Abstract 11083: Pulse Oximetry Overestimates Arterial Oxy-Hemoglobin in Neonates With Known Critical Congenital Heart Disease (CCHD) - Implications for Newborn Screening
Background: Pulse oximetry is a key part of the clinical management of CCHD and more recently, in 2011, the United States adopted recommendations to use pulse oximetry to screen for critical congenital heart disease (CCHD). CCHD is estimated to occur in about 0.25% of live births. Recently, the accuracy of the pulse oximetry in PICU patients has been questioned, especially in hypoxemic patients.
Objective: Compare SpO2 and oxy-Hgb values in patients with known CCHD to evaluate the precision of non-invasive measurements of arterial saturations in CCHD management and screening.
Methods: Single-Center Retrospective Study. Inclusion criteria: AHA defined CCHD and a post ductal arterial blood gas in first 72 hours of life. 71 patients with 466 measurements of SpO2 (Masimo Inc., Irvine, CA) were analyzed to determine correlation with arterial oxy-hemoglobin (Siemens, Erlangen, GE). Paired T-Test and ANOVA utilized to compare data points (SPSS v21).
Results: SpO2 overestimates arterial oxy-Hgb by a mean difference of 4.6% in all patients with CCHD. The mean variance was statistically significant with all groups including those with SpO2 >95% (5.6%), SpO2 90-94% (5.7%), SpO2 80-89% (4.6%), SpO2 70-79% (4.5%), and SpO2 >=6% in 46% of all paired measurements. Only 4% of SpO2 measurements underestimated oxy-Hgb by >3%. Hour of life was not statistically significant.
Conclusion: This data set raises concern that present pulse oximeters might have a meaningful false negative rate in CCHD screening. In infants less than 72 hours old with CCHD, pulse oximetry was found to significantly overestimate oxy-Hgb with a mean variance of 4.6%. For example, Truncus Arteriosis and Hypoplastic Left Heart Syndrome both had average SpO2 readings of 94% in the first 72 hours of life while OxyHgb measured between 87.6%-88.8%. Clinical decisions are frequently made based on non-invasive pulse oximetry, which may lead to inaccurate predictions of qp/qs and inappropriate escalation in therapy. These results raise concern that present pulse oximetry algorithms may be sub-optimal in CCHD newborns.
Author Disclosures: D.P. Murphy: None. J.P. Cleary: None.
- © 2014 by American Heart Association, Inc.