For example, the Adult study found that for a High Glucose alarm level set to 200 mg/dL:
99.2% of the time a high glucose alarm was received when blood glucose was indeed above the
alarm level but 0.8% of the time a high glucose alarm was received when blood glucose wasn’t
actually above the alarm level.
97.1% of the time blood glucose was above the alarm level and a high glucose alarm was
appropriately issued but 2.9% of the time the glucose event was missed and no alarm was issued.
Table 5c: High Glucose Alarm Performance (Adult; n=144)
High Glucose
Alarm level
(mg/dL)
Alarm Rate
Detection Rate
Number of
Events (n)
True Alarm
Rate (%)
False Alarm
Rate (%)
Number of
Events (n)
Correct
Detection
Rate (%)
Missed
Detection
Rate (%)
120
105544
99.1
0.9
11417
98.2
1.8
140
93574
99.1
0.9
10152
98.1
1.9
180
74290
99.2
0.8
8080
97.8
2.2
200
66039
99.2
0.8
7269
97.1
2.9
220
57549
99.0
1.0
6390
96.9
3.1
240
48733
98.4
1.6
5550
95.6
4.4
300
21512
96.3
3.7
2672
90.0
10.0
204