Use of Intermittent vs Continuous Pulse Oximetry for Nonhypoxemic Infants and Young Children Hospitalized for Bronchiolitis: A Randomized Clinical Trial
McCulloh R, Koster M, Ralston S, et al
JAMA Pediatr. 2015;169:898-904
Study Summary
Although use of pulse oximetry in hospitalized children with respiratory infections is common, the data supporting its use are limited. Recommendations from the American Academy of Pediatrics in 2006 and 2014 suggest that intermittent (rather than continuous) pulse oximetry is an appropriate use of the technology, and they specifically recommend intermittent pulse oximetry measurements for children who are not receiving oxygen.
This study randomly assigned children to either intermittent or continuous pulse oximetry monitoring, with approximately 80 children in each group. The children were enrolled from those hospitalized at children's hospitals during the 2009 through 2014 respiratory seasons. All children were ≤24 months old at enrollment and were enrolled within 24 hours of admission for acute bronchiolitis. Infants who were premature, had severe chronic medical conditions, were admitted to the intensive care unit (ICU) initially, or who had immune-compromising conditions were excluded. Because of the nature of the study, parents as well as the investigators and other treating providers were not blind to study assignment.
Patients in the continuous monitoring group received continuous pulse oximetry throughout their hospital stay. Patients in the intermittent group were monitored intermittently with pulse oximetry once they were no longer hypoxic (resting oxygen saturation ≥90%). If children in the intermittent group required oxygen for persistent saturations <90%, they were placed on continuous pulse oximetry until they were weaned from oxygen, when intermittent monitoring began again. The oxygen saturation levels of the "intermittent group" children were checked with each vital signs assessment, at parental request, or any time that the child experienced a clinically significant change in condition. The primary outcome of interest was the length of stay (LOS). The investigators also evaluated duration of oxygen use, how frequently the children required an escalation in care, laboratory data, medication use, intravenous hydration, as well as symptoms upon presentation.
A total of 161 children were included in this analysis, 80 of whom were in the continuous pulse oximetry group. Group differences were present at enrollment, with larger proportions of the intermittent group having been in daycare (18.5% vs 10%) and having a family history of asthma of (55.6% vs 37.5%). Children in the intermittent group were also more likely to have received oral corticosteroids or antibiotics prior to admission. Overall, there was a mean LOS difference of 2.7 hours between the two groups, but this difference did not reach statistical significance (95% confidence interval, -7.6 to 13.0 hours). Laboratory testing appeared to occur more commonly in the children receiving intermittent monitoring, but these differences likewise did not reach statistical significance. As far as untoward outcomes, similar percentages of children in both groups received supplemental oxygen (43%), corticosteroids, and antibiotics, and approximately 5% of both groups required transfer to the ICU during their hospital stay. The authors concluded that intermittent, compared with continuous, pulse oximetry monitoring in infants and children with bronchiolitis did not reduce LOS.
Viewpoint
It is undoubtedly difficult to do a trial such as this, utilizing multiple institutions that probably vary in their care of infants with bronchiolitis. However, I interpret the results a little more cautiously than do the authors, in that I believe that we can only view this as a "pilot" study. The authors acknowledged that their study was only powered to detect a difference in LOS ≥18 hours. Moreover, the randomization did not result in balanced groups. To the authors' credit, and perhaps to the detriment of this trial, the imbalances tended to bias against being able to show a shorter LOS in the intermittent group, with greater percentages of the intermittent group being male, having daycare exposure, or having a family history of asthma or reactive airway disease. In addition, the trial was not powered for any of the secondary outcomes, including such safety outcomes as requiring higher levels of care or death. That said, I think these are valuable data that can guide additional studies on how to implement a larger trial.
Abstract