Acute exacerbation of COPD: quick reference

Key messages

Two things set the COPD exacerbation apart from many other causes of acute dyspnoea: oxygen must be titrated cautiously and the blood gas determines the need for ventilatory support. At the same time, exacerbation is a clinical syndromic diagnosis, not a sufficient explanation for all acute dyspnoea in a patient with COPD. Pneumonia, heart failure, pulmonary embolism, pneumothorax and acute coronary syndrome must be actively considered [1].

Do the following early:

1. Assess airway, work of breathing, level of consciousness and circulation.
2. Give controlled oxygen with a target of 88 to 92 percent until the risk of hypercapnia has been assessed.
3. Take an arterial blood gas in respiratory compromise, low saturation, suspected carbon dioxide retention, reduced consciousness, or when NIV is being considered.
4. Give short-acting bronchodilators.
5. Give a systemic corticosteroid if the exacerbation is more than mild.
6. Assess the indication for antibiotics on the basis of sputum purulence, the other cardinal symptoms and the severity of the illness.
7. Start NIV without undue delay in persistent hypercapnic respiratory acidosis, if invasive ventilation is not immediately required.
8. Document the ceiling of care and the decision on intubation early.

Initial assessment

Confirm that the presentation fits an exacerbation

A COPD exacerbation is an acute or subacute worsening of dyspnoea, cough and sputum in particular that exceeds the patient's normal variation and requires a change of treatment. Common precipitants are viral infection, bacterial infection and air pollution, but often no definite trigger can be identified [1,2].

Ask specifically about:

• Onset and time course.
• Increased dyspnoea compared with usual.
• Increased sputum volume and new sputum purulence.
• Fever, rigors and respiratory tract symptoms.
• Chest pain, haemoptysis, syncope and palpitations.
• Orthopnoea, weight gain and peripheral oedema.
• Previous hypercapnia, NIV or invasive ventilation.
• Home oxygen and the patient's usual saturation.
• Previous exacerbations and hospital admissions.
• Current inhaled treatment and adherence.
• Opioids, benzodiazepines and other sedating drugs.
• Thromboembolic risk, for example immobilisation, cancer and previous venous thromboembolism.
• A documented decision on the ceiling of care.

Signs of a serious condition

Oxygen targets

Give oxygen if the patient is hypoxaemic. Oxygen is not a treatment for dyspnoea as such and must not be given uncontrolled simply because the patient feels breathless.

Excessive oxygen delivery can raise pCO₂. The main mechanisms are impaired ventilation–perfusion matching, increased physiological dead space and the Haldane effect. It is therefore misleading to explain the whole phenomenon as abolition of the patient's hypoxic respiratory drive.

In a randomised prehospital trial, mortality was 9 percent with high-flow oxygen and 4 percent with titrated oxygen. Among patients with spirometry-confirmed COPD the corresponding mortality was 9 and 2 percent respectively. Titrated oxygen also reduced hypercapnia and respiratory acidosis [3]. A Cochrane review found the same direction of effect but judged the evidence limited, since the conclusion rested mainly on a single trial [4].

Practical delivery

• Usually start with a Venturi mask, for example 24 or 28 percent, or a low-flow nasal cannula.
• Titrate to 88 to 92 percent while awaiting the blood gas.
• Where possible use compressed air, not oxygen, to drive the nebuliser in patients at risk of hypercapnia. Give separately titrated oxygen.
• Do not abruptly reduce oxygen to zero in a severely hypoxaemic patient. Reduce it in a controlled way and follow the saturation.
• Take a repeat blood gas about 30 to 60 minutes after a clinically significant change in oxygen therapy, and sooner if the patient deteriorates.
• If pCO₂ rises but pH is stable, consider whether this represents chronic compensated hypercapnia.
• If pH falls at the same time as pCO₂ rises, alveolar ventilation is deteriorating. Consider NIV immediately.

Blood gas

When is an arterial blood gas needed?

Take an arterial blood gas in everyone with clearly compromised breathing, and particularly in:

• Saturation below target.
• A need for oxygen.
• Marked tachypnoea or use of accessory respiratory muscles.
• Fatigue, confusion or other impairment of consciousness.
• Previous hypercapnic respiratory failure.
• Suspicion of a metabolic component.
• Consideration of NIV or invasive ventilation.
• Clinical deterioration despite treatment.

The arterial blood gas is the reference method when oxygenation, ventilatory failure and the indication for NIV are to be assessed. A venous blood gas can be used as an initial screening test for pH and hypercapnia in some more stable patients, but venous pO₂ cannot be used to assess oxygenation. A normal venous pCO₂ and a normal venous pH make marked hypercapnic acidosis less likely, whereas abnormal values need arterial confirmation. The evidence for the venous blood gas as an adequate substitute is uncertain and its specificity for hypercapnia is low [5].

Interpretation at a glance

Compare with earlier blood gases where possible. A single high pCO₂ is less alarming in an alert patient with known chronically raised pCO₂ and a normal pH than a rapidly rising pCO₂ with a falling pH.

Acute treatment

Bronchodilators

Give an inhaled short-acting beta-2 agonist. Usually add ipratropium in a moderate or severe exacerbation. Doses and dosing intervals should follow the local emergency protocol.

A nebuliser is practical in severe dyspnoea, poor coordination or when NIV is needed at the same time. A metered-dose inhaler with a spacer is an adequate alternative for patients who can cooperate. A Cochrane review found no definite clinically important difference between nebuliser and metered-dose inhaler with a spacer, but the evidence base was small and of low quality [6].

Bear the following in mind:

• Oxygen-driven nebulisation can deliver an unintentionally high oxygen concentration. Use compressed air if possible.
• Tachycardia, tremor and hypokalaemia can be caused or aggravated by repeated beta-2 agonists.
• Follow potassium and the ECG with frequent doses, arrhythmia or concomitant diuretic treatment.
• A dry powder inhaler is unsuitable if the patient cannot generate a sufficient inspiratory flow.
• Methylxanthines are not recommended routinely because of limited benefit and the risk of toxicity.
• Intravenous magnesium is not standard treatment in a COPD exacerbation. The evidence is insufficient [7].

Systemic corticosteroids

Systemic corticosteroids reduce treatment failure and relapse and hasten improvement in lung function and symptoms. In a Cochrane review about nine patients needed to be treated to prevent one treatment failure. Treatment did not, however, reduce short-term mortality and increased the risk of adverse effects, particularly hyperglycaemia [8].

Oral treatment is first choice when intestinal absorption is reliable. Intravenous treatment has not been shown to be more effective than oral treatment and may cause more hyperglycaemia [8]. Intravenous treatment is therefore reserved mainly for patients who cannot swallow, are vomiting, have suspected malabsorption or are critically ill.

Five days is usually sufficient. In the REDUCE trial, prednisone 40 mg daily for 5 days was non-inferior to 14 days with respect to a further exacerbation during the following 6 months, while total steroid exposure was considerably lower [9]. A later Cochrane review likewise found no definite advantage of 10 to 14 days compared with about 5 days [10].

The exact Swedish dose may differ between regional recommendations and between the choice of prednisolone and betamethasone. Follow the local protocol. Tapering is not normally needed after a short five-day course, but individual assessment is required with long-term steroid treatment or suspected adrenal suppression.

Monitor in particular:

• Blood glucose.
• Delirium, sleep disturbance and psychiatric adverse effects.
• Fluid retention and blood pressure.
• Cumulative steroid exposure with frequent exacerbations.

Eosinophil-guided steroid treatment can reduce steroid exposure. In CORTICO-COP the median duration of treatment fell from 5 to 2 days with no difference in the primary outcome, but the strategy is not yet obvious routine practice in Swedish emergency care and is best suited to a validated local protocol [11]. Inhaled corticosteroids cannot be recommended generally as a substitute for systemic treatment in a severe exacerbation, since the evidence is still uncertain [12].

The indication for antibiotics

Antibiotics should not be given routinely in every COPD exacerbation. A bacterial cause is more likely with new sputum purulence, above all together with increased dyspnoea or increased sputum volume.

Antibiotics are usually given in:

• Increased dyspnoea, increased sputum volume and purulent sputum.
• Purulent sputum plus one of the two other cardinal symptoms.
• A need for invasive or non-invasive ventilatory support.
• Pneumonia, which is then treated according to the pneumonia recommendation.
• Another strong clinical suspicion of bacterial infection in a severely ill patient.

The overall benefit is greatest in very severe exacerbations and in those requiring intensive care. In a Cochrane review antibiotics reduced treatment failure in outpatients, whereas the effect in other inpatients was uncertain. In intensive care patients there was a clear reduction in treatment failure and mortality, but this analysis rested on a single small trial [13]. A more recent meta-analysis found an overall lower risk of treatment failure with antibiotics but no definite mortality benefit [14].

Empirical treatment

The table reproduces the stated recommendation from Strama Stockholm 2025/2026. Local protocol, allergy, renal function, culture results and the regional resistance situation take precedence. The Swedish recommendation could not be verified in the international databases underlying the other references in this article.

Purulent sputum alone in a patient with otherwise healthy lungs is not in itself an indication for antibiotics. In a patient with confirmed COPD, purulence is by contrast an important part of the antibiotic assessment, particularly together with increased dyspnoea, increased sputum volume or a need for ventilatory support.

When are cultures needed?

Take a sputum culture before antibiotics where possible in:

• A severe exacerbation or a need for intensive care.
• Bronchiectasis.
• Frequent exacerbations or repeated courses of antibiotics.
• Recent hospital care.
• Previous isolation of resistant Gram-negative bacteria or Pseudomonas aeruginosa.
• Treatment failure.
• Immunosuppression.

Blood cultures are taken in sepsis, high fever, rigors or severe pneumonia. Viral testing may be relevant during the season and may influence isolation and antiviral treatment, but a positive viral test does not exclude concomitant bacterial infection.

CRP can support the overall assessment but must not by itself determine antibiotic treatment. Procalcitonin-guided treatment has been studied, but the evidence does not support routinely replacing clinical assessment with a single biomarker value [1].

Non-invasive ventilation

Bilevel NIV is the first-line treatment in acute hypercapnic respiratory failure due to a COPD exacerbation, provided the patient does not need immediate intubation.

NIV reduces both mortality and the need for intubation. In a Cochrane review of 17 randomised trials, mortality fell by 46 percent and the need for intubation by 65 percent. About 12 patients needed to be treated to prevent one death and 5 to prevent one intubation [15].

Starting and monitoring

• Start NIV in a monitored setting with staff who can adjust the treatment and intubate promptly if needed.
• Sit the patient upright.
• Choose a mask that gives as little leak as possible without causing unnecessary pressure or claustrophobia.
• Start with tolerable pressures and titrate the inspiratory support according to tidal volume, respiratory rate, comfort, leak and blood gas.
• Titrate sufficient expiratory pressure to counteract airway collapse and intrinsic PEEP, but avoid excessive pressure and hyperinflation.
• Give oxygen through the circuit and maintain the target of 88 to 92 percent.
• Continue bronchodilators, corticosteroid and other treatment of the underlying cause.
• Use continuous or frequent monitoring of saturation, respiratory rate, pulse, blood pressure and level of consciousness.
• Take a repeat blood gas after about 1 hour, at the latest within 2 hours, and immediately on deterioration.

Signs of a response to NIV are a rising pH, falling pCO₂, a lower respiratory rate, less accessory muscle use and improved alertness. Improvement in pH may be seen as early as one hour [15].

Low pH

The lower the pH, the higher the risk of NIV failure. A pH below 7.25 is not in itself an absolute reason to withhold NIV, but treatment should then be given where immediate intubation is possible, if intubation is part of the goal of care. NIV must not be used in a way that delays a necessary intubation.

Contraindications and obstacles

Strong contraindications, or situations where invasive ventilation should usually take priority, are:

• Respiratory arrest or agonal breathing.
• Inability to protect the airway.
• Active vomiting or a very high risk of aspiration.
• Severe haemodynamic instability.
• Facial burns, extensive facial trauma or an impossible mask fit.
• An undrained pneumothorax.
• Marked sputum retention that the patient cannot manage.
• An immediate need for intubation for another reason.

Impaired consciousness is an important warning sign but not always an absolute contraindication. A patient who is drowsy from reversible hypercapnia can respond rapidly to NIV, provided the airway can be protected and monitoring is adequate.

Signs of NIV failure

• Falling pH or failure of pH to improve.
• A continuing rise in pCO₂.
• Worsening hypoxaemia.
• Increasing fatigue or falling level of consciousness.
• Haemodynamic instability.
• Inability to tolerate the mask despite correctable measures.
• Large leaks that cannot be remedied.
• Secretion problems or aspiration.
• A new pneumothorax.

If there is no improvement within 1 to 2 hours, reassess mask fit, leak, pressure settings, secretions, pneumothorax and differential diagnoses. Then make an active decision about continued NIV, intubation or symptomatic treatment within a documented goal of care.

High-flow nasal cannula

A high-flow nasal cannula can be more comfortable than NIV and can reduce dead space, but it should not routinely replace NIV in hypercapnic respiratory acidosis. A meta-analysis of nine randomised trials found no definite difference in mortality or intubation, but the high-flow nasal cannula showed a tendency towards more treatment failure, 24.4 percent compared with 15.4 percent, despite better tolerance [16]. NIV therefore remains first choice in an acidotic hypercapnic COPD exacerbation.

Invasive ventilation

Consider intubation in:

• Respiratory arrest.
• Severe impairment of consciousness with an unprotected airway.
• Refractory hypoxaemia.
• Severe or rapidly progressive acidosis despite adequate NIV.
• Haemodynamic instability.
• NIV intolerance where ventilatory support is still needed.
• Marked sputum retention or aspiration.
• Another concomitant diagnosis requiring invasive ventilation.

Severe COPD is not in itself a reason to withhold invasive ventilation. The decision should be based on reversibility, functional level before the deterioration, comorbidity, frailty, previous course, the patient's wishes and the likelihood of returning to an acceptable quality of life.

Ceiling-of-care decisions

Make and document the decision on intensive care, intubation and resuscitation early, together with the patient where possible. Look for previously documented decisions and involve relatives if the patient lacks capacity.

A decision not to intubate does not automatically mean that NIV should be withheld. NIV can be used as the ceiling of treatment if the aim is to treat potentially reversible hypercapnic failure. If NIV does not improve the condition and intubation is not part of the goal of care, the focus should shift to effective symptom relief and good communication.

Red flags and differential diagnoses

A patient may have both a COPD exacerbation and another acute illness.

Pulmonary embolism is particularly important not to dismiss. In a meta-analysis the pooled prevalence was about 18 percent among the patients studied with a suspected COPD exacerbation, but the estimate varied widely with the population selected and the diagnostic strategy. Where CT was performed only after clinical selection, the prevalence was considerably lower than when everyone was investigated [17]. The finding argues for clinical probability assessment, not routine CT pulmonary angiography in everyone.

Investigations

Baseline investigations in an exacerbation requiring hospital care

Consider or usually take:

• Arterial blood gas.
• Full blood count.
• CRP.
• Electrolytes and creatinine.
• Glucose.
• ECG.
• Chest radiograph.
• Viral testing according to season and clinical picture.
• Sputum culture in a severe or complicated exacerbation.
• Troponin where myocardial injury or ischaemia is suspected.
• NT-proBNP where heart failure is suspected.

D-dimer should be used after clinical probability assessment. A COPD exacerbation, inflammation and old age can all give non-specifically raised values. Use a validated diagnostic strategy according to local protocol.

A chest radiograph can demonstrate pneumonia, pneumothorax, pulmonary congestion and pleural effusion. A normal chest radiograph does not exclude pulmonary embolism. Lung ultrasound can be a valuable adjunct for pneumothorax, interstitial oedema and pleural effusion where the expertise is available.

Spirometry should not be used to grade the acute exacerbation. If the diagnosis of COPD has never been confirmed, spirometry should be planned for when the patient is clinically stable again.

Admission or discharge

Factors favouring admission

• Marked or persistent dyspnoea after initial treatment.
• A new or increased oxygen requirement.
• Respiratory acidosis or a need for NIV.
• New confusion.
• Haemodynamic compromise or significant arrhythmia.
• Pneumonia, pneumothorax, heart failure or another serious concomitant diagnosis.
• Inadequate ability to eat, drink, walk or manage medication.
• Significant comorbidity or frailty.
• Repeated emergency attendances or recent hospital admission.
• Inadequate support at home.

Previous hospital admission is one of the clearest risk markers for further admission. Reported readmission rates vary widely between populations and healthcare systems, but readmission is consistently linked to worse survival and quality of life [18].

Discharge can be considered when

• Dyspnoea has improved to a manageable level.
• The saturation is stable on air or on the patient's usual oxygen.
• There is no respiratory acidosis.
• The patient can walk, eat and sleep at an acceptable level.
• Inhaled treatment can be taken correctly.
• Medication and follow-up have been arranged.
• The patient understands when and where to seek further care.
• Serious differential diagnoses have been sufficiently assessed.

Before discharge

Use a structured discharge check. Discharge bundles usually cover smoking cessation, optimised treatment, rehabilitation and continuity of follow-up, even if trial results with respect to readmission are not entirely consistent [19,20].

Check the following:

• Inhaler technique has been observed, corrected and demonstrated.
• The patient has access to working inhalers and a spacer where one is needed.
• Maintenance treatment has been restarted or optimised.
• The indication for an inhaled corticosteroid has been weighed against exacerbation history, eosinophils and the risk of pneumonia.
• Smoking cessation has been offered with counselling and pharmacological support.
• A written exacerbation action plan has been provided.
• Any rescue prescription for a corticosteroid or antibiotics follows local practice and is combined with clear criteria for starting it.
• Vaccination status has been assessed.
• Nutrition, physical function, falls risk and frailty have been assessed where needed.
• The patient's mental health and cognitive ability have been considered.
• The need for oxygen has been assessed without prescribing long-term oxygen on the sole basis of transient hypoxaemia during the exacerbation.
• Ceiling-of-care decisions have been documented where relevant.
• Follow-up and points of contact are booked, not merely recommended.

A self-management plan alone after discharge has not been shown to reduce mortality or all hospital admissions reliably, but it can improve disease-specific quality of life. Support therefore needs to be tailored to the patient's ability and combined with active clinical follow-up [21].

Follow-up and rehabilitation

Plan early clinical follow-up after an exacerbation requiring hospital care, often within a few weeks and sooner for frail patients, patients with a new oxygen requirement or a high risk of readmission.

At follow-up the following should be assessed:

• Symptoms and recovery.
• Saturation and any persisting hypercapnia.
• Inhaler technique and adherence.
• Maintenance treatment.
• Smoking and exposures.
• Exacerbation frequency.
• Comorbidity, particularly cardiac disease, anxiety and depression.
• The need for spirometry if the diagnosis is not firmly confirmed.
• The indication for long-term oxygen once the patient is stable.
• The need for long-term NIV in persistent marked hypercapnia, after specialist assessment.
• Ceiling of care and future treatment preferences.

Offer pulmonary rehabilitation after an exacerbation requiring hospital care. A meta-analysis of 20 randomised trials found reduced readmission, improved walking distance, less dyspnoea and better quality of life when rehabilitation was started during the admission or within 4 weeks of discharge. No definite mortality benefit could be shown, and several trials had a high risk of bias [22]. An earlier meta-analysis also found lower mortality and readmission, but the results should be interpreted in the light of small trials and varying programmes [23].

Long-term oxygen

Assessment for long-term oxygen should be made when the patient is stable. Hypoxaemia during an acute exacerbation may improve after treatment. Patients discharged on temporary oxygen therefore need planned reassessment. Oxygen should be prescribed on the basis of a blood gas and established criteria, not on dyspnoea alone.

Alternative models of care

Hospital at home can be an alternative for carefully selected patients without acidosis, unstable comorbidity, a need for invasive investigation or inadequate social support. Such programmes require rapid access to clinical reassessment and clear criteria for transfer back to hospital [24].

Common mistakes

• Giving high-flow oxygen without a target prescription.
• Interpreting a normal saturation after a lot of oxygen as meaning the patient is stable.
• Missing respiratory acidosis because the patient looks relatively calm.
• Allowing oxygen-driven nebulisation to raise the inspired oxygen concentration uncontrollably.
• Giving intravenous corticosteroid routinely despite adequate oral treatment.
• Prolonging the steroid course without a specific reason.
• Giving antibiotics for every exacerbation.
• Delaying NIV when pH is falling and pCO₂ rising.
• Continuing ineffective NIV for so long that a necessary intubation is delayed.
• Using a high-flow nasal cannula as a routine substitute for NIV in respiratory acidosis.
• Assuming that all acute dyspnoea in a patient with COPD is caused by COPD.
• Missing pneumothorax, pulmonary embolism, heart failure or acute coronary syndrome.
• Discharging without checked inhaler technique, a rehabilitation plan and a booked follow-up.
• Newly prescribing long-term oxygen during an unstable phase without planned reassessment.
• Postponing ceiling-of-care decisions until the patient can no longer take part in the conversation.

References

1. Doğan NÖ et al. 2021 Guideline for the Management of COPD Exacerbations: Emergency Medicine Association of Turkey and Turkish Thoracic Society Clinical Practice Guideline Task Force. Turkish Journal of Emergency Medicine 2021. PMID: 34849428
2. Crisafulli E et al. Management of severe acute exacerbations of COPD: an updated narrative review. Multidisciplinary Respiratory Medicine 2018. PMID: 30302247
3. Austin MA et al. Effect of high flow oxygen on mortality in chronic obstructive pulmonary disease patients in prehospital setting: randomised controlled trial. BMJ 2010. PMID: 20959284
4. Kopsaftis Z et al. Oxygen therapy in the pre-hospital setting for acute exacerbations of chronic obstructive pulmonary disease. Cochrane Database of Systematic Reviews 2020. PMID: 31934729
5. Byrne AL et al. Peripheral venous blood gas analysis for the diagnosis of respiratory failure, hypercarbia and metabolic disturbance in adults. Cochrane Database of Systematic Reviews 2025. PMID: 40557830
6. van Geffen WH et al. Bronchodilators delivered by nebuliser versus pMDI with spacer or DPI for exacerbations of COPD. Cochrane Database of Systematic Reviews 2016. PMID: 27569680
7. Ni H et al. Magnesium sulfate for acute exacerbations of chronic obstructive pulmonary disease. Cochrane Database of Systematic Reviews 2022. PMID: 35616126
8. Walters JAE et al. Systemic corticosteroids for acute exacerbations of chronic obstructive pulmonary disease. Cochrane Database of Systematic Reviews 2014. PMID: 25178099
9. Leuppi JD et al. Short-term vs conventional glucocorticoid therapy in acute exacerbations of chronic obstructive pulmonary disease: the REDUCE randomized clinical trial. JAMA 2013. PMID: 23695200
10. Walters JAE et al. Different durations of corticosteroid therapy for exacerbations of chronic obstructive pulmonary disease. Cochrane Database of Systematic Reviews 2018. PMID: 29553157
11. Sivapalan P et al. Eosinophil-guided corticosteroid therapy in patients admitted to hospital with COPD exacerbation: the CORTICO-COP trial. Lancet Respiratory Medicine 2019. PMID: 31122894
12. Papadopoulou E et al. Inhaled versus systemic corticosteroids for acute exacerbations of COPD: a systematic review and meta-analysis. European Respiratory Review 2024. PMID: 38508668
13. Vollenweider DJ et al. Antibiotics for exacerbations of chronic obstructive pulmonary disease. Cochrane Database of Systematic Reviews 2018. PMID: 30371937
14. Suzuki Y et al. Antibiotic treatment for patients with exacerbation of chronic obstructive pulmonary disease: a systematic review and meta-analysis. Respiratory Investigation 2024. PMID: 38761481
15. Osadnik CR et al. Non-invasive ventilation for the management of acute hypercapnic respiratory failure due to exacerbation of chronic obstructive pulmonary disease. Cochrane Database of Systematic Reviews 2017. PMID: 28702957
16. Zhong Z et al. Comparison of clinical outcomes between high-flow nasal cannula and non-invasive ventilation in acute exacerbation of COPD: a meta-analysis of randomized controlled trials. BMC Pulmonary Medicine 2025. PMID: 40859250
17. Li M et al. The prevalence and risk factors of pulmonary embolism in patients with chronic obstructive pulmonary disease: a systematic review and meta-analysis. Thrombosis Journal 2025. PMID: 40301960
18. Njoku CM et al. Risk factors and associated outcomes of hospital readmission in COPD: a systematic review. Respiratory Medicine 2020. PMID: 33190738
19. Miravitlles M et al. Implementing an Evidence-Based COPD Hospital Discharge Protocol: A Narrative Review and Expert Recommendations. Advances in Therapy 2023. PMID: 37537515
20. Salvi S et al. A 7-point evidence-based care discharge protocol for patients hospitalized for exacerbation of COPD: consensus strategy and expert recommendation. NPJ Primary Care Respiratory Medicine 2024. PMID: 39706845
21. Majothi S et al. Supported self-management for patients with COPD who have recently been discharged from hospital: a systematic review and meta-analysis. International Journal of Chronic Obstructive Pulmonary Disease 2015. PMID: 25995625
22. Lu HY et al. Effects of Early Pulmonary Rehabilitation on Hospitalized Patients with Acute Exacerbation of Chronic Obstructive Pulmonary Disease: A Systematic Review and Meta-Analysis. International Journal of Chronic Obstructive Pulmonary Disease 2023. PMID: 37215744
23. Ryrsø CK et al. Lower mortality after early supervised pulmonary rehabilitation following COPD-exacerbations: a systematic review and meta-analysis. BMC Pulmonary Medicine 2018. PMID: 30219047
24. McCurdy BR. Hospital-at-home programs for patients with acute exacerbations of chronic obstructive pulmonary disease: an evidence-based analysis. Ontario Health Technology Assessment Series 2012. PMID: 23074420
25. Vitacca M et al. Use of inhaled devices during a hospital exacerbation of COPD: a summary of an interdisciplinary audit. Monaldi Archives for Chest Disease 2020. PMID: 32072800