Abstract
Objectives: To develop risk prediction models for overall and moderate-severe bronchopulmonary dysplasia (BPD) on Days 1, 7, and 21, and compare these to 2011 NICHD model, in a cohort of VLBW infants all receiving prophylactic caffeine treatment.
Methods: This retrospective chart review included 448 infants born between 2012-2018 with gestational age <30 weeks. Models were constructed using step-wise forward logistic regressions, and predictive performances assessed using c-statistic.
Results: BPD developed in 215 (47.9%) infants. “Gestational age,” “type of ventilation,” and “average FiO2 levels” were predictive of BPD (and moderate-severe BPD) on Days 1 and 7; only the latter two factors were predictive on Day 21. Our model’s performance was comparable to the 2011 NICHD model for Days 7 (p=.07) and 21 (p=.14) but was lower for Day 1 (p<.001).
Conclusion: We developed simple, and reliable models for BPD and moderate-severe BPD that neonatologists could easily implement in fastpaced NICU settings.
Methods: This retrospective chart review included 448 infants born between 2012-2018 with gestational age <30 weeks. Models were constructed using step-wise forward logistic regressions, and predictive performances assessed using c-statistic.
Results: BPD developed in 215 (47.9%) infants. “Gestational age,” “type of ventilation,” and “average FiO2 levels” were predictive of BPD (and moderate-severe BPD) on Days 1 and 7; only the latter two factors were predictive on Day 21. Our model’s performance was comparable to the 2011 NICHD model for Days 7 (p=.07) and 21 (p=.14) but was lower for Day 1 (p<.001).
Conclusion: We developed simple, and reliable models for BPD and moderate-severe BPD that neonatologists could easily implement in fastpaced NICU settings.
Original language | English (US) |
---|---|
Journal | Global Journal of Pediatrics & Neonatal Care |
Volume | 4 |
Issue number | 4 |
DOIs | |
State | Published - Mar 1 2024 |
Keywords
- BPD
- Prematurity
- VLBW
- Caffeine treatment
- Prediction model