You are on page 1of 5

JOURNAL OF APPLIED BEHAVIOR ANALYSIS 2013, 46, 539–543 NUMBER 2 (SUMMER 2013)

AN EVALUATION OF TEXTURE MANIPULATIONS


TO INCREASE SWALLOWING
HEATHER KADEY, CATHLEEN C. PIAZZA, KRISTI M. RIVAS, AND JASON ZELENY
UNIVERSITY OF NEBRASKA MEDICAL CENTER’S MUNROE-MEYER INSTITUTE

The purpose of the current investigation was to identify an appropriate texture for initial treatment
of 1 child’s feeding problem. Variability in mouth clean (a product measure of swallowing) during a
texture assessment suggested that individual pureed foods differentially affected feeding behavior. A
single-food assessment identified individual problematic foods. We used the results of these
assessments to inform treatment, resulting in high levels of mouth clean across foods.
Key words: feeding disorders, mouth clean, packing, pediatric feeding disorders, texture

Children with feeding problems may not be risk for aspiration if fed a texture they cannot
able to manage age-appropriate textures effec- manage. On the other hand, decreasing texture
tively during initial treatment. For example, unnecessarily may delay progression to age-
Patel, Piazza, Layer, Coleman, and Swartzwelder typical feeding.
(2005) conducted a texture assessment with three In the current investigation, we identified a 5-
children during treatment for feeding problems. year-old girl with a feeding problem and a history
The highest texture at which each child exhibited of packing (holding food in the mouth without
competent feeding skills was different, but all swallowing) pea-sized pieces of table food. We
were competent on textures that were lower than extended the work of Patel et al. (2002, 2005) by
expected based on age. These findings can be conducting a systematic texture assessment,
contrasted with the fact that children without which suggested that mouth clean (a product
feeding problems consume a diet exclusively of measure of swallowing) was differentially affected
table food by the age of 3 years. by food texture and possibly by food type; a
Children with feeding problems may respond single-food assessment to identify whether
differentially to foods as a function of texture and individual foods differentially affected mouth
type. For example, Patel, Piazza, Santana, and clean; and a treatment assessment informed by
Volkert (2002) treated the feeding problem of a the texture and single-food assessments to
3-year-old girl who consumed purees but introduce individual foods in the appropriate
expelled higher textures. Anecdotal observation texture.
suggested that expulsions were more frequent
when meats were presented. Decreasing the
METHOD
texture of meats resulted in fewer expulsions.
Taken together, the results of Patel et al. Participant and Setting
(2002, 2005) suggest that identification of the Margaret, a 5-year-old girl, had been admitted
appropriate texture is an important step in to a pediatric feeding disorders day-treatment
treatment of children with feeding disorders. program. She received approximately 70% of her
Some children with feeding problems may be at caloric needs via gastrostomy tube. Her mother
reported that Margaret chewed chopped food for
Correspondence should be addressed to Heather Kadey, long periods and held it in her mouth. She took
Kelberman Behavior and Feeding Program, SUNY Upsta- lansoprazole for gastroesophageal reflux. Her
teMedical University 600 E. Genesee St. Suite 124,
Syracuse, New York 13202 (e-mail: kadeyh@upstate.edu). physician and speech therapist cleared her as an
doi: 10.1002/jaba.33 appropriate and safe oral feeder. Therapists

539
540 HEATHER KADEY et al.

conducted sessions in a room (4 m by 4 m) until she accepted the bite or until 30 s elapsed.
equipped with a one-way observation panel, Acceptance during the texture and single-food
sound monitoring, and session materials (e.g., assessments was defined as Margaret opening her
booster seat). Observers sat in the observation mouth to allow the therapist to deposit the bite
room and collected data on laptop computers. within 5 s of presentation. In the treatment
assessment, Margaret was allowed to engage in
General Procedure and Response Measurement the more age-typical response of self-feeding
Margaret’s mother indicated that a treatment (because acceptance was 100% during the
goal was to increase Margaret’s consumption of previous assessments). Acceptance was defined
16 foods: applesauce, asparagus, bread, broccoli as Margaret picking up the spoon and placing the
with cheese, potatoes, carrots, chicken nuggets, bite into her mouth within 5 s of bite presenta-
fruit cocktail, green beans, grilled chicken, tion. If Margaret did not accept the bite (one
mashed potatoes, pancakes, peanut butter and occurrence), the therapist removed the bowl and
jelly sandwich, peaches, pears, and strawberry presented the next bite.
yogurt. Trained therapists conducted five daily The therapist said, “show me,” 30 s after the
meal blocks, with at least 1 hr between meal bite entered Margaret’s mouth. Observers scored
initiation (e.g., 9:00 a.m., 10:30 a.m.). The mouth clean if no food larger than a pea was in
therapist presented five (texture assessment) or her mouth, unless the absence of food was the
three (single-food and treatment assessments) result of expulsion, and scored pack if food pea
bites in each session. The number of bites size or larger was in her mouth. The therapist
presented per session was decreased from five to then presented the next bite; that is, regardless if
three to allow subsequent analyses to occur more she had swallowed, the next bite was presented
efficiently. In addition, the number of sessions (i.e., we moved on). The food was removed only
was limited by medical recommendations that if Margaret was packing after the therapist
Margaret’s consumption not exceed 15 g in a conducted the 30-s check for the last bite of
given meal. This restriction resulted in some the session. If an expulsion occurred, the therapist
sessions with fewer bite presentations. The provided no differential consequence and pre-
therapist conducted two to five sequential sented the next bite when the 30-s interval
sessions within a meal, with cumulative session elapsed. Praise was delivered for acceptance and
length (the total amount of time to conduct all mouth clean.
sessions in the meal) ranging from 4 to 9 min.
Bolus was a level spoonful on a small spoon. The Texture Assessment
feeder randomly selected five (texture assess- The texture assessment was an evaluation of
ment), one (single-food assessment), or three increasingly smoother textures to identify the
(treatment assessment) foods to present in each highest texture Margaret consumed with high
session. The order of foods was randomized levels of mouth clean. Speech-therapist-recom-
within sessions. mended textures included chopped (table food
The therapist presented bites approximately hand chopped into 2-mm by 2-mm pieces); wet
every 30 s. A presentation consisted of the ground (table food blended in a food processor to
therapist (a) placing the spoon approximately produce small chunks of food in a liquid
5 cm from the midline of Margaret’s lips in the medium); and puree (table food blended in a
texture and single-food assessments, or (b) food processor until relatively smooth). Depend-
placing a bite of food on a spoon in a bowl in ing on the type of food, water (e.g., green beans)
front of Margaret in the treatment assessment. or milk (e.g., pancakes) was added to the foods to
The bite remained in its presentation position ensure that all foods were equally viscous. Recipes
INCREASING SWALLOWING 541

that stated the exact amounts of food and liquid of texture, single-food, and treatment assess-
were developed for each food to ensure consis- ments, respectively. Interobserver agreement was
tency across trials. All 16 foods were presented at calculated for mouth clean on a trial-by-trial
each texture, with the exceptions of applesauce basis. Mean agreement was 94% (range, 63% to
and strawberry yogurt (presented as is from their 100%), 99% (range, 92% to 100%), and 97%
original containers) and instant mashed potatoes (range, 82% to 100%) for the texture, single-
(prepared following the recipe on the box). food, and treatment assessments, respectively.

Single-Food Assessment
RESULTS AND DISCUSSION
The purpose of the single-food assessment was
to evaluate whether variability in levels of mouth Figure 1 (top) depicts mouth clean during the
clean in the texture assessment with pureed foods texture assessment. Mouth clean was highest with
was due to specific foods (Patel et al., 2002). Two pureed textures, but was variable (range, 20% to
to three sessions (depending on the impact of the 100%). One possible explanation for this
15-g volume cap) were conducted, with each of variability was that the probability of mouth
the 16 foods presented in a random order. The clean differed across foods. Although pureed food
results of the single-food assessment suggested is relatively smooth, specific foods will leave a
that mouth clean was less than 80% for peaches, shell residue (e.g., peas) or will be gritty (e.g.,
peanut butter and jelly sandwich, potatoes, chicken), which may increase the difficulty of
chicken nuggets, green beans, asparagus, grilled swallowing. This question was evaluated during
chicken, and broccoli and cheese. We included the single-food assessment.
these eight foods and bread (due to therapist Figure 1 (middle) depicts mouth clean during
error) in the treatment assessment. the single-food assessment. Each bar represents
mouth clean across all sessions for the individual
Treatment Assessment food (e.g., mean of Session 1 pears plus mean of
We evaluated treatment using a BABAB(C) Session 2 pears plus mean of Session 3 pears
design (B ¼ Magic Bullet texture; A ¼ pureed divided by 3). Mouth clean was below 80% for
texture). The purpose of this assessment was to eight pureed foods. Figure 1 (bottom) depicts
determine whether presentation of foods at a mouth clean during the treatment assessment.
smoother texture than puree would result in an Mouth clean was higher in the Magic Bullet
increase in mouth clean. The eight pureed foods texture phases than in the pureed texture phases.
that were associated with low levels of mouth clean Mouth clean was close to 100% during
in the single-food assessment were prepared using presentation of the nine Magic Bullet and seven
the same recipes and were blended using a Magic pureed foods.
Bullet. The Magic Bullet is a type of food processor Results of the current study replicated those of
that has the ability to grind particles of food and Patel et al. (2005), in that the texture at which
food shells more effectively than a standard food Margaret competently ate during the texture
chopper and creates a smoother texture than puree. assessment was lower than expected based on her
After the single-food assessment, the therapist age. Results also are similar to those of Patel et al.
presented all 16 foods (C), nine at the Magic Bullet (2002), in that individual foods differentially
texture and seven at the pureed texture. affected eating. Whereas the child in Patel et al.
(2002) was more likely to expel a specific food
Interobserver Agreement group, Margaret was more likely to pack specific
A second observer simultaneously but inde- individual pureed foods (e.g., chicken, broccoli).
pendently collected data for 39%, 65%, and 47% Further, we extended Patel et al. (2002) by
542 HEATHER KADEY et al.

Wet
Chopped Ground Puree

100
90
Percentage of Mouth Clean

80
70
60
50
40
30
20
10
0
0 10 20 30
Sessions

100
Mean Percentage of Mouth Clean

90
80
70
60
50
40
30
20
10
0

Puree Puree Nine Magic Bullet +


Magic Bullet Magic Magic Bullet Seven Pureed Foods
Bullet
100
90
Percentage of Mouth Clean

80
70
60
50
40
30
20
10
0
0 10 20 30 40 50 60 70 80
Sessions

Figure 1. Percentage of mouth clean for the texture assessment (top), single-food assessment (middle), and treatment
assessment (bottom).
INCREASING SWALLOWING 543

conducting a single-food assessment to evaluate (e.g., in the texture assessment in the current
systematically how food types affect eating, as study). The disadvantage of this method is that
opposed to relying on anecdotal information. not all children exhibit feeding behavior that is
Recall that in the current study, bites were differentially affected by individual foods, and
presented every 30 s, independent of whether single-food assessments are time intensive.
Margaret swallowed. We moved on so that Another alternative would be to require the
packing did not produce avoidance of bites. The child to swallow before presenting subsequent
disadvantage of moving on is that there may be bites. It is unknown if waiting for the child to
carry over from bite to bite. For example, assume swallow before presenting the next bite would
that the child always packs chicken. If presenta- function as reinforcement for packing, and
tion order is chicken then pears, packing is likely this should be the subject of future investiga-
to be 100% for the two bites because the presence tions. Of note is that we used the move-on
of packed chicken might increase the difficulty of procedure only when personnel (e.g., speech
swallowing pears; the child would have to swallow therapist) were available to determine if the
both chicken and pears. These data suggest that child has a competent swallow, per-session bite
both foods were problematic. By contrast, if the presentations were limited, and bolus sizes were
presentation order was pears then chicken, and small.
the child swallowed pears, then packing for the Ultimately, the goal for any child with a
two bites would be 50%, suggesting that chicken, feeding disorder is to progress them to age-typical
but not pears, was problematic. However, it eating. Although the procedures described in this
would not be possible or efficient to analyze how study did not increase Margaret’s consumption of
particular presentation sequences (e.g., asparagus, age-appropriate textures, it provided an avenue
potatoes, peaches) affect mouth clean because for her to meet her nutritional needs as well as a
food presentation order was randomized both starting point for increasing texture.
across and within sessions and not all possible
sequences were presented.
The single-food assessment allowed us to REFERENCES
evaluate levels of mouth clean for each food Patel, M. R., Piazza, C. C., Layer, S. A., Coleman, R., &
individually and systematically. After we identi- Swartzwelder, D. M. (2005). A systematic evaluation of
fied foods that were associated with low levels of food textures to decrease packing and increase oral
intake in children with pediatric feeding disorders.
mouth clean, we manipulated the texture of Journal of Applied Behavior Analysis, 38, 89–100. doi:
problematic foods. Decreasing the texture of 10.1901/jaba.2005.161-02
problematic foods resulted in increased levels of Patel, M. R., Piazza, C. C., Santana, C. M., & Volkert, V. M.
(2002). An evaluation of food type and texture in the
mouth clean for those foods. We then were able treatment of a feeding problem. Journal of Applied
to present all 16 foods in textures (i.e., either Behavior Analysis, 35, 183–186. doi: 10.1901/
Magic Bullet or puree) that resulted in high levels jaba.2002.35-183
of mouth clean across foods.
Received June 1, 2012
An alternative method would be to present Final acceptance November 13, 2012
foods individually during initial treatment Action Editor, Joel Ringdahl

You might also like