Evaluation of the performance of three pulse oximeters at different probe positions in awake rabbits

Abstract

In rabbits, pulse oximetry plays a crucial role in monitoring oxygen saturation (SpO2) and pulse rate (PR), but it can be challenging due to their small size, variable anatomy, limited measurement sites, and potential motion artifacts during measurements. This prospective randomized study aimed to assess the reliability of three pulse oximeters on the forelimb and tail of fifty awake, healthy European rabbits. Two pulse oximeters, the Masimo Rad-5 (device 1) and the Edan VE-H100B, were tested, with the latter using both a Y-clip (device 2) and the Nellcor OxiMax adhesive finger sensor (device 3). Reliable values of SpO2 and PR were considered ≥ 95% and ≥ 151 bpm, respectively, which are normal values in a healthy rabbit. Success rates of the devices were calculated as the ratio of reliable to unreliable/missing values and compared using χ2 tests. Correlation and agreement between PR and clinically measured heart rate (cHR) were assessed using the Pearson correlation coefficient (r) and Bland-Altman analysis, respectively. Success rates differed significantly among devices (p < 0.001). The SpO2 success rates were 46%, 95.6%, and 94.4% for devices 1, 2, and 3, respectively. The PR success rates were 54.6%, 63%, and 72.8% for devices 1, 2, and 3, respectively. Success rates were higher on the forelimb (>66%) than on the tail (>26%) across devices. Correlation between PR and cHR was stronger on forelimb (r > 0.80) than on tail (r < 0.70). Agreement between PR and cHR was similar among devices, with a mean bias ranging from -6.6 to 6.7 bpm, and limits of agreement from -33.8 to 41.5 bpm. The Edan devices showed superior ability to detect reliable values of SpO2 and PR compared to the Masimo in rabbits. The forelimb appears to be a more reliable site for pulse oximetry in rabbits than the tail.

Fig 1. The three pulse oximeters tested in the study on rabbits, applied to the forelimb and tail.

(A) Masimo Rad-5 with a Y-clip sensor (device 1), (B) Edan VE-H100B with a Y-clip sensor (device 2), and (C) Edan VE-H100B with a Nellcor OxiMax adhesive finger sensor (device 3).

Fig 2. Application of pulse oximeter probes to the forelimb and tail of a rabbit.

(A) Masimo Rad-5 Y-clip sensor (device 1) applied to the forelimb, (B) Edan VE-H100B Y-clip sensor (device 2) applied to the forelimb, (C) Edan VE-H100B Nellcor OxiMax adhesive finger sensor (device 3) applied to the forelimb, (D) Masimo Rad-5 Y-clip sensor (device 1) applied to the tail, and (E) Edan VE-H100B Nellcor OxiMax adhesive finger sensor (device 3) to the tail.

Fig 3. Limits of agreement (Bland–Altman) plot showing differences between mean clinical heart rate (mcHR) and reliable pulse rate (PR) values in 50 rabbits.

Differences are shown using the Masimo Rad-5 with a Y-clip (device 1) on the forelimb (A) and on the tail (B); using the Edan VE-H100B with a Y-clip (device 2) on the forelimb (C) and on the tail (D); and using the Edan VE-H100B with Nellcor OxiMax adhesive finger sensor (device 3), on the forelimb (E) and on the tail (F).