Table VII.
Neonatal outcomes: Multivariate analyses examining the effects of treatment assignment on neonatal outcomes
| Multivariable model† |
||
|---|---|---|
| Outcomes | Relative risk (95% CI)‡ | P value |
| Primary outcome | ||
| Ligation or cardiology follow-up | 0.73 (0.50-1.04) | .083 |
| PDA ligation | 0.94 (0.62-1.43) | .774 |
| Cardiology follow-up, outpatient | 0.62 (0.32-1.21) | 0.161 |
| Secondary outcomes | ||
| NEC | 0.89 (0.60-1.32) | .574 |
| BPD | 0.89 (0.60-1.33) | .582 |
| BPD or death | 0.98 (0.72-1.34) | .913 |
| Death | 1.23 (0.75-2.01) | .405 |
| Mean difference between groups (95% CI)§ | ||
| Days until enteral intake 120 mL/kg/d * | +0.85 (0.77-0.93) | <.001 |
| Days until last gavage feeding* | +0.93 (0.89-0.97) | .002 |
| Mean difference between groups (95% CI)§ | ||
| Daily weight gain, g/kg* | −0.26 (−1.95 to 1.46) | .769 |
| Other outcomes | Relative risk (95% CI)‡ | |
| PDA (moderate/large) at 10 d after randomization* | 0.53 (0.35- 0.81) | .003 |
| Rescue criteria met | 0.49 (0.32-0.76) | .001 |
| Received rescue treatment | 0.39 (0.25-0.59) | <.001 |
| Pulmonary hemorrhage* | 1.13 (0.10-12.9) | .916 |
| sIVH | 1.03 (0.52-2.01) | .942 |
| PVL (cystic) | 0.90 (0.48-1.70) | .744 |
| ROP (treated) | 1.03 (0.40-2.68) | .940 |
| Pneumonia* | 0.67 (0.29-1.55) | .351 |
| Bacteremia* | 1.25 (0.82-1.91) | .295 |
| Bacteremia non-CONS* | 0.88 (0.47-.64) | .677 |
| Received dopamine ≥3 d* | 0.48(0.21-1.10) | .082 |
| Received corticosteroids ≥7 d* | 0.77 (0.46-1.29) | .317 |
| Received furosemide ≥14 d* | 0.77 (0.50-1.18) | .225 |
| Mean difference between groups (95% CI)§ | ||
| Days until discharge* | 1.02 (0.99-1.05) | .278 |
Reported outcome is for the incidence or time interval that occurred after randomization.
Multivariate model: generalized estimating equations were used to account for clustering within center, gestational age (<26 wk vs ≥26 wk), multiple birth, and early-onset bacteremia (see Methods). An interaction term between treatment assignment and gestational age was also included in models for the outcomes of death, bacteremia non-CONS, and days until enteral intake 120 mL/kg/day, because the interaction between treatment assignment and gestational age for these 3 outcomes reached a level of significance of P< .15.
Relative risk and 95% CI in the ERT group compared with the CT group.
Mean difference and 95% CI of Early Treatment compared with Conservative Treatment group.