Visualization

Visualization of swallowing function is critical for a thorough understanding of biomechanics, airway protection, anatomy and mucosal integrity. There are numerous methods for dynamic imaging of swallowing, namely VFSS (Logemann, 1983), videoendoscopic evaluation of swallowing (VEES; Bax, McFarlane, Green, & Miles, 2014; Langmore, Schatz, & Olsen, 1988; Leder & Murray, 2008; Lim et al., 2001), ultrasonography (Macrae, Jones, Myall, Melzer, & Huckabee, 2013), magnetic resonance imaging (MRI; Kreeft et al., 2012) and dynamic CT (Fujii et al., 2011; Inamoto et al., 2011). These

techniques can be thought of as complementary, with individual strengths and limitations. For example, although dynamic CT imaging provides unparalleled three-dimensional analysis of swallowing, it is limited by expense and substantial radiation exposure, and is therefore limited to use in specialized research laboratories (Fujii et al., 2011; Inamoto et al., 2011; Kobayashi, Koshida, Suzuki, & Katada, 2012). Contrastingly, ultrasonography is non- invasive and widely available in health care settings, however, it only reliably provides a narrow insight into lingual surface movement, hyoid movement and two-dimensional muscle diameter (Macrae, Doeltgen, Jones, & Huckabee, 2012; Macrae et al., 2013).

The two most widely utilized visualization techniques are VFSS and VEES (Langmore, 2003). Both VFSS and VEES can be performed across the lifespan, in nearly all dysphagic aetiologies and has been proven to be superior than clinical bedside evaluations alone (Wilson & Howe, 2012). VEES which consists of trans-nasal insertion of a fibreoptic endoscope to record intra-luminal swallowing (Bax et al., 2014; Lim et al., 2001; Tohara et al., 2010). VEES examinations have superior capability for evaluation of mucosal and anatomic integrity given direct visualization of the soft-tissue structures. With no radiation exposure, VEES can be used as a training and biofeedback tool during rehabilitation without contraindications inherent in repeating procedure (Langmore et al., 1988). However, VEES has notable limitations, including a ‘white-out’ period during peak pharyngeal swallowing due to intraluminal constriction obscuring the recording from the endoscope. Further, VEES cannot provide information regarding oral and oesophageal phases of swallowing or evaluate critical biomechanics such as hyoid movement. Research indicates ratings of aspiration (Kelly, Drinnan, & Leslie, 2007) and residue (Kelly, Leslie, Beale, Payten, & Drinnan, 2006) are significantly higher following evaluation with VEES as compared to VFSS.

VFSS is largely considered the gold-standard in evaluation of deglutition as it can visualize all stages of swallowing as an integrated process (Rugiu, 2007) and has been utilized in research and clinical practice for over thirty years (Logemann, 1983). VFSS uses ionizing radiation to visualize a bolus impregnated with a contrast agent, typically barium. This instrumentation provides two-dimensional, dynamic, radiographic images of swallowing, allowing frame-by-frame analysis of ingestive biomechanics (Feinberg, 1993). VFSS evaluates temporal characteristics of swallowing, including duration and onset of swallowing

kinematic events (Bardan, Kern, Arndorfer, Hofmann, & Shaker, 2006; Humbert et al., 2013; Macrae, Anderson, Taylor-Kamara, & Humbert, 2014; Sia, Carvajal, Carnaby-Mann, & Crary, 2012), integrity of airway protection (Feinberg, 1993; Hind et al., 2009; Kelly et al., 2007; Robbins, Coyle, Rosenbek, Roecker, & Wood, 1999) and effects of compensation (Baylow et al., 2009; Bülow, Olsson, & Ekberg, 1999, 2001, 2002). Further, patient performance on VFSS has been linked with important predictive outcomes. In a study comparing patients with dysphagia (n = 26) to case matched controls (n = 33), Schmidt et al. (1994) found that of patients who aspirated thickened liquids or more solid consistencies on VFSS, the odds ratio for developing pneumonia was 5.6 times greater compared with those who did not aspirate, or who aspirated thin liquids only. Further, the odds ratio for death was 9.2 times greater for patients who aspirated thickened liquids and more solid consistencies compared with those who did not aspirate or who aspirated thin liquids only (Schmidt, Holas, Halvorson, & Reding, 1994).

Despite the undisputed utility of VFSS, there are limitations. It exposes patients to ionizing radiation, which can have short and long-term health implications. Dose is related to screening time and acquisition/frame rate, which is dependent on the complexity of the swallowing assessment and patient compliance (Wright et al., 1998). Nevertheless, doses recorded in reviews of exposure in clinical practice are < 1.23 mSv (SD = 0.64) (Kim et al., 2013), which compares favourably to a single full-body CT-scan of 10 to 30 mSv (Brenner & Hall, 2007). Results have indicated more than 40 VFSS examinations would need to be administered per year to exceed an individual’s annual radiation dose exposure limit (Kim et al., 2013).

Additionally, researchers have highlighted pronounced concerns regarding reliability in interpretation of VFSS. Ekberg et al. (1998) investigated reliability of radiologists scoring VFSS. While the highest reliability was found for identification of aspiration (k = 0.83), the lowest concurrence was for critical swallowing parameters such as decreased or absent pharyngeal constriction and delayed opening of the UES (k < 0.40). Even in implementation of standard protocols, reliability has been found to be similarly poor (k = 0.01–0.56) (Stoeckli, Huisman, Seifert, & Martin-Harris, 2003). This poor agreement has been replicated with Speech-Language Pathologists in similar studies (Kuhlemeier, Yates, & Palmer, 1998; Scott, Perry, & Bench, 1998). In a recent systematic review, Baijens et al. (2013) reported

measurements in VFSS varied considerably with ICC between 0.22–0.84, depending on method of measurement, pre-experimental training and bolus consistency used. The authors report studies investigating intra- and interrater reliability VFSS measurements were of poor methodological quality (Baijens, Barikroo, & Pilz, 2013). Increases in inter-rater reliability to greater that 80% have been achieved with standard protocols (Martin-Harris et al., 2008) but these programmes require criterion-referenced training and the development of the rating protocol was not available for stringent peer review due to commercialisation concerns, raising concerns regarding validity. Further, VFSS cannot provide information on the underlying nature of impairment, such as weakness, spasticity, apraxia, or other neuromuscular change. Although impaired biomechanics may be observed, this observation cannot define the underlying pathophysiology. Although undoubtedly a valuable tool for visualising swallowing biomechanics, the widespread dependence on VFSS in isolation may be contributing to misdiagnosis in the evaluation of dysphagia. There is a need for more objective adjuncts to make the current gold standard diagnostic imaging technique less biased and increasingly robust.