Day-and-night glycaemic control with closed-loop insulin delivery versus conventional insulin pump therapy in free-living adults with well controlled type 1 diabetes: an open-label, randomised, crossover study

作者：Lia Bally, Hood Thabit, Harald Kojzar, Julia Katharina Mader, Jehona Qerimi-Hyseni, Sara Hartnell, Martin Tauschmann, Janet M. Allen, Malgorzata E. Wilinska, Thomas Rudolf Pieber, Mark L. Evans, Roman Hovorka · 发表于：The Lancet Diabetes & Endocrinology · 年份：2017 · DOI：10.1016/s2213-8587(17)30001-3 · 被引用次数：151 · 研究领域：Diabetes Management and Research、Pancreatic function and diabetes、Hyperglycemia and glycemic control in critically ill and hospitalized patients

Background Tight control of blood glucose concentration in people with type 1 diabetes predisposes to hypoglycaemia. We aimed to investigate whether day-and-night hybrid closed-loop insulin delivery can improve glucose control while alleviating the risk of hypoglycaemia in adults with HbA 1c below 7·5% (58 mmol/mol). Methods In this open-label, randomised, crossover study, we recruited adults (aged ≥18 years) with type 1 diabetes and HbA 1c below 7·5% from Addenbrooke's Hospital (Cambridge, UK) and Medical University of Graz (Graz, Austria). After a 2–4 week run-in period, participants were randomly assigned (1:1), using web-based randomly permuted blocks of four, to receive insulin via the day-and-night hybrid closed-loop system or usual pump therapy for 4 weeks, followed by a 2–4 week washout period and then the other intervention for 4 weeks. Treatment interventions were unsupervised and done under free-living conditions. During the closed-loop period, a model-predictive control algorithm directed insulin delivery, and prandial insulin delivery was calculated with a standard bolus wizard. The primary outcome was the proportion of time when sensor glucose concentration was in target range (3·9–10·0 mmol/L) over the 4 week study period. Analyses were by intention to treat. This study is registered with ClinicalTrials.gov, number NCT02727231, and is completed. Findings Between March 21 and June 24, 2016, we recruited 31 participants, of whom 29 were randomised. One participan...