1 Kinderkrankenhaus auf der Bult, Janusz- Korczak-Allee 12, D-30173 Hannover, Germany, 3 Olgahospital Stuttgart, Kinderklinik, Bismarckstr. 8, D-70176 ...
INFLUENCE OF CGM ON THE VALUE OF THE TECHNICAL POSSIBILITIES IN MODERN INSULIN PUMPS T. Danne1, O. Kordonouri1, K. Remus1, M. Holder2, T. Wadien2, H. Haberland3, S. Golembowski3, S. Zierow3, R. Hartmann4, A. Thomas5 1 Kinderkrankenhaus 3
auf der Bult, Janusz- Korczak-Allee 12, D-30173 Hannover, Germany, 3 Olgahospital Stuttgart, Kinderklinik, Bismarckstr. 8, D-70176 Stuttgart, Germany, Sana-Klinkum Lichtenberg, Sozialpädiatrisches Zentrum / Diabetologie, Gotlindestr. 2-20, D-10365 Berlin, Germany, 4 Buchenallee 17, D-16341 Panketal, Germany 5 Medtronic GmbH, Meerbusch, Germany
Background: Severe hypoglycemia are an important problem and barrier for a nearly normal glycemic control. The sensor-augmented insulin pump therapy (SaP) with the “Paradigm®VEO” system (PRVEO) offers a novel automatic insulin shut-off mechanism (LGS) possibly preventing severe hypoglycemia. To assess the effectiveness of the algorithm so far little experience exists especially with children. Aim: Observation study to investigate how the LGS algorithm reduce the frequency of hypoglycemic episodes. Of primary interest is the frequency of LGS activations under daily life conditions. Method: • Observation study over 8 weeks in 3 pediatric centers in Germany • Patients with experience of CSII, but naive in respect of using LGS • First phase (2 weeks): SaP without LGS (alert switched off) • Second phase (6 weeks): patients using LGS function (after education) • Setting alerts: LGS - 70 mg/dl (3.9 mmol/L), Hypoglycemia - 75 mg/dl (4.2 mmol/L) • Analysis: - Frequency of LGS - Frequency of hypoglycemic excursions (< 70 mg/dl and < 40 mg/dl) - AUC and time in hypoglycemic region (70 mg/dl (3.9 mmol/L)) • Feedback from patients with a special questionnaire for the evaluation of pump and CGM management
Method: Design of observation study
33
40
23
0,1
28 28
23
20
11
7
4
9 3
8
7
8
6
>115-120
>110-115
>105-110
>90-95
Intervals with time of LGS interruption (min)
>100-105
10pm to 06am 06am to 10pm
>95-100
Night time: Day time:
9
>85-90
all
9
0
>80-85
0
10
>75-80
Night time (10pm to 06am)
14 6
>70-75
0.032
20 20
84%
>65-70
Day time: 06am to 10pm
54
>55-60
Night time: 10pm to 06am
60
>45-50
all
16%
80
>35-40
0
Day time (06 am to 10 pm)
0.175
>30-35
Night time (10pm to 06am)
0.206 0,2
>25-30
43%
108 100
>20-25
0,2
120
Frequency of interruption per day time
>15-20
0.38+0.32
0,4
ns (0.30) ns (0.48) ns (0.15) ns (0.50) ns (0.30) 0.05 0.002 0.001 0.01 0.005
• Distribution of LGS interruptions for all 21 patients (≥ 5 min intervals)
5-10
57%
p
Phase 1 Phase 2 (without LGS) (with LGS) 145.29 ± 22.99 148.43 ± 19.44 54.67 ± 10.03 55.62 ± 10.75 25.51 ± 14.38 28.20 ± 13.99 651 ± 240 639 ± 182 11.53 ± 2.40 12.51 ± 1.77 0.76 ± 0.61 0.53 ± 0.37 101 ± 68 58 ± 33 0.053 ± 0.029 0.021 ± 0.008 1.27 ± 0.75 0.95 ± 0.49 0.28 ± 0.18 0.13 ± 0.14
>10-15
0.49+0.43
End of week 8 download of data’s in CareLink measurement: HbA1c Questionnaires patients/parents
Week 3 download of data’s in CareLink
Frequency
Day time (06 am to 10 pm)
0,6
Using PVEO with LGS: LGS alert under hypo alert Setting: hypoglycemia alert: 75 mg/dl LGS alert: 70 mg/dl
Using PVEO without LGS
Week 1 measurement: HbA1c case history
• Frequency of LGS interruptions over 120 minutes (following automatically disruption) Frequency of interruptions over 120 min/patient/day
Alerts with LGS activation per time (night and day):
0,8
LGS alerts / day
Experience with CSII age: 1 - 21 years
0,3 0.89+0.67
Doctor's round 4 End of week 8 end of UE
24 patients with type 1 diabetes
Mean glucose value (mg/dl) Standard deviation (mg/dl AUC > 140 mg/dl (mg/dl x d) time > 140 mg/dl / Tag (min) AUC x t > 140 mg/dl (mg/dl) AUC < 70 mg/dl (mg/dl x d) time < 70 mg/dl / Tag (min) AUC x t< 70 mg/dl (mg/dl) excursions < 70 mg/dl / day excursions ≤ 40 mg/dl / day
Results • Frequency of LGS (all alerts in summery): 1298 - From it: 853 shorter than 5 minutes (reaction by patients immediately, no interruption of insulin delivery) • Frequency per patient and day - including alerts ≤ 5 minutes: 2.56 ± 1.86 - excluding alerts > 5 minutes: 0.89 ± 0.67
1
Doctor's round 3 Week 5 doctor’s round (telefon call)
Doctor's round 2 Week 3 Education LGS and their application (pump training)
Results: • Comparison of glycemic parameters during investigation
Patients: • 24 children with type 1 diabetes, Data analysis was possible in 21 patients • Age: 10.84 ± 3.77 years • diabetes duration: 5.89 ± 2.95 years • experience with CSII: 3.65 ± 1.68 years (no experience with LGS) • basic HbA1c: 7.77 ± 1.13 % • height: 1.47 ± 0.21 m, weight 42.82 ± 16.96 kg • BMI: 18.05 ± 2.30 kg/m2
• Frequency of alerts with LGS activation
Doctor’s round 1 Week 1 education of patients of PVEO (pump training)
• Examples Successful avoidance of severe hypoglycemia's by using of LGS option Stability in respect of patients mistakes (didn’t change sensor after 6 day using) – LGS interruption is a safe option:
Conclusion: LGS prevent heavy hypoglycemia‘s. An alert level of 70 mg/dl is effective for prevention of hypoglycemia's. LGS over 120 minutes is often during sleep. Nobody got a heavy hypoglycemia in this case. In many cases the patients didn’t notice this event. The present study provides evidence for reducing the risk for hypoglycemia with LGS without compromising the safety of SaP therapy.
1: wrong calibration 2: LGS because sensor signal is low (blood sugar not) 3: insulin delivery switch on automatically after 120 min 4: after increasing basal rate (4.15 am) next LGS alert, patient switched off 5: next LGS interruption 6: insulin delivery switch on automatically after 120 min 7: hyperglycemia (blood glucose) but no DKA
