Effects of Puree Type and Color on Ratings of Pharyngeal Residue, Penetration, and Aspiration during FEES: A Prospective Study of 37 Dysphagic Outpatient Adults

Flexible endoscopic evaluation of swallowing (FEES) is an instrumental procedure used to assess pharyngeal residue, penetration, and aspiration in people with known or suspected swallowing impairments. Like videofluoroscopic swallow studies, the swallowing protocols used during FEES typically involve the presentation of a variety of consistencies, including but not limited to thin liquids, pureed foods, and regular solid foods [1, 2]. Each of these consistencies can assist with identifying different types of swallowing impairments, and thus, are each important to test as part of routine clinical care [3‒5]. Furthermore, rating scales such as the Modified Barium Swallow Study Impairment Profile (MBSImP), the Dynamic Imaging Grade of Swallowing Toxicity (DIGEST), and the DIGEST-FEES incorporate thin liquids, purees, and regular chewable solids as part of their standardized protocol to facilitate ratings of a person’s overall level of swallowing impairment [2, 6‒9].

For FEES, the type and color of liquids that are used has been found to affect ratings of pharyngeal residue, penetration, and aspiration. Specifically, ratings for pharyngeal residue, penetration, and aspiration tend to be more severe for: barium and milk compared to (non-coating) water [3, 10‒12]; for liquids with a coating effect compared to liquids with no coating effect [12, 13]; for liquids that are opaque compared to liquids that are clear [10, 12, 13]; and for liquids that are colored blue or green compared to liquids with no added food coloring [12, 14].

While it has been well established that ratings of pharyngeal residue, penetration, and aspiration can change as an effect of liquid type and color during FEES, less is known if similar effects are seen with pureed foods. This is important to determine since a variety of puree types and colors are used during FEES. Some puree types used during FEES include applesauce and pudding [1, 2, 9, 15]. During FEES, purees can also be presented naturally (i.e., without any food coloring added by the evaluating clinician) or with food coloring added by the evaluating clinician [2, 9, 13]. To our knowledge, however, no research has examined if and how ratings of pharyngeal residue, penetration, and aspiration change during FEES as an effect of puree type and color. Determining if puree type and color affects ratings of pharyngeal residue, penetration, and aspiration is important for determining if there is a need to standardize the type and color of purees used during FEES.

Given the above, the aims of this study were to assess the effects of pureed food type (applesauce vs. pudding) and color (natural, blue, green) on ratings of pharyngeal residue, penetration, and aspiration during FEES. Based on similar research with thin liquids, we hypothesized that differences in ratings of pharyngeal residue, penetration, and aspiration would be present as an effect of both puree type and color.

This was a prospective, quality improvement, cross-sectional study. Patients were recruited consecutively within the workflow of an outpatient otolaryngology dysphagia clinic located at a tertiary academic medical center between May 2023 and August 2023. Inclusion criteria for study enrollment were patients who: (1) received a FEES as part of routine clinical care for assessment of dysphagia; and (2) ≥21 years of age. Exclusion criteria included an incomplete FEES for the puree portion of the testing protocol.

All patients were blinded to the aims of this study. Patient demographics, medical diagnosis, and level of swallow impairment were recorded for each patient. The Dynamic Imaging Grade of Swallowing Toxicity for FEES (DIGEST-FEES) [9, 16, 17] was used to characterize level of impairments in swallowing safety, efficiency, and overall pharyngeal swallow function using the following standardized bolus trials: 5 mL thin liquid, 10 mL thin liquid, 20 mL thin liquid, self-selected volume of thin liquid (two trials), 5 mL puree (two trials), and selected volume of cracker (two trials).

FEES was used to record pharyngeal residue, penetration, and aspiration. All FEES were performed by speech-language pathologists with greater than 10 years of clinical experience and dysphagia expertise. FEES equipment consisted of a 2.4 mm diameter flexible distal chip laryngoscope (VNL8-J10, PENTAX Medical), a video processor with an integrated xenon light source (DEFINA EPK-300, PENTAX Medical), and an LCD display. Approximately 0.1 mL of a liquid mixture solution of lidocaine and neo-synephrine was aerosolized to each naris approximately 3 min prior to the start of the FEES [18].

Participants were seated in an upright position during the FEES and were instructed to avoid talking during the exam unless otherwise instructed or unless needing to express pain or discomfort. The flexible laryngoscope was passed transnasally with the distal end of the endoscope positioned between the uvula and posterior pharyngeal wall (i.e., in the “high” position). The laryngoscope was advanced into the oropharynx and laryngeal vestibule after the swallow as needed for closer inspection of post-swallow residue patterns. A standardized protocol was planned for each patient which included thin liquid (n = 7), puree (n = 4), and regular solid (n = 2) trials, which deviations to protocol as needed based on their clinical needs. Only the puree trials of each FEES were included for analysis in this study.

Clear water was presented prior to the start of puree trials to reduce the presence of pre-existing thin liquid residue. As needed, clear water was also used to wash away residue between puree trials until <5% of the valleculae and piriforms were filled with oropharyngeal and hypopharyngeal residue. Each patient was presented with four trials of puree: (1) natural applesauce (Mott’s^® no sugar added applesauce); (2) colored (blue or green) applesauce; (3) natural vanilla pudding (Snack Pack^®); and (4) colored (blue or green) vanilla pudding. The order of which puree type (applesauce or pudding) was presented first was randomized between patients. The order of which color condition (natural vs. colored) was presented first was also randomized within each set of applesauce and pudding trials.

The food color (blue or green) that was added for one puree type was set to be different from the food color added to the second puree type. For example, a patient was presented with blue pudding and natural pudding trials would also be presented with green applesauce and natural applesauce trials, but never blue applesauce and natural applesauce trials. This was done to optimize the ability to differentiate residue patterns across puree trials.

The testing of puree consistencies was completed according to the International Dysphagia Diet Standardisation Initiative (IDDSI), with each puree condition measured to be IDDSI level 4. The colored purees were prepared by mixing four drops of either blue or green food coloring (Chef-O-Van) to a 4.0-ounce cup of pudding or a 4.0-ounce cup of applesauce. Separate 10 mL syringes were used to extract 5 mL of the room temperature puree from each puree cup. The 5 mL puree was injected onto a spoon using the syringe, which was then given to the patient for oral consumption. Trials were self-administered, non-held, and non-cued. Patients were instructed to be complete the puree trial within a single swallow. The instruction for single swallows was given to reduce the frequency of piecemeal deglutition [13] in an attempt to ensure the 5 mL of puree that was dispensed in the mouth was the same amount that was swallowed. Instructions for single swallows were given before the start of each trial but were not given once self-feeding began so as to avoid any effects of verbal cueing or bolus holding [19‒22]. Video clips were stored digitally for subsequent post-assessment analysis.

The Visual Analysis of Swallowing Efficiency and Safety (VASES) [23, 24] was used to facilitate standardized ratings of oropharyngeal residue, hypopharyngeal residue, and the Penetration-Aspiration Scale (PAS). Residue ratings are visual-perceptual in nature and range from 0% to ≥100%, with higher ratings indicating that a greater amount of residue is filling the oropharynx and hypopharynx. For this study, oropharyngeal residue ratings were made by estimating how filled the valleculae would be with all residue observed within the boundary of the oropharynx and epiglottis. Similarly, hypopharyngeal residue ratings were made by estimating how filled the piriforms would be with all residue observed within the boundary of the hypopharynx. The PAS is an 8-point scale used to describe the depth and reaction to penetration and aspiration, with higher scores suggesting worse swallowing safety and airway protection [25].

FEES videos were blindly analyzed using a duplicate rating method by pairs of independent raters. These raters were graduate interns in speech-language pathology who were trained by the first author to interpret FEES using VASES. The graduate interns were blinded to the aims of this study and were blinded to each other’s ratings. Paired PAS ratings agreed if matching exactly. Paired residue ratings agreed if both ratings indicated that residue was either absent (0%) or present (>0%), and if present, amount differed by ≤10%. Paired residue ratings that agreed were averaged together and used for data analysis. Paired PAS and residue ratings that disagreed were replaced by ratings made by the first author, who was blinded to the original paired ratings. The first author is a speech-language pathologist with board specialty certification in swallowing and over 10 years of experience performing and interpreting FEES. As one of the original developers of VASES, they also frequently train students and practicing clinicians on the use of VASES for FEES interpretation.

Statistical analyses were conducted using R version 4.2.3 [26]. An alpha level of p < 0.05 was used to determine the level of statistical significance for each predictor variable. Unadjusted (raw) p values are outlined in the text and tables, with a “*” in the tables denoting statistically significant findings after adjusting for multiple comparisons.

Oropharyngeal residue, hypopharyngeal residue, and PAS were treated as the outcome variables. Puree type and color were treated as predictor variables for separate statistical models. Furthermore, separate models were run for each puree type when assessing the effects of puree color on swallowing outcomes. Number of swallows was treated as a covariate for all with oropharyngeal and hypopharyngeal residue models. Number of swallows included all swallows completed within the “during the swallow” temporal phase (as defined by VASES), up to the point when the residue and PAS ratings were made. Patient was treated as a random intercept effect for all statistical models.

Ordinal cumulative link multilevel models were used to examine the effects of puree type and color on PAS. Multilevel beta regressions were used to examine the effect of puree type and color on oropharyngeal and hypopharyngeal residue ratings. To perform the multilevel beta regressions, residue ratings of 0% were transformed to 0.001%. This occurred for 8 (5%) of the oropharyngeal residue ratings and 19 (12%) of the hypopharyngeal residue ratings. Residue ratings of ≥100% were transformed to 99.999%. This was not needed for any of the oropharyngeal or hypopharyngeal residue ratings.

Fifty percent (n = 76) of the final dataset was repeated for analysis by a separate rater to examine inter-rater reliability. Descriptive and inferential statistics were used for inter-rater reliability assessment. Intraclass correlation coefficients (ICC) were used for all continuous measures, including oropharyngeal and hypopharyngeal residue ratings. ICC estimates and their 95% confidence intervals were calculated based on a single rating (k = 1), absolute-agreement, 2-way mixed-effects model. Kappas were used to examine reliability for all categorical measures, including the number of swallows and PAS. Cohen’s Kappa (κ) was used for number of swallows, whereas quadratic weighted kappa (κ_w) was used for PAS.

Thirty-seven dysphagic patients were included in this study (Table 1), resulting in an analysis of 147 puree swallows. 32.4% of patients presented with no swallow impairment (DIGEST 0), 48.6% presented with mild swallow impairment (DIGEST 1), 13.5% present with moderate swallow impairment (DIGEST 2), and 2.7% presented with severe swallow impairment (DIGEST 3), and 2.7% presented with profound swallow impairment (DIGEST 4). Reliability results are outlined in Table 2. Inter-reliability coefficients were 0.83 (oropharynx), 0.50 (hypopharynx), 0.34 (PAS), and 0.60 (number of swallows). Descriptively, the median difference in residue reliability was 2.5% for the oropharynx, and 0.0% for the hypopharynx. Additionally, 89.4% of PAS reliability ratings were in exact agreement and 78.9% of number of swallows reliability ratings were in exact agreement.

There was a statistically significant difference in oropharyngeal residue ratings as an effect of puree type (p = 0.015). Specifically, the odds of a higher oropharyngeal residue rating were 1.38 times greater for pudding compared to applesauce (OR = 1.38; 95% CI: 1.07–1.78). However, there were no statistically significant differences between pudding and applesauce for hypopharyngeal residue ratings (p = 0.989; OR = 1.00; 95% CI: 0.76–1.31) or PAS ratings (p = 0.819; OR = 1.18; 95% CI: 0.28–4.95).

There was a statistically significant difference in oropharyngeal residue ratings for applesauce as an effect of color (p = 0.011). Specifically, the odds of a higher oropharyngeal residue rating were 1.62 times greater for blue-colored applesauce bolus compared to a natural applesauce (OR = 1.62; 95% CI: 1.13–2.32; Fig. 1). However, there was no statistically significant difference in oropharyngeal residue ratings when comparing natural applesauce to green-colored applesauce (p = 0.278; OR = 1.27; 95% CI: 0.82–1.96).

A statistically significant difference was also present in hypopharyngeal residue ratings for applesauce as an effect of color (p = 0.0001). Specifically, the odds of a higher hypopharyngeal residue rating was 3.33 times greater for blue applesauce compared to natural applesauce (OR = 1.62; 95% CI: 1.13–2.32; Fig. 2), and 2.69 times greater for green applesauce compared to natural applesauce (p = 0.0005; OR = 2.69; 95% CI: 1.56–4.65; Fig. 2).

There were no statistically significant differences in oropharyngeal residue ratings for pudding as an effect of color (p > 0.05). Specifically, there was no statistically significant differences in oropharyngeal residue ratings when comparing natural pudding to blue-colored pudding (p = 0.595; OR = 1.09; 95% CI: 0.79–1.50), nor when comparing natural pudding to green-colored pudding (p = 0.246; OR = 1.17; 95% CI: 0.89–1.55).

However, there was a statistically significant difference in hypopharyngeal residue ratings for pudding as an effect of color. Specifically, the odds of a higher hypopharyngeal residue rating were 2.35 times greater for green pudding compared to natural pudding (p = 0.0002; OR = 2.35; 95% CI: 1.52–3.65; Fig. 2). No statistically significant differences in hypopharyngeal residue ratings were present when comparing natural pudding to blue-colored pudding (p = 0.096; OR = 1.55; 95% CI: 0.92–2.60).

Only 6% of the swallowing trials had a Penetration-Aspiration Scale (PAS) >1. Due to the limited variability in PAS ratings, statistical models assessing the effects of puree color on PAS could not be run separately for applesauce and pudding. Therefore, only one statistical model was used which incorporated both applesauce and pudding trials. PAS was not significantly different for blue- (p = 0.162; OR = 4.27; 95% CI: 0.56–32.71) or green-dyed (p = 0.067; OR = 6.47; 95% CI: 0.88–47.52) purees compared to natural purees (Fig. 3).

This study contributes to an evolving understanding of the effects of puree type and color on ratings of functional swallowing outcomes during FEES. Our findings demonstrate that puree type and color appear to influence ratings of pharyngeal residue. These results are important for informing clinical and research practice patterns associated with FEES protocols and pureed foods. Within the clinical practice patterns of the current authors, the findings from this quality improvement study have led to the standardization of both the type and color of purees used during FEES. Specifically, the authors now routinely use blue-colored pudding.

In this study, pudding was associated with higher oropharyngeal residue ratings, albeit with a relatively small effect size, when compared to applesauce. This finding could be due to several factors. One possible explanation is that pudding may be more opaque than applesauce – a visual feature that has been associated with higher residue ratings for liquids [12]. Another potential explanation is that, although both applesauce and pudding are puree consistencies, they likely have differing rheological properties that influence bolus clearance. Future research is needed to confirm these underlying mechanisms. Despite this, the finding that pudding may lead to higher oropharyngeal residue ratings underscores the need for careful consideration of the type of pureed foods used during FEES. Given that differences in pharyngeal residue ratings were present between applesauce and pudding, clinicians should determine which type of pureed food is most appropriate for the standardized protocols used within their practice patterns.

Applesauce typically has an bumpy and irregular texture which may vary considerably within and between manufacturing companies as an effect of multiple variables including food content, cooking conditions, and storage [27‒30]. The irregular texture of applesauce may more accurately represent the rheological properties of puree and regular solid foods present after oral processing. If true, then using applesauce may increase the ecological validity of puree swallows during FEES. However, the texture of applesauce can vary between and within manufacturing brands [27]. This variability in food texture could negatively affect test-retest reliability of FEES outcomes within and across patients. Pudding, conversely, may have a more smooth and consistent texture [31]. Importantly, the smooth texture of pudding may also more closely resemble the specific texture of standardized barium products (Varibar^® Pudding) which are commonly used to test puree consistencies during modified barium swallow studies. If true, then using pudding (as opposed to applesauce) during FEES could enhance the generalizability in testing methods between FEES and modified barium swallow studies for puree swallow trials.

Findings from the present study suggest that adding food coloring to puree foods can lead to higher pharyngeal residue ratings compared to their natural counterparts. These data support our original hypothesis that adding food coloring might alter ratings of swallowing outcomes during FEES. This observation is also similar to what has been previously reported with thin liquids and added food coloring during FEES [12‒14]. The reasons why color influenced residue ratings for pudding and applesauce purees are not fully understood. One possible explanation is that the addition of blue or green colorants may enhance the visual contrast between the puree, secretions, and the surrounding anatomical structures. However, further research is needed to confirm this hypothesis. Nonetheless, give that food color can alter pharyngeal residue ratings of puree boluses during FEES, there is need to standardize FEES protocols with respect to the use of colorants with pureed foods to ensure consistency and reliability of residue ratings in clinical and research settings.

This study is not without limitations. The small-to-medium effect sizes observed indicate that, while statistically significant, the clinical relevance of these findings might be limited. Future research should identify if the differences in pharyngeal residue ratings that were present as an effect of puree type and color are clinically meaningful and influence clinical recommendations and management planning. Second, this study was not designed to compare blue and green food coloring to each other, but instead, to only compare colored purees to noncolored purees. Therefore, future research should expand on the present study design to allow for direct comparisons between blue and green food coloring with purees. This will be important to ultimately determine if the type of color added to purees might also need to be standardized. Third, more than half of the patients in this study had either no impairment, or only mild impairment, in swallowing safety and efficiency. Including people with more severe impairments in swallowing safety and efficiency may have changed the results of this study (e.g., puree type and color affecting PAS ratings). Therefore, future research should work to include people with moderate to profound impairments in swallowing safety and efficiency. Lastly, future research should expand on the present study by including a broader range of puree types, different types of coloring agents, and more people with greater impairments in swallowing efficiency and swallowing safety.

In conclusion, this study provides valuable insights into how puree type and color influence pharyngeal residue during FEES. While the impact on penetration-aspiration was not statistically significant, differences in pharyngeal residue ratings were observed. These findings support the need to standardize the types and colors of pureed foods used during FEES to improve consistency between and within patients. Further research is needed to explore the broader implications of these findings and to refine FEES protocols for more accurate and reliable dysphagia assessments.

The authors would like to thank the patients who were involved in this quality improvement study. The authors would also like to thank Crystal Diaz and Theresa Lee who assisted with data collection and FEES analysis.

This study was approved by the Weill Cornell Medicine Institutional Review Board at Weill Cornell Medical College of Cornell University, approval No. 23-11026705 and was conducted in accordance with the amended Declaration of Helsinki. Written and verbal consent from the patients were not required by the IRB for this quality improvement study.

The authors have no conflicts of interest to declare.

This study was not supported by any sponsor or funder.

All authors met the minimum criteria for authorship status, as proposed by the International Committee of Medical Journal Editors. Authorship contributions are outlined below using the following CRediT roles (https://credit.niso.org/): conceptualization, data curation, formal analysis, funding acquisition, investigation, methodology, project administration, resources, software, supervision, validation, visualization, writing – original draft, and writing – reviewing and editing. James A. Curtis, PhD, CCC-SLP: conceptualization, data curation, formal analysis, investigation, methodology, project administration, resources, software, visualization, and writing – reviewing and editing; Anaïs Rameau, MD, MPhil: resources, validation, and writing – reviewing and editing, Valentina Mocchetti: conceptualization, investigation, methodology, project administration, and writing – reviewing and editing.

All data and R code associated with this study are openly available in the Open Science Framework repository at https://osf.io/v7x58/. Further inquiries can be directed to the corresponding author.