Postextubation use of certain modes of Noninvasive Respiratory Support was associated with fewer incidents of extubation failure in critically ill children, a meta-analysis found.

Compared with conventional oxygen therapy, both continuous positive airway pressure (CPAP; OR 0.43, 95% credible interval [CrI] 0.17-1.0) and high-flow nasal cannula (HFNC; OR 0.64, 95% CrI 0.24-1.0) appeared more effective at reducing extubation failure, the requirement for reintubation within 48 to 72 hours of removing the artificial airway, reported Narayan Prabhu Iyer, MBBS, of the University of Southern California, Los Angeles and coauthors.

As for treatment failure — defined as reintubation or use of another mode noninvasive Respiratory Support — CPAP (OR 0.27, 95% CrI 0.11-0.57) and HFNC (OR 0.34, 95% CrI 0.16- 0.65) were both more effective compared with conventional oxygen therapy, the meta-analysis published in JAMA Pediatrics found.

While not statistically significant, bilevel positive airway pressure (BiPAP) trended toward being a numerically better treatment option for preventing treatment failure and extubation failure in the pediatric ICU (PICU).

Out of the various types of Noninvasive Respiratory support, CPAP treatment had the highest probability of being the best option for minimizing extubation failure and treatment failure.

However, both CPAP and BiPAP were associated with an approximately 3% increase in nasal injury and abdominal distension.

Researchers noted that no method of noninvasive respiratory support has been proven to be the “optimal method;” its use varies in different settings as rescue or planned prior to extubation. Moreover, it is unclear which critically ill children are likely to need planned breathing support, if they need it at all.

Martin Kneyber, MD, of the University of Groningen, the Netherlands, commented that the present meta-analysis did not include individuals who did not require postextubation respiratory support.

“Now, the most challenging aspect of all of these studies is that they have included many patients in whom pre-emptive use of non-invasive respiratory support post-extubation was ordered. This is not my practice; I extubate 99% and wait and see,” he told MedPage Today in an email. “So, in my view, future studies should really identify the subset of patients that benefits from post-extubation non-invasive respiratory support.”

Jeremy Loberger, MD, of the University of Alabama, Birmingham, however, commented that CPAP has been particularly successful in infants 6 months old or younger, especially when it is included as part of extubation strategy right from the start. “I’ve seen greater success with supporting them more (with HFNC, CPAP, or BiPAP) at the start rather than trying to rescue them when they show signs of failing on conventional oxygen therapy,” he said.

For the meta-analysis, Iyer’s group included nine randomized clinical trials featuring 1,421 pediatric patients. Participants in the studies were critically ill, born no earlier than 37 weeks gestational age, and were supported with mechanical ventilation for over 24 hours followed by postextubation noninvasive respiratory support. Most were under 48 months old.

Participants with certain noncardiac congenital abnormalities were not included in the meta-analysis, so the report may have limited applicability to older pediatric patients and those coming out of heart surgery, cautioned Hunter Wilson, MD, of Children’s Healthcare of Atlanta, who was not involved with the meta-analysis.

“In my experience older patients often have more robust respiratory mechanics than younger patients and so extubating an older cohort of patients to CPAP may be too conservative and actually prolonged hospital length of stay and need for sedation, etc.,” he told MedPage Today.

“I think it will be difficult to clearly translate these findings to the population of patients who are extubated following heart surgery. There are some additional ways in which positive pressure can support heart function and so it may be difficult to translate findings from a general PICU population to this population in particular,” Wilson added.

Other limitations to the meta-analysis include low-certainty results due to bias in some included studies and a lack of data on outcomes or resource utilization.

“When choosing a NRS [noninvasive respiratory support] mode, considerations of equipment availability, associated costs to patients and the health care system, and the need for a high level of nursing care are also important,” Iyer and colleagues wrote. “These factors vary across health systems and geographic regions and are likely to have an impact on the relative efficacy of different NRS modes.”