# Early lung ultrasound score combined with umbilical cord-blood procalcitonin improves 1-year prognostic stratification in preterm neonates with respiratory distress syndrome

## TL;DR

Combining lung ultrasound scores and cord-blood procalcitonin improves 1-year outcome prediction in preterm infants with respiratory distress syndrome.

## Contribution

A novel combined model using lung ultrasound and cord-blood procalcitonin improves prognostic stratification in preterm neonates with RDS.

## Key findings

- The combined model achieved an AUC of 0.87, outperforming LUS (AUC 0.76) and PCT (AUC 0.78) alone.
- The high-risk category from the model was independently associated with a 2.9-fold higher hazard of the composite outcome.
- Bootstrap optimism-corrected AUC remained at 0.86, confirming robustness.

## Abstract

Respiratory distress syndrome (RDS) remains a major cause of morbidity in very preterm infants. Lung ultrasound score (LUS) provides a bedside assessment of lung aeration and has demonstrated utility for early respiratory decision-making, but its prognostic performance for long-term outcomes is only moderate. Procalcitonin (PCT) measured in umbilical cord blood may reflect perinatal inflammatory exposure and risk of infection-related complications.

We conducted a single-center prospective cohort study enrolling infants born at 24 + 0–33 + 6 weeks' gestation who were admitted to the NICU within 6 h of birth and were clinically diagnosed with RDS. Within 6 h after delivery, a standardized 12-zone LUS and umbilical cord-blood PCT were obtained. The primary endpoint was a composite of bronchopulmonary dysplasia, severe intraventricular hemorrhage, necrotizing enterocolitis, culture-proven sepsis occurring after 72 h of age, or all-cause death within 12 months' corrected age. Discrimination was evaluated using ROC analysis and DeLong tests. Time-to-first-event associations were examined using multivariable Cox regression. Internal validation used bootstrap optimism correction.

Among 290 infants, 110 (37.9%) reached the composite endpoint (event-free proportion 62.1%). LUS alone achieved an AUC of 0.76 (95% CI 0.70–0.82), and PCT alone an AUC of 0.78 (0.72–0.84). A logistic model combining LUS and log-transformed PCT improved discrimination to an AUC of 0.87 (0.83–0.92), outperforming each single marker (paired DeLong p < 0.001). At the Youden-optimal operating point, sensitivity was 82% and specificity 80%. In multivariable Cox analysis, the high-risk category defined by the combined model was independently associated with higher hazard of the composite outcome (HR 2.9, 95% CI 2.0–4.1), alongside lower gestational age, lower birthweight, and early-onset infection. Bootstrap optimism-corrected AUC was 0.86.

In preterm infants with RDS, early integration of 12-zone LUS and cord-blood PCT improves prediction of 12-month major morbidity or death compared with either marker alone. This bedside approach may support early risk stratification. External validation and impact studies are needed before score-guided management is recommended.