Get started
Work

Respiratory severity score as a predictor for need for tracheostomy in infants with severe bronchopulmonary dysplasia

Year: 2026
Type: article
Study design: Observational Study
Abstract: Bronchopulmonary Dysplasia (BPD) is a major sequela of extreme prematurity. A subset of patients with BPD requires tracheostomy placement and chronic ventilatory support [ 1 ]. Patient selection and decision-making regarding tracheostomy placement can be challenging. Timely and selective tracheostomy placement in infants with severe BPD can improve lung growth, decrease the need for sedation, and improve the ability to participate in developmental therapies. Respiratory severity score (RSS), the product of mean airway pressure (MAP) and fraction of inspired oxygen (FiO2), is a marker of respiratory illness severity. In this study, we evaluated whether the RSS at 36–42 weeks postmenstrual age (PMA) can serve as a predictor for tracheostomy placement or death before discharge in extremely preterm infants with severe BPD. This is a retrospective cohort study of preterm infants born at 23 0/7 to 28 6/7 weeks with severe BPD (grade 2 to 3 BPD according to 2019 Neonatal Research Network BPD definition [ 2 ]) in the Prematurity and Respiratory Outcomes Program [ 3 ]. Infants were excluded if they died prior to 36 weeks’ PMA, had cardiopulmonary anomalies, had no BPD or grade 1 BPD, or had unknown RSS data. RSS was calculated or estimated as previously described [ 4 ]. Daily RSS for each infant as documented at 12 noon on each day was averaged over the number of days by completed PMA week (average weekly RSS). Median of average weekly RSS (median RSS) and RSS trend were calculated for each infant from 36 to 42 weeks PMA. Univariate logistic regression was used to assess whether the median RSS and RSS trend (estimated using linear quantile mixed models) over 36-42 weeks PMA were associated with the primary outcome of death before discharge or tracheostomy placement. The University of Wisconsin-Madison IRB determined that this work is not research involving human subjects as defined by DHHS and FDA regulations. A total of 119 infants (89 with grade 2 BPD and 30 with grade 3 BPD) from 6 academic centers in the United States were included in the study. Mean gestational age and birth weight were 25.8 weeks and 774 g, respectively. Ninety-two percent (110/119) of the study cohort had average RSS values available for at least four of the seven weeks. Death before discharge or tracheostomy occurred in 20 (17%) infants that included 12 (10%) who received tracheostomy placement. Infants who experienced the primary outcome had a noticeably higher median RSS (3.78, IQR 2.31–6.38) compared to those who did not (0.71, IQR 0.42–1.38) (Fig. 1 ). Univariate logistic regression further demonstrated that each 2-point increase in median RSS was associated with nearly a threefold increase in the odds of death or tracheostomy (OR = 2.96; 95% CI: 1.82–5.29; p < 0.001). Area under the curve (AUC) was 0.90 (95% CI: 0.82–0.94) and a median RSS ≥ 2.14 yield sensitivity of 90% and specificity of 85%. The RSS trend was associated with the primary outcome to a lesser extent (OR = 1.13 per 0.25 points per week PMA; 95% CI: 1.03–1.32, p = 0.01). AUC was 0.52 (95% CI: 0.34–0.70). When the analysis was repeated for the outcome variable of tracheostomy among survivors, the median RSS still showed a significant association (OR = 2.14; 95% CI: 1.23–3.93, per each 2 point increase in RSS; p = 0.007). However, the RSS trend was no longer significant (OR = 0.76; 95% CI: 0.51—1.11, per each 0.25 point increase in RSS per week, p = 0.19). Plotting symbols (square or star) denote the median average weekly RSS with whiskers marking a 95% confidence interval for the median.
DOI: 10.1038/s41372-026-02720-y
Language: English
Field: Medicine
Open Access status: hybrid
Estimated APC paid: $4,790
OpenAlex ID: W7165967253
FWCI:
Cites: 3
Cited by:
Related to: