Oxygen Therapy and Noninvasive Ventilation in Chronic Obstructive Pulmonary Disease

Both hypoxemic and hypercapnic respiratory failure occur in patients with progressive chronic obstructive pulmonary disease (COPD). The presence of respiratory failure predicts worse prognosis and higher mortality. Supplemental oxygen therapy (SOT) and noninvasive ventilation (NIV) have been increasingly used to treat these abnormalities, aiming to improve both prognosis and quality of life. This review provides an overview of the evidence and current recommendations for the use of SOT and NIV in COPD.

Key points

  • Supplemental oxygen therapy is recommended in patients with severe hypoxemia, and also those with moderate hypoxemia if there is concomitant pulmonary hypertension, heart failure or polycythemia.

  • Noninvasive ventilation (NIV) improves the prognosis in patients with hypercapnic chronic obstructive pulmonary disease (COPD) exacerbation and may allow invasive ventilation to be avoided.

  • Data for long-term NIV in chronic hypercapnic respiratory failure are not conclusive, but the body of evidence for benefit is growing.

  • After COPD exacerbation, long-term NIV should be prescribed only to patients with persistent hypercapnia.

  • In this setting, it is important to target NIV to achieve a significant reduction in carbon dioxide levels.


Progressive chronic obstructive pulmonary disease (COPD) is associated with the development of respiratory failure, which worsens disease prognosis and predicts high mortality. The main feature of respiratory failure is inadequate blood oxygenation and/or decarboxylation by the lungs. Supplemental oxygen therapy (SOT) and mechanical ventilation are important components of therapeutic strategies for respiratory failure, and can improve oxygenation and decarboxylation. However, although these are simple therapeutic concepts, currently available evidence to support their usage in COPD is complex. Furthermore, the heterogenicity of COPD and the variety of associated respiratory failure phenotypes make definitive unified clinical judgments about these therapies very difficult.

This review provides an overview of the best published evidence relating to the use of SOT and noninvasive ventilation (NIV) in the management of COPD. The most recent recommendations are also presented, although it is important to note that these are not always consistent. However, we have tried to unify the recommendation when possible or represent the most obvious recommendation with the strongest evidence when there are inconsistencies in current guidelines.

Supplemental oxygen therapy in chronic obstructive pulmonary disease

The general indications, technical details and different application systems are outside the scope of this article. The focus here is on use of SOT in COPD.

Pathophysiology and Burden of Hypoxemia in Chronic Obstructive Pulmonary Disease

The pathophysiology of hypoxemia in COPD is complex and multifactorial ( Box 1 ); however, uncorrected chronic hypoxemia is associated with the development of adverse sequelae of COPD, including secondary pulmonary hypertension (PH), secondary polycythemia, systemic inflammation, neurocognitive dysfunction, and skeletal and respiratory muscle dysfunction. , A combination of these factors leads to reduced exercise tolerance and quality of life (QoL), and increased cardiovascular morbidity and death.

Box 1

Pathophysiological causes of hypoxemia in chronic obstructive pulmonary disease

Data from Refs.

Abbreviations: COPD, chronic obstructive pulmonary disease; OSA, obstructive sleep apnea; PH, pulmonary hypertension; V/Q, ventilation/perfusion.

  • 1.

    V/Q mismatch , :

    • The most important cause!

    • Due to 2 mechanisms:

      • a.

        Airflow limitation.

      • b.

        Emphysematous destruction of the pulmonary capillary bed.

  • 2.

    Alveolar hypoventilation/diminished ventilatory response to hypoxemia :

    • Due to 2 disorders:

      • a.

        Defective neural control mechanisms.

      • b.

        Malfunctioning inspiratory muscle/dynamics and hyperinflation (more importantly!).

  • 3.

    Pulmonary vascular remodeling/secondary PH ,

  • 4.

    Obesity , :

    • Increasingly prevalent among patients with COPD.

    • Leads to absolute and relative hypoxemia in many mechanisms:

      • a.

        Small airway dysfunction

      • b.

        Worsening chest wall compliance

      • c.

        Worsening V/Q mismatch

      • d.

        Increasing peripheral oxygen consumption

      • e.

        Associated sleep-disordered breathing (nocturnal hypoxemia).

  • 5.

    Nocturnal hypoxemia :

    • Complex mechanisms, but mainly due to:

      • a.

        Alveolar hypoventilation (multifactorial)

      • b.

        Commonly coexisting OSA (ie, overlap syndrome)

Positive Effects of Supplemental Oxygen Therapy in Chronic Obstructive Pulmonary Disease

From a physiologic point of view, SOT affects not only blood oxygenation but also tissue/cellular oxygenation. Arterial lactate levels are elevated in hypoxemic patients with COPD, which reflects a greater reliance on anaerobic metabolism and is a strong mortality predictor. SOT reduces lactate levels, which indirectly indicates improved tissue oxygenation. Consequently, SOT leads to improvements of multiple physiologic functions ( Table 1 ).

Table 1

Physiologic effects of oxygen therapy on different organ functions

Data from Refs. ,


  • Reduces exercise minute ventilation.

  • Reduces ventilatory requirements at submaximal work and therefore improves exercise endurance and reduces dynamic hyperinflation.

  • No stable effect on resting minute ventilation.

Cardiovascular system , ,

  • Reduces pulmonary artery pressure and increases cardiac output (through eliminating hypoxic vasoconstriction).

  • Reduces the resting heart rate, without affecting the ejection fraction.


  • Improves renal blood flow with resulting diuresis.

Bone marrow ,

  • Reverses secondary polycythemia.

  • May improve platelet survival time.

Central nervous system

  • Neuropsychiatric improvement.

From a clinical point of view, 2 clinical trials conducted in the early 1980s investigated the effect of long-term oxygen therapy (LTOT) on survival and established its use in COPD: the Nocturnal Oxygen Therapy Trial (NOTT) and the Medical Research Council (MRC) Long-Term Domiciliary Oxygen Therapy Trial ( Table 2 ).

Table 2

Study populations and key findings of the NOTT and MRC trials

Data from Refs. ,

Inclusion criteria

  • Pa o 2 ≤55 mm Hg (≤7.3 kPa), or

  • Pa o 2 ≤59 mm Hg (≤7.87 kPa) accompanied by either edema, P-pulmonale (electrocardiogram) or raised hematocrit (≥55%)

  • Pa o 2 40–60 mm Hg (5.3–8 kPa) and at least 1 recorded episode of heart failure with ankle edema

Results The 24-mo mortality rate was significantly lower in patients randomized to continuous oxygen therapy (mean usage ∼18 h/d) than in those randomized to 12 h/d of nocturnal oxygen therapy (22.4% vs 40.8%, respectively; P <.01) The 5-y mortality rate was significantly lower in the LTOT group (usage ≥15 h/d) than in control subjects who did not receive LTOT (risk of death 12%/y vs 29%/y, respectively; P = .04)

Abbreviations: LTOT, long-term oxygen therapy; MRC, Medical Research Council long-term domiciliary oxygen therapy trial; NOTT, nocturnal oxygen therapy trial; Pa o 2 , arterial partial pressure of oxygen.

These were the only trials for decades showing positive effects of LTOT in a very selected group of patients with COPD (with cardiac comorbidity), and with the most positive effects in patients with mild hypercapnia. , Other trials assessing LTOT in patients with moderate hypoxemia (moderate resting, isolated moderate exercise-induced or isolated sleep-related hypoxemia) were not able to reproduce LTOT beneficial effects on survival, or on pulmonary hemodynamics. , In a recent randomized controlled trial (RCT) (The Long-Term Oxygen Treatment Trial [LOTT]), LTOT was not associated with a reduction in the combined endpoint (death and first hospitalization) event rate in patients with moderate resting or exercise-induced moderate hypoxemia. However, this trial has some serious limitations, making it difficult to form clear conclusions about the value of LTOT in moderate hypoxemia.

To the best of our knowledge, the populations with moderate hypoxemia and polycythemia and/or PH that were included in the NOTT and MRC trials were arbitrarily defined, and the reduction in mortality in patients with COPD with these characteristics is still not clearly established. Nevertheless, according to expert opinion, LTOT is indicated in these situations.

Other clinical benefits of LTOT include improved QoL (but with some concerns regarding the negative aspects of cylinder gas use, such as social stigma, social isolation, lack of self-confidence, depression and fear of dependence), , and reduced breathlessness , (partially due to a placebo effect).

Negative Effects of Supplemental Oxygen Therapy and Hyperoxia in Chronic Obstructive Pulmonary Disease

The adverse consequences of SOT and higher than normal oxygen levels need to be taken into account when prescribing LTOT. In his book Evidence-Based Critical Care , Paul E. Marik wrote the following under the heading “ Too much oxygen kills ”: “ It is essentially impossible to get a Pa o 2 of much about 100 mm Hg if you only have 21% oxygen to breathe, which is all we had for millennia. There’s no evolutionary response to deal with hyperoxia.

Oxygen toxicity has been the subject of many studies, and it is now well known that hyperoxia leads to the formation of reactive oxygen species. Although these have positive roles in mediating cell signaling and enhancing phagocytic antimicrobial action and inflammatory response, high concentrations cause cellular damage and degrade bioactive nitric oxide, triggering cellular necrosis and apoptosis. Breathing supplemental oxygen in high concentrations for a long time may be associated with toxic side effects ( Table 3 ). These side effects, especially the pulmonary toxicities, are obviously very counterproductive in patients with COPD.

Table 3

Toxic effects of hyperoxia and hyperoxia-generated oxygen-derived radicals

Data from Refs. , ,

Neurologic toxicity , , Visual disturbance
Ear problems
Retina damage
Decreased cerebral blood flow
Ischemia-induced brain damage
Pulmonary toxicity , , Tracheobronchial irritation with impaired mucociliary clearance
Decline in lung function
Resorptive atelectasis
Alveolar protein leakage
Enhanced expression of leukotrienes by alveolar macrophages
Increases in alveolar neutrophils
Diffuse alveolar damage
Acute respiratory distress syndrome
Chronic pulmonary fibrosis and emphysema
Systemic toxicity , Increased vascular resistance and decreased cardiac output
Impaired innate immune responses resulting in increased susceptibility to infection

Oxygen inhalation might precipitate or worsen hypercapnic ventilatory failure via several mechanisms. First, some patients with COPD (those with chronic hypercapnia) are largely dependent on hypoxemia for stimulating breathing via peripheral chemoreceptor activity (carotid bodies). Removal of the hypoxemia stimulus reduces the work of breathing (WOB) despite worsening Pa co 2 (arterial partial pressure of carbon dioxide) values. , As hypercapnia worsens, this stimulus loses its natural ability to increase ventilatory drive via central chemoreceptors in upper brainstem in most patients with COPD. , Second, oxygen reduces airway resistance and improves respiratory system impedance, which itself can contribute to reduced ventilatory drive. Other mechanisms may also have roles in oxygen-induced hypercapnia in COPD, including the Haldane effect, which refers to oxygen displacing CO 2 from hemoglobin, and (more importantly) impaired hypoxic pulmonary vasoconstriction (Euler-Liljestrand mechanism) with a resulting increase in ventilation/perfusion (V/Q) mismatch and an increase in CO 2 content in shunted blood. , ,

P. Marik mentioned the term “ permissive hypoxemia” in the following context : “ hyperoxia appears to be more harmful to the host than hypoxemia. The human has developed elaborate mechanisms to compensate for hypoxemia and may tolerate permissive hypoxemia with minimal adverse effects. However, the human has not evolved to deal with hyperoxia which appears to be associated with significant adverse effects.” He also mentioned that “ there appears to be a dose-dependent association between supranormal oxygen tension and risk of adverse outcomes.

Long-Term Oxygen Therapy Indications in Chronic Obstructive Pulmonary Disease

It is first important to note that oxygen is a treatment for hypoxemia, not breathlessness. Thus, although improvements in dyspnea and exercise tolerance could be important benefits of LTOT, LTOT remains an expensive treatment modality and, like other medical treatments, has potential side effects. On the other hand, determining the “safe degree of hypoxemia” for an individual subject is exceedingly difficult. Interestingly, Magnet and colleagues noted that the NOTT and MRC trials were performed in the late 1970s, when patients with COPD had fewer comorbidities than they do currently and therapeutic approaches were different, meaning that the findings of these studies might not necessarily be applicable today.

When prescribing LTOT it is important to consider current guidelines and to make a careful risk-benefit evaluation. Guidelines are useful in some areas, but lacking in others. Furthermore, some need to be updated with the latest evidence. However, the current recommendations regarding LTOT indications are summarized in Tables 4 , 5 and Fig. 1 .

Table 4

Indications for LTOT in COPD from different published guidelines/recommendations

Data from Refs. , , , ,

Recommending Society 1st Indication: Severe Resting Hypoxemia (Classical Criteria) 2nd Indication: Milder Hypoxemia
BTS Pa o 2 ≤55 mm Hg (≤7.3 kPa)

  • Pa o 2 >55–60 mm Hg (>7.3–8.0 kPa) plus:

    • peripheral edema

    • secondary polycythemia (HCT ≥55%), or

    • evidence of PH

ACCP Pa o 2 ≤55 mm Hg (≤7.3 kPa)

  • Pa o 2 55–59 mm Hg (7.3–7.8 kPa) plus:

    • dependent edema

    • P-pulmonale on ECG (>3 mm in standard lead II, III, or aVF), or

    • HCT >56%


  • Pa o 2 55–59 mm Hg or SaO 2 <89% plus:

    • Congestive heart failure,

    • ECG evidence of right atrial enlargement, or

    • HCT ≥55%


  • Pa o 2 56–59 mm Hg (7.4–7.8 kPa) plus evidence of hypoxic organ damage, including

    • Clinical, electrocardiographic, or echocardiographic evidence of PH, or

    • Episodes of right heart failure, or

    • HCT >55%

GOLD Pa o 2 ≤55 mm Hg (≤7.3 kPa) or SaO 2 ≤88%

  • Pa o 2 55–60 mm Hg (7.3–8.0 kPa) or SaO 2 88% plus

    • Peripheral edema

    • Evidence of PH, or

    • HCT >55%

ACP Pa o 2 ≤55 mm Hg (≤7.3 kPa) or SaO 2 ≤88%

  • Pa o 2 55–59 mm Hg plus

    • Cor pulmonale, or

    • Polycythemia

Abbreviations: ACCP, American College of Chest Physicians; ACP, American College of Physicians; ATS, American Thoracic Society; BTS, British Thoracic Society; DGP, Deutsche Gesellschaft für Pneumologie und Beatmungsmedizin; ECG, electrocardiogram; GOLD, Global Initiative for Chronic Obstructive Lung Disease; HCT, hematocrit; PaO 2 , arterial partial pressure of oxygen; PH, pulmonary hypertension; SaO 2 , arterial oxygen saturation; TSANZ, Thoracic Society of Australia and New Zealand.

Table 5

Indications of AOT and NOT

Data from Refs. , , , , , ,


  • AOT should only be offered to patients already on LTOT if they are

    • Mobile outdoors (so they can achieve the required duration of daily usage), or

    • Too symptomatic to leave home without AOT.

  • Two of 3 of the following criteria should be met to prescribe AOT (expert consensus, no solid evidence):

    • SpO2 can be kept constantly >90% during exercise with oxygen

    • Improvement of 6MWT by ≥10%

    • Improvement of dyspnea (≥1 point improvement on the Borg scale)

  • Regarding patients with COPD with exercise-induced desaturation who do not have the classic criteria at rest:

    • They probably do not benefit from AOT regarding mortality, exercise capacity or dyspnea during exercise , ,

    • Because AOT transiently improves their exercise capacity, they should get AOT as part of a rehabilitation program. , ,


  • Regarding patients who do not meet the classic LTOT criteria at rest,

    • NOT is not recommended in most guidelines

    • In many trials, NOT did not improve pulmonary hemodynamics, survival, QoL, or sleep quality , ,

    • An ongoing trial (INOX; NCT01044628 ) is evaluating the effects of NOT in patients with COPD; results are expected soon

    • NOT can be considered for hypoxemic patients receiving LT-NIV for treatment of hypercapnic respiratory failure, but the evidence and practical details are lacking.

Abbreviations: 6MWT, six-minute walk test; AOT, ambulatory oxygen therapy; COPD, chronic obstructive pulmonary disease; LT-NIV, long-term noninvasive ventilation; LTOT, long-term oxygen therapy; NOT, nocturnal oxygen therapy; QoL, quality of life; SpO 2 , oxygen saturation on pulse oximeter.

Although respiratory failure is usually defined as an arterial partial pressure of oxygen (Pa o 2 ) less than 60 mm Hg, in reality it is the arterial oxygen saturation (SaO 2 ) that should be used to make clinical decisions and to titrate SOT. As shown by the arterial oxygen content equation ( Box 2 ), oxygen delivery is almost entirely dependent on SaO 2 and cardiac output rather than Pa o 2 (which contributes to the small amount of oxygen dissolved in serum).

Box 2

Arterial oxygen content and arterial oxygen delivery equations

Abbreviations: CaO 2 , oxygen content of arterial blood; DaO 2, arterial oxygen delivery; Hb, hemoglobin level; Pa o 2 , arterial partial pressure of oxygen; SaO2, arterial oxygen saturation.

CaO 2 [mL/L] = (Hb [g/dL] × 1.34 [mL] × SaO 2 [%] ) + (Pa o 2 [mm Hg] × 0.031)

DaO 2 [mL/min] = CaO 2 [mL/L] × cardiac output [L/min]

1.34 = oxygen combining capacity of hemoglobin.

0.031 = solubility coefficient of oxygen in human plasma.

Further details about the diagnostic approach before prescribing LTOT, and about titration of therapeutic oxygen flow rates are presented in Tables 6 and 7 .

Table 6

Timing of LTOT initiation and first-line diagnostics

Data from Refs. , , , , ,

When to Initiate LTOT?
Timing of LTOT initiation in stable patients

  • Patients should undergo 2 blood gas analyses at least 3 wk apart during a period of apparent clinical stability and while receiving adequate COPD therapy (expert opinion based on MRC protocol, no data from controlled trials)

LTOT in post-exacerbation COPD patients

  • After an exacerbation, patients should not be discharged from hospital with LTOT unless they are breathless, cannot manage without oxygen, and have an SpO 2 <92% while breathing room air

  • Even then, the LTOT indication should be reassessed after ≥8 wk of stability (patients who become hypoxemic during COPD exacerbation could have improvements in gas exchange after 8 wk) , , ,

What is the best first-line diagnostic tool?

  • Recommended as first-line diagnostic tool for hypoxemia , ,

  • Not always practical, especially in an ambulatory setting ,


  • Underestimates Pa o 2 (patients could receive LTOT inappropriately)

  • A combination of CBG and oximetry can be used, even as primary approach ,

Pulse oximetry

  • Pulse oximeters have a bias of 0.2%–1% and a standard deviation (precision) of <2% when SaO 2 is >80% ,

  • SpO 2 of ≤92% can be safely used for screening patients (SpO 2 ≤94% with clinical evidence of peripheral edema, polycythemia or PH)

  • Oximetry cannot be used alone for formal assessment without at least a CBG analysis ,

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Aug 16, 2020 | Posted by in GENERAL | Comments Off on Oxygen Therapy and Noninvasive Ventilation in Chronic Obstructive Pulmonary Disease

Full access? Get Clinical Tree

Get Clinical Tree app for offline access