Table 2.

Numerical interrelationships among number of subjects, experiments, data sets and glucose data segments

	Meal study	Comments
Subjects	7	One sensor per subject
Experiments	14	Two experiments per subject (one high glycaemic index, one low glycaemic index)
Data sets	14	One sensor data set per experiment
Rising glucose data segments	13	Of the 14 data sets, there were 11 that had one evaluable rising segment, one that had two evaluable rising segments and two that had no evaluable rising segment
Rising segments after combination of duplicate sensors	13	(No change, single sensor per experiments)
Falling glucose data segments	12	Of the 14 data sets, there were six that had one evaluable falling segment, three that had two evaluable falling segments and five that had no evaluable falling segment
Falling segments after combination of duplicate sensors	12	(No change, single sensor per experiments)

Closed-loop study	Comments
14	Two sensors per subject
23	Five subjects had one experiment, nine subjects had two experiments
46	Two sensor data sets per experiment
34	Of the 46 data sets, there were nine that had one evaluable rising segment, five that had two evaluable rising segments, five that had three evaluable rising segments and 27 that had no evaluable rising segment
23	In 11 sensor pairs, both sensors met all inclusion criteria and thus were combined (averaged), leaving 23
36	Of the 46 data sets, there were seven that had one evaluable falling segment, four that had two evaluable falling segments, three that had three evaluable falling segments, three that had four evaluable falling segments and 29 that had no evaluable falling segment
24	In 12 sensor pairs, both sensors met all inclusion criteria and thus were combined (averaged), leaving 24

Combined data from both studies
Total rising data segments	36	13 (meal) + 23 (closed loop) = 36
Total falling data segments	36	12 (meal) + 24 (closed loop) = 36