Consumer Baby Hypoxemia Monitors Unreliable, Study Finds

Troy Brown, RN

August 21, 2018

Researchers tested two consumer baby monitors designed to detect hypoxemia in sleeping infants. One monitor inconsistently detected low oxygen levels when they occurred in hospitalized infants, whereas the other monitor missed every episode of low oxygen detected by a US Food and Drug Administration (FDA)–approved monitor. The monitor that failed to detect episodes of low oxygen also falsely indicated infants were experiencing low pulse rates, a study found.

Christopher P. Bonafide, MD, MSCE, from the Division of General Pediatrics, Children's Hospital of Philadelphia, Pennsylvania, and colleagues tested the only two smartphone-integrated consumer baby monitors that use pulse oximetry currently on the market: the Owlet Smart Sock 2 (consumer monitor A) and Baby Vida (consumer monitor B). They describe their findings in an article published online today in JAMA.

The monitors are being marketed as a way to help parents feel confident that their infant is safe, yet there no data support that claim. "We don't have any evidence that suggests that home monitoring is helpful in a SIDS [sudden infant death syndrome][ scenario or in a near-SIDS scenario," Bonafide told Medscape Medical News. "Another thing that I worry about a little is that these monitors can be used in situations where parents have chosen to make decisions about sleep that are not particularly safe."

The study included 30 infants aged 6 months and younger hospitalized on cardiology and general pediatrics units at Children's Hospital of Philadelphia. The most frequent diagnoses were bronchiolitis (27%), apnea or brief resolved unexplained event (10%), hypoplastic left heart syndrome (10%), and double outlet right ventricle (10%). Half the children were female and one third were black. The median gestational age at birth was 39 weeks.

"[We tested this in a range of babies, from babies who were well and were in for say a fever or bronchiolitis…all the way up to babies who had congenital heart disease so had very abnormally low oxygen saturations," Bonafide explained.

Each infant wore a consumer monitor on one foot and an FDA-approved reference monitor (Radical-7, Masimo) on the other foot.

While monitor A was being tested, the reference monitor detected hypoxemia in 12 patients. Monitor A detected one or more simultaneous hypoxemia readings for each of the affected patients; however, five of those patients had at least one stable normal reading on monitor A during hypoxemia episodes.

"With oxygen saturation, we did see some inconsistencies where the monitor would not consistently display low readings at times of low oxygen levels," Bonafide said.

At no time did monitor A erroneously display bradycardia when the rate was normal on the reference monitor.

For monitor A, the sensitivity and specificity were 88.8% and 85.7%, respectively, for hypoxemia, and 0.0% and 100.0%, respectively, for bradycardia.

During testing of  monitor B, 14 babies had hypoxemia according to the reference monitor; however, the consumer monitor did not show simultaneous hypoxemia readings for any of the events. In fact, all oxygen saturation (SpO2) readings on that consumer monitor were within normal limits.

"[W]hat we found for the oxygen saturation for the Baby Vida monitor was very worrisome," Bonafide told Medscape Medical News. "Regardless of what the baby's oxygen saturation was…the Baby Vida monitor never ever detected an oxygen saturation that was below about 92% or 91%. It always thought the oxygen saturation was normal…even in the face of pretty profoundly low oxygen saturations."

He continued, "That's very worrisome to me because if a parent is at home and has a baby who is truly quite sick and has a low oxygen saturation from something like pneumonia or bronchiolitis…that monitor could falsely reassure a parent. Even though they may be worried about the way the baby looks or the way the baby's breathing, they may be falsely reassured by a number that suggests their baby's oxygen levels are okay when they might actually be in the 80s or even in the 70s. That was a very worrisome surprise that we found in the data," Bonafide explained.

Moreover, 14 infants using monitor B falsely displayed bradycardic pulse rates when pulse rates were normal on the reference monitor.

"[T]he Baby Vida thought that the pulse rate was dangerously low at times when it was actually just perfectly fine. So we could imagine a scenario where a parent might see this reading of a pulse rate of 55 in a baby who otherwise seems to be doing okay and maybe they would call their pediatrician for that. I think most pediatricians with a parent calling about a low pulse rate would refer that baby for medical attention and might even refer them to the emergency room or even advise the parents to call an ambulance," Bonafide said.

For monitor B, the sensitivity for hypoxemia was 0.0% and the specificity was 100.0%, respectively. The sensitivity for bradycardia was 0.0% and the specificity was 82.3%.

Neither company has provided accuracy data, according to Bonafide. "Owlet published one paper but it was really on the experiences of parents with the monitors and not reporting any accuracy data," he added.

The study's main limitation is that it used a pulse oximeter as the reference standard when arterial blood gas measurements would be preferable; however, use of arterials blood gas measurements would "limit feasibility," the authors write.

When reached for comment, Kurt Workman, co-founder and CEO of Owlet, said "Owlet sensor accuracy has been validated in studies performed by independent laboratories and when compared to arterial blood gas measurements the sensor performed well within international standards for pulse oximetry. The Owlet accuracy study on ClinicalTrials.gov can be found here."  

Workman continued, "Owlet has also tested SpO2 accuracy in babies between 0 and 17 months old against Massimo Radical-7 with both sensors on one foot and it reached desired industry accuracy standards and compared favorably to Massimo Radical-7. The accuracy and performance of the Owlet Smart Sock is something that we take very seriously."

Bonafide told Medscape Medical News that the results of such a test should be published. "Ours is the first peer-reviewed scientific publication to report the accuracy of the Owlet.… We reviewed the ClinicalTrials.gov entry referenced by Owlet's CEO in response to our paper. It appeared to be a study of Owlet's accuracy in 11 adults ages 18 years old to 50 years old, and there were no accuracy results included in the registry record," Bonafide explained.

"In order to provide evidence to parents and pediatricians to support their accuracy claims, we would encourage Owlet to publish their accuracy data using infants in a peer-reviewed scientific journal and pursue FDA clearance for the consumer version of the Owlet," Bonafide added.

Medscape Medical News reached out to the makers of the Baby Vida monitor but did not hear back from the company before publication.

What to Tell Parents?

Bonafide advises clinicians to ask parents if they plan to use one of these monitors.

For parents who may be considering using one of the monitors, clinicians should recognize the anxiety they feel and listen to their concerns, Bonafide said. Clinicians should tell parents, "[W]e do have evidence that there is a potential for this to actually compound things and make the anxiety worse, if we're getting either false reassurance or false alarms that are interrupting sleep and leading folks to seek unnecessary care."

When communicating with parents who are already using these monitors, Bonafide suggests exploring "What if?" scenarios. For example, "[W]hat should they do if this alarm goes off in the middle of the night for a low oxygen level or for a low heart rate? What are the ways that they can potentially verify and figure out in a quick way if the baby is okay or if the baby is not okay, and who they call?" he explained.

The authors have disclosed no relevant financial relationships.

JAMA. Published online August 21, 2018. Abstract

For more news, join us on Facebook and Twitter

Comments

3090D553-9492-4563-8681-AD288FA52ACE
Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.

processing....