Table 2.
Glycemia and Insulin Therapy after Randomization, According to Study Group.*
| Variable | Lower Target (N = 349) | Higher Target (N = 349) | P Value† |
|---|---|---|---|
| First qualifying blood glucose level — mg/dl | 0.25 | ||
|
| |||
| Median | 189 | 182 | |
| IQR | 165–243 | 164–232 | |
|
| |||
| Duration from first qualifying blood glucose level to randomization — hr | 0.37 | ||
| Median | 19.7 | 19.4 | |
| IQR | 12.5–28.8 | 11.9–26.2 | |
|
| |||
| Blood glucose level at start of intervention — mg/dl | 0.80 | ||
| Median | 133 | 131 | |
| IQR | 110–160 | 107–165 | |
|
| |||
| Treated with insulin therapy — no. of patients (%) | 344 (98.6) | 215 (61.6) | <0.001 |
|
| |||
| No. of days of insulin therapy | <0.001 | ||
| Median | 5 | 1 | |
| IQR | 3–9 | 0–4 | |
|
| |||
| Average daily insulin dose — units/kg/day | <0.001 | ||
| Median | 0.74 | 0.01 | |
| IQR | 0.37–1.20 | 0.00–0.14 | |
|
| |||
| Adherence to protocol recommendations — no. of recommendations/total no. (%) | 49,835/51,212 (97.3) | 49,921/50,329 (99.2) | <0.001 |
|
| |||
| No. of average daily glucose measurements | <0.001 | ||
| Median | 17.4 | 7.0 | |
| IQR | 13.9–19.6 | 5.5–11.5 | |
|
| |||
| Time to the target range — hr | <0.001 | ||
| Median | 5.5 | 1.5 | |
| IQR | 2.5–11.5 | 0.5–3.0 | |
|
| |||
| Time in the target range — % of time | <0.001 | ||
| Median | 57 | 91 | |
| IQR | 43–67 | 81–96 | |
|
| |||
| Time-weighted glucose average — mg/dl | |||
| Median | 109 | 123 | <0.001 |
| IQR | 102–118 | 108–142 | |
|
| |||
| Hypoglycemia — no. of patients (%)‡ | |||
|
| |||
| Severe | |||
|
| |||
| All | 18 (5.2) | 7 (2.0) | 0.03 |
| Unrelated | 5 (1.4) | 6 (1.7) | 0.71 |
| Related | 13 (3.7) | 1 (0.3) | 0.01 |
|
| |||
| Any | |||
|
| |||
| All | 79 (22.6) | 33 (9.5) | <0.001 |
| Unrelated | 26 (7.4) | 29 (8.3) | 0.62 |
| Related | 64 (18.3) | 5 (1.4) | <0.001 |
|
| |||
| Hypokalemia — no. of patients (%)§ | 76 (21.8) | 64 (18.3) | 0.37 |
|
| |||
| Insulin-dosing error — no. of patients (%) | 4 (1.1) | 1 (0.3) | 0.37 |
To convert values for glucose to millimoles per liter, multiply by 0.05551.
For hypoglycemia and hypokalemia, P values for the comparison between treatment groups were calculated with the use of logistic regression with adjustment for age group and PRISM III-12 score. For other variables, P values were calculated with the use of the Wilcoxon rank-sum or Fisher’s exact test, as appropriate, in the per-protocol population.
Severe hypoglycemia was defined as a blood glucose level below 40 mg per deciliter (2.2 mmol per liter), and any hypoglycemia as a blood glucose level below 60 mg per deciliter (3.3 mmol per liter). Hypoglycemia was considered by the investigators to be related or unrelated to insulin administration according to prospectively defined criteria.16 Patients may have had both unrelated and related hypoglycemia events, so the sums of the values in those categories may exceed the overall total.
Hypokalemia was defined as a potassium level below 2.5 mmol per liter.