Table 1.
Males | Females | P-value | |
---|---|---|---|
Demographic statistics | |||
# Patients | 91 (63%) | 54 (37%) | |
Age | 67 [57, 77] | 67 [58, 74] | 0.63a |
Mortality | 18% | 19% | 1.0b |
APACHE II score | 20 [16 27] | 19.5 [17, 26] | 0.98a |
First day SOFA score | 6 [4 8] | 5.5 [4, 8] | 0.46a |
ICU length of stay (h) | 108 [67.2, 188.4] | 127.2 [64.8, 213.6] | 0.91a |
SPRINT duration (h) | 83 [45.5, 157.3] | 86.5 [39, 167] | 0.81a |
T2DM (%) | 13 (14%) | 11 (20%) | 0.4b |
Per-patient GC statistics | |||
Median BG (mmol/L) | 5.7 [5.2, 6.1] | 6.0 [5.3, 6.4] | 0.06a |
Median % BG 4.4–8.0 mmol/L | 83 [72, 90] | 82 [67, 89] | 0.3a |
Median % BG < 4.0 mmol/L | 1.4 [0, 5.5] | 1.4 [0, 6.9] | 0.42a |
Median %BG < 2.2 mmol/L | 0 [0, 0] | 0 [0, 0] | 1.0a |
BG measurements/day | 15.8 [14.4, 17.5] | 15.7 [14.5, 18.2] | 0.47a |
Median insulin (U/h) | 3 [2, 3] | 3 [2, 3] | 0.26a |
Median feed excl. hours not fed (g/h) | 3.5 [2.1, 5.5] | 2.8 [1.8, 3.9] | < 0.01a |
Median feed excl. hours not fed (%GF) | 51 [30, 80] | 51 [30, 75] | 0.61a |
GF (g/h) | 6.5 [6.5, 7.4] | 5.2 [5.2, 5.7] | < 0.01a |
Cohort statistics | |||
Cohort BG (mmol/L) | 5.6 [4.9, 6.6] | 5.9 [5.0, 6.9] | < 0.01ac |
% BG 4.4–8.0 mmol/L | 80.6 | 76.9 | < 0.01b |
% BG < 4.0 mmol/L | 3.2 | 3.8 | 0.11b |
%BG < 2.2 mmol/L | 0 | 0 | 1.0b |
Insulin (U/h) | 3 [1 4] | 3 [2 4] | < 0.01a |
Hours not fed | 1914 (35%) | 861 (27%) | < 0.01a |
Feed excl. hours not fed (g/h) | 3.8 [2.0, 5.8] | 2.4 [1.7 3.9] | < 0.01a |
Feed excl. hours not fed (%GF) | 55 [29 83] | 47 [29 75] | < 0.01a |
P-values are not adjusted for multiple comparisons. Median [IQR] is given where appropriate. T2DM = pre-diagnosed type 2 diabetes mellitus, GF = goal feed, and BG = blood glucose. BG is hourly resampled to allow fair comparison.
Statistical difference is shown using (a) the Wilcoxon rank-sum test or (b) Fisher exact test where appropriate. (c) indicates clinical equivalence regardless of statistical significance, as further explained in the methods