Intended for healthcare professionals

Education And Debate

Oxygen treatment for acute severe asthma

BMJ 2001; 323 doi: (Published 14 July 2001) Cite this as: BMJ 2001;323:98
  1. David Inwald (D.Inwald{at}, MRC clinical training fellowa,
  2. Mark Roland, clinical research fellowb,
  3. Lieske Kuitert, consultant in respiratory medicinec,
  4. Sheila A McKenzie, consultant in paediatric respiratory medicined,
  5. Andy Petros, consultant in paediatric intensive care medicinee
  1. a Portex Department of Anaesthesia, Intensive Care and Respiratory Medicine, Institute of Child Health, London WC1N 1EH
  2. b Department of Respiratory Medicine, St Bartholomew's and the Royal London School of Medicine and Dentistry, London EC1A 7BE
  3. c Royal London Hospital, London E1 1BB
  4. d Queen Elizabeth Children's Service, Royal London Hospital, London E1 1BB
  5. e Paediatric Intensive Care Unit, Great Ormond Street Hospital, London WC1N 3JF
  1. Correspondence to: D Inwald
  • Accepted 8 March 2001

Deaths from asthma in England and Wales dropped by about 6% a year in people aged 5-64 from 1983 to 1995,1 but about 20 children and 1600 adults still die in the United Kingdom every year from acute asthma. Profound hypoxaemia may be a readily preventable cause of some of these deaths. The British Thoracic Society's asthma guidelines advise oxygen as first line treatment in hospital for all patients in cases of acute severe asthma.2 However, the guidelines do not advise treatment with oxygen in primary care in children and do not insist on its use in adults. The recent death of a child in a primary care setting after administration of salbutamol nebulised with air prompted us to question whether treatment with oxygen should be recommended in all cases of acute severe asthma, including those presenting in primary care.

Summary points

Asthma causes 1600 deaths in the United Kingdom every year

Progressive hypoxaemia is probably an important cause of death

Oxygen should be the first treatment given to any patient with acute severe asthma

Nebulisation of β2 agonists with air during severe attacks may worsen hypoxaemia

Patients with acute severe asthma should receive β2 agonists nebulised with oxygen


A systematic review was not possible as there have never been any randomised controlled trials of oxygen in acute severe asthma. We therefore present a traditional literature review and our opinion based on it. We selected 24 publications from our personal collections and from Medline searches. Using the search terms “asthma” and “hypoxaemia or hypoxemia or hypoxia or oxygen” and “salbutamol or albuterol” yielded 204 papers, of which 11 were included in this review. The other 13 papers that we evaluated were from our personal collections.

Results and discussion

In acute severe asthma, narrowing of the airway occurs as a result of bronchospasm, mucosal oedema, and hypersecretion. The homoeostatic response to this is to decrease blood flow to underventilated lung units, thus maximising oxygenation by matching pulmonary perfusion with alveolar ventilation. This was shown in an elegant series of studies spanning nearly two decades that used the multiple inert gas technique to study ventilation-perfusion relations and gas exchange in asthmatic patients with varying degrees of disease severity.35 The pattern of ventilation-perfusion is bimodal in acute severe asthma, ranging from normally perfused areas to areas of hypoxic pulmonary vasoconstriction. However, the percentage of cardiac output to areas with a low ventilation-perfusion ratio increases with worsening severity of the acute flare, ranging from 0.4% in chronic stable asthma to 28-36% in the most severe acute flares, 5 6 including those requiring mechanical ventilation.

Treatment with inhaled β2 agonists is often given to relieve bronchospasm and improve oxygenation. In acute severe asthma, nebulisation of β2 agonists without oxygen can cause or worsen hypoxaemia. The mechanism for this has been understood since 1967, when it was shown that isoprenaline, a β agonist then in common use for asthma, when nebulised with compressed air resulted in pulmonary vasodilatation, increasing perfusion to poorly ventilated lung units and ventilation-perfusion mismatch, and thus worsening hypoxaemia.7 Since then it has been found that salbutamol can also worsen ventilation-perfusion mismatch by causing pulmonary vasodilatation and increasing cardiac output.8

Findings in children

The first report of such effects in children was in 1969, when a significant fall in arterial blood oxygen saturation was found in asthmatic children after they inhaled salbutamol, even though this improved forced expiratory volume in one second in the same period.9 A later study recorded a fall in arterial oxygen saturation of more than 5% in nine out of 18 asthmatic children aged 2-15 years who were treated with salbutamol nebulised with air.10 A small randomised crossover study failed to show a significant decrease in oxygenation during nebulisation with air in 27 episodes of acute severe asthma; however, in 10 cases arterial oxygen saturation decreased by 2-6%, with the biggest drops occurring 10-15 minutes after nebulisation.11 More recently, in a study of 111 children with acute severe asthma, six children with pneumonic consolidation in addition to asthma who were treated with salbutamol nebulised with oxygen became profoundly hypoxaemic when oxygen was discontinued after nebulisation.12 Taken together, these studies provide strong evidence that a small number of children develop important hypoxaemia related to salbutamol administration during acute episodes of asthma if the drug is administered without oxygen.

Findings in adults

The findings in children contrast with those in adults. Over the past 30 years many studies have assessed the effects of salbutamol on oxygenation in asthmatic adults, either as a primary or secondary end point. 3 8 1318 Most of these studies were small, uncontrolled, and of widely varying design, and some had conflicting results. Most published data show that salbutamol does not have a clinically important effect on oxygenation in asthmatic adults. This seems to be true for both stable and acute asthma, across a range of doses administered by various routes. However, the studies are limited by their exclusion of the most severe exacerbations, particularly those accompanied by marked hypoxaemia.18 Even though the average change in oxygenation in response to salbutamol in these studies may not be significant, the fact that considerable variability exists among patients has ramifications for those in extremis. In 1974 Choo-Kang postulated that this variability might be due to variations between patients in the response of vascular smooth muscle to stimulation of β2 adrenoceptors,18 and it is now known that polymorphism of the β2 adrenoceptor gene is an important inheritable determinant of smooth muscle response to agonists in the airway.19

Features of severe asthma

Age 1-5 years
  • Tachycardia

  • Flaring

  • Use of accessory muscles

  • Recession

  • Head retraction

  • Inability to feed

Age >5 years
  • Tachycardia

  • Use of accessory muscles

  • Recession

  • Peak expiratory flow <50% of best

Features of life threatening asthma

Age 1-5 years
  • Cyanosis

  • Silent chest

  • Fatigue

  • Reduced level of consciousness

Age >5 years
  • Cyanosis

  • Inability to speak

  • Silent chest

  • Fatigue

  • Reduced level of consciousness

  • Peak expiratory flow <33% of best

Deaths from asthma

Most asthma deaths occur in the community, often in patients whose symptoms have been poorly controlled for days or even weeks before the fatal attack.20 Two hypotheses have been postulated for the cause of these deaths. Firstly, cardiac arrhythmias may contribute to some of the observed mortality, particularly in adults. The risk of arrhythmia is theoretically greatly increased by hypokalaemia and prolongation of the QTc interval, both well described side effects of β2 agonists and theophyllines.21 However, in a series of admissions of patients whose asthma attacks were nearly fatal, few arrhythmias other than sinus tachycardias and bradycardias were documented.22

A far more likely hypothesis is that deaths occur as a result of hypoxaemia.22 As the forced expiratory volume in one second falls below 25% of predicted (or peak expiratory flow rate is <30% of predicted), progressively worsening hypoxaemia occurs secondarily to ventilation-perfusion mismatch and alveolar hypoventilation.23 Giving air driven, high dose nebulised salbutamol to genetically susceptible patients at this point may transiently increase ventilation-perfusion mismatch sufficiently to cause a further critical acute desaturation.

The British Thoracic Society's asthma guidelines advise oxygen as first line treatment in hospital for all patients in cases of acute severe asthma. Ideally oxygen should be given before and concurrently with a nebulised bronchodilator to maximise alveolar oxygenation in areas of poor ventilation and should then be continued after nebulisation.2

The guidelines for treatment of acute severe asthma in general practice imply that general practitioners should be prepared to treat acute asthma of all severities but do not advise the use of oxygen for children or insist on its use in adults. This may be because many general practices do not keep an oxygen cylinder, relying instead on air driven nebulisers or metered dose inhalers and holding chambers. The use of metered dose inhalers and holding chambers rather than nebulisers for drug delivery was reviewed recently by the Cochrane Collaboration, which concluded that outcomes were similar whatever the method of drug delivery and that spacers may even have some advantages in children. However, these recommendations were not intended for drug delivery in life threatening asthma.24

It is impossible to prove that the continuing trickle of deaths from asthma in Britain is a result of hypoxaemia caused by air driven nebulisers. Many factors may contribute to hypoxaemia in these patients, including bacterial pneumonia, previous poor control, extreme bronchospasm, and mucus plugging. The important point is that asthmatic patients are still dying during acute attacks—and the use of oxygen before, during, and after nebulised β2 agonist therapy in primary care and in the community is rational and could save lives. Refillable portable oxygen cylinders are readily available and can be used to drive nebulisers if they are fitted with high flow valves. Oxygen is also useful in many other medical emergencies. As long as resources, training, and safety procedures are adequate, oxygen should be available in every general practice. Patients with severe disease could be provided with oxygen cylinders with high flow valves for emergency use at home. This is already the practice in some units that deal with patients with difficult asthma.


There are two important caveats to our suggestion. Firstly, whether oxygen should be available for home visits requires careful consideration, including a risk-benefit analysis. This would be best done by the British Thoracic Society in conjunction with primary care colleagues when the guidelines are next updated. Secondly, administration of oxygen is clearly beneficial in children and young adults with asthma; however, patients over 45 with a history of chronic obstructive pulmonary disease should receive salbutamol nebulised with air to avoid carbon dioxide narcosis.


Treatment of mild and moderate asthma attacks should continue as at present, with either air driven nebulisers or metered dose inhalers and holding chambers. This should not cause hypoxaemia. However, if the signs of a severe or life threatening attack are present (see box), oxygen before and after treatment with a β2 agonist nebulised with oxygen should be the standard treatment wherever the patient happens to be. We urge the British Thoracic Society to review this issue when it updates its guidelines.


Contributors: DI and AJP conceived the idea for the paper. All authors discussed the core ideas and contributed to the writing and editing of the paper. DI is guarantor.


  • Funding DI is funded by the Medical Research Council.

  • Competing interests None declared.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
View Abstract