251 - Hypotension and Race Independently Impact Pulse Oximeter Accuracy in Preterm Infants

Publication Number: 4247.251

Zachary A. Vesoulis, Washington University School of Medicine, St. Louis, MO, United States; Halana Whitehead, Washington University in St. Louis School of Medicine, St. Louis, MO, United States; Daniel Sprehe, Washington University in St. Louis School of Medicine, St. Louis, MO, United States; Callie Marshall, Washington University in St. Louis School of Medicine, SAINT LOUIS, MO, United States; Stephanie Diggs, Washington University in St. Louis School of Medicine, Saint Louis, MO, United States

Background: Pulse oximetry (SpO2) monitoring is standard of care in NICUs, but racial disparities and the low perfusion state of premature infants decrease its accuracy. Blood pressure may independently influence pulse oximetry accuracy and potentially exacerbate known racial disparities.

Objective: To determine whether blood pressure status independently affects SpO2 accuracy in preterm infants and whether hypotension differentially impacts accuracy by race.

Design/Methods: VLBW infants born < 32 weeks gestation and admitted to the NICU at St. Louis Children's Hospital were screened for a retrospective study. Infants were included if they had simultaneous a) invasive arterial blood pressure data, b) pulse oximetry data, and c) arterial blood gas (ABG) samples. Data triplets compromising ABG SaO2 and the 30-second averages of mean arterial blood pressure (MABP) and SpO2 preceding ABG sampling were collected.

Hypotension was defined using MABP (in mmHg) minus PMA in weeks and categorized in the following groupings: normotensive (≥0 mmHg), moderate hypotension (-10 to 0 mmHg below expected), and severe hypotension (< -10 mmHg below expected). Pulse oximeter bias was calculated as SpO2 minus the gold standard SaO2 on ABG (where positive bias represents overestimation, negative bias underestimation). Mixed-effects models were used to examine blood pressure effects on absolute bias, controlling for SaO2 and PMA. Race-stratified analysis compared Black and White patients for interaction effects.

Results: Analysis included 6,113 triplets from 304 VLBW infants (mean GA=26.5 weeks; mean BW=1008g; 63% White, 37% Black); 7% occurred during hypotension. Hypotension was independently associated with greater absolute bias (p=0.032). Further analysis revealed Black patients had greater baseline bias (+0.45%, p=0.041), as previously reported. However, severe hypotension (<-10 mmHg below expected) paradoxically showed opposite effects by race: White patients demonstrated worsened bias (+4.63%) while Black patients showed improved accuracy (bias +0.76%), with significant interaction (p=0.003). This differential effect was specific to severe hypotension only.

Conclusion(s): Hypotension independently affects pulse oximetry accuracy in preterm infants. Severe hypotension paradoxically impacts racial bias in pulse oximetry, with White patients showing worsening accuracy with decreasing MABP, while Black patients show improving accuracy during severe hypotensive episodes. These findings suggest that complex physiological mechanisms influence SpO2 accuracy in preterm infants with a differential impact of race based on patient hemodynamics.

Table 1. Patient Characteristics


Figure 1. Mean Bias by Blood Pressure and SpO2
This heat shows the mean bias (inaccuracy) of the SpO2 as compared to gold standard SaO2 by blood pressure groups (x axis) and SpO2 level (y axis). Note that accuracy decreases as SpO2 decreases (as measured by large positive numbers). There is limited impact of hypotension on accuracy at high SpO2 but it worsens accuracy at low SpO2.

Figure 2. Interaction of BP and SpO2 on Accuracy
For white infants, accuracy of SpO2 (against SaO2 reference) worsens as SpO2 decreases. In contract, Black infants have a paradoxical increase in accuracy during the most severe hypotensive periods.