Skip to main content

Table 2 Summary of primary and secondary outcomes dealing with glycaemic variability

From: Effect of insulin infusion line on glycaemic variability in a perioperative high dependency unit (HDU): a prospective randomised controlled trial

Parameters Standard infusion line Optimised infusion line Between-group difference
N Values N Values Mean (95% CI) P value
Glucometer
 GLI over 48 h, ITT analysis* mean (95% CI) 86 5.37 (5.21; 5.53) 86 5.46 (5.27; 5.64) −0.09 (−0.34; 0.16) 0.49ǂ
 GLI over 48 h, complete case analysis* mean (95% CI) 79 5.35 (5.21; 5.49) 80 5.44 (5.26; 5.62) −0.09 (−0.32; 0.13) 0.40ǂ
 Capillary blood glucose average over 48 h, mg dl−1 79 134.8 ± 21.0
132.0 (117.0–149.0)
80 138.7 ± 26.2
133.5 (123.0–146.0)
0.58#
 Capillary standard deviation over 48 h, mg dl−1 79 31.0 ± 13.3
29.0 (25.0–36.0)
80 32.5 ± 18.3
29.0 (22.5–37.5)
0.93#
 MAGE over 48 h, mg dl−1 79 9.8 ± 6.2
9.0 (6.0–12.0)
80 10.1 ± 7.4
9.0 (5.0–12.0)
0.98#
 At least one hypoglycaemia, n (%) 79 11 (13.9) 80 8 (10.0) 0.45Ɨ
 At least one hyperglycaemia, n (%) 79 51 (64.6) 80 48 (60.0) 0.55Ɨ
CGM
 iPro2 glucose average over 48 h, mg dl−1 74 128.0 ± 21.3
124.5 (111.0–142.0)
70 133.5 ± 25.3
129.0 (117.4–139.0)
0.26#
 iPro2 glucose standard deviation over 48 h, mg dl−1 74 23.3 ± 8.9
21.0 (18.0–25.0)
70 25.1 ± 12.3
21.0 (17.0–29.0)
0.75#
 At least one hypoglycaemia detected by iPro2, n (%) 74 19 (25.7) 70 9 (12.9) 0.052Ɨ
 Time spent in hypoglycaemia for 1000 h recorded with iPro2 (h) 74 9.7 ± 25.0
0.0 (0.0–1.74)
70 4.4 ± 14.8
0.0 (0.0–0.0)
0.059#
 At least one hyperglycaemia detected by iPro2, n (%) 74 45 (60.8) 70 46 (65.7) 0.54Ɨ
 Time spent in hyperglycaemia for 1000 h recorded with iPro2 74 64.3 ± 97.3
15.6 (0.0–97.2)
70 92.7 ± 181.1
13.9 (0.0–104.17)
0.78#
 Time in target range recorded with iPro2 74 71 ± 26
80 (54–92)
70 72 ± 25
80 (64–78)
0.82#
  1. Values are expressed as mean ± standard deviation and median (interquartile range) unless otherwise indicated
  2. ITT Intended-to-treat
  3. * Primary outcome: means and their 95% CI were calculated after log transformation of GLI values
  4. After the mean of 20 imputations to treat the missing values
  5. ǂStudent t test
  6. #Mann–Whitney U test
  7. ƗChi-square test