Abstract
Aerophagia is characterized by excessive air swallowing and can have serious negative effects on one’s health. We present the assessment and treatment of a 16-year-old girl, Khloe, with developmental disabilities and aerophagia. The initial assessment indicated that aerophagia occurred primarily to access attention in a divided attention context; however, our function-based treatment did not result in a clinically significant reduction in problem behavior. We then conducted a second assessment that indicated that Khloe’s aerophagia indeed occurred primarily in a divided attention context, but that it persisted independent of social consequences. We concluded that the divided attention context served as a motivating variable for aerophagia. Our second treatment consisted of differential reinforcement of other behavior, noncontingent access to competing stimuli, and graduated exposure to contextual variables (i.e., people and divided attention) that occasioned aerophagia. The treatment was successful in reducing rates of aerophagia. We discuss implications for assessment and treatment, as well as recommendations for clinicians and students.
1. Theoretical and Research Basis for Treatment
Some behaviors are not harmful in and of themselves, but they can be harmful when occurring at unusually high rates. One such example is swallowing air—it happens for everyone when drinking and eating and, in moderation, it is safe and helpful for the person (e.g., helps with digestion). When someone swallows air excessively, it can have serious negative effects on one’s health. Aerophagia is a gastrointestinal disorder that consists of excessive air swallowing to the point that it causes abdominal distension. The most notable symptom of aerophagia is repetitive and prolonged belching, which can occur at a rate of 1200 per hr (as compared to 8 to 10 per hr in the general population; Chitkara et al., 2006).
Untreated aerophagia results in serious medical complications, such as arrhythmia, abdominal pain, dizziness, nausea, vomiting, anorexia, constipation, and weight loss (Barrett et al., 1987). Approximately 8.8% of adults with developmental disabilities are also diagnosed with aerophagia, and even fewer typically developing individuals (i.e., 1.3–7%; Murphy et al., 2017). Nonetheless, developing effective assessment and treatment approaches for aerophagia is critically important due to the severe health complications caused by the behavior.
Medical interventions for aerophagia entail dietary recommendations, such as restricting the intake of carbonated drinks. Since the literature suggests that medical interventions oftentimes have little effect on aerophagia, behavioral interventions are recommended (Murphy et al., 2017). Such interventions have been demonstrated to be effective across a wide range of problematic behaviors in individuals with developmental disabilities, including self-injury, aggression, and disruption (Beavers et al., 2013). Typically, an assessment is conducted to determine the environmental variables that evoke and maintain the problematic behavior. The assessment results inform the development of a function-based, individualized treatment. This is important because the way each individual interacts with their environment is unique; for example, one might demonstrate aggression to escape from demands, while another person might demonstrate aggression to get attention from others. The treatment for these two individuals will be completely different, even though the form of the behavior might be the same. As such, identifying the function of problematic behavior and developing a treatment to match that function is critical.
Minimal extant research describes the process of assessment, treatment, and evaluation of the effectiveness of treatment for aerophagia (Cigrang et al., 2006; Flaisher, 1995; Garcia et al., 2001; Holburn & Dougher, 1985; Murphy et al., 2017). Differential outcomes are reported across studies, with no discernable characteristic of one specific treatment to account for an effective reduction in aerophagia, when a reduction occurred. One potential explanation for these differential outcomes is that, except for Garcia et al. (2001), none of these studies conducted an assessment (in this case, called functional analysis), followed by the development of a function-based treatment. Instead, the procedures employed in these studies were aligned with the hypothesis that aerophagia is automatically maintained, in that it occurs independent of socially mediated consequences (i.e., it is a form of self-stimulation). However, numerous studies show that the same form of behavior can be maintained by different environmental variables across individuals (for a review, see Beavers et al., 2013). For example, pica has long been conceived as being automatically maintained. However, Ness et al. (2020) determined that pica was sensitive to contingent attention in a 3-year-old neurotypical girl. Based on these assessment results, the authors developed a treatment that effectively suppressed pica and increased adaptive behavior that served the same function (i.e., asking for attention). It is, thus, important not to assume, but rather identify these maintaining environmental variables. This, in turn, allows us to effectively manipulate these environmental variables to suppress the frequency of aerophagia during treatment.
The purpose of this study is (a) to conduct a functional analysis to identify the function of aerophagia in a 16-year-old female diagnosed with a developmental disability, and (b) to develop and implement a function-based treatment to decrease aerophagia and increase prosocial behaviors that serve the same function.
2. Case Introduction
The participant, Khloe, was a 16-year-old female, diagnosed with Autism, seizure disorder, intellectual disability, and disruptive behavior disorder. Khloe was admitted to an outpatient clinic for assessment and treatment of severe problem behavior including aerophagia, self-injury, aggression, disruption, and elopement. For this paper, only data on aerophagia will be reported. Separate assessments and treatments were developed for all other problematic behaviors.
Khloe communicated vocally, in sentences. Instances of aerophagia were reported as occurring every 3–5 minutes, for less than 1 s each. Khloe resided in a group home for individuals with developmental disabilities. Her parents visited her at the group home regularly and were actively involved in her treatments.
All assessment and treatment sessions were 10 min and were conducted in treatment rooms (approximately 3 m × 3 m) at an outpatient clinic for severe behavior disorders. The treatment rooms were equipped with a one-way observation window and audio-video recording capabilities. All sessions were conducted in the afternoon, 2 days a week for 2 hrs each throughout an 11-month admission.
3. Presenting Complaints
Khloe’s cardiologist and her caregivers reported that their primary behavioral concern was aerophagia, due to the severe medical complications it was associated with. Most notably, the cardiologist diagnosed Khloe with arrhythmia and referred her to our clinic because he thought that aerophagia was a causal or contributing factor to this diagnosis at such a young age. The parents and the residential staff members reported that Khloe constantly sought attention, and frequently demonstrated problematic behavior to get a strong reaction; that is, she preferred extremes reactions (e.g., raised tone, elevated pitch, and reprimands) over moderate reactions, or positive social interaction (e.g., praise). Khloe enjoyed individual attention and did not respond positively to situations where the staff members’ attention was divided.
4. History
Khloe had a long and complicated medical history, spending a considerable portion of her childhood in and out of medical centers, and receiving extensive medical care during childhood. She reportedly preferred attention from medical care providers over anyone else’s, and she frequently made inaccurate statements about her health (e.g., “I have a headache,” “I am not feeling well,” “I need to see a doctor”), presumably to receive medical attention. It was unclear at the time of intake for how long Khloe showed signs of aerophagia, and reports indicated a graduated exposure development of overt signs such as belching or abdominal distention. At the time of intake, Khloe’s abdomen was extremely distended, which stood in contrast with her petite, thin figure. Khloe took the following medications throughout her admission: Aripiprazole, Simethicone, Divalproex, Benadryl, and Risperidone.
Throughout Khloe’s admission, we communicated with her medical doctor and psychiatrist regularly. This interdisciplinary approach to her treatment allowed all parties involved to make informed decisions that complemented each other. For example, the psychiatrist always reviewed the behavioral data before each consultation with Khloe and her parents and discussed treatment progress with our team. Similarly, the psychiatrist always informed us of any medication change (i.e., in Khloe’s case, these consisted of small changes in dosage) and expected behavioral effects. This information allowed us to tailor treatment accordingly. For example, if the psychiatrist stated that medication would make Khloe tired, we gave her more frequent breaks and allowed her to choose activities that were not effortful (e.g., sit and talk rather than do school work in between sessions).
5. Assessment
Response Measurement
For the free operant preference assessment, the target behavior consisted of engagement with toys. Engagement was defined as making physical contact with a toy or multiple toys for 5 s or more. We measured the duration of engagement by starting the timer when Khloe made contact with a toy and stopping the timer stopped making contact. At the end of the assessment, we calculated the total engagement with each toy and established a hierarchy of preference based on these data (i.e., longer duration of engagement indicated higher preference).
For the multiple stimulus without replacement preference assessment, the target behavior was choosing an item. Choice was defined as pointing toward or grabbing the item, stating the name of an item, or a combination of these responses within 5 s of the presentation of the items. Across trials, preference was indicated by the order in which items were chosen: the sooner an item was chosen, the more preferred it was considered to be. We calculated the number of choice responses for each item. At the end of the assessment, we established a hierarchy of preference from the most to least preferred based on the order of choice responses, as indicated above.
For the functional analysis, we collected data on swallowing air (aerophagia). Aerophagia was defined as tilting the head back, followed by swallowing air, which could or could not be audible. The experimenter collected data on each instance of swallowing air and converted the data in rate per min.
During treatment, the behavior targeted for increase was communication, defined as the participant pointing to the card on the communication board while making the targeted statement (i.e., “I am not feeling well”, or “I want to talk about my day”). The experimenter collected data on each type of communicative response separately and converted the data in rate per min.
During the functional analysis and treatment, the following measures of collateral changes were used: belching and abdominal circumference. Belching was defined as expelling air through the mouth from the stomach that resulted in audible noise. The experimenter collected data on each instance of belching and converted the data to a rate (responses per min). The experimenter measured the participants’ abdomen before (pre-) and after (post-) each functional analysis session. Then, the experimenter subtracted the pre-measurement from the post-measurement to calculate the difference. We collected and reported data on belching and abdominal circumference for sessions where aerophagia occurred (regardless of rate per min), and sessions where we did not observe any instances of aerophagia. This distinction allowed us to identify any possible correlation between aerophagia and belching, or between aerophagia and a change in abdominal circumference, respectively.
Interobserver Agreement
Two observers collected data simultaneously and independently on laptop computers, through a one-way mirror. The data collection software (BDataPro; Bullock et al., 2017) was used to calculate interobserver agreement (IOA). To calculate IOA, the total number of disagreements was divided by the total number of agreements and then multiplied by 100. An agreement was defined as both primary and secondary observers recording the same data within a 10-s interval. We collected IOA data for 5% of the functional analysis, and the agreement was 99.2% (range, 92.5–100%). We collected IOA data for 46% of the antecedent assessment, and the agreement was 100%. Finally, we collected IOA for 11.5% of the treatment sessions, and the agreement was 99.45% (range, 95.8–100%).
Treatment Integrity
All experimenters and caregivers were trained to a predetermined mastery criterion using behavioral skills training before conducting the study (Ward-Horner & Sturmey, 2012). Behavioral skills training consisted of an explanation of the procedure, followed by modeling, practice, and feedback. We collected treatment integrity data for each component of the treatment for 28% of the sessions. The average treatment integrity for the treatment was 99.96% (range, 97.9–100%).
Free Operant Preference Assessment
The free operant preference assessment was used to identify items that may serve as reinforcers for alternative, socially appropriate behaviors (Sautter et al., 2008). During the assessment, Khloe was left alone in the room with preferred items and the experimenter measured the duration of engagement with each toy. During the preference assessment, the only toy Khloe engaged with was the plush toy (74.6% of the time). Therefore, it was determined that the plush toy was the most preferred.
Multiple Stimulus without Replacement Preference Assessment
We conducted a multiple stimulus without replacement preference assessment (DeLeon & Iwata, 1996) to establish a hierarchy of preferences for edible items. The edible items consisted of different types of candy and gum. These were subsequently used during the competing stimulus assessment described in one of the following sections. The experimenter allowed Khloe to sample each stimulus before conducting the preference assessment, by offering a small portion of each at a time, stating its name (e.g., bubble gum), and allowing her 10 s to consume it. If Khloe chose not to consume an item, the experimenter removed it after 10 s. Once the preference assessment was initiated, the experimenter presented all stimuli and asked Khloe to choose one. Once Khloe chose an item, she was allowed to consume a small amount. Regardless of whether Khloe consumed the item or not, after 10 s the respective item was removed from the array. The experimenter rearranged the remaining stimuli on the table and asked Khloe to make another choice. This process was repeated until there was only one stimulus left, at which point the experimenter asked Khloe to consume it. Khloe consumed all but the last two items during the assessment. The top three choices were included in the second treatment described below.
Experimental Design
We used single-subject research designs for the functional analysis and treatment evaluation (Kratochwill et al., 2010). Single-subject research design allows the evaluation of experimental control even with a small number of participants, such as the one reported in this study. The functional analysis (described below) was initiated in a multielement design, where all conditions were alternated from session to session in randomized order. After several series, a pairwise design was implemented. In the pairwise design only two conditions were compared at a time—attention and toy play, or divided attention and toy play. We changed from a multielement to a pairwise design to evaluate which of two test conditions (i.e., attention or divided attention) occasioned aerophagia. We used a multielement design for the assessment of antecedent variables and a reversal design for treatment (both described below). For treatment, the initial baseline consisted of the data from the divided attention condition collected during the functional analysis.
Functional Analysis
The functional analysis was conducted to examine possible environmental determinants of aerophagia. We employed the following conditions based on Iwata et al. (1982/1994a): attention, escape, alone, and toy play (Iwata et al., 1982/1994a). We supplemented the analysis with a divided attention condition (Fahmie et al., 2013) and a tangible condition (Iwata et al., 1994b). Khloe demonstrated the highest rate of aerophagia (average number of responses per min of 1.88) in the divided attention condition. Across sessions, there was an increasing trend in responding. She demonstrated low and stable rates of aerophagia in the other conditions (range, 0.05 to 0.37 responses per min). These results suggested that aerophagia was maintained by attention and occurred primarily in the divided attention context. The data and detailed description of functional analysis conditions are available upon request from the first author.
We collected data on belching and changes in abdominal circumference throughout the functional analysis. We compared rates of belching and change in abdominal circumference for sessions where aerophagia occurred and sessions where aerophagia did not occur. Across sessions with aerophagia, Khloe demonstrated belching at a rate of 0.07 per min (range, 0 to 0.3) and her abdominal circumference increased by 19.5 cm (range, −3 to 20). Across sessions without aerophagia, Khloe demonstrated belching at a rate of 0.01 per min (range, 0 to 0.1) and her abdominal circumference decreased by 12.7 cm total (range, −0.1 to −8.9). Overall, we identified an inverse relation in belching and abdominal circumference when aerophagia occured. Nonetheless, due to the correlational nature of the data, conclusions regarding the precise relations among these three variables should be made with caution. After the functional analysis, we discontinued collecting these data because, whenever we measured Khloe’s abdomen, she made statements about her physical wellbeing and aerophagia (e.g., “Does my belly hurt because I swallow air?”). Therefore, we were concerned that the measurements could increase rates of aerophagia by bringing attention to it.
Antecedent Analysis
To further parametrically evaluate levels of attention that may occasion aerophagia, as well as to evaluate whether attention served as a motivating variable rather than a maintaining consequence for this behavior, we implemented an antecedent analysis. The analysis included the following conditions: alone, low attention one experimenter, low attention two experimenters, and high attention two experimenters. Each session was 10 min long. Khloe had access to preferred toys across conditions and sessions. There were no programmed consequences for any behaviors. In the alone condition, Khloe was left alone in the treatment room, while the experimenters observed from an adjacent observation room. This served as a control condition for forms of social interaction (remaining conditions). In the low attention one experimenter condition, Khloe interacted with an experimenter for 2 min, after which the experimenter stated “I am busy (reading a magazine/working on my phone), please play by yourself and when I finish we can talk.” Thereafter, the experimenter pretended to read a magazine or work on the phone. In the low attention two experimenters condition, Khloe interacted with two experimenters for 2 min, after which one of the experimenters stated “We need to do some work, please play by yourself and when we finish we can talk.” Thereafter, each experimenter pretended to be busy and did not interact with Khloe or with one another. In the high attention two experimenters condition, Khloe interacted with two experimenters for 2 min, after which one of them stated, “We need to talk to each other, please play by yourself and when we finish we can talk.” Thereafter, the experimenters talked to each other and did not provide any attention to Khloe. Across conditions, there were no programmed consequences for aerophagia—in other words, different levels and magnitudes of attention were arranged as antecedents, whereas no forms of attention ever consequated instances of aerophagia.
Khloe demonstrated high rates of aerophagia exclusively in the condition where two experimenters interacted with one another (high attention two experimenters; Figure 1), even though they did not provide any programmed consequences for aerophagia. Across sessions, the rates of aerophagia increased. These results indicate that aerophagia occurred independently of social consequences in divided attention contexts, possibly because divided attention functions as a motivating variable (Cengher & Miguel, 2019). Antecedent Assessment Results.
Competing Stimulus Assessment
Competing stimulus assessments are designed to identify stimuli that, when provided for free, compete and, thus, reduce problematic behavior (Hagopian, 2020). In our case, we wanted to identify edible items that competed with aerophagia. We selected items to include in this assessment based on the results of the multiple stimulus without replacement preference assessment described above. Items consisted of different types of gum and candy. Each trial lasted 2 min, and a different type of gum or candy was presented each trial.
The following types of trials were conducted: test trials and control trials (Control I and II). During Control I, Khloe was alone in the room with no access to any tangible objects while the experimenter observed from an adjacent observation room. This control condition allowed us to evaluate whether the baseline level of aerophagia in the absence of attention and competing stimuli (provided as antecedents). During Control II, Khloe was in the room with two experimenters who talked to each other but otherwise did not provide any programmed consequences to Khloe. This control condition allowed us to evaluate the baseline level of aerophagia in the absence of competing stimuli. Test trials were conducted similarly to Control II, with the exception that Khloe had free access to candy or gum (a different type presented each session). We conducted two series of test trials, and each series consisted of the presentation of each aforementioned stimulus. During the first series, Khloe had free access to the items during trials. During the second series, Khloe had free access to the items and the experimenter provided a verbal reminder that she can eat candy every 30 s; therefore, in this series we prompted engagement.
Khloe demonstrated the highest engagement and lowest rate of aerophagia in three conditions that entailed the presentation of different candy (e.g., Tic Tac). These stimuli were used as competing stimuli during subsequent treatment evaluations.
6. Case Conceptualization
Khloe demonstrated the highest rate of aerophagia in the divided attention condition, followed by the attention condition. These results suggested that aerophagia was maintained by attention, primarily in a divided attention context. It is important to note that oftentimes statements about health and wellbeing (e.g., “I am not feeling well,” “I have a headache,” “I should stop swallowing air”) co-occurred with aerophagia and followed the experimenters’ statements of concern. As such, we hypothesized that Khloe’s aerophagia occurred primarily in a divided attention context, was maintained by attention, and the form of attention that she sought was statements of concern or discussions about physical health and wellbeing.
Requests for medical attention and reports about poor health and wellbeing are only appropriate if there is correspondence between one’s physical health (e.g., one indeed has a headache) and the verbal report (e.g., “I have a headache”). If such correspondence does not exist, reports about poor health and wellbeing and requests for medical attention can result in unnecessary medical care that is costly and may have detrimental effects on one’s health (see Munchausen diagnosis symptoms and treatment; Aadil et al., 2017). However, at the onset of treatment, requests for medical attention seemed a more appropriate behavior as compared to aerophagia. Besides, based on the functional analysis data, we hypothesized that requests for medical attention and aerophagia served the same purpose—namely, to access attention primarily in the form of discussions about health and medical care. As such, during treatment we attempted to teach Khloe to (a) ask for medical attention (as a replacement for aerophagia), and (b) ask or talk about anything other than medical attention. During treatment and contingent on decreasing rates of aerophagia, we attempted to decrease the frequency of her requests for medical attention while increasing conversations around other topics. We implemented this arrangement in an attempt to teach and reinforce a more appropriate communication repertoire. However, this treatment was unsuccessful as rates of aerophagia persisted despite our attempts at differential reinforcement and extinction (described below).
During the first treatment, the persistence of aerophagia when social consequences were not provided suggested that either attention was not the primary maintaining consequence for aerophagia, or that other co-occurring consequences played a role in its maintenance. Thus, we hypothesized that a divided attention context may serve as a motivating variable for aerophagia—specifically, divided attention established the motivation to engage in aerophagia, and this behavior persisted independent of social consequences (i.e., aerophagia continued despite withholding social consequences, albeit occurring under strict motivational variables). We designed our second treatment largely based on this hypothesis. The treatment consisted of differential reinforcement of other behavior, noncontingent access to competing stimuli, and graduated exposure fading of social interaction (described below).
7. Course of Treatment and Assessment of Progress
Functional Communication Training for Attention
The experimenter conducted functional communication training (FCT) to teach Khloe to request two different forms of attention. The experimenter placed the communication board in front of Khloe and implemented a time delay procedure to teach the communication response. Each session consisted of 10 opportunities to respond (trials). For the first two FCT sessions, the experimenter provided a hand-over-hand prompt as soon as she placed the communication board in front of Khloe. The experimenter provided statements of praise and on-topic conversation contingent on a functional communication response. For the following two sessions, the experimenter presented the communication board and allowed 3 s for Khloe to respond. If Khloe engaged in the targeted FCT response within 3 s, the experimenter provided attention (3–4 statements). If Khloe did not engage in the targeted FCT response within 3 s, the experimenter provided the hand-over-hand prompt with very brief access to attention (one short statement). The criterion to increase the time interval between the presentation of the communication board and the prompt was two consecutive sessions with 90% or more correct and independent FCT responses. The delay to the prompt increased in 3-s increments. The training was conducted across two different FCRs: requests for medical forms of attention and other topics of discussion. If Khloe pointed to the card corresponding to medical forms of attention, the experimenter gave Khloe attention consisting of (a) questions about her health, and (b) statements of concern about her health, or (c) follow-up conversation about her health. If Khloe pointed to the card corresponding to other forms of attention (positive attention), the experimenter made positive statements (e.g., “I like your dress”), but refrained from making any statements about Khloe’s health. If Khloe attempted to talk about her physical health after choosing the card corresponding to other forms of attention, the experimenter did not provide any programmed consequences. There were no programmed consequences for any problematic behavior. The mastery criterion was 80% or more correct independent responses for two consecutive sessions. Khloe met the mastery criterion in 4 sessions.
Treatment Evaluation I for Aerophagia
Based on criteria used in previous studies (e.g., Hagopian et al., 2018), we determined that a clinically significant reduction in problematic behavior was 80% or more as compared to baseline (i.e., the functional analysis test condition that yielded the highest rate of problematic behavior). The sessions were 10 min long and were initially conducted by experimenters in a treatment room. Khloe had the communication board available on the table. Two experimenters were in the room talking to each other and not attending to Khloe unless she demonstrated the functional communication response. When Khloe pointed to the card corresponding to medical forms of attention, the experimenter gave Khloe attention consisting of (a) questions about her health, and (b) statements of concern about her health, or (c) follow-up conversation about her health. Khloe did not receive medical care (e.g., medical tests and medication administration) from our research team during treatment. When Khloe pointed to the card corresponding to other forms of attention (positive attention), the experimenter made positive statements (e.g., “I like your dress”), but refrained from making any statements about Khloe’s health. Khloe was free to choose between these two topics of conversation (data in Figure 2, in responses per min). There were no programmed consequences for any instances of aerophagia. Due to excessive use of the “I don’t feel well” card, at session 55, the team made a modification: Khloe had only three opportunities to use the “I don’t feel well” card during a 10-min session, while “I want to talk” was available continuously. Treatment Evaluation. Note. BSL stands for baseline and TX stands for treatment. The red dotted horizontal line represents the 80% reduction in rates of aerophagia as compared to baseline.
Khloe demonstrated a reduction in aerophagia as compared to baseline; however, the reduction in aerophagia did not reach clinically significant levels of 80% or more as compared to baseline. Therefore, we discontinued this treatment at session 142. We hypothesized that social interaction may evoke aerophagia, but that it may occur in the absence of socially mediated consequences (i.e., automatically maintained, where divided attention serves as a motivational variable). To verify that the aerophagia was evoked by divided attention and persisted in the absence of socially mediated consequences, the antecedent analysis was conducted. The results were used to inform the second treatment evaluation, described below.
Treatment Evaluation II for Aerophagia
At first, the sessions were 10 min long. Sessions were initially conducted by the experimenters in a treatment room. Khloe had the communication board available on the table. A whole-session different reinforcement of other behavior (DRO) procedure was implemented. With a whole-session DRO, the experimenter provides reinforcers contingent on the absence of a specific behavior throughout the session (i.e., in this case, swallowing air; Poling & Ryan, 1982). We supplemented the whole-session DRO with a token economy. Tokens are initially neutral stimuli that are paired with established reinforcers in an attempt to create a schedule that maintains responding under thin and delayed reinforcement conditions (Hackenberg, 2018). For example, instead of accessing a preferred game after completing a worksheet, the experimenter can create a token economy whereby a participant may receive a token for each worksheet completed, and only after receiving all five tokens, they may exchange them for access to the preferred game.
Khloe received a token contingent on the absence of aerophagia for a predetermined amount of time. Whenever Khloe demonstrated aerophagia, the timer was reset. Once she received five tokens, she was allowed to exchange them for access to laminated pictures of preferred characters. The initial time interval for the DRO was 30 s; this initial time interval was determined based on baseline levels of aerophagia observed during the functional analysis. In other words, Khloe demonstrated aerophagia on average every 30 s during the functional analysis; therefore, we determined that this time interval represents an optimal starting point for treatment as it would allow Khloe to contact reinforcement from its onset.
In addition to the DRO, Khloe was offered a choice of one (out of her top three) competing stimuli on a time-based schedule (i.e., noncontingent reinforcement). With noncontingent reinforcement on a time-based schedule, a preferred stimulus is provided after a predetermined amount of time has passed, regardless of what behavior occurred throughout that time (Marcus & Vollmer, 1996). The delivery of competing stimuli was initiated at session 189. We programmed the DRO and noncontingent reinforcement so that the delivery of reinforcers did not occur simultaneously. For example, it could be that during a session the token for the DRO was provided 5 min into the session, while the noncontingent reinforcement was implemented at minute 6.
As treatment progressed and Khloe demonstrated decreasing rates of aerophagia, we made changes to make the treatment more amendable and generalizable to non-clinical settings, and to make the procedures easier to implement for caregivers (and, subsequently, to increase treatment adherence). First, the aforementioned DRO and noncontingent reinforcement procedures began with a 30-s interval. The time interval for the DRO and delivery of competing stimuli was increased gradually up to 5 min. The criterion for increasing the time interval was two consecutive sessions with rates of aerophagia of 20% or less as compared to baseline. The time interval was increased in geometrical progression (i.e., 30 s, 1 min, and 2 min). At discharge, Khloe demonstrated low rates of aerophagia under a 5-min DRO and noncontingent reinforcement.
“The second phase consisted of graduated exposure to the contextual variables that evoked aerophagia”. This method of gradually introducing contexts that evoke problematic behavior is consistent with graduated exposure therapy (e.g., Allen & Kupzyk, 2016 [medical procedures]; Schmidt et al., 2013 [nonpreferred activities]; Szalwinski et al., 2019 [dental procedures]), where the presumably aversive stimulus is introduced gradually and approach responses are reinforced. In our case, we increased the number of experimenters present during the session from one to two (at session 77). Once Khloe demonstrated low rates of aerophagia in this context, the experimenters began talking to each other during sessions (i.e., divided attention). These changes were implemented to recreate the condition from the assessment of antecedent variables that evoked the highest rates of aerophagia. Our goal was for Khloe to demonstrate a reduction in aerophagia in this most problematic divided attention context, which is prevalent in the natural environment.
The third change consisted of a shift from a whole-session DRO to a momentary DRO. The change in DRO procedures occurred at session 219. With a momentary DRO, the experimenter provided the programmed consequence only at the end of the interval, regardless of what happened throughout the interval (although we collected data on aerophagia throughout the sessions; Repp et al., 1983). For example, if one were to implement a 5-min momentary DRO, they would reinforce the participant if no aerophagia occurred at the 5-min mark, regardless of what happened before. This change was implemented to make the treatment more amendable to non-clinical settings, as parents and other caregivers would likely not be able to constantly monitor Khloe’s behavior. The change in the DRO procedure did not increase the rates of aerophagia.
The third procedural detail that contributed to the generalization of treatment effects consisted of caregiver training and caregiver-implemented treatment. Caregivers were involved in treatment design and implementation from its onset. The experimenters implemented behavioral skills training with four different caregivers (i.e., both parents and two residential staff members). Each component of the treatment package was introduced during behavioral skills training, described in a previous section. As indicated in the treatment integrity section, the caregivers’ and staff members’ implementation of procedures was good (average of 99.97%).
8. Complicating Factors
The main complicating factor for this case involved adherence to treatment recommendations across staff members. Khloe interacted with numerous staff members at the group home and school. Some of these staff members participated in the assessment and treatment from the onset of Khloe’s admissions, while others did not. Non-adherence to treatment was due to (a) difficulty in coordinating schedules, (b) some staff members’ lack of concern regarding aerophagia, and (c) insufficient staff members in the group home. Specifically, Khloe had a one-to-one aide at school at all times who would implement treatment; however, staff members supervised multiple teenagers in the group home and, thus, found it difficult to implement our labor-intensive, individualized treatment. Therefore, it is likely that our treatment, while successful, was not implemented consistently across providers and settings. Future research should explore procedures to increase treatment adherence.
The second related complicating factor was that it was difficult to schedule generalization visits in the natural environment, primarily because there were so many staff members involved in Khloe’s care. In total, during the admission, the first author conducted two group home visits and two school visits. Generalization is an active process, and it is important that future research and clinicians program for it by conducting regular sessions outside of the therapeutic setting, implementing treatments across providers, and introducing new contexts to treatment. Relatedly, Khloe changed group homes 7 months after discharge from our clinic. Her new group home was located in another state, and our license did not allow us to continue follow-up services. Therefore, we only conducted a few follow-up visits post-discharge (described below).
9. Access and Barriers to Care
Khloe’s assessment and treatments were costly and time-consuming—she received services in our clinic for 11 months, at an intensity that ranged from 2-hr sessions twice a week, to 5-hr sessions five times a week. This prolonged duration of the admission was partially caused by the other problematic behaviors that we treated, as well as the relatively treatment-resistant nature of Khloe’s aerophagia. To illustrate this latter point, we accomplished our treatment goal (i.e., a reduction in aerophagia of 99.09% as compared to baseline) in 226 sessions. This level of investment in assessment and treatment (i.e., time and money) is most likely difficult to accomplish for most caregivers. In addition, clinics specialized in the assessment and treatment of severe problematic behaviors such as ours are few across the country. For this reason, many of the patients seen at our clinic come from other states or countries, but, likely, some families cannot afford to travel far for treatment or take extensive time off from work or other responsibilities. One of the few positive consequences of COVID-19 is that it opened the door for such therapeutic services being provided remotely, and we recommend that future research investigate whether remote services are as effective as services provided in person. Another important avenue for future research is to explore abbreviated therapeutic approaches for aerophagia that may result in a suppression of the behavior to clinically significant levels using fewer resources.
10. Follow-Up (How and How Long)
We conducted follow-up visits for 7 months after discharge. Follow-up visits were initially conducted bi-monthly, and their frequency was then decreased to monthly and every other month. The decrease in the frequency of visits was due to Khloe demonstrating low rates of aerophagia. We conducted these follow-up visits at Khloe’s group home and school. The purpose of these visits was to collect behavioral data to monitor the effects of treatment and its maintenance, as well as to discuss treatment progress with staff members and to conduct ongoing training in treatment implementation. During these visits, Khloe’s rates of aerophagia maintained at a rate of 0.1 or below.
11. Treatment Implications of the Case
This case study provides an example of assessment and treatment for aerophagia in a 16-year-old girl with multiple diagnoses. We began with a functional analysis that indicated that aerophagia was maintained by attention and that it occurred primarily in a divided attention context. Our first treatment evaluation was matched to this function and it entailed attention extinction for aerophagia (i.e., no attention provided contingent on this behavior), as well as attention (presumably a reinforcer) for communication. Given that Khloe made frequent inaccurate statements about her health and that she preferred to talk about her wellbeing over other topics, we attempted to shift the focus of the conversation from medical care to other, more socially appropriate contexts. Our treatment did not result in a meaningful reduction in aerophagia. We then conducted an additional assessment that confirmed that Khloe engaged in aerophagia in a divided attention context, but the behavior persisted independent of social consequences. As such, we conceptualized that a divided attention context serves as a motivational variable for aerophagia. Our second treatment included a DRO and noncontingent access to competing stimuli, consisting of preferred mints and candy. We took several measures to ensure treatment success and generalization, including fading in the number of people present during treatment and the graduated exposure introduction of the divided attention context. Our treatment resulted in a reduction in aerophagia of 99.09% as compared to baseline (i.e., divided attention condition of the functional analysis).
The main clinical implication that we want to highlight is that a function-based assessment is crucial even for response topographies that are traditionally conceived as being automatically maintained (e.g., aerophagia, pica—see Ness et al., 2020). Even though we found that Khloe’s aerophagia occurred independent of social consequences, the assessment results also allowed us to learn that (a) it occurred exclusively in a divided attention context, and (b) that divided attention context served as a motivational variable. This information was crucial for the selection of our treatment components (e.g., graduated exposure fading of the divided attention context, competing stimuli). In addition, we were mindful of Khloe’s motivation to receive medical attention and, to avoid potentiating this issue, we incorporated a DRO procedure in treatment that allowed us to reinforce the absence of aerophagia without bringing any attention to it.
12. Recommendations to Clinicians and Students
The first recommendation we have for clinicians and students is to always begin with a functional assessment to evaluate the function(s) of behavior. The assessment results can not only clarify what environmental variables maintain the behavior but also what the starting point of treatment should be and what the treatment should consist of. For example, our first goal for the DRO (i.e., a 30-s interval) was informed by the rate of aerophagia observed during the assessment. Similarly, during the assessment, we observed that aerophagia occurs primarily in a divided attention context, and, as such, we gradually faded in other people and conversation during treatment to ensure Khloe’s success.
Second, we had an interdisciplinary approach to assessment and treatment that involved collaboration with multiple professionals (e.g., school staff members, group home members, medical doctors, and psychiatrists). Such an interdisciplinary approach can help all parties involved and, most importantly, help design an effective and generalizable treatment. We recommend that clinicians seek input from all relevant professionals involved in their patients’ care.
Third, we collected data not just on the target behaviors, but also on collateral variables (i.e., belching and abdominal circumference). We recommend that clinicians collect such data whenever possible, as they can provide additional evidence for treatment effectiveness. For example, a reduction in abdominal circumference that co-occurs with a reduction in rates of aerophagia provides further confirmation that treatment is effective. In addition, collateral measures may replace behavioral data when the therapists cannot reliably collect such data. For example, Khloe’s aerophagia was relatively easy to observe and collect data on; however, other individuals may swallow air in a less observable way. In such cases, data on collateral measures can play a critical role in evaluating treatment effectiveness.
Fourth, we noted that the assessment and treatment approaches outlined in this case study are amendable for behaviors that traditionally have been conceived as occurring independently of social consequences, such as aerophagia or pica. However, such assessment and treatment approaches have been demonstrated to be highly successful for a wide range of topographies of behavior, including different other forms of self-injury, aggression, and disruption (Beavers et al., 2013). We specifically want to highlight that the use of graduated exposure procedures can be highly effective for escape-maintained behaviors that are common in individuals diagnosed with phobias (e.g., Allen & Kupzyk, 2016 [medical procedures]; Schmidt et al., 2013 [nonpreferred activities]; Szalwinski et al., 2019 [dental procedures]). Our study extended this graduated exposure literature by demonstrating the effectiveness of these procedures on behaviors that are not escape maintained. As such, we recommend that clinicians consider replicating our procedures with other individuals that demonstrate aerophagia, as well as with individuals that demonstrate other topographies of problematic behavior.
It is important to highlight the versatile role attention can play in our everyday interactions. Khloe, for example, constantly sought medical attention possibly due to a long history of interacting with medical providers. For Khloe, we also found that a divided attention context evoked aerophagia. Other studies found that certain forms of attention can be punishing (Cengher et al., 2021b), can reduce one’s speech to the point where they are diagnosed with selective mutism (Cengher et al., 2021a), and can function as a reinforcer for extremely dangerous problematic behavior (e.g., pica, Ness et al., 2020). Collectively, these studies demonstrate that attention can play many roles. It is, thus, important for clinicians to be mindful that attention has so many formal and functional properties. Knowledge about the formal and functional properties of attention should help one design more nuanced assessments and treatments, individualized based on each patient’s profile.
There are three primary limitations to this case study. The first, which is a limitation of all research on the topic of aerophagia, is that we have limited evidence of generality of treatment effects given that we only employed one participant. However, given that aerophagia is not common, it would be difficult for researchers to employ large-N designs. In addition, the function-based approach to assessment and treatment of problematic behavior has vast empirical support across numerous problematic behaviors (Beavers et al., 2013). To further increase the generality of research on aerophagia specifically, we recommend that researchers conduct meta-analyses to quantify the effectiveness of treatments, or to conduct controlled retrospective case series designs (Hagopian, 2020) if having access to a larger database with patients. Second, we have limited generalization data. There are several variables described above that contributed to this limitation. Future research should aim to demonstrate the generalization and maintenance of treatment effects. Third, Khloe’s aerophagia was relatively resistant to treatment. Given the limited number of studies focused on this dependent variable and the variability in populations and procedures implemented, it is difficult at this point to verify whether this form of behavior is generally resistant to treatment. However, a long history of engaging in the behavior or failed medical trials to treat the behavior can be indicative of resistance to treatment. In such cases, the therapist can ensure that they have the resources necessary available to ensure the success of treatment. This could take the form of devoting sufficient time to treatment, as well as having discussions with stakeholders early on about the time commitment this treatment would likely entail.
In conclusion, this case study exemplifies the effectiveness of an assessment-based approach to the treatment of aerophagia, the importance of interdisciplinary collaborations, and the utility of collecting collateral measures when possible. All of these approaches can help clinicians develop and provide effective and generalizable treatments.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
