Abstract
BACKGROUND
Identifying accurate predictors of decannulation outcome is essential to ensure safe and timesaving decannulation. Studies indicated hypopharyngeal secretion retention is closely associated with compromised airway protection and patients with it are less likely to be decannulated. However, data verifying the link between secretion retention and decannulation outcome are lacking.
AIM
The aim of this study was to identify the association between hypopharyngeal secretion-retention (indicated by Murray Secretion Scale [MSS]) and decannulation outcomes in a large cohort of severe acquired brain injury (sABI) patients with tracheotomy. To test the diagnostic performance of secretion retention in decision making of decannulation.
DESIGN
Retrospective cross-sectional study.
SETTING
The setting of the study is neurorehabilitation department in a tertiary teaching hospital.
POPULATION
A total of 144 adult patients with sABI and tracheostomy were retrospectively selected from the database from September 1, 2019, to August 31, 2021.
METHODS
The results of hypopharyngeal secretion-retention observed by fiberoptic endoscopy on the day that decision on decannulation was made were collected. The association between severity of secretion retention and decannulation outcomes was investigated through logistic regression, which was used to adjust covariates, including presence of food/liquid aspiration, decreased laryngeal sensation and conscious level. The optimal cut-off values of MSS for decannulation status prediction was determined by maximizing the Youden Index.
RESULTS
One hundred twenty-one patients were included in the sample. The age was 55.6±15.2 years, 84 (69.4%) patients were male. Eighty-four (69.4%) of them were successfully decannulated during their hospital stay. Multivariable logistic regression analysis indicated severe secretion retention (MSS level3) was independently associated with prolonged tracheostomy (adjusted odds ratio 65.23, 95% CI 6.58-646.35, P<0.001). The sensitivity and specificity of MSS level3 to assess the probability of prolonged decannulation were 78.4% and 96.4%, respectively. The area under the curve was 0.894 (95% CI 0.819-0.969).
CONCLUSIONS
Our results add to the evidence supporting screening of secretion retention severity in sABI population to identify patients at risk of prolonged tracheostomy. Whether decreasing secretion retention increases probability of successful decannulation deserves to be investigated by further study.
CLINICAL REHABILITATION IMPACT
This study could provide evidence for establishing objective decannulation criteria based on fiberoptic endoscopy and be helpful for implementing targeted rehabilitation interventions to promote successful decannulation.
Key words: Tracheostomy, Rehabilitation, Airway management, Deglutition disorders
Tracheostomy is a life-saving procedure increasingly performed in patients with severe acquired brain injury (sABI) to improve airway protection and decrease mechanical-ventilation duration.1, 2 After the tracheostomy’s underlying indication has resolved, decannulation readiness must be evaluated. However, there is no universally accepted protocol for its performance and optimal decannulation criteria and timing need further explored.3, 4 Premature decannulation potentially results in airway obstruction, recurrent aspiration pneumonia, respiratory failure, and even death.5 Identifying accurate predictors of decannulation outcome – successful decannulation or prolonged tracheostomy – is essential to ensure safe and timesaving decannulation.
SABI patients potentially experience dysphagia because of various neurological changes, reduced consciousness, respiratory problems and tracheostomy, which may impact their ability to manage secretions.6 The poor pharyngeal clearance of secretions could lead to laryngeal penetration and aspiration, predisposing the patients to aspiration and aspiration pneumonia.7, 8 Several studies have employed observation of hypopharyngeal secretion retention as an essential screening step to exclude the patients without sufficient airway protection from decannulation in neurological Intensive Care Unit.9, 10 But the performance of hypopharyngeal secretion retention in diagnosis of readiness for tracheostomy removal has not been fully tested. Additionally, previous studies have reported more severe secretion retention in patients with a tracheostomy tube than in those without. And severe secretion retention has been associated with poor tube-capping tolerance in patients with tracheostomy, indicating a lower likelihood of progression to decannulation.11, 12 However, data verifying the link between hypopharyngeal secretion retention and decannulation outcome are lacking.
Fiberoptic endoscopy has been reliably applied in assessment of readiness for decannulation in the sABI population.9, 10, 13 It has been shown to be an objective and sensitive tool to allow the direct observation of secretions in the pharynx, and to evaluate swallowing function in patients with dysphagia.14 Food/liquid aspiration and decreased laryngeal sensitivity detected by fiberoptic endoscopy are closely linked to compromised airway protection, a potential cause of prolonged tracheostomy or decannulation failure in patients with sABI; however, current studies have not discussed them as potential decannulation-outcome predictors.15, 16 Furthermore, studies on the predictive power of a reduced conscious state on decannulation outcomes remain inconclusive.2, 3, 5, 17
This study aimed to determine the association between hypopharyngeal secretion retention severity and decannulation outcomes in a large cohort of patients with sABI who underwent tracheostomy. Secretion retention severity is rated using the Murray Secretion Scale (MSS), which was developed in 1997 with high inter-rater reliability.7 We hypothesized that severe secretion retention (MSS level 3) is a significant predictor of prolonged tracheostomy with high diagnostic sensitivity and specificity. It could provide evidence for establishing objective and safe decannulation criteria based on fiberoptic endoscopy and be helpful for implementing targeted rehabilitation interventions to promote successful decannulation.
Materials and methods
This retrospective, cross-sectional study was conducted in a tertiary teaching hospital. Eligible patients were adults with sABI (age≥18 years) undergoing tracheostomy who were admitted to our neurorehabilitation department for rehabilitation between September 1, 2019, and August 31, 2021. The exclusion criteria were as follows: hemodynamic instability, high risk of mechanical ventilation or neurologic deterioration, candidacy for neurosurgical surgery, neuromuscular-disease history, incomplete information on fiberoptic endoscopic evaluation of swallowing (FEES), and significant airway obstruction confirmed by fiberoptic endoscopy. Significant airway obstruction was considered if the patient presented with glottic stenosis or tracheal-lumen reduction >50% of the airway diameter during fiberoptic endoscopy examination.18 Owing to the study’s retrospective design, informed consent was waivered, and the study was approved by local Hospital Ethics Boards and Commissions on March 7, 2022, with protocol number 2022FXHEC-KSP009. The chairperson of the ethics committee is J S Zhang.
Decannulation protocol
During the studied period, from September 1, 2019, to August 31, 2021, all patients underwent postadmission routinely screening to determine decannulation readiness according to an internal decannulation protocol based on capping trail (Figure 1).
Figure 1.
—Flow chart of the study.
This protocol was undertaken by two attending physicians responsible for final decision making regarding decannulation. Successful decannulation was defined as without need for recannulation or oral intubation due to respiratory distress or desaturation 72h after decannulation. Successfully decannulated patients were defined as the decannulation group (D group). Those who did not progress to predischarge decannulation were assigned to the prolonged tracheostomy group (PT group).
FEES examination
On the day that tracheostomy decannulation was decided (the day of decannulation in group D and that before discharge in the PT group), FEES was routinely performed to confirm suitable food consistencies for the subsequent oral feeding training or to evaluate the treatment effect of swallowing therapy by a senior speech and language therapist (ST). All examinations were video recorded and performed with the tracheostomy occluded. Patients were evaluated at the bedside in a sitting position, with the bed head elevated to ≥70°. The endoscope was inserted into the nostril, anaesthetized using 0.2 mL of 4% lidocaine, and pushed forward along the nasal floor through the velopharyngeal port. The endoscope’s tip was advanced into the hypopharynx to observe secretion retention. Laryngeal sensation was tested by gently touching the aryepiglottic region with the endoscope’s tip. Swallowing function was assessed using a standardized set of food and liquid trials. Patients were administered the following bolus consistencies to swallow: 1) 3 mL puree consistency jelly; 2) 5 mL puree consistency jelly; 3) 3 mL blue-colored thin water; and 4) 5 mL blue-colored thin water. All boluses were delivered using a syringe that was initially placed on the tongue’s front half. When patients could not deliver food/liquid because of an inefficient oral phase, the food/liquid was placed at the back of the tongue. If aspiration was noted on any bolus, the second trial of that consistency (larger volume) was not administered.
Data collection
Baseline data were collected from electronic medical records as follows: demographic information (age and sex), time from acute events, length of hospital stay in our department, primary diagnosis (classified into four categories: stroke, traumatic brain injury, anoxic encephalopathy, and others, including brain tumors or encephalitis), brain-injury sites (categorized into two groups: supratentorial brain injury only and either supratentorial or infratentorial injury or that with isolated infratentorial involvement; this was based on the initial neuroimaging report available following acquired brain injury), and comorbidities (including hypertension, diabetes mellitus, coronary heart disease and stroke).
FEES outcomes on the day of deciding on tracheostomy decannulation were reviewed by two speech and language therapists to determine several factors. The first factor is represented by secretion-retention severity, based on the MSS, which has been used in dysphagia patients with or without tracheostomy to evaluate the secretion management ability and the related risk of food or liquid aspiration and pneumonia.7-10 The MSS is a 4-point scale ranging from 0 (normal, without secretion retention) to 3 (most severe). Level 1 indicates secretions evident in the laryngeal-vestibule surroundings. Level 3 is associated with secretion penetration into the laryngeal vestibule. Level 2 refers to any transition from level 1 to level 3 during the observation period. The second factor was laryngeal sensation: decreased laryngeal sensation was considered if the patient did not demonstrate bilateral laryngeal reflex (LAR) after touching each arytenoid’s mucosa with the endoscope’s tip.19 The third factor was represented by the presence of food/liquid aspiration: during the food/liquid swallowing examination, aspiration was determined if food/liquid entry into the airway below the true vocal-fold level was observed. Glasgow Coma Scale (GCS) scores on the same day were extracted from the electronic medical records. The GCS is a clinical scale used to reliably measure a person’s consciousness level after brain injury and ranges from 3 to 15 points. Patients with scores of 3-8 are usually considered to be in a coma.
Statistical analysis
Based on our previous study on decannulation, the proportion of patients with successful decannulation was approximately 60%.20 Sample size for logistic regression was calculated according to the rule of “10 outcome events per variable” per predictor variable. The estimated sample size was 100 subjects. Descriptive statistics were used to quantify patient characteristics, GCS, and FEES findings. Data are presented as frequencies for categorical variables and medians (interquartile ranges) for continuous variables. The independent t-test, Mann-Whitney U Test, χ2 test, and Fisher’s Exact Test were used to evaluate differences in baseline information between the D and PT groups. MSS inter-rater reliability was evaluated using the weighted kappa correlation coefficient. Univariate logistic regression was used to compare the GCS scores and three FEES parameters between the D and PT groups. To investigate the relationship between secretion-retention severity and prolonged tracheostomy, binary logistic regression was used to adjust for covariates. This procedure was performed using an entry strategy. Logistic regression results are presented as adjusted odds ratios (ORs) and their 95% confidence intervals (CIs). The optimal MSS cutoff values for prolonged tracheostomy prediction were determined by maximizing the Youden Index. Receiver operating characteristic (ROC) curves were generated for the MSS cutoff values, and areas under the ROC curves (AUCs) were computed. All statistical tests for the hypotheses were two-sided, and statistical significance was set at P<0.05. SPSS statistics (version 24.0; IBM Corp., Armonk, New York, NY, USA) was used for statistical analysis.
Results
A total of 144 patients with sABI who underwent tracheostomy were transferred to our department between September 1, 2019, and August 31, 2021. Twenty-three patients were excluded: two with hemodynamic instability; three with respiratory failure; two awaiting surgery; seven with incomplete FEES examination outcomes, including two unable to undergo the swallowing examination, three in whom laryngeal-sensation evaluation was infeasible because of a limited laryngeal view, and two in whom FEES examinations were not performed due to unplanned discharge; and nine with significant airway obstruction, including three with vocal-cord bilateral paresis, two with tracheomalacia, three with subglottic granulomas, and one with severe arytenoid mucosa edema (Figure 2).
Figure 2.
—Decannulation protocol.
Finally, 121 patients were included; the mean age was 55.6±15.2 years, 84 (69.4%) patients were male, the mean time from acute events to admission was 50.00 (33.50-93.50) days, and length of hospital stay was 46.00 (36.00–60.00) days. Most patients had stroke (64.5%). Traumatic brain injury (21.5%), anoxic encephalopathy (4.1%), and other conditions, including brain tumors or encephalitis (9.9%), were less frequent as primary diagnoses. Eighty-four (69.4%) patients were decannulated during their stay in our department, among whom none required predischarge re-cannulation, indicating that they underwent successful decannulation. The baseline characteristics of the final sample of 121 patients are presented in Table I. There were no statistically significant differences between the D and PT groups.
Table I. —Baseline characteristics.
| Characteristic | Total (N.=121) | D group (N.=84) | PT group (N.=37) | P value |
|---|---|---|---|---|
| Demographic | ||||
| Age, median (IQR), years | 55.00 (45.00, 66.00) | 57.50 (43.00, 65.00) | 61.00 (50.50, 67.00) | 0.104* |
| Male (N., %) | 84 (69.4) | 55 (65.5) | 29 (78.4) | 0.156# |
| Primary diagnosis (N., %) | 0.525† | |||
| Stroke | 78 (64.5) | 55 (65.5) | 23 (62.2) | |
| Brain trauma | 26 (21.5) | 19 (22.6) | 7 (18.9) | |
| Anoxia | 5 (4.1) | 2 (2.4) | 3 (8.1) | |
| Others | 12 (9.9) | 8 (9.5) | 4 (10.8) | |
| The sites of brain injury (N., %) | 0.150# | |||
| Supratentorial | 89 (73.6) | 65 (77.4) | 24 (64.9) | |
| Combined with infratentorial lesions | 32 (26.4) | 19 (22.6) | 13 (35.1) | |
| Comorbidity | ||||
| Hypertension | 77 | 50 | 27 | 0.156# |
| Diabetes mellitus | 39 | 26 | 13 | 0.458# |
| Coronary heart disease | 24 | 16 | 8 | 0.774# |
| Stroke | 23 | 13 | 10 | 0.136# |
| Time from the acute events, median (IQR), days | 50.00 (33.50, 93.50) | 48.50 (31.50, 99.25) | 50.50 (36.00, 87.75) | 0.665‡ |
| Length of hospital stay, days | 46.00 (36.00, 60.00) | 45 (36.00, 58.00) | 47 (38.50, 63.00) | 0.205‡ |
*Independent t-test; #χ2 test; †Fisher’s Exact Test; ‡Mann-Whitney U Test.
MSS was level 0 in 40 patients (33.1%), level 1 in 24 (19.8%), level 2 in 25 (20.7%), and level 3 in 32 (26.4%). We observed an increasing tendency toward aspiration with higher MSS levels (Table II).
Table II. —Descriptive and univariable analysis of GCS and the parameters observed by FEES in the D group and LTT group.
| Parameter N. (%) | D group (N.=84) | PT group (N.=37) | Crude OR (95% CI) | P value |
|---|---|---|---|---|
| GCS≤8 | 7 (8.3) | 12 (32.4) | 5.28 (1.87-14.87) | 0.001 |
| MSS | <0.001 | |||
| 0 | 37 (44.0) | 3 (8.1) | 1 (referent) | |
| 1 | 23 (27.4) | 1 (2.7) | 0.54 (0.05-5.47) | 0.599 |
| 2 | 21 (25.0) | 4 (10.8) | 2.35 (0.48-11.52) | 0.292 |
| 3 | 3 (3.6) | 29 (78.4) | 119.22 (22.39-634.91) | <0.001 |
| Decreased laryngeal sensation | 24 (28.6) | 24 (64.9) | 4.62 (2.02-10.53) | <0.001 |
| Food or liquid aspiration | 37 (44.0) | 35 (94.6) | 22.23 (5.02-98.50) | <0.001 |
GCS: Glasgow Coma Scale; MSS: Murry’s Secretion Scale; OR: odds ratio.
We calculated the Pearson correlation coefficient to assess the relationship between MSS level and food/liquid aspiration and found a moderate correlation between the two (R=0.677, P<0.001). Of the 84 successfully decannulated patients, 37 (44.0%) experienced food/liquid aspiration. Among these 37 patients, MSS was level 0 in six (16.2%), level 1 in nine (24.3%), level 2 in 19 (51.4%), and level 3 in three (8.1%) (Table III). Substantial agreement was observed in the interrater reliabilities of the MSS (kappa=0.905; P<0.001). Differences in MSS ratings, laryngeal sensation, and food/liquid aspiration were discussed by raters to reach a consensus.
Table III. —MSS level and aspiration condition.
| MSS | All patients enrolled | D group | ||
|---|---|---|---|---|
| Aspiration (N.=72) | Without aspiration (N.=49) | Aspiration (N.=37) | Without aspiration (N.=47) | |
| 0 | 7 (17.5%) | 33 (72.5%) | 6 (16.2%) | 31 (83.8%) |
| 1 | 10 (41.7%) | 14 (58.3%) | 9 (39.1%) | 14 (60.9) |
| 2 | 23 (92%) | 2 (8%) | 19 (90.5%) | 2 (9.5%) |
| 3 | 32 (100%) | 0 (0%) | 3 (100%) | 0 (0%) |
MSS: Murry’s Secretion Scale.
Univariate analysis revealed that the GCS scores and all three FEES parameters were significantly associated with prolonged tracheostomy: GCS≤8 (OR: 5.28; 95% CI: 1.87-14.87; P=0.001), MSS level 3 (OR: 119.22; 95% CI: 22.39-634.91, P<0.001), decreased laryngeal sensation (OR: 4.62; 95% CI: 2.02-10.53, P<0.001), and food/liquid aspiration (OR: 22.23; 95% CI: 5.02-98.50, P<0.001). Four multiple logistic regression models were used to calculate secretion-retention severity ORs with prolonged tracheostomy probability after adjusting for other factors (Table IV).
Table IV. —Multiple logistic regression of MSS related to long-term decannulation.
| MSS level (referent 0) | Adjusted OR (95% CI) | P value | |
|---|---|---|---|
| Model 1* | 1 | 0.54 (0.05-5.47) | 0.599 |
| 2 | 2.35 (0.48-11.52) | 0.292 | |
| 3 | 119.22 (22.39-634.91) | <0.001 | |
| Model 2** | 1 | 0.51 (0.05-5.68) | 0.507 |
| 2 | 3.13 (0.55-17.67) | 0.197 | |
| 3 | 154.54 (24.67-968.18) | <0.001 | |
| Model 3† | 1 | 0.41 (0.03-4.90) | 0.479 |
| 2 | 1.41 (0.16-12.22) | 0.755 | |
| 3 | 64.93 (6.62-636.82) | <0.001 | |
| Model 4‡ | 1 | 0.41 (0.03-4.97) | 0.484 |
| 2 | 1.41 (0.16-12.26) | 0.755 | |
| 3 | 65.22 (6.58-646.35) | <0.001 |
OR: odds ratio. *Model 1: unadjusted; **model 2: multiple logistic regression adjusted for GCS; †model 3: multiple logistic regression adjusted for GCS and aspiration; ‡model 4: multiple logistic regression adjusted for GCS, aspiration, and laryngeal sensation.
In model 4, after adjusting for food/liquid aspiration, laryngeal sensation, and GCS score, multiple logistic regression of secretion-retention severity yielded an MSS-level-3 adjusted OR of 65.23 (95% CI: 6.58-646.35, P<0.001), confirming the strong relationship between severe secretion retention and prolonged tracheostomy. The Hosmer-Lemeshow Goodness-Of-Fit Test was not significant (P=0.807), indicating a good model fit to the data. In the multivariable analysis, besides MSS level 3, GCS≤8 (OR: 8.25; 95% CI: 1.51-45.16; P=0.015) remained in the regression model. Finally, the MSS AUC for predicting decannulation outcome was 0.894 (95% CI: 0.819-0.969) (Figure 3).
Figure 3.
—ROC curve for MSS level 3 as a predictor of the decannulation outcome in the entire study population. ROC curve area: 0.894(95% CI 0.81-0.969).
The optimal MSS level for predicting decannulation outcome was level 3. The sensitivity, the proportion of PT group that was correctly diagnosed by MSS level 3, was 29/37=78.4%. And the specificity, the proportion of DT group that was correctly identified by MSS level 3 was 81/89=96.4% (Table V).21
Table V. —MSS with different cut-off values.
| Cutoff values | MSS | D group | LTT group | Sensitivity (%) | Specificity (%) | PPV (%) | NPV (%) | RR |
|---|---|---|---|---|---|---|---|---|
| 1 | 0 | 37 | 3 | 91.9 | 44.0 | 92.5 | 42.0 | 1.59 |
| 1-3 | 47 | 34 | ||||||
| 2 | 0-1 | 60 | 4 | 89.2 | 71.4 | 93.8 | 57.9 | 2.23 |
| 2-3 | 24 | 33 | ||||||
| 3 | 0-2 | 81 | 8 | 78.4 | 96.4 | 91.0 | 90.6 | 9.71 |
| 3 | 3 | 29 |
MSS: Murry’s Secretion Scale; PPV: positive predictive value; NPV: negative predictive value; RR: relative risk.
Discussion
We demonstrated that severe secretion penetration (MSS level 3) was independently associated with prolonged tracheostomy. Furthermore, MSS level 3 was the optimal level for predicting decannulation outcome with a high diagnostic value (sensitivity: 78.4%, specificity: 96.4%, and AUC: 0.894). These results provide evidence supporting hypopharyngeal severe secretion retention (MSS level 3) as a potentially useful predictor of prolonged tracheostomy risk in the sABI patient population. Since a high secretion-retention prevalence has been reported in patients with tracheostomy, these findings may be crucial for establishing objective and safe decannulation criteria based on fiberoptic endoscopy and implementing appropriate therapeutic strategies that promote safe and timesaving decannulation.10, 22
Decreased secretion management ability and associated airway protection deficits are the common reasons for prolonged tracheostomy and decannulation failure in sABI patients.2 Presently clinical diagnosis of secretion management ability is challenging and often delivers unreliable results. For example, the Blue Dye Test’s low diagnostic accuracy and potential safety issues hindered its clinical application; the clinical swallowing evaluation (CSE) is also unreliable at detecting secretion penetration or aspiration.23, 24 Although secretions could be directly observed by FEES, the diagnosis of secretion aspiration is difficult when a clear view of the larynx for determining secretion aspiration is limited or the volume of secretion aspirated is too small to view.22, 25 Secretion retention in hypopharyngeal is also a sign of decreased secretion management ability.7 Some studies have employed secretion retention assessment as an important step to determine decannulation readiness with low decannulation failure rate.12, 13 However, they are mainly non-comparative observational researches conducted in neurological Intensive Care Units. By means of selecting a protocol-based capping trial as the reference standard, our study verified the good diagnostic performance of severe secretion retention for decannulation outcome in a large cohort neurological rehabilitation patient population. In addition, food/liquid delivery or touching the laryngeal structure during secretion evaluation is unnecessary, remarkably minimizing the requirement for special expertise, examination time, and aspiration risk. These further supported the screening of secretion-retention in the sABI population in both acute and chronic phases, guiding the early identification of decannulation in everyday clinical care.
Food/liquid aspiration is also an important factor for decreased airway protection.15 In our study, food/liquid aspiration was associated with prolonged tracheostomy in univariate analysis; however, in multivariable analysis, it was excluded from the model as a confounder. Higher MSS levels were correlated with food/liquid aspiration, concurring with other researches’ findings, in which accumulated laryngeal-vestibule secretions were linked to food/fluid aspiration.7, 8, 26 This is partly explained by the fact that severe secretion retention and food/liquid aspiration share common swallowing pathophysiology involving the pharyngeal phase (i.e., decreased laryngeal sensation, incomplete pharyngeal constriction, and impaired upper esophageal sphincter opening).27, 28 Aspiration events could also be caused by swallowing impairment during the oral swallowing phase, such as bolus propulsion and premature spillage. However, compared with food/liquid swallowing, spontaneous swallowing that manage oropharyngeal secretions bypassed the participation of oral phase more frequently.29 Then many patients can manage their secretions or protect their airway from mild secretion penetration, they still aspirate food/liquid.7 In our study, among successfully decannulated patients, 37 (44.0%) experienced food/liquid aspiration, with MSS level 0 in 6 patients (16.2%), level 1 in 9 (24.3%), and level 2 in 19 (51.4%). In the remaining three (8.1%) MSS-level-3 patients, secretions consistently penetrated the laryngeal vestibule without spontaneous clearance due to bulbar paralysis and compromised cough reflex. However, with prompting, secretions could be ejected from the laryngeal vestibule by voluntary coughing, thus compensating for the decreased laryngeal and pharyngeal self-cleaning abilities. The application of a protocol that rules out food/liquid aspiration potentially exposes patients with adequate secretion-management ability to delayed-decannulation risk.
Although several published international surveys of decannulation practice rated consciousness level as one of the most important factors influencing patient decannulation, studies on the predictive power of reduced consciousness on decannulation outcomes are conflicting.2, 3, 5, 17 Our findings are consistent with those of Hashem et al. and Reverberi et al. in that reduced consciousness is a predictor of prolonged tracheostomy.30-32 First, coughing ability is closely related to consciousness level.33 The cough reflex can be blunted if it cannot be perceived cortically and cognitively. Moreover, patients with reduced consciousness predominantly rest in bed. The supine position often leads to somewhat reduced pulmonary function compared with standing and sitting, including reduced lung volumes and suboptimal expiratory-muscle length; thus, higher intrathoracic and expiratory pressures, which are essential for effective coughing, cannot be generated.34 Second, spontaneous-swallowing frequency is closely related to consciousness level.33 Swallowing frequency is considerably lower in the sleeping and anesthesia states than in the awake state.29 Lower swallowing frequency plays an important role in secretion retention.7, 35
Strengths and limitations of the study
The strengths of this study were objectively evaluation of secretion retention by FEES and the application of multiple logistic regression to confirm the relationship between severity of secretion retention and prolonged tracheostomy after adjusting for covariates that related to airway protection impairments. This study has certain limitations. First, its retrospective nature has inherent limitations by design; nevertheless, there are few missing data related to the patients’ outcomes or assessed predictors. This is related to the electronic nature of the records and FEES-examination completeness in our department. Further prospective studies including more confounding factors should be performed to validate our findings. Second, the study’s cross-sectional design prohibited the establishment of a causal relationship between severe secretion retention and prolonged tracheostomy. However, the fact that many patients with sABI were tracheostomized because of decreased secretion-management ability, in turn, indicated that severe secretion retention might have been a reason for prolonged tracheostomy in this patient population. Whether decreasing secretion retention increases successful-decannulation probability warrants further investigation in clinical trials. Third, our study selected a protocol-based capping trial as the reference standard for diagnosing decannulation readiness. Consensus regarding the optimal clinical practice for decannulation is lacking. However, capping trials are commonly used methods for determining whether a critically ill patient with a tracheostomy tube is ready for decannulation, both in Intensive Care Units and rehabilitation faculties.3-5, 17, 30, 31 Protocol-based capping trials have been proven to have high sensitivity and positive predictive value for successful decannulation.36 Finally, the patients in this study originated from a single center with a large stroke population; therefore, selection bias should be considered.
Conclusions
This study found that severe secretion retention (MSS level 3) is a good predictor of prolonged tracheostomy in patients with sABI with high diagnostic value. These data augment current evidence supporting the screening of secretion-retention in sABI population to determine decannulation readiness. Whether decreasing secretion retention would increase successful-decannulation probability need further investigation in clinical trials.
References
- 1.Krishnamoorthy V, Hough CL, Vavilala MS, Komisarow J, Chaikittisilpa N, Lele AV, et al. Tracheostomy After Severe Acute Brain Injury: trends and Variability in the USA. Neurocrit Care 2019;30:546–54. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=30919303&dopt=Abstract 10.1007/s12028-019-00697-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Bösel J. Tracheostomy in stroke patients. Curr Treat Options Neurol 2014;16:274. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=24357462&dopt=Abstract 10.1007/s11940-013-0274-1 [DOI] [PubMed] [Google Scholar]
- 3.Singh RK, Saran S, Baronia AK. The practice of tracheostomy decannulation-a systematic review. J Intensive Care 2017;5:38. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=28649385&dopt=Abstract 10.1186/s40560-017-0234-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Medeiros GC, Sassi FC, Lirani-Silva C, Andrade CR. Criteria for tracheostomy decannulation: literature review. CoDAS 2019;31:e20180228. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=31800881&dopt=Abstract 10.1590/2317-1782/20192018228 [DOI] [PubMed] [Google Scholar]
- 5.Stelfox HT, Crimi C, Berra L, Noto A, Schmidt U, Bigatello LM, et al. Determinants of tracheostomy decannulation: an international survey. Crit Care 2008;12:R26. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=18302759&dopt=Abstract https://doi.org/ 10.1186/cc6802 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Checklin M, Etty-Leal M, Iseli TA, Potter N, Fisher S, Chapman L. Saliva management options for difficult-to-wean people with tracheostomy following severe acquired brain injury (ABI): A review of the literature. Brain Inj 2015;29:1–10. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=25313955&dopt=Abstract 10.3109/02699052.2014.967298 [DOI] [PubMed] [Google Scholar]
- 7.Murray J, Langmore SE, Ginsberg S, Dostie A. The significance of accumulated oropharyngeal secretions and swallowing frequency in predicting aspiration. Dysphagia 1996;11:99–103. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=8721067&dopt=Abstract https://doi.org/ 10.1007/BF00417898 [DOI] [PubMed] [Google Scholar]
- 8.Ota K, Saitoh E, Baba M, Sonoda S. The secretion severity rating scale: a potentially useful tool for management of acute-phase fasting stroke patients. J Stroke Cerebrovasc Dis 2011;20:183–7. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=20621518&dopt=Abstract 10.1016/j.jstrokecerebrovasdis.2009.11.015 [DOI] [PubMed] [Google Scholar]
- 9.Muhle P, Suntrup-Krueger S, Burkardt K, Lapa S, Ogawa M, Claus I, et al. Standardized Endoscopic Swallowing Evaluation for Tracheostomy Decannulation in Critically Ill Neurologic Patients - a prospective evaluation. Neurol Res Pract 2021;3:26. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=33966636&dopt=Abstract 10.1186/s42466-021-00124-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Warnecke T, Suntrup S, Teismann IK, Hamacher C, Oelenberg S, Dziewas R. Standardized endoscopic swallowing evaluation for tracheostomy decannulation in critically ill neurologic patients. Crit Care Med 2013;41:1728–32. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=23774336&dopt=Abstract 10.1097/CCM.0b013e31828a4626 [DOI] [PubMed] [Google Scholar]
- 11.Donzelli J, Brady S, Wesling M, Craney M. Predictive value of accumulated oropharyngeal secretions for aspiration during video nasal endoscopic evaluation of the swallow. Ann Otol Rhinol Laryngol 2003;112:469–75. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=12784989&dopt=Abstract 10.1177/000348940311200515 [DOI] [PubMed] [Google Scholar]
- 12.Donzelli J, Brady S, Wesling M, Theisen M. Secretions, occlusion status, and swallowing in patients with a tracheotomy tube: a descriptive study. Ear Nose Throat J 2006;85:831–4. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=17240710&dopt=Abstract 10.1177/014556130608501216 [DOI] [PubMed] [Google Scholar]
- 13.Hales PA, Drinnan MJ, Wilson JA. The added value of fibreoptic endoscopic evaluation of swallowing in tracheostomy weaning. Clin Otolaryngol 2008;33:319–24. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=18983340&dopt=Abstract 10.1111/j.1749-4486.2008.01757.x [DOI] [PubMed] [Google Scholar]
- 14.Langmore SE. History of Fiberoptic Endoscopic Evaluation of Swallowing for Evaluation and Management of Pharyngeal Dysphagia: changes over the Years. Dysphagia 2017;32:27–38. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=28101663&dopt=Abstract 10.1007/s00455-016-9775-x [DOI] [PubMed] [Google Scholar]
- 15.Finlayson O, Kapral M, Hall R, Asllani E, Selchen D, Saposnik G, Canadian Stroke Network ; Stroke Outcome Research Canada (SORCan) Working Group. Risk factors, inpatient care, and outcomes of pneumonia after ischemic stroke. Neurology 2011;77:1338–45. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=21940613&dopt=Abstract 10.1212/WNL.0b013e31823152b1 [DOI] [PubMed] [Google Scholar]
- 16.Kaneoka A, Pisegna JM, Inokuchi H, Ueha R, Goto T, Nito T, et al. Relationship Between Laryngeal Sensory Deficits, Aspiration, and Pneumonia in Patients with Dysphagia. Dysphagia 2018;33:192–9. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=28866750&dopt=Abstract 10.1007/s00455-017-9845-8 [DOI] [PubMed] [Google Scholar]
- 17.Enrichi C, Battel I, Zanetti C, Koch I, Ventura L, Palmer K, et al. Clinical Criteria for Tracheostomy Decannulation in Subjects with Acquired Brain Injury. Respir Care 2017;62:1255–63. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=28698267&dopt=Abstract 10.4187/respcare.05470 [DOI] [PubMed] [Google Scholar]
- 18.Rodrigues LB, Nunes TA. Importance of flexible bronchoscopy in decannulation of tracheostomy patients. Rev Col Bras Cir 2015;42:75–80. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=26176672&dopt=Abstract 10.1590/0100-69912015002003 [DOI] [PubMed] [Google Scholar]
- 19.Kaneoka A, Pisegna JM, Krisciunas GP, Nito T, LaValley MP, Stepp CE, et al. Variability of the Pressure Measurements Exerted by the Tip of Laryngoscope During Laryngeal Sensory Testing: A Clinical Demonstration. Am J Speech Lang Pathol 2017;26:729–36. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=28732098&dopt=Abstract 10.1044/2017_AJSLP-16-0006 [DOI] [PubMed] [Google Scholar]
- 20.Lu SO, Yong WA, Hai-Dong LI, Jie LI, Zheng LI, Wen-Jing ZH, et al. Predictors of Tracheostomy Decannulation in Patients with Severe Acquired Brain Injury. Chinese Journal of Stroke. 2021;16:699–704. [Google Scholar]
- 21.Akobeng AK. Understanding diagnostic tests 1: sensitivity, specificity and predictive values. Acta Paediatr 2007;96:338–41. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=17407452&dopt=Abstract 10.1111/j.1651-2227.2006.00180.x [DOI] [PubMed] [Google Scholar]
- 22.Kang Y, Chun MH, Lee SJ. Evaluation of salivary aspiration in brain-injured patients with tracheostomy. Ann Rehabil Med 2013;37:96–102. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=23525725&dopt=Abstract 10.5535/arm.2013.37.1.96 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Therese M. O’Neil-Pirozzi, Deborah J Lisiecki, K Jack Momose, Jennifer J Connors, Mary P Milliner. Simultaneous modified barium swallow and blue dye tests: a determination of the accuracy of blue dye test aspiration findings. Dysphagia 2003;18:32–8. 10.1007/s00455-002-0081-4 [DOI] [PubMed] [Google Scholar]
- 24.Linden P, Kuhlemeier KV, Patterson C. The probability of correctly predicting subglottic penetration from clinical observations. Dysphagia 1993;8:170–9. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=8359036&dopt=Abstract 10.1007/BF01354535 [DOI] [PubMed] [Google Scholar]
- 25.Rodrigues B, Nóbrega AC, Sampaio M, Argolo N, Melo A. Silent saliva aspiration in Parkinson’s disease. Mov Disord 2011;26:138–41. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=21322025&dopt=Abstract 10.1002/mds.23301 [DOI] [PubMed] [Google Scholar]
- 26.Kuo CW, Allen CT, Huang CC, Lee CJ. Murray secretion scale and fiberoptic endoscopic evaluation of swallowing in predicting aspiration in dysphagic patients. Eur Arch Otorhinolaryngol 2017;274:2513–9. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=28286927&dopt=Abstract 10.1007/s00405-017-4522-y [DOI] [PubMed] [Google Scholar]
- 27.Shapira-Galitz Y, Shoffel-Havakuk H, Halperin D, Lahav Y. Association Between Laryngeal Sensation, Pre-swallow Secretions and Pharyngeal Residue on Fiberoptic Endoscopic Examination of Swallowing. Dysphagia 2019;34:548–55. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=30911836&dopt=Abstract https://doi.org/ 10.1007/s00455-019-10001-4 [DOI] [PubMed] [Google Scholar]
- 28.Yamaguchi T, Mikushi S, Ayuse T. Evaluation of swallowing function in patients with oropharyngeal secretions. Clin Exp Dent Res 2019;5:557–65. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=31687191&dopt=Abstract 10.1002/cre2.223 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Mélotte E, Maudoux A, Panda R, Kaux JF, Lagier A, Herr R, et al. Links Between Swallowing and Consciousness: A Narrative Review. Dysphagia 2022. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=35773497&dopt=Abstract 10.1007/s00455-022-10452-2 [DOI] [PubMed]
- 30.Hakiki B, Draghi F, Pancani S, Portaccio E, Grippo A, Binazzi B, et al. Decannulation After a Severe Acquired Brain Injury. Arch Phys Med Rehabil 2020;101:1906–13. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=32428445&dopt=Abstract 10.1016/j.apmr.2020.05.004 [DOI] [PubMed] [Google Scholar]
- 31.Alhashemi H, Algarni M, Al-Hakami H, Seebran N, Hussain T, Bhutto T, et al. An Interdisciplinary Approach to the Management of Individuals with Tracheostomy. Respir Care 2021;respcare.08869. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=34344718&dopt=Abstract [DOI] [PubMed]
- 32.Reverberi C, Lombardi F, Lusuardi M, Pratesi A, Di Bari M. Development of the Decannulation Prediction Tool in Patients With Dysphagia After Acquired Brain Injury. J Am Med Dir Assoc 2019;20:470–475.e1. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=30455047&dopt=Abstract 10.1016/j.jamda.2018.09.022 [DOI] [PubMed] [Google Scholar]
- 33.Troche MS, Brandimore AE, Godoy J, Hegland KW. A framework for understanding shared substrates of airway protection. J Appl Oral Sci 2014;22:251–60. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=25141195&dopt=Abstract 10.1590/1678-775720140132 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Katz S, Arish N, Rokach A, Zaltzman Y, Marcus EL. The effect of body position on pulmonary function: a systematic review. BMC Pulm Med 2018;18:159. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=30305051&dopt=Abstract https://doi.org/ 10.1186/s12890-018-0723-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Brady SL, Wesling MW, Donzelli JJ, Kaszuba S. Swallowing frequency: impact of accumulated oropharyngeal secretion levels and gustatory stimulation. Ear Nose Throat J 2016;95:E7–10. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=26930346&dopt=Abstract 10.1177/014556131609500203 [DOI] [PubMed] [Google Scholar]
- 36.Pandian V, Miller CR, Schiavi AJ, Yarmus L, Contractor A, Haut ER, et al. Utilization of a standardized tracheostomy capping and decannulation protocol to improve patient safety. Laryngoscope 2014;124:1794–800. https://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=24473939&dopt=Abstract 10.1002/lary.24625 [DOI] [PubMed] [Google Scholar]