To the best of this author’s knowledge this is the first study conducted on patients with traumatic cervical SCI that has used a stringently defined VFSS protocol and a standardized measure (Rosenbek et al., 1996) to determine the severity and recovery of laryngeal penetration-aspiration. Furthermore, the current study is the first that has focused on the clinical signs and the risk factors of laryngeal penetration- aspiration during the acute phase in patients with traumatic SCI. Additionally, a standardized method was used to describe a functional oral feeding outcome of the patients (Crary et al., 2005). The bedside evaluation and swallowing trial were administered to the patients in an established manner as described in Chapter 4.4.1;
this was designed to be brief and thus appropriate for patients treated in intensive care settings. In addition, all the patients were evaluated by the same speech therapist and therefore inter-evaluator bias was eliminated. However, intra-evaluator bias was not examined. Furthermore, the exact time frames between the injury and the different assessments points have been presented in detail in contrast to the majority of prior studies as seen in Table 1 on pages 29–30.
In addition, the current study sample consists of a consecutive series of adult patients with traumatic cervical SCI recruited with rigorous exclusion criteria (see Figure 4 on page 36 for details). Premorbid conditions such as previous diseases or surgical procedures that could cause dysphagia were carefully excluded to enhance the validity of the data. The current study sample – even though relatively small – can also be considered as being representative of Finnish patients with traumatic cervical SCI as it examined a consecutive series of patients admitted to one of the three University Hospitals responsible for the care of this patient group in Finland.
The distributions of age and gender as well as the injury severity and mechanisms distributions are comparable with those published by Koskinen and co-researchers (Koskinen et al., 2014) who assessed the incidence of traumatic SCI injuries in Finland. Still, it can be argued that the incidence of penetration-aspiration would have been higher if the excluded cases, especially those with premorbid conditions, had been included in the study.
Naturally, this study also has some limitations and methodological issues that need to be discussed. First, the small sample size, restricted the statistical analyses and further limited the generalizability of the results. For example, the small sample size reduced the possibility to perform extensive statistical testing (e.g. multivariable
in the different subgroups more sophisticated tests would have been underpowered and would have not provided reliable evidence. Second, since no control group was available, it is not possible to differentiate the course of spontaneous recovery from the consequences of treatment. Thus, the speech therapeutic interventions were only described and summarised and no assumptions can be drawn concerning the effectiveness of these interventions. Based on ethical considerations (e.g. not to expose people to radiation unnecessarily), a healthy control group could not be enrolled in this study. This naturally led to the absence of subsequent normative data.
Furthermore, one source of weakness in this study was that our x-ray equipment allowed only a resolution of 15 images per second. In the literature, a frame rate of 30 images per second is recommended in order not to miss any significant pathology (Belafsky & Kuhn, 2014; Martin-Harris & Jones, 2008). For example, Bonilha and colleagues (2013) explored the agreement between PAS ratings at different temporal resolutions; they revealed that one obtained different PAS ratings if one applied a high resolution of 30 images per second as compared to those at lower pulse rates (4, 7.5 and 15) in 80% of the patients studied. However, the difference in PAS ratings between 30 and 15 images per second was not statistically significant. Thus, 15 frames per second can be considered as satisfactory.
Next, the VFSS protocol included only measured boluses (5 ml, 10 ml, and 20 ml) of a thin, liquid consistency in order to reduce the amount of radiation exposure in those patients who did not show signs of dysphagia and penetration-aspiration. A thin liquid was chosen since it is the most sensitive consistency for detecting penetration-aspiration: thicker consistencies are clinically used as compensation strategies for aspiration of thin liquids (Clave et al., 2006; Leonard, White, McKenzie,
& Belafsky, 2014; Logemann et al., 2008). Naturally, a systematic use of a wider range of volumes and different consistencies, as for example applied in the MBSImp – protocol (Martin-Harris et al., 2008) in the whole study sample would have enabled a more extensive evaluation of the swallowing function.
The limited approach to determining dysphagia chosen for this study may also be criticized. The present study focused only on penetration-aspiration findings emerging from the bedside examination and VFSS, although dysphagia is a much broader phenomenon including problems in various phases of swallowing (preparatory, oral, pharyngeal and esophageal). However, this choice is defensible as, in acute clinical settings, it is crucial to identify those patients at risk of penetration-
pulmonary complications (Vazquez et al., 2013; Winslow & Rozovsky, 2003;
Zimmer et al., 2007).
The decision to include only the patient’s worst (i.e. highest) PAS score in the statistical analyses may also be criticized. In comparison, some studies have used the mean of PAS scores as an outcome measure (Langmore et al., 2016; Pearson, Davidoff, Smith, Adams, & Langmore, 2016). Nonetheless, as Steele and Grace- Martin (2017) have emphasized, two individuals may have the same mean score but one of them may still suffer from a clinically more serious condition. For instance, one patient scoring 5,3,3,5,5 and the second scoring 8,5,3,3,2 both have a mean score of 4.2. However, it is apparent that the second patient with silent aspiration (PAS 8) probably needs more speech therapeutic interventions than the first patient. Steele and Grace-Martin (2017) suggested that at least for clinical purposes, the most informative way to represent PAS scores may be to report both the most typical (mode) and the worst score across a set of swallows. Unfortunately, the information provided by Steele and Grace-Martin (2017) was not available when designing the study protocol for the current study. Nonetheless, the decision to emphasize the worst PAS score is defensible especially in acute and intensive care settings, which was the context of this study. It is well established that VFSS verified aspiration is associated with an increased risk of pneumonia in acute and sub-acute settings (Holas, DePippo, & Reding, 1994; Kidd, Lawson, Nesbitt, & MacMahon, 1995;
Pikus et al., 2003). In addition, a few studies have reported that the incidence of pneumonia is significantly higher in cervical SCI patients with dysphagia than in those without dysphagia (Abel et al., 2004; Chaw et al., 2012; Shem et al., 2011; Shem et al., 2012b). One important goal in the acute treatment of traumatic cervical SCI is to prevent pulmonary complications.
In addition, the PAS scoring system itself needs to be discussed. Scores 4 and 6 appear to be rarely observed (Kelly, Drinnan, & Leslie, 2007; Martin-Harris et al., 2008; Molfenter & Steele, 2014; Rosenbek et al., 1996) as is also apparent in the current study. This naturally affects the distribution of the scores. Furthermore, Steele and Grace-Martin (2017) have raised the question if the PAS is an linear scale i.e. a scale in which each increasing score is understood to represent greater severity than the previous score. The authors have questioned if PAS scores of 3 and 5 should be considered more severe than scores of 4 and 6 because penetration into the supraglottic space (PAS 3 or 5) does not trigger an immediate swallow response with effective clearance of the penetrated material. Therefore, Steele and Grace-Martin
function, and PAS scores of 3, 5 and 6 would refer to abnormal penetration. A PAS score of 7 would reflect a failure of airway protection mechanisms with some residual sensory integrity. Finally, a PAS score of 8 would refer to impaired sensory integrity and ineffective cough responses to aspiration. Further research and expert opinions on this subject are warranted.