Pulse Oximetry in Veterinary Patients: Limitations and Troubleshooting

By Dr. Zubair Khalid, DVM, MS, PhD · 2026-08-11

Key Takeaways

• Pulse oximetry relies on a two-wavelength (typically 660/900 nm) absorption ratio of pulsatile arterial blood, assuming only oxyhemoglobin and deoxyhemoglobin are present; deviations occur when dyshemoglobins like carboxyhemoglobin (COHb) or methemoglobin (MetHb) are present.
• Carboxyhemoglobin causes pulse oximetry to overestimate true oxyhemoglobin saturation by summing COHb and oxyhemoglobin absorption, potentially masking severe hypoxia in smoke inhalation or carbon monoxide exposure cases.
• Methemoglobinemia leads to a plateaued SpO2 reading around 84-86% as MetHb concentration increases, rendering the device unresponsive to further desaturation and requiring co-oximetry for accurate assessment.
• Signal quality is paramount; a reliable plethysmographic waveform and adequate perfusion index are essential before trusting a numeric SpO2 value, with low amplitude indicating poor perfusion or motion artifact.
• In hyperoxemic patients, pulse oximetry exhibits significant detection latency (up to 180 seconds) for airway obstruction, making capnography the superior early warning system for ventilation failure.
• Conventional 660/900 nm sensors degrade in accuracy below approximately 70% SpO2, with 735/890 nm sensors offering improved low-saturation performance, necessitating confirmation with arterial blood gas analysis for critical low readings.

Pulse oximetry is the most frequently applied continuous monitor of arterial oxygenation in small animal practice, yet its interpretation demands more than reading a displayed percentage. This article addresses the practicing veterinarian who uses pulse oximetry in anesthetic, emergency, and critical care settings and who needs to distinguish a reliable SpO2 value from an artifact or a physiologically misleading one. The focus is deliberately narrow: sensor placement, signal quality assessment, recognition of dyshemoglobin interference, and systematic troubleshooting. Other methods of oxygen monitoring, including blood gas analysis, co-oximetry, and transcutaneous techniques, are discussed only where they clarify the limitations of pulse oximetry itself.

The clinical question this reference answers is practical: when the pulse oximeter reads a value, what does that value actually represent, and when should it be disregarded? The answer requires understanding the device's two-wavelength design, its calibration assumptions, and the optical behavior of hemoglobin species that are not oxyhemoglobin. The sections that follow build the scientific foundation first, then apply it to common failure modes and corrective actions.

At a Glance

Principles of Pulse Oximetry

Pulse oximetry measures the ratio of pulsatile light absorption at two wavelengths, conventionally red (approximately 660 nm) and infrared (approximately 900 nm). The device isolates the arterial component by subtracting the constant absorption of tissue, venous blood, and capillary blood from the pulsatile signal. The resulting ratio is mapped to saturation through an empirically derived calibration curve. The calibration assumes that only two hemoglobin species are present in significant concentration: oxyhemoglobin and deoxyhemoglobin.

This two-species assumption is the root of most interpretive errors. When a third hemoglobin species is present, the device cannot distinguish it and reports a value that reflects the combined optical absorption instead of the true fractional oxyhemoglobin saturation. The magnitude and direction of the error depend on which dyshemoglobin is present and at what concentration.

Wavelength Selection and Calibration Limits

The accuracy of pulse oximetry is not uniform across the saturation range. Conventional sensors using 660 and 900 nm emitters perform well at high saturation but show progressive inaccuracy below approximately 70% SpO2. Numerical modeling and animal testing demonstrate that sensors using 735 and 890 nm emitters maintain more stable accuracy at low saturation because the fractional change in photon path length is better matched between the two wavelengths under conditions of tissue perturbation. For the clinician, this means that a reading of 65% from a standard sensor should be treated as an approximation with wide confidence limits, not a precise measurement.

Dyshemoglobin Interference

Carboxyhemoglobin

Carbon monoxide binds hemoglobin with an affinity far exceeding that of oxygen, and the resulting carboxyhemoglobin absorbs light at the red wavelength in a manner similar to oxyhemoglobin. The pulse oximeter therefore reads COHb as if it were O2Hb. In a canine model where COHb was varied from 0% to 75%, pulse oximeters continued to display saturation values above 90% while true oxyhemoglobin fell below 30%. The displayed SpO2 approximated the sum of COHb and O2Hb.

Clinical confirmation comes from a retrospective series of 30 patients with severe carbon monoxide poisoning and COHb levels above 25%. SpO2 consistently overestimated fractional arterial oxygen saturation, and the magnitude of the error increased with rising COHb levels. The practical implication for veterinary patients is direct: any animal with suspected smoke inhalation, exhaust exposure, or house fire involvement may have significant COHb, and a normal or high SpO2 reading cannot exclude profound oxygen content deficiency. Transcutaneous PO2, by contrast, falls linearly as COHb increases and reflects the true oxygen delivery impairment.

Methemoglobin

Methemoglobin results from oxidation of the heme iron to the ferric state, which cannot carry oxygen. Its optical absorption at both pulse oximetry wavelengths creates a distinctive error pattern. In a canine model with progressively induced methemoglobinemia, SpO2 overestimated fractional saturation by an amount proportional to MetHb concentration until MetHb reached approximately 35%. Beyond that level, SpO2 plateaued at 84% to 86% and did not decrease further, regardless of additional MetHb accumulation.

The plateau creates a clinical trap. A patient with severe methemoglobinemia may display an SpO2 in the mid-80s that appears stable and reassuring, while true oxygen carrying capacity is critically reduced. Furthermore, when additional desaturation was induced at fixed MetHb levels, SpO2 changed by only 16% to 32% of the change in true saturation. The pulse oximeter becomes progressively less responsive to genuine hypoxemia as MetHb rises. Suspected methemoglobinemia requires co-oximetry for confirmation, and the visual appearance of chocolate-brown blood should prompt this diagnostic even when SpO2 appears only moderately depressed.

Detection Latency in Airway Emergencies

Pulse oximetry is a downstream monitor. It detects the consequence of hypoxemia, not the cause, and its response time depends on the oxygen reserve in the lungs and blood. In a hyperoxemic porcine model with initial PaO2 above 400 mm Hg, complete airway obstruction produced no change in SpO2, heart rate, or systolic blood pressure during the full 180-second observation period. Capnography, by contrast, detected complete obstruction within 6 to 8 seconds.

This finding has direct anesthetic relevance. A patient breathing a high inspired oxygen fraction has a substantial oxygen reservoir, and SpO2 will remain normal for several minutes after an endotracheal tube becomes kinked, obstructed, or dislodged into the hypopharynx. The pulse oximeter cannot serve as an early warning device for airway compromise in the hyperoxemic patient. Capnography, where available, provides the earliest detection of ventilation failure, and its absence should be recognized as a monitoring gap instead of compensated by more frequent pulse oximetry checks.

Sensor Selection and Placement

The choice of sensor type and application site materially affects signal quality and accuracy. Clip sensors designed for human fingers often fit poorly on veterinary patients. Wrap sensors with adhesive or hook-and-loop fasteners conform better to tapered limbs, tails, and ears. Reflectance sensors, which place emitter and detector on the same surface, suit flat application sites such as the buccal mucosa, tongue, or rectal mucosa in species where transmissive sensors cannot be secured.

Perfusion determines which site yields a usable waveform. The tongue is the most reliable site in anesthetized dogs and cats because it is well perfused and easily accessible. The pinna works in many dogs but is less reliable in cats with small or cold ears. The prepuce, vulva, lip, and rectal mucosa are alternatives when limb or ear sites fail, though mucosal readings can lag behind arterial desaturation during rapid changes. In birds, the wing web, base of the tibiotarsus, and the cere have been used, but published validation data are sparse and the clinician should interpret readings with caution.

Secure the sensor away from direct surgical fields, tourniquets, and blood pressure cuffs. Ambient light, particularly from surgical lamps and heat lamps, can overwhelm the photodetector. Cover the sensor with an opaque drape when interference is suspected. Motion artifact from shivering, panting, or surgical manipulation remains the most common cause of spurious readings in awake patients.

Recognizing Artifact and Signal Failure

A pulse oximeter reports a value only when it detects a pulsatile signal. The plethysmographic waveform, not the numeric display, is the primary indicator of signal quality. A smooth, consistent waveform with a clear dicrotic notch suggests reliable measurement. A low-amplitude, irregular, or flat waveform indicates poor perfusion or motion artifact, and the displayed SpO2 should be disregarded until the signal improves.

The perfusion index, available on many modern devices, quantifies the pulsatile component of the signal. A falling perfusion index can precede loss of the waveform and may reflect deteriorating cardiac output, vasoconstriction, or hypovolemia. When the perfusion index declines, reassess the patient instead of repositioning the sensor alone.

Common failure modes include:

• Low signal amplitude: hypothermia, hypotension, vasoconstriction, cardiac arrest. Warm the site, improve perfusion, or select a more central site.
• Motion artifact: shivering, panting, seizures, patient movement. Use a site with less motion, increase averaging time if the device allows, or switch to a reflectance sensor on the tongue.
• Optical interference: bright surgical lights, infrared heating lamps, or fluorescent lighting. Cover the sensor with opaque material.
• Venous pulsation: occurs with tight sensor application, venous congestion, or right-sided heart failure. The waveform appears broad and the SpO2 reads artificially low. Loosen the sensor or change sites.
• Sensor malposition: partial detachment or folding of the sensor creates a short optical path that reads low. Reapply the sensor.

When the signal is lost entirely, the device typically displays a default value or a warning. Do not chart a numeric value without a visible plethysmographic waveform.

Troubleshooting Flowchart

The following sequence applies when SpO2 is absent, unstable, or inconsistent with the clinical picture.

1. Check the patient first. Assess mucous membrane color, pulse quality, heart rate, and respiratory effort. A patient that appears well perfused and pink with a low SpO2 reading warrants equipment scrutiny. A patient that is tachycardic, pale, or dyspneic warrants immediate oxygen supplementation and further assessment regardless of the monitor.
2. Inspect the sensor and connection. Verify that the sensor is properly seated on the site, the cable is connected, and the battery is adequate. Reapply if needed.
3. Evaluate the waveform. If the plethysmographic trace is absent or poor, improve perfusion or change sites. If the trace is present but irregular, suspect motion artifact.
4. Eliminate optical interference. Cover the sensor and dim nearby lights.
5. Compare with clinical assessment. If the SpO2 does not match the patient's clinical status, obtain an arterial blood gas sample for co-oximetry. This is the definitive step when dyshemoglobins are suspected or when the reading remains questionable.
6. Consider dyshemoglobinemia. A patient with known smoke inhalation, recent nitrate or benzocaine exposure, or suspected methemoglobinemia may have a falsely elevated or plateaued SpO2. Co-oximetry is required for accurate oxygen saturation measurement in these cases.

Interpreting SpO2 Values

Normal SpO2 in healthy dogs and cats breathing room air is 95% to 100% at sea level. Values below 95% warrant investigation, and values below 90% indicate significant hypoxemia requiring intervention. In patients receiving supplemental oxygen, the target SpO2 depends on the underlying disease and the fraction of inspired oxygen. For most critically ill patients, maintaining SpO2 above 94% is reasonable while avoiding hyperoxia where oxygen toxicity is a concern.

The pulse oximeter measures functional oxygen saturation, which is the ratio of oxyhemoglobin to the sum of oxyhemoglobin and deoxyhemoglobin. It does not detect carboxyhemoglobin or methemoglobin. In the presence of carbon monoxide, SpO2 approximates the sum of oxyhemoglobin and carboxyhemoglobin, so a patient with 30% COHb and 70% O2Hb may display an SpO2 near 100% despite profound tissue hypoxia. This was demonstrated in an experimental model where pulse oximeters continued to read above 90% while actual oxyhemoglobin fell below 30% during progressive carbon monoxide exposure. Similarly, methemoglobin causes SpO2 to overestimate true oxygen saturation, with readings plateauing near 84% to 86% as methemoglobin levels rise, regardless of further desaturation.

Species differences affect normal values and clinical interpretation. Horses and ruminants have higher resting oxygen affinity in some contexts, but pulse oximetry thresholds for intervention remain similar across mammals. In birds, normal SpO2 values are less well established, and the clinician should correlate readings with blood gas analysis when available.

Documentation and Monitoring Strategy

Record the SpO2 value, the site used, the presence or absence of a plethysmographic waveform, and the fraction of inspired oxygen at the time of measurement. Note any interventions performed to obtain the reading, such as site changes or warming. In anesthetic records, chart SpO2 at intervals appropriate to the patient's stability, typically every 5 minutes during stable anesthesia and continuously during induction, recovery, or deterioration.

When SpO2 and arterial blood gas results disagree, the blood gas with co-oximetry is the reference standard. Document the discrepancy and the corrective action taken. Serial trends are more informative than single readings. A gradual decline in SpO2 with a stable waveform suggests progressive respiratory deterioration, while an abrupt drop with a poor waveform more often indicates sensor or perfusion problems.

For patients with suspected dyshemoglobinemia, pulse oximetry should not be used as the sole monitor of oxygenation. The RECOVER cardiopulmonary resuscitation guidelines emphasize that during cardiac arrest, pulse oximetry is unreliable because pulsatile flow is absent, and capnography or blood gas analysis should guide resuscitation efforts. In the post-arrest period, pulse oximetry becomes useful again once spontaneous circulation is restored, but readings should be confirmed against clinical assessment and blood gas values.

The MSD Veterinary Manual provides species-specific reference ranges and monitoring guidance that can supplement the thresholds presented here, particularly for exotic and production animal species where published data are limited.

Recognized Complications and Early Detection

Pulse oximetry failure modes fall into three categories: signal acquisition failure, inaccurate reading despite adequate signal, and delayed detection of physiologic change. Each requires a different monitoring response.

Signal acquisition failure is the most common and least dangerous. Motion artifact, ambient light interference, and poor sensor contact produce either no waveform or an erratic one. Most modern devices display a low-perfusion or weak-signal message before reporting a value. The correct response is to reposition the sensor, select a different site, or switch to a reflectance probe for a haired or pigmented patient. Do not accept a displayed value without a consistent plethysmographic waveform.

Inaccurate readings with adequate signal are more insidious. Dyshemoglobin interference is the classic example. Carboxyhemoglobin absorbs light at the 660 nm wavelength used by conventional sensors, causing the device to read the sum of oxyhemoglobin and carboxyhemoglobin instead of true oxyhemoglobin saturation. In an experimental model, pulse oximeters continued to display saturation above 90% while actual oxyhemoglobin fell below 30% as carboxyhemoglobin rose to 75%. The same principle applies to methemoglobin, which drives readings toward a plateau of 84% to 86% regardless of true saturation. When smoke inhalation, benzocaine exposure, or oxidant drug administration is suspected, co-oximetry on an arterial blood sample is the only reliable measurement.

Delayed detection of airway compromise is a limitation of the method itself. In a hyperoxemic porcine model of complete airway obstruction, heart rate, blood pressure, and pulse oximetry saturation did not change during the 180 second observation period, while capnography detected obstruction within seconds. Pulse oximetry measures the result of inadequate ventilation, not the event itself. In a patient breathing a high inspired oxygen fraction, the oxygen reservoir in the functional residual capacity masks desaturation for a clinically significant interval. Capnography should be used whenever airway patency or ventilation is the primary concern.

Common Errors and Corrective Actions

Less experienced clinicians frequently misinterpret a normal SpO2 as evidence of adequate ventilation. The value reflects oxygenation only. A patient with profound hypoventilation and supplemental oxygen can maintain a normal SpO2 while PaCO2 rises to dangerous levels. Ventilation must be assessed by capnography or blood gas analysis.

A second common error is treating the number instead of the waveform. A low SpO2 with a robust, regular waveform in a patient with known lung disease warrants oxygen therapy and reassessment. A low SpO2 with a dampened or irregular waveform is more likely artifact, and the sensor should be moved before therapy is escalated. Conversely, a normal SpO2 with a poor waveform should not reassure the clinician.

A third error is failing to account for the calibration limits of the device. Conventional sensors using 660 and 900 nm emitters lose accuracy below 70% saturation. Sensors using 735 and 890 nm emitters perform better at low saturation, but few practices stock them. When the displayed value falls below 70%, treat it as an approximation and confirm with arterial blood gas analysis if the clinical decision depends on the exact value.

Limitations of the Evidence

The dyshemoglobin interference data derive largely from experimental animal models and human clinical studies. The porcine and canine models establish the direction and magnitude of error, but species differences in tissue perfusion and hemoglobin absorption may alter the exact values in clinical patients. The RECOVER guidelines provide evidence-evaluated consensus for CPR monitoring, but they do not address pulse oximetry accuracy across the full range of veterinary species.

Expert opinion still differs on the threshold for confirming a low SpO2 with blood gas analysis. Some clinicians accept a stable waveform and a value above 90% without confirmation. Others recommend co-oximetry for any value below 94% in a patient with suspected dyshemoglobinemia. The safest approach is to confirm with laboratory measurement whenever the SpO2 does not match the clinical picture, when the value will change a treatment decision, or when a dyshemoglobin is suspected.

Referral and Escalation Criteria

Referral or specialist consultation is warranted when pulse oximetry cannot be stabilized despite appropriate sensor selection and placement, when the patient requires continuous monitoring beyond the capacity of the practice, or when the underlying condition requires advanced oxygen delivery systems. Patients with suspected carbon monoxide or methemoglobinemia should have laboratory co-oximetry performed before treatment decisions are finalised, and referral to a facility with hyperbaric oxygen capability may be indicated for severe carbon monoxide poisoning.

Laboratory involvement is required whenever co-oximetry is needed, when arterial blood gas analysis is necessary to assess ventilation, or when the pulse oximetry reading conflicts with clinical assessment. The MSD Veterinary Manual provides species-specific reference ranges for blood gas values that support interpretation.

Regulatory reporting obligations vary by jurisdiction. The World Organization for Animal Health terrestrial animal health standards address disease surveillance and reporting obligations that may apply when monitoring reveals a notifiable condition. The American Veterinary Medical Association practice resources describe professional standards for monitoring and documentation that apply across clinical settings. Clinicians should consult their regional veterinary board for specific reporting requirements.

Frequently Asked Questions

What should I do when a pulse oximeter is unavailable or fails repeatedly?

When pulse oximetry is unavailable or unreliable, shift to clinical assessment and alternative monitoring. Serial mucous membrane color, capillary refill time, heart rate, respiratory rate and effort, and mentation provide a crude but continuous oxygenation assessment. Arterial blood gas analysis remains the reference method for quantifying oxygenation and ventilation, and it also permits direct measurement of dyshemoglobins when co-oximetry is available. In ventilated patients, capnography detects airway obstruction and accidental extubation far earlier than pulse oximetry, as demonstrated in a hyperoxemic animal model where SpO2 remained unchanged for 180 seconds after complete airway obstruction while capnography waveforms flattened within seconds. For patients with suspected carbon monoxide exposure, pulse oximetry is unreliable and transcutaneous PO2 or co-oximetry should be used instead.

How does pulse oximetry perform in neonatal, avian, or exotic patients?

Most pulse oximeters are calibrated using adult mammalian hemoglobin and tissue models, so accuracy in neonates, birds, reptiles, and small exotic mammals is not guaranteed. Sensor size and placement are the primary practical limitations. In neonates, peripheral vasoconstriction and motion artifact degrade signal quality rapidly. In birds, the lack of a conventional peripheral circulation and the presence of nucleated erythrocytes do not inherently alter the optical signal, but the small tissue volumes and rapid heart rates challenge most commercial sensors. Reptiles present additional problems because of variable heart rates, shunting, and profound temperature-dependent perfusion. For any species outside standard calibration data, treat SpO2 as a trend monitor instead of an absolute value, and confirm critical readings with arterial blood gas analysis whenever feasible. The MSD Veterinary Manual provides species-specific guidance on normal oxygen parameters and monitoring approaches.

What is the most cost-effective way to maintain pulse oximetry capability in a busy practice?

Invest in one high-quality veterinary-specific or human pediatric pulse oximeter with reusable and disposable sensors instead of multiple inexpensive devices. Reusable clip sensors for dogs and cats, wrap sensors for small patients, and a reflectance sensor for rectal or esophageal use cover most clinical scenarios. Calibrate or verify the device against co-oximetry on arterial samples at least annually, and document the comparison. Keep spare batteries and sensors in the anesthesia and emergency areas. For practices with limited budgets, a single reliable device used consistently during anesthesia, recovery, and critical care outperforms several cheap devices with erratic readings. The AVMA practice resources offer guidance on equipment standards and maintenance planning for veterinary facilities.

How should I document SpO2 readings and signal quality in the medical record?

Record the numeric SpO2 value, the waveform quality, the sensor site, and the fraction of inspired oxygen at the time of measurement. A reading without a plethysmographic waveform should be documented as unverified or suspect, not as a definitive saturation. Note the time of each reading, any interventions performed, and the patient's response. In anesthetic records, record SpO2 at least every five minutes during stable anesthesia and continuously during induction, recovery, or instability. When SpO2 conflicts with clinical assessment or blood gas results, document both values and the reasoning for the interpretation. This documentation supports clinical decision-making and provides a defensible record if complications arise. The RECOVER guidelines emphasize structured monitoring and documentation during CPR and post-arrest care.

How do I explain a falsely reassuring or falsely alarming SpO2 reading to a concerned client?

Explain that the pulse oximeter estimates oxygen saturation by measuring light absorption through tissue, and that certain conditions distort this estimate. Use a concrete example: in carbon monoxide poisoning, the device cannot distinguish oxygen-carrying hemoglobin from carbon monoxide-bound hemoglobin, so it may read near normal while the patient is actually hypoxemic. In methemoglobinemia, readings tend to plateau around 84 to 86 percent regardless of true saturation. Reassure the client that the monitor is one tool among several, and that blood gas analysis or co-oximetry provides the definitive measurement when the reading does not match the patient's clinical appearance. Frame the explanation in terms of what the device measures versus what it cannot measure.

When should I escalate monitoring from pulse oximetry to more invasive methods?

Escalate when SpO2 is persistently below 94 percent despite supplemental oxygen, when the reading is unstable or inconsistent with clinical assessment, when dyshemoglobinemia is suspected, or when the patient requires mechanical ventilation. Arterial blood gas analysis with co-oximetry is indicated whenever SpO2 and clinical status diverge, before initiating or adjusting vasopressor therapy, and during weaning from mechanical ventilation. In cardiac arrest or severe hemodynamic instability, pulse oximetry often fails because of poor perfusion, capnography and blood gas analysis become the primary monitoring tools. For patients with carbon monoxide or methemoglobin exposure, co-oximetry is mandatory because standard pulse oximetry overestimates true oxygen saturation. Document the escalation and the rationale in the medical record.

Related Clinical & Scientific Guides

• Toxicology in Emergency Practice: Common Poisons and Diagnostic Approach
• Veterinary Cardiopulmonary Resuscitation: Post-Cardiac Arrest Care
• Fluid Therapy Guidelines for Dogs and Cats: A Practical Update

References and Further Reading

• Pulse oximetry in severe carbon monoxide poisoning.. 1998.
• The effect of carbon monoxide inhalation on pulse oximetry and transcutaneous PO2.. 1987.
• Continuous, noninvasive, and localized microvascular tissue oximetry using visible light spectroscopy.. 2004.
• Effects of methemoglobinemia on pulse oximetry and mixed venous oximetry.. 1989.
• Utility of monitoring capnography, pulse oximetry, and vital signs in the detection of airway mishaps: a hyperoxemic animal model.. 1998.
• Wavelength selection for low-saturation pulse oximetry.. 1997.
• RECOVER Initiative Veterinary CPR Guidelines. Veterinary Emergency and Critical Care Society.
• AAHA/AAFP Fluid Therapy Guidelines for Dogs and Cats. AAHA.
• MSD Veterinary Manual, Professional Edition. MSD Veterinary Manual.

Related Articles

• Troubleshooting Pulse Oximetry Failures in Critically Ill Patients
• Complications of Oxygen Therapy in Veterinary Patients
• Veterinary Mechanical Ventilation Weaning and Troubleshooting
• Failure Modes in Mechanical Ventilation of Veterinary Patients
• Hyperbaric Oxygen Therapy in Veterinary Patients: Evidence and Applications

This article is educational professional reference material for veterinary audiences. It is not a substitute for veterinary diagnosis, individual clinical judgment, current product labeling, or applicable regulatory requirements.