Efficacy of open-source automated insulin delivery systems compared with commercial hybrid closed-loop insulin pumps: real-world data from a tertiary center

Result code in IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00023001%3A_____%2F23%3A00084381" target="_blank" >RIV/00023001:_____/23:00084381 - isvavai.cz</a>

Result on the web

—

DOI - Digital Object Identifier

—

Result language

angličtina

Original language name

Efficacy of open-source automated insulin delivery systems compared with commercial hybrid closed-loop insulin pumps: real-world data from a tertiary center

Original language description

Background and aims: Open-source automated insulin delivery (osAID)platforms emerged in response to prolonged development and certifcationof commercial closed loop systems and, to a knowledgeable and dedicateduser, they ofer excellent customizability enabling to more precisely tailorinsulin delivery to individual needs and preferences. However, as thesesystems did not undergo formal testing in larger randomized controlledtrials, data on their performance are scarce. The aim of our study was tocompare the efcacy of open-source AID systems with the current mainstay of commercial hybrid closed loop (HCL) systems.Materials and methods: Seventeen subjects with type 1 diabetes mellitus (T1DM) using an open-source AID system and 3 equally sizedgroups on commercial HCL pumps (Medtronic 780G with SmartGuard, Tandem t:slim with ControlIQ and YpsoPump with CamAPSFX) matched according to age and diabetes duration were selected fromthe outpatient department of a tertiary diabetes care center. Evaluatedparameters included HbA1C, average sensor glucose and its standarddeviation (SD), time spent in (TIR), above (TAR) and below (TBR)the designated target range of 3.9-10 mmol/l and daily insulin dose.Results: Compared with all 3 commercial HCL systems subjects onopen-source AIDs had signifcantly lower HbA1c (46 vs. 55 vs. 53vs. 60 mmol/mol for osAID vs. 780G vs. ControlIQ vs. CamAPS,p=0.003), average sensor glucose (7.0 ± 0.7 vs. 8.4 ± 1.8 vs. 8.6 ±1.2 vs. 8.3 ± 1.2 mmol/l, p&lt;0.001) and SD (2.1 ± 0.6 vs. 2.7 ± 1.3vs. 2.8 ± 1.5 vs. 3.0 ± 0.8, p=0.004). Also, TIR was increased (84vs. 78 vs. 77 vs. 72% for osAID vs. 780G vs. ControlIQ vs. CamAPS,p=0.004) and TAR decreased (12 vs. 20 vs. 22 vs. 20%, p=0.001) inosAID relative to HCL groups, while TBR was the lowest in 780Gand ControlIQ subjects (3 vs. 1 vs. 1 vs. 4% for osAID vs. 780G vs.ControlIQ vs. CamAPS, p=0.011). In contrast, daily doses of total,basal and bolus insulin did not difer between the groups (46.1 ± 26.7vs. 48.6 ± 20.6 vs. 49.0 ± 15.7 vs. 41.2 ± 15.7 IU/day for osAID vs.780G vs. ControlIQ vs. CamAPS, p=0.656).Conclusion: In our study group of subjects from a tertiary diabetescare center open-source automated insulin delivery systems showedsuperior performance to commercial HCL systems in almost allparameters of glucose control. Apart from technical differencesin the applied algorithms, these results might be at least partially attributed to the higher level of engagement and knowledgerequired to successfully use an open-source AID system.

Czech name

—

Czech description

—

Type

O - Miscellaneous

CEP classification

—

OECD FORD branch

30202 - Endocrinology and metabolism (including diabetes, hormones)

Project

<a href="/en/project/LX22NPO5104" target="_blank" >LX22NPO5104: National Institute for Research of Metabolic and Cardiovascular Diseases</a><br>

Continuities

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Publication year

2023

Confidentiality

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů