The Penetration and Aspiration Scale (PAS) evaluates effectiveness in protecting the upper airway from aspiration using 8 degrees of dysphagia (from 1 to 8; see caption; Table 2).12 Scores are determined primarily by the depth to which material passes in the airway and by whether or not the material is expelled.
The sample size was calculated to observe a clinically significant incidence of severe swallowing impairment at 3 μg/mL TCI propofol target, which, in our experience, is the most common target in a relevant clinical setting. We arbitrarily judged that a 5% incidence of severe swallowing impairment would be clinically significant, whereas a 1% incidence would not be.
“Severe swallowing impairment” indicated DSS = 3 (moderate-severe or severe dysphagia) or PAS = 7–8. These scores imply that the test liquid passes below the vocal folds and is not expelled (i.e., it is inhaled).
To differentiate a 5% versus a 1% incidence of severe swallowing impairment with α = 0.05 and power 80%, we calculated a sample size of 76 patients, which we later increased to 80 patients. Data were analyzed using dedicated Stata 13.0 software (Copyright 1985–2013; StataCorp. LP, StataCorp, College Station, TX).
Continuous variables are reported as median (interquartile range) because none is normally distributed (Shapiro-Wilk W test for normality). Discrete variables are reported as number (percentage). Odds ratios are reported as OR (95% confidence interval).
Statistical analysis was performed with the aim of building ordinal logistic regression models with correlated data,13 using the proportional odds (or cumulative logit) model and taking DDS and PAS as the dependent variables, respectively. DDS has 4 ordered outcome categories (see note to Table 1). PAS has 9 ordered outcome categories (see note to Table 2), but after data inspection we recoded this variable to a “Reduced” Penetration-Aspiration Scale (rPAS) with only 4 categories (1–2 vs 3–4 vs 5–6 vs 7–8). This was because some scores of the original scale were not represented at the 4 μg/mL TCI target evaluation (see Table 2), precluding correct proportional odds ordinal logistic regression modeling. Next, the proportional odds assumption was not verified (employed test described later) for the 3 μg/mL TCI target evaluation of the original scale, whereas it was for every analysis performed on the reduced score.
Each of the following variables was tested separately for variable selection in univariable analysis. TCI target was considered a cluster-specific covariate with 3 levels (2, 3, and 4). Age (years) and body mass index (BMI) were considered continuous cluster-level covariates. Other cluster-level covariates were gender (M versus F), ASA physical status (I vs II–III), smoking habits (nonsmoker versus smoker), and minor glottis abnormalities (no versus yes) that were all included using reference-cell (dummy) coding with 2 categories.
Variables exhibiting P < 0.25 were considered candidates for the multivariable models, and their interactions were checked. In building the final models, P < 0.05 was considered statistically significant.
Assuming that the observations are, in fact, not correlated, multicollinearity of variables included in the final model was excluded with the variance inflation factor evaluation, and linearity in the logit of continuous variables (namely age and BMI) was assessed using the fractional polynomial method. Fitting 2-degree fractional polynomials on a set of 8 powers (−2, −1, −0.5, 0 [i.e., logarithmic transformation], 0.5, 1, 2, 3), 44 models were fitted. No significant difference was found between linear models and the other models in any predictive model for both continuous variables.
To verify the proportional odds (or parallel regression) assumption, an approximate likelihood ratio test of whether coefficients are equal across categories was performed. This test compares the fit of the proportional odds model with that of a nominal logistic regression model. It should be noted that a significant P value rejects the null hypothesis that coefficients are equal across categories, whereas a nonsignificant P value is only suggestive that we have an appropriate model for the data.
Model comparison was performed with the Akaike information criterion (AIC).14 Given 2 models, not necessarily nested, the one with the smaller AIC fits the data better than the one with the larger AIC.
The 80 patients enrolled in the study gave written, informed consent and completed the swallowing evaluation according to the study protocol.
Median age was 69 (56.5–73) years, with a median BMI of 24.6 (22.0–26.9; range 16.4–33.8): 41 (51%) patients were female, 16 (20%) had ASA physical status score I, and 52 (65%) were nonsmokers.
We did not observe major glottis abnormalities, which would have led to patient exclusion from the study. Minor glottis abnormalities were found in 8 (10%) patients: 2 bilateral vocal fold prolapses, 1 bilateral inflammatory vocal fold nodule, 1 small vocal fold polyp, 2 minor bilateral Reinke edema, 1 light vocal cord hypotonia, and 1 unilateral vocal cord adductor paresis.
At the 2 μg/mL TCI, the OAAS score was 2 (patient responds to mild prodding or shaking) in 21 (26%) patients and 1 (patient does not respond to mild prodding or shaking) in 59 (73%). The OAAS score was 1 in all patients reaching the 3 and 4 μg/mL TCI target.
All the 21 patients with OAAS score 2 had normal swallowing (DSS = 0, PAS = 1). The DSS found at the different TCI propofol targets is shown in Table 1. At the 3 μg/mL TCI propofol target, 19 (24%) patients showed severe swallowing impairment (DSS = 3).
The PAS found at the different TCI propofol targets is shown in Table 2. At the 3 μg/mL TCI propofol target, 18 (23%) patients presented severe swallowing impairment (PAS = 7–8).
Cough occurred in all patients exhibiting penetration or aspiration, with the exception of the 13 patients exhibiting aspiration with PAS = 8 (5 patients at the 3 μg/mL TCI target and 8 patients at the 4 μg/mL TCI target), in whom the test bolus did not evoke either swallowing or cough reflexes. Laryngospasm and bronchospasm were never observed.
Patients did not experience clinically significant hemodynamic or heart rhythm alterations. Desaturation occurred at each TCI target in some patients (see below) and always after completion of the corresponding TCI target swallowing test (Table 3).
At 2 μg/mL TCI, 2 patients (2.5%) exhibited desaturation. Their swallowing test showed DSS = 2 and PAS = 5, and they were not studied at higher TCI targets. Two other patients, who were scored DSS = 3 and PAS = 3, could not eject material entering the airway (but not reaching the vocal folds) during the first minute after bolus injection but were kept for further swallowing evaluation because any residual cleared thereafter before reaching the 3 μg/mL TCI target.
Seventy-eight patients were studied at the 3 μg/mL TCI target. Twelve (15%) exhibited desaturation and were not studied at the 4 μg/mL TCI target. Their DSS score was 2, whereas their PAS score was 3 in 2 patients, 4 in 6 patients, 5 in 2 patients, and 6 in 2 patients.
Sixty-six patients were studied at the 4 μg/mL TCI target. Fifty-one (77%) exhibited desaturation after the swallowing test. Eight patients had DSS = 2 and PAS = 4, 10 patients had DSS = 3 and PAS = 6, 30 patients had DSS = 3 and PAS = 7, and 3 patients had DSS = 3 and PAS = 8.
DSS Predictive Model
Univariate analysis indicated 4 possible predictive variables for DSS: age, BMI, TCI target, and OAAS (with no interaction). Although significant on the univariate analysis (OR 0.03 [0.003–0.21]; P = 0.001), OAAS did not enter the final multivariate model, which included only 3 variables: DSS was associated with increasing age (5-year OR 1.53 [1.22–1.93]; P < 0.001), BMI (OR 1.24 [1.08–1.42]; P = 0.002), and TCI target (OR 15.80 [7.76–32.20]; P < 0.001; likelihood ratio test for the model P < 0.0001; approximate likelihood ratio test of proportionality of odds across response categories P = 0.3911).a
Given the importance of OAAS, an alternative model incorporating OAAS instead of TCI target was built as follows: DSS was associated with increasing age (5-year OR 1.13 [1.02–1.24]; P = 0.014) and BMI (OR 1.08 [1.02–1.15]; P = 0.006) and with decreasing OAAS (OR 0.05 [0.006–0.36]; P = 0.003; likelihood ratio test for the model P < 0.0001; approximate likelihood ratio test of proportionality of odds across response categories P = 0.1215).
In our specific study setting, TCI targets are better predictors of DSS than OAAS (model AIC 461 vs 551, respectively), so it seems appropriate to have greater confidence in the TCI target model.
rPAS Predictive Model
Univariate analysis indicated 4 possible predictive variables for rPAS: age, BMI, TCI target, and OAAS (with no interaction). Although significant on the univariate analysis (OR 0.03 [0.004–0.23]; P = 0.001), OAAS did not enter the final multivariate model, which included only 3 variables: rPAS was associated with increasing age (5-year OR 1.09 [1.04–1.15]; P < 0.001), BMI (OR 1.23 [1.07–1.41]; P = 0.003), and TCI target (OR 15.23 [7.45–31.16]; P < 0.001; likelihood ratio test for the model P < 0.0001; approximate likelihood ratio test of proportionality of odds across response categories P = 0.1806).b
Given the importance of OAAS, an alternative model incorporating OAAS instead of TCI target was built as follows: rPAS was associated with increasing age (5-year OR 1.14 [1.04–1.26]; P = 0.007) and BMI (OR 1.09 [1.02–1.15]; P = 0.006) and decreasing OAAS (OR 0.05 [0.006–0.41]; P = 0.005; likelihood ratio test for the model P < 0.0001; approximate likelihood ratio test of proportionality of odds across response categories P = 0.2614).
In our specific study setting, TCI targets are better predictors of rPAS than OAAS (model AIC 459 vs 547, respectively), so it seems appropriate to have greater confidence in the TCI target model.
In our study, we addressed the issue of swallowing impairment during propofol sedation at clinically relevant doses and studied the risk factors for swallowing deficit in deeply sedated patients. We used 2 scales to better depict pharyngeal function: the DSS to evaluate the swallowing mechanism and the PAS to evaluate its effectiveness in protecting the upper airway from inhalation.11,12 As expected, scoring distribution on the 2 scales was similar, as were their predictive models.
A first result of our study is that severe swallowing impairment is apparent in approximately 25% of patients sedated with 3 μg/mL propofol TCI and exhibiting an OAAS score of 1 (patient does not respond to mild prodding or shaking but responds to painful stimulation), which is a common situation during clinical practice.
Previous studies showed how the swallowing reflex is impaired during nitrous oxide sedation in volunteers,15 during neuroleptanalgesia,16 and during recovery from midazolam or propofol general anesthesia.4 Propofol general anesthesia strongly alters electromyography of the muscles at the base of the tongue.6 Propofol at subhypnotic sedation caused pharyngeal dysfunction in up to 58% of healthy volunteers (more than with isoflurane or sevoflurane), by reducing the upper esophageal sphincter resting tone and its peak contraction amplitude.5
Another major result of our study is the quantification of predictive models for swallowing impairment during deep sedation. OAAS is predictive of swallowing impairment together with patient’s age and BMI, but in our series the propofol TCI target concentration performed better than OAAS in predicting swallowing impairment. TCI targets are correlated with both clinical scales and physiological indices of sedation depth, such as the OAAS and the bispectral index, respectively.17
The observed effect of age and BMI on swallowing is not surprising. It is well known that elderly patients have a high incidence of pharyngeal dysfunction,18 and it is conceivable that they could be more prone to swallowing impairment when undergoing sedation. Moreover, older age is associated with increased pharyngeal airway collapsibility during normal sleep, independent of BMI.19 McKay et al.20 showed that a higher BMI delays airway reflex recovery after prolonged sevoflurane and desflurane administration.21,22 Moreover, obesity is known to be associated with an increased frequency of sedation-related complications, regardless of the type of anesthetic agent used.23
In evaluating the clinical relevance of our results, it is conceivable that in our patients, who exhibited desaturation, subclinical hypoventilation proportional to sedation depth occurred. Both coughing and bronchial reactivity provoked by bolus water coming into contact with the vocal folds, or being aspirated, may have precipitated impending desaturation in these cases. Although we kept desaturation as a safety criterion to stop further swallowing tests, severe aspiration without cough may actually not cause desaturation when 2 mL water test boluses are used: only 3 of 13 patients with the worst PAS score exhibited desaturation, and these 3 patients were sedated at the highest TCI target (4 μg/mL). Although the desaturation episodes were not directly caused by aspiration, the swallowing impairment we observed deserves attention. In fact, although no material is normally voluntarily administered orally, aspiration of abnormal materials such as blood, vomit, or reflux can occur during sedation, and patient’s airway protection is important in preventing both acute and delayed respiratory problems, such as acute airway obstruction and aspiration pneumonia, respectively.24
Our study presents some limitations. Our results are applicable only for propofol TCI when administered with the Marsh model because different pharmacokinetic/pharmacodynamic models yield different sedation profiles. Of particular interest is that the Marsh TCI model does not consider patient age. This could have contributed to the effect of increasing age on swallowing we observed. Other TCI models, such as the Schneider model, consider age as a covariate. Nevertheless, we chose to use the widely diffused Marsh model both in our clinical practice and in our study because of its better correlation with OAAS and bispectral index.17
In setting our TCI infusion system, we entered patients’ total body weight. Although some studies support the use of total body-weight–based schemes during propofol TCI, even in obese patients,25 recent evidence suggest that adequately adjusted body weight improves the performance of propofol TCI models.26 The possible underestimation of propofol concentration in higher BMI patients could have contributed to the effect of increasing BMI on swallowing impairment in our study, although we did not observe any severely obese patient in our series, and only 5 patients exhibited moderate obesity with a BMI >30 (maximum, 33.8).
We sequentially increased TCI target from 2 to 3 and to 4 μg/mL and did not randomize the sequence of propofol TCI targets in a crossover trial. Having performed the swallowing tests only after reaching the desired TCI propofol concentration limits the possibility of carryover effect.
In performing FEES, the injected bolus should be <5 mL and kept at the minimal volume able to elicit reflex swallowing.27 Although, in awake seated patients, 1 mL boluses are frequently used, we injected 2 mL boluses according to our routine practice in evaluating patients with impaired consciousness because 1 mL boluses are often insufficient to elicit swallowing in these cases. However, using a standard 2 mL bolus, we did not score the volume of aspiration, which would be an interesting measure of the clinical relevance of the swallowing impairment. Moreover, no evaluation of spontaneous swallowing was performed.
Although widely validated and used,9,28 both DSS and PAS are subjective scales. More invasive or complex diagnostic tools, such as videofluorography or swallowing electromyography, provide more objective data about swallowing. However, their use in a large series of patients poses feasibility problems in the clinical setting. Moreover, each examination was performed by an otolaryngologist with extensive experience in swallowing disorder evaluation. With respect to this issue, it should be noted that the otolaryngologist was not blind to the ongoing TCI target.
We bypassed the oral phase of swallowing by injecting the test bolus at the base of the tongue because injecting the test liquid on the surface of the tongue in unconscious patients lying on their side would result in partial or total loss of the bolus through the mouth.27 Lateral decubitus was adopted to limit desaturation because of propofol-induced airway obstruction and unrelated to swallowing impairment, which could have limited our study of swallowing.
Our study did not address the question whether delayed respiratory complications are associated with sedation-induced swallowing impairment. We feel that this issue needs further attention in future studies.
The sedation-induced swallowing deficit we observed may be the result of impaired laryngeal motor function, laryngeal sensitive impairment, or both. We did not perform sensory testing for glottic reflex that would be helpful to isolate motor and sensory components. Muscle weakness is a well-recognized risk factor for pharyngeal dysfunction leading to aspiration.29 Although, according to our exclusion criteria, we did not enroll patients with critical illness or other potential causes of muscle weakness, we cannot exclude that coughing during a previous swallowing test yielded a fatigue effect contributing to swallowing deficit in subsequent tests.
We did not study swallowing after prolonged exposure to propofol infusion that might lead to progressively deeper levels of sedation and to more severe swallowing impairment if a milligram per kilogram per hour infusion pattern is used rather than TCI, which guarantees constant propofol serum concentration over time.30
Our data show that aspiration because of swallowing impairment may occur during propofol deep sedation at commonly used TCI targets. TCI targets are better predictors of swallowing impairment than OAAS, whereas increased age and high BMI are concomitant risk factors. Every effort should be made to avoid the presence of abnormal oropharyngeal material (e.g., blood, vomit, reflux) during sedation.
Name: Marco Gemma, MD.
Contribution: This author helped design the study, analyze the data, and prepare the manuscript.
Attestation: Marco Gemma attests to the integrity of the original data and the analysis reported in this manuscript.
Name: Laura Pasin, MD.
Contribution: This author helped conduct the study.
Attestation: Laura Pasin attests to the integrity of the original data and the analysis reported in this manuscript.
Name: Alessandro Oriani, MD
Contribution: This author helped conduct the study and prepare the manuscript.
Attestation: Alessandro Oriani attests to the integrity of the original data and the analysis reported in this manuscript, and is the archival author.
Name: Massimo Agostoni, MD.
Contribution: This author helped conduct the study.
Attestation: Massimo Agostoni attests to the integrity of the original data and the analysis reported in this manuscript.
Name: Francesca Palonta, MD.
Contribution: This author helped conduct the study.
Attestation: Francesca Palonta attests to the integrity of the original data and the analysis reported in this manuscript.
Name: Barbara Ramella, BA.
Contribution: This author helped conduct the study.
Attestation: Barbara Ramella attests to the integrity of the original data and the analysis reported in this manuscript.
Name: Mario Bussi, MD.
Contribution: This author helped conduct the study.
Attestation: Mario Bussi attests to the integrity of the original data and the analysis reported in this manuscript.
Name: Luigi Beretta, MD.
Contribution: This author helped design the study, and prepare the manuscript.
Attestation: Luigi Beretta attests to the integrity of the original data and the analysis reported in this manuscript.
This manuscript was handled by: Peter S. Glass, MBChB, FFA (S.A.).
a The interpretation of these ORs is that, keeping the other predictor variables constant, for a 1-unit increase in a predictor variable (e.g., TCI target from 2 to 3 or from 3 to 4), the odds of having a DSS score greater than a certain value versus having that value or less is proportional to the corresponding OR. As a sensitivity analysis, we build a similar predictive model considering the TCI target a categorical variable, the relevant TCI target effect being 3 vs 2 OR = 16.67 (6.49–42.88), P < 0.001; 4 vs 3 OR = 14.95 (5.77–38.73), P < 0.001(age and BMI effect unchanged).
b The interpretation of these ORs is that, keeping the other predictor variables constant, for a 1-unit increase in a predictor variable (e.g., TCI target from 2 to 3 or from 3 to 4), the odds of having a rPAS score greater than a certain value versus having that value or less is proportional to the corresponding OR. As a sensitivity analysis, we build a similar predictive model considering the TCI target a categorical variable, the relevant TCI target effect being 3 vs 2 OR = 14.95 (5.82–38.45), P < 0.001; 4 vs 3 OR = 15.53 (5.92–40.77), P < 0.001(age and BMI effect unchanged).
1. Fanti L, Agostoni M, Casati A, Guslandi M, Giollo P, Torri G, Testoni PA. Target-controlled propofol infusion during monitored anesthesia in patients undergoing ERCP. Gastrointest Endosc. 2004;60:361–6
2. Gemma M, de Vitis A, Baldoli C, Calvi MR, Blasi V, Scola E, Nobile L, Iadanza A, Scotti G, Beretta L. Functional magnetic resonance imaging (fMRI) in children sedated with propofol or midazolam. J Neurosurg Anesthesiol. 2009;21:253–8
3. Agostoni M, Fanti L, Gemma M, Pasculli N, Beretta L, Testoni PA. Adverse events during monitored anesthesia care for GI endoscopy: an 8-year experience. Gastrointest Endosc. 2011;74:266–75
4. D’Honneur G, Rimaniol JM, el Sayed A, Lambert Y, Duvaldestin P. Midazolam/propofol but not propofol alone reversibly depress the swallowing reflex. Acta Anaesthesiol Scand. 1994;38:244–7
5. Sundman E, Witt H, Sandin R, Kuylenstierna R, Bodén K, Ekberg O, Eriksson LI. Pharyngeal function and airway protection during subhypnotic concentrations of propofol, isoflurane, and sevoflurane: volunteers examined by pharyngeal videoradiography and simultaneous manometry. Anesthesiology. 2001;95:1125–32
6. Rimaniol JM, D’Honneur G, Duvaldestin P. Recovery of the swallowing reflex after propofol anesthesia. Anesth Analg. 1994;79:856–9
7. Chernik DA, Gillings D, Laine H, Hendler J, Silver JM, Davidson AB, Schwam EM, Siegel JL. Validity and reliability of the Observer’s Assessment of Alertness/Sedation Scale: study with intravenous midazolam. J Clin Psychopharmacol. 1990;10:244–51
8. Langmore SE, Schatz K, Olsen N. Fiberoptic endoscopic examination of swallowing safety: a new procedure. Dysphagia. 1988;2:216–9
9. Aviv JE, Kaplan ST, Thomson JE, Spitzer J, Diamond B, Close LG. The safety of flexible endoscopic evaluation of swallowing with sensory testing (FEESST): an analysis of 500 consecutive evaluations. Dysphagia. 2000;15:39–44
10. Langmore SE, Schatz K, Olson N. Endoscopic and videofluoroscopic evaluations of swallowing and aspiration. Ann Otol Rhinol Laryngol. 1991;100:678–81
11. Daniels SK, Ballo LA, Mahoney MC, Foundas AL. Clinical predictors of dysphagia and aspiration risk: outcome measures in acute stroke patients. Arch Phys Med Rehabil. 2000;81:1030–3
12. Rosenbek JC, Robbins JA, Roecker EB, Coyle JL, Wood JL. A penetration-aspiration scale. Dysphagia. 1996;11:93–8
13. Hosmer D, Lemeshow S Applied Logistic Regression. 20092nd ed New York, NY John Wiley & Sons, Inc.
14. Sawa T. Information criteria for discriminating among alternative regression models. Econometrica. 1978;46:1273–91
15. Nishino T, Takizawa K, Yokokawa N, Hiraga K. Depression of the swallowing reflex during sedation and/or relative analgesia produced by inhalation of 50% nitrous oxide in oxygen. Anesthesiology. 1987;67:995–8
16. Brock-Utne JG, Winning TJ, Rubin J, Kingston HG. Laryngeal incompetence during neuroleptanalgesia in combination with diazepam. Br J Anaesth. 1976;48:699–701
17. Barakat AR, Sutcliffe N, Schwab M. Effect site concentration during propofol TCI sedation: a comparison of sedation score with two pharmacokinetic models. Anaesthesia. 2007;62:661–6
18. Ekberg O, Feinberg MJ. Altered swallowing function in elderly patients without dysphagia: radiologic findings in 56 cases. AJR Am J Roentgenol. 1991;156:1181–4
19. Eikermann M, Jordan AS, Chamberlin NL, Gautam S, Wellman A, Lo YL, White DP, Malhotra A. The influence of aging on pharyngeal collapsibility during sleep. Chest. 2007;131:1702–9
20. McKay RE, Malhotra A, Cakmakkaya OS, Hall KT, McKay WR, Apfel CC. Effect of increased body mass index and anaesthetic duration on recovery of protective airway reflexes after sevoflurane vs desflurane. Br J Anaesth. 2010;104:175–82
21. Eshima RW, Maurer A, King T, Lin BK, Heavner JE, Bogetz MS, Kaye AD. A comparison of airway responses during desflurane and sevoflurane administration via a laryngeal mask airway for maintenance of anesthesia. Anesth Analg. 2003;96:701–5
22. La Colla L, Albertin A, La Colla G, Mangano A. Faster wash-out and recovery for desflurane vs sevoflurane in morbidly obese patients when no premedication is used. Br J Anaesth. 2007;99:353–8
23. Wani S, Azar R, Hovis CE, Hovis RM, Cote GA, Hall M, Waldbaum L, Kushnir V, Early D, Mullady DK, Murad F, Edmundowicz SA, Jonnalagadda SS. Obesity as a risk factor for sedation-related complications during propofol-mediated sedation for advanced endoscopic procedures. Gastrointest Endosc. 2011;74:1238–47
24. Cooper GS, Kou TD, Rex DK. Complications following colonoscopy with anesthesia assistance: a population-based analysis. JAMA Intern Med. 2013;173:551–6
25. La Colla L, Albertin A, La Colla G, Ceriani V, Lodi T, Porta A, Aldegheri G, Mangano A, Khairallah I, Fermo I. No adjustment vs. adjustment formula as input weight for propofol target-controlled infusion in morbidly obese patients. Eur J Anaesthesiol. 2009;26:362–9
26. Cortínez LI, De la Fuente N, Eleveld DJ, Oliveros A, Crovari F, Sepulveda P, Ibacache M, Solari S. Performance of propofol target-controlled infusion models in the obese: pharmacokinetic and pharmacodynamic analysis. Anesth Analg. 2014;119:302–10
27. Langmore SE. An important tool for measuring quality of life. Dysphagia. 2000;15:134–5
28. Aviv JE, Kim T, Thomson JE, Sunshine S, Kaplan S, Close LG. Fiberoptic endoscopic evaluation of swallowing with sensory testing (FEESST) in healthy controls. Dysphagia. 1998;13:87–92
29. Mirzakhani H, Williams JN, Mello J, Joseph S, Meyer MJ, Waak K, Schmidt U, Kelly E, Eikermann M. Muscle weakness predicts pharyngeal dysfunction and symptomatic aspiration in long-term ventilated patients. Anesthesiology. 2013;119:389–97
© 2016 International Anesthesia Research Society
30. Coppens M, Van Limmen JG, Schnider T, Wyler B, Bonte S, Dewaele F, Struys MM, Vereecke HE. Study of the time course of the clinical effect of propofol compared with the time course of the predicted effect-site concentration: Performance of three pharmacokinetic-dynamic models. Br J Anaesth. 2010;104:452–8