To determine the reliability of peripheral oxygen haemoglobin saturation (SpO2), measured by a Nonin PalmSAT 2500A pulse oximeter with 2000T transflectance probes
at four attachment sites (third eyelid, cheek, rectum and tail), by comparing these
measurements to arterial oxygen haemoglobin saturation (SaO2), measured by an AVOXimeter 4000 co-oximeter reference method in immobilized white
rhinoceros (Ceratotherium simum).
Randomized crossover study.
A convenience sample of eight wild-caught male white rhinoceros.
White rhinoceros were immobilized with etorphine (0.0026 ± 0.0002 mg kg–1, mean ± standard deviation) intramuscularly, after which the pinna was aseptically
prepared for arterial blood sample collection, and four pulse oximeters with transflectance
probes were fixed securely to their attachment sites (third eyelid, cheek, rectum
and tail). At 30 minutes following recumbency resulting from etorphine administration,
the animals were given either butorphanol (0.026 ± 0.0001 mg kg–1) or an equivalent volume of saline intravenously. At 60 minutes following recumbency,
insufflated oxygen (15 L minute–1 flow rate) was provided intranasally. In total, the SpO2 paired measurements from the third eyelid (n = 80), cheek (n = 67), rectum (n = 59) and tail (n = 76) were compared with near-simultaneous SaO2 measurements using Bland-Altman to assess bias (accuracy), precision, and the area
root mean squares (ARMS) method.
Compared with SaO2, SpO2 measurements from the third eyelid were reliable (i.e., accurate and precise) above
an SaO2 range of 70% (bias = 1, precision = 3, ARMS = 3). However, SpO2 measurements from the cheek, rectum and tail were unreliable (i.e., inaccurate or
Conclusions and clinical relevance
A Nonin PalmSAT pulse oximeter with a transflectance probe inserted into the space
between the third eyelid and the sclera provided reliable SpO2 measurements when SaO2 was > 70%, in immobilized white rhinoceros.