A clinical profile of otitis media

An otitis media with effusion algorithm developed by Paradise et al and tested by Cantekin et al has become the basis for many studies of otitis media. However, a two-point scale algorithm (otitis media with effusion-no otitis media with effusion) may be too ambitious (ie, low specificity) and too optimistic (ie, absence of fluid does not necessarily mean normal ear). We propose a four-point profile that characterizes the condition of the middle ear, and we report the validation of the profile against findings at myringotomy. Statistically, a four-point scale profile would substantially increase powers of statistical tests, compared with a two-point scale algorithm (in studies of the same size), when used to study changes of the middle-ear condition over time or to compare treatment effects.