Intended for healthcare professionals

Clinical Review Extracts from “Clinical Evidence”

Acute asthma

BMJ 2001; 323 doi: (Published 13 October 2001) Cite this as: BMJ 2001;323:841
  1. Mark FitzGerald (markf{at}, respiratory physician
  1. Centre for Clinical Epidemiology and Evaluation, Vancouver General Hospital, Vancouver, BC, Canada V5Z 1L8



    Spacer devices for delivering inhaled medications from pressurised metered dose inhalers in acute asthma (as good as nebulisers)

    Short courses of oral corticosteroids for acute exacerbations

    Ipratropium bromide added to ß2 agonists for acute exacerbations

    Likely to be beneficial:

    Continuous nebulised delivery of bronchodilators for acute asthma (better than intermittent treatment)

    Oxygen supplementation for acute asthma (no direct randomised evidence available)

    Intravenous magnesium sulphate for people with more severe acute asthma

    Mechanical ventilation for people with near fatal asthma (no direct randomised evidence available)

    Specialist versus generalist care for acute exacerbations

    Asthma education for people with acute asthma

    Unknown effectiveness:

    Intravenous versus nebulised delivery of short acting ß2 agonists for acute asthma


    Definition Asthma is characterised by dyspnoea, cough, chest tightness, wheezing, variable airflow obstruction, and airway hyperresponsiveness. The diurnal variation of peak expiratory flow rate (PEFR) is increased in people with asthma. Chronic asthma is defined as asthma requiring maintenance treat- ment, and will be dealt with in a separate “Extract from Clinical Evidence.1 2 Acute asthma is defined here as an exacerbation of underlying asthma requiring urgent or emergency treatment.

    Incidence/prevalence Reported prevalence of asthma is increasing worldwide. About 10% of people have had an attack of asthma. 3 4

    Aetiology/risk factors Most people with asthma are atopic; exposure to certain stimuli initiates inflammation and structural changes in airways, causing airway hyperresponsiveness and variable airflow obstruction, which in turn cause most asthma symptoms. Stimuli include environmental allergens, occupational sensitising agents, and respiratory viral infections. 5 6

    Prognosis About 10-20% of people presenting to the emergency department with asthma are admitted to hospital. Of these, fewer than 10% receive mechanical ventilation, 7 8 although previous ventilation is associated with a 19-fold increased risk of ventilation for a subsequent episode.9 It is unusual for people to die unless they have had respiratory arrest before reaching hospital.10 One prospective study of 939 people discharged from emergency care found that 17% (95% confidence interval 14% to 20%) had relapsed by two weeks.11

    Aims To minimise or eliminate symptoms; to maximise lung function; to prevent exacerbations; to minimise the need for medication; to minimise adverse effects of treatment; and to provide enough information and support to facilitate self management of asthma.

    Outcomes Symptoms (daytime and nocturnal); lung function (PEFR and forced expiratory volume in one second (FEV1)); need for rescue medication such as inhaled ß2 agonists; variability of flow rates; activities of daily living; adverse effects of treatment.


    Clinical Evidence update search and appraisal September 2000. Additional sources identified by experts.

    Question What are the effects of treatments for acute asthma?

    Option Spacer devices or holding chambers versus nebulisers for delivering ß2 agonists

    Summary One systematic review of randomised controlled trials (RCTs) found no difference in efficacy between nebulisers and holding chambers with metered dose inhalers for delivering ß2 agonists in people with acute but not life threatening asthma.


    We found one systematic review (search date 1999, 13 RCTs, non-hospitalised adults and children with acute asthma) comparing holding chambers plus metered dose inhalers against nebulisation for delivering ß2 agonists.12 Results in adults and children were analysed separately. In adults, there was no significant difference in rates of hospital admission (odds ratio (OR) 1.12, 95% confidence interval 0.45 to 2.76), length of time spent in the emergency department (weighted mean difference (WMD) 0.02 hours, −0.40 to 0.44 hours), or in PEFR or FEV1. There was still no significant difference when the three studies involving the most severely affected people (FEV1 <30% predicted) were included (WMD for FEV1 holding chamber v nebuliser −1.5% predicted, −8.3% to 5.3%). Symptoms were measured on different scales and findings could not be combined.


    The review found no significant difference in heart rates between the two methods (WMD with holding chamber v nebuliser 1.6% of baseline, –2.4% to 5.5% of baseline).


    The review found no evidence of publication bias. To overcome possible dose confounding, the review was confined to studies that used multiple treatment doses titrated against the individuals' responses. As studies excluded people with life threatening asthma, results may not generalise to such people.

    Option Oral corticosteroids in acute asthma

    Summary Two systematic reviews have found that systemic corticosteroids taken at the start of an acute exacerbation reduce rates of admission and relapse, and the need for additional ß2 agonists, without increasing adverse effects. One systematic review found no significant difference in rates of hospital admission. There was no significant difference between oral corticosteroids compared with inhaled or intramuscular corticosteroids. We found no evidence on the optimal dose or duration of treatment.


    Rates of admission: We found three systematic reviews.1315 The most recent systematic review (search date 2000, 7 RCTs, 1204 people) compared oral and inhaled corticosteroids and found no significant difference in the admission rate (2 RCTs, OR 1.0, 0.4 to 2.5).13 The second systematic review (search date 1997, 7 RCTs in about 320 people) compared oral corticosteroids and placebo (4 RCTs), oral and intramuscular corticosteroids (2 RCTs), and intramuscular corticosteroids and placebo (1 RCT).14 It found that systemic corticosteroids given during an acute asthma exacerbation reduced the number of relapses requiring additional care (first week: relative risk (RR) v placebo 0.39, 0.21 to 0.74; number needed to treat (NNT) 10; first 21 days: RR 0.47, 0.25 to 0.89) and reduced hospital admissions (RR 0.35, 0.13 to 0.95). Corticosteroids reduced the use of ß2 agonists (WMD –3.3 activations/day, −5.5 to −1.0). The review found no clear difference between intramuscular and oral corticosteroids. The third, earlier systematic review (search date 1991, 5 RCTs, 422 people) compared systemic corticosteroids and placebo.15 It also found that early use of systemic corticosteroids reduced hospital admissions and relapses in both adults and children (OR of hospital admission in adults for corticosteroids v placebo 0.47, 0.27 to 0.79). Stopping treatment: We found no systematic review. One RCT (35 people admitted to hospital with acute asthma) compared tapering of prednisolone over a week against abrupt cessation. It found that 0.5-1 mg/kg a day for 10 days was effective, and once asthma control was re-established prednisolone could be stopped without tapering.16 Optimal dose and duration of treatment: We found no systematic review or RCTs. The optimal duration of treatment is likely to depend on the individual, the severity of the exacerbation, and concomitant use of drugs.


    Systemic corticosteroids can cause the same adverse effects in asthma as in other diseases, even when given for a short time.


    We found no reliable evidence about the role of oral corticosteroids in acute asthma after admission to hospital, nor is it likely that a placebo controlled RCT would be conducted in acute severe asthma. One RCT (413 adults presenting to general practitioners with acute asthma) found no difference in rates of treatment failure between a short course of oral steroids and a high dose of inhaled fluticasone.17

    Option: Continuous versus as needed short acting ß2 agonists for acute asthma

    Summary One RCT found that continuous versus as needed ß2 agonists in acute asthma was associated with greater improvement in FEV1. Other smaller RCTs found no significant difference. In adults with more severe airflow obstruction, RCTs consistently found that continuous nebulised treatment improved outcome more than intermittent treatment.


    We found no systematic review. We found seven RCTs. 8 1823 The first, largest RCT (99 people) found that, in a subgroup of 69 people with more severe asthma, continuous aerosol delivery compared with as needed delivery increased PEFR at 120 minutes (296 l/minute, 266 to 329 with continuous delivery v 244 l/minute, 216 to 272 with as needed delivery).18 Hospital admissions were significantly lower with continuous delivery (11/35 (28%) v 19/34 (57%), P=0.03). The post hoc nature of the analysis weakens these results. One RCT (38 people) found no significant difference in improvement in FEV1 between continuous and as needed salbutamol. Subgroup analysis found a greater improvement in FEV1 with continuous treatment in people with lower initial FEV1.20 Another RCT (165 people) compared four regimens of salbutamol in a factorial design: high (1.5 mg) versus standard (0.5 mg) doses, and continuous versus as needed delivery. It found greater improvement in FEV1 at two hours with continuous versus as needed delivery at both high and standard doses.22


    Commonly reported, mild adverse effects associated with frequent dosing include tachycardia, tremor, and headache. Metabolic upsets are less common, and include hypokalaemia. One RCT found the highest rate of adverse effects with high dose as needed treatment. The most common adverse effect was tremor (24% as needed high dose, 20% continuous high dose, 9.3% hourly standard dose, and 2.5% continuous standard dose).22


    We also found one RCT (46 adults in hospital), which addressed the slightly different but related question of regular nebulised salbutamol (5 mg every 4 hours) versus on demand salbutamol (2.5-5 mg).24 It found that on demand dosage was significantly associated with shorter hospital stay (3.7 days v 4.7 days), reduced number of nebulisations (geometric mean 7.0 v 14, P=0.003), and fewer palpitations (P=0.049).

    Option Intravenous versus nebulised delivery of short acting ß2 agonists for acute asthma

    Summary RCTs found conflicting evidence on the effect of intravenous versus nebulised salbutamol. Intravenous salbutamol was associated with more adverse effects.


    We found no systematic review, and results from three RCTs comparing intravenous versus nebulised salbutamol were conflicting.2527 The largest trial (76 people who had not responded to nebulised salbutamol after 30 minutes) compared nebulised treatment at 30 minutes and two hours against intravenous salbutamol. It found greater bronchodilation in the intravenous group (FEV1 improved by 25% with intravenous salbutamol v 14% with nebulised salbutamol; difference 11%, 2.4% to 19%).27 The second, multicentre RCT (47 people) found that nebulised salbutamol (total 10 mg) was more effective than intravenous salbutamol (0.5 mg) over one hour (19/22 responded with nebulised v 12/25 with intravenous salbutamol, P=0.006).28 The third RCT (16 people) found no significant difference in FEV1 but greater self reported improvement in symptoms among people treated with inhaled salbutamol.29


    All trials found more adverse effects with intravenous delivery.2527 In the largest RCT, two of 39 people in the intravenous group withdrew because of tachyardia.27 In the nebulised salbutamol group none withdrew because of adverse effects, but three people withdrew because of lack of effect.



    Option Addition of ipratropium bromide to ß2 agonists in acute asthma

    Summary One systematic review and one subsequent RCT have found that combining short acting ß2 agonists with ipratropium bromide significantly reduces hospital admissions and improves lung function in people with more severe acute asthma.


    We found one systematic review (search date 1999, 5 RCTs evaluating hospital admissions)28 and one subsequent RCT.29 The review compared salbutamol plus inhaled ipratropium bromide against salbutamol alone and found that the addition of ipratropium significantly reduced hospital admissions (OR 0.62, 0.44 to 0.88; NNT 18, 11 to 77). Meta-analysis of the four trials that evaluated people with severe airflow obstruction (FEV1 <35%) found that additional treatment with ipratropium bromide improved FEV1 over 90 minutes (effect size 0.38, 0.05 to 0.67). The subsequent RCT (180 people with acute asthma, mean FEV1 <50%) compared salbutamol plus placebo against salbutamol plus ipratropium.29 It found that the addition of ipratropium significantly improved PEFR (difference in improvement with ipratropium v placebo 21%, 2.6% to 38%), and FEV1 (difference in improvement with ipratropium v placebo 48%, 20% to 76%). People taking ipratropium were significantly less likely to require hospital admission at the end of the three hour trial period (20% v 39%, P=0.01).


    Addition of ipratropium bromide had no significant effect on adverse effects.29



    Option Oxygen

    Summary We found no systematic review or RCTs of oxygen in acute asthma. However, experience and pathophysiology suggest that its role is vital in acute asthma. Small RCTs found mixed evidence that combining oxygen with helium may improve PEFR.


    Oxygen alone: We found no systematic review and no RCTs. Oxygen with helium: We found three RCTs of combined helium (70% or 80%) and oxygen (30% or 20%) in adults with acute asthma.3032 One RCT included 27 people; PEFR <250 l/minute despite treatment, pulsus paradoxus >15 mm Hg. Breathing a helium-oxygen mixture (80:20) compared with breathing room air alone reduced pulsus paradoxus and improved peak flow (results presented only as a graph).30 The second RCT (23 people) found the helium-oxygen combination compared with 30% oxygen increased PEFR (58% with helium-oxygen v 10% with oxygen).31 The third RCT (205 people) found no evidence of benefit from helium plus oxygen, but it was limited by a brief intervention (15 minutes), single blinding, and inclusion of people with mild to moderate acute asthma.32


    We found no evidence of adverse effects associated with oxygen alone or with helium-oxygen in acute asthma.


    The most severe stages of acute asthma are respiratory failure, cardiopulmonary arrest, and death. 9 10 Studies of near fatal asthma suggest that hypoxia, rather than arrhythmias, accounts for asthma deaths. It seems reasonable that supplemental oxygen should continue to form a critical part of management even though we found no RCTs providing direct evidence for this. Peak flow readings vary depending on the viscosity of the gas being delivered (helium is less dense than oxygen, so non-standardised measures of peak flow will increase relative to air, even if the mixture has no effect on airway narrowing). It was not clear in all studies whether peak flow readings were standardised for air and for helium-oxygen mixtures.

    Option Intravenous magnesium sulphate

    Summary Valid subgroup analysis from one systematic review suggests that, in people with more severe acute asthma, adding intravenous magnesium sulphate to usual treatment may reduce rates of hospital admission. One small subsequent RCT found that nebulised magnesium sulphate improved PEFR significantly more than saline.


    We found one systematic review (search date 1999, 5 RCTs in adults, 3 RCTs in children, 665 people),33 and one subsequent RCT.34 The review compared intravenous magnesium sulphate against placebo added to usual treatment, and found no significant difference in hospital admissions (OR 0.31, 0.09 to 1.02). Prespecified subgroup analysis of adults with more severe airflow obstruction (sample size not given; FEV1 <30% at presentation, failure to respond to initial treatment, or failure to improve beyond 60% in FEV1 after 1 hour) found that people receiving magnesium sulphate had better PEFR and reduced rates of hospital admission. The subsequent RCT (35 people) compared salbutamol plus saline against salbutamol plus magnesium sulphate through a nebuliser. It found that magnesium sulphate compared with saline significantly increased PEFR (increase in PEFR after 10 minutes: 61% v 31%; difference 30%, 3 to 56%; P=0.03).34


    We found no significant adverse effects associated with treatment.


    Further studies are needed to clarify the role of intravenous magnesium sulphate in acute asthma. Two of the studies involved treatment with aminophylline and one with ipratropium, both of which have been found to affect hospital admission rates without affecting the degree of airflow obstruction.35 The subgroup analysis involved intergroup and intragroup analyses specified before the trial was conducted, and so provides reasonably strong evidence of an effect.

    Option Mechanical ventilation

    Summary We found no RCTs comparing mechanical ventilation against no ventilation for severe acute asthma. Evidence from cohort studies and case series support its use, despite a high level of morbidity related to the intervention.


    We found no systematic reviews or RCTs.


    Mechanical ventilation is associated with hypotension, barotrauma, infection, and myopathy, especially when prolonged paralysis is required with muscle relaxants and systemic corticosteroids.36 Adverse effects reported in one retrospective study of 88 episodes of mechanical ventilation were hypotension (20%), pulmonary barotrauma (14%), and arrhythmias (10%).37


    Experience suggests that mechanical ventilation is a life saving intervention needed by a small minority of people with severe acute asthma. Cohort studies 38 39 and one case series40 found fewer deaths with controlled hypoventilation compared with ventilation in which carbon dioxide levels were normalised (for which historical cohorts and case series have reported mortality rates of 7.5-23%). 37 4143 Non-invasive ventilation has been used in people with acute exacerbations of chronic obstructive lung disease44 but requires prospective validation in people with acute asthma. Future research should also focus on delivery of bronchodilators, optimal use of muscle relaxants, and dose of corticosteroids.

    Option Specialist versus generalist care

    Summary Non-randomised trials suggest that specialist care is more effective than generalist care.


    We found no systematic review and no RCTs. One non-systematic review of controlled trials (search date 1999) found that “expert based” care was associated with improved outcomes.45 One trial quasi-randomised people (on the basis of day of attendance) referred from the emergency department to specialist care and compared this with routine general medical follow up.46 It found that people receiving specialist care were significantly less likely to wake at night (OR 0.24, 0.11 to 0.52), suffer relapse requiring emergency admission by six months (for one admission RR 0.56, 0.34 to 0.95; for two admissions RR 0.30, 0.16 to 0.60), or have multiple relapses. They were more likely to use inhaled corticosteroids (OR 3.6, 1.9 to 6.6) and cromolyn (RR 2.2, 1.9 to 2.5).


    We found no harms associated with specialist compared with generalist care.



    Option Asthma education for people with acute asthma

    Summary One systematic review of RCTs has found that education to facilitate self management of asthma in adults reduces hospital admission, unscheduled visits to the doctor, and days off work.


    We found one systematic review (search date 1999, 22 RCTs) of self management of asthma in adults.47 The review found that education about asthma to facilitate self management, whether initiated from a specialist or generalist setting, significantly reduced the risk of hospital admission (RR 0.62, 0.41 to 0.96; NNT 38, 20 to 382), unscheduled visits to the doctor (RR 0.74, 0.63 to 0.90; NNT 12, 8 to 36), and days off work (RR 0.75, 0.63 to 0.90; NNT 7, 5 to 13). Best results were achieved in people who had written care plans.


    Diurnal variation A characteristic of people with asthma is increased variation in peak flow rates and FEV1 during the day. The diurnal variation is sometimes expressed as the difference between maximum and minimum values expressed as a fraction of the maximum value.

    Forced expiratory volume in one second (FEV1) The volume breathed out in the first second of forceful blowing into a spirometer, measured in litres.

    Peak expiratory flow rate (PEFR) The maximum rate at which air is expired from the lungs when the patient blows into a peak flow meter or a spirometer. It is measured at an instant, but the units are expressed as litres per minute.

    Pulsus paradoxus A measure of the severity of asthma based on the difference in systolic pressure during inspiration and expiration. The blood pressure normally falls a little during inspiration (<10 mm Hg), but in acute severe asthma (and in some other conditions) the fall of systolic pressure in inspiration is greater.

    Salbutamol A short acting ß2 agonist known as albuterol in the United States.


    None reported.




    • Competing interests MF has received honorariums for lectures and research funds from GlaxoSmithKline, Merck, AstraZeneca, Novartis, Boehringer Ingelheim, Byk Canada, Schering Canada, and 3M.

    • Embedded Image

      This article is part of the “Asthma” topic in issue 5 of Clinical Evidence (

      Clinical Evidence is published by BMJ Publishing Group. The fifth issue is available now, and Clinical Evidence will be updated and expanded each month on the website. Individual subscription rate, issues 5 and 6 £75/$110; institutional rate £160/$240; student rate £55/$80. For more information including how to subscribe, please visit the Clinical Evidence website at


    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.
    25. 25.
    26. 26.
    27. 27.
    28. 28.
    29. 29.
    30. 30.
    31. 31.
    32. 32.
    33. 33.
    34. 34.
    35. 35.
    36. 36.
    37. 37.
    38. 38.
    39. 39.
    40. 40.
    41. 41.
    42. 42.
    43. 43.
    44. 44.
    45. 45.
    46. 46.
    47. 47.
    View Abstract