Klaske van Sluis

152 9.1. Voice and speech outcomes Automated acoustic outcome measures The studies included in the systematic literature review presented in Chap- ter 2 use a great number of acoustic outcome measures. Therefore, in this thesis we chose to evaluate a range of traditional acoustic measures includ- ing fundamental frequency (F0), jitter, shimmer, intensity, and spectral tilt, as well as signal-to-noise ratio computations including Harmonics to Noise Ratio (HNR), and percentage of voicedness (%voiced) and maximum phonation time (MPT) [1, 2]. These outcome variables are frequently used in the literature, al- though the validity and clinical utility of many of these acoustic measures has been strongly debated [3–5]. Some studies discussed in Chapter 2only included excellent speakers and excluded audio samples of tracheoesophageal speakers which were too aperiodic to perform acoustic analysis [6–10]. With the ape- riodic characteristics of tracheoesophageal speech, the use of these standard acoustic outcomes is debatable [1, 2, 11, 12]. It is troublesome to compare groups of speakers (over time, between studies etc.) with this high number of acoustic parameters. Each study applies different assessment protocols, record- ing equipment, and software for analysis which reduces the comparability of outcomes between studies. In chapter 3 and chapter 5, we therefore decided to use the Acoustic Voice Quality Index. To date, AVQI has not often been utilised for evaluating voice quality after total laryngectomy, however it provides several benefits in struc- tured voice evaluation and therefore is a promising tool. Since it produces one outcome score rather than many individual scores, it has a better usability and comparability within the subject as well as in-between subjects’ study designs. Chapter 3 shows the need for separate validation of this scale for use in tra- cheoesophageal speech. The AVQI scale has originally been developed to assess dysphonia in laryngeal speakers [5]. In our prospective study we show that pa- tients admitted to total laryngectomy already have abnormal AVQI values be- fore surgery, which worsen after surgery when evaluating the tracheoesophageal speech samples. The AVQI scores do correlate well with the perceptual rated voice quality and intelligibility scores. Therefore, AVQI provides consistent in- formation on both perceived voice quality and intelligibility. Patient-reported outcome measures In this thesis chapters 2 and 3 include patient-reported outcome measures (PROMs). With the help of these PROM questionnaires the patients’ function- ing has been assessed. This is an essential part in evaluating communicative functioning, including voice problems and intelligibility issues [13]. The studies in the systematic review show the use of different questionnaires, including VR- QOL, VHI-30, and VHI-10. None of these questionnaires have been specifically validated for total laryngectomy patients, but since these questionnaires are the best available instruments and widely used in dysphonia research, they are also used often in this patient group. However, there is a need for a questionnaire