Head and neck surgery recommendations during the COVID-19 pandemic - Author's reply.

We thank Thomas J Galloway and colleagues for their letter. We agree that available evidence, not solely expert opinion, should always be considered. However, the COVID-19 pandemic has resulted in an unprecedented situation in which there are little or no available data, and where the extrapolation of existing evidence is not appropriate in many cases. Hence, there is a need for robust expert opinion recommendations.

We suggest that the small amount of evidence available, which the authors reference, is more nuanced. For example, although reporting that early cancers were more affected by treatment delay, Colin Murphy and colleagues also reported that oral cancer outcomes were much less affected by delays in treatment than were laryngeal cancer outcomes. Murphy and colleagues also showed that the threshold for treatment delay that resulted in significant detriment was 67 days.

We also urge Galloway and colleagues to read our recommendations more carefully, including the explanatory text, as they are more nuanced than described in their letter. None of our recommendations support delay of surgery beyond 60 days. Indeed, for all early cancers, the recommendations only supported delay beyond 30 days, and only half of the experts supported delay up to 60 days. There was strong agreement not to delay up to 90 days.

Furthermore, our recommendations pertain to delays in surgery, which should not be conflated with delays in treatment, as many tumours can be treated by alternative, equally effective, non-surgical methods. For example, for laryngeal cancer, which is more affected by delays than oral cancer, there was agreement that if surgery was delayed beyond 30 days, radiation should be given immediately instead.

Similarly, the section of our Review on overall prioritisation of patients for surgery is highly nuanced, factoring in the important issues that should be considered when deciding on priority. The availability of alternative effective treatment modalities, for example, chemoradiotherapy for T2 N1 oropharyngeal cancer or radiotherapy for T1 N0 laryngeal cancer, results in deprioritisation of these cases for surgery (not for treatment) during severe resource constraint compared with cases in which surgery is the only, or significantly better, option (eg, for advanced disease or T1 N0 oral cancer).