Oxygen therapy, like all technology-based treatments, is continuously evolving. It is defined as the therapeutic use of oxygen (O2) administered at concentrations higher than those of room air to treat or prevent the effects of hypoxemia.1–3 The aim of administering oxygen is to maintain oxygenation levels above the range of respiratory failure (O2 blood pressure [PaO2]>60mmHg, arterial oxygen saturation measured by pulse oximetry [SpO2]>90%). It is widely used in different care areas in patients with acute and chronic respiratory problems, at home and in the hospital.4 Although O2 is considered a medication by health authorities, its widespread use and image as a benign, uncomplicated treatment have resulted in careless dosing and little concern over side effects, in spite of the available scientific evidence.5,6 As a medication, O2 should be used with the same rigor as any other drug, and indications and dosage must be evidence-based. Only in this way will we achieve the main objectives of this therapy: improved survival, proper correction of hypoxemia, and avoidance of the risks of excessive oxygenation. In this review, we will address the need for oxygen titration, the advantages of automatic and manual titration, the required features of automatic titration devices, the devices available on the market and the evidence supporting them, and, finally, the limitations and shortcomings of existing equipment.

Both hypoxemia and hyperoxia resulting from exposure to excessive O2 flows can have serious consequences for patients,7,8 as demonstrated by data from children and adults with acute and chronic respiratory failure.

There is ample clinical evidence of the undesirable effects of oxygen therapy in the neonatal population. An insufficient supply of O2 has deleterious effects on neurological development and pulmonary circulation,9–15 while an excessive supply is associated with systemic oxidative damage in different organs.16–19 For this reason, maintaining the right oxygenation levels is essential. Traditionally, oxygen titration was performed manually, by adjusting O2 flows to keep SpO2 within the desired range. However, children titrated manually have shown SpO2 levels below the range for more than 50% of the time and above the range for at least 33% of the time. This could be related to both the ability of health personnel to respond to fluctuations in SpO2, and the excessive tolerance of children to high levels of SpO2.20–22

In adults, the goal of oxygen therapy is also to avoid the consequences of hypoxemia, but the well-known harmful effects of hyperoxia are, if anything, even more frequently neglected.23

In cases of acute respiratory failure, hypoxemia has been associated with an increase in mortality following myocardial infarction and serious head injuries.24–26 It is also associated with increased mortality in patients admitted to intensive care units (ICU) and a greater number of complications in the postoperative period.27,28 Suboptimal use of oxygen therapy has also been observed in patients with chronic obstructive pulmonary disease (COPD) exacerbations, in whom the effects of hypoxemia are particularly significant and widely known.29,30 These patients are sometimes poorly oxygenated over long periods.31

Hyperoxia also entails a greater risk for patients with acute respiratory failure, not because the risk is unknown, but because it is minimized by healthcare professionals.6,7 Its presence is associated, irrespective of hypoxemia, with an increased risk of death after cardiorespiratory arrest and increased mortality in patients with serious head injuries.25,32,33 Nearly 25% of patients with COPD exacerbations are exposed to significant periods of hyperoxia, which is related to higher mortality and worse clinical outcomes that can be prevented by careful adjustment of O2 flows.34–36 In spite of all this evidence, the clinical reality is that the vast majority of patients receive high-flow O2 during COPD exacerbations, an intervention that carries an increased risk of respiratory acidosis.37

Patients with stable chronic respiratory failure often present long-term hypoxemia, this being the main reason for reduced exercise tolerance and multiple, well-characterized complications (pulmonary hypertension, right heart failure, polycythemia, and increased mortality).38,39 The use of continuous home oxygen therapy (CHOT) in COPD patients has shown beneficial effects on survival, quality of life, and exercise tolerance,2,40–42 and its indication has been extended, by analogy, to other chronic lung diseases.43–47 In CHOT, oxygen flow is usually fixed. This is not always optimal,48 since the needs of patients vary throughout the day. More than half have O2 desaturation during their usual activities.49,50

All these data show that oxygen therapy must be adapted to individual patient requirements. The available evidence indicates that the proper use of oxygen therapy is an important factor in clinical outcomes,51 and that a prescription based on continuous monitoring can significantly increase the percentage of time with an appropriate SpO2, while decreasing the amount of O2 prescribed.52 Essential objectives in the adjustment of oxygen therapy should be:

• •

  To optimize therapy and safety, minimizing the number of episodes of hypoxemia and hyperoxia. The prescription should be aimed at achieving a saturation range, rather than establishing high O2 flows, regardless of arterial O2 levels.35

• •

  To tailor O2 flow to the real needs of patients in their daily activities. O2 flow selection does not always meet the needs of the patient, and may be insufficient when demand increases.53

• •

  To optimize the consumption of O2 to increase time on portable devices; the technological limitations of these devices underline the need for measures that adapt therapy to real life.

• •

  To improve morbidity and mortality outcomes.

• •

  To reduce the excessive costs generated by inappropriate use.

The current situation shows that these goals are difficult to achieve in clinical practice.54 As already mentioned, the traditional method of manual titration has significant limitations. Trained personnel are needed both in the hospital and at home. Healthcare personnel in ICUs dedicate a significant amount of a time to titration, and the process may be burdensome for patients, and can contribute to poor therapeutic compliance. Moreover, the adjustment achieved is suboptimal and episodes of under- and overdosing are common, decreasing the efficiency of the treatment and jeopardizing patient safety.

All these findings, concerns, and limitations have led in recent years to the development of physiological, controlled, closed-circuit medical devices for the automatic titration of O2 flows. These are generally defined as devices that use 1 or several sensors to automatically change a physiological variable (O2 flow), thus maintaining the oxygenation level determined for each patient55 (Fig. 1). Automatic O2 titration consists of 3 components: a system for monitoring oxygenation; an algorithm that determines the O2 flow settings to achieve the appropriate oxygenation level; and an O2 source.53

• 1)

  Monitoring oxygenation levels. Various systems can be used, depending on the variable selected. The use of permanent PaO2 electrodes and transcutaneous O2 pressure (PtcO2) electrodes is limited due to the need for an invasive catheter in the first case and frequent calibrations in the second. SpO2, as a noninvasive method that does not require calibration, has thus become the preferred monitoring system.

• 2)

  Algorithms. Algorithms that determine O2 flow settings required to maintain established SpO2 figures compare continuous or intermittent SpO2 results with target levels and adjust O2 flow accordingly. A major challenge is to determine when hypoxemia is a real event and when it is due to movement artifacts or a weak pulse oximeter signal, in order to avoid the risk of O2 overdosing and hyperoxia. Moreover, the response to hypoxemia should be sufficient to correct the situation, but without producing rebound hyperoxia. Algorithms are specific to each device and hard to compare, so they must all be validated clinically before use.

• 3)

  Source of O2. O2 can come from a wall outlet or a bottle of compressed O2, or from any other source supplying a flow that can be controlled by an external mechanism. The O2-enriched gas can then be administered to the patient via mechanical ventilators, continuous positive airway pressure (CPAP) devices, high-flow equipment or oxygen therapy devices, such as nasal prongs. Venturi masks are not appropriate for O2 automatic titration, as the fraction of inspired O2 (FiO2) will depend on the size of the mask aperture, and not on the flow supplied. Ventilators and high-flow equipment must have accurate, robust mixers that can withstand the required changes in O2 flow.

Fig. 1.

Design of a controlled closed-loop physiological medical device.

(0.18MB).

The mode of operation is easy and intuitive. The pulse oximetry signal is analyzed on a continuous basis by the central unit, which increases or decreases the flow of O2 as needed to maintain the predetermined level of SpO2.

These devices have the potential to optimize oxygen therapy by reducing the risk of over- and underdosing and overburdening caregivers, and improving patients’ clinical outcomes. Compared to the erratic adjustment of manual titration, these devices can achieve appropriate, sustained oxygenation levels by continuous, automated adjustment of the O2 flow53 (Fig. 2). The results of a recent meta-analysis show how these devices increase the time during which SpO2 is maintained within the range predefined by the clinician compared with manual titration,56 opening up new possibilities in the area of oxygen therapy titration: automatic titration or automated titration.

Fig. 2.

Differences between the manual and automatic titration.

(0.17MB).

Despite the numerous studies on automatic titration, many questions remain. Do automatic titration devices work with any interface? Are the oximetry sensors sufficiently sensitive and reliable? Is there a risk of hardware malfunction due to development defects or electrical problems? What is their role in oxygen therapy? In the acute patient? In home therapy? During exertion?

In most studies, the circuit feedback variable was SpO2, and patients were continuously connected to a pulse oximeter. This is the only source of information that feeds the oxygen adjustment algorithm. In manual titration, however, the doctor has several sources of information (monitors, patient response, etc.), thus increasing patient safety. In this regard, the instability of the sensor to movements or physiological delay in the measurements is significant, so improvements in pulse oximetry will be key in the development of clinically effective monitors.

Moreover, the effectiveness of oximetry in assessing hyperoxia is very limited, as is its ability to promptly detect respiratory depression, even when is associated with significant CO2 retention. New lines of research, such as the oxygen reserve index (ORI), aimed at improving the reliability of measurements are of considerable interest.76 ORI is a non-invasive continuous variable provided by a generation of pulse oximeters that use multi-wavelength pulse co-oximetry. The ORI reflects oxygenation in the moderate hyperoxic range. This provides an early warning of deteriorated oxygenation well before any changes occur in SpO2, and is therefore a good complement to pulse oximetry the management of patients with oxygen therapy.76

Another problem of automatic titration is that an increase in O2 flow may not be an appropriate response when hypoxemia is due to hypoventilation. Moreover, to minimize the risk of masking continued deterioration in respiratory function by automatically increasing in O2 flow, automatic titration devices should warn of persistent increases in the O2 flow, even if SpO2 remains within the range.

In conclusion, automatic titration systems improve the dosage of oxygen therapy while avoiding the deleterious consequences of hypoxemia and hyperoxia in acute and chronic respiratory failure. However, they cannot replace clinical monitoring, particularly in light of their limitations in correcting hypoventilation-induced hypoxemia. Despite these drawbacks, the clinical potential of these devices is considerable, and they may help doctors tailor oxygen therapy to the needs of individual patients. Let us not forget that, although oxygen therapy saves lives, when used inappropriately it can cause serious damage. Technological progress in this field will, without doubt, lead to the development more reliable equipment with more sensitive sensors with a capacity for faster response; however, new avenues of research in this area of knowledge are still needed. At the present time, automatic titration of oxygen therapy must remain a subject of research.

Sagrario Mayoralas Alises is the medical director of Oximesa Grupo Praxair, supplier of domiciliary respiratory therapies. Oximesa does not sell any of the devices described in the manuscript. The other authors state that they have no conflict of interests with regard to this manuscript.