342 - Pulse Oximetry Low Signal Quality Within a Home Setting for Preterm Infants: Does it Matter?

Poster Number: 342
Publication Number: 342.346

Ernie Tao, Geisel School of Medicine at Dartmouth, Hanover, NH, United States; Katherine R. Edeburn, UMass Medical School, Worcester, MA, United States; Heather White, University of Massachusetts Medical School, Worcester, MA, United States; Lawrence Rhein, University of Massachusetts Medical School, Waban, MA, United States; Tyler Kenneth. Hartman, Dartmouth Hitchcock Medical Center, Lebanon, NH, United States

Background: The Recorded Home Oximetry (RHO) Program utilizes recorded oximetry uploaded from preterm infants who require home oxygen therapy (HOT) to assist in the titration of supplemental oxygen (O₂) flow rates between clinic visits. This program was widely implemented after a multi-center trial found that it was safe and effective in shortening HOT duration. The RHO algorithm describes recorded time below target thresholds of 93% and 96% oxygen saturation (SpO₂) to provide recommendations on managing infants’ O₂ flow rate. Newer models of pulse oximeters have algorithms to remove low signal quality (L-SIQ) potentially caused by motion artifacts. Prior studies have not evaluated these algorithms in home oximetry recordings and how this may impact medical decision-making dependent on SpO₂ analyses.

Objective: The purpose of this study was to evaluate how the removal of L-SIQ from recorded home oximetry may impact O₂ management recommendations for preterm infants who are on HOT.

Design/Methods: Preterm infants discharged on HOT are eligible for the RHO Program at 14 participating sites. We obtained recorded oximetry data from infants discharged between 2019 and 2022 from Dartmouth Hitchcock Medical Center and the University of Massachusetts Memorial Medical Centers. Pulse oximetry signal quality was determined via Masimo algorithms which calculate relative confidence output signals to flag potentially erroneous data (Masimo Corporation, Irvine, CA). Second-by-second oximeter recordings, perfusion index, and pulse rate were additionally recorded. RHO algorithm recommendations when including and excluding L-SIQ were then compared.


Results: 54 preterm infants’ data were eligible for analysis. On average, 6.8% of the sampled time was identified as L-SIQ.  On average, 9.3% of sampled data flagged as normal SpO₂ quality was below 96% SpO₂ compared to 34.1% of sampled data flagged as L-SIQ (p < 0.001). Of these 54 preterm infants enrolled in RHO while on HOT, 14 (25.9%) would have had different O₂ flow rate recommendations after removing L-SIQ. The recommendations changed from “maintain O₂” to “decrease O₂” for 10 (18.5%) infants. The recommendation changed from “increase O₂” to “decrease O₂” or “maintain O₂” for 4 (7.41%) infants.

Conclusion(s): A significant number of supplemental O₂ flow rate recommendations for preterm infants on HOT would have changed after the removal of sampled time flagged as L-SIQ. Oximetry assessments that do not factor in L-SIQ seem to overestimate oxygen desaturations, potentially unnecessarily prolonging the duration of HOT.