Evaluation of a Wearable Device for Cardiorespiratory Monitoring in Healthy Anaesthetised Cats

ABSTRACT

Objective

The aim of this study was to evaluate the accuracy and agreement of heart rate (HR) and peripheral oxygen saturation (SpO₂) measurements obtained using the wearable Owlet Smart Sock (OSS) in anaesthetised cats, compared with a reference veterinary bedside monitor (Monitor 1).

Methods

Twenty healthy cats aged 1–2 years and weighing 3.2 and 5.1 kg were enrolled. All animals underwent a standardised general anaesthesia protocol. HR and SpO₂ were recorded simultaneously using the OSS and Monitor 1 at 1‐minute intervals for 20 min. A total of 307 paired measurements were analysed. Agreement was assessed using Bland–Altman analysis and clinical equivalence was evaluated using the two one‐sided tests (TOST) procedure, with predefined equivalence margins of ±10 bpm for HR and ±3% for SpO₂.

Results

A very strong correlation was observed between OSS and Monitor 1 for HR measurements (r = 0.995; p < 0.0001). The mean difference was 0.18 bpm (95% CI: −0.38–0.40), demonstrating clinical equivalence (p < 0.0001). For SpO₂ measurements, a statistically significant linear correlation was identified (r = 0.634; p < 0.0001). The mean difference was ‐0.5% (95% CI: −0.74 to −0.29), remaining within the predefined ±3% equivalence margin and indicating clinical equivalence (p < 0.0001).

Conclusion

In healthy cats, the OSS demonstrated high agreement with the reference monitor for HR measurements and provided clinically acceptable SpO₂ values within established equivalence limits. Further validation under diverse clinical conditions and in different patient populations is warranted.

The OSS was evaluated for heart rate and peripheral oxygen saturation monitoring in healthy anaesthetised cats. A very strong correlation was observed between OSS and a veterinary reference monitor for heart rate measurements and peripheral oxygen saturation values were clinically equivalent within predefined equivalence limits. The OSS may be considered a reliable non‐invasive tool for heart rate and peripheral oxygen saturation monitoring in healthy cats.

1. Introduction

Heart rate (HR) and peripheral oxygen saturation (SpO₂) are essential physiological parameters used in the assessment of cardiovascular and respiratory function in both human and veterinary medicine (Cugmas et al. 2019; Yanmaz et al. 2023). HR represents a principal indicator during clinical evaluation and is routinely applied in determining disease prognosis and assessing therapeutic response. SpO₂ serves as a critical measure for monitoring a patient's oxygenation status (Yanmaz et al. 2023).

In veterinary practice, the measurement of HR and SpO₂ constitutes a standard component of physical examinations, surgical procedures and intensive care management (Bednarski et al. 2011; Cugmas et al. 2019). These parameters can be obtained from various anatomical sites, including the tongue, pinnae and paws and provide valuable clinical information in both conscious and anaesthetised animals (Matthews et al. 2003; Cugmas et al. 2019; Dörfelt et al. 2021).

Hypoxaemia remains a significant cause of mortality worldwide. Monitoring SpO₂ enables early, non‐invasive detection of reductions in blood oxygen levels and the potential onset of respiratory insufficiency (Humm and Kellett‐Gregory 2016; Hasan and Negulescu 2020).

The Owlet Smart Sock (OSS) is a wearable device developed to monitor HR and SpO₂ in human neonates during sleep. The system incorporates a hospital‐grade, non‐invasive pulse oximetry sensor embedded within a soft sock designed to fit the infant's foot. Recorded data are transmitted via Bluetooth to a base station and subsequently delivered to the user through a mobile application and cloud‐based platform (Young and Flores 2020; Hasan and Negulescu 2020).

To date, no studies have reported on the accuracy or clinical agreement of the OSS for measuring HR and SpO₂ in cats. Accordingly, the objective of this study was to evaluate the accuracy and agreement of HR and SpO₂ measurements obtained using the OSS in anaesthetised cats, in comparison with a reference veterinary bedside monitor (Monitor 1).

2. Materials and Methods

2.1. Animals

The study population consisted of 20 clinically healthy mixed‐breed cats presented to Atatürk University Animal Hospital for routine dental cleaning. The animals were between 1 and 2 years of age, with body weights ranging from 3.2 to 5.1 kg.

Prior to enrolment, all cats underwent a comprehensive physical examination and routine haematological testing. Physical status was classified according to the criteria of the American Society of Anaesthesiologists (ASA), and only cats categorised as ASA I were included in the study.

The study protocol was approved by the Local Ethics Committee for Animal Experiments of Atatürk University (decision no: 2026/02). Written informed consent was obtained from all owners. Preoperatively, food was withheld for 6 h and water for 2 h before anaesthesia.

2.2. Premedication and Anaesthesia

A standardised anaesthetic protocol was implemented in all cats. Premedication consisted of tramadol (2 mg/kg, intramuscularly) and medetomidine (0.08 mg/kg, intramuscularly). Anaesthesia was induced with propofol (4 mg/kg, intravenously). Following premedication, an intravenous catheter was placed in the cephalic vein for fluid support and 0.9% NaCl solution was administered at a maintenance rate of 3 mL/kg/h throughout the procedure. Following induction with propofol and endotracheal intubation, anaesthesia was maintained with sevoflurane in 100% oxygen, with the vaporiser setting adjusted between 1.5% and 2.5% according to clinical depth.

2.3. Measurement Devices and Application

For HR and SpO₂ assessment, a veterinary patient monitor (Monitor 1; Cardell 9405, Sharn Veterinary Inc., Tampa, Florida) was used as the reference method.

HR was measured using the integrated electrocardiography (ECG) function, while SpO₂ values were obtained using the monitor's transmittance pulse oximetry (TPO) sensor positioned on the tongue. Electrodes were positioned according to the manufacturer's recommendations, with leads attached to the right forelimb, left forelimb and left hindlimb. HR values were recorded directly from the monitor display.

To verify the accuracy of Monitor 1, simultaneous manual auscultatory HR measurements were performed using a stethoscope over a 30‐s window and extrapolated to beats per minute (bpm). Respiratory rate was determined concurrently by counting visual thoracic excursions over a full 60‐s period to confirm stable respiratory status prior to data pooling.

The OSS (Owlet Smart Sock‐2, Owlet Baby Care Inc., Lehi, UT, USA) was used as the wearable monitoring device. To minimize potential optical artefacts between the sensor and the skin, hair at the tibial application site was clipped using a No.40 blade. The device was placed on the tibial region without the use of additional adhesive tape or bandaging. Its sock‐based design provided adequate fixation at the application site (Figure 1). OSS Size 1 was used in all cats because it provided the most appropriate fit for the body size and distal limb circumference of the study population, ensuring stable sensor positioning and consistent skin contact during monitoring.

FIGURE 1.

The tibial region where HR and SpO2 measurements were obtained using the OSS.

During the 20‐min data collection window, all cats were maintained in stable right lateral recumbency on a padded surgical table. The OSS was uniformly applied to the left pelvic (the non‐dependent leg) to avoid any weight‐bearing artefacts or compromised localised microcirculation associated with the dependent limb.

For the purposes of statistical analysis and graphical representation, the OSS was designated as ‘Device’, whereas the reference veterinary patient monitor (Cardell 9405) was designated as ‘Monitor 1’.

2.4. Data Collection Procedure

HR and SpO₂ measurements were recorded simultaneously using both devices at 1‐min intervals over a 20‐min period. Data acquisition was conducted concurrently by three investigators: one recorded OSS measurements, one documented values obtained from Monitor 1, one performed and recorded auscultatory HR measurements using a stethoscope.

All collected data were transferred into an electronic spreadsheet for subsequent statistical analysis. A total of 307 paired measurements were included in the final analysis.

2.5. Statistical Analysis

All statistical analyses were performed using MedCalc (v. 20.013, MedCalc Software Ltd, Ostend, Belgium). Continuous variables were expressed as mean ± standard deviation (SD) with corresponding 95% confidence intervals (CI). The assumption of normal distribution was assessed using visual inspection methods and descriptive statistics.

Agreement between devices was evaluated using Bland–Altman analysis, with calculation of the mean difference (bias) and the 95% limits of agreement (mean ±1.96 SD). The relationship between measurement differences and the magnitude of measurements was examined using regression analysis.

Clinical equivalence was assessed using the two one‐sided tests (TOST) procedure. Equivalence margins were predefined as ±10 bpm for HR, ±3% for SpO₂. A p‐value < 0.05 was considered statistically significant for all analyses.

Linear association between devices was analysed using Pearson correlation analysis, and correlation coefficients (r) together with corresponding p‐values were reported.

3. Results

A total of 20 cats were included in the study. Among the measurements obtained using the OSS and Monitor 1, 307 paired HR and SpO₂ data points were included in the statistical analysis. The remaining data points were excluded due to synchronisation discrepancies during data collection. Only complete and temporally matched measurements were included in the final analysis.

3.1. Heart Rate

A very strong and statistically significant linear correlation was identified between the OSS and Monitor 1 for HR measurements (r = 0.995, p<0.0001; Table 1, Figure 2)

TABLE 1.

Mean difference (bias), 95% limits of agreement derived from Bland–Altman analysis, and correlation coefficients for HR and SpO₂ measurements between the OSS and Monitor 1.

Parameters	Mean difference (bias)	Lower limit (%)	Upper limit (%)	Correlation (r)	p (r)
HR	0.224	−3.36	3.81	0.995	<0.0001
SpO2	−0.538	−4.78	3.7	0.634	<0.0001

FIGURE 2.

Correlation analysis of HR measurements between OSS (Device) and Monitor 1.

In the Bland–Altman analysis comparing Monitor 1 and the OSS, the mean difference was calculated as 0.224 bpm and was statistically significant (p = 0.0326). The 95% limits of agreement ranged from −3.36 to 3.81 bpm (Table 1; Figure 3).

FIGURE 3.

Bland–Altman analysis of HR measurements between the OSS and Monitor 1.

Regression analysis demonstrated a statistically significant slope (p = 0.0011).

In the TOST analysis (±10 bpm equivalence margin), the mean difference was 0.18 bpm (95% CI: −0.38–0.40), demonstrating clinical equivalence (p < 0.0001; Table 2).

TABLE 2.

Results of the TOST equivalence analysis of HR and SpO₂ measurements between the OSS and Monitor 1.

Comparison	Mean Difference	95% CI (bias)	TOST p‐value	Clinical equivalence
HR Monitor 1—Device (bpm)	0.18	−0.38–0.40	<0.0001	Equivalent
SpO2 Monitor 1—Device (%)	−0.5	−0.74 to −0.29	<0.0001	Equivalent

3.2. Peripheral Oxygen Saturation

A statistically significant linear correlation was identified between the OSS and Monitor 1 for SpO₂ measurements (r = 0.634; p < 0.0001; Table 1; Figure 4).

FIGURE 4.

Correlation analysis of SpO2 measurements between the OSS and Monitor 1, including heat map and regression curve.

In the Bland–Altman analysis, the mean differences were calculated as −0.538 bpm, with 95% limits of agreement ranging from ‐4.78 to 3.7 bpm (Table 1; Figure 5).

FIGURE 5.

Bland–Altman analysis of SpO2 measurements between the OSS and Monitor 1.

In the TOST analysis (±3% equivalence margin), the mean difference was ‐0.5% (95% CI: −0.74 to −0.29), indicating clinical equivalence (p < 0.0001; Table 2).

4. Discussion

HR and SpO₂ are fundamental parameters for assessing cardiopulmonary status in veterinary clinical practice. There is increasing interest in the continuous and non‐invasive monitoring of such vital variables (Cugmas et al. 2019; Yanmaz et al. 2023). Particularly during general anaesthesia, surveillance of these parameters is critical for the early identification of anaesthesia‐related complications (Li et al. 2015; Tamura et al. 2015; Pypendop et al. 2019; Dörfelt et al. 2022). Cats are considered to be at higher risk of anaesthesia‐associated mortality compared with dogs (Dörfelt et al. 2022). Accordingly, the evaluation of reliable monitoring systems is of considerable clinical importance.

Wearable technologies have gained increasing attention in veterinary healthcare in recent years. These systems offer potential advantages, particularly for the non‐invasive and continuous monitoring of cardiorespiratory parameters (Mitek et al. 2022). Photoplethysmography (PPG)‐based systems measure HR and SpO₂ by optically detecting variations in tissue blood volume (Cugmas et al. 2019; Hasan and Negulescu 2020). The OSS device operates according to this principle (Hasan and Negulescu 2020). However, fur and pigmented skin may influence light absorption and thereby reduce signal quality; consequently, clipping of the measurement site has been recommended (Cugmas et al. 2019). In the present study, clipping of the tibial region prior to device placement and conducting measurements under anaesthesia were intended to minimise potential optical artefacts and motion‐related signal disturbances.

In human medicine, the OSS device has been reported to provide acceptable accuracy in measuring HR and SpO₂ in neonates (Dangerfield et al. 2017). In addition, its use in the early detection of a rare case of ventricular tachycardia has been documented (Young and Flores 2020). However, no studies in the veterinary literature have specifically evaluated the use of the OSS in cats; existing publications primarily address the broader applications of wearable technologies (Mitek et al. 2022).

In the present study, a very strong linear correlation was observed between the OSS and Monitor 1 for HR measurements in healthy cats under general anaesthesia (r = 0.995; p < 0.0001). Bland–Altman analysis demonstrated a mean difference of 0.224 bpm, with 95% limits of agreement ranging from −3.36 to 3.81 bpm. The mean difference was statistically significant (p = 0.0326). In the TOST analysis, clinical equivalence was confirmed within the predefined ±10 bpm equivalence margins (p < 0.0001). These findings indicate that, under controlled anaesthetic conditions, the OSS can provide HR measurements that are consistent with those obtained from Monitor 1.

For SpO₂ measurements, a statistically significant linear correlation was identified between the two devices (r = 0.634; p < 0.0001). Bland–Altman analysis revealed a mean difference of −0.538 bpm, with 95% limits of agreement ranging from −4.78 to 3.7 bpm. The mean difference was statistically significant. In the TOST analysis, clinical equivalence was demonstrated within the predefined ±3% equivalence margins (p < 0.0001). However, the lower correlation coefficient compared with HR and the wider limits of agreement suggest greater variability in SpO₂ measurements.

In cats, cardiovascular disorders such as hypertrophic cardiomyopathy (Kitz et al. 2019; Şahin et al. 2025), congenital heart diseases (Abboud et al. 2025) and arterial thromboembolism (Kitz et al. 2019; Şahin et al. 2025), as well as respiratory conditions including feline asthma and chronic bronchitis (Chalifoux et al. 2021; Abboud et al. 2025), various forms of pneumonia (Boiron et al. 2019; Abboud et al. 2025), acute lung injury and acute respiratory distress syndrome (Balakrishnan et al. 2017), upper respiratory tract diseases (Chalifoux et al. 2021; Abboud et al. 2025) and pleural effusion (Abboud et al. 2025), may result in clinically significant alterations in HR and SpO₂. Accordingly, reliable monitoring of these parameters may facilitate early detection of clinical deterioration in cats at risk of cardiorespiratory disease. However, the findings of the present study are limited to healthy cats under general anaesthesia and should not be directly extrapolated to diseased populations.

The present study has several limitations. The sample size was limited, and only clinically healthy cats classified as ASA I were included. Measurements were performed under general anaesthesia in controlled conditions with minimal movement. Therefore, device performance may differ in conscious and active cats. The OSS was evaluated exclusively on a single anatomical site (the tibia). While the tibial region provided adequate fixation, alternative locations routinely used in veterinary monitoring (such as the metatarsus, digits or base of the tail) were not explored. These sites differ in tissue thickness and vascularity, which may influence light absorption and PPG signal quality. A further limitation is that peripheral perfusion was not objectively quantified using a perfusion index (PI). Because medetomidine may induce peripheral vasoconstriction, the potential influence of altered peripheral perfusion on OSS performance could not be specifically assessed.

5. Conclusion

The present study demonstrates that, in healthy cats under general anaesthesia, the OSS exhibits high agreement with the reference monitor for HR measurements and provides clinically acceptable SpO₂ values within predefined equivalence margins. These findings indicate that the device may represent a useful non‐invasive tool for monitoring cardiorespiratory parameters under controlled anaesthetic conditions. Nevertheless, because the study was limited to healthy cats, a single sensor placement site and controlled anaesthetic conditions, further investigations involving larger sample sizes, diverse clinical populations, alternative anatomical locations and varying perfusion states are required to confirm and extend these findings.

Author Contributions

Mümin Gökhan Şenocak: methodology, formal analysis. Şule Melek: conceptualization, methodology, supervision, writing – review and editing, writing – original draft. Ömer Tarık Orhun: conceptualization, methodology, data curation, validation. Taner Arslan: data curation.

Funding

This study was supported by the Türkiye Bilimsel ve Teknolojik Araştırma Kurumu (TÜBİTAK).

Ethics Statement

The authors confirm that the ethical policies of the journal, as noted on the journal's author guidelines page, have been adhered to and the appropriate ethical review committee approval has been received. The study protocol was approved by the Atatürk University Animal Experiments Local Ethics Committee (Approval No: 2026/02)

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The article data are available from the corresponding author upon request.