Part of being a competent speech-language pathologist is knowing about and using the evidence that supports or refutes what we do with our patients and clients. One service provided by SLPs—managing patients with dysphagia who have tracheostomy tubes—may be particularly challenging and therefore might be an especially important area in which to consult available evidence.
In addition to understanding the clinical considerations typical for all patients—such as medical comorbidities, current medical status, and mental state—an SLP needs to be aware of additional factors pertaining to a patient's tracheostomy. These factors include the presence of a cuff and the effects of cuff inflation status; the size of the tracheostomy tube in relation to the patient's probable tracheal diameter; whether mechanical ventilation is in use; the ventilation mode and patient progress toward weaning; the patient's tolerance of expiratory occlusion and success in establishing exhalation airflow to the larynx and oral-nasal exits; the patient's tolerance of a one-way speaking valve; and secretion management. As clinicians, how do we know which of these factors will positively or negatively affect a patient's swallowing function? How do we know whether these factors will result in impaired biomechanics that lead to aspiration and associated morbidity? How can we determine through assessment, and minimize through intervention, the likelihood of any adverse effects?
Which of our assessment procedures most accurately predicts or identifies aspiration? Multiple studies have demonstrated high rates of silent aspiration in patients breathing through a tracheostomy tube (Ding & Logemann, 2005; Davis et al., 2002; Davis et al., 2004; Elpern et al., 2000). Is stand-alone clinical testing sufficiently robust, how sensitive are these clinical screening tools in predicting aspiration, and should they stand alone as diagnostic studies for this population? Many argue that identification of aspiration is best accomplished by instrumental testing (e.g., Elpern et al., 2000; Ding & Logemann, 2005), but which instrumental test is superior?
For the majority of patients with tracheostomies who are seen by SLPs, the ultimate goal is eventually to restore independent upper aerodigestive tract function. When SLPs are asked if these patients are "safe" to eat, or whether speech can be restored, SLPs assume that these goals are clinical endpoints. Resuming an oral diet and restoring verbal communication may be the main treatment goals for these patients, but often the ultimate clinical goal for the medical team and the patient is to achieve decannulation. Our rationale for assessments and management should be driven by this ultimate goal whenever possible, in addition to our typical short-term goals of "eliminating or reducing" aspiration and re-establishing speech.
With the short-term goal of eliminating aspiration, SLPs should focus on understanding that aspiration is one of the many risk factors that influence the pathogenesis of pneumonia in the tracheostomy population. Patients who have tracheotomies develop pneumonia at a much higher rate than patients without a tracheostomy—whether or not they are eating or drinking. Contamination of the respiratory system by oropharyngeal pathogens can occur when a patient aspirates food or liquid, saliva, or refluxed gastric contents, regardless of cuff inflation status. One recent study examining pneumonia rates in intensive care unit patients with tracheostomies showed a 67% pneumonia incidence at 28 days (Blot et al., 2008). Their counterparts who were orally intubated had a pneumonia rate of 59%, suggesting that the presence of the tracheostomy tube somehow increased these patients' risk of acquiring pneumonia beyond the already-high risk of pneumonia associated with mechanical ventilation.
Patients who are on a ventilator are six to 21 times more likely to develop pneumonia than those who are not and their risk increases by about 1% each day that they stay on the ventilator (Craven et al., 1986; Craven et al., 1988). These statistics underscore the importance of the ultimate goal of weaning and decannulation.
In the Centers for Disease Control and Prevention's Guidelines for Preventing Healthcare Associated Pneumonia (2004), the first recommendation is to remove the tracheostomy (or endotracheal tube) as soon as possible. The sooner patients are decannulated, the faster their risk for acquiring pneumonia decreases. It is paramount that SLPs consider how to help restore the oral intake of food and liquids and how to help speech so that as many of these patients can be decannulated as soon as possible.
The following is a brief review of some of the most pertinent literature that addresses key factors in the assessment of patients with tracheostomy tubes. There is no universally accepted protocol or set of considerations that SLPs rely on when managing dysphagia in patients with tracheostomy tubes. This review identifies important factors to consider when making decisions about assessment and management.
Tracheostomy Tube and Dysphagia
Does the presence of a tracheostomy tube cause dysphagia? Maybe not. Terk et al. (2007) gathered videofluoroscopic data to evaluate the hypothesis that a tracheostomy tube tethers the larynx, thereby reducing hyolaryngeal excursion, contributing to pharyngeal phase dysfunction. Among seven adult participants who had tracheostomies, normal swallowing, and who could tolerate decannulation, there were no differences observed in hyoid bone movement or laryngeal excursion in any of the three conditions tested:
- tracheostomy tube in and open with a 5-cc air-inflated cuff
- tracheostomy tube in and capped with a deflated cuff
- tracheostomy tube out (decannulated)
Donzelli et al. (2005) investigated this same question by examining aspiration using fiberoptic endoscopic evaluation of swallowing (FEES) in patients with a tracheostomy tube in place, and then once again immediately following decannulation. Among the 37 patients tested, 35 demonstrated consistent aspiration status with and without tracheostomy tubes in place. The overall rate of aspiration in this study was only 8% with the tube in place and 13.5% with the tube removed, rate much lower than the previously reported rates of 42–53% expected for this population (e.g., Brady et al., 1999; Donzelli et al., 2001; O'Neil-Pirozzi et al., 2003). This study indicates that it was possible that the patients were significantly recovered from their underlying illness and thus candidates for decannulation, indicating probable significant recovery of dysphagia before evaluation. Given the contradictory results in terms of much lower than expected aspiration rates, the timing of evaluation in relation to patient recovery cannot be ignored.
Patients undergo tracheotomies because of the nature of their underlying medical conditions. These patients often have neuromuscular disorders, head and neck cancers, respiratory diseases, unexpected postoperative complications, sepsis, and other medical conditions that can predispose them to short- or long-term dysphagia. The presence of dysphagia is likely due to the medical diagnosis and not to the presence of the tracheostomy tube. It remains possible, though, that certain tracheostomy tube conditions could contribute to or exacerbate swallowing dysfunction among some patients with tracheostomies. More research is needed to investigate the possible interactions among patient factors, primary medical condition, tracheostomy tube condition, and swallowing status.
Considerations During Assessment
Several factors regarding a patient's tracheostomy tube need to be considered to assess dysphagia in this population. For example, should the tracheostomy tube cuff be inflated or deflated? Does the presence or absence of a one-way speaking valve matter in reducing aspiration risk? Will a blue dye test be informative if a patient is aspirating? While these questions are often raised, there is little consensus on some of the answers.
1. Cuff Status
It is not uncommon to hear physicians, nurses, or respiratory therapists on the hospital units say, "keep his cuff up so he doesn't aspirate." Because of the location of the tracheostomy tube, however, an inflated cuff sitting below the glottis will not prevent aspiration, but at most, delay the arrival of the aspirated material into the lungs. Several studies have sought to examine the effect of cuff inflation status on aspiration risk.
Ding and Logemann (2005) retrospectively examined the effects of cuff status on aspiration rate in a group of patients with varying medical diagnoses. After examining the modified barium swallow (MBS) results of the 623 patients, the researchers observed that the overall rate of aspiration was nearly 65% and of those who aspirated, 30% did so silently. Using a regression analysis, they determined that patients with inflated cuffs were more likely to aspirate and were also at higher risk to aspirate silently than those whose cuffs were deflated. Unfortunately, due to the retrospective nature of the study, the authors were not able to compare the two cuff conditions within patients to see if the cuff status made a difference for individual patients.
In another study that examined the effect of cuff status on aspiration of various consistencies of food and liquid, Davis et al. (2002) prospectively examined 12 patients with videofluoroscopy under both cuff conditions. The rate of aspiration was 17.8% when the cuff was inflated and dropped to 6.5% when the cuff was deflated. According to the researchers, cuff status was an independent predictor of aspiration in their sample. Both studies proposed that laryngeal desensitization and lack of glottal airflow could have contributed to the higher rates of aspiration in the cuff inflated conditions. Table 1 [PDF] includes additional details about both of these studies.
2. Occlusion Status
Several studies also have examined the effectiveness of reducing or eliminating aspiration related to swallowing when a one-way speaking valve is used. Table 2 [PDF] provides additional details about these studies. One of the criteria for one-way valve usage is that a patient must be able to tolerate cuff deflation or have a cuffless tracheostomy tube. Does this mean that patients who cannot wear a one-way valve are at higher risk of aspiration and therefore should not eat or drink until they are able to tolerate a valve? The evidence is not entirely clear on this.
Suiter et al. (2003) found a reduction in the rate of aspiration of thin liquids when a one-way speaking valve was used compared to an unoccluded tracheostomy tube condition. This difference was significant at 21.4% in the valve-on condition versus 71.4% and 100% for no valve with cuff deflated and inflated conditions, respectively. Similarly, both Elpern et al. (2000) and Stachler et al. (1996) found a reduction in aspiration with a one-way speaking valve on. Elpern et al. found that using a one-way valve eliminated aspiration in five of their seven patients who had aspirated with an unoccluded tube during their study. In the Stachler et al. study, eight of 11 post-treatment head and neck cancer patients demonstrated a reduction in aspiration while wearing a one-way valve versus an unoccluded tracheostomy tube. Although aspiration was not eliminated for any of their participants, the amount aspirated was reduced. This is one of few studies in which the amount of aspiration is mentioned. As it is not clear how much aspiration is enough to raise a patient's pneumonia risk, it is critically important that a fairly precise measure of the amount of aspiration be developed, standardized, and reported in future research so that the relationship between the amount of aspiration observed during assessment and the degree of pneumonia risk could be clarified.
The researchers in these three studies all hypothesized that a one-way valve may reduce the risk of aspiration because it restores some of the subglottic pressure that is lost through the tracheostomy tube at the onset of swallowing-related laryngeal closure. Most people swallow at the onset of exhalation due to elastic recoil of the chest wall and alveoli; when an unoccluded tracheostomy tube is present, airflow can escape below the true vocal folds even though they are adducted during the swallow, subsequently dropping or eliminating positive subglottic air pressure (Gross et al., 2006). Although it is somewhat self-evident that the pressure in the airway is higher than the atmospheric pressure during exhalation, the hypothesis that restoring positive subglottic pressure and airflow through the larynx and pharynx is beneficial to airway protection has been investigated by several researchers.
Gross et al. (2006) examined subglottic pressure changes during swallowing in a healthy non-tracheotomized person and found that during normal swallowing, subglottic pressure is typically higher than atmospheric pressure. This can be altered by the amount of air present in the lungs at the time of the swallow, resulting in negative subglottic pressure when lung volume is low (i.e., below functional residual capacity) during the swallow, or if the patient swallows after beginning to inhale. In another study that examined the swallowing function of a patient with a tracheostomy, Gross et al. (2003) found that the amount of supra-atmospheric subglottic pressure during the swallow was reduced when the participant's tracheostomy tube was unoccluded and was greater when it was occluded.
Several other differences were noted in terms of swallowing physiology within patients between occluded and unoccluded tube conditions. In addition to the role that positive subglottic air pressure plays in improving airway protection, airflow through the glottis may improve sensation and thereby help to maintain sufficient sensory integrity to detect material that enters the laryngeal vestibule. Additionally, sufficient subglottic pressures are needed to throat clear and to expel material before it is aspirated. Thus, the increased subglottic pressure associated with a one-way valve is hypothesized to allow for improved airway protection during the swallow as well as during the post-swallow clearance of material.
Because aspiration, and more particularly silent aspiration, is a common occurrence in patients with tracheostomy tubes (Ding and Logemann, 2005; Davis et al. 2002; 2004; Elpern et al., 2000), thorough assessment and evaluation of all risk factors is indicated prior to the initiation of oral intake to reduce the likelihood of complications associated with aspiration. While instrumental testing is essential in providing the most accurate assessment of swallowing function (e.g. Elpern et al., 2000; Ding & Logemann, 2005), a clinical assessment may be the only type of assessment some patients undergo. O'Neil-Pirozzi et al. (2003) reported that a clinical swallowing exam which utilizes blue dye testing can serve as an aspiration screening prior to instrumental testing. They also report that the clinical exam can be used when the patient is not stable enough for an instrumental exam or because access to an instrumental exam is time or cost prohibitive. In cases where it is the only exam performed, clinicians should keep in mind the low sensitivity of the clinical exam procedures discussed in the following section, in detecting aspiration in this population.
The Evan's Blue Dye Test (EBDT; Cameron et al., 1973) and Modified Evan's Blue Dye Test (MEBDT; Thompson-Henry and Braddock, 1995) were developed as screening tools for detecting aspiration of oral secretions and food or liquids, respectively. In the EBDT, the patient's oral secretions are stained by placing blue dye on the tongue and observing the patient over a period of time, and in the MEBDT, the blue contrast is added to boluses of food or liquid, with both involving serial suctioning of tracheal secretions. The presence of blue dye in the tracheal secretions is a positive sign that aspiration has occurred, but its absence does not necessarily indicate no risk of aspiration and should not be taken to mean that the patient did not or will not aspirate.
A number of studies have compared the accuracy of the MEBDT in identifying aspiration to the results of simultaneous or successive instrumental exams. Overall, the studies assessing the accuracy of the MEBDT in identifying aspiration show much disagreement with instrumental testing (0-82% overall reported accuracy) because of the lack of a standard methodology for investigating this research question. Table 3 [PDF] provides additional study details. These studies have sought to compare the MEBDT and instrumental swallowing assessments to determine the sensitivity (accuracy of detecting aspiration when it does occur) and specificity (accuracy of correctly identifying the absence of aspiration when it does not occur) of the blue dye method.
A major flaw in most studies examining the use of blue dye tests as a screening tool is their non-concurrent designs. When the blue dye and instrumental tests are not performed simultaneously, this design limitation weakens the reader's confidence in the reported sensitivity and specificity results. For example, Thompson-Henry and Braddock (1995) completed a FEES or MBS on five tracheostomized patients within four to 22 days following the MEBDT. They found that all of the patients aspirated during the instrumental assessments, but none of the MEBDTs were positive. In this first study that used blue dye in food and liquid consistencies, the researchers identified potential false negative findings using the blue dye methodology. Belafsky et al. (2003) completed FEES exams 24 hours following MEBDT on 30 patients. All of the patients had either cuffless tracheostomy tubes or their cuff was deflated; none were wearing one-way speaking valves; and 10 were on the ventilator during the assessments. The overall rate of aspiration for the cohort was 73% with FEES, and the sensitivity and specificity of the MEBDT in detecting aspiration was 38% and 82%, respectively.
Three studies using concurrent MEBDT and instrumental testing deserve attention. Donzelli et al. (2001) performed simultaneous FEES and MEBDT on 15 patients and found an overall 50% sensitivity, and aspiration was identified with blue dye only when it occurred in more than trace amounts. Large amounts of aspiration were detected more accurately (67% of the time) than small aspiration volumes (0%). Unfortunately data on the seven patients that did not aspirate were not included to determine the specificity of the MEBDT. A possible methodological limitation, which may have impacted their findings, was the use of FEES as their instrumental method. Aspiration during the swallow may go undetected during FEES due to the loss of endoscopic visualization that occurs during the swallow, leaving the endoscopist to look for evidence of aspiration that may have eluded their line of sight by the time visualization is restored. However these investigators actually inserted the endoscope into the trachea to directly visualize the subglottal area to confirm whether aspiration had indeed occurred. Similar results were observed by Brady et al. (1999) who performed concurrent MEBDT and videofluoroscopic swallowing (VFS) evaluation were with 20 patients.
Using simultaneous videofluoroscopy and MEBDT (both blinded), O'Neil-Pirozzi et al. (2003) evaluated 50 patients. They found 62% sensitivity and 79% specificity (the published study erroneously reversed the values of these two results). They also were able to identify "trace" and "gross" amounts of aspiration using the MEBDT procedure in their study. From these studies it seems that small-volume aspiration is not easily detected by MEBDT, and that overall, about 30% of aspiration goes undetected.
Many argue that the gold standard for instrumental swallowing assessments is the videofluoroscopic swallowing (VFS) evaluation, while others argue that fiberoptic endoscopic evaluation of swallowing (FEES) is equivalent without the radiation exposure and access issues that VFS evaluation may have. Some patients cannot be transported to radiology or undergo fiberoptic testing for various reasons, and in most cases clinical evaluation is performed before instrumental testing. Each instrumental evaluation method has its advantages and disadvantages and both provide valuable information that overshadows clinical observations in terms of precision. The debate regarding superiority of VFS or FEES notwithstanding, instrumental testing should be performed in the majority of patients with tracheostomies, unless the clinician is reasonably certain that swallowing function is a highly unlikely source of increased risk of adverse medical outcomes.
A Reduction in Morbidity
The aforementioned studies focus on the incidence of aspiration, but not as much on the amount of aspiration. While identifying aspiration and reducing or eliminating its occurrence seems like an intuitively reasonable goal, unfortunately it remains unclear whether small amounts of aspiration matter and for what materials. We do know that having a tracheostomy tube is a pneumonia risk factor and decannulation should be a goal that SLPs help work towards to eliminate this risk factor. This population is often tenuous and fragile, and their day-to-day—or even hour-to-hour—performance can vary dramatically. The majority of these patients are seen by SLPs in acute care settings and are progressing toward decannulation and eventual restoration of upper aerodigestive tract integrity. These efforts toward decannulation should take precedence over efforts to restore oral intake in patients who are making good progress toward decannulation.
In an attempt to identify the main factors SLPs should consider when assessing the swallowing function of this population, we should consult the literature whenever possible. The evidence supporting the use of blue-dye based screening for the presence of aspiration underscores two important points. First, it is best considered a screening test that can identify gross amounts of aspiration with a fair degree of precision, but for patients who do not aspirate gross amounts, there is a considerable risk that they will be misidentified as non-aspirators. Given the high variability among patients, it can also be used on a case-by-case basis to make clinical decisions regarding such things as candidacy or readiness for instrumental evaluation. The overwhelming majority of patients with tracheotomies should undergo instrumental testing because they are likely to have dysphagia and because they exhibit a high rate of silent aspiration. The research literature has also provided evidence that supports the notion that a one-way speaking valve reduces aspiration risk for some, but not all, patients. In addition to the benefit of enabling verbal communication, a one-way valve may also a benefit some patients by improving swallowing function. When feasible, one-way speaking valve assessments should be considered in addition to standard clinical assessments of dysphagia which should include evaluation of the sensorimotor integrity of cranial nerves V, VII, and IX-XII, screening of cognitive and language function, and observation of trial swallows in this population, and evaluated on a case-by-case basis.
Aspiration absolutely matters when patients develop pulmonary complications or are unable to meet their nutrition and hydration needs as a result. Clinical experience suggests that some patients who aspirate do not get sick while others do. More research needs to be conducted to identify the most predictive factors concerning which patients will be more likely to suffer adverse events as a result of aspirating. As professionals, our focus should shift toward reducing morbidity—and be less narrowly focused on just eliminating aspiration, because this often cannot be achieved in some of the patients we assist in managing.
We can make an educated guess as to whether a patient with a tracheostomy aspirates while swallowing and make efforts to minimize it. But what does that mean in terms of reducing their risk of developing pneumonia? As suggested above, the history of mechanical ventilation and the presence of the tracheostomy tube itself are independent predictors of pneumonia. So, when a patient with a tracheostomy develops pneumonia, is it aspiration pneumonia, ventilator-associated pneumonia, or hospital-acquired non-aspiration-related pneumonia?
The answer is not simple, and is discussed in the accompanying article in (insert hyperlink to article online). Among factors to consider are the duration of mechanical ventilation, number of invasive bronchoscopic procedures, bacteriology of protected sputum specimens, nature of the patient's underlying conditions—particularly pulmonary disease, immunological competence, duration of endotracheal intubation preceding tracheostomy, and others. In the final analysis, it may be less risky to some patients with upper digestive conditions or other factors predisposing to aspiration of gastric contents, to aspirate a little while swallowing than to undergo feeding tube placement. In this population (and in the population at large), the risk of pneumonia cannot be lowered to zero. The best we can do is to reduce it to the lowest acceptable level for our patient.