Article Text


Difficult asthma
  1. B D W Harrison
  1. Department of Respiratory Medicine, Norfolk and Norwich University Hospital NHS Trust, Colney Lane, Norwich NR4 7UY, UK; brian.harrison{at}

Statistics from

It is hoped that the systematic approach to managing patients with therapy resistant asthma reported in this issue of Thorax will encourage others to study this difficult group of patients and to test hypotheses about improving their management.

Central to any description of difficult asthma1 is a disconnection between expectations and outcome. Difficult asthma may be defined as being present in a patient with a confirmed diagnosis of asthma whose symptoms and/or lung function abnormalities are poorly controlled with treatment which experience suggests would usually be effective. This immediately begs the questions of who confirmed the diagnosis, how the diagnosis was made, whether the symptoms and lung function abnormalities are due entirely to the diagnosis of asthma, and whose “experience” is being used. It is certainly wise when seeing a patient with difficult asthma to question the diagnosis. If it is confirmed, are there any co-existing organic respiratory conditions such as COPD or bronchiectasis or psychogenic conditions such as hyperventilation or vocal cord dysfunction with wheeze? If there are co-existing problems, are these the main cause of the uncontrolled symptoms as in pseudo-steroid resistant asthma?2 It is also wise to be alert when there is discordance between the patient’s symptoms and objective lung function assessment, with the poor perceiver on the one hand3,4 and the over reactor on the other. Be aware, too, of the mood enhancing properties of oral steroids and the placebo effect of any new medication in patients at the over reactor end of the spectrum. The combination of supramaximal doses of inhaled steroid and multiple β2 agonist preparations in patients referred with asthma should always raise alarm bells. Difficult asthma can occur in patients with objectively mild, moderate, or severe disease, but the consequences are most dramatic in patients with severe asthma.

Accepting the above definition of difficult asthma, assuming the diagnosis has been confirmed, and having taken account of co-existing physical diseases, the asthma may be difficult for the patient, for the clinician, or both because of one or more of the following:

  • Disease factors—for example, brittle asthma or genuine steroid resistant asthma, both of which are rare.

  • Doctor or nurse therapist factors—for example, inexperience or inappropriate therapies, both of which are common.

  • Patient factors—for example, poor compliance, other behavioural or personality factors, adverse psychological or social factors, which are also common.

Every chest physician will see patients with difficult asthma in their practice and at least six centres have established special clinics for these patients. In the current issue of Thorax Heaney et al report their experience with therapy resistant asthma (TRA) from such a clinic.5 Patients were recruited on the basis of having a confirmed diagnosis of asthma, persisting refractory symptoms prompting specialist referral, high dose inhaled steroid therapy coupled with inhaled long acting β2 agonists, and at least one course of systemic steroids in the preceding 12 months. The protocol included assessment by a respiratory physician, a psychiatrist, and an ENT specialist with appropriate investigations including a high resolution CT scan of the thorax, 24 hour dual probe ambulatory oesophageal pH monitoring, a DEXA scan, and induced sputum evaluation. Patients were then managed according to the BTS guidelines and followed up for a minimum period of 12 months. Seventy three patients were evaluated. At the end of 12 months the 39 whose asthma symptoms were controlled were classified as having non-therapy resistant asthma (non-TRA) and were discharged. Thirty four patients had TRA, which was defined as persisting asthma symptoms despite high dose inhaled steroids plus long acting β2 agonist therapy with the requirement for either maintenance systemic steroids or at least two rescue courses of steroids during the follow up period of 12 months despite trials of other add-on therapies. Of the 80 subjects initially recruited, two had psychogenic breathlessness or vocal cord dysfunction, another had very poor compliance, one did not have asthma, and three were lost for a variety of other reasons. Twenty five had an additional diagnosis including 14 with chronic airways disease and 10 with psychogenic respiratory problems. Fifty seven of the 60 (95%) who had an ENT examination had one or more abnormalities, and 32 of the 65 (49%) reviewed by a psychiatrist had an ICD10 psychiatric diagnosis which was unrecognised in 27. Oesophageal reflux was found in 31 of the 54 (57%) in whom it was measured. These results emphasise the importance of looking for co-existing diseases in patients with severe or difficult asthma, but no differences were found in any of these areas between the TRA and the non-TRA groups except that the non-TRA group had significantly more additional diagnoses. The high frequency of psychiatric illness corresponds to the high levels of psychosocial adversity found in studies of near fatal asthma6 and asthma deaths7 and, indeed, in the control groups of patients admitted to hospital with severe asthma in both of those studies.6,7

Heaney et al5 identify a number of significant differences between the TRA and non-TRA groups of patients. Those with TRA had a longer period of instability before referral and during follow up, a higher dose of inhaled steroid at referral, more rescue courses of prednisolone in the 12 months before referral and during follow up, more frequently required maintenance systemic steroids, were more likely to have attended a previous specialist, had lower FEV1 % predicted on referral and during follow up and a lower FEV1/FVC ratio, had fewer additional diagnoses as noted above, had a greater incidence of osteoporosis, and recorded a lower asthma quality of life score during follow up. The authors derived a probability of having TRA from the following: inhaled steroids >2000 μg beclomethasone or equivalent at referral, previous specialist attendance, and a pre-bronchodilator FEV1 % predicted of <70%. Patients with all three of these had a probability of 0.93 of having TRA. The authors are now exploring the value of using this approach to detect patients with TRA.


Some patients with difficult asthma can be helped by optimising asthma treatment by addressing co-existing morbidities and by accepting which symptoms are due to asthma and which are not. Other patients, most of whom might be identified by the presence of the three factors noted above, have limited benefit from this approach. However, even in this group there was objective evidence of some benefit in that the pre-bronchodilator FEV1 at enrolment was 65% and the best during follow up was 83%. It would be of interest to know the number of co-morbidities or adverse factors present in individuals in the TRA and the non-TRA groups and their duration. Our work suggests that patients with large numbers of co-morbidities, including psychosocial adversities which have been present for longer, respond less well to attempts to optimise medical treatment and other supportive measures (S Mildenhall, M Noble, personal communication). In these circumstances the secondary gain from attributing all symptoms to asthma can be very powerful. Patients may then not be prepared to accept the psychogenic contribution to their symptoms. When they do, the problems may be so deep seated that even long term psychological therapies, which are rarely available, might be ineffective. A qualitative research study in this group of patients with TRA would be worth exploring.

In view of the paper by Green et al8 on the potential value of monitoring induced sputum eosinophilia in the management of asthma, it is unfortunate that sputum eosinophil counts were only available in 31 of the 73 patients (42%) with the remaining subjects failing to expectorate an adequate sample or to tolerate the induction procedure. There is currently a great deal of interest in approaching asthma management through indices of inflammation and in no group is this likely to be greater than in patients with difficult asthma.


In an earlier study of patients with difficult to control asthma Irwin and colleagues9 concluded that the two most useful interventions were diagnosis and treatment of gastro-oesophageal reflux and treatment with inhaled corticosteroids. A recent Cochrane review10 of 12 double blind controlled trials found no symptomatic or physiological benefit in asthma following treatment of gastro-oesophageal reflux when both conditions were present. The fact that patients seen by Irwin et al throughout the 1980s were receiving prednisolone but not inhaled corticosteroids reflects the differences in practice between the USA and northern Europe, Australasia and Canada. What they also concluded was that non-adherence to treatment was the most likely reason for the asthma continuing to be difficult to control.

We have reported our experience of helping patients with difficult asthma and psychosocial adversity in a clinic run jointly by a psychiatrist and a chest physician.11 In a recently completed systematic review of psycho-educational interventions for difficult asthma, we have found very few well conducted trials (J Smith, personal communication). Again, more well designed research is needed.

The systematic approach described by Heaney et al5 will encourage others to look at this difficult group of patients, many of whom are patients with asthma and co-existing difficulties which will not only increase our awareness of them but also help us to test hypotheses about improving their management.

It is hoped that the systematic approach to managing patients with therapy resistant asthma reported in this issue of Thorax will encourage others to study this difficult group of patients and to test hypotheses about improving their management.


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles

  • Airwaves
    Wisia Wedzicha