Klaske van Sluis

Objective and subjective voice outcomes: a systematic review 31 Figure 2.4: Left violin plot displaying the distribution of the mean jitter outcomes of eight studies [27, 32, 35, 38, 42, 44, 49, 52] ES n =87, TES n =102, H n =105. Right violin plot displaying the distribution of the mean shimmer outcomes of five studies ES n =65, TES n =65, H n =95 [27, 32, 38, 44, 49]. Abbreviations: ES = esophageal speakers; TES = tracheoesophageal speakers; H=Healthy Speaker; ELS = electrolarynx speakers; HNR = harmonics to noise ratio; MPT = maximum phonation time. definite trend is seen between the speech rehabilitation methods ES and TES ( n =3, Table 2.3) [27, 49, 52]. Intensity Four level B categorized studies [34, 41, 49, 52] ( n =113) reported intensity scores (Table 2.3). Intensity in decibel is not an absolute value, and therefore comparing the mean outcome is irrelevant. A higher intensity is indicated as being better [1]. Highest intensity scores are found for healthy speakers [34, 49]. Between substitute voices four studies [34, 41, 49, 52] found higher intensity scores for TES than for ES. In only one of these studies the higher intensity for TES compared to ES was significant [52]. Spectral tilt Two studies [32, 37] ( n =64) calculated the ratio between the energy above 4 kHz and the energy in the lower frequencies. One study was categorized as level A [32], the other as level B [37]. A larger ratio is correlating with better voice quality [55]. The level A rated study of Shim et al. [32] reported a larger ratio in healthy speakers than in ES. In the other study [37] the spectral tilt ratio was found to be larger for TES than for ES. It is not possible to draw overall conclusion from this due to the small number of studies reporting this outcome measure.