LUS assessment
LUS is performed using GE Voluson S10 or Philip EP5 ultrasound equipment . The frequency of the linear array probe was 10 to 13 MHz.
Lung ultrasound was performed at one of the following timings after birth: 0.5h, 1h, 2h, 4, 6h. And the performer was a physician who received two months of formal training. This training includes a 2-day course and practice on no less than 100 cases under senior supervision14.
The scanning protocol was based on an adult adaptive method20 and adjusted for infants. The LUS was performed in totally ten regions, as shown in Figure S1. Scanning was performed continuously in each region in case of missing and finished quacky in case of being disturbed by the motion of infants. A similar protocol can be seen in other studies21 but we reduced total twelve regions in theirs to ten regions due to the area of two lateral(left and right) regions is nearly as large as other eight regions. The infants may be posed in a supine position, later position, or prone position if needed. In every region, if any of ”high-risk” patterns were detected, then the region was marked as ”high-risk”. By contrast, only when the whole region has no ”high-risk” patterns, this region can be defined as ”low-risk”. The number of ”high-risk” regions and ”low-risk” regions were used to assess the reliability of prediction. The definitions of ”high-risk” patterns and ”low-risk” patterns were determined by a pre-experiment and were shown in Figure 2.
Results were recorded on a dedicated sheet, not included in patients’ files and masked to other clinicians: this was the best way to mask the clinical conditions to the colleagues performing LUS and the LUS results to other clinicians14. The images and its interpretations for each participant were recorded by the LUS examiner and were linked to a serial number(SN) from1 to 310(after excluded those without qualified data). The SNs were then corresponded to infants outcomes by the physician who offer NICU treatment or follow up these infants. Finally, an independent data analyst who has no idea of lung ultrasound analyzed the data.