Introduction

The ability to eat is a human necessity, as food supplies our bodies with fuel and nutrition to function in everyday life. Approximately 35% of children with neurodevelopmental disabilities and 25% of typically developing children have feeding concerns (Arts-Rodas & Benoit, 1998). The three main categories of feeding concerns are food refusal, refusal to self-feed, and food selectivity (Tereshko et al., 2021). When children exhibit feeding concerns, it can be a common source of stress for the caregivers, and approximately 40% of caregivers of children on the autism spectrum have reported mealtime difficulties (Sharp et al., 2010). Further, research has shown that approximately 89% of children on the autism spectrum exhibit a strong preference for certain foods based on color, texture, etc., and engage in higher rates of disruptive behaviors (e.g., getting up from the table, pushing away food, throwing items) at mealtime compared to typically developing peers (Sharp et al., 2010). For many children, feeding concerns are multifaceted and include medical, behavioral, and oral motor concerns (Piazza, 2008). Specifically, many children may experience pain or discomfort after eating and develop behaviors effective at avoiding eating and the subsequent discomfort. Additionally, children who do not engage in typical eating behaviors may have limited opportunities to develop oral motor skills (Piazza, 2008). Although for many children biological factors impact their food refusal, some research has indicated that feeding habits for children on the autism spectrum and/or with other neurodevelopmental disabilities are not always related to the organic factors or gastrointestinal problems but are often considered behavioral concerns (Sharp et al., 2010).

Evidence-Based Interventions

A systematic review of feeding interventions has identified several intervention approaches used to address increasing food acceptance. These include contingent rewards, non-removal of the spoon (escape extinction), stimulus shaping and fading, re-presentation of expelled food, non-contingent reinforcement, response shaping, simultaneous presentation, behavioral momentum, visual supports, provision of negative consequences, and choice (Ledford et al., 2018). As many inappropriate mealtime behaviors have been observed to occur in order to escape from feeding demands, escape extinction, or non-removal of the spoon, is an intervention that has been found to be effective (LaRue et al., 2011; Sharp et al., 2010; Smith et al., 2019). One modification to escape extinction includes a time cap (Smith et al., 2019). A time cap involves an agreed-upon end time of the session. If the child does not engage in the required feeding behavior before the time elapses, the session is terminated, and the child may receive possible negative reinforcement by the removal of the demand (Smith et al., 2019).

Although escape extinction has a significant evidence basis through research and many caregivers have found it to be an important treatment component (Taylor & Taylor, 2022; Woods & Borrero, 2019), there are some considerations for this approach in outpatient treatment, especially in programs with fewer and/or briefer sessions. Some caregivers may find escape extinction difficult to implement, especially when it is challenging to keep the child at the table in the home setting (Tereshko et al., 2021). Furthermore, research has indicated that feeding interventions that include antecedent and reinforcement strategies in the absence of more intrusive procedures, such as escape extinction, have been successful at improving feeding behavior (Tereshko et al., 2021; Turner et al., 2020).

Interdisciplinary Treatment

While there has been significant support for behavioral interventions to address feeding concerns, feeding problems are often multifaceted and best addressed by an interdisciplinary team (McComish et al., 2016). Disciplines that may be included in interdisciplinary feeding work include psychologists, nutritionists, occupational therapists, speech and language pathologists, social workers, medical providers, and other professionals (McComish et al., 2016). Specifically, a psychologist on the team may address behavioral and mental health concerns, such as fear associated with feeding (Silverman, 2010). They may also provide evidence-based behavioral strategies to decrease disruptive behaviors and increase appropriate feeding behaviors during mealtime (Silverman, 2010). The occupational therapist will often provide assessment and treatment recommendations to address difficulties with oral motor skills, which could include difficulties with food textures or gagging (Beaudry-Bellefeuille et al., 2021). Expertise in sensory functioning also allows the occupational therapist to assess and provide treatment recommendations for feeding concerns that may be sensory in nature (Beaudry-Bellefeuille et al., 2021).

Research on interdisciplinary treatment has resulted in improved feeding behaviors of children with pediatric feeding disorders in day treatment programs and inpatient programs, including children with severe feeding difficulties that have resulted in the use of a feeding tube (Seiverling et al., 2020; Sharp et al., 2017; Volkert et al., 2021). Improved outcomes from interdisciplinary treatment programs have included decreases in inappropriate mealtime behaviors, decreases in caregiver stress, and increases in food volume and variety (Greer et al., 2008; Seiverling et al., 2020; Sharp et al., 2017). While positive outcomes from interdisciplinary feeding treatment from day treatment and inpatient programs have been prevalent in the feeding literature, there has been less research on the use of interdisciplinary approaches in more traditional outpatient settings.

Caregiver Training

Caregivers are responsible for providing most meals to their children; however, there is limited research examining the caregivers as intervention implementers, as opposed to researchers or clinicians (Ledford et al., 2018; Seiverling et al., 2012). A study by Johnson et al. (2019) evaluated the caregiver training to address feeding concerns and found that caregiver treatment fidelity was high. Additionally, it was found that caregivers found the intervention to be feasible and were satisfied with results (Johnson et al., 2019). Caregiver training via telemedicine has also been found to result in high caregiver treatment fidelity and improvements in feeding behaviors of children (Bloomfield et al., 2019, 2021). Based on previous research, it has been indicated that there is a need to work with caregivers to identify specific feeding goals for their child (Ledford et al., 2018). Additionally, as the goal is to improve feeding behaviors in natural settings (e.g., home), evaluating generalization of caregiver skills to home and other natural settings is important. However, to date, much of the previous clinic-based research has not addressed generalization to other settings (Ledford et al., 2018; Tereshko et al., 2021). When working with caregivers, it is important that the social validity of treatments, in addition to treatment outcomes, is addressed. Measuring social acceptability of feeding interventions has been identified as an area in need of additional research (Ledford et al., 2018; Tereshko et al., 2021).

Pilot Program Overview and Aims

The interdisciplinary outpatient feeding treatment program was designed to address feeding concerns of children with developmental disabilities in a hospital medical center, including patients who had previously been unsuccessful in one time per week, single disciplinary outpatient services at the same facility or with outside services (e.g., speech and language therapy, occupational therapy, and behavioral therapy). The current program included the combination of two disciplines (i.e., psychology and occupational therapy) to address the behavioral and oral motor concerns of participants. These disciplines were selected for the current study because (1) most children referred were already receiving medical care (e.g., gastroenterology, pediatrician) for biological causes of feeding concerns as needed but continued to display behavioral and/or oral motor concerns that required additional support, and (2) psychologists and occupational therapists were the primary providers of feeding treatment services at the current clinic and the pilot program was designed to integrated the two treatment approaches into one service for children for whom this was appropriate.

The first aim of this pilot program evaluation was to evaluate if changes in feeding behavior of participants in the program were observed including (1) percentage of bites accepted in clinic and at home; (2) rate of inappropriate mealtime behaviors in clinic and at home; (3) percentage of individualized treatment goals met for participants; and (4) reported number of foods consumed by caregivers. Additionally, the second aim of this pilot program was to evaluate if the program was (1) feasible for caregivers to implement based on a treatment fidelity checklist; (2) if caregivers reported a change in level of concern related to their child’s feeding behavior and their confidence to address their child’s feeding behaviors; and (3) acceptability of the treatment program based on a feedback survey. The final aim was to determine whether there was a difference in treatment outcomes based on the length of the program (12 vs 8 weeks).

Method

Recruitment

Participants for the pilot program were recruited at a hospital medical center in the Midwestern United States. All participants were identified as having a developmental disability and Avoidant/Restrictive Food Intake Disorder (ARFID) or other specified feeding disorder. All participants were referred to the hospital medical center for treatment of feeding concerns. Prior to participation in the program, each child had an intake appointment with an occupational therapist and psychologist. During this intake appointment, caregivers were asked questions about their child’s feeding behaviors and concerns, and a meal observation was conducted. At the end of this assessment, the occupational therapist and psychologist provided their treatment recommendations to the family. If the child was recommended for the interdisciplinary outpatient feeding treatment program, a description of the program was provided, and caregiver questions were answered. If the caregiver was interested in their child participating in the program, the child was added to the waitlist for this program. Families were contacted in the order they were added to the waitlist regarding openings in the program. All families who were placed on the waitlist and offered a spot in the pilot program agreed to participate.

Participants

Seven children participated in the 12-week pilot feeding program, and five children participated in the 8-week pilot feeding program. The median age of participants for the 12-week program was 5.7 years (3–11 years of age). The median age of participants for the 8-week program was 5.8 years (3–10 years of age). Participants in the 12-week program were 86% male and 100% male in the 8-week program. Participants in the 12-week program were 71% White and 29% Black. Participants in the 8-week program were 80% White and 20% Black. All participants in the 12-week program had a diagnosis of autism spectrum disorder. Four of the five participants (80%) in the 8-week program had a diagnosis of autism spectrum disorder. The other participant had a diagnosis of Down syndrome. At least one primary caregiver attended all the appointments.

A total of eight participants were considered to have completed the 8-week or 12-week program (participant pseudonyms: Maximus, Damari, Kinley, Asher, Marley, Kody, Kendrick, and Samuel). An additional four children who started the pilot feeding program did not attend enough appointments to meet the parameters of program “completion” and are not included in the pilot program outcomes. Specifically, participants were expected to attend at least 10 sessions in total (two per week equaling at least five weeks in the program) in order to “complete” the program. One child who did not complete the program participated in the 12-week program. This child had frequent illnesses which resulted in frequent cancelations of sessions and, ultimately, the family elected to discontinue the program before the end of the 12 weeks. All participants in the 12-week program participated prior to the COVID-19 pandemic, while the participants in the 8-week program participated during the pandemic. Three children in the 8-week program did not complete the program. Reasons for cancelations included complications related to the COVID-19 pandemic, illness, and schedule conflicts. One family elected to discontinue appointments before the end of the 8-week program due to difficulties with telemedicine sessions (family elected not to come to clinic setting due to the pandemic). The other two participants completed the appointments they were able to make during the 8-week program, but due to frequent cancelations of sessions were not considered to have completed the program. Table 1 includes the participant information for the six participants who completed the 12-week program and the two participants who completed the 8-week program.

Table 1 Participant information
Full size table

Setting

This was an outpatient program that included caregiver training at an outpatient hospital medical center. Appointments were conducted in a private therapy room with different sized tables for feeding (based on the size and age of the child). Rifton chairs and/or highchairs were used with participants who currently ate in either of these seating options at home. Telehealth appointments were conducted once or twice a week (depending on family preference) for the 8-week participants only. Families were in their homes, typically in the kitchen or dining area, for telehealth appointments.

Procedures

Baseline Assessment

To assess the challenging behavior related to feeding, three to four observations of feeding behaviors in the clinic and in the home setting (via telemedicine or video recording) occurred. Caregivers conducted each of these sessions and were instructed to present four different foods, five times each. All foods were provided by the caregivers, who were instructed to use foods that were common to their home setting. Two foods were reported to be the participant’s preferred foods, and two foods were reported to be non-preferred foods. The caregiver was instructed to present one bite at a time (or for self-feeders presenting all food but verbally prompting to “take a bite of (food)” for each bite presentation). During this phase, the caregiver was told they could use any materials or activities typically used during mealtime and respond to their child as they typical would once bites were presented. For all participants, inappropriate mealtime behaviors (e.g., turning head, pushing spoon away, crying, walking away) were observed when bites of non-preferred foods were presented, and these behaviors stopped when the bite was removed. Based on these observations, escape was hypothesized to be a maintaining function of inappropriate mealtime behavior for all participants. For some participants, increased caregiver attention was observed in response to inappropriate mealtime behaviors. Providing access to preferred tangible objects was never observed by caregivers following inappropriate mealtime behavior; however, at times access to a more preferred food was provided after the non-preferred bite was removed.

Treatment Program

The interdisciplinary outpatient feeding treatment program was developed to address feeding concerns of children with developmental disabilities who were previously unsuccessful in single discipline treatment services that occurred one time per week or every other week. This program was initially piloted as a 12-week program. Many participants of the program met treatment goals more quickly than expected. There were also clinician time constraints and wait list factors, which are common systems barriers in hospital medical centers. Therefore, this led to the decision to also pilot an 8-week program.

Appointments occurred two times per week for approximately 50 min each. One appointment was a co-treatment appointment with the occupational therapist and psychologist. This appointment occurred on-site in the clinic or via telehealth. The second appointment occurred with only the psychologist. This appointment occurred on-site in the clinic or via telehealth. The goal of the co-treatment appointment each week was to update the individual feeding treatment protocols for the child as needed based on progress (e.g., adjust bite number required for reinforcement, introduce a new food, alter an antecedent or differential reinforcement strategy). The second appointment with only the psychologist was focused on caregiver training on the treatment protocols. Telehealth appointments were only available after the start of the pandemic and, therefore, were only available for the 8-week program participants. Caregivers were expected to implement the treatment protocols in the home setting daily.

Following collection of baseline data, a preference assessment was conducted, individualized treatment goals were identified with caregivers, and an initial treatment protocol was developed. All treatment protocols included antecedent-based strategies (e.g., choices, visual supports, behavioral momentum) and differential reinforcement (e.g., reinforcement provided for a target skill such as sitting appropriately, taking a bite, or cooperating with demand completion depending on currently targeted behavior for participant).

Behavioral shaping was typically used, in which initially approximations of the terminal behavior were reinforced (e.g., taking a bite from an empty spoon or preferred food) and gradually expectations were increased as the child was successful (e.g., gradually increasing bolus on spoon until taking a bite from a full spoon). Behavioral momentum was also often used. In some instances, this was done using high probability non-food related demands initially and gradually adding in food related demands as cooperation with demands increased (e.g., completion of matching task not related to food, completion of matching task of pictures of food, completion of matching tasks using actual food, completion of taking a bite of a food) to access reinforcement.

Partial escape extinction was also often used (not in the typical form of non-removal of the spoon). For participants who were self-feeders (Marley, Kody, Kendrick, and Samuel), once a bite was presented, it remained on the table until consumed or a specific time cap was met. For participants who were not self-feeders (Maximus, Damari, Kinley, Asher), the caregiver would present the bite via stationary spoon (i.e., holding a bite in the same spot) until the bite was consumed or a specific time cap was met. As the bite was not removed contingent upon challenging behaviors, escape was not provided for inappropriate behaviors but rather upon bite acceptance or the end of the time cap (i.e., partial escape extinction). When a time cap was recommended, the therapists and caregiver would discuss what was feasible at home and possible benefits from using a time cap. A decision would then be made collaboratively on the length of the time cap, ensuring it (1) provided an adequate amount of time for the child to potentially take the bite while (2) being feasible for the caregiver to implement consistently at home. Time caps for self-feeders (presentation of bite on plate) ranged from 5 to 10 min for non-preferred foods. Time caps for meal sessions (presentation of previously mastered foods) extended up to 30 min for self-feeders. For non-self-feeders, the time cap for stationary spoon ranged from 1 to 5 min. Only strategies that a caregiver could feasibly implement in the home setting were recommended in the treatment protocol. The caregivers were provided copies of any visual supports to use at home (and/or school if appropriate) during mealtimes.

The same behavioral procedures (i.e., antecedent strategies and differential reinforcement) were used across participants, but individualizations to protocols were made based on the specific goals, skills, and needs of participants and caregivers. For example, most participants had visuals used as part of their treatment protocol (e.g., number board, visual reminder of steps in self-feeding), but three participants (Maximus, Damari, and Asher) did not. For antecedent strategies, all participants had choice and behavioral momentum as treatment components. For Maximus, texture fading was also used as a strategy (i.e., gradually increasing the texture of the food from a smooth puree), as he had previously only consumed smooth, puree textures. For Kinley, simultaneous presentation and fading (i.e., pairing a preferred with non-preferred food and gradually decreasing volume of preferred food) were used with lower texture foods that could be blended well. Her bite acceptance was higher when preferred and non-preferred foods were initially blended and gradually transitioned to the non-preferred food only, which is why this strategy was used with her. Additionally, Kinley engaged in expelling of non-preferred foods at baseline. To address this concern, expelled bites were re-presented (i.e., same bite that was expelled or new bite of same food, depending on appropriateness of situation) immediately following expulsion via stationary spoon (2-min time cap). Other individualizations to protocol included self-feeder vs non-self-feeder bite presentation, bolus size (amount of food presented), specific foods presented, total number of bites presented, and reinforcement provided. These specific protocol details were updated week to week as needed, based on the progress of the participant.

After each session, the caregiver was provided a written copy of the child’s feeding protocols that were to be practiced at home between appointments. This was provided after each session, even if no changes to the protocol were made during the session (i.e., continuing plan from previous week). At the completion of the program, a recommendation was made for follow-up services. This included follow-up services with the occupational therapist, psychologist, or another provider as appropriate, based on the individual needs of each participant at the time of completion of the program.

Measures

Outcome measures of the pilot feeding program included direct observation of feeding behaviors, measured progress toward individual treatment goals, caregiver-reported food list, caregiver treatment fidelity, caregiver ratings, and a social validity questionnaire.

Direct Observation

Direct observation data were collected during each session with the participant using either paper, pencil, and stop watches or BData Pro software. During baseline sessions, the caregiver was instructed to present five bites of four different foods, one at a time. Two of the foods were identified by the caregiver as foods the child consistently consumed at home (i.e., preferred) and two were foods the child did not consume at home at the time (i.e., non-preferred). During the intervention, the psychologist directly coached the caregiver to present a specific sequence of preferred and non-preferred bites in accordance with the participant’s individual treatment goals. The psychologist and/or a behavioral assistant collected the data during baseline sessions and at the end of intervention. Data were collected on the following variables: bite presentation, bite acceptance, latency of acceptance, self-fed bites, inappropriate mealtime behavior, aggression, self-injurious behavior, packing, gagging, expelling, and time in seat. Due to limited staff, and because this was a pilot of a clinical program, inter-observer agreement was not able to be collected for all participants. However, to ensure all three observers were consistent in data collection, agreement was obtained on 25% of all sessions across both baseline and intervention phases via video recordings (home sessions) and/or live data collection in the clinic room or over telehealth. Inter-observer agreement was 88% (range 79–95%).

For the purposes of this evaluation of the feeding program, two direct observation measures are reported, as they were relevant to all participant goals. The first is percentage of bites consumed (bites consumed divided by bites presented times 100). A bite presentation consisted of any bite presented to the child’s lips and/or on the table in front of the child and/or a verbal prompt to “take a bite.” A bite acceptance was defined as a bite that passed the child’s lips and was swallowed. The second was the responses per minute of inappropriate mealtime behaviors. Inappropriate mealtime behaviors were defined as any verbal or non-verbal behavior that indicated refusal to take a bite, or that interfered with bite consumption. This included turning head from a bite, pushing the bite away, saying “no,” throwing food, or pushing away from the table.

Frequency data were collected on the occurrences of inappropriate mealtime behavior, bites presented, and bites accepted for all participants. Using the frequency data, collected percentages were calculated of percentage of bites accepted (bites accepted divided by bites presented times 100). Inappropriate mealtime behavior frequency data were presented in responses per minute.

Data were collected during three to four baseline sessions. At least one of these sessions occurred in the home setting either via telemedicine (8-week program participants only) or by the family recording a meal session on a camera that was provided to them, and the data were collected by the psychologist and/or behavior assistant from viewing the video recording. Data were then collected at in-person sessions throughout the intervention and at the end of the program, as well as data from at least one home session (via telemedicine or video recording by the family).

Treatment Goal Progress

After baseline data were collected, the occupational therapist, psychologist, and caregiver decided on specific treatment goals. Caregiver input on goals was most important in development of these goals, with input from the therapists on ensuring goals were measurable and achievable. Three behavior-specific goals and three to four occupational therapy-specific goals were developed per participant. Behavior-specific and occupational therapy-specific goals were related and both therapists worked on comprehensive treatment protocols to address all goals. Individual discipline goals were set for billing purposes and to clarify the necessity for both disciplines and the specific focus of each clinician during therapy sessions. Caregiver treatment fidelity was included as a behavior-specific and occupational therapy-specific goal for each participant. See Table 2, for a list of all behavioral goals, and Table 3, for all occupational therapy goals. At the last appointment in the program, progress toward all goals was assessed via direct observation data to determine whether each goal was met or not met. The percentage of goals met was reported for each participant and shared with the caregiver at the time of program completion.

Table 2 Participant behavioral goals
Full size table
Table 3 Participant occupational therapy goals
Full size table

Food List

Prior to the start of the program, each caregiver was asked to list all the foods their child currently consumed. Food consumed included any food the child would accept or independently consume at least 80% of the time when presented. A specific portion was not specified, so if the child consumed any portion of the food (e.g., child friendly portion or smaller), this food was included in the list. Therefore, foods inconsistently consumed were not included on this list. Throughout the program, foods that were introduced that the child would consume in the home setting at least 80% of the time based on caregiver report were added to the list. Any food that was introduced during the program but consumed less than 80% of the time at completion of the program (based on direct observation data and caregiver report) was not added to the list. The number of foods consumed at baseline and completion of the program were calculated for each participant.

Treatment Fidelity

Individualized treatment protocols were developed for each participant, but core behavioral strategies that were consistent across most protocols were developed into a treatment fidelity checklist. Six behavioral strategies were observed to either occur, not occur, or were not applicable during a meal observation. See Table 4, for the list of strategies. To calculate the treatment fidelity, the total number of strategies that occurred was divided by the total number of behavioral strategies that occurred plus did not occur (excluding any determined to be not applicable) and multiplied by 100.

Table 4 Treatment fidelity components
Full size table

Caregiver Ratings

Prior to intervention and at the end of treatment, each participant’s caregiver rated their concern for their child’s feeding on a scale of 1–10 (1 being highest level of concern and 10 being no concern). Caregivers were also asked how confident they felt addressing their child’s feeding goals on a scale of 1–10 (1 being not at all confident, 10 feeling completely confident) prior to treatment and again at the end of the program.

Social Validity

At the end of the treatment program, each participant’s caregiver was asked to provide ratings on their satisfaction with the program. They were provided a questionnaire with 7 total questions and provided ratings from 1 (strongly disagree) to 5 (strongly agree). Questions were developed specific for this intervention to evaluate the utility of the program to continue in the clinical setting. Additionally, caregivers were asked to provide any additional comments about the program. See Table 5, for the social validity statements used in the social validity survey.

Table 5 Social validity questionnaire
Full size table

Results

Direct Observations

During direct observations, the percentage of bites accepted and responses per minute of inappropriate mealtime behaviors were recorded. Participants in the 12-week program accepted an average of 45.55% of bites at baseline and 99.64% of bites during intervention. Participants in the 8-week program increased in bite acceptance from an average of 40.40% at baseline to 95.00% during intervention (Fig. 1). For the participants in the 12-week program, a decrease in inappropriate mealtime was observed from an average rate of 1.59 min−1 at baseline to 0.24 per minute during intervention. For the 8-week program participants, a decrease in inappropriate mealtime behavior was observed from an average rate of 1.11 min−1 at baseline to 0.46 min−1 during intervention (Fig. 2). Figures 3 and 4 show the observation data on average responses per minute of inappropriate mealtime behaviors and percentage of bites accepted for each participant. Data collected in the home setting via telemedicine or video recording by caregiver are indicated on these figures.

Fig. 1
figure 1

Average percentage of bites accepted

Full size image
Fig. 2
figure 2

Average rate of inappropriate mealtime behaviors

Full size image
Fig. 3
figure 3

Rate of inappropriate mealtime behaviors for each participant

Full size image
Fig. 4
figure 4

Percentage of bites accepted for each participant

Full size image

Treatment Goals

See Tables 2 and 3, for list of all participant goals and goal mastery. Participants in the 12-week program met 97.67% of treatment goals (range 86–100%). Participants in the 8-week program met 91.5% of treatment goals (range 83–100%). In total, only two treatment goals were not obtained (one goal each for Damari and Kendrick). Both goals that were not met were specific to the number of non-preferred foods introduced and accepted. For both participants, another goal was more significant to the caregiver (e.g., self-feeding, sitting during meals), and the focus of treatment was directed to the preference of the caregiver. Both participants increased their food variety, but not by the specific number of foods initially indicated at the outset of treatment, and this goal was still ongoing at the end of the program.

Food List

Participants in the 12-week program averaged 4.67 foods at baseline (range 0–11) and 11.00 foods at the end of the program (range 2–25). This is an increase of 6.33 foods during the 12-week program. Participants in the 8-week program averaged 10.00 foods at baseline (range 9–11) and 15.00 foods at the end of the program. This is an increase of 5 foods during the 8-week program.

Treatment Fidelity

Treatment fidelity for all caregivers was 100% at the completion of the treatment program. See Table 4, for a list of components and percentage of steps completed or not applicable. Some components were considered not applicable based on the individual treatment protocol of each participant. For all participants, the first three strategies (i.e., clear presentations, contingent reinforcement, and partial escape extinction) were consistently implemented and considered appropriate. The fourth strategy, ignoring inappropriate behavior or removal of preferred items contingent on inappropriate behavior (based on treatment plan), was considered not applicable for two of the eight participants (Kody and Asher), as inappropriate behaviors did not occur during the observed sessions. The fifth strategy, use of visual supports, was not applicable for three of the eight participants (Maximus, Damari, and Asher), as visuals were not used in their individualized treatment protocols. Finally, the sixth strategy, use of verbal, gesture, or physical prompts, was considered not applicable for two of the eight participants (Asher and Kendrick), as prompting strategies were not needed during the observed sessions.

Caregiver Ratings

Caregivers were asked to rate at baseline how much of a problem feeding was for their child, with 1 being “feeding is always a problem” to 10 being “it is never a problem.” For participants in the 12-week program, the average caregiver rating of feeding as a problem was 2.67 at baseline and 5.83 at the completion of the program. This was a change of 3.17 in their rating from the start to the end of the program. For participants in the 8-week program, the average caregiver rating of feeding as a problem was 2.00 at baseline and 6.00 at the completion of the program, with a change of 4.00 across the program. Caregivers were also asked to rate their confidence addressing their child’s feeding problems with 1 being “not at all confident” and 10 being “completely confident.” The average caregiver rating on confidence of participants in the 12-week program was 5.50 at baseline and 7.83 at the completion of the program, which is a change of 2.33. For the 8-week program, the average caregiver rating on confidence was 2.00 at baseline and 5.00 at the completion of treatment, with a change of 3.00.

Social Validity

On the social validity form developed for the pilot program, all caregivers across both the 12- and 8-week program rated the program highly. They answered that they agreed to strongly agreed with all statements (average response 4.75–5). See Table 5 for the average responses to each of the social validity statements. All caregivers strongly agreed (rating of 5) that the staff made practical recommendations that could be followed in session and at home, they would continue to use the strategies learned during the feeding program after it was completed, and that they were overall satisfied with the feeding program.

Caregivers were also given the opportunity to provide comments on the social validity form. On the open-ended question requesting suggestions for improvement of the feeding program, three caregivers provided responses. The first response was: “I’m so happy with the program. I’m not sure there is much to suggest differently.” The second response was: “More videos of him eating, so we can show him eating since he is motivated by videos.” The third response was: “None at this time.”

In response to the open-ended question about how caregivers felt their child progressed during their time in the feeding program, seven caregivers provided responses. One caregiver wrote: “My son was only eating three to five snack foods and was eliminating even the small group of foods even more. Refusing to try new things or even sit at the table. He would gag watching others eat and run away. He is now eating at the kitchen table and able to tolerate others eating. He is open to trying new foods and is adding new foods to his diet daily. He comes to the table easily and is not distressed anymore. It’s unbelievable and so amazing! So happy with this program!” The second comment from a caregiver was: “My son has made incredible progress. He went from being anxious just being around food to actually taking bites. He mastered all the goals we set for him at the beginning of the program.” The third caregiver response was: “The variety of foods she will eat and her technique (eating off a spoon and drinking out of a cup) have greatly improved!” The fourth caregiver response was: “He is eating more food and choices between a mixture of foods.” The fifth caregiver comment was: “He wasn’t biting into any food at all, now he bites into cereal and graham crackers. He can eat a whole pudding cup. He sits when eating.” The sixth caregiver response was: “Him being open to trying new foods and his sitting have improved. They gave us (parents) the skills, so he knows what to expect when trying new foods.” The final caregiver response was: “He’s eating at home now and sitting.”

Discussion

For practitioners working with children with feeding disorders, the evaluation of this pilot program provides an example of how to develop a clinical program that includes core evidence-based treatment components but allows for individualization of treatment to meet specific needs of each child. This pilot program is unique in several ways. First, it incorporated the expertise of both psychologists and occupational therapists. Collaboration between multiple disciplines is often ideal to address feeding concerns as these challenges can be a result of medical, oral motor, and behavior difficulties (Piazza, 2008). By combining evidence-based behavioral interventions with occupational therapy skill building techniques related to oral, fine motor, and adaptive skills, the participants of this pilot program were successfully able to meet many of their feeding program goals. Previous research has found that behavioral feeding interventions that incorporate two or more disciplines have been effective in day treatment settings (Seiverling et al., 2020). The current evaluation of a bi-weekly interdisciplinary outpatient feeding intervention found that the combination of both psychology and occupational therapy services within this outpatient program was beneficial. Having both a psychologist and occupational therapist allowed for the program to address a wide range of feeding concerns including increasing bite acceptance, food variety, drinking from an open cup or straw, self-feeding, chewing, and advancing food texture. It also allowed the clinical resources to provide comprehensive caregiver training, which we hypothesize may improve sustainability of new feeding skills overtime.

Regarding practical considerations in the medical setting, this intervention allowed for individualization of treatment protocols to address the specific concerns of each participant in a manner that was feasible for caregivers to implement in the home setting. While each protocol was individualized, specific evidence-based techniques were used across all participants and allowed for a treatment fidelity form to be developed and used for this program. Previous research on feeding interventions has found that few studies measured treatment fidelity (Tereshko et al., 2021), which was evaluated during the pilot program.

Participant Outcomes

The first aim of this pilot program evaluation was to evaluate if changes in feeding behavior of participants in the program were observed including (1) percentage of bites accepted in clinic and at home; (2) rate of inappropriate mealtime behaviors in clinic and at home; (3) percentage of individualized treatment goals met for participants; and (4) reported number of foods consumed by caregivers. Less frequent appointments paired with caregiver training allowed for improvement in feeding behavior in a relatively short amount of time. Evidence-based strategies were utilized, including antecedent strategies and differential reinforcement. For the eight participants who completed this program, these strategies were successful at meeting most of their individualized treatment goals, decreasing inappropriate mealtime behaviors, and increasing bite acceptance. As a pilot study, the results should be evaluated with caution. Future research should evaluate the significance of outcomes using single-case research designs and/or group designs. The outcomes of this pilot program evaluation are consistent with previous research suggesting that antecedent and differential reinforcement are effective interventions to address the feeding concerns (Tereshko et al., 2021).

Caregiver Training

The second aim of this pilot program evaluation was to evaluate if the program was (1) feasible for caregivers to implement based on a treatment fidelity checklist; (2) if caregivers reported a change in level of concern related to their child’s feeding behavior and their confidence to address their child’s feeding behaviors; and (3) acceptability of the treatment program based on a feedback survey. Caregiver involvement and training were the focus of this program and an essential component of most feeding interventions (Allen & Warzak, 2000). First, individual feeding goals were developed for each participant through collaboration with caregivers, as has previously been recommended in research (Ledford et al., 2018).

Further, understanding not only the effectiveness of feeding interventions, but also their social validity to caregivers is essential. Previous research indicates that social validity is not often addressed in feeding intervention research (Tereshko et al., 2021). The current pilot program evaluated social validity based on caregiver feedback and found that caregivers were both satisfied with the program and the outcomes. High caregiver treatment fidelity in the pilot program is aligned with previous research findings that caregivers can implement feeding interventions with fidelity when given appropriate training (Johnson et al., 2019). To measure acceptability of the program, the pilot program evaluation aimed to determine how caregivers rated their child’s feeding concerns and confidence in addressing those feeding concerns at post-treatment, and whether caregivers rated the pilot feeding program as effective and valuable. Caregivers in both program lengths reported decreased concern about feeding, increased confidence in addressing feeding concerns, and satisfaction with the program and outcomes for their child.

Treatment Dosage

The final aim was to determine whether there was a difference in treatment outcomes based on the length of the program (12 vs 8 weeks). Although the participants who completed the 8-week treatment program showed improvements in feeding behaviors, high caregiver treatment fidelity, and high caregiver satisfaction, during the 8-week program more than 50% of families who started this program were unable to complete it in the 8 weeks due to unexpected absences and challenges. Although this is a high attrition rate, it is uncertain whether this was due to the specific circumstances of the COVID-19 pandemic (which was not occurring during the 12-week pilot) or if it was more challenging to complete the minimum of 10 appointments required to “complete” the program during the 8-week duration, compared to the 12 weeks.

As a result of this pilot project, the interdisciplinary outpatient feeding treatment program continues to be implemented as a clinical program. Based on the two different durations piloted, it was decided that the duration of the program would be 10 weeks. This duration was selected as most participants in the 12-week program had met their goals before the end of the program, indicating a shorter program length may be appropriate, but during the 8-week program, it was difficult for many families to attend even 10 total appointments which made it difficult to meet goals. Additional research about the appropriate length of program should be conducted.

Generalization

An important goal of behavior change is often success in the natural environment; however, this is not often easily assessed in feeding research conducted in clinical settings (Ledford et al., 2018; Tereshko et al., 2021). Therefore, a main goal of the feeding program was to observe improvements in feeding behaviors in the home setting in addition to the clinical setting. Different methods were used to observe mealtime at home including having the family videotape at home and/or telemedicine appointments that occurred from home. Telemedicine offered one method for practitioners in this clinical setting to address generalization to natural settings, which can be used to effectively to address feeding concerns of children (Bloomfield et al., 2019, 2021). When this was not feasible (as was not available during the 12-week pilot of this program), video recordings of feeding sessions at home were another way for practitioners to evaluate generalization to natural settings. All changes observed using direct observation measures in the clinic were also observed in the home setting. Generalization across settings was possible due to the strong focus on caregiver training.

Limitations and Future Directions

This was an evaluation of a pilot interdisciplinary outpatient feeding treatment program that exists in a hospital medical center. As it was conducted as clinical programming, there were several limitations that are important when considering results. First, as this was a pilot study the participant numbers were small, especially for the 8-week group. Therefore, outcomes should be interpreted with caution and further research is needed to determine the significance of treatment effects. Additionally, the current program only included two disciplines. This pilot study combined the treatment of psychologists and occupational therapists to address the behavioral and oral motor concerns related to feeding for children with developmental disabilities, primarily autism spectrum disorder. For many children, other disciplines are also needed in their care. Future research on outpatient services should examine models for incorporating other disciplines (e.g., medical professionals, speech and language pathologists, nutritionists) into these services.

The current program was limited in training other caregivers, such as teachers or school staff. Strategies for school were discussed, including sending to school only foods that had already been added to the child’s diet at home, using already established visual supports in the school setting, and caregivers modeling feeding strategies with school staff members. Future studies should evaluate methods for generalizing feeding treatments to school settings, if possible, as most children eat at least one time during the school day, and eating in the school setting can have implications for behavior and learning in that setting. Additionally, as this was a clinical program, inter-observer agreement was limited based on staff availability. As the program could only run for the 12 and 8 weeks, they were scheduled based on staff availability and, therefore, when a family missed an appointment, it could not be rescheduled. Although each participant in the two durations of the program were offered the same length of time in the program, they attended a different number of appointments due to family availability and circumstances.

Another limitation is that this program evaluation only included pre- and post-intervention data. Therefore, changes in feeding behavior, because of time only were not assessed. The current research did not include data collection on the sustainability of each participant’s newly attained feeding skills overtime. Future research should use experimental designs that control for maturation, such as multiple-baseline or waitlist control designs like the Johnson et al. 2019 study, and design that can include follow-up at 3-, 6-, or 12-month intervals to measure for sustainability of skills.

Another limitation of the current evaluation was that a formal functional analysis was not conducted. Future research should formally assess the function of inappropriate mealtime behaviors that interfere with food consumption to better address the function of feeding concerns. Future research should include measures of social validity for participants, in addition to caregivers, whenever possible. Additionally, the current study did not collect demographic information such as family socioeconomic status or family structure. This information would be valuable to collect in future research on feeding outpatient services.

The pilot program was conducted both before and during the COVID-19 pandemic and, therefore different procedures were used at different times. The participants in the 12-week program all participated prior to the start of the pandemic. Telehealth appointments were not approved at the hospital medical center where the program was conducted at that time and, therefore, video recording was used to evaluate feeding behaviors at home. Participants in the 8-week program participated during the pandemic when telehealth appointments were approved at the hospital medical center and behaviors at home could be assessed in real time through telehealth appointments. This difference between groups should be considered when reviewing the intervention outcomes. Finally, the current evaluation did not include any data from participants who started but did not complete the program. While reasons for attrition included illness and challenges related to the COVID-19 pandemic, it is also possible that poor progress in the program was a contributing factor to attrition and should be evaluated in future research.

Summary

The current evaluation provides outcomes from an interdisciplinary outpatient feeding treatment program for children with developmental disabilities that included two disciplines. This pilot program evaluation illustrates how goals related to feeding can be connected and collaborated across the disciplines of psychology and occupational therapy. Additionally, the pilot program utilized primarily antecedent and differential reinforcement-based procedures, providing further evidence to the growing literature that these strategies are effective at improving feeding behaviors (Tereshko et al., 2021). The pilot feeding program resulted in improvements in feeding behaviors for the eight children who completed the program. The current program also evaluated both caregiver treatment fidelity and social validity based on caregiver feedback, finding that caregivers were able to implement evidence-based strategies and were satisfied with the treatment and outcomes for their child. Finally, the pilot program conducted in a medical setting focused on ensuring improvements in feeding behaviors generalized to the home setting. This was possible through telemedicine, and before that was available, video recordings. Although this program has several limitations related primarily to constraints of the medical setting in which it was conducted and that it was implemented as a clinical program, the current evaluation is an example of how evidence-based treatments can be implemented to address feeding concerns of children with developmental disabilities in an outpatient setting with relatively few appointments.