Predictors of speech and swallowing function following primary surgery for oral and oropharyngeal cancer

Sir, Head and neck cancer strikes at some of the most basic human functions – verbal communication, eating and breathing. In their article, Zuydam et al. describe their experience of prospectively evaluating quality of life (QOL) in a large group of patients with HNC using the University of Washington QOL questionnaire – a validated, disease-specific QOL assessment instrument. They focus on two domains – speech and swallowing – and correlate these with various clinical parameters, which are proposed as predictors of outcome. Whilst the authors are to be congratulated on this work, the results ought to be viewed with caution. The group of patients is largely an oral cancer group with only 12% having oropharyngeal tumours. The conclusions arrived at are therefore less applicable to the oropharyngeal group. The latter are more likely to have impaired swallowing and speech than the oral cancer group. Indeed, the group as a whole when asked to rank the most important functional domains gave chewing and saliva more importance than speech and swallowing. As with many studies of this kind, this work is weakened by the poor rates of response – 71% at baseline then 63%, 74% and 66% at 6, 12 and 18 months respectively – after adjusting for mortality. Patients who do not complete QOL questionnaire are more likely to be doing poorly. It would be useful to know how these patients with incomplete data for the various time points were analysed. Finally, but critically, the authors have not stated which version of the University of Washington QOL questionnaire was used. They have not mentioned version 3 but do mention version 4, which is undergoing prospective validation. The scoring system quoted (scores of 100, 70, 30 and 0) do not concur with those cited for version 3 and 4 by Weymuller’s group at the University of Washington (scores of 100, 67, 33 and 0). If, however, the authors have used different versions over the study period, then it would be interesting to know the valid adjustments that they have made. Although the University of Washington QOL questionnaire is relatively simple and easy to administer, we agree with the authors that it may lack sensitivity in detecting the full range of speech and swallowing problems experienced by head and neck cancer patients. Quoting the results of specific domains overcomes the cancellation effect that is implicit in a global score, but is a poor substitute for specific questionnaires. Kazi, R., Kanagalingam, J., Al-Mutairy, A., Nutting, C.M., Clarke, P., Rhys-Evans, P.H. & Harrington, K.J. Royal Marsden Hospital, Fulham Road, London SW3 6JJ, UK, E-mail: rehan.kazi@rmh.nhs.uk