Table 2

Insulin usage, glucose control, and safety of an intensive insulin protocol stratified by study period


Period I
(n = 2,366)
1 Mar 01 to 28 Feb 02
Period II
(n = 3,322)
1 Mar 02 to 30 Jun 03
Period III
(n = 4,786)
1 Jul 03 to 28 Feb 05
P valuea

Patients receiving insulin infusion, n (%)
206 (8.71)
817 (24.59)
2,027 (42.51)
<0.01
Patients receiving subcutaneous insulin, n (%)
307 (14.21)
349 (13.93)
237 (8.65)c
<0.01
Patients receiving any insulin, n (%)
513 (21.68)
1,166 (35.10)b
2,264 (47.48)c
<0.01
Average daily dose of insulin, units/24 hours, mean ± SDd
72.4 (84.9)
55.7 (58.7)
47.5 (43.1)
<0.01
Central lab 6 a.m. blood glucose
n = 2,339
n = 3,278
n = 4,643

     Mean ± SD
144.0 ± 38.1
138.7 ± 34.4
129.3 ± 30.3
<0.01
     Median (IQR)
138 (120–160)
134 (117–153)
125 (112–140)

Average daily blood glucose, mean ± SD
146.6 ± 41.5
142.0 ± 37.4
132.6 ± 30.8
<0.01
Average daily blood glucose, median (IQR)
139 (121–161)
136 (120–156)
128 (115–143)

Average point-of-care blood glucose
n = 700
n = 1,407
n = 2,610

     Mean ± SD
167.4 ± 46.7
148.0 ± 36.8
131.6 ± 31.2
<0.01
Hypoglycemia <40 mg/dL, n (%)e
24 (1.01)
53 (1.60)
103 (2.15)c
<0.01
Hypoglycemia <65 mg/dL, n (%)e
115 (4.86)
351 (10.57)
810 (16.99)
<0.01
High-concentration glucose replacement, n (%)f
86 (3.63)
244 (7.34)
462 (9.69)
<0.01
Hyperglycemia >200 mg/dL, n (%)
330 (14.1)
367 (11.1)
339 (7.26)
<0.01

aAnalysis of variance or chi-square comparing the three study periods; period I is used as reference. bP <0.05 compared with period I (Wald test). cP <0.01 compared with period I (Wald test). dPatients receiving insulin infusion only. eNumber of patients with at least one episode of hypoglycemia; blood glucose values include both point-of-care and central laboratory measurements. fHigh-concentration glucose indicates administration of 50 mL of dextrose 50% (25 g). IQR, interquartile range; SD, standard deviation.

Treggiari et al. Critical Care 2008 12:R29   doi:10.1186/cc6807