Comparison of three clinical scoring tools for bronchiolitis to predict the need for respiratory support and length of stay in neonates and infants up to three months of age

Domenico Umberto De Rose; Chiara Maddaloni; Ludovica Martini; Annabella Braguglia; Andrea Dotta; Cinzia Auriti

doi:10.3389/fped.2023.1040354

Comparison of three clinical scoring tools for bronchiolitis to predict the need for respiratory support and length of stay in neonates and infants up to three months of age

Front Pediatr. 2023 Feb 17:11:1040354. doi: 10.3389/fped.2023.1040354. eCollection 2023.

Authors

Domenico Umberto De Rose¹, Chiara Maddaloni¹, Ludovica Martini¹, Annabella Braguglia², Andrea Dotta¹, Cinzia Auriti¹

Affiliations

¹ Neonatal Intensive Care Unit, Bambino Gesù Children's Hospital IRCCS, Rome, Italy.
² Neonatal Sub-Intensive Care Unit and Follow-up, Bambino Gesù Children's Hospital IRCCS, Rome, Italy.

Abstract

Background: Bronchiolitis severity can be assessed using different clinical scores. Some of the most used are the Wang Bronchiolitis Severity Score (WBSS), the Kristjansson Respiratory Score (KRS), and the Global Respiratory Severity Score (GRSS), calculated on the vital parameters and the clinical conditions.

Objective: To assess which of the three clinical scores better predicts the need for respiratory support and length of hospital stay in neonates and infants younger than three months, admitted to neonatal units for bronchiolitis.

Methods: Neonates and infants younger than three months admitted to neonatal units from October 2021 to March 2022 were included in this retrospective study. The scores were calculated in all patients soon after admission.

Results: Ninety-six patients (of whom 61 neonates) admitted for bronchiolitis were included in the analysis. Median WBSS at admission was 4.00 (interquartile range, IQR 3.00-6.00), median KRS was 4.00 (IQR 3.00-5.00), and median GRSS 4.90 (IQR 3.89-6.10). We found significant differences in all three scores between infants who needed respiratory support (72.9%) and those who did not (27.1%) (p < 0.001). A value >3 for WBSS, > 3 for KRS, and >3.8 for GRSS were accurate in predicting the need for respiratory support, with a sensitivity of 85.71%, 75.71%, and 93.75% and a specificity of 80.77%, 92.31%, and 88.24%, respectively. The three infants who required mechanical ventilation had a median WBSS of 6.00 (IQR 5.00-6.50), a KRS of 7.00 (IQR 5.00-7.00), and a GRSS of 7.38 (IQR 5.59-7.39). The median length of stay was 5 days (IQR 4-8). All three scores were significantly correlated with the length of stay, although with a low correlation coefficient: WBSS with an r² of 0.139 (p < 0.001), KRS with an r² of 0.137 (p < 0.001), and GRSS with an r² of 0.170 (p < 0.001).

Conclusion: Clinical scores WBSS, KRS, and GRSS calculated on admission accurately predict the need for respiratory support and the length of hospital stay in neonates and infants younger than three months with bronchiolitis. The GRSS score seems to better discriminate the need for respiratory support than the others.

Keywords: RSV; bronchiolitis; clinical scoring tools for bronchiolitis; influenza; newborns; respiratory infections; rhinovirus; viruses.