Effects of non-invasive ventilation on middle ear function in healthy volunteers

Objectives: To evaluate the effects of non-invasive ventilation (NIV) with facial mask or helmet on middle ear (ME). Design: Prospective, randomised study. Setting: University hospital. Participants: Ten healthy subjects randomly allocated in two groups of five subjects each. Interventions: NIV for I h, with helmet (group H) or facial mask (group M). Flow-triggered pressure support was 10 cmH(2)O, PEEP 5 cmH(2)O, FiO(2) 0.21. Measurements and results: Impedenzometry was performed before NIV and 5 min after NIV ended; it was repeated 60 min later. In group H the acoustic compliance increased after NIV from 2.0+/-.6 ml to 2.3+/-.6 ml (P<.01), suggesting that the tympanic membrane became less stiff; I h later the compliance returned to basal values (2.0+/-.7 ml); in group M the compliance was unaffected (from 2.0+/-.5 ml to 2.0+/-.4 ml; 1.9+/-.4 ml I h later). The acoustic reflex, i.e., the contraction of the stapedial muscle in response to an auditory stimulus, involving the acoustic and facial nerves, was also evaluated during impedenzometry at 250 Hz, 500 Hz, 1,000 Hz, and 4,000 Hz; no significant change of the threshold was observed. Conclusions: The tympanic membrane is tighten by the tensor tympani and a reversible loosening suggests muscle fatigue in response to the application of intermittent positive pressure applied to the external ear during NIV with helmet. The loss of tensor tympani protective action could theoretically predispose the middle and inner ear to mechanical damage during NIV with helmet, suggesting the use of protective devices (ear plugs) in selective cases requiring long-term, high-pressure treatment.