Oxygen Therapy Delivery Methods in Veterinary Critical Care

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

Key Takeaways

• The choice of oxygen delivery method in veterinary critical care significantly impacts the fraction of inspired oxygen (FiO2) achieved, patient work of breathing, and tolerance, with devices ranging from flow-by (25-40% FiO2) to oxygen cages and high-flow nasal oxygen (HFNO) (50-90% FiO2).
• Hyperoxia, defined as SpO2 >98%, is a recognized complication associated with increased mortality in specific human critical care subgroups and animal models, necessitating titration of oxygen therapy to the lowest FiO2 maintaining adequate oxygenation (target SpO2 94-98%).
• Device selection should be guided by patient status, with flow-by or hoods as first-line for stable, mildly hypoxemic patients, oxygen cages for moderate hypoxemia, and HFNO or immediate intubation for severe hypoxemia, considering species-specific tolerance (e.g., cats and brachycephalics often tolerate hoods/cages better than masks).
• Common failure modes include flow-by dilution, mask rebreathing (rising CO2), oxygen cage CO2 accumulation or interrupted delivery, and HFNO nasal irritation or gastric distension, requiring serial monitoring via pulse oximetry and arterial blood gas analysis.
• Brachycephalic breeds and cats may benefit from hoods or cages due to upper airway obstruction, while HFNO can partially bypass resistance but requires careful prong placement to avoid gastric distension and nasal irritation.
• Documentation of oxygen therapy is critical, including device, flow rate, FiO2, SpO2, respiratory parameters, patient tolerance, and any changes made, forming the basis for titration and escalation decisions.

Oxygen supplementation is among the most immediate interventions available to the veterinary criticalist, yet the choice of delivery device materially alters the fraction of inspired oxygen (FiO2) achieved, the work of breathing imposed on the patient, and the tolerance of long-term therapy. This article provides a procedural comparison of oxygen delivery methods used in veterinary critical care, including flow-by administration, mask systems, oxygen hoods, oxygen cages, and high-flow nasal oxygen therapy. It is written for practicing veterinarians who must select and monitor oxygen delivery systems in dogs, cats, and other small animal patients with respiratory compromise.

The clinical questions addressed are practical: what FiO2 can each device realistically deliver, which patients tolerate each method, what complications attend prolonged use, and how should therapy be monitored and adjusted? The article also examines the physiological rationale for titrating oxygen therapy against the risks of hyperoxia, a concern supported by human critical care literature and increasingly recognized in veterinary patients.

At a Glance

Physiology of Oxygen Supplementation

Oxygen delivery to tissues is the product of cardiac output and arterial oxygen content, the latter determined by hemoglobin concentration, hemoglobin saturation, and dissolved oxygen. In the dyspnoeic patient, the immediate goal of supplemental oxygen is to raise arterial oxygen tension (PaO2) and hemoglobin saturation while the underlying disease is diagnosed and treated. The relationship between inspired oxygen and arterial oxygenation is governed by the alveolar gas equation, which accounts for barometric pressure, water vapour pressure, and arterial carbon dioxide tension. Patients with hypoventilation, diffusion impairment, ventilation-perfusion mismatch, or right-to-left shunting each respond differently to increased FiO2, and the delivery device must be chosen with these mechanisms in mind.

The oxygen-hemoglobin dissociation curve is sigmoid, so small increases in PaO2 produce large gains in saturation when the patient is on the steep portion of the curve. This explains why modest FiO2 increases from flow-by oxygen can be clinically meaningful in a hypoxemic patient, even when the delivered FiO2 is imprecise. Conversely, once hemoglobin is fully saturated, further increases in FiO2 add only the small amount of oxygen carried in physical solution, and the risk of oxygen toxicity begins to accrue.

Risks of Hyperoxia

The assumption that higher oxygen fractions are always beneficial has been challenged by evidence from human critical care. A systematic review and meta-analysis of cohort studies in critically ill adults found an association between arterial hyperoxia and increased mortality in specific subgroups, particularly after cardiac arrest and in mechanically ventilated patients, although the authors noted substantial methodological heterogeneity and an absence of dedicated studies in sepsis, acute lung injury, and trauma populations. Animal data support this concern. A systematic review and meta-analysis of controlled animal trials modeling cardiac arrest found that ventilation with 100% oxygen after return of spontaneous circulation produced significantly worse neurological deficit scores and increased histological neuronal damage compared with lower oxygen concentrations.

These findings have direct relevance to veterinary resuscitation and post-arrest care. The RECOVER Initiative veterinary CPR guidelines incorporate evidence evaluation of oxygen management during and after cardiopulmonary resuscitation, and they emphasize the need to titrate oxygen delivery to measured saturation instead of defaulting to 100% oxygen indefinitely. The practical implication is that oxygen therapy should be viewed as a drug with a dose-response relationship, and the lowest FiO2 that maintains adequate oxygenation should be used.

Determinants of Delivered FiO2

The FiO2 actually delivered by any device depends on oxygen flow rate, the patient's minute ventilation, the volume of the reservoir or chamber, and the degree of entrainment of room air. In flow-by systems, the oxygen stream is diluted by room air before it reaches the nares, and the effective FiO2 falls as the distance between the tubing and the patient's face increases. Mask systems create a larger reservoir but also allow rebreathing of exhaled gas if the flow rate is insufficient to flush the mask volume. Oxygen cages and hoods enclose the patient in a controlled atmosphere, but the FiO2 changes slowly when the patient enters or when the door is opened, and carbon dioxide can accumulate if fresh gas flow is inadequate.

The MSD Veterinary Manual provides species-specific guidance on oxygen flow rates and delivery techniques, and the reader should consult current editions for device-specific recommendations. In all systems, the delivered FiO2 should be estimated from the device characteriztics and then confirmed by arterial blood gas analysis or pulse oximetry wherever possible, because patient factors such as panting, struggling, or brachycephalic conformation can substantially alter the achieved oxygen fraction.

Patient Assessment and Oxygenation Targets

Before selecting a delivery method, quantify the patient's oxygenation deficit. Pulse oximetry provides a continuous estimate of hemoglobin saturation, but it loses reliability when perfusion is poor, when the probe site is poorly perfused, or when methaemoglobin is present. Arterial blood gas analysis remains the reference standard for calculating the alveolar-arterial oxygen gradient and for distinguishing hypoventilation from diffusion impairment or shunt. Venous blood gas samples can approximate arterial pH and bicarbonate but should not be used to estimate PaO2.

The target SpO2 for most veterinary patients is 94% to 98%. Values below 90% correspond to a PaO2 near 60 mmHg and mandate immediate intervention. Values above 98% on supplemental oxygen suggest unnecessary hyperoxia, which carries its own risks. A systematic review and meta-analysis of cohort studies in critically ill adults found an association between arterial hyperoxia and increased mortality in specific subgroups, particularly after cardiac arrest and in mechanically ventilated patients Association Between Arterial Hyperoxia and Outcome in Subsets of Critical Illness. Animal trials modeling cardiac arrest and resuscitation have shown worse neurological deficit scores in subjects ventilated with 100% oxygen compared with lower concentrations The effect of hyperoxia following cardiac arrest. These findings support a strategy of titrating oxygen to a defined target instead of administering the highest possible fraction.

Serial assessment determines whether the chosen device is adequate. Recheck SpO2 within 5 to 10 minutes of any change in delivery method or flow rate. If the target is not met, escalate to a higher-FiO2 device instead of simply increasing flow on the current one, because each device has a ceiling beyond which additional flow adds little.

Flow-By Oxygen

Flow-by oxygen is the least invasive method and the most appropriate first step in a dyspnoeic patient that resists handling. A tube or mask is held 1 to 2 cm from the nares, delivering an FiO2 that varies widely with flow rate, distance, and the patient's respiratory pattern. Reported FiO2 values range from approximately 25% to 40%, with higher flows and closer positioning producing higher fractions. The method is inherently unstable: any movement of the patient or the operator changes the delivered concentration.

Flow-by is best used during initial triage, while venous access is obtained, or during brief procedures such as radiography. It is not suitable for prolonged oxygen dependence because it consumes large volumes of gas, requires continuous staff attention, and provides no humidification. Cats tolerate flow-by better than masks in many cases, but a stressed cat may still resist the noise and air movement. In brachycephalic breeds, flow-by directed at the nares may be partially ineffective because upper airway obstruction limits the inspiratory flow that reaches the lower airways.

Oxygen Masks and Hoods

Oxygen masks can deliver FiO2 from 40% to 60% when a tight seal is achieved, but most veterinary patients do not tolerate a sealed mask for more than a few minutes. A loose-fitting mask with high flow rates (3 to 10 L/min) provides an FiO2 closer to 35% to 50%. The mask should be transparent to allow visualization of mucous membrane color and any vomitus or secretions. Masks are appropriate for short-term stabilization, pre-oxygenation before intubation, and during recovery from anesthesia.

The Elizabethan collar hood method uses a plastic cone with the narrow end covered in plastic wrap and a port cut for the oxygen line. This creates an oxygen-enriched chamber around the head with an FiO2 of approximately 40% to 50% at flow rates of 1 to 2 L/min. Hoods are better tolerated than masks in many cats and small dogs because they do not contact the face. They also allow the patient to assume a comfortable position. Disadvantages include CO2 retention if flow is too low, fogging that obscures visualization, and the risk of hyperthermia in small patients if the hood is not vented. Hoods are unsuitable for patients that paw at the device or that cannot maintain sternal recumbency.

Oxygen Cages

Oxygen cages provide a controlled environment with FiO2 up to 60% in most commercial units, with some capable of higher fractions. They are the method of choice for patients that require continuous oxygen but that become distressed with masks or hoods. The cage allows the patient to move freely, reduces handling stress, and permits concurrent nursing care through access ports.

The principal limitation of oxygen cages is the loss of FiO2 every time the door is opened. Each opening effectively returns the cage to room air, and re-establishing the target fraction takes several minutes. This makes cages poorly suited to patients that need frequent interventions, repeated blood sampling, or close monitoring. Cages also isolate the patient from the clinician, making subtle changes in respiratory effort harder to detect. Temperature and humidity must be monitored, because recirculated oxygen dries the airways and can cause hyperthermia in small patients. Condensation on the walls obscures observation and may indicate inadequate ventilation of the chamber.

Oxygen cages are appropriate for stable patients that need moderate oxygen supplementation over hours to days, such as those with pneumonia, pulmonary contusions, or mild upper airway obstruction. They are not appropriate for unstable patients, patients requiring frequent suction or physiotherapy, or patients that are dyspnoeic enough to need immediate intervention when the door opens.

High-Flow Nasal Oxygen

High-flow nasal oxygen (HFNO) delivers heated, humidified oxygen through nasal prongs at flow rates that exceed the patient's peak inspiratory flow. This produces a delivered FiO2 of up to 100% while providing low-level positive airway pressure, flushing of nasopharyngeal dead space, and improved mucociliary clearance. HFNO requires specialised equipment, including a flow generator, heated humidifier, and nasal cannulae sized to the patient.

The main advantages of HFNO over cages and masks are the stability of the delivered FiO2, the ability to monitor the patient continuously without breaking the oxygen circuit, and the reduced work of breathing from the positive pressure effect. HFNO is particularly useful in patients with hypoxemic respiratory failure that are not candidates for immediate intubation, and in the post-extubation period. It is also valuable in cats with severe bronchial disease, where the heated humidification reduces airway irritation.

Disadvantages include the cost of equipment, the need for patient cooperation, and the risk of gastric distension if flow rates are excessive or the prongs are misplaced. Nasal mucosal drying and epistaxis can occur despite humidification. HFNO is not appropriate for patients with complete nasal obstruction, severe facial trauma, or those that cannot maintain a patent airway.

Device Selection by Patient Status

The decision algorithm begins with the patient's stability. A patient in respiratory distress that cannot be safely handled should receive flow-by oxygen while the clinician assesses the cause. If the patient tolerates the flow-by and SpO2 improves to target, continue with flow-by or transition to a hood. If SpO2 does not improve within 5 to 10 minutes, escalate to a mask or cage. A patient that is already collapsed, cyanotic, or apnoeic should be intubated and ventilated with 100% oxygen instead of attempting any non-invasive device.

Species and conformation alter the correct choice. Cats and brachycephalic dogs tolerate hoods and cages better than masks. Deep-chested dogs with large tidal volumes may exhaust the FiO2 capacity of a cage more quickly than small patients. Patients with thoracic trauma and suspected pneumothorax require concurrent thoracocentesis, oxygen therapy alone will not resolve the hypoxemia. The MSD Veterinary Manual provides species-specific guidance on oxygen therapy and respiratory support that should be consulted when adapting these principles to non-canine, non-feline patients, including exotic and production animals.

Document the device, flow rate, estimated FiO2, SpO2, respiratory rate and effort, and the patient's tolerance at each assessment. Record the time of any device change and the response to that change. This documentation supports titration decisions and provides a basis for escalation to mechanical ventilation if non-invasive methods fail.

Recognized Complications and Early Detection

Oxygen therapy failure typically presents as persistent hypoxemia, progressive hypercapnia, or patient intolerance that limits delivered oxygen. Each device class has characteriztic failure modes that the clinician should anticipate.

Flow-by oxygen fails when the patient moves out of the oxygen stream, when flow rates are set too low for the patient's minute ventilation, or when nasal discharge and panting dissipate the oxygen plume. Detection relies on serial pulse oximetry and blood gas analysis. A patient that appears comfortable but maintains SpO2 below 94% despite apparently adequate flow-by supplementation should prompt immediate reassessment of delivery geometry, not an increase in flow rate alone.

Oxygen masks fail through rebreathing when the mask volume is too small or the flow rate insufficient to flush exhaled carbon dioxide. The discriminating finding is rising end-tidal or arterial CO2 with stable or improving oxygenation. Masks also provoke anxiety in dyspneic patients, and the resulting tachypnea increases oxygen consumption and work of breathing. A patient that fights the mask is receiving less oxygen than one that rests calmly with a lower nominal FiO2.

Oxygen cages fail through carbon dioxide accumulation when soda lime is exhausted, through humidity and temperature drift that increases respiratory heat loss, and through the unavoidable interruption of oxygen delivery during handling, medication administration, and cage cleaning. The cage also isolates the patient from observation, and a deteriorating patient may be discovered only when the transparent walls are examined closely. Intermittent access for procedures means that critically hypoxemic patients experience repeated episodes of relative hypoxia.

High-flow nasal oxygen fails through tube dislodgement, through flow rates that cannot meet the patient's inspiratory demand, and through gastric distension when the patient swallows air. Nasal mucosal drying and crusting can obstruct the cannulae, and the audible noise of high-flow systems can distress cats in particular. Detection requires direct inspection of the nares and cannula position, not reliance on the oxygen source flow meter alone.

Hyperoxia is a recognized complication of aggressive oxygen supplementation. A systematic review and meta-analysis of cohort studies in critically ill adults found associations between arterial hyperoxia and increased mortality in several subgroups, including cardiac arrest survivors and mechanically ventilated patients Helmerhorst et al., 2015. Animal models of cardiac arrest resuscitation show worse neurological outcomes with 100% oxygen compared with lower concentrations Pilcher et al., 2012. These findings support the practice of titrating oxygen to a target SpO2 range instead of administering the highest achievable FiO2.

Common Errors and Corrective Actions

Less experienced clinicians frequently set flow-by oxygen at rates appropriate for a mask, typically 1 to 3 L/min, which produces a negligible FiO2 in a panting dog. The corrective action is to calculate flow based on the patient's inspiratory flow rate and to verify the effect with pulse oximetry instead of assuming the device is working.

A second common error is interpreting a normal SpO2 as evidence of adequate ventilation. Oxygenation and ventilation are independent variables, and a patient with progressive hypercapnia can maintain a normal SpO2 until respiratory failure is advanced. Capnography and blood gas analysis are required in any patient whose respiratory effort is increasing.

A third error is leaving a patient in an oxygen cage without scheduled reassessment. The cage creates a false sense of security because the patient appears quiet. The clinician should establish a monitoring interval based on the patient's stability, and should remove the patient for brief assessment of respiratory effort, mucous membrane color, and mentation at regular intervals.

Limitations of the Evidence and Areas of Disagreement

The veterinary evidence base for oxygen delivery devices consists largely of small physiological studies and clinical experience instead of randomised controlled trials. The RECOVER initiative provides consensus guidance for CPR and post-arrest care, but does not make device-specific recommendations for oxygen delivery in non-arrest respiratory failure RECOVER Initiative. Expert opinion differs on the optimal FiO2 target for dogs and cats with pulmonary parenchymal disease, with some authorities accepting SpO2 values as low as 90% to avoid oxygen toxicity and others recommending a higher threshold.

The human literature on hyperoxia is extensive but conflicting. The systematic review by Helmerhorst and colleagues noted that no studies dedicated to patients with acute lung injury, sepsis, shock, or multiple trauma could be included, and that the definition of hyperoxia varied across studies Helmerhorst et al., 2015. Extrapolation of these findings to veterinary patients requires caution, particularly for species differences in pulmonary oxygen toxicity susceptibility. The role of normobaric oxygen in conditions such as stroke remains an area of active investigation, with some evidence suggesting that early oxygen therapy may extend the window for reperfusion therapies Singhal, 2007.

Referral, Consultation, and Reporting

Referral to a specialist critical care service is warranted when a patient requires high-flow nasal oxygen, when hypoxemia persists despite maximal conventional delivery methods, or when the underlying disease requires advanced diagnostics such as thoracic ultrasound, CT, or bronchoscopy. Patients that fail to improve within 6 to 12 hours of appropriate oxygen therapy should be evaluated for non-pulmonary causes of hypoxemia, including cardiac disease, pulmonary thromboembolism, and anemia.

Laboratory involvement is indicated for serial blood gas analysis, for assessment of acid-base status in patients with suspected hypercapnia, and for evaluation of underlying disease processes. Regulatory reporting obligations vary by jurisdiction and by the underlying diagnosis. The World Organization for Animal Health maintains international standards for notifiable diseases that may present with respiratory signs WOAH terrestrial animal health code. Clinicians should consult their local veterinary authority for current reporting requirements.

Frequently Asked Questions

How should I select an oxygen delivery method when only limited equipment is available?

When an oxygen cage or high-flow nasal oxygen unit is unavailable, prioritize the method that delivers the highest reliable FiO2 with the least patient stress. Flow-by oxygen at 2 to 5 L/min in dogs and 1 to 3 L/min in cats is the minimum intervention, but delivered FiO2 is highly variable and depends on distance from the nares and panting behavior. An oxygen mask with a rubber diaphragm creates a partial seal and outperforms flow-by for most patients. For fractious cats, an oxygen hood constructed from an Elizabethan collar and clear plastic wrap provides a low-stress alternative. Reassess arterial oxygenation or pulse oximetry within 10 to 15 minutes of any device change, and escalate to a more efficient method if targets are not met.

What is the role of oxygen therapy in the post-arrest period?

The RECOVER veterinary CPR guidelines recommend titrating oxygen to normoxaemia after return of spontaneous circulation instead of continuing 100% oxygen empirically. Animal models of cardiac arrest show that ventilation with 100% oxygen after resuscitation produces worse neurological deficit scores and greater histological neuronal injury than lower oxygen concentrations, as reported in a systematic review and meta-analysis of animal trials hyperoxia following cardiac arrest. The mechanism is thought to involve reperfusion injury amplified by hyperoxia. Once spontaneous circulation is restored, reduce inspired oxygen in steps while monitoring SpO2, aiming for 94% to 98% in dogs and cats. Avoid both hypoxemia and hyperoxia, as the optimal range is narrow.

How does oxygen therapy differ in brachycephalic and toy breed patients?

Brachycephalic dogs and cats frequently present with upper airway obstruction, and oxygen therapy is supportive while the obstruction is managed. These patients often hyperventilate and may become distressed by masks, which worsens airway edema. An oxygen cage is usually better tolerated initially. For toy breeds, oxygen hoods and masks add dead space, which can increase rebreathing of carbon dioxide in patients with already compromised ventilation. Monitor capnography or venous blood gases if available. High-flow nasal oxygen can bypass upper airway resistance partially, but placement of nasal prongs in stenotic nares may be difficult. Sedation to reduce distress must be used cautiously because it can depress respiratory drive in these patients.

What documentation should accompany oxygen therapy in the medical record?

Record the indication for oxygen therapy, the delivery device, flow rate or FiO2 setting, and the time of initiation. Document the patient's respiratory rate and effort, SpO2 or arterial blood gas values, and the assessment of perfusion before starting therapy and at each reassessment. Note the patient's tolerance of the device and any changes made in response to deterioration or improvement. Record the oxygen weaning plan, including the target SpO2 range and the stepwise reduction schedule. If the patient is discharged on oxygen, document the home environment assessment and client instructions. The AVMA practice resources provide general guidance on medical record standards that apply to oxygen therapy documentation.

How do I explain oxygen therapy options to a client whose pet is not improving?

Explain that oxygen therapy is supportive, not curative, and that the underlying disease must be treated concurrently. Describe the device in terms the client can observe, such as the oxygen cage as a quiet space where the pet can rest, or nasal prongs as a small tube that delivers oxygen directly. State clearly that oxygen supplementation does not treat the cause of low oxygen levels, and that diagnostic testing is needed to identify why the pet is hypoxemic. If the pet is not improving despite oxygen, explain that the next steps may include advanced imaging, bronchoscopy, or referral. Be honest about prognosis while maintaining a professional and compassionate tone. Document the conversation in the medical record.

When should I consider referral for hyperbaric oxygen therapy?

Hyperbaric oxygen therapy is not a first-line intervention in veterinary critical care. It requires specialised chambers and trained personnel, and availability is limited to referral centers. The evidence base in human medicine supports its use for refractory wounds, radiation injuries, and compromised grafts, with mechanisms based on reactive oxygen and nitrogen species signaling, as reviewed in the hyperbaric oxygen mechanisms and efficacy literature. In veterinary patients, consider referral when a condition known to respond to hyperbaric oxygen is present and conventional therapy has failed, such as non-healing wounds or ischemic tissue. Discuss the risks of transport and the delay in ongoing critical care with the receiving facility before transfer.

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

• Association Between Arterial Hyperoxia and Outcome in Subsets of Critical Illness: A Systematic Review, Meta-Analysis, and Meta-Regression of Cohort Studies.. 2015.
• A review of oxygen therapy in ischemic stroke.. 2007.
• Hyperbaric oxygen: its mechanisms and efficacy.. 2011.
• Retinal degeneration and local oxygen metabolism.. 2005.
• Hyperbaric oxygen therapy.. 1990.
• The effect of hyperoxia following cardiac arrest - A systematic review and meta-analysis of animal trials.. 2012.
• 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

• Veterinary Oxygen Therapy: Flow Rates and Delivery Devices
• Capnography in Veterinary Emergency and Critical Care
• Complications of Oxygen Therapy in Veterinary Patients
• Veterinary Electrocardiography in Emergency and Critical Care
• 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.