Article Text

Download PDFPDF

Outcomes in children treated for persistent bacterial bronchitis
  1. Catherine Byrnes1,
  2. Elizabeth Edwards1
  1. 1Department of Respiratory Medicine, Starship Children’s Hospital, Auckland, New Zealand
  1. Correspondence to:
    Dr Catherine Byrnes
    Paediatric Department, Faculty of Health & Medical Sciences, Auckland University, Private Bag 92019, Auckland, New Zealand; c.byrnes{at}
  1. D Donnelly2,
  2. A Critchlow2,
  3. M L Everard2
  1. 2Paediatric Respiratory Unit, Sheffield Children’s Hospital, Sheffield, UK

Statistics from

We write with regard to the retrospective chart review by Donnelly published recently in Thorax.1 The review covers a 5 year period and it is presented as a cross-sectional survey of their personal practice of children with persistent bacterial bronchitis. We are unsure how often the patients were seen over the time of the review and it does not examine outcomes longitudinally.

We agree that there is a lack of clarity regarding the definition of chronic bronchitis. However, like all diagnoses of exclusion, it is a difficult diagnosis to make. The definition of chronic bronchitis in adults is more specific: “the presence of chronic productive cough for 3 months in each of two successive years and a patient under whom other causes of chronic cough have been excluded”.2 Comments regarding the diagnosis from two recently published paediatric respiratory text books are illuminating. In the first3 the adult definition of chronic bronchitis is given but “whether this definition can be applied to childhood chronic bronchitis remains unclear” and “it has the potential to divert the paediatrician from detecting a more specific respiratory condition”. The authors of the chapter suggest that: “The diagnosis of chronic bronchitis should occur in two phases. The first is consideration and identification of several well defined respiratory disorders according to a staged management protocol. The second but simultaneous phase is elimination or modification of exogenous factors that produce or maintain the child’s illness.” The second paediatric text4 also notes that “the definition of chronic bronchitis in children is less clear …”.

Our concern with this paper, based on our own experience, is that the label of “chronic bronchitis” is given to children without adequate exclusion of other diagnoses. We acknowledge that our centres may differ because of the referral pattern, with most of the patients in this review from primary rather than secondary care. The definition in this article is very reliant on a persistent “wet cough”, with no other anomalies. However, in this group of 81 children, 68% had abnormal chest radiographs (of the 98% in whom they were done), but only 17% had chest CT scans and 23% underwent bronchoscopic examination. The other paper5 with which the authors align themselves had 15 patients diagnosed with chronic bronchitis, but all of these had a negative chest CT scan and all had positive bronchoscopy so a more accurate diagnosis is likely. One of the diagnostic criteria used in this article was response to antibiotics—analogous to response to asthma treatment. However, while only asthma responds to asthma treatment, many conditions may respond to antibiotic treatment and time alone may lead to some symptom resolution, so it is not a discriminating diagnostic characteristic. We agree that persistent bacterial bronchitis is often misdiagnosed as asthma although the two conditions may coexist and, although 30% of families smoked, there was no discussion on the impact this may have had.

We agree that chronic bronchitis is a real entity and that it may be underdiagnosed at the current time. Untreated, this condition may progress to bronchiectasis in an (unknown) percentage of children and prospective work in this area is needed to confirm this. However, we need to proceed cautiously so as not to miss already established bronchiectasis or other diseases without adequate investigations. It is hard to label all of these children as chronic bronchitis when 59% had symptoms for more than 1 year, 13% required six courses of 4–6 weeks of antibiotics before improvement, and so few had chest CT scans or bronchoscopic examinations. Those that improved after two courses of antibiotics with no chest radiographic changes would be understandable to label as more likely to have chronic bronchitis. In this paper, however, that would amount to 51% of the patients.

In addition, we disagree that the children with an underlying immunodeficiency frequently have bronchiectasis that resolves. The only groups in which we have seen this happen regularly are those with bronchiectasis secondary to foreign body inhalation or those diagnosed during immunosuppressive treatment for an oncological disease; other examples of reported circumstances are referenced.6,7

The authors state that most publications (not referenced) relating to idiopathic bronchiectasis focus on diagnosis and management and ignore the antecedent stages of the disease; this is not true.10,8,9 Interestingly, while both the paediatric texts cite a number of differential diagnoses that require exclusion before using the term “chronic bronchitis”, neither specifically mentions bronchiectasis although both mention cystic fibrosis, yet the former is far more common in our experience and in other populations.8,9,10,11

The diagnosis of chronic bronchitis still needs to be made with care. While we believe it is a true and often under-recognised entity, there is still a risk of incorrectly ruling out other underlying problems which may well result in a child re-presenting with greater lung damage at a later stage. We recognise that it remains a difficult decision between assuming chronic bronchitis or subjecting a child to the radiation of a high resolution CT scan and the possible morbidity associated with bronchoscopy. In a child with persistent wet cough for more than 4 weeks, the features we believe should indicate more aggressive ascertainment of a definitive diagnosis would be:

  • repeated antibiotic courses (>3 in 1 year) with only partial or temporary resolution of symptoms;

  • persisting chest radiographic changes;

  • a definitive episode of two or more pneumonias requiring hospital admission;

  • referral from hospital or community paediatricians rather than primary care.

We agree that research and discussion need to continue in this area with regard to development of disease, diagnosis and nomenclature.


Authors’ reply

We thank Drs Byrnes and Edwards for their comments regarding our paper1 and would certainly agree that the diagnosis of persistent bacterial bronchitis/persistent bacterial endobronchial infection can be difficult to make. However, without recognition that the condition exists, the diagnosis cannot be made. In our earlier letter (see page 921) we highlight the difficulties in reaching a correct diagnosis in a child with a chronic cough due to the limited repertoire of responses shown by the lungs when inflamed.2 Failure to identify children with persistent endobronchial infection results in a huge burden of unnecessary morbidity due to the disease and due to inappropriate treatment. In a significant proportion it probably leads to bronchiectasis, although this may take decades. Our experience is that persistent endobronchial infection in children is curable, irrespective of the CT appearance, providing there is not a significant ongoing underlying problem such as cystic fibrosis or severe immunodeficiency. Bronchiectasis is not a diagnosis; rather, it is represents a radiological finding at one end of the spectrum from normality through minor peribronchial wall thickening and patchy non-specific changes. We have major concerns regarding the use of CT scans by those who do not understand the natural history of the disease. We have seen a number of patients who were noted to have significant quantities of secretions on bronchoscopy and a heavy growth of one or two organisms in the lavage fluid who were then largely left untreated because the CT scan did not show bronchiectasis.

The letter from Drs Byrnes and Edwards highlights the problem of using a non-specific term such as “chronic bronchitis”. This is why we and Anne Chang’s group3 have deliberately adopted the terms “persistent bacterial bronchitis/persistent bacterial endobronchial infection” which highlight the fact that this is persistent endobronchial bacterial infection and is quite distinct from adult “chronic bronchitis” associated with cigarette smoke. Many adult patients with chronic obstructive pulmonary disease (COPD) are plagued by recurrent/persistent bacterial endobronchial infection with the same organisms we see in children but this, as in the children, is a secondary phenomenon resulting from impaired mucocilary clearance. These patients have two ongoing pathologies—one (COPD) predisposing to acquisition of the second (persistent endobronchial bacterial infection). We would go further and speculate that the continuing symptoms and decline in lung function in a significant proportion of ex-smokers is due to ongoing inflammation secondary to persistent endobronchial bacterial infection.

As noted above, we believe that persistent endobronchial bacterial infection is not a primary diagnosis but represents colonisation secondary to impaired clearance of the airways. This may be due to cystic fibrosis or an immunodeficiency but, most commonly, is secondary to a “hit and run” insult such as a significant viral lower respiratory tract infection or—much less commonly these days—pertussis. Other causes of impaired clearance such as mucus plugging in asthma, tracheomalacia or even pulmonary vascular congestion with congenital heart disease may allow Haemophilus influenzae in particular to colonise the lower airways.

Finally, we would wish to clarify some of the misconceptions in the letter by Drs Byrnes and Edwards. We did not say that bronchiectasis frequently resolves in those with immunodeficiency but mention that it has been reported. Patients did not take up to six courses of prolonged antibiotics to improve. This was the time taken to affect a cure and a few will take longer. As previously noted, in the vast majority of cases the cough resolves within 10–14 days on high-dose antibiotics and failure to show a dramatic response calls the diagnosis into question. However, a small minority do take longer and occasionally do not clear even with 2 weeks of intravenous antibiotics, but have subsequently cleared with nebulised colistin which is active against Haemophilus. Our approach is based on the belief that the lack of a cough suggests that there is no active inflammation and that, under these conditions, the airways are healing themselves. When the typical cough returns, we aim to treat it aggressively and early until the condition resolves. We believe the main focus of research should be in how to identify the condition early in order to prevent the need for long and, in some cases, recurrent courses of antibiotics once the infection has been present for months or years.


View Abstract


  • Competing interests: None.

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