CARIM School for Cardiovascular Diseases, Maastricht University, Maastricht, The Netherlands.
Department of Endocrinology and Nephrology, Nordsjællands Hospital, Hillerød, Denmark.
Diabetes Obes Metab. 2024 May;26(5):1950-1961. doi: 10.1111/dom.15521. Epub 2024 Mar 19.
To conduct a systematic review with meta-analysis to provide a comprehensive synthesis of randomized controlled trials (RCTs) and prospective cohort studies investigating the effects of currently available bolus advisors on glycaemic parameters in adults with diabetes.
An electronic search of PubMed, Embase, CINAHL, Cochrane Library and ClinicalTrials.gov was conducted in December 2022. The risk of bias was assessed using the revised Cochrane Risk of Bias tool. (Standardized) mean difference (MD) was selected to determine the difference in continuous outcomes between the groups. A random-effects model meta-analysis and meta-regression were performed. This systematic review was registered on PROSPERO (CRD42022374588).
A total of 18 RCTs involving 1645 adults (50% females) with a median glycated haemoglobin (HbA1c) concentration of 8.45% (7.95%-9.30%) were included. The majority of participants had type 1 diabetes (N = 1510, 92%) and were on multiple daily injections (N = 1173, 71%). Twelve of the 18 trials had low risk of bias. The meta-analysis of 10 studies with available data on HbA1c showed that the use of a bolus advisor modestly reduced HbA1c compared to standard treatment (MD -011%, 95% confidence interval -0.22 to -0.01; I = 0%). This effect was accompanied by small improvements in low blood glucose index and treatment satisfaction, but not with reductions in hypoglycaemic events or changes in other secondary outcomes.
Use of a bolus advisor is associated with slightly better glucose control and treatment satisfaction in people with diabetes on intensive insulin treatment. Future studies should investigate whether personalizing bolus advisors using artificial intelligence technology can enhance these effects.
系统评价和荟萃分析，以综合评估目前可用的推注剂量设定器对糖尿病成人血糖参数的影响，纳入随机对照试验（RCT）和前瞻性队列研究。
2022 年 12 月，对 PubMed、Embase、CINAHL、Cochrane 图书馆和 ClinicalTrials.gov 进行电子检索。使用修订后的 Cochrane 偏倚风险工具评估偏倚风险。（标准化）均数差（MD）用于确定组间连续结局的差异。采用随机效应模型进行荟萃分析和荟萃回归。本系统评价在 PROSPERO（CRD42022374588）上进行了注册。
纳入了 18 项 RCT，共纳入 1645 名成年患者（50%为女性），糖化血红蛋白（HbA1c）中位数为 8.45%（7.95%-9.30%）。大多数参与者患有 1 型糖尿病（N=1510，92%），并接受多次每日胰岛素注射（N=1173，71%）。18 项试验中有 12 项具有低偏倚风险。对 10 项有 HbA1c 可用数据的研究进行荟萃分析显示，与标准治疗相比，使用推注剂量设定器可适度降低 HbA1c（MD-0.11%，95%置信区间-0.22 至-0.01；I=0%）。这种效果伴随着低血糖指数和治疗满意度的微小改善，但与低血糖事件的减少或其他次要结局的变化无关。
在接受强化胰岛素治疗的糖尿病患者中，使用推注剂量设定器可略微改善血糖控制和治疗满意度。未来的研究应探讨使用人工智能技术个性化推注剂量设定器是否可以增强这些效果。
