Table 3. Trial Outcomes Based on Design-Based Stopping.
| Outcome | SHINE frequentist (n = 936) | Shadow SHINE (n = 800) | ||
|---|---|---|---|---|
| Intensive (n = 483) | Standard (n = 453) | Intensive (n = 408) | Standard (n = 392) | |
| Primary efficacy outcome | ||||
| Favorable, No. (%)a | 102 (21.1) | 102 (22.5) | 89 (21.8) | 89 (22.7) |
| Missing, No. (%)b | 18 (3.7) | 12 (2.7) | 14 (3.4) | 11 (2.8) |
| Risk difference, % (95% CI) | ||||
| Unadjusted | −1.4 (−6.7 to 3.9) | −0.9 (−6.7 to 4.9) | ||
| Adjustedc | −0.1 (−2.6 to 2.5) | 0.3 (−2.5 to 3.0) | ||
| P value for adjusted analysis | .97 | .86 | ||
| Secondary efficacy outcome | ||||
| Favorable NIHSS score (0 or 1), No./total No. (%) | 125/293 (42.7) | 144/306 (47.1) | 109/247 (44.1) | 131/270 (48.5) |
| Unadjusted risk difference, % (95% CI) | −4.4 (−12.4 to 3.6) | −4.4 (−13.0 to 4.2) | ||
| Favorable Barthel Index (range, 95-100), No./total No. (%) | 222/408 (54.4) | 218/387 (56.3) | 193/344 (56.1) | 194/336 (57.7) |
| Unadjusted risk difference, % (95% CI) | −1.9 (−8.8 to 5.0) | −1.6 (−9.1 to 5.8) | ||
| SSQOL | ||||
| No. of patients | 366 | 353 | 312 | 304 |
| Median (IQR) score | 3.73 (2.86 to 4.39) | 3.70 (3.07 to 4.46) | 3.71 (2.84 to 4.39) | 3.73 (3.06 to 4.50) |
| Difference, medians (95% CI) | 0.03 (−0.18 to 0.23) | −0.02 (−0.23 to 0.23) | ||
| Primary safety outcomes | ||||
| Deaths, No. (%) | 44 (9.1) | 49 (10.8) | 37 (9.1) | 41 (10.5) |
| Risk difference, % (95% CI) | −1.7 (−5.6 to 2.1) | −1.4 (−5.5 to 2.7) | ||
| Severe hypoglycemia (glucose <40 mg/dL [<2.22 mmol/L]), No. (%) | 12 (2.5) | 0 | 10 (2.5) | 0 |
| Risk difference, % (95% CI) | 2.5 (1.1 to 3.9) | 2.5 (1.0 to 4.0) | ||
Abbreviations: NIHSS, National Institutes of Health Stroke Scale; SHINE, Stroke Hyperglycemia Insulin Network Effort; SSQOL, Stroke Specific Quality of Life.
Favorable for the primary efficacy outcome is defined as a modified Rankin scale (mRS) score of 0 in patients with mild stroke (baseline NIHSS score, 3-7), mRS score of 0 to 1 in patients with moderate stroke (baseline NIHSS score, 8-14), and mRS score of 0 to 2 in patients with severe stroke (baseline NIHSS score, 15-22).
The analysis for the primary outcome included missing data imputed by a multiple-imputation method.
The adjusted risk difference for the primary outcome was adjusted for baseline stroke severity (NIHSS scores of 3-7 [mild], 8-14 [moderate], and 15-22 [severe]) and thrombolysis use (yes or no).