Results
Eighteen preterm infants born at a median (range) 27+4(23+4–34+4) weeks of gestation with a median (range) birth weight of 1,100 (460–2,820) g were included in the study. A total of 354 patient days of ventilator data were collected: 269 patient days during invasive NAVA and 85 patient days during noninvasive NAVA. The percentage of excluded data was a median (IQR) of 0.8% (0.4–1.6%) per each patient day: 0.9% (0.4–1.7%) during invasive NAVA and 0.7% (0.3–1.0%) during noninvasive NAVA. Baseline characteristics and ventilator settings at the time of data collection are described in Table 1. On the day of data collection, the infant’s age in days, PMA and body weight ranged from 1 to 90 days, 24+2 to 36+6 weeks and 550 to 3,260 g, respectively. Their median (range) NAVA level was 1.5 (0–2.5) cmH2O/μV. The preset median (range) apnea time and backup ventilation during apnea were 3 (2–10) seconds and 30 (16–40) breaths per minute, respectively (Table 1).
During both invasive and noninvasive NAVA, the time on backup ventilation (%/min) significantly decreased according to an increase in PMA (ρ = -0.491, p < 0.001 and ρ = -0.257, p = 0.017, respectively) (Figure 1A and 1C). Consistently, nRR significantly increased by increasing PMA during both invasive and noninvasive NAVA (ρ = 0.428, p < 0.001 and ρ = 0.225, p = 0.038, respectively) (Figure 1B and 1D). During invasive NAVA, the median time on backup ventilation decreased from 25.1%/min for infants below 26+0 weeks PMA to less than 15%/min for infants above 26+0 weeks PMA. The median nRR increased from 33.0 breaths/min for infants below 26+0 weeks PMA to more than 45 breaths/min in infants above 26+0 weeks PMA. Several outliers were found in both percentage time on backup ventilation and nRR in most PMA groups (Figure 1A and 1B). During noninvasive NAVA, the median time on backup ventilation decreased from 5.6%/min to less than 3%/min, and the median nRR increased from 45 breaths/min to more than 55 breaths/min for infants below 26+0 weeks and above 30+0 weeks PMA (Figure 1C and 1D).
There were 7 episodes of clinical deterioration in 6 patients. The episodes were caused by hemodynamically significant patent ductus arteriosus, clinical sepsis, ventilator-associated pneumonia and necrotizing enterocolitis (Table 2). Relative backup ventilation significantly increased up to the episode of clinical deterioration (Figure 2). The percentage of backup ventilation increased by 1.3 (0.8–2.0) fold from 24 to 18 hours, 1.7 (1.2–3.1) fold from 18 to 12 hours, 2.0 (1.3–4.6) fold from 12 to 6 hours, and 3.4 (2.3–6.4) fold from 6 to 0 hours prior to the episodes compared to the baseline (Friedman test’s χ2=20.69, p <0.001). After stabilization, all 6 infants returned to NAVA support.
Except for the 7 episodes described above, ventilatory support with invasive NAVA continued without interruptions until extubation. Noninvasive NAVA was changed to invasive NAVA if the infant needed intubation or discontinued when the infant stabilized to cope with nasal continuous positive airway pressure or high-flow nasal cannula.