Preliminary Study Of Efficacy, Safety And Accuracy Of Sensor-augmented Insulin Pump In Diabetes Mellitus

Objectives: The purpose of this study was to understand the efficacy and safety ofsensor-augmented insulin pump (SAP) treatment and the accuracy of real-time continuousglucose monitoring system(RT-CGMS).Subjects and Methods:1.60patients with type2diabctes were randomly assigned to3groups of treatmentwith SAP, continuous subcutaneous insulin infusion (CSII)，or multiple daily injection(MDI) therapy for6days. Parameters of glycemic control that were determined includedmean blood glucose (MBG),standard deviation of blood glucose(SDBG),mean amplitudeof glycaemic excursion (MAGE),absolute means of daily differences(MODD),area undercurvel10(BG＞10mmol/L,AUC10),area under curvel3.9(BG＜3.9mmol/L,AUC3.9), andlow blood glucose index(LBGI).2.40patients with type1diabetes were randomized to receive one of the flowing twotreatments for6days: SAP or CSII combined with CGMS therapy.In the first treatmentgroup，the insulin does was calculated using the Bolus Wizard，real-time continuous glucosemonitoring was performed with the high or low glucose alarm on，and individualized patienteducation was carried out through the CareLinkTmPersonal software. In the secondtreatment group，the monitoring data were retrieved once a day，and the insulin dosing wasadjusted accordingly.3.79patients with diabctes mellitus wore RT-CGMS3days, matching rate was usedto assess the accuracy of RT-CGM.MAGE, SDBG, MBG, AUC10, AUC3.9, and effectiveglycaemic excursion in frequency(NGE) were used to assess blood glucose fluctuation.Weanalyzed correlation of the matching rate with blood glucose fluctuations parameters bycorrelation analysis and multiple linear regression analysis. Then we grouped correspondingmatch rate by MAGE and MBG, compared the difference among groups. Assuming thematching rate75%as the cutoff point to calculate the ROC curve and Youden index of MAGE and MBG. Doing ANOVA of RAD (%) of fasting,postprandial,and night sensorvalues.Results:1.No significant differences were observed among the three groups in terms of MBG,SDBG, MAGE, or MODD at the beginning of treatment. The MBG, SDBG, MAGE,MODD, and total AUC10of the SAP group improved over the4days of the interventioncompared with the CSII and MDI groups; however, no significant differences wereobserved among the three groups in terms of total AUC3.9and LBGI.2.On the last day （day6） of treatment， mean glucose concentration {MBG，(6.80±0.72) mol/L vs.(7.87±0.88) mmol/L}，standard deviation of blood glucose {SDBG，(1.19±0.33) mmol/L vs.(1.85±0.62) mmol/L}，mean amplitude of glycemic excursions{MAGE，(3.73±0.64) mmol/L vs.(4.71±1.29) mmol/L}，fasting blood glucose {FBG，(6.68±0.85) mmol/L vs.(7.31±1.02) mmol/L}， postprandial blood glucose{PPG，(7.89±1.12)mmol/L vs.(8.44±1.26) mmol/L}， low blood glucose index {LBGI，(0.48±0.35) vs.(1.11±0.84)}，frequency of hypoglycemia {(0.90±0.85) vs.(2.10±1.07)}， and area under thecurve of bllod glucose <3.9mmol/L {AUC3.9，(0.02±0.03) mmol/L h vs.(0.31±0.27)mmol/L h） were all better in patients receiving SAP therapy than patientsreceiving continuous subcutaneous insulin infusion therapy.3. RAD (%) and ISO criteria (%) of all day,fasting,postprandial,and night sensor valueshowed good consistency.No significant differences of RAD (%) of fasting,postprandial,andnight sensor values.Correlation analysis showed that the matching rate was negativelycorrelated with the MAGE（r=-0.464）、SDBG（r=-0.442）、MBG（r=-0.409）、AUC10（r=-0.411）MAGE, SDB. MAGE, MBG were into the regression equation, MAGE<3.9mmol/L, matching rate was81%,3.9≤MAGE <7mmol/L, matching rate was67%,MAGE≥7mmol/L, the matching rate was52%, the difference among3groups wasstatistically significant (P <0.01). MBG <7.8mmol/L, matching rate was78%,7.8≤MBG <10mmol/L, matching rate was68%, MBG≥10mmol/L, the matching rate was57%, the difference among3groups was statistically significant (P <0.01). When Youdenindex was0.409and MAGE was less than3.75mmol/L, the matching rate could be up to75%. When Youden index was0.369and MBG was less than8.38mmol/L, the matchingrate was more than75%. Conclusions:1.Compared with CSII and MDI therapy, SAP therapy was able to rapidly lower meanBG and reduce BG level fluctuations with no increased risks of hypoglycemia in patientswith T2DM.2.Compared with retrospective CGMS, RT-CGMS can timely regulate therapeutic regimen,prospective regulate blood glucose, reduce MBG, Fasting plasma glucose (FPG),Postprandial plasma glucose (PPG). Bolus wizards (BW) and dual-square wave functioncan try to make the blood glucose curve coincides with the curve of insulin action,effectively reduce PPG. Hypoglycemia/Hyperglycemia early warning System (EWS) caneffectively reduce hypoglycemia and severe hyperglycaemia, reduce blood glucosefluctuations. CareLinkTmpersonal software can regulate therapeutic regimen, guide diet,exercise and diabetes education individualized.3.The sensor values agree closely with the interphalangeal glucose values，which couldbe used for blood glucose assessment. Blood glucose fluctuations influences accuracy ofRT-CGM, the larger of the blood glucose fluctuations, the larger of the mean glucoseconcentration, the smaller of the accuracy.