Observer variation in computed tomographic diagnosis of bronchiectasis

• Research type

  Research Study

• Full title

  Interobserver variability among radiologists for the high-resolution CT (HRCT) diagnosis of bronchiectasis

• IRAS ID

  277819

• Contact name

  Sujal R Desai

• Contact email

  s.desai@rbht.nhs.uk

• Sponsor organisation

  Imperial College London

• Duration of Study in the UK

  0 years, 10 months, 1 days

• Research summary

  In this study the principal question being asked is whether there is significant disagreement among radiologists, of different levels of seniority/experience, in making a diagnosis of bronchiectasis. This is an important area to consider for a number of reasons: firstly, bronchiectasis (in which the airways in the lungs are permanently and abnormally dilated) is an important clinical diagnosis to make or, indeed, exclude in patients with particular respiratory symptoms. Secondly, the 'modern' diagnosis of bronchiectasis is entirely reliant on high-resolution computed tomography (HRCT). Against this, it is clear that the diagnosis of bronchiectasis on HRCT can be challenging and prone to error; in part, this will relate to variation among radiologists in confirming or refuting the diagnosis.
  Because the clinical features and symptoms of bronchiectasis (cough productive of sputum, recurrent infections, weight loss, haemoptysis) are non-specific, false positive and negative diagnoses of bronchiectasis are problematic: thus, for instance, an erroneous diagnosis of bronchiectasis on HRCT in a patient without the condition may conceivably lead to delay in the diagnosis of more sinister conditions (e.g. lung cancer). Equally, a false negative diagnosis will lead to inevitable delays in instituting the appropriate management for bronchiectasis and, importantly, preventing progression.
  The aim of this study is to determine the magnitude of observer variation. If variation is unacceptably high, this may lead to steps for further education and training for radiologists when faced with patients with a clinical suspicion of the diagnosis.

• REC name

  N/A

• REC reference

  N/A