@ERSpublications Why do only some asthmatic children exacerbate, what is the underlying pathology, and how should we treat them? http://ow.ly/Mrik304eIVUMany common airway diseases are characterised by chronic symptoms with interspersed acute and unpredictable deteriorations, conventionally called "pulmonary exacerbations". Until relatively recently, these so-called exacerbations were a rather Cinderella subject, relatively neglected in favour of treating day-to-day problems. This may in part be because in asthma, for example, we have treatments such as inhaled corticosteroids (ICS) which are enormously effective in treating chronic symptoms, but prevention of exacerbations is much less successful. However, it is becoming increasingly clear that pulmonary exacerbations carry much more sinister overtones than once we thought, to the extent that the term "lung attacks" has been proposed to describe them [1,2]. In cystic fibrosis (CF), treatment of lung attacks even with intravenous antibiotics may fail to restore spirometry to baseline [3], and are associated with an accelerated decline in lung function [4] and a worse prognosis [5]. Failure of response is also described in primary ciliary dyskinesia [6]. Lung attacks, as we shall continue to describe them, have been used as clinical trial end-points in CF [7] and bronchiectasis [8], and may be a more sensitive marker of response than spirometry [9]. They are very far from being a minor and reversible inconvenience to the patient with airways disease.Exacerbations of asthma cause considerable morbidity and mortality; and, as recently confirmed yet again by the UK National Review of Asthma Deaths (NRAD) [10], a previous acute attack is a very strong predictor of asthma death. It has long been known that, physiologically, loss of asthma control and asthma attacks are different phenomena [11]. Furthermore, asthma attacks are predictive both of failure of normal airway growth [12], and an accelerated rate of decline of spirometry [13]. It is thus very clear that understanding why patients have asthma attacks, and preventing them, is a high priority. It is also abundantly clear that our definitions of "severe" asthma are woefully inadequate, a point to which we return; nearly 70% of asthma deaths were in people who did not have "severe" asthma [10], despite most people believing that asthma death is a pretty bad outcome! However, seemingly paradoxically, only a subset of asthmatics appear to exacerbate [14], and this is certainly an area in which risk stratification is needed. Reasons for asthma attacks may be considered as behavioural and biological, and so prevention needs a multifaceted approach. On the behavioural side, failure to attend follow-up appointments, overuse of short-acting β2-agonists and underuse of ICS are among the most important factors [10]. Biologically, the combination of viral infection, allergen sensitisation and high environmental allergen exposure are strongly predictive of attacks needing hospitalisation [15]. Adult data, in particular, have...