Intermittent Monitoring of Oxygen Saturation in Infants and Children With Acute Bronchiolitis

Abstract

Peekaboo is always fun. Or at least it should be. Confidence that what is revealed will be in line with your expectation is the underlying principle—shatter that and the trust has gone. So the element of surprise is abalancedestimate—whatever’s there won’t be too scary. In this issueof JAMAPediatrics,McCullohandcolleagues1play peekaboo with pulse oximetry for infants and children with bronchiolitis. The intermittentmonitoring anddisplayof oxygen saturation values to clinical staff had similar outcomes to the continuous monitoring and display of oxygen saturation values. No nasty surprises. Prior to the widespread use of pulse oxygen saturation monitoring, itused tobeeasy;youwerepinkorblue.Ataround 85% oxygen saturation, the clinical determination of cyanosis is a blunt tool.A low-cost, technologically reliablepulseoximeter is now ubiquitous bedside furniture in our hospitals. Staff members like that it is easy to use and easy to read, and theyhaveadopted it as a surrogatebabysitter. Parents toohave faith that the illuminatednumber that it displays is the key indicator of their child’s health. But all this comes at a cost; the active incorporation of the percentile scale of pulse oxygen saturation into clinical practice has displaced clinical acumen and made a dent in one’s own clinical confidence. With no evidence of increased virulence, admissions for acute viralbronchiolitishave increasedover the last20yearswhileoutcomes, in particular mortality rates, remain unchanged. In emergencydepartments, twice asmany infantswith bronchiolitis would be admitted to the hospital if oxygen saturation levelswere 92% rather than 94%.2 If oxygen saturation is displayed 3 percentage points higher than measured, clinicians assessing bronchiolitis in emergency departments use their clinical acumen to amuchgreater extent in decisionmaking.3 With the increasing awareness of the potential limitations of pulse oxygen saturation to discriminate and predict health anddisease,wearenowonapath to retrofit clinical understanding to this technology; small differentials in oxygen saturationare important for outcomes inpreterm infants,4 desaturation during an asthma exacerbation heralds loss of control,5 self-resolving fluctuationofoxygensaturation is common in normal infants,6 children with obstructive sleep apnoea can have self-resolving desaturation without cognitive sequelae,7 and children at home with acute respiratory diseaseprobablyhave lower-than-normal oxygen saturation levelsduring recovery.8One sizedoesnot fit all.Oneoxygensaturationmanagementstrategydoesnotsuitall clinical scenarios— even within the same disease process. Hypoxia is commonin infantswithbronchiolitis, but so too is intermittent self-resolvingbriefdesaturation: the former important, the latter probably not. McCulloh and colleagues1 try to address the latter byusing intermittentmonitoring to avoid displaying brief self-resolving desaturation events, and they suggest that there is no difference in outcomes among infants andchildrenmonitoredcontinuouslyor intermittently.But the limitations of their study constrain the application of the results—the opportunities to identify potentially important “peekaboo” clinical surprises were very limited. Recruitment and randomization occurred up to 24 hours from admission, limiting the potential for the intervention (if infants and children had already recovered to an oxygen saturation as measured by pulse oximetry [SpO2] of ≥90% by that time); we are not informed of the time at which infants and children were transitioned to intermittentmonitoring and subsequentlydischargedhome (ie,whatwas the true“timesaved,”andwas this sufficiently long toanticipateproblemsoccurring?); thestudy’s clinical exclusions were large, limiting the generalizability of the results;we arenot informedhowmanyof the intermittent group were subsequently noted to have an SpO2 of less than 90%. Importantly, too, therewasnoassessmentofparentalperspectives—did parents remain as confident with clinical decisionmaking in the intermittent group (and, if not, what could be done to address that problem?), and were staff persuaded that patient safety had not been compromised? So where do we go now with intermittent vs continuous monitoring of pulse oxygen saturation in acute bronchiolitis? More evidence is required (and may be coming [https: //clinicaltrials.gov/ct2/show/NCT01646606]), but canwemake bestuseof current evidence toplaceoxygen saturation in context and support clinical confidence? Guidelines for themanagement of bronchiolitis vary in their target oxygen saturation levels for starting and stopping supplemental oxygen,9-11 andnonehave supportingevidence, althoughanoxygen saturation level of 90% is gaining consistency as a watershed target, and evidence may soon support that target (http://www .isrctn.com/ISRCTN28405428).There isevidencethatonceoral feeding is reestablished, respiratory deterioration is uncommon.12Onthisbasis, supplementaloxygencouldbeprovided to infants with an oxygen saturation level below 90% monitoredby local choiceof continuousor intermittentmonitoring—until evidence supports one over the other (and if inRelated article page 898 Opinion