Table 3.
Relative Risk for Continuous Clinical Outcomes of all Study Infants
| SGA | AGA | |||||||
|---|---|---|---|---|---|---|---|---|
| Continuous outcomes | Vitamin A (n = 50) | Control (n = 54) | p value for treatment difference within SGA | Vitamin A (n = 355) | Control (n = 348) | p value for treatment difference within AGA | p value for GA group and treatment interaction | p-value for difference between SGA and AGA groups | 
| Fraction of inspired oxygen, median (5th, 95th percentiles) | ||||||||
| 36 wk | 0.21 (0.21, 1.0) | 0.26 (0.21, 1.0) | 0.50 | 0.23 (0.21, 1.0) | 0.25 (0.21, 1.0) | 0.98 | 0.48 | |
| 28 d | 0.30 (0.21, 0.58) | 0.32 (0.21, 0.75) | 0.14 | 0.34 (0.21, 0.87) | 0.32 (0.21, 0.80) | 0.74 | 0.64 | |
| Length of hospitalization, median (5th, 95th percentiles) | ||||||||
| Infants who survived | 91 (54, 176) | 90 (55, 242) | 0.85 | 95 (62, 168) | 94 (57, 166) | 0.40 | 0.07 | 0.13 | 
| Infants who died | 20 (6, 213) | 20 (5, 141) | 0.56 | 20 (3, 226) | 19 (5, 152) | 0.37 | 0.09 | 0.51 | 
Abbreviations: AGA, average for gestational age; GA, gestational age; SGA, small for gestational age.
Note: For FIO2, p values were obtained using Wilcoxon rank sum tests adjusting for gestational age. For length of hospitalization, a Poisson model assuming an interaction between treatment and GA group was used to obtain p values. The model controlled for GA.