Concerns that a largely anecdotal increase in post tonsillectomy haemorrhage rates was related to the introduction of disposable instruments have prompted much investigation. The result has been, rather, to highlight other variables influencing this risk, but especially to insist on the following:

1. 1 Training in traditional ‘cold’ techniques.

2. 2 Regular departmental audit of haemorrhage rates.

3. 3 Presentation of such data to patients to ensure informed consent.

This audit demonstrates the pitfalls in interpretation of crude data, unadjusted for case-mix, in predicting individual patient risk and in national ranking of unit performance.