Toffanin Chiara, Zisser Howard, Doyle Francis J, Dassau Eyal
Department of Information and Industrial Engineering, University of Pavia, Pavia, Italy.
J Diabetes Sci Technol. 2013 Jul 1;7(4):928-40. doi: 10.1177/193229681300700415.
Insulin-on-board (IOB) estimation is used in modern insulin therapy with continuous subcutaneous insulin infusion (CSII) as well as different automatic glucose-regulating strategies (i.e., artificial pancreas products) to prevent insulin stacking that may lead to hypoglycemia. However, most of the IOB calculations are static IOB (sIOB): they are based only on approximated insulin decay and do not take into account diurnal changes in insulin sensitivity.
A dynamic IOB (dIOB) that takes into account diurnal insulin sensitivity variation is suggested in this work and used to adjust the sIOB estimations. The dIOB function is used to correct the dosage of insulin boluses in light of this circadian variation.
Basal-bolus as applied by pump users and model predictive control therapy with and without dIOB were evaluated using the University of Virginia/Padova metabolic simulator. Three protocols with four meals of 1 g carbohydrate/kg body weight were evaluated: a nominal scenario and two robustness scenarios, one in which insulin sensitivity was 15% greater than estimated and the other where the lunch is 30% less than announced. In the nominal and robustness scenarios, respectively, the dIOB led to 6% and 24% and 40% less hypoglycemia episodes than approaches without IOB. The new approach was also compared with the sIOB to evaluate the improvements with respect to the previous approach.
Improved glucose regulation was demonstrated using the dIOB where circadian insulin sensitivity is used to adjust IOB estimation. Use of diurnal variations of insulin sensitivity appears to promote effective and safe insulin therapy using CSII or artificial pancreas. Clinical trials are warranted to determine whether nocturnal hypoglycemia can be reduced using the dIOB approach.
在现代胰岛素治疗中，包括持续皮下胰岛素输注（CSII）以及不同的自动血糖调节策略（即人工胰腺产品），都会使用胰岛素存量（IOB）估算来预防可能导致低血糖的胰岛素蓄积。然而，大多数IOB计算都是静态IOB（sIOB）：它们仅基于近似的胰岛素衰减，并未考虑胰岛素敏感性的昼夜变化。
本研究提出了一种考虑昼夜胰岛素敏感性变化的动态IOB（dIOB），并用于调整sIOB估算值。dIOB函数用于根据这种昼夜变化来校正胰岛素大剂量的用量。
使用弗吉尼亚大学/帕多瓦代谢模拟器评估了泵使用者应用的基础-大剂量胰岛素方案以及有无dIOB的模型预测控制疗法。评估了三种含四餐、每餐碳水化合物摄入量为1 g/kg体重的方案：一个标称方案和两个稳健性方案，其中一个方案的胰岛素敏感性比估计值高15%，另一个方案的午餐量比宣布的量少30%。在标称方案和稳健性方案中，与不使用IOB的方法相比，dIOB分别使低血糖发作次数减少了6%、24%和40%。还将新方法与sIOB进行了比较，以评估相对于先前方法的改进情况。
使用dIOB（利用昼夜胰岛素敏感性来调整IOB估算）证明了血糖调节得到改善。利用胰岛素敏感性的昼夜变化似乎可以促进使用CSII或人工胰腺进行有效且安全的胰岛素治疗。有必要进行临床试验以确定使用dIOB方法是否可以减少夜间低血糖。
