Limitations of the computerized laryngeal analyzer (CLA) in the assessment of the swallowing mechanism.

This study was devised to assess the repeatability of measurements obtained from the Computerized Laryngeal Analyzer (CLA), a commercial system which monitors global laryngeal activity through a piezoelectric transducer positioned at the level of the thyroid cartilage. Twenty-two healthy subjects (37+/-7 yr) were asked to repeat three consecutive times the deglutition of three types of bolus: dry (saliva only), liquid (10 ml water) and solid (1 cm(3) biscuit). The whole recording session was repeated 4+/-3 days apart. Two parameters were considered: amplitude and duration of the deglutition. Both intra- and inter-sessions repeatability were assessed by the standard error of measurement. The amplitude measurements were often found to be clipped at 100 mV, making duration the only parameter suitable for analysis. The duration was homogenous for the three types of bolus and no trend effect was found either intra- or inter-sessions. Inter-sessions repeatability was significantly lower than the intra-session one (P<0.05). As a practical implication, a change in duration >/=1 s (=65% of the observed mean value) between two separate measurements is required in order to be 95% confident that it is not due to chance alone. We conclude that the observed limitations in the measurements provided by the CLA system, in conjunction with its intrinsic characteristic of providing only an uncalibrated measurement of global laryngeal activity, seriously question the possibility of using the CLA system either as a diagnostic tool or as a substitute of the gold standard method (videofluorography).