This retrospective study included early preterm infants born at Pusan National University Hospital between January 2017 and December 2021. Preterm infants with a gestational age of < 32 weeks who were intubated, received MV for > 24 h, and underwent attempted extubation at least once were enrolled. The exclusion criteria were as follows: 1) death before the first extubation attempt; and 2) transfer to another hospital.
The clinical team decided on intubation/re-intubation using the following criteria: 1) fraction of inspired oxygen (FiO2) on noninvasive respiratory support (NRS) ≥ 0.2 above the initial FiO2 value required to maintain a peripheral oxygen saturation of 90% ; 2) Silverman Anderson retraction score > 4, indicating respiratory failure ; 3) repetitive apneas within 3 h; 4) insufficient respiratory drive; and 5) respiratory acidosis with a partial pressure of CO2 (PaCO2) > 65 mmHg and pH < 7.2. The decision to extubate was based on the following criteria: 1) hemodynamic stability; 2) sufficient respiratory drive; 3) oxygen saturation > 90% with a mean airway pressure (MAP) < 8 cmH2O and FiO2 < 0.3; and 4) absence of any other pathologic conditions, such as shock and sepsis.
After extubation, all patients received nasal continuous positive airway pressure (nCPAP) or high-flow nasal cannula (HFNC) as NRS. According to our unit’s policy and available equipment, patients whose corrected age was < 28 weeks’ gestation or whose body weight was < 1,500 g at the time of the first extubation were placed on nCPAP: otherwise, they were administered HFNC. During nCPAP, the FiO2, positive end-expiratory pressure, and flow rate were set at 0.3, 5-6cmH2O and 5-8L/min, respectively. The FiO2 and flow rate of the HFNC were started at 0.3 and 2 L/kg/min, respectively.
We matched ES patients for gestational age and birth weight to EF patients. Demographic, neonatal, maternal, and ventilation-associated data and clinical outcomes were collected. Demographic and neonatal factors included gestational age, sex, birth weight, 1- and 5-min Apgar scores, use of inotropic agents within 7 days of age, administration and number of pulmonary surfactants administered through an endotracheal tube, presence of patent ductus arteriosus (PDA), and persistent pulmonary hypertension of the newborn. Ventilation-associated factors were: respiratory severity score (RSS) at birth, 1 day, 3 days, 1 week, 2 weeks, 3 weeks, 4 weeks of age, and extubation; corrected age; pH; PCO2; body weight at the first extubation attempt; and oxygen supply after the first extubation. Regarding the EF group, we added the reason for re-intubation, extubation duration after the first extubation, hospital day at the second extubation attempt, and the total number of extubation attempts.
Clinical outcomes were collected, including BPD severity, age at full enteral feeding (> 100 mL/kg/day), duration of oxygen supply, duration of hospital stay, corrected age, body weight and percentile at discharge, tracheostomy, need for home oxygen at discharge, and other co-morbidities, such as IVH and retinopathy of prematurity (ROP).
Maternal data included age at delivery, presence of pathologic chorioamnionitis, and the use and number of antenatal steroids.
Extubation success was defined as survival for ≥ 7 days without the need for intubation, whereas EF was defined as the need for intubation within 7 days of extubation. The chief clinician decided on the initial intubation, timing of extubation, method of post-extubation respiratory support, and need for re-intubation. If high-frequency ventilation was used, it was converted to conventional MV mode before extubation. The percentage body weight was calculated based on Fenton (2013) . Intrauterine growth retardation was defined as < 10th birth weight percentile for gestational age.
A pediatric cardiologist diagnosed the patients with PDA, pulmonary hypertension, and/or other congenital heart diseases using echocardiography. Persistent pulmonary hypertension of newborns was defined as requiring treatment with nitrogen monoxide after confirmation of a bidirectional shunt through the ductus arteriosus or patent foramen ovale, tricuspid regurgitation, or flattening of the interventricular septum on echocardiography. IVH was diagnosed by a pediatric radiologist as grade 3–4 germinal matrix hemorrhage on cranial ultrasonography. PDA, which showed hemodynamic instability, such as severe pulmonary hemorrhage, severe metabolic acidosis, and increased ventilator settings, was treated with ibuprofen or surgery. Patients with contraindications to or who failed ibuprofen treatment underwent surgery. RSS was defined as the MAP multiplied by FiO2, ROP was defined as stage 3 with a plus sign, and stage 4 or 5 of ROP by a pediatric ophthalmologist. The diagnosis and severity of BPD were based on the Jobe-Bancalari criteria . If the oxygen saturation (SpO2) was persistently < 90% despite an FiO2 > 0.4 and a chest radiograph showed diffuse ground-glass opacity, air bronchogram, or a total white-out, respiratory distress syndrome (RDS) was diagnosed . Pulmonary surfactants were administered when RDS was confirmed. Systemic steroid use for weaning from MV was chosen when patients had no central catheter and no signs of infection. Per protocol, a total of 1.1 mg of dexamethasone was administered intravenously or orally over 10 days.
Antenatal steroid was defined as dexamethasone 5 mg twice daily for 2 days at a gestational age of > 24 weeks. Steroid administration not completed 24 h before delivery was considered not having used antenatal steroids.
We matched patients in 1:1 ratio from the ES and EF groups using the propensity score matching method. Using logistic regression analysis, we calculated the propensity score using gestational age and birth weight as covariates. However, we excluded five patients in the ES group because they showed a gestational age and birth weight that was more than two standard deviations above the average gestational age and birth weight of the EF group. Finally, we included 19 patients in the ES group and 24 in the EF group.
All statistical analyses were performed using SPSS version 22 (IBM Corp., Armonk, NY, USA). When comparing the two variables, continuous and non-continuous variables were analyzed using the t-test and chi-squared test, respectively. In addition, the area under the receiver operating characteristic (ROC) curve analysis was performed. Finally, the optimal RSS cutoff values at 1 and 4 weeks of age were evaluated based on the ROC.