Intended for healthcare professionals

Clinical Review

Difficult to treat asthma in adults

BMJ 2009; 338 doi: (Published 24 February 2009) Cite this as: BMJ 2009;338:b494
  1. Graeme P Currie, consultant chest physician1,
  2. J Graham Douglas, consultant chest physician1,
  3. Liam G Heaney, consultant chest physician2
  1. 1Aberdeen Royal Infirmary, Aberdeen AB25 2ZN
  2. 2Belfast City Hospital, Belfast BT9 7AB
  1. Correspondence to: G P Currie Graeme.currie{at}

    Summary points

    • Difficult asthma is present when patients using treatment at steps 4 or 5 of the British Thoracic Society and Scottish Intercollegiate Guideline Network guidelines remain symptomatic and have exacerbations

    • Determine whether the patient truly has asthma, whether prescribed treatment is being taken, and whether an alternative or coexisting diagnosis or aggravating factors exist

    • After taking a detailed history and examination, further directed investigations may be needed

    • Assessment and further management should ideally take place within a multidisciplinary clinic setting

    There is no universally accepted definition of difficult asthma. However, it is reasonable to consider it present when people have persistent symptoms and frequent exacerbations, despite being treated at steps 4 or 5 of the British Thoracic Society and Scottish Intercollegiate Guideline Network (BTS/SIGN) guidelines (fig 1).1 Such patients typically receive high dose inhaled steroids (≥800 µg beclometasone equivalent), a long acting β2 agonist, plus add-on treatment. The prevalence of difficult asthma is uncertain, but it may account for 5-10% of adults with asthma.2 Morbidity and health costs are disproportionately high in these patients,w1 and they are at greater risk of fatal and near fatal exacerbations.w2 In addition, frequent, intermittent, or continuous courses of oral prednisolone (plus regular high dose inhaled steroids) increase the risk of steroid related adverse effects.

    Sources and selection criteria

    All authors performed a comprehensive search of articles published up to January 2009 using PubMed and Medline. Keywords and phrases used were “asthma”, “exacerbations”, “symptoms”, “difficult asthma”, “lung function”, “diagnosis”, “bronchial hyper-responsiveness”, “nitric oxide”, “sputum eosinophils”, “treatment”, “monitoring”, “inhaled steroids”, “long acting β2 agonists”, “leukotriene receptor antagonists”, “bronchial thermoplasty”, “immunoglobulin E therapy”, and “tumour necrosis factor”.


    Fig 1 Simplified diagram of the pharmacological management of chronic asthma in adults1 w1

    Why are some people with asthma difficult to treat?

    Several key questions must be considered before prescribing add-on treatments and higher doses of inhaled or oral steroids to patients with difficult to treat asthma.

    Do they really have asthma?

    To answer this question, the patient’s history and objective support for a diagnosis of asthma need to be reviewed (boxes 1 and 2). In patients with airflow obstruction, bronchodilator reversibility to 400 μg of inhaled salbutamol should be performed. In patients without airflow obstruction, other objective ways of confirming the diagnosis—such as exercise testing, measuring exhaled nitric oxide, and bronchial challenge testing—may be necessary.1 According to BTS-SIGN guidelines, the diagnosis should be reconsidered in patients with persistent symptoms who do not show airway hyper-responsiveness to bronchoconstrictor stimuli such as methacholine (or in the future mannitol).1 w3 Box 2 outlines other tests that may help establish whether asthma is present. Not all tests will be needed in each patient—this will depend on the clinical situation.

    Box 1 Important clinical aspects to cover in patients with difficult asthma

    • Details of original diagnosis: “who, when, and how,” previous demonstration of airflow obstruction, documented response to treatment

    • Morbidity: length of history, severity of symptoms, exacerbating factors, frequency of exacerbations, previous admissions to intensive care with abnormal airway inflation pressures

    • Presence of other respiratory symptoms (sputum production, chest pain, haemoptysis)

    • Smoking history

    • Family history of respiratory disorders

    • Past and present occupation

    • Potential aggravating medical conditions (features of rhinosinusitis, nasal polyps, depression or anxiety, gastro-oesophageal reflux disease, relation to menstrual cycle)

    • Drugs (β blockers, aspirin, non-steroidal anti-inflammatory drugs, angiotensin converting enzyme inhibitors)

    • Adherence to treatment

    • Psychosocial circumstances

    • Respiratory system: look for signs that may or may not be consistent with asthma

    • Cardiovascular system: look for signs that may indicate an alternative diagnosis

    • Body mass index: obesity is associated with difficult to control asthma, and uncertainty may arise in differentiating between breathlessness caused by obesity and that caused by asthma

    Box 2 Investigations that may be needed in patients with difficult asthma

    Blood tests
    • Blood eosinophil count

    • Total serum IgE and allergen skin prick testing (or radioallergosorbent test) especially to Aspergillus

    • Blood theophylline or prednisolone concentrations (where appropriate) to support adherence

    • Other blood tests depending on clinical features (α1 antitrypsin, angiotensin converting enzyme, immunoglobulins and functional antibody testing, antineutrophil cytoplasmic antibody, Aspergillus serology)

    Pulmonary function
    • Spirometry

    • Airflow obstruction present: test for bronchodilator reversibility

    • Airflow obstruction absent: consider exercise test, bronchial challenge test (using inhaled methacholine or mannitol), exhaled nitric oxide (raised in eosinophilic airway inflammation), sputum eosinophil count

    • Flow volume loop, lung volumes, and transfer factor to identify features not consistent with asthma (where appropriate)

    • Chest radiography

    • High resolution computed tomography of thorax (to look for bronchiectasis or interstitial lung disease)

    • Computed tomography of sinuses (to look for sinus disease after nasal endoscopy or a trial of nasal steroids)

    Other tests (as clinically indicated)
    • Nasal endoscopy (to look for polyps and features of rhinosinusitis)

    • Bronchoscopy (to look for tumours or foreign bodies)

    • Laryngoscopy during symptomatic episodes (to look for vocal cord dysfunction)

    • Echocardiogram (where wheeze may be caused by “cardiac asthma”)

    • Psychiatric assessment

    • Cardiopulmonary exercise testing

    Do coexisting conditions exacerbate the asthma?

    Asthma often occurs alongside other conditions. In two case series, coexisting disorders with asthma-like symptoms were found in 19%3 and 34%4 of patients with difficult asthma (fig 2). If coexisting conditions are correctly identified and managed, it may be possible to improve symptom control without escalating treatment. Moreover, bronchiectasis, gastro-oesophageal reflux disease, rhinosinusitis, and psychological disorders are more common in people with frequent versus infrequent severe exacerbations of asthma.w4 Vocal cord dysfunction (paradoxical adduction during inspiration) is an important disorder that can mimic or coexist with asthma.w5 w6 It may masquerade as wheeze and breathlessness, with episodes beginning and remitting abruptly (unrelated to treatment). Its prevalence is uncertain, and diagnosis requires a high index of suspicion and direct visualisation of the vocal cords when symptomatic.w7


    Fig 2 The frequency and type of coexistent diagnoses in two studies of patients with difficult asthma. Adapted, with permission, from Heaney and Robinson5

    What aggravating factors might be considered?

    Psychological factors

    Difficult to control asthma often causes considerable psychological stress, especially if a life threatening episode has occurred; psychological factors (especially anxiety) may also worsen asthma control. Adverse psychological factors are associated with hospital admission with acute asthma,w8 w9 and psychosocial morbidity has also been linked to fatal and near fatal episodes of asthma.1 In a study where sequential patients referred to a difficult asthma service were assessed by a psychiatrist, 32 out of 65 had an ICD-10 (international classification of diseases, 10th revision) psychiatric diagnosis (usually depression).4 In the same study, the hospital anxiety and depression scale (HADS) had a good negative predictive value for depression. In a cohort of 56 patients, 33 had a psychiatric component to their asthma, and in 10 this was thought to be “major.”3 It is unclear whether psychological morbidity negatively affects difficult asthma, and whether treating concomitant psychiatric morbidity improves asthma control. Although patients often identify acute stress and depression as triggers, psychological factors may cause poor adherence to treatment, rather than have a direct effect on asthma severity.

    Upper airway disease

    The upper and lower airways have a direct anatomical connection, share a similar epithelial lining, and release similar inflammatory mediators. Asthma and allergic rhinitis represent a continuation of the same inflammatory process so it is not surprising that they commonly coexist.6 Indeed, treating allergic airway inflammation in the nose can improve markers of asthma control.w10-w13 The allergic rhinitis and its impact upon asthma (ARIA) guidelines emphasise the importance of identifying symptoms of asthma in people with rhinitis and vice versa. 7 Management strategies for allergic rhinitis consist of allergen avoidance, immunotherapy, intranasal steroids, and systemic or topical antihistamines.7

    Gastro-oesophageal reflux disease

    The incidence of gastro-oesophageal reflux disease is higher in patients with asthma than in the general population, although its relation to difficult asthma is not clear.w14 w15 A systematic review of 12 studies found no convincing evidence to support improved asthma control with treatment for gastro-oesophageal reflux disease,8 and in difficult asthma, identification and treatment of the disease failed to improve control.w16

    Adverse drug effects

    Non-steroidal anti-inflammatory drugs, β blockers (including eye drops), and aspirin can exacerbate asthma. Indeed, aspirin sensitive asthma may exist in up to 20% of people with asthma.9 Chronic cough caused by angiotensin converting enzyme inhibitors may also mimic less well controlled asthma.


    Increased sensitisation to aeroallergens has been associated with greater airway hyper-responsiveness,10 w17 and sensitisation to fungal allergens has been linked to life threatening asthma.w18 However, little evidence exists to suggest that reducing exposure to house dust mite and other ubiquitous allergens reduces symptom scores and exacerbations or improves peak expiratory flow.11

    Occupational factors

    Many patients with difficult asthma may have stopped work because of symptoms. Because occupational asthma may account for as much as 10-15% of adult onset disease,1 questioning patients about occupational exposure to airway irritants may uncover a trigger. Those who are better on days away from work or when on holiday should be investigated for occupational asthma.

    Cigarette smoking

    Cigarette smoking is associated with persistent asthma, an accelerated decline in lung function, and higher mortality after admission with an episode of near fatal asthma.w19-w21 People who currently smoke or previously smoked have reduced airway sensitivity to the effects of inhaled steroids compared with non-smokers.12 13


    Accumulating data associate obesity with persistent asthma and asthma related visits to the emergency department, and obese women seem to have an increased risk of having asthma.14 Although the association between asthma and obesity is not fully understood, weight loss should be encouraged.

    Are patients taking their treatment?

    In a cross sectional observational study of non-adherence in difficult asthma (n=182), 34% of people were collecting less than 50% of their prescriptions for inhaled combination therapy.w22 In a case series, 50% of patients prescribed oral steroids were found to be non-adherent when assessed by plasma prednisolone and cortisol concentrations.3 Thus, despite persistent symptoms, many patients choose not to take their prescribed treatment and reasons for this need to be explored (box 3).

    Box 3 Potential reasons for patients not adhering to treatment

    • Treatment considered unnecessary, ineffective, or dangerous

    • Lack of immediate effect after taking inhaled steroids

    • Poor understanding of the treatment regimen

    • Poor inhaler technique

    • Resentment about the need for treatment

    • Economic restriction on access to health care and drugs

    • Demographic factors such as sex and ethnicity

    • Forgetfulness and stress

    • Secondary gain from persistent symptoms

    Should you refer a patient with difficult asthma?

    Patients who are refractory to guideline based treatment should be referred to a specialised clinic. This is important if the diagnosis is in doubt or when excessive amounts of drugs, especially oral steroids, are used. Ideally, referring doctors should have access to a multidisciplinary clinic where experienced personnel (respiratory physicians, psychologists, and specialist nurses) have the necessary tools (such as bronchial challenge test kits and fibreoptic nasoendoscopy) and infrastructure available for assessment. However, a UK postal questionnaire sent to more than 600 consultant members of the BTS (50% response rate) found that only 23% of responding respiratory physicians had a dedicated difficult asthma clinic in their hospital.15

    What new approaches are available?

    Anti-immunoglobulin E

    Many people with asthma are atopic, with the consequence that aeroallergens interact with IgE and cause the release of inflammatory mediators.w23 Omalizumab is a humanised monoclonal antibody that can be given subcutaneously; its dose is determined by baseline IgE and body weight. Omalizumab cannot be given if the total IgE is more than 700 IU/l, which effectively excludes highly atopic patients. A randomised controlled trial evaluated the addition of omalizumab in people with severe asthma who had persistent symptoms despite using inhaled steroids and long acting β2 agonists.16 The clinically significant asthma exacerbation rate was 0.68 with omalizumab and 0.91 with placebo over 28 weeks (P=0.042), and compared with placebo, active treatment significantly reduced severe asthma exacerbation rates (0.24 v 0.48; P=0.002) and emergency visit rates (0.24 v 0.43; P=0.038). The National Institute for Health and Clinical Excellence advises that omalizumab should be considered only for patients who have had at least two severe exacerbations requiring hospital admission within the previous year, whereas the Scottish Medicines Consortium advises its restriction to patients requiring maintenance oral steroids when all other treatments have failed.

    Other biological treatments

    Tumour necrosis factor α is thought to have a role in some chronic inflammatory conditions. Despite initial promise, however, a randomised double blind study evaluating etanercept (a tumour necrosis factor antagonist) over 12 weeks found no significant improvement in major outcomes in patients with steroid dependent asthma.17 Moreover, a trial of 24 weeks’ treatment with gomilumab (a humanised monoclonal antibody against tumour necrosis factor) also found no important differences in asthma outcome. The trial was stopped early because of the significantly increased numbers of infections and incidence of malignancies in the active treatment arm.18 Whether other biological agents such as anti-interleukin 13 antibody or anti-neutrophilic strategies such as anti-CXCR1/R2 will be beneficial in refractory asthma remains to be seen.

    Other drugs

    Various drugs such as ciclosporin, methotrexate, gold, and subcutaneous terbutaline have been tried with various degrees of success in difficult asthma. These agents are not in widespread use but may be considered under specialist supervision.

    Bronchial thermoplasty

    An increase in airway smooth muscle mass is thought to be an important factor in severe or fatal asthma.w24 w25 Bronchial thermoplasty—where controlled thermal energy is delivered to the airway wall during several bronchoscopy procedures—results in prolonged reduction of smooth muscle mass. In people with moderate to severe asthma, this procedure reduces symptoms, use of relievers, and exacerbations, and it improves quality of life and lung function.19 Larger long term studies are needed to evaluate this new technique fully.

    How should patients with difficult asthma be monitored?

    It is unclear how best to monitor patients in the community, although regular follow-up by doctors and nurses trained in asthma management may be beneficial. This also provides an opportunity to assess inhaler technique and switch to a more efficient device if necessary or one that the patient can use more easily. Non-specific questions tend to underestimate symptoms,20 but asthma control can be assessed by well validated questions, such as those in the asthma control test (table).

    Asthma control test20

    View this table:

    It is difficult to assess adherence to inhaled drugs, although surrogate markers such as prescription filling of maintenance inhalers may be helpful. Direct measurements such as plasma theophylline concentrations or plasma prednisolone and cortisol measurements can provide supportive evidence.

    In the future it may be increasingly important to monitor disease activity—and thereby titrate treatment—using surrogate inflammatory biomarkers; examples of these include airway hyper-responsiveness (using indirect challenge tests such as mannitol), induced sputum eosinophil counts, and exhaled nitric oxide.21 Indeed, asthma control may be improved when a surrogate inflammatory biomarker is incorporated into algorithms (along with conventional measures of asthma control) by which treatment is altered.22 23 24 However, in one study, the use of exhaled nitric oxide as an indicator of asthma control resulted in higher doses of inhaled steroids being used, without a clinically important reduction in symptoms.25


    Managing asthma that is refractory to usual treatment requires a systematic approach to ensure a correct diagnosis, identify coexisting disorders, tailor treatment, and evaluate adherence. The BTS has established UK registries of adults and children with difficult to control asthma to standardise and optimise assessment protocols across UK centres. In the future, this may facilitate research into severe asthma phenotypes and disease mechanisms in an attempt to define best practice.

    Additional educational resources

    Additional educational resources for healthcare professionals
    Additional educational resources for patients
    • Asthma UK (—A leading UK asthma charity organisation

    • British Lung Foundation (—A leading UK charity helping those with respiratory disorders

    • ASH (—A public health charity dedicated to limiting harm caused by smoking

    Tips for non-specialists

    • Always check that the diagnosis of asthma is correct

    • Consider further tests to help confirm the diagnosis; refer to a specialist if the diagnosis remains questionable

    • Always ask about adherence to treatment

    • Assess inhaler technique at every opportunity

    • Always encourage smoking cessation and weight reduction


    Cite this as: BMJ 2009;338:b494


    • References numbered w1 to w25 are on, labelled as extra

    • Contributors: GPC had the original idea for the manuscript and wrote it along with JGD and LGH; GPC is guarantor.

    • Competing interests: GPC has received funding and honorariums from AstraZeneca, Merck Sharp and Dohme, and GlaxoSmithKline; JGD from GlaxoSmithKline, AstraZeneca, and Novartis; and LGH from AstraZeneca, Merck Sharp and Dohme, Novartis, and GlaxoSmithKline for attending postgraduate educational meetings and giving talks.

    • Provenance and peer review: Commissioned; externally peer reviewed.


    View Abstract