Smartphone-integrated consumer baby monitors that measure vital signs are popular among parents but are not regulated by the US Food and Drug Administration (FDA).1-4 This study measured the accuracy of pulse oximetry–based consumer baby monitors using an FDA-cleared oximeter as a reference.
We purchased the only 2 currently marketed smartphone-integrated consumer baby monitors that use pulse oximetry, the Owlet Smart Sock 2 (consumer monitor A) and Baby Vida (consumer monitor B). We enrolled infants aged 0 to 6 months hospitalized in general pediatrics and cardiology wards at the Children’s Hospital of Philadelphia from July through December 2017. Infants were excluded if born before 34 weeks’ gestation, critically ill, anemic (hemoglobin <10 g/dL), febrile (≥38.0°C), hypothermic (<36.0°C), hypotensive (systolic blood pressure <60 mm Hg if 0-28 days old or <70 mm Hg if 29 days–6 months old), or had compromised perfusion. Written informed consent was obtained from infants’ parents. The Children’s Hospital of Philadelphia’s institutional review board approved the study.
On 1 foot, infants were monitored using a Masimo Radical-7 with 16-second averaging (reference monitor). Each consumer monitor was applied to the other foot of all infants in a random sequence for 60 minutes while asleep or awake and calm.
We identified “stable” paired reference monitor–consumer monitor points that met criteria: (1) for the reference monitor, during the prior 30 seconds, oxygen saturation (Spo2) varied by 1 percentage point or less in either direction for Spo2 comparisons or pulse rate varied by 5 beats/min or less in either direction for pulse rate comparisons and (2) for the consumer monitor, during the prior 30 seconds, no dropout (defined as failure to display a value). To minimize ascertainment bias, we randomly selected up to 10 stable points per patient for analysis. We generated scatterplots and calculated sensitivity and specificity for hypoxemia (Spo2<91%)5 and bradycardia (pulse rate <90 beats/min),6 accounting for clustering within patients. We used R (R Foundation), version 3.3.3, and Stata (StataCorp), version 15.1, for analysis.
Of the 30 infants (50% female; 33% black; median birth at 39 weeks’ gestation [interquartile range {IQR}, 38-40]; median age, 50 days [IQR, 26-90]), the most common diagnoses were bronchiolitis (27%), apnea or brief resolved unexplained event (10%), hypoplastic left heart syndrome (10%), and double outlet right ventricle (10%).
We recorded 2466 stable Spo2 and 1801 stable pulse rate points. Using monitor A, 5 patients had fewer than 10 stable Spo2 points and 10 patients had fewer than 10 stable pulse rate points. Using monitor B, 3 patients had fewer than 10 stable pulse rate points.
Figure 1 and Figure 2 display scatterplots for Spo2 and pulse rate. During testing of monitor A, 12 patients experienced hypoxemia according to the reference monitor and all 12 had at least 1 simultaneous hypoxemia reading on monitor A, although 5 of the 12 each had at least 1 stable normoxemic reading on monitor A during hypoxemia. During testing of monitor B, 14 patients experienced hypoxemia according to the reference monitor, but none had simultaneous hypoxemia readings on monitor B. All Spo2 readings on monitor B were in the normal range. Monitor A had 0 instances of falsely displaying bradycardic pulse rates when the reference monitor rate was normal. However, in 14 patients, monitor B falsely displayed bradycardic pulse rates when the reference monitor rate was normal.
For monitor A, the sensitivity and specificity for hypoxemia were 88.8% and 85.7%, respectively, and for bradycardia were 0.0% and 100.0%, respectively.
For monitor B, the sensitivity and specificity for hypoxemia were 0.0% and 100.0%, respectively, and for bradycardia were 0.0% and 82.3%, respectively.
Accuracy testing of 2 Spo2-based baby monitors that are not FDA-regulated revealed concerning findings. Monitor A detected hypoxemia but performed inconsistently. Monitor B never detected hypoxemia and also displayed falsely low pulse rates. Beyond their accuracy, other concerns about consumer monitor use include the lack of medical indications for monitoring infants at home, the absence of FDA oversight, and the potential for unintended consequences.1 The main study limitation is the use of a pulse oximeter as the reference standard; arterial blood gas measurements would be preferred but limit feasibility.
As more neonate and infant vital sign monitors emerge in this largely unregulated market, physicians and parents should exercise caution incorporating data from these monitors into medical decisions.
Accepted for Publication: June 6, 2012.
Corresponding Author: Christopher P. Bonafide, MD, MSCE, Children's Hospital of Philadelphia, 3401 Civic Center Blvd, Room 12NW80, Philadelphia, PA 19104 (bonafide@email.chop.edu).
Author Contributions: Dr Bonafide had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. The analysis was performed by Drs Orenstein and Bonafide with oversight from Dr Localio.
Concept and design: All authors.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Bonafide, Localio, Orenstein.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Bonafide, Localio, Ferro, Orenstein, Lavanchy.
Obtained funding: Bonafide.
Administrative, technical, or material support: Orenstein, Jamison, Lavanchy, Foglia.
Supervision: Bonafide, Ferro.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Foglia reports grants from Zoll Foundation, outside the submitted work. No other disclosures were reported.
Funding/Support: This work was funded by the Children’s Hospital of Philadelphia Foerderer Award for Excellence (Dr Bonafide) and the Division of General Pediatrics.
Role of the Funder/Sponsor: The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: We thank Caroline Boyle and Shannon Stemler, RN, BSN (Children’s Hospital of Philadelphia), for their assistance with data collection, for which they received compensation.
2.Dangerfield
MI, Ward
K, Davidson
L, Adamian
M. Initial experience and usage patterns with the Owlet Smart Sock monitor in 47 495 newborns [published online December 4, 2017].
Glob Pediatr Health. doi:
10.1177/2333794X17742751Google Scholar 5.Ralston
SL, Lieberthal
AS, Meissner
HC,
et al; American Academy of Pediatrics. Clinical practice guideline: the diagnosis, management, and prevention of bronchiolitis.
Pediatrics. 2014;134(5):e1474-e1502. doi:
10.1542/peds.2014-2742PubMedGoogle ScholarCrossref 6.American Heart Association. Systematic approach to the seriously ill or injured child. In: Pediatric Advanced Life Support Provider Manual. Dallas, TX: American Heart Association; 2016:48.