Following informed parental consent 93 children underwent bilateral grommet insertion. Tympanometry was performed pre-operatively, and immediately prior to myringotomy. A standardized anaesthetic was used. At myringotomy the presence or absence of fluid was recorded, as well as the time since induction of the general anaesthetic.

A pre-operative type B tympanogram predicted a middle-ear effusion at myringotomy in 92 per cent of patients. A pre-operative type C2 tympanogram predicted a middle-ear effusion at myringotomy in 39 per cent of patients. Sixty tympanograms (30 per cent) changed following a general anaesthetic. Fourteen type B tympanograms changed to type A and eight of these had effusions. The duration of the general anaesthetic did not influence the probability of a middle-ear effusion being present at myringotomy. A pre-operative type B tympanogram is a good predictor of middle-ear fluid. The duration of the general anaesthetic is not significant in predicting the presence of a middle-ear effusion.