Prognosis prediction for impaired consciousness recovery in stroke patients using videofluoroscopic swallowing study: A retrospective observational study : Medicine

Journal Logo

Research Article: Observational Study

Prognosis prediction for impaired consciousness recovery in stroke patients using videofluoroscopic swallowing study: A retrospective observational study

Jang, Sung Ho MDa; Kwak, Soyoung PhDa; Lee, Min Young PhDa,*

Author Information
Medicine 102(20):p e33860, May 19, 2023. | DOI: 10.1097/MD.0000000000033860
  • Open


1. Introduction

Impaired consciousness is caused by various brain pathologies, including stroke, hypoxic-ischemic brain injury, and traumatic brain injury.[1,2] According to stroke registry data, 4% to 38% of stroke patients experience a decreased consciousness level or mental status.[3] Impaired consciousness after stroke is reported to increase long-term mortality and risk of complications, which can significantly burden society, the patient, and the patient’s family.[4,5] Prognosis prediction of impaired consciousness is clinically important for establishing therapeutic strategies, determining a rehabilitative goal and functional outcome, and estimating rehabilitative therapy duration.[6] In addition, because the natural progression of consciousness recovery varies substantially across individual patients with impaired consciousness, any evidence useful in prognosis prediction is crucial for individual patients.[7]

A prognosis of impaired consciousness depends on the brain pathology, duration from the stroke onset, and the severity of consciousness impairment.[8] Previous studies of prognosis prediction of impaired consciousness have assessed demographic factors and conducted clinical examinations, electrophysiological studies, and neuroimaging studies, including functional magnetic resonance imaging, diffusion tensor imaging, and position emission tomography (PET).[9–19] Moreover, several other studies have reported a relationship between swallowing ability and impaired consciousness prognosis based on clinical evaluations and naso-endoscopy of swallowing function.[20–23] However, these methods have not provided sufficient objective information on swallowing function because clinical evaluations involve subjective judgement of examiners, and naso-endoscopy is dependent on the examiner’s skills and does not show the whole swallowing sequence.[24]

Videofluoroscopic swallowing study (VFSS) is a standard diagnostic tool for dysphagia which is performed using lateral viewing fluoroscopic images with the patients in a sitting position.[25] It is capable of providing useful information on swallowing ability because it provides reliable and objective quantitative data on the penetration and aspiration swallowing events obtained by videofluoroscopy and quantified by using the penetration-aspiration scale (PAS).[26] In addition, VFSS can be performed relatively easily and non-invasively to detect airway invasion of food without requiring techniques of the examiners, so it is widely used for swallowing assessment.[24–27] Swallowing process can occur in the patients with impaired consciousness either as reflexive or volitional movements. With abnormal swallowing reflex and inadequate behavioral responses to swallowing-inducing stimuli, it is followed by tracheal aspiration, which means VFSS findings can reflect the conscious level of the patients with impaired consciousness. However, no study on prognosis prediction of impaired consciousness based on VFSS has been reported so far.

In this study, we hypothesized that better swallowing responses during VFSS would be associated with better recovery of consciousness, and the assessment of airway invasion and the presence of aspiration would have a predictive value for recovery from impaired consciousness. Since prognosis prediction of consciousness recovery is important for both clinical and socioeconomic aspects, we, investigated the predictive value of VFSS findings for impaired consciousness prognosis in stroke patients.

2. Methods

2.1. Subjects and clinical evaluations

Fifty-one consecutive patients (22 men, 29 women; mean age: 62.06 ± 15.41 years, range: 23–87 years) admitted to the rehabilitation department of a university hospital between 2017 and 2021 were recruited in this retrospective observational study according to the following inclusion criteria: first-ever stroke, less than 8 weeks after stroke onset, underwent VFSS examination within 8 weeks from onset, diagnosed as either vegetative state (VS) or minimally conscious state (MCS) based on a coma recovery scale-revised (CRS-R) assessment at the time of the VFSS,[28] and no premorbid illnesses affecting swallowing function. This study was conducted retrospectively, the study protocol and the exemption of informed consent were approved by the institutional review board of university hospital.

The CRS-R assessment tool was used to evaluate patients’ consciousness level at the time of their VFSS exam (initial CRS-R) and at 3 months after the VFSS exam (second CRS-R). The change in scores between the 2 CRS-R was calculated and considered as 3-month improvement of consciousness in each patient. From the same CRS-R assessments, the scores for 6 CRS-R subscales (i.e., auditory, visual, motor, oromotor, communication, and arousal function subscales) were also calculated for each patient. All these scores were assessed by a well-trained physiatrist without knowing the VFSS results.

VFSS examinations were performed at an average of 6.02 ± 1.90 weeks after stroke onset. The VFSS was performed using an X-ray flat panel detector system (FPD, Zexira®, Toshiba, Tokyo, Japan) and according to the revised Logemann protocol, in which the patient is in a sitting position, allowing the lateral view to be observed during the swallowing process.[27] All patients could maintain a sitting position in a suitable wheelchair by themselves or with some support by the examiner’s hands during VFSS. Patients in difficulty with neck control were tested while in a recliner wheelchair equipped with neck support cushion. Bonorex was used as a VFSS contrast medium and each patient was given 3 mL of liquid bonorex by spoon. If the patient was not able to open mouth voluntarily, one examiner helped mouth opening and the other put the liquid material into the mouth to induce a swallowing reflex. Then, the patient’s PAS score was assessed by 2 trained physiatrists together.

The PAS is a useful tool to describe airway invasion events during VFSS. The PAS is a multidimensional scale that assesses 3 factors: depth of airway invasion, material remaining after swallow completion, and the patient’s sensory response to aspiration.[29] The scale was originally validated in adults by Rosenbek et al,[26] and since then, its reliability and validity are well established so far.[29–31] We evaluated airway invasion by liquid material based on PAS scores, ranging from a minimum of 1 point to a maximum of 8 points. In this study, a PAS score within the 6 to 8 range indicated the patient was aspiration-positive, whereas a score in the 1 to 5 range indicated an aspiration-negative status.[26]

Demographic and clinical data of the patients are summarized in Table 1. Among fifty-one patients, forty-four had hemorrhagic stroke, and 7 had ischemic stroke (cortical lesion; 20 patients, subcortical lesion; 12 patients, brainstem lesion; 2 patients, and mixed-lesion; 17 patients). According to conscious state assessed by CRS-R at the time of the VFSS, there were 4 of VS patients, and forty-seven of MCS patients. Based on the VFSS assessment of the ability to swallow liquid material, the patients were assigned to 1 of 2 groups based on PAS score: the aspiration-positive group (PAS score ≥ 6), and the aspiration-negative group (PAS score < 6). Twenty-seven patients were assigned to the aspiration-positive group (12 men, 15 women; mean age: 60.41 ± 16.00 years), and the other twenty-four patients to the aspiration-negative group (10 men, 14 women; mean age: 63.92 ± 14.85 years). There were no significant differences between the 2 groups in their demographic and clinical characteristics (i.e., gender, age, cause, and time to VFSS after onset; P ≥ .05), including the mean initial CRS-R scores at the time of VFSS examination (aspiration-positive group: 12.67 ± 1.78; aspiration-negative group: 11.50 ± 5.32; P ≥ .05) (Table 1).

Table 1 - Demographic data of patients in the aspiration-positive and -negative groups.
Variables Total patients Aspiration-positive group Aspiration-negative group
Patients, n 51 27 24
Age, yr 62.06 ± 15.41 60.41 ± 16.00 63.92 ± 14.85
Sex, male:female 22:29 12:15 10:14
Cause, n (%)
 Hemorrhagic stroke 44 (86.3%) 25 (92.6%) 19 (79.2%)
 Ischemic stroke 7 (13.7%) 2 (7.4%) 5 (20.8%)
Time to VFSS after onset, weeks 6.02 ± 1.90 5.81 ± 2.02 6.25 ± 1.78
Consciousness at the time of VFSS
 VS/MCS 4/47 2/22 2/25
 CRS-R score 12.12 ± 4.95 12.67 ± 1.78 11.50 ± 5.32
Values are presented as the mean ± standard deviation or number (percentage).
CRS-R = coma recovery scale-revised, MCS = minimally conscious state, VFSS = videofluoroscopic swallowing study, VS = vegetative state.
*Significantly different between the 2 groups, at P < .05.

2.2. Statistical analysis

Statistical analysis was performed using SPSS 23.0 software (SPSS Inc, Chicago, IL). Comparisons of clinical factor averages and CRS-R score improvement between the 2 groups (aspiration-positive and -negative) were conducted by using independent sample t tests. Categorical variables were evaluated using the chi-squared test or Fisher’s exact test. Pearson’s correlation was calculated to examine the relationship between the liquid material PAS score and the CRS-R score change after 3 months. Similarly, correlation analysis was also conducted for each of the 6 CRS-R subscales to determine which CRS-R subscale had the strongest relationship with the PAS score. The correlation coefficient (r value) was used to indicate the relative strength and direction of the linear relationship between the 2 factors (weak: r = 0.1–0.3, moderate: r = 0.3–0.5, strong: r = >0.5).[32] The null hypothesis of no difference was rejected if the P value was less than .05. Descriptive statistics are reported as a mean and standard deviation (SD), and P values < .05 were considered statistically significant.

3. Results

Comparisons of the PAS scores for swallowing liquid material and the increases in the total and subscale CRS-R scores between the 2 aspiration groups are summarized in Table 2. The mean PAS score of the aspiration negative group (PAS score < 6) was 1.63 ± 0.92, whereas the mean PAS score of the aspiration positive group (PAS score ≥ 6) was 7.56 ± 0.51. Three months after VFSS examinations, the total CRS-R score was improved in all patients (mean increase 6.76 ± 4.85). The increase in total CRS-R score was significantly greater in the aspiration-negative group than that in the aspiration-positive group (9.63 ± 4.84 vs 4.22 ± 3.20, P < .05). Regarding the CRS-R subscale scores, the aspiration-negative group showed significantly greater score increases at 3 months after VFSS examination than those shown in the aspiration-positive group in 5 of the 6 subscales: auditory, motor, oromotor, communication, and arousal scores (P < .05).

Table 2 - Comparison of the increase in total and subscale coma recovery scale-revised scores in the aspiration-positive and -negative groups.
Aspiration-positive group (n = 27) Aspiration-negative group (n = 24) P value
PAS (liquid) 7.56 ± 0.51 1.63 ± 0.92 .000*
CRS-R after 3 months 16.89 ± 4.54 21.13 ± 2.19 .000*
Increase in total CRS-R score 4.22 ± 3.20 9.63 ± 4.84 .000*
Increase in subscale CRS-R scores
 Auditory 0.63 ± 0.74 1.88 ± 1.33 .000*
 Visual 1.15 ± 1.17 1.79 ± 1.59 .103
 Motor 1.00 ± 1.14 2.17 ± 1.44 .002*
 Oromotor 0.56 ± 0.57 1.38 ± 0.97 .001*
 Communication 0.33 ± 0.48 1.21 ± 0.66 .000*
 Arousal 0.56 ± 0.80 1.21 ± 0.72 .004*
Values are presented as the mean ± standard deviation.
CRS-R = coma recovery scale-revised, PAS = penetration-aspiration scale.
*Significantly different between the 2 groups, at P < .05.

Table 3 summarizes the correlations between liquid material PAS score and the increase in total and subscale CRS-R scores. The calculated Pearson’s correlation coefficients (r) revealed a moderate negative correlation between liquid material PAS score and the increase in total CRS-R score (r = −0.499, P < .05). Among the 6 CRS-R subscales, a strong negative correlation was observed between liquid material PAS score and the increase in communication (r = −0.563, P < .05) score while moderate negative correlations were detected for the increases in auditory (r = −0.465, P < .05), motor (r = −0.372, P < .05), oromotor (r = −0.426, P < .05), and arousal (r= −0.368, P < .05) scores (Fig. 1).

Table 3 - Correlations between liquid material penetration-aspiration scale score and the increase in coma recovery scale-revised score.
Increase in total CRS-R score Increase in subscale CRS-R scores
Auditory Visual Motor Oromotor Communication Arousal
PAS (liquid)
r −0.499** −0.465** −0.174 −0.372** −0.426** −0.563*** −0.368**
P .000* .001* .222 .007* .002* .000* .008*
CRS-R = coma recovery scale-revised, PAS = penetration-aspiration scale.
*Statistically significant at P < .05.
The correlation coefficient (r value) represents the strength of the linear relationship (0.1–0.3, weak; 0.3–0.5, moderate
**; >0.5, strong

Figure 1.:
Scatter plots showing the correlation of the penetration-aspiration scale (PAS) score with the increase of total and subscale coma recovery scale-revised (CRS-R) scores. The PAS score shows negative correlation with the increase in total CRS-R score and 5 subscale scores (auditory, motor, oromotor, communication, and arousal) (P < .05), Δ: changes in total CRS-R score and subscale CRS-R scores.

4. Discussion

In the current study, we investigated the predictive value of VFSS findings for recovery from impaired consciousness in patients with the early stage of stroke and found out that the increase in CRS-R score during 3 months was greater in the aspiration negative group and the PAS score had a negative correlation with the increase of CRS-R score.

If airway penetration occurs during the swallowing process, patients without aspiration can expel or squeeze the penetrated material back out of the larynx and protect the airway, which happens as either a reflexive movement or through conscious behavior. By contrast, patients with aspiration have an abnormal swallowing reflex and inadequate behavioral responses to swallowing-inducing stimuli such as a touching sensation associated with the material inside the oropharyngeal cavity or auditory stimulation by verbal commands of the feeder.[33] As a result, they may fail to clear the penetrated material, which is followed by tracheal aspiration of the material. Patients in our aspiration-negative group showed better stimulus-response and swallowing reflexes during VFSS and, compared to the aspiration-positive group, showed better recovery of CRS-R after 3 months. Our results suggest that the absence of aspiration during the early stage of stroke is indicative of a better prognosis of impaired consciousness.

Concerning the correlations between the PAS score and CRS-R recovery, a moderate negative correlation was observed between the liquid material PAS score and the increase in the total CRS-R score. The PAS score reflects the severity of the airway penetration and aspiration event based on an objective quantitative scale comprising 1 to 8 points; a PAS score of 1 reflects healthy swallowing with no entry of material into the airway, whereas a score of 8 is the most severe swallowing condition, which is identified clinically as a silent aspiration.[33] Therefore, it can be said that the lower the PAS score, the more-preserved the patient’s swallowing ability, indicating suitable responses to material inside the oropharyngeal cavity and less invasion of the airway, either as a result of reflexive or volitional movements.

Taken together, our results indicate that patients with a well-preserved swallowing function (i.e., PAS < 6) on VFSS showed a better prognosis for recovery from impaired consciousness. Regarding the 6 CRS-R subscales, we found that the prognosis for recovery from impaired consciousness was strongly and negatively correlated with the communication subscale score and moderately negatively correlated with the auditory, motor, oromotor, and arousal subscale scores. These subscales are deeply involved in the swallowing process of VFSS. During the test, various stimulations are given to the patients such as voice, touch sensation of the examiners, swallowing material stimuli into the mouth and the patients need to respond to them with auditory perception and oromotor behavior along with some communication function and arousal state. Among the 6 subscales, visual scale does not have significant involvement in the process of VFSS, so it might have resulted in no correlation with the PAS score. On the other hand, during the swallowing process, verbal and gestural commands are repeatedly given by the feeder, which requires communication ability to induce responses such as oral movement or double swallow behavior. Our results indicate that, among the PAS subscales, communication ability is the most important factor contributing to swallowing function responses in patients with impaired consciousness and recovery from impaired consciousness.

Several previous studies have demonstrated a relationship between swallowing ability and impaired consciousness.[20–23] However, only a few studies have reported the prognostic value of swallowing function in the recovery of impaired consciousness.[21,22] In 2004, Formisano et al,[22] using the Glascow Coma Scale and the Barthel Index at 1-year follow-up, reported that the time interval to the first safe oral feeding provided the best prediction of the final outcome in disorders of consciousness patients with severe traumatic brain injury. In 2019, Wang et al[21] reported that, in disorders of consciousness patients with both traumatic or non-traumatic brain injury, early initiation of a swallowing response to familiar stimuli, such as a spoon, was indicative of a better prognosis for consciousness state, as evaluated by CRS-R at a 6-month follow-up. However, these studies focused on clinical observations and bedside examinations, which are insufficient for assessing airway invasion of food, and those approaches are unable to detect silent aspiration. As a result, to the best of our knowledge, this is the first study to demonstrate the predictive value of VFSS results for subsequent recovery of impaired consciousness in stroke patients.

However, some limitations of this study should be considered. First, the sample size might be small (fifty-one patients) to obtain enough statistical power although it showed statistical significance, and the follow-up duration was only 3 months. Second, the study included much larger number of MCS patients than VS patients. Third, because the study was retrospective and observational, medications could not be controlled, and each patient was taking a different type and dosage of neurotropic drugs. Fourth, although there was no statistically significant difference in the initial CRS-R mean score between the 2 groups, the lower mean CRS-R score in the aspiration-positive group might have affected the outcome. Therefore, further studies should be encouraged to overcome the above limitations. Fifth, among the VFSS results, only the PAS score was used, and the other swallowing sequence was not reflected in this study. Sixth, the lesion size and location of stroke which could affect swallowing were not considered in this study.

5. Conclusion

In conclusion, we found that patients with better score of penetration-aspiration scale on videofluoroscopic swallowing study had better recovery of impaired consciousness in the early stage of stroke. Our results suggest that evaluation of swallowing function by VFSS can be helpful for prognosis prediction of impaired consciousness in stroke patients.

Author contributions

Conceptualization: Sung Ho Jang.

Formal analysis: Min Young Lee.

Funding acquisition: Sung Ho Jang.

Investigation: Min Young Lee.

Methodology: Soyoung Kwak.

Resources: Soyoung Kwak.

Supervision: Sung Ho Jang.

Validation: Soyoung Kwak.

Writing – original draft: Sung Ho Jang, Min Young Lee.

Writing – review & editing: Sung Ho Jang, Min Young Lee.


coma recovery scale-revised
minimally conscious state
penetration-aspiration scale
videofluoroscopic swallowing study
vegetative state


[1]. Zeman A. Consciousness. Brain. 2001;124:1263–89.
[2]. Eapen BC, Georgekutty J, Subbarao B, et al. Disorders of consciousness. Phys Med Rehabil Clin N Am. 2017;28:245–58.
[3]. Li J, Wang D, Tao W, et al. Early consciousness disorder in acute ischemic stroke: incidence, risk factors and outcome. BMC Neurol. 2016;16:140.
[4]. Dostović Z, Smajlović D, Dostović E, et al. Stroke and disorders of consciousness. Cardiovasc Psychiatry Neurol. 2012;2012:429108.
[5]. Rajsic S, Gothe H, Borba HH, et al. Economic burden of stroke: a systematic review on post-stroke care. Eur J Health Econ. 2019;20:107–34.
[6]. Zasler ND. Prognostic indicators in medical rehabilitation of traumatic brain injury: a commentary and review. Arch Phys Med Rehabil. 1997;78:S12–6.
[7]. Pignat JM, Mauron E, Jöhr J, et al. Outcome prediction of consciousness disorders in the acute stage based on a complementary motor behavioural tool. PLoS One. 2016;11:e0156882.
[8]. Patel S, Hirsch N. Coma. Contin Educ Anaesth Crit Care Pain. 2013;14:220–3.
[9]. Steppacher I, Fuchs P, Kaps M, et al. A tree of life? multivariate logistic outcome-prediction in disorders of consciousness. Brain Inj. 2020;34:399–406.
[10]. Choi SC, Barnes TY, Bullock R, et al. Temporal profile of outcomes in severe head injury. J Neurosurg. 1994;81:169–73.
[11]. Wijdicks EFM. Predicting the outcome of a comatose patient at the bedside. Pract Neurol. 2020;20:26–33.
[12]. Lucca LF, Lofaro D, Pignolo L, et al. Outcome prediction in disorders of consciousness: the role of coma recovery scale revised. BMC Neurol. 2019;19:68.
[13]. Edlow BL, Claassen J, Schiff ND, et al. Recovery from disorders of consciousness: mechanisms, prognosis and emerging therapies. Nat Rev Neurol. 2020;14:1–22.
[14]. Comanducci A, Boly M, Claassen J, et al. Clinical and advanced neurophysiology in the prognostic and diagnostic evaluation of disorders of consciousness: review of an IFCN-endorsed expert group. Clin Neurophysiol. 2020;131:2736–65.
[15]. Jang SH, Kwon YH. Neuroimaging characterization of recovery of impaired consciousness in patients with disorders of consciousness. Neural Regen Res. 2019;14:1202–7.
[16]. Bao W, Li X, Luo B. A novel prognostic approach to predict recovery in patients with chronic disorders of consciousness. Neurosci Bull. 2019;35:953–4.
[17]. Wu X, Zou Q, Hu J, et al. Intrinsic functional connectivity patterns predict consciousness level and recovery outcome in acquired brain injury. J Neurosci. 2015;35:12932–46.
[18]. Galanaud D, Perlbarg V, Gupta R, et al.; Neuro Imaging for Coma Emergence and Recovery Consortium. Assessment of white matter injury and outcome in severe brain trauma: a prospective multicenter cohort. Anesthesiology. 2012;117:1300–10.
[19]. Stender J, Gosseries O, Bruno MA, et al. Diagnostic precision of PET imaging and functional MRI in disorders of consciousness: a clinical validation study. Lancet. 2014;384:514–22.
[20]. Bremare A, Rapin A, Veber B, et al. Swallowing disorders in severe brain injury in the arousal phase. Dysphagia. 2016;31:511–20.
[21]. Wang J, Wang J, Hu X, et al. The initiation of swallowing can indicate the prognosis of disorders of consciousness: a self-controlled study. Front Neurol. 2019;10:1184.
[22]. Formisano R, Voogt RD, Buzzi MG, et al. Time interval of oral feeding recovery as a prognostic factor in severe traumatic brain injury. Brain Inj. 2004;18:103–9.
[23]. Mélotte E, Maudoux A, Delhalle S, et al. Swallowing in individuals with disorders of consciousness: a cohort study. Ann Phys Rehabil Med. 2020;18:101403.
[24]. Kim SB, Lee SJ, Lee KW, et al. Usefulness of early videofluoroscopic swallowing study in acute stroke patients with dysphagia. Ann Rehabil Med. 2018;42:42–51.
[25]. Chang MC, Kwak S. Videofluoroscopic swallowing study findings associated with subsequent pneumonia in patients with dysphagia due to frailty. Front Med (Lausanne). 2021;8:690968.
[26]. Rosenbek JC, Robbins JA, Roecker EB, et al. A penetration-aspiration scale. Dysphagia. 1996;11:93–8.
[27]. Palmer JB, Kuhlemeier KV, Tippett DC, et al. A protocol for the videofluorographic swallowing study. Dysphagia. 1993;8:209–14.
[28]. Giacino JT, Kalmar K, Whyte J. The JFK coma recovery scale-revised: measurement characteristics and diagnostic utility. Arch Phys Med Rehabil. 2004;85:2020–9.
[29]. Borders JC, Brates D. Use of the penetration-aspiration scale in dysphagia research: a systematic review. Dysphagia. 2020;35:583–97.
[30]. Everton LF, Benfield JK, Michou E, et al. Reliability of the penetration-aspiration scale and temporal and clearance measures in poststroke dysphagia: videofluoroscopic analysis from the swallowing treatment using electrical pharyngeal stimulation trial. J Speech Lang Hear Res. 2022;65:858–68.
[31]. Wick EH, Johnson K, Demarre K, et al. Reliability and construct validity of the penetration-aspiration scale for quantifying pediatric outcomes after interarytenoid augmentation. Otolaryngol Head Neck Surg. 2019;161:862–9.
[32]. Cohen J. Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Hillsdale, NJ: L. Erlbaum Associates; 1988.
[33]. Steele CM, Grace-Martin K. Reflections on clinical and statistical use of the penetration-aspiration scale. Dysphagia. 2017;32:601–16.

coma; consciousness; deglutition; stroke

Copyright © 2023 the Author(s). Published by Wolters Kluwer Health, Inc.