Lung ultrasound features predict admission to the neonatal intensive care unit in infants with transient neonatal tachyp
- PDF / 331,018 Bytes
- 8 Pages / 595.276 x 790.866 pts Page_size
- 91 Downloads / 190 Views
ORIGINAL ARTICLE
Lung ultrasound features predict admission to the neonatal intensive care unit in infants with transient neonatal tachypnoea or respiratory distress syndrome born by caesarean section Antonio Poerio 1 & Silvia Galletti 2 & Michelangelo Baldazzi 1 & Silvia Martini 2 & Alessandra Rollo 2 & Sofia Spinedi 2 & Francesco Raimondi 3 & Maurizio Zompatori 1 & Luigi Corvaglia 2 & Arianna Aceti 2 Received: 24 June 2020 / Revised: 17 August 2020 / Accepted: 20 August 2020 # The Author(s) 2020
Abstract We aimed to evaluate the reliability of lung ultrasound (LU) to predict admission to the neonatal intensive care unit (NICU) for transient neonatal tachypnoea or respiratory distress syndrome in infants born by caesarean section (CS). A prospective, observational, singlecentre study was performed in the delivery room and NICU of Sant’Orsola-Malpighi Hospital in Bologna, Italy. Term and late-preterm infants born by CS were included. LU was performed at 30’ and 4 h after birth. LU appearance was graded according to a previously validated three-point scoring system (3P-LUS: type-1, white lung; type-2, black/white lung; type-3, normal lung). Full LUS was also calculated. One hundred infants were enrolled, and seven were admitted to the NICU. The 5 infants with bilateral type-1 lung at birth were all admitted to the NICU. Infants with type-2 and/or type-3 lung were unlikely to be admitted to the NICU. Mean full-LUS was 17 in infants admitted to the NICU, and 8 in infants not admitted. In two separate binary logistic regression models, both the 3P- and the full LUS proved to be independently associated with NICU admission (OR [95% CI] 0.001 [0.000–0.058], P = .001, and 2.890 [1.472–5.672], P = .002, respectively). The ROC analysis for the 3P-LUS yielded an AUC of 0.942 (95%CI, 0.876–0.979; P 0.05 for both models. The reliability of the 3P-LUS to predict NICU admission was first calculated by considering as true-positive only infants with both type 1 lungs: this score had a specificity of 100% (95% CI 95.4–100%) and a sensitivity of 71.4% (95% CI 61.4–79.8%); PPV was 100% (95% CI 95.4–100%), and NPV 97.9% (95% CI 92.1–99.6%). Negative LR was 0.286 (95% CI 0.179–0.456); no calculation of positive LR was possible, as the number of FP infants was zero. When analysing data to clarify the potential role of a single type 1 lung in predicting NICU admission, specificity was 94.6% (95% CI 87.7–97.9%), and sensitivity was 85.7% (95% CI 77.0–91.6%); PPV was 54.5% (95% CI 44.3– 64.4%), and NPV was 98.9% (95% CI 93.6–99.9%). Negative LR was 0.15 (95% CI 0.11–0.22) and positive LR was 15.94 (95% CI 11.74–21.66). Four hours after birth, none of the infants presented with an inhomogeneous white lung. Most infants (69%) showed a normal LUS, with at least one type 3 lung, while the remaining 31% had a bilateral type 2 lung. Back-sliding (LUS worsening over time [4]) was seen in only one infant who was not admitted to the NICU and scored 3 at birth and 2 at 4 h of life on both lungs. The ROC analysis for the 3P-LUS yielded an AUC of 0.
Data Loading...