Background: Clinicians have recognised the role of narrow band imaging (NBI) in the management of head and neck cancer in several studies. However, a recent systematic review was unable to pool the data on diagnostic efficacy in this setting owing to the heterogeneity in the published data.
Methods: Secondary analysis of data, utilising Bayes' theorem, from meta-analyses and randomised trials.
Results: In patients with a histological diagnosis of mild dysplasia who show no abnormalities on NBI, the post-test probability of malignancy is estimated to be 2.3%, compared to 10.3% with conventional white light imaging (WLI). For severe dysplasia, similar post-test probabilities after NBI and WLI are estimated to be 8.0% and 29.7%, respectively. Post-test probabilities in this setting indicate the chance of missing malignancy following a negative NBI or WLI in patients who undergo no further intervention. This study also provides a nomogram designed for use in this setting.
Conclusions: This study identifies the evidence base for use of NBI in the follow-up for laryngeal dysplasia.
Keywords: general; outcomes; research; systematic reviews.
© 2018 John Wiley & Sons Ltd.