乌鸦传媒

Need for immediate, full intubation in infants born before 28 weeks' gestational age was lower, as was mortality in the first 30 minutes, when they received breathing support via a novel intranasal device versus a standard face mask in a randomized trial.

Rates of delivery-room intubation or death within 30 minutes of birth -- the 's primary outcome -- were 33.1% with the intranasal device compared with 45.1% for neonates given the , a statistically significant difference after adjusting for stratification variables, reported Snorri Donaldsson, MD, of Karolinska University Hospital in Stockholm, Sweden, and colleagues in .

The , which uses short binasal prongs sized for extremely premature infants to deliver air with positive pressure, is made by the British firm Inspiration Healthcare and is sold under the name rPAP. (The company had no role in funding or designing the study, according to the published report, although Donaldsson helped develop the product and receives royalties on sales.)

CORSAD -- the Comparison of Respiratory Support After Delivery on Infants Born Before 28 Weeks Gestational Age -- was set up to evaluate the system in seven neonatal intensive care units in five northern European countries. Enrollment totaled 250 neonates, with 246 born live and actually treated. They were randomized in the delivery room 1:1 to the rPAP device or the T-piece mask system. Blinding was not possible, of course.

Treatments began as soon as the newborn was put on the resuscitation table and continued for 10-30 minutes. Starting parameters were to be 20-25 cm H₂O for positive inflating pressure and 5-8 cm H₂O for positive end-expiratory pressure. Heated humidification was allowed, but centers choosing to use it had to do so for infants in both study arms. Other routine care such as stimulation and plastic wrapping was provided normally as indicated.

Outcomes for the primary endpoint evaluation were determined when infants were intubated, deemed stable, and breathing adequately after at least 10 minutes of the assigned support, or at 30 minutes if clinicians decided continued support was needed. No crossover was allowed. Full intubation was prescribed when any of the following were detected: heart rate fell below 100 bpm, persistent apnea or "poor respiratory effort," respiratory distress, or stubbornly low oxygenation.

Secondary outcomes, tracked for 72 hours, included time to intubation, use of other respiratory support measures such as mechanical ventilation, and treatment with surfactants. The interventions' safety and mechanical performance were assessed as well.

Actual gestational age averaged just under 26 weeks in both groups, with roughly 10% born younger than 24 weeks. Mean infant weight was 800 g and ranged from 411 to 1,420 g. The groups were well balanced for other characteristics such as infant sex, use of antenatal steroids, type of delivery, and presence of eclampsia or preeclampsia.

For the primary endpoint of early intubation or death, the advantage for rPAP was even greater than the raw data indicated after adjustment for these variables, with an odds ratio for these adverse outcomes of 0.53 (95% CI 0.30-0.94). Three infants receiving rPAP died in the delivery room versus one given T-piece support, but the difference was not statistically significant.

Secondary endpoints did not differ significantly between arms, although there was a weak trend toward less surfactant use in the rPAP group (35.5% vs 45.9%), the researchers reported.

One potentially important difference, Donaldsson and colleagues noted, was that "problems with protocol adherence, adverse events, and technical issues" occurred more frequently with the rPAP system (seven versus one infant). Nine infants in the rPAP group were switched to face masks "as rescue and according to protocol," the authors wrote; this appeared to stem from difficulties in fitting the nasal prongs in the smallest newborns, although "introduction of a new and unfamiliar device in a stressful situation" may have played a role, too.

On the other hand, "no deaths or serious adverse events [were] related to the trial protocol or devices used" when evaluated by the trial's independent monitoring committee, the report stated.

CORSAD was not without limitations, Donaldsson and colleagues acknowledged. Besides the lack of blinding, these included the consent procedure, which took place prior to birth and thus excluded babies born without clear warning. "These infants often have a poorer prognosis and often are not exposed to antenatal steroids. This potentially affects the generalizability of our results," the researchers wrote.

Comment