🚨 O2 Subset Feeding Timing
Executive Summary¶
Analysis Population: Infants requiring O2 support at 36 weeks PMA (n=60)
Research Question: Among respiratory-compromised infants, does starting oral feeding earlier vs later affect time to full oral feeding?
Key Findings: - No significant difference in time to full feeding between early and late feeders (p=0.3614) - No clear optimal timing identified within data range
Sample Description¶
O2 Subset Characteristics¶
Total sample: 60 infants with respiratory support at 36 weeks PMA
Timing group cutpoints used: - Median time to first feeding: 57.6 days: Early n=30, Late n=30 - PMA at first feeding < 35.0 weeks: Early n=33, Late n=27
Baseline Characteristics¶
Baseline Characteristics by Feeding Timing Group¶
Continuous Variables:
| Variable | Early (mean ± SD) | Late (mean ± SD) | p-value | Cohen's d |
|---|---|---|---|---|
| gestational_age_weeks | 28.5 ± 1.8 | 25.6 ± 1.3 | < 0.0001 | 1.82 |
| birth_weight_grams | 1169.5 ± 357.6 | 718.5 ± 190.9 | < 0.0001 | 1.57 |
| time_to_first | 40.5 ± 12.1 | 71.0 ± 12.7 | < 0.0001 | -2.45 |
Categorical Variables:
| Variable | Chi-square | p-value | Cramér's V |
|---|---|---|---|
| medicalcomplexity | 4.32 | 0.1153 | 0.27 |
| mechanical_ventilation | 6.28 | 0.0122 | 0.32 |
| baby_sex | 1.88 | 0.1709 | 0.18 |
Primary Outcome Analysis¶
Time to Full Oral Feeding: Early vs Late Feeders¶
Early feeders (n=30): - Mean: 19.1 days (SD: 11.9) - Median: 18.0 days (IQR: 10.5-27.6)
Late feeders (n=30): - Mean: 22.2 days (SD: 13.7) - Median: 19.8 days (IQR: 11.9-30.3)
Statistical Test: - t-statistic: -0.920 - p-value: 0.3614 - Cohen's d: -0.238 (small)
Outcome Distribution by Group¶
Distribution of time to full oral feeding by early vs late feeding group.
Feeding Trajectory Analysis¶
Time-Based Feeding Trajectory Analysis¶
Sample: n=60 with complete data Correlation (time_to_first vs time_to_FOF): r = 0.052
Linear Model: - R²: 0.003 - Slope: 0.034 (p = 0.6945) - Interpretation: Each additional day to first feeding associated with 0.03 day change in time to FOF
Quadratic Model (testing for optimal timing): - R²: 0.045 - Linear coefficient: 0.595 (p = 0.1057) - Quadratic coefficient: -0.0048 (p = 0.1164)
No optimal timing identified within data range (quadratic coefficient ≤ 0 or optimal outside range)
Trajectory Visualization¶
Relationship between time to first feeding and time to full oral feeding within O2 subset.
Stratified Analysis¶
Stratified Analysis by Gestational Age Category¶
Moderate Preterm: - Early feeders (n=20): mean 15.9 days - Late feeders (n=3): mean 18.3 days - p-value: 0.7225 - Cohen's d: -0.223
Extremely Preterm: - Early feeders (n=10): mean 25.5 days - Late feeders (n=27): mean 22.6 days - p-value: 0.5591 - Cohen's d: 0.218
Summary¶
Summary¶
This analysis examined feeding timing effects within the respiratory-compromised subpopulation (infants requiring O2 at 36 weeks PMA).
Methodological Note: This complements Report 13 which compared O2 vs non-O2 groups. This report focuses on whether timing matters within the O2-requiring group specifically.
Statistical Conclusion: No statistically significant difference in time to full oral feeding between early and late feeders within the respiratory-compromised subgroup.