptide levels of 0.2–0.5 pmol/ml). 7

Persistence of residual functioning beta-cells, measured by C-peptide secretion, is associated with improved glycaemic control, reduced risk of hypoglycaemia and lower incidence of microvascular complications. 11,12 In the DCCT, those who had ≥ 0.20 pmol/ml C-peptide initially or sustained over a year had markedly fewer complications – a 79% decrease in the relative risk of retinopathy. 12 Importantly, these benefits were seen in the face of less hypoglycaemia events. Individuals in the intensive treated group with ≥ 0.20 pmol/ml C-peptide had about the same frequency of severe hypoglycaemia as those in the standard care group; a 30% reduction as compared to those in intensive therapy without this level of C-peptide. A linear relationship between frequency of retinopathy progression and C-peptide as low as 0.03 pmol/ml has been reported. 11 Islet transplant studies have also shown that even small amounts of residual beta-cell function are clinically important. Vantyghem et al. showed that, while significant beta-cell function was required to improve mean glucose, lower glucose excursions, and result in insulin independence, participants who maintained minimal beta-cell function experienced almost no severe hypoglycaemic events. 13 Interventions which can preserve endogenous insulin secretion prior to and following clinical diagnosis of type 1 diabetes are therefore clinically important. 14,15

Assignment to the intensively managed group in the DCCT reduced the risk for loss of C-peptide by 57% over the mean 6.5 years of study. This suggests that metabolic control soon after the onset of type 1 diabetes may have a significant effect on preservation of residual islet cell function. However, intensification of insulin therapy inevitably hits the barrier of hypoglycaemia. 16

Previous studies have investigated whether an early period of islet cell rest achieved by intensive glycaemic control following diagnosis of type 1 diabetes can prevent the decline in endogenous insulin secretion with conflicting results. An early exploratory study in adolescents reported improved stimulated C-peptide levels (0.51 pmol/ml) at 12 months following a period of intensive insulin treatment in hospital for 2 weeks after diagnosis. 17 However, a more recent study applying a short period of hybrid closed-loop (CL) within 7 days of diagnosis, followed by sensor-augmented pump therapy, did not alter C-peptide secretion at 12 months compared with standard care, but there was no difference in glucose control between the two treatment groups over the 12-month study period. 18

It has yet to be determined whether sustained intensive glycaemic control following diagnosis can ameliorate the decline in endogenous insulin secretion in youth with type 1 diabetes.

Closed-loop technology

The emergence of new technologies including continuous glucose monitoring (CGM),19 sensor-augmented pump therapy (SAP)20 and threshold pump suspend21,22 provides new opportunities to improve outcomes for people with type 1 diabetes. The most promising approach is CL insulin therapy23 which combines real-time CGM with insulin pump therapy to achieve glucose responsive subcutaneous (s.c. ) insulin delivery. The vital component of such a system, also known as an artificial pancreas (AP), is a computer-based algorithm. The role of the control algorithm is to translate, in real time, the information it receives from the CGM and to compute the amount of insulin to be delivered by the pump. The other components include a real-time continuous glucose monitor and an infusion pump to titrate and deliver insulin.

Rationale for the study

The purpose of the study is to test the impact of continued intensive metabolic control using CL insulin delivery after diagnosis on preservation of C-peptide residual secretion. The study enrols children aged 10 and older, as they are characterised by higher residual C-peptide secretion at diagnosis compared to younger children. The present study will also test the feasibility and acceptance of this therapy so that it could be considered as a standard treatment modality in the future.

External funding has been secured to continue with the treatment allocated by randomisation for 48 months to allow ongoing assessment of the impact of continued intensive metabolic control using CL insulin delivery on residual C-peptide and will test acceptability of this therapy over a longer duration.

Hypothesis

We hypothesised that a sustained period of intensive glucose control with hybrid CL for 12 months following diagnosis of type 1 diabetes in children and adolescents can preserve C-peptide secretion compared to standard insulin therapy.

Primary objective

The primary objective is to evaluate the effect of continued intensive metabolic control using CL insulin delivery after diagnosis on preservation of C-peptide residual secretion by comparing the area under the stimulated C-peptide curve (AUC) of a mixed-meal glucose tolerance test conducted at the 12-month visit in participants receiving CL insulin delivery with those receiving standard therapy – that is, multiple daily injections applying basal bolus regimen.

The objective of the internal pilot phase is to carry out preliminary evaluation of recruitment, randomisation, treatment and follow-up assessments at the five participating sites.

Secondary objectives

Internal pilot phase

The purpose of the internal pilot study was to estimate the rate of recruitment, and to pilot randomisation, treatment and follow-up assessments at the initial participating sites. During the pilot phase, we aimed to recruit at least 10 subjects – that is, 2 per site. All participants recruited during the pilot phase proceeded to the full study.

Full study

Following the internal pilot phase and consecutive re-evaluation of recruitment procedures and follow-up assessments, recruitment for the study resumed at full rate, and all randomised subjects were followed up until study completion. External funding was secured to allow participants to continue with the treatment allocated at randomisation for a further 12 months. Study processes are therefore referred to over the 24-month study period, but this report focuses on the 12-month results which includes the primary end point. The study flow chart is outlined in Figure 1.

FIGURE 1.

Study flow chart.

Extension phase

At 24 months, all participants were invited to participate in an optional extension phase to continue with their current treatment (automated CL glucose control or standard insulin therapy) for a further 24 months. Permission has also been sought from participants and their carers for ongoing submission of routine clinical data to the research team for a further 9 years to enable long-term outcomes to be reported.

Approval was received from Cambridge East Research Ethics Committees (16/EE/0286) and Medicines and Healthcare products Regulatory Agency. Safety aspects were overseen by an independent data safety monitoring board. The study is registered with clinicaltrials.gov (NCT02871089).

Study participants

Participants were recruited from seven paediatric diabetes clinics in the UK (Cambridge, Edinburgh, Leeds, Liverpool, Nottingham, Oxford, Southampton). Eligible participants were identified by clinical teams at each centre. Participants aged 16 years and parents/guardians of participants < 16 years gave written informed consent. Written assent was obtained from participants < 16 years.

Following the initial 24 months of the study, all participants opting to continue with the extension phase of the study were re-consented.

Randomisation

Eligible participants were randomised after baseline mixed-meal tolerance test (MMTT) using block randomisation (block sizes of two and four were used with equal probability) and central randomisation software to CL therapy or standard insulin therapy. Randomisation was stratified by site and age (10–13 years and 14–16 years); the randomisation ratio was 1 : 1 within each stratum.

Closed-loop system

The Cambridge MPC algorithm (version 0.3.71) was run in two hardware configurations, the initial FlorenceM configuration followed by the CamAPS FX (CamDiab, Cambridge, UK) configuration to improve usability and therapy adherence (Figure 2).

FIGURE 2.

Hybrid CL configurations used in the CL group. Panel A shows FlorenceM configuration and panel B shows CamAPS FX configuration.

FlorenceM

The FlorenceM configuration comprised a locked smartphone (Samsung Galaxy S4, South Korea) running an app with the Cambridge control algorithm (version 0.3.71), a Medtronic prototype phone enclosure with an embedded modified Carelink USB to allow the smartphone to wirelessly communicate with a modified Medtronic MiniMed^TM 640G insulin pump (Medtronic, Northridge, CA, USA). This pump had low glucose suspend enabled and received glucose sensor data from the Medtronic MiniMed^TM GuardianTM 3 sensor, which requires finger-stick calibrations.

CamAPS FX

The CamAPS FX configuration superseded FlorenceM in July 2019. The CamAPS FX system comprised an unlocked smartphone (Samsung Galaxy S8, South Korea) hosting the CamAPS FX app running the Cambridge control algorithm (version 0.3.71), which communicated wirelessly with both the Dana Diabecare RS insulin pump (Sooil, Seoul, South Korea) and Dexcom G6 transmitter (Dexcom, San Diego, CA, USA).

In both configurations, algorithm-driven insulin delivery was adjusted automatically every 8–12 minutes, with the app-based control algorithm communicating the insulin infusion rate to the insulin pump wirelessly. The control algorithm was initialised using total daily insulin dose and body weight, and incorporated adaptive learning with regard to total daily insulin requirements, diurnal variations, meal patterns, and duration of insulin action.

In both configurations, when auto mode was not operational, the insulin pump reverted to pre-programmed basal rates. The treat-to-target adaptive control algorithm had a nominal glucose target level of 5.8 mmol/l, which was adjustable in the CamAPS FX configuration between 4.0 mmol/l and 11.1 mmol/l across different times of day. The CamAPS FX app contained a bolus calculator to initiate bolus delivery from the phone, a user-selectable ‘Ease-off’ mode to reduce insulin delivery around activity/exercise, and a ‘Boost’ mode to intensify insulin delivery when insulin needs were elevated. The CamAPS FX app streamed data to Diasend data ecosystem (Glooko/Diasend, Sweden).

Procedures

Assays

C-peptide and glucose were measured centrally (Swansea University, Swansea, UK); C-peptide was measured using a sensitive, luminescence immunoassay (IV2-004, Invitron, UK) and glucose using a glucose oxidase method (YSI 2300 stat plus, YSI Life Sciences, USA). HbA1c was measured centrally (Swansea University) using an International Federation of Clinical Chemistry and Laboratory Medicine (IFCC)-aligned method and following National Glycohemoglobin Standardisation Program (NGSP) standards; Tosoh GX (Tosoh Bioscience, UK). Lipid profile was measured locally.

Study end points

The primary end point was the between-group difference in mean stimulated C-peptide AUC of the 12-month MMTT. Key secondary end points included time in target glucose range 3.9–10.0 mmol/l, HbA1c, and time < 3.9 mmol/l at 12 months tested sequentially to control the type I error. Sensor glucose end points were based on data from a masked glucose sensor worn for 14 days.

Additional secondary end points included mean stimulated C-peptide at 6 and 24 months, plasma glucose and fasting C-peptide divided by fasting glucose at 6, 12 and 24 months during MMTT, HbA1c at 6 and 24 months, percentage of participants in each group with HbA1c < 7.5% (58 mmol/mol) at 12 and 24 months. Secondary end points based on masked sensor glucose data collected every 3 months included time in range, mean glucose, standard deviation (SD) and coefficient of variation of glucose, time spent in hyperglycaemia, time with glucose < 3.5 mmol/l, < 3.0 mmol/l and < 2.8 mmol/l and AUC of glucose < 3.9 mmol/l and < 3.5 mmol/l. Insulin metrics (total, basal and bolus insulin dose), BMI SD score, blood pressure and lipid profile were compared between groups at 12 and 24 months.

Safety evaluation comprised the frequency of severe hypoglycaemia and diabetic ketoacidosis (DKA), and other AEs and SAEs.

Power calculation

The primary analysis compares the difference between groups in the 2-hour AUC-mean using the ln(mean C-peptide + 1). The residual SD of the ln(x + 1) transformed C-peptide AUC analysis of covariance is referred to by TrialNet44 as the root mean squared error (RMSE). The back-transform, exp(y) – 1, of the mean of the transformed values is referred to by TrialNet as the geometric-like mean. In the DirecNet/TrialNet new-onset studies,44 the point estimate for RMSE was 0.18 (transformed scale) and was used in the power calculations. As in the TrialNet sample size calculations, a 50% improvement was assumed in the geometric-like mean C-peptide AUC. The sample size depends on the geometric-like mean value in the control group. The original TrialNet sample size calculations assumed a value of 0.37 pmol/ml for the control group based on the lower 90% confidence limit from previous data. 44 The present power calculation applied the same value. A 50% increase in the intervention group of the geometric-like mean C-peptide AUC gives 0.37*1.50 = 0.555 pmol/ml. After ln(x + 1) transformation, the mean values in the control and treatment groups are 0.315 and 0.441 (transformed scale), respectively, the treatment effect is 0.441–0.315 = 0.126. The treatment effect of 0.126 with a SD of 0.18 requires 44 subjects per group at 90% power for a two-sided test at the 0.05 level. Allowing for 10% loss to follow-up means we aimed to recruit a total of 96 randomised participants (48 per group).

Statistical analysis

Primary analyses were performed on an intention-to-treat basis with each participant analysed according to the treatment assigned by randomisation. All randomised participants were included in the intention-to-treat population and all enrolled participants were included in the safety cohort. Comparison of safety outcomes between the two treatment groups only included those events occurring on or after randomisation. All randomised participants with at least one CGM reading were included in the CGM analyses following a modified intention-to-treat approach. Treatment interventions were compared using a longitudinal mixed effects linear model adjusting for baseline value, gender, presence/absence of DKA at diagnosis and age as fixed effects, and clinical site as a random effect. A 95% CI was reported for the difference between the interventions based on the model. The log (C-peptide AUC + 1) values at 12 months post randomisation were compared for the primary analysis. For highly skewed residuals for secondary outcomes winsorisation was used. Mixed effects regression models addressed missing data by using maximum likelihood estimation incorporating data from all randomised participants, which assumes data were missing at random.

A per-protocol analysis restricted to participants in the CL group who used the system at least 60% of the time and those in the control group who did not start insulin pump therapy was conducted.

The primary and key end points comparing between-group differences at 12 months post diagnosis were tested sequentially to maintain a type 1 error rate of 5%. If C-peptide AUC was significant at α = 0.04, then key end points were tested at α = 0.05. Otherwise, they were tested at α = 0.01. If all three key end points were significant at α = 0.01, then this alpha was recycled to the primary outcome and C-peptide AUC was tested at α = 0.05. Secondary end points were adjusted for multiple comparisons to control the false discovery rate using the two-stage adaptive Benjamini–Hochberg method. 45 Analyses were conducted with SAS software version 9.4 (SAS Institute, Inc.).

Health economics analysis was not included in the NIHR Efficacy and Mechanism Evaluation contract and is therefore not included in this report.

Participant withdrawal criteria

The following pre-randomisation withdrawal criterion will apply:

• Subject/Family is unable to demonstrate safe application of MDI during run-in period as judged by the investigator.

The following pre- and post-randomisation withdrawal criteria will apply:

• Subject is unable to demonstrate safe use of MDI or study insulin pump and/or CGM during post-randomisation training period as judged by the investigator.

• Subject fails to demonstrate compliance with MDI therapy or study insulin pump and/or CGM during post-randomisation training period.

• Subjects may terminate participation in the study at any time without necessarily giving a reason and without any personal disadvantage.

• Significant protocol violation or non-compliance.

• Recurrent severe hypoglycaemia events not related to the use of the CL system.

• Recurrent severe hyperglycaemia event/DKA unrelated to infusion site failure and related to the use of the CL system.

• Decision by the investigator or the Sponsor that termination is in the subject’s best medical interest.

• Allergic reaction to insulin.

• Allergic reaction to adhesive surface of infusion set or glucose sensor.

• If patient cannot be contacted in 12 weeks subject will be considered lost to follow-up.

Protocol and statistical analysis plan amendments can be found in Appendices 2 and 3.

Recruitment

Full study recruitment

Table 3 shows the recruitment success rate overall and by individual site. We applied no exclusions at enrolment such as technology propensity or healthcare professional considerations about suitability, minimising selection bias. The study population is therefore representative of the general population of youth newly diagnosed with type 1 diabetes, improving generalisability of the findings.

TABLE 3 - Recruitment success rate

T1D, type 1 diabetes mellitus.

Site	Number of newly diagnosed T1D patients aged 10–16.9 years	Number of patients approached (given PIS)	Number of patients eligible for CLOuD	Number of participants recruited to CLOuD	Recruitment success rate (%)
1	37	35	33	19	57.6
2	15	15	14	6	42.9
3	23	23	23	12	52.2
4	25	21	20	17	85.0
5	29	26	26	21	80.8
6	32	30	27	14	51.9
7	21	21	19	12	63.2
Total	182	171	162	101	62.4

The reasons potential participants declined to take part in CLOuD were recorded and are shown in Figure 3. The most common reason for declining to take part in the study was concerns using an insulin pump (21%) or being uninterested in research (16%). Recruitment completed in July 2019. The cumulative rate of recruitment is shown in Figure 4.

FIGURE 3.

Reasons given for declining to participate in the CLOuD study (%).

FIGURE 4.

Rate of recruitment.

Between 6 February 2017 and 18 July 2019, 101 participants were enrolled. Four participants withdrew prior to randomisation; 97 participants were randomised [mean ± SD age 12 ± 2 years, 44% (n = 43) female, baseline HbA1c 93 ± 18 mmol/mol (10.6 ± 1.7%) and 29% (n = 28) presented in DKA at diagnosis], 51 to the CL group and 46 to control group (Table 4). Mean ± SD time to randomisation from diagnosis was 9.5 ± 6.2 days (Figure 5).

TABLE 4 - Characteristics of study participants at baseline and according to treatment group

DKA, diabetic ketoacidosis.

Note

Data are n (%) or mean ± SD.

	Overall (N = 97)	Closed-loop (N = 51)	Control (N = 46)
Age (years)	12 ± 2	12 ± 2	12 ± 2
Distribution, n (%)
10–13 years	79 (81)	41 (80)	38 (83)
14–< 17 years	18 (19)	10 (20)	8 (17)
Sex, n (%)
Female	43 (44)	25 (49)	18 (39)
Male	54 (56)	26 (51)	28 (61)
BMI percentile	52 ± 31	53 ± 29	51 ± 34
Ethnicity, n (%)
White	79 (81)	44 (86)	35 (76)
Black	3 (3)	1 (2)	2 (4)
Asian	6 (6)	2 (4)	4 (9)
More than one race	5 (5)	4 (8)	1 (2)
Unknown/not reported	4 (4)	0 (0)	4 (9)
Baseline glycated haemoglobin (mmol/mol)	93 ± 18	94 ± 20	91 ± 17
Baseline glycated haemoglobin (%)	10.6 ± 1.7	10.7 ± 1.8	10.5 ± 1.6
Presence of DKA at diagnosis, n (%)	28 (29)	17 (33)	11 (24)

FIGURE 5.

Participant flow CONSORT diagram.

Retention

There were 10 post-randomisation withdrawals by 12 months, 4 in the CL group and 6 in the control group. Two participants, one in each treatment group, were withdrawn by the clinic due to safety concerns and the other eight participant withdrawals were voluntary (Table 5).

Primary and key end points

Primary and key end points for all randomised participants at 12 months are shown in Table 6. There was no difference in C-peptide AUC between treatment groups at 12 months {primary end point: geometric mean CL 0.35 pmol/ml [interquartile range (IQR) 0.16–0.49] compared with control 0.46 pmol/ml (IQR 0.22–0.69); mean adjusted difference –0.06 pmol/ml (95% CI –0.14 to 0.03)}.

The proportion of time in target range 3.9–10.0 mmol/l based on masked LibrePro sensor glucose data at 12 months was 10 percentage points (95% CI 2 to 17) higher in the CL group (mean ± SD 64 ± 14%) compared to control group (mean ± SD 54 ± 23%) (Table 6). As this end point did not reach the threshold of 0.01 in the analysis, other key secondary end points were not tested for statistical significance. HbA1c was lower in the CL group by 4 mmol/mol (0.4%) [95% CI 0 to 8 mmol/mol (0.0% to 0.7%)] at 12 months. The mean difference in time spent < 3.9 mmol/l between groups was 0.9 percentage points (95% CI –1.0 to 2.8).

Secondary end points

Secondary end points for all randomised participants at 12 months are shown in Table 6.

C-peptide end points

C-peptide AUC declined following diagnosis in both treatment groups (see Table 6, Figures 6 and 7). Plasma glucose AUC was similar between groups at 12 months. There was no difference in fasting C-peptide divided by fasting glucose between treatment groups at 12 months. The proportion of participants with negative C-peptide stimulation in response to mixed-meal test (defined as non-fasted C-peptide < 0.2 pmol/ml), increased in both groups from baseline to 12 months but was similar between treatment groups and is provided in Table 7.

FIGURE 6.

The area under the curve for plasma C-peptide in response to a mixed-meal tolerance test at baseline, 6 and 12 months post diagnosis of type 1 diabetes. Panel A; geometric mean [IQR] and the glycated haemoglobin from baseline to 12 months. Panel B; median [IQR]. C-peptide AUC at baseline (n = 49 CL, n = 45 control), 6 months (n = 45 CL, n = 38 control), 12 months (n = 46 CL, n = 37 control). HbA1c at baseline (n = 51 CL, n = 46 control), 3 months (n = 48 CL, n = 44 control), 6 months (n = 46 CL, n = 42 control), 9 months (n = 46 CL, n = 38 control), 12 months (n = 46 CL, n = 39 control).

FIGURE 7.

Stimulated C-peptide at MMTT at baseline (panel A), 6 (panel B) and 12 (panel C) months. Circles indicate the median, and the bars represent the 25th and 75th percentiles. Baseline (n = 49 CL, n = 45 control), 6 months (n = 45 CL, n = 38 control), 12 months (n = 46 CL, n = 37 control).

TABLE 7 - Participants with negative C-peptide stimulation in response to MMTT

Note

Data are presented as n (%).

Negative C-peptide stimulation is defined as non-fasted C-peptide < 0.2 pmol/ml.

	Closed-loop	Control
Baseline	(n = 51)	(n = 46)
	1 (2)	1 (2)
6 months	(n = 45)	(n = 41)
	4 (9)	3 (7)
12 months	(n = 46)	(n = 38)
	10 (22)	5 (13)

Glycaemic end points

End points by treatment group at 6 months is shown in Table 8. C-peptide AUC was lower in the CL group compared to the control group (0.51 vs. 0.70 pmol/ml). Time in target glucose range was higher (70 vs. 65%) and mean glucose lower (8.0 vs. 8.9 mmol/l) in the CL group compared to control group at 6 months but HbA1c was similar. Time in hypoglycaemia (< 3.9 mmol/l) was higher (6.1 vs. 4.2%) in the CL group compared to control group at 6 months.

TABLE 8 - End points by treatment group at 6 months

Reporting geometric mean as the analysis was done using log transformed values.

Measured during MMTT.

Variable winsorised at the 10th and 90th percentiles.

Based on masked sensor glucose data provided by Freestyle LibrePro over up to 14 days.

Note

Data are mean ± SD or geometric mean (IQR).

	Closed-loop	Control
C-peptide AUC (pmol/ml)a,b	(N = 45)	(N = 38)
	0.51 (0.36–0.71)	0.70 (0.44–0.89)
Time spent at glucose level 3.9–10.0 mmol/l (%)	(N = 44)	(N = 37)
	70 ± 15	65 ± 22
	(N = 46)	(N = 42)
HbA1c (%)	6.7 ± 0.9	6.7 ± 0.9
HbA1c (mmol/mol)	49 ± 9	50 ± 9
Time spent at glucose level < 3.9 mmol/l (%)c	(N = 44)	(N = 37)
	6.1 ± 5.7	4.2 ± 3.9
CGM end pointsd	(N = 44)	(N = 37)
Mean glucose (mmol/l)	8.0 ± 1.9	8.9 ± 2.7
Standard deviation (mmol/l)	3.1 ± 0.9	3.3 ± 1.3
CV of glucose (%)	38 ± 6	36 ± 7
Time spent at glucose level (%)
< 3.5 mmol/lc	3.4 ± 3.4	2.3 ± 2.4
< 3.0 mmol/lc	1.4 ± 1.7	0.8 ± 0.9
< 2.8 mmol/lc	1.0 ± 1.3	0.5 ± 0.6
> 10.0 mmol/lc	23 ± 16	30 ± 21
> 16.7 mmol/lc	2.5 ± 3.6	4.8 ± 6.4
Area above curve 3.9 mmol/l (mmol/l)	0.7 ± 0.7	0.4 ± 0.4
Area above curve 3.5 mmol/l (mmol/l)	0.3 ± 0.4	0.2 ± 0.2
HbA1c < 7.5%, n (%)	(N = 46)	(N = 42)
	38 (83)	33 (79)
Insulin end points (U/kg/day)	(N = 45)	(N = 40)
Total daily insulin	0.73 ± 0.33	0.69 ± 0.41
Total daily basal insulin	0.33 ± 0.16	0.29 ± 0.20
Total daily bolus insulin	0.40 ± 0.23	0.40 ± 0.24
Fasting C-peptide divided by fasting glucose (pmol/ml per mmol/l)b	(N = 44)	(N = 41)
	0.04 ± 0.02	0.05 ± 0.04
Plasma glucose AUC (mmol/l)b	13.0 ± 2.7	13.3 ± 2.9
	(N = 42)	(N = 40)
BMI percentile (%)	66 ± 26	60 ± 33
Blood pressure (mmHg)	(N = 46)	(N = 41)
Systolicc	111 ± 7	109 ± 10
Diastolicc	64 ± 8	66 ± 8

Day and night glucose control is shown in Table 9. Daytime glucose control deteriorated in both groups from baseline to 12 months. Daytime glucose control measured by time in target glucose range and mean glucose was better in the CL group than the control group at 12 months. Night-time glucose control deteriorated in the control group from baseline to 12 months (72% at baseline and 57% at 12 months) but was maintained in the CL group (75% at baseline and 71% at 12 months).

TABLE 9 - Day and night glucose control by treatment group

Variable winsorised at the 10th and 90th percentiles.

Note

Date are presented as mean ± SD using masked sensor glucose data provided by Freestyle LibrePro over up to 14 days.

	Daytime (8:00–23:59)				Night-time (00:00–07:59)
	Baseline		12 months		Baseline		12 months
	Closed-loop	Control	Closed-loop	Control	Closed-loop	Control	Closed-loop	Control
	(N = 50)	(N = 43)	(N = 44)	(N = 33)	(N = 50)	(N = 43)	(N = 44)	(N = 33)
Time in range 3.9–10.0 mmol/l (%)	73 ± 14	72 ± 13	60 ± 15	52 ± 23	75 ± 16	72 ± 17	71 ± 15	57 ± 26
Mean glucose (mmol/l)	7.3 ± 1.7	7.2 ± 1.7	8.9 ± 1.7	10.1 ± 3.4	7.0 ± 1.6	6.5 ± 1.9	7.5 ± 1.6	9.1 ± 3.1
Glucose SD (mmol/l)	2.8 ± 0.7	2.7 ± 0.8	3.7 ± 1.0	4.0 ± 1.5	2.4 ± 0.8	2.3 ± 0.9	2.9 ± 0.9	3.0 ± 1.2
Time < 3.0 mmol/l (%)a	1.5 ± 1.6	1.9 ± 1.9	0.9 ± 0.9	1.5 ± 1.8	2.6 ± 4.1	4.0 ± 4.8	2.2 ± 2.4	1.7 ± 2.5

Longitudinal sensor glucose end points are shown in Table 10 and Figure 8. Although glucose control as measured by sensor glucose metrics time in range, time above range and mean glucose deteriorated in both groups from baseline to 12 months, this was more pronounced in the control group. There was a trend towards a reduction in the proportion of time spent in hypoglycaemia in both groups over time.

TABLE 10 - Longitudinal analyses of continuous glucose monitor metrics by treatment group

Variable winsorised at 10th and 90th percentiles.

Note

Data presented as mean ± SD using masked sensor glucose data provided by Freestyle LibrePro over up to 14 days.

	Baseline	3 months	6 months	9 months	12 months
	Closed-loop (N = 50), control (N = 43)	Closed-loop (N = 46), control (N = 41)	Closed-loop (N = 44), control (N = 37)	Closed-loop (N = 42), control (N = 35)	Closed-loop (N = 44), control (N = 33)
Time in range 3.9–10.0 mmol/l (%)
Closed-loop	74 ± 14	73 ± 13	70 ± 15	67 ± 14	64 ± 14
Control	72 ± 13	73 ± 15	65 ± 22	65 ± 19	54 ± 23
Mean glucose (mmol/l)
Closed-loop	7.2 ± 1.6	7.6 ± 1.6	8.0 ± 1.9	8.0 ± 1.6	8.5 ± 1.6
Control	7.0 ± 1.6	7.4 ± 1.7	8.9 ± 2.7	8.8 ± 2.0	9.8 ± 3.3
Glucose SD (mmol/l)
Closed-loop	2.7 ± 0.7	2.8 ± 0.9	3.1 ± 0.9	3.3 ± 1.0	3.6 ± 0.9
Control	2.7 ± 0.8	2.7 ± 0.9	3.3 ± 1.3	3.4 ± 1.1	3.7 ± 1.4
Glucose CV (%)
Closed-loop	38 ± 7	37 ± 7	38 ± 6	41 ± 8	42 ± 7
Control	39 ± 7	37 ± 6	36 ± 7	38 ± 7	39 ± 8
Time < 3.9 mmol/l (%)a
Closed-loop	9.1 ± 6.3	6.3 ± 5.3	6.1 ± 5.7	7.5 ± 5.2	6.2 ± 3.8
Control	10.7 ± 7.1	7.0 ± 5.4	4.2 ± 3.9	4.4 ± 3.2	5.4 ± 4.7
Time < 3.5 mmol/l (%)a
Closed-loop	5.2 ± 4.8	3.7 ± 3.8	3.4 ± 3.4	4.6 ± 4.0	3.6 ± 2.6
Control	6.6 ± 5.2	3.8 ± 3.7	2.3 ± 2.4	2.4 ± 2.1	3.3 ± 3.1
Time < 3.0 mmol/l (%)a
Closed-loop	2.0 ± 2.5	1.6 ± 1.9	1.4 ± 1.7	2.2 ± 2.6	1.4 ± 1.2
Control	2.8 ± 2.8	1.3 ± 1.8	0.8 ± 0.9	1.1 ± 1.3	1.5 ± 1.6
Time < 2.8 mmol/l (%)a
Closed-loop	1.3 ± 1.8	1.1 ± 1.4	1.0 ± 1.3	1.6 ± 2.1	0.9 ± 0.9
Control	1.9 ± 2.2	0.8 ± 1.1	0.5 ± 0.6	0.7 ± 0.9	1.1 ± 1.4
Area over curve < 3.9 mmol/l (mmol/l)a
Closed-loop	1.0 ± 0.9	0.7 ± 0.8	0.7 ± 0.7	0.9 ± 0.8	0.7 ± 0.5
Control	1.3 ± 1.1	0.7 ± 0.7	0.4 ± 0.4	0.5 ± 0.4	0.6 ± 0.6
Area over curve < 3.5 mmol/l (mmol/l)a
Closed-loop	0.5 ± 0.5	0.4 ± 0.5	0.3 ± 0.4	0.5 ± 0.5	0.3 ± 0.3
Control	0.6 ± 0.6	0.3 ± 0.4	0.2 ± 0.2	0.2 ± 0.3	0.3 ± 0.4
Time > 10.0 mmol/l (%)a
Closed-loop	15 ± 9	19 ± 14	23 ± 16	24 ± 11	29 ± 14
Control	14 ± 10	18 ± 13	30 ± 21	30 ± 16	40 ± 25
Time > 16.7 mmol/l (%)a
Closed-loop	1.0 ± 1.6	1.9 ± 2.7	2.5 ± 3.6	2.5 ± 2.7	4.2 ± 3.8
Control	1.0 ± 1.5	1.4 ± 2.4	4.8 ± 6.4	3.1 ± 3.1	10.0 ± 12.4

FIGURE 8.

Longitudinal glycaemic control from baseline to 12 months based on masked sensor glucose data collected by Freestyle LibrePro for up to 14 days. Panel A shows mean glucose levels. Panel B shows time in target glucose range 3.9–10.0 mmol/l. Panel C shows time with glucose above 10.0 mmol/l. Panel D shows time with glucose below 3.9 mmol/l. Full circles indicate the median, and the bars represent the 25th and 75th percentiles. Baseline (n = 50 CL, n = 43 control), 3 months (n = 46 CL, n = 41 control), 6 months (n = 44 CL, n = 37 control), 9 months (n = 42 CL, n = 35 control), 12 months (n = 44 CL, n = 33 control).

C-peptide AUC by HbA1c, coefficient of variation of glucose, and time in target range 3.9–10.0 mmol/l are shown in Table 11 and Figure 9. Higher C-peptide AUC was associated with lower glucose variability as measured by coefficient of variation of glucose, and higher time in target glucose range (see Figure 9). There was no clear relationship between C-peptide AUC quartiles and HbA1c quartiles (see Table 11).

TABLE 11 - Quartiles of C-peptide AUC by quartiles of HbA1c among participants using CL

Note

Data are presented as n (%).

HbA1c (mmol/mol) quartiles	C-peptide AUC (pmol/ml) quartiles
	< 0.16	0.16–< 0.32	0.32–< 0.49	≥ 0.49
< 46	1 (9)	5 (42)	2 (17)	4 (36)
46–< 50.5	6 (55)	1 (8)	1 (8)	3 (27)
50.5–< 57	3 (27)	2 (17)	4 (33)	2 (18)
≥ 57	1 (9)	4 (33)	5 (42)	2 (18)

FIGURE 9.

Relationship between glucose CV, time in range and C-peptide AUC at 12 months (CL and control participants combined). The black line represents the least squares linear regression line. Baseline (n = 48 CL, n = 42 control), 12 months (n = 44 CL, n = 32 control).

Per-protocol analysis

The primary end point was similar in a per-protocol analysis using data from randomised participants in the CL group with at least 60% CL use and those in the control group who did not start insulin pump therapy (Table 12). The improved glycaemic control in the CL group compared to the control group was slightly more pronounced in the per-protocol analysis of key secondary end points.

Adverse events

Safety-related events are summarised in Table 13. Three severe hypoglycaemia events occurred in two participants randomised to the CL group and one severe hypoglycaemia event occurred in one participant randomised to control group. There was one DKA in the CL group and none in control group. Details of the events are in Table 14. Two non-treatment related SAEs occurred in the CL group and four in control group. A total of 71 other AEs (34 in the CL group, 37 in control group) were reported (Table 15).

Questionnaires

Surveys and tests used in this trial are listed in Table 18. Participants/guardians completed the questionnaires at time points as indicated in the table. Additionally, feedback questionnaires on CL specific experience were distributed to participants/guardians randomised to the CL intervention arm.

Questionnaire results

Responses to the HFS (Table 19), PedsQL (Table 20), PAID (Table 21) and SDQs (Table 22) were similar in both children and parents between treatment groups at 12 months. Scores for the INSPIRE questionnaire were high in children, teenagers and parents suggesting positive expectancies regarding automated insulin delivery in this population (Table 23).

Closed Loop from Onset in Type 1 Diabetes: qualitative substudy

Publications

The primary manuscript has been accepted for publication in the New England Journal of Medicine (July 2022). The protocol paper was published in BMJ Open. The qualitative evaluation has led to five publications in high impact peer reviewed journals. Data from CLOuD study participants have also been included in several secondary analyses with several more underway.

• Boughton CK, Allen JM, Ware J, Wilinska ME, Hartnell S, Thankamony A, et al. ; on behalf of the CLOuD Consortium. Closed-loop and preservation of C-peptide secretion in type 1 diabetes. Accepted in NEJM.

• Boughton C, Allen JM, Tauschmann M, Hartnell S, Wilinska ME, Musolino G, et al. ; CLOuD Consortium. Assessing the effect of closed-loop insulin delivery from onset of type 1 diabetes in youth on residual beta-cell function compared to standard insulin therapy (CLOuD study): a randomised parallel study protocol. BMJ Open 2020;10(3):e033500.

• Rankin D, Kimbell B, Hovorka R, Lawton J; on behalf of the CLOuD Consortium. Adolescents’ and their parents’ experiences of using a closed-loop system to manage type 1 diabetes in everyday life: qualitative study. Chronic Illn 2021:1742395320985924.

• Kimbell B, Rankin D, Ashcroft NL, Varghese L, Allen JM, Boughton CK, et al.; on behalf of CLOuD Consortium. What training, support and resourcing do health professionals need to support people using a closed-loop system? A qualitative interview study with health professionals involved in the Closed Loop from Onset in Type 1 Diabetes (CLOuD) trial. Diabetes Technol Ther 2020;22(6):468–75.

• Lawton J, Kimbell B, Rankin D, Ashcroft NL, Varghese L, Allen JM, et al.; on behalf of the CLOuD Consortium. Health professionals’ views about who would benefit from using a closed-loop system: qualitative study. Diabet Med 2020;37(6):1030–7.

• Rankin D, Kimbell B, Allen JM, Besser REJ, Boughton CK, Campbell F, et al. ; on behalf of the CLOuD Consortium. Adolescents’ experiences of using a smartphone application hosting a closed-loop algorithm to manage type 1 diabetes in everyday life: qualitative study. J Diabetes Sci Technol 2021;15(5):1042–51.

• Lawton J, Hart RI, Kimbell B, Allen JM, Besser REJ, Boughton CK, et al.; on behalf of the CLOuD Consortium. Data sharing while using a closed-loop system: qualitative study of adolescents’ and parents’ experiences and views. Diabetes Technol Ther 2021;23(7):500–7.

• Chen NS, Boughton CK, Hartnell S, Fuchs J, Allen JM, Willinska ME, et al. ; AiDAPT, AP@home04, CLOuD, DAN05, DAN06, and KidsAP consortia. User engagement with the CamAPS FX hybrid closed-loop app according to age and user characteristics. Diabet Care 2021;44(7):e148–50.

Presentations

The primary results were presented as an oral presentation at the American Diabetes Association in New Orleans in June 2022. Results from the qualitative study were presented at the Diabetes Professional Care Conference in London in November 2019. The Edinburgh team have given a series of presentations at the CLOuD Annual Meetings. We extended the invitation to the CLOuD Annual Meeting in 2020 and 2021 to include a wider audience of relevant stakeholders and other interested parties:

• Boughton CK, Allen JM, Ware J, Wilinska ME, Hartnell S, Thankamony A, et al. ; on behalf of the CLOuD Consortium. Effect of 24 Months of Optimised Glucose Control on Residual C-peptide Secretion in Youth with New-onset Type 1 Diabetes. American Diabetes Association, New Orleans, June 2022.

• Kimbell B. Closed-loop Technology: Who Should Be Given Priority and What Training, Resourcing and Support Will Healthcare Professionals Need? Diabetes Professional Care Conference, London, November 2019.

• Rankin D, Kimbell B, Lawton J. Adolescents’ Experiences of Using the CamAPS FX Closed-loop. Closed-loop from Onset of type 1 Diabetes – Annual Meeting, Cambridge, November 2020.

• Kimbell B, Rankin D. Qualitative Interviews – Staff Study. Closed-loop from Onset of type 1 Diabetes – Annual Meeting, Cambridge, November 2019.

• Rankin D, Lawton J. Initial Findings from Qualitative Interviews with CLOuD Participants and Their Parents. Closed-loop from Onset of type 1 Diabetes – Annual Meeting, Cambridge, November 2018.

Webinars

Eleven webinars were delivered on the hybrid CL system for healthcare professionals, people living with diabetes and teachers/education support staff. These are hosted by the Cambridge Diabetes Education Platform (CDEP) and are publicly available on YouTube and have been watched over 10,000 times. CLOuD participants have featured in these webinars sharing their experience of the impact CL technology has had on their lives including managing sporting activities and mental health. David Rankin (Edinburgh) also presented findings from the CLOuD participants experience in a webinar focussing on CL insulin therapy and quality of life.

Social media

• The CLOuD website (http://cloud.mrl.ims.cam.ac.uk/) has been kept updated with recruitment data.

• A CLOuD participant from Nottingham took to Twitter to share their experience of the MMTT.

• A CLOuD participant from Oxford features in an NIHR case study https://local.nihr.ac.uk/case-studies/artificial-pancreas-helps-jack-manage-his-diabetes/23305.

Newsletters

Two newsletters have been sent out to participants to date to share updates on the trial progress, recent trial related publications and news from other trial participants.

Contributions of authors

Charlotte K Boughton (https://orcid.org/0000-0003-3272-9544) (Clinical Lecturer) co-designed the study, recruited and provided support for trial participants, carried out or supported data analysis, including the statistical analyses, prepared the results for publication and wrote the final report.

Janet M Allen (Research Nurse) co-designed the study, screened and enrolled participants, provided patient care and took study samples.

Julia Ware (https://orcid.org/0000-0002-4497-0979) (Clinical Research Fellow) screened and enrolled participants, provided patient care and took study samples.

Malgorzata E Wilinska (https://orcid/0000-0002-1564-8083) (Clinical Research Associate) screened and enrolled participants, provided patient care, took study samples, carried out or supported data analysis, including the statistical analyses.

Sara Hartnell (Diabetes Educator) screened and enrolled participants, provided patient care and took study samples.

Ajay Thankamony (https://orcid.org/0000-0001-6290-3553) (Principal Investigator, Cambridge) screened and enrolled participants, provided patient care and took study samples.

Tabitha Randell (https://orcid.org/0000-0002-1703-1589) (Principal Investigator, Nottingham) co-designed the study, screened and enrolled participants, provided patient care and took study samples.

Atrayee Ghatak (Principal Investigator, Liverpool) co-designed the study, screened and enrolled participants, provided patient care and took study samples.

Rachel EJ Besser (https://orcid.org/0000-0002-4645-6324) (Principal Investigator, Oxford) co-designed the study, screened and enrolled participants, provided patient care and took study samples.

Daniela Elleri (Principal Investigator, Edinburgh) co-designed the study, screened and enrolled participants, provided patient care and took study samples.

Nicola Trevelyan (https://orcid.org/0000-0002-5991-2691) (Principal Investigator, Southampton) co-designed the study, screened and enrolled participants, provided patient care and took study samples.

Fiona M Campbell (https://orcid.org/0000-0002-7618-6759) (Principal Investigator, Leeds) co-designed the study, screened and enrolled participants, provided patient care and took study samples.

David Rankin (https://orcid.org/0000-0002-5835-3402) (Research Fellow) undertook and supported analysis of the qualitative interviews.

Barbara Kimbell (https://orcid.org/0000-0003-4510-9862) (Research Fellow) undertook and supported analysis of the qualitative interviews.

Julia Lawton (https://orcid.org/0000-0002-8016-7374) (Professor of Health and Social Science) undertook and supported analysis of the qualitative interviews.

Judy Sibayan (https://orcid.org/0009-0005-9298-133X) (Study coordinator) supported study set-up and randomisation.

Peter Calhoun (https://orcid.org/0000-0002-5325-7200) (Statistician) carried out or supported data analysis, including the statistical analyses.

Ryan Bailey (Statistician) carried out or supported data analysis, including the statistical analyses.

Gareth Dunseath (https://orcid.org/0000-0001-6022-862X) (Senior Research Officer) undertook sample analysis.

Roman Hovorka (https://orcid.org/0000-0003-2901-461X) (Chief Investigator) co-designed the study, designed and implemented the glucose controller, carried out or supported data analysis, including the statistical analyses and wrote the report.

All authors critically reviewed the report and contributed to the interpretation of the results. Charlotte K Boughton, Peter Calhoun, Malgorzata E Wilinska and Roman Hovorka are the guarantors of this work and, as such, had full access to the data and take responsibility for the integrity of the data and the accuracy of the data analysis. All authors reviewed the paper prior to publication.

We are grateful to study volunteers for their participation. We acknowledge support by the staff at the Addenbrooke’s Wellcome Trust Clinical Research Facility. We thank the members of the TSC (see Appendix 4) and Data Safety Monitoring Board (see Appendix 5) for their oversight of the trial. The study was co-coordinated by Cambridge Clinical Trial Unit.

Disclosure of interests

Full disclosure of interests: Completed ICMJE forms for all authors, including all related interests, are available in the toolkit on the NIHR Journals Library report publication page at https://doi.org/10.3310/KTFR5698.

Primary conflicts of interest: Charlotte K Boughton has received consulting fees from CamDiab and speaker honoraria from Ypsomed. Julia Ware reports receiving speaker honoraria from Ypsomed. Malgorzata E Wilinska reports patents related to closed-loop and being a consultant at CamDiab. Sara Hartnell serves as a member of Medtronic advisory board, is a director of Ask Diabetes Ltd providing training and research support in healthcare settings, and reports having received training honoraria from Medtronic and Sanofi and consulting fees for CamDiab. Tabitha Randell receives consultancy fees from Abbott Diabetes Care and has received honoraria from NovoNordisk for delivering educational meetings. Rachel EJ Besser reports receiving speaker honoraria from Eli Lilly and Springer Healthcare, and reports sitting on the NovoNordisk UK Foundation Research Selection Committee on a voluntary basis. Julia Lawton was a member of HTA General Committee 2018–9. Roman Hovorka reports receiving speaker honoraria from Eli Lilly, Dexcom and Novo Nordisk, receiving licence and/or consultancy fees from B Braun and Abbott Diabetes Care; patents related to closed-loop, and being director at CamDiab. Janet M Allen, Ajay Thankamony, Atrayee Ghatak, Daniela Elleri, Nicola Trevelyan, Fiona M Campbell, David Rankin, Barbara Kimbell, Judy Sibayan, Peter Calhoun, Ryan Bailey and Gareth Dunseath declare no competing financial interests exist.

Patient data statement

This work uses data provided by patients and collected by the NHS as part of their care and support. Using patient data is vital to improve health and care for everyone. There is huge potential to make better use of information from people’s patient records, to understand more about disease, develop new treatments, monitor safety and plan NHS services. Patient data should be kept safe and secure, to protect everyone’s privacy, and it is important that there are safeguards to make sure that they are stored and used responsibly. Everyone should be able to find out about how patient data are used. #datasaveslives You can find out more about the background to this citation here: https://understandingpatientdata.org.uk/data-citation

Data-sharing statement

De‐identified data sets will be made available on a case-by-case basis on reasonable request for research purposes. All requests should be addressed to the corresponding author.

Ethics statement

Approval was received from Cambridge East Research Ethics Committee on the 5 August 2016 (16/EE/0286) and Medicines and Healthcare products Regulatory Agency.

Information governance statement

The University of Cambridge is committed to handling all personal information in line with the UK Data Protection Act (2018) and the General Data Protection Regulation (EU GDPR) 2016/679.

Under the Data Protection legislation, the University of Cambridge is the Data Controller, and you can find out more about how we handle personal data, including how to exercise your individual rights and the contact details for our Data Protection Officer here: https://www.research-integrity.admin.cam.ac.uk/research-governance.

Department of Health and Social Care disclaimer

This publication presents independent research commissioned by the National Institute for Health and Care Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, MRC, NIHR Coordinating Centre, the EME programme or the Department of Health and Social Care.

This monograph was published based on current knowledge at the time and date of publication. NIHR is committed to being inclusive and will continually monitor best practice and guidance in relation to terminology and language to ensure that we remain relevant to our stakeholders.

Publications

Boughton C, Allen JM, Tauschmann M, Hartnell S, Wilinska ME, Musolino G, et al. Assessing the effect of closed-loop insulin delivery from onset of type 1 diabetes in youth on residual beta-cell function compared to standard insulin therapy (CLOuD study): a randomised parallel study protocol. BMJ Open 2020;10(3):e033500.

Kimbell B, Rankin D, Ashcroft NL, Varghese L, Allen JM, Boughton CK, et al.; on behalf of CLOuD Consortium. What training, support and resourcing do health professionals need to support people using a closed-loop system? A qualitative interview study with health professionals involved in the Closed Loop from Onset in Type 1 Diabetes (CLOuD) trial. Diabetes Technol Ther 2020;22(6):468–75.

Lawton J, Kimbell B, Rankin D, Ashcroft NL, Varghese L, Allen JM, et al.; on behalf of the CLOuD Consortium. Health professionals’ views about who would benefit from using a closed-loop system: qualitative study. Diabet Med 2020;37(6):1030–7.

Chen NS, Boughton CK, Hartnell S, Fuchs J, Allen JM, Willinska ME, et al. User engagement with the CamAPS FX hybrid closed-loop app according to age and user characteristics. Diabetes Care 2021;44(7):e148–50.

Lawton J, Hart RI, Kimbell B, Allen JM, Besser REJ, Boughton CK, et al. ; on behalf of the CLOuD Consortium. Data sharing while using a closed-loop system: qualitative study of adolescents’ and parents’ experiences and views. Diabetes Technol Ther 2021;23(7):500–7.

Rankin D, Kimbell B, Allen JM, Besser REJ, Boughton CK, Campbell F, et al. Adolescents’ experiences of using a smartphone application hosting a closed-loop algorithm to manage type 1 diabetes in everyday life: qualitative study. J Diabetes Sci Technol 2021;15(5):1042–51.

Rankin D, Kimbell B, Hovorka R, Lawton J; on behalf of the CLOuD Consortium. Adolescents’ and their parents’ experiences of using a closed-loop system to manage type 1 diabetes in everyday life: qualitative study. Chronic Illn 2021:1742395320985924.

Boughton CK, Allen JM, Ware J, Wilinska ME, Hartnell S, Thankamony A, et al. Closed-loop and preservation of C-peptide secretion in type 1 diabetes. Accepted in NEJM.

Disclaimers

This article presents independent research. The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, the MRC, the EME programme or the Department of Health and Social Care. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, the EME programme or the Department of Health and Social Care.