This study was designed as a prospective and observational study at an academic tertiary NICU between 2018–2021. The present study was approved by our institutional ethical committee (KA-20029) and written consent was obtained from parents. The protocol was registered with ClinicalTrials.gov Identifier NCT04483492.
Only babies receiving nasal continuous positive airway pressure (nCPAP) who were assigned caffeine with less than or equal to 32 weeks’ gestation at birth were enrolled to the study. Infants with congenital anomalies and perinatal asphyxia were excluded. Fifty-six patients were enrolled to ensure that the power of the test was 80% to detect a difference at a 0.05 significance level. Our caffeine protocol includes 20-minute loading dose of caffeine citrate (20 mg/kg) to the babies less than or equal to 32 weeks’ gestation required mechanic ventilation or the babies without any respiratory support, but having more than one episode of apnea 18.
Demographic data were collected including gender, antenatal exposure to corticosteroids, maternal history of preeclampsia, gestational age, mode of delivery, multiple gestation, birth weight, APGAR, mode of ventilation prior to caffeine administration.
Ultrasonographic examinations to evaluate diaphragm function were done with measuring diaphragm thickness, amplitude of excursion and diaphragmatic velocity of movement. US was performed before and within 5 minutes after caffeine loading dose to be able to show the caffeine affect by two unblinded observers having experience in diaphragm ultrasound (observer 1-pediatric radiologist/fellow and observer 2-neonatologist/fellow). Right and left sides of diaphragm thickness were assessed using B-mode with a 11.4 MHz broadband linear transducer (VF13-5, Acuson X300, Siemens, Erlangen, Germany) placed on the intercostal space in the anterior axillary line while baby was in the supine position. The diaphragm was viewed between the two echogenic layers of the pleura and peritoneum in the transition zone from lung to liver or spleen. Diaphragm thickness was measured as the perpendicular distance between the pleural and peritoneal reflections (Fig. 1). The amplitude of diaphragmatic excursion was measured on the vertical axis of the M-mode US tracing from the baseline to the point of maximum inspiration with a 2.7 MHz broadband sector array transducer (P8-4, Acuson X300, Siemens, Erlangen, Germany) placed below the costal margin between the mid-clavicular and anterior axillary line for the only right diaphragm, with the liver serving as an acoustic window (Fig. 2) (Video 1). There was difficulty in imaging the left diaphragm due to the lung obscuring the view and smaller window of the spleen as compared with the liver window on M-mode US. Therefore, only right side excursions and velocities were measured. On the horizontal axis, the time of right diaphragmatic contraction was noted. It begins at the beginning of inspiration and ended when the peak was reached. The diaphragmatic velocity of movement (mm/s) was calculated by dividing the amplitude of excursion to the time of contraction. The two observers independently recorded their measurements and each observer was blinded to the measurements of the other.
Primary outcome was to investigate changes in the ultrasound measures of diaphragm muscle after caffeine administration. These measurements included both sides of diaphragmatic thicknesses, amplitudes of right diaphragm excursion and right diaphragm velocities obtained before and after the therapy were compared with each other. Also interobserver variability was assessed. Secondary outcome was to show the utility of ultrasound to evaluate diaphragm muscle functions in preterm babies.
The assumption of normality was assessed with Shapiro-Wilk test. If the data was normally distributed, data are presented with mean (standard deviation) and differences assessed for statistical significance using paired sample t-test for changes after caffeine administration or interobserver variability. If the data was non-normally distributed, data are presented with median (interquartile range) and differences evaluated for statistical significance using Wilcoxon rank-sum test for changes after caffeine administration. The intraclass correlation coefficient (ICC) was used to assess inter-observer reproducibility. SPSS 26.0 program (SPSS Inc., Chicago, IL, USA) was used in the statistical analyzes and p < 0.05 was considered statistically significant.