Table II.
Rates of bronchopulmonary dysplasia or death by quartile of risk
| BPD or death | ||||
|---|---|---|---|---|
|
| ||||
| Predicted probability of BPD or death | Vitamin A n (%) | Control n (%) | Relative Risk (95% CI) | Risk Difference % (95% CI) |
| Quartile 1 | 30/105 (28.5) | 39/96 (40.6) | 0.69 (0.47 – 1.00) | −12.8 (−25.7 – 0.0) |
| Quartile 2 | 39/97 (40.2) | 56/105 (53.3) | 0.84 (0.64 – 1.10) | −8.2 (−20.9 – 4.6) |
| Quartile 3 | 71/104 (68.3) | 70/98 (71.4) | 0.96 (0.81 – 1.15) | −2.7 (−14.9 – 9.4) |
| Quartile 4 | 82/99 (82.8) | 83/103 (80.6) | 1.04 (0.91 – 1.18) | 3.2 (−7.5 – 13.9) |
| All | 222/405 (54.8) | 248/402 (61.7) | 0.89 (0.80 – 0.99) | −6.9 (−13.0 - −0.7) |
|
| ||||
| Death | ||||
|
| ||||
| Predicted probability of BPD or death | Vitamin A n (%) | Control n (%) | Relative Risk (95% CI) | Risk Difference % (95% CI) |
|
| ||||
| Quartile 1 | 3/105 (2.9) | 5/96 (5.2) | 0.56 (0.13 – 2.33) | −2.3 (−7.8 – 3.3) |
| Quartile 2 | 5/97 (5.2) | 6/105 (5.7) | 0.90 (0.26 – 3.20) | −0.5 (−6.9 – 5.9) |
| Quartile 3 | 16/104 (15.4) | 16/98 (16.3) | 0.82 (0.45 – 1.52) | −3.3 (−14.1 – 7.5) |
| Quartile 4 | 35/99 (35.4) | 28/103 (27.2) | 1.26 (0.83 – 1.91) | 7.6 (−5.9 – 21.0) |
| All | 59/405 (14.6) | 55/402 (13.7) | 1.07 (0.77 – 1.48) | 0.9 (−3.8 – 5.6) |
|
| ||||
| BPD among survivors at 36 weeks’ PMA | ||||
|
| ||||
| Predicted probability of BPD or death | Vitamin A n (%) | Control n (%) | Relative Risk (95% CI) | Risk Difference % (95% CI) |
|
| ||||
| Quartile 1 | 27/102 (26.5) | 34/91 (37.4) | 0.70 (0.46 – 1.05) | −11.4 (−24.4 – 1.6) |
| Quartile 2 | 34/92 (37.0) | 50/99 (50.5) | 0.85 (0.64 – 1.14) | −6.9 (−19.7 – 5.8) |
| Quartile 3 | 55/88 (62.5) | 54/82 (65.8) | 0.94 (0.74 – 1.19) | −3.8 (−18.3 – 10.8) |
| Quartile 4 | 47/64 (73.4) | 55/75 (73.3) | 1.04 (0.84 – 1.29) | 3.2 (−12.5 – 18.8) |
| All | 163/346 (47.1) | 193/347 (55.6) | 0.86 (0.75 – 0.98) | −7.9 (−14.7 - −1.1) |
The relative risk and risk difference calculations accounted for the variables used for stratified randomization (center and birth weight strata [401–750 vs 751–1000 g]) with the Mantel-Haenszel procedure consistent with the original trial manuscript.14 Analyses of the components of the primary outcome accounting for stratified randomization were not shown in the original trial manuscript but are shown here.
BPD=bronchopulmonary dysplasia; PMA=postmenstrual age; CI=confidence interval