Effect of lingual gauze swab placement and light exclusion on pulse oximeter agreement with arterial haemoglobin-oxygen saturation derived from blood-gas analysis in healthy anaesthetised horses.

Research Overview

• This study investigated how placing a lingual gauze swab and using light exclusion affect the accuracy of pulse oximeter readings in measuring arterial oxygen saturation in healthy anesthetized horses.
• The main goal was to compare readings from pulse oximetry with the gold-standard arterial blood gas measurements under various conditions.

Study Purpose and Background

• Pulse oximetry (SpO) is a non-invasive method used to estimate arterial hemoglobin-oxygen saturation, but its accuracy can be influenced by various factors.
• In horses, the tongue is a common site for measuring SpO, but factors such as the presence of a gauze swab and ambient lighting might affect reading accuracy.
• Arterial blood-gas analysis (SaO) provides an accurate, but invasive, measurement of arterial oxygen saturation and is considered the gold standard.
• This study aimed to quantify how the use of a lingual gauze swab (dry or wet) and the exclusion of ambient light impact the agreement between SpO and SaO values.

Methods

• Design: Prospective method-comparison study involving 60 healthy adult horses undergoing elective general anesthesia.
• Subjects: Horses classified as ASA physical status I-II, indicating they were healthy or had mild systemic disease.
• Measurements: SpO readings were taken under six conditions during anesthesia:
  • No swab under ambient light
  • No swab with light exclusion
  • Dry gauze swab under ambient light
  • Dry gauze swab with light exclusion
  • Wet gauze swab under ambient light
  • Wet gauze swab with light exclusion
• Comparison: SpO readings were compared against SaO values measured via arterial blood-gas analysis.
• Analysis: Bland-Altman method for repeated measurements was used to assess agreement.
  • Bias (SpO minus SaO) and limits of agreement (LoA) were estimated through mixed-effects models.
  • 95% confidence intervals were obtained using cluster bootstrapping to account for repeated measures within anesthetic episodes.

Results

• Data: 154 paired SpO-SaO observations from 62 anesthetic episodes were analyzed per condition.
• Oxygen Saturation Range:
  • SaO values ranged from 94.4% to 100%.
  • Partial pressure of oxygen (PaO2) ranged from 62.6 to 595.8 mmHg (8.3 to 79.4 kPa).
• Bias:
  • SpO readings consistently underestimated SaO across all conditions.
  • Bias ranged from -1.6% to -2.7%.
  • The smallest bias (-1.6%, 95% CI -2.1 to -1.1) occurred with a dry gauze swab under ambient light.
• Limits of Agreement:
  • LoA widths ranged between 10.3% to 13.3% across all configurations.
  • Light exclusion caused a more negative bias and wider LoA than ambient light, but the differences were minor relative to overall variability.

Conclusions

• Placing a lingual gauze swab during pulse oximetry in anesthetized horses did not harm reading accuracy but also did not significantly improve agreement with arterial blood gas values.
• Pulse oximeter readings underestimated true arterial oxygen saturation and demonstrated imprecision regardless of swab use or lighting conditions.
• Excluding ambient light slightly worsened bias and precision but not to a clinically meaningful extent.
• Overall, pulse oximetry in this context is imprecise and cannot replace arterial blood gas analysis for accurate assessment of arterial oxygenation during anesthesia in healthy horses.