This was a retrospective study of infants born between 23 0/7 and 31 6/7 weeks gestational age (GA) who did not require endotracheal intubation in the delivery room, were admitted to the NICU on bubble CPAP, and subsequently received surfactant replacement therapy for RDS in the NICU. Infants requiring intubation for clinical indications either during delivery room stabilization or in NICU, and those born with congenital anomalies of the airway were excluded. The study was conducted at two NICUs (Cohen Children’s Medical Center, New Hyde Park, NY and North Shore University Hospital, Manhasset, NY), which together are comprised of 99 beds. Both NICUs are covered by the same medical staff and use the same clinical guidelines.

We compared outcomes of premature infants who received surfactant via VL-assisted LISA (January 2021–December 2021) with infants who had received surfactant via DL-assisted InSurE prior to the introduction of LISA (January 2019–December 2020). Eligible infants were identified using the NICU electronic database. The number of infants enrolled in the InSurE cohort was limited due to data access issues related to a change in the electronic medical record (EMR) platform. Criteria for surfactant therapy were the same for both cohorts. Based on a standard NICU written guideline, surfactant was administered when FiO2 reached 30% for infants ≤28 weeks GA or 40% for infants >28 weeks during both time periods. Escalation of CPAP was based on the work of breathing and performed at the clinician’s discretion. Similarly, intubation for MV, extubation from MV to non-invasive ventilation (NIV), and further weaning of respiratory support followed generally accepted practice standards. The procedures were completed by skilled neonatologists, neonatology fellows and advanced care providers. Data on delivery room management, respiratory support, and short- and long-term outcomes were collected from the EMR. Bronchopulmonary dysplasia (BPD) was defined according to the NICHD 2018 definition [13]. Pulmonary hemorrhage was defined as an acute event characterized by respiratory deterioration with blood coming from the upper respiratory tract or the endotracheal tube [14].

Infants in the historical cohort were treated with the standard InSurE method described by Verder et al. This technique involves surfactant administration after ETT intubation using direct laryngoscopy (DL), followed by brief MV, and early extubation typically within 1–2 h [15].

LISA was administered using strict guidelines that include patient eligibility, procedure location, contraindications, equipment, preparation instructions, VL technique, postprocedural monitoring, equipment care and documentation. Our standard LISA equipment includes neonatal VL (Peak Medical, Netanya, Israel), 16-gauge 5.25-inch sterile angiocatheter in lieu of thin catheter, marker, measuring tape and 5 mL surfactant syringe. Bubble nasal CPAP was continued during the procedure, with escalation of the CPAP level as indicated based on the work of breathing, chest x-rays and clinical condition. We did not use analgesia or sedation since VL-assisted LISA decreases the number of intubation attempts and is quicker than conventional DL-assisted LISA [12, 16]. Infants who did not tolerate the procedure, defined by oxygen saturation below 80% or heart rate below 80 beats/min for 10 s, were allowed to recover by withdrawing the VL blade or by slowing down surfactant delivery, and adjusting oxygen. If a second surfactant dose was necessary, it was delivered by LISA in infants who remained clinically stable and were breathing spontaneously.

Evaluation of the health care costs was performed by using the cost-consequence model described by Yao et al. [17]. This model was developed based on previously published literature to assess economic inputs that influence total NICU costs. The key input of cost per day for NICU hospitalization was estimated based on national data. The average total cost of care (cost per patient per hospital stay) was based on the average cost per day with MV and the average cost per day without MV. Total hospital costs for each infant were generated using the number of days with and without MV, and the average cost of NICU hospitalization for each cohort was calculated by dividing the sum of patient hospital costs by the number of patients in each cohort. These results were compared overall and across GA between the two cohorts. Complications, such as BPD, IVH, ROP, and pneumothorax are the main modifiers of the total cost. However, these modifiers were not included in the calculation because there was no difference in these outcomes between the cohorts. All costs were adjusted to the inflation index between the two periods as reported by the Consumer Price Index.

Descriptive statistics provided are mean and SD for continuous variables, or frequencies and percentages for categorical variables, unless otherwise stated. Missing data were omitted from analysis. Statistical analysis included both comparisons between the entire cohorts as well as comparisons between subgroups after stratification based on GA. Each continuous variable was tested for normality within each cohort using a Shapiro-Wilk test. A Wilcoxon rank-sum test was performed for most demographic data, delivery room data, and respiratory outcomes, where normality was not assumed. Other tests include Fisher’s exact test or chi-square test for categorical variables, and Student’s t-test for continuous variables. The total cost of care was calculated by the evaluated cost per day with MV and the cost per day without MV. Any p-value less than 0.05 was considered statistically significant. Statistical analysis was performed using R Statistical Software (R Core Team 2022, V.4.2.1.).