Description of the Population

Of 1023 users, 896 met eligibility criteria (Fig. 1). Population characteristics are shown in Table 1. Mean age (± standard deviation, SD) was 23.1 ± 20 years and 472 users (52.7%) were female. Most (95.1%, n = 852) had T1D and 4.9% (n = 44, all adults) had T2D. HbA1c levels (7.43 ± 0.99%) were slightly above recommended targets (< 7% [10, 11]). HbA1c levels of HCL users were 0.3% to 0.6% lower than those of other IDS users.

Fig. 1Table 1 Sociodemographic and clinical description of diabetes in the analysed population according to the device currently used

Participants were evenly distributed across age subgroups (34.0% children, 30.4% adolescents, and 35.6% adults) and IDS subgroups (23.2% HCLs, 30.4% tubeless pumps, 24.0% tubed pumps, and 22.4% MDI.

Tubeless pump and HCL users were predominantly female (58.5% and 57.2% female, respectively), and MDI users were predominantly male (43.8% female).

Choosing a Specific Type and Model of IDS

The main reason for choosing tubed pumps and HCLs was primarily based on physician recommendations and medical criteria (“diabetologist’s recommendation”, “better glycaemic control”), whereas tubeless pumps were most frequently chosen for the experience they offer (“belief that the system would be more practical”) (Fig. 2). “Better glycaemic control” was selected significantly more often by HCL users (66.3%) compared with tubed pump users (37.2%, p < 0.001) or tubeless pump users (43%, p < 0.001). “Belief that the new system would be more practical” was selected significantly more often for tubeless pump users (53.7%) and HCL users (42.8%) than for tubed pump users (34%, p < 0.001 and p = 0.005, respectively). “Making exercise easier” was also chosen more often for tubeless pump users (21.3%) and HCLs users (16.8%) than for tubed pump users (7%) (both p < 0.001).

Fig. 2

Reasons for initiating the current insulin delivery systems (IDS) by IDS group. This figure illustrates responses to the question addressed to users with a pump or a hybrid closed loop: “You use a pump or HCL because…”. The bars show the percentage of respondents for each item. Patients could give several reasons. The p value of the independent Chi2 test is indicated for each item. Post hoc Chi2 tests were performed and are presented in Supplementary Material. The p value and level of significance are displayed as follows: ns, p ≥ 0.05; *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001; ****p ≤ 0.0001. HCL hybrid closed loop system

Healthcare providers were frequently involved in the decision for a specific model of IDS (Fig. 3). Physicians played a pivotal role, especially for tubed pumps, where just over half (51.6%) of users reported that their physician chose the model. Conversely, personal research played a greater role in the choice of model for tubeless pumps and HCLs compared with tubed pumps (25.5% for HCLs and 26.5% for tubeless pumps versus 11.6% for tubed pumps; p < 0.001 for both comparisons).

Fig. 3

Reasons for choosing a specific model of pump or hybrid closed loop (HCL) by insulin delivery systems (IDS) group. This figure illustrates responses to the question addressed to users with a pump or a hybrid closed loop: “Your pump model was chosen because…”. Patients could give several reasons. The bars show the percentage of respondents for each item. The p value of the independent Fisher test is indicated for each item. Post hoc Chi2 tests were performed and are presented in Supplementary Material. The p value and level of significance are displayed as follows: ns, p ≥ 0.05; *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001; ****p ≤ 0.0001. No statistical test was carried out for the item “Suggested by a pharmacist” because of the very low number of respondents who selected this item

Physicians played a more important role in the choice of IDS type and model for children than for other age groups (Fig. S1).

Satisfaction with the DeviceOverall Satisfaction Score

Overall satisfaction (VAS) was 7.3 ± 1.7 for MDI, 7.9 ± 1.2 for tubeless pumps, 8.0 ± 1.5 for tubed pumps, and 8.4 ± 1.4 for HCLs (Fig. 4a). Satisfaction was lowest for MDI (p < 0.001, all comparisons) and highest for HCLs (p < 0.001 versus tubeless pumps and p < 0.05 versus tubed pumps). The difference between tubed and tubeless pumps was not significant (p ≥ 0.05).

Fig. 4

Overall satisfaction, Insulin Delivery System Rating Questionnaire (IDSRQ) satisfaction, and IDSRQ interference scores presented by insulin delivery systems (IDS) group. a Overall satisfaction with the IDS, measured on a visual analog scale from 1 (not at all satisfied) to 10 (totally satisfied). The question was “Overall, how satisfied are you with your current device?”. b IDSRQ Satisfaction Score. A value of 0 indicates low satisfaction and a value of 100 indicates high satisfaction with the IDS. c IDSRQ Interference Score. A value of 0 indicates no interference and a score of 100 indicates a high level of interference of the IDS. For each score, we conducted an overall assessment of statistical differences and, where appropriate, performed pairwise comparisons between all IDS. The tests used are presented on the figure. The p value and level of significance are displayed as follows: ns, p ≥ 0.05; *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001; ****p ≤ 0.0001

IDSRQ Satisfaction Scores

IDSRQ satisfaction scores were 53.1 ± 23.0 for MDI, 64.1 ± 21.6 for tubed pumps, 66.6 ± 19.1 for tubeless pumps, and 70.9 ± 17.7 for HCLs (Fig. 4b). Satisfaction with MDI remained significantly lower compared with all other IDS (p < 0.001). In fact, this group exhibited the lowest satisfaction levels on each of the 15 IDSRQ items (Fig. 5).

Fig. 5

Items of the Insulin Delivery System Rating Questionnaire (IDSRQ) satisfaction subscale. a Heat map presenting the percentage of respondents who were “Completely satisfied” or “Very satisfied” (score ≥ 66.66) for each item of the IDSRQ satisfaction subscale. The graduated coloured scale ranges from the minimum (purple) to the maximum (green) percentage of satisfied respondents. Users answered the following question: “How satisfied are you with your current device in the following areas?”. b Main differences (at least 20% difference) between multiple daily injections (MDI) and other insulin delivery systems (IDS). c Main differences (at least 10% difference) between tubed pumps (orange), tubeless pumps (yellow), and hybrid closed loop systems (HCLs) (blue). The p value of the independent Chi2 test is indicated for each item. Post hoc Chi2 tests were performed and presented in Supplementary Material (File S1). The p value and level of significance are displayed as follows: ns, p ≥ 0.05; *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001; ****p ≤ 0.0001

Regression analysis confirmed that the probability of being satisfied with the device was significantly higher for HCLs (OR 6.18; 95% CI 3.81, 10.01), tubeless pumps (OR 3.88; 95% CI 2.54, 5.92), and tubed pumps (OR 2.58; 95% CI 1.66, 4.03) compared with MDI. Other factors independently associated with higher satisfaction included age (adults vs children—OR 3.53; 95% CI 2.48, 5.04) and sex (male vs female—OR 1.36; 95% CI 1.02, 1.82) (Fig. S2A, Table S2A).

Key Items Contributing to (Dis)satisfaction

Overall, most users were satisfied with ease of learning to use the IDS (79%), ease of taking insulin (78%), difficulty of taking all the insulin prescribed (75%), and complexity of use (74%) (Fig. 5). Conversely, the fewest users were satisfied with the amount of pain (51%), skin irritation or bruises (47%), and difficulty of carrying all the supplies (45%). Of note, women (78.2%) were more satisfied than men (71.8%) with the difficulty of taking all the insulin prescribed (p < 0.05) (Fig. S3A).

For HCL systems, 13 out of 15 items reached at least a 65% satisfaction rate (Fig. 5a). For tubeless pumps, 11 out of 15 items reached at least a 65% satisfaction rate, versus 7 for tubed pumps and 2 for MDI. The strongest differences between MDI and other IDS were in terms of “how easy it is to take insulin”, “difficulty in dosing accurately”, and the “occurrence of skin infections” (Fig. 5b).

Focusing on pumps and HCLs, large differences (≥ 10%) were observed for pain, convenience, and ease of learning to use the device. More users were satisfied with the amount of pain when using HCLs (71.6%) compared with tubed pumps (55.8%, p = 0.002) and tubeless pumps (45.6%, p < 0.001) (Fig. 5c). Dissatisfaction with the amount of pain was also reported more frequently by children (64.3%) and adolescents (56.2%) than by adults (27.3%) (p < 0.001) (Fig. S4A).

Impacts of Insulin Therapy on Daily LifeIDSRQ Interference Scores

IDSRQ interference scores were 42.2 ± 24.3 for MDI, 38.6 ± 26.0 for tubed pumps, 37.0 ± 25.5 for HCL pumps, and 31.1 ± 24.8 for tubeless pumps (Fig. 4c). Tubeless pumps interfered significantly less than tubed pumps (p < 0.01) and MDI (p < 0.0001); however, there was no significant difference between tubeless pumps and HCLs (p = 0.07) (Fig. 4c). There was also no difference observed between HCLs, tubed pumps, and MDI (all p ≥ 0.05).

Regression analysis showed that the probability of an IDS interfering less with users’ daily life was significantly higher for tubeless pumps (OR 2.78; 95% CI 1.88, 4.11), HCLs (OR 2.30; 95% CI 1.49, 3.57), or tubed pumps (OR 1.63; 95% CI 1.08, 2.46) compared with MDI (Fig. S2B, Table S2B). Male users were also more likely to have a lower interference score than female users (OR 1.40; 95% CI 1.06, 1.84) (no effect observed from time using device or age).

Key Items of Interference

Overall, IDS interfered most frequently with “taking care of yourself when travelling” (53%), “wearing the clothes you want” (46%), and “doing the type of exercise you want” (40%) (Fig. 6a).

Fig. 6

Items of the Insulin Delivery System Rating Questionnaire (IDSRQ) interference subscale. a Heat map presenting the percentage of respondents who reported “Some interference” or “A lot of interference” (score > 33.33) for each item of the IDSRQ interference subscale. The graduated coloured scale ranges from the minimum (purple) to the maximum (green) percentage of respondents who reported at least some interference. Users answered the following question: “To what extent does your current device interfere with the following activities?”. b Main differences (at least 10% difference) between multiple daily injections (MDI) and other insulin delivery systems (IDS). c Main differences (at least 10% difference) between tubed pumps (orange), tubeless pumps, (yellow), and hybrid closed loop systems (HCLs) (blue). The p value of the independent Chi2 test is indicated for each item. Post hoc Chi2 tests were performed and presented in Supplementary Material (File S1). The p value and level of significance are displayed as follows: ns, p ≥ 0.05; *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001; ****p ≤ 0.0001

MDI interfered less than other IDS with being able to “wear the clothes you want” (24.4% vs 52.1%, p < 0.001). Conversely, MDI interfered much more frequently with items related to eating, be it “eating what you want” or “when you want”, compared to other IDS (p < 0.001) (Fig. 6b).

Tubeless pumps exhibited the least interference, with only 3 out of 10 items having more than 35% of users reporting interference. The next best profile was HCLs with 6 out of 10 items followed by tubed pumps and MDI (both 8 out of 10) (Fig. 6a).

Focusing on pumps and HCLs, tubeless pumps interfered less often with “doing the type of exercise you want” (30.9% vs HCLs 45.2%, p = 0.005; and tubed pumps 42.8%, p = 0.016) and “wearing the clothes you want” (44.5% vs HCLs 57.2%, p = 0.023; and tubed pumps 56.7%, p = 0.023) (Fig. 6c). Of note, women reported interference regarding clothing choice significantly more often than men (53.6% vs 37.4%, p < 0.001) (Fig. S3A). Tubeless pumps were the only advanced technology where the gender gap was not of statistical significance (p = 0.054) (Fig. S5).

Other Aspects of Daily Life

Significantly more tubeless pumps users reported their IDS did not interfere with “using the restroom” (81.6%) than MDI (70.6%), tubed pump (60.0%) or HCL users (61.1%) (all p < 0.001) (Fig. S6).

Two items were specific to children: “eating at the school cafeteria” and “leaving your child with someone else”. For the former, 51.4% and 51.3% of tubeless pumps and HCL users, respectively, reported no interference, compared to 34.4% of tubed pump users and 32.2% of MDI users (significant overall effect of the device; p < 0.001). Similarly, for the latter, a higher proportion of tubeless pump (55.2%) or HCL users (55.5%) reported little or no interference, compared to tubed pumps (34.3%) or MDI users (41.3%) (significant overall effect of the device; p = 0.002) (Fig. S6).

Among other bespoke items, “Embarrassment because the disease is visible due to the treatment” was rated significantly more problematic for children (4.2 ± 2.7) and adolescents (4.3 ± 3.0) than adults (3.3 ± 2.8) (p < 0.0001). Similarly, female users exhibit greater embarrassment from the visibility of their illness due to therapy (4.1 ± 2.9) compared with men (3.6 ± 2.73) (p = 0.014) (Fig. S7).

Switching to a Hypothetical HCL System

Very few people said they would refuse a HCL if it were offered to them, except MDI users, for whom 21.4% (n = 43) would refuse (Fig. 7a). MDI users who would refuse HCL were older (24.7 ± 15.8 years) and had a higher IDSRQ satisfaction score (63.3 ± 23.8), compared with those who would accept this system (age 17.8 ± 15.9 years, p < 0.001; IDSRQ satisfaction score 49.5 ± 20.1, p < 0.001).

Fig. 7

Type of pump respondents would choose as part of their ideal hybrid closed loop system (HCL). a Users of tubed pumps (orange), tubeless pumps (yellow), and multiple daily injections (MDI) (grey) answered the following question: “If you could switch to a hybrid closed loop tomorrow, would you choose…”. b HCL users answered the following question: “If you had the opportunity tomorrow to change the type of pump in your hybrid closed loop, you would choose…”. The p value of the independent Chi2 test is indicated for each item. Post hoc Chi2 tests were performed and presented in Supplementary Material (File S1). The p value and level of significance are displayed as follows: ns, p ≥ 0.05; *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001; ****p ≤ 0.0001

When asked to choose the pump for their ideal HCL, of the 208 HCLs users, 20.2% would choose a tubed pump, 34.1% would choose a tubeless pump, and 45.7% did not know (Fig. 7b). Of the 272 tubeless pump users, only 1.5% would choose a tubed pump whereas 84.6% would choose a tubeless pump, and 11.8% did not know (Fig. 7a). Of the 215 tubed pump users, 43.3% would choose a HCL with a tubed pump, 24.7% would choose a HCL with a tubeless pump, and 30.2% did not know.