A re-examination of the Carhart effect.

It is well established that in individuals with a conductive hearing loss the bone conduction thresholds do not directly indicate cochlear function, and are artificially depressed as witnessed by their improvement following successful reconstructive surgery. This depression, which we define as the 'Carhart effect' is due to changes in the middle and external ear contributions to bone conduction. Published figures for the Carhart effect, inferred from pre- and post-operative bone conduction thresholds vary considerably but the discrepancies are markedly reduced when the appropriate selection criteria are used. However, theoretical considerations would suggest that values based on a comparison of bone conduction before and after surgery underestimate the size of the Carhart effect. This is confirmed at least at low frequencies, by a series of experiments on six normal individuals in whom an artificial conductive hearing loss was created by increasing the ear canal pressure. The change in bone conduction threshold was 5 dB at 250 Hz, and 15 dB at 500 Hz, changes greater than those estimated from surgical data.