Reporting treatment processes and outcomes for paediatric feeding disorders: A current view of the literature

ABSTRACT

Empirically supported treatment for paediatric feeding disorders is behaviour-analytic, often taking place in controlled hospital settings. We evaluated reporting of treatment processes and outcomes within behaviour-analytic studies to consider recommendations for practitioners, as well as future research. Strengths of the literature were reporting of directly observed behavioural outcome data demonstrating the effectiveness of empirically supported treatment. However, there was variability in reporting aspects of the overall treatment process. With respect to multidisciplinary involvement, gaps in reporting included the frequency of involvement in assessment and treatment, as well as the roles of each discipline involved. Further, there were gaps in reporting of treatment goal progression, such as food variety, volume, texture and independence, as well as nutritional outcomes (oral intake, growth). Our findings support recommendations to improve reporting in the literature, to increase standardisation of processes and improved dissemination to the wider research community. Further, improved research reporting will increase practitioner knowledge of empirically supported treatment for paediatric feeding disorders. To this end, we provide recommendations and resources regarding multidisciplinary involvement and systematic monitoring of treatment goal progression.

KEYWORDS:

• Treatment progress monitoring
• competency
• paediatric feeding disorders
• avoidant/restrictive food intake disorder (ARFID)
• tube weaning
• inappropriate mealtime behaviour
• multi/interdisciplinary team

Paediatric feeding disorders are highly prevalent (Kovacic et al., 2021). There is an extensive body of literature summarising the causes, classification and treatment of paediatric feeding disorders (more recently including diagnostic terminology of avoidant/restrictive food intake disorder [ARFID]). Given the risks of prolonged feeding problems to child health and development, timely evaluation and access to empirically supported treatment (EST) is vital (Taylor & Taylor, 2021). The EST for paediatric feeding problems is behaviour-analytic, with 50 years of evidence (Kerwin, 1999). Access to specialised professional training and intensive treatment for severe cases is limited to a few locations in the United States where this research literature originated (Babbitt et al., 1994). These locations are behaviour-analytic hospital/medical school feeding programmes equipped with therapy and observation rooms, providing intensive treatment (3–5 meals/day, 8 weeks) conducted by a team of highly trained therapists under behaviour-analytic supervision (e.g., Milnes & Piazza, 2013). Training opportunities for behaviour analysts and psychologists are provided in these settings through predoctoral internships and postdoctoral fellowships. Due to barriers to training and treatment access, there may be a lack of understanding of the EST for feeding problems and translation from research to practice, especially outside of controlled hospital settings (Taylor & Taylor, 2021). Recently, behaviour analysts specialising in feeding have reported intensive empirically supported interventions translated to home and community settings, but more details are needed on the varied processes and methods employed (Patel et al., 2022; Taylor et al., 2019, 2020).

Multiple previous reviews have been conducted describing aspects of treatments for paediatric feeding problems, such as those covering intensive multidisciplinary programmes (Sharp et al., 2017), group design studies (Lukens & Silverman, 2014), treatments for tube dependency (Taylor et al., 2019), specific treatment procedures (e.g., blending; King & Burch, 2020; high-p sequence; King et al., 2019; physical guidance; Rubio et al., 2020; antecedent procedures; Seubert et al., 2014) and empirically supported treatments (Kerwin, 1999; Sharp et al., 2010; Volkert & Piazza, 2012). Recommendations for the literature have included clarifying the contributions of disciplines, standardising nutritional or behavioural outcomes and better describing participant characteristics. However, because of breadth of literature, there has not been a detailed review of treatment processes and outcomes such as intensity, team qualifications and roles, treatment goals, replicable descriptions of treatment procedures, full progression from admission to follow-up or comprehensive outcomes. Reviewing such details could help to clarify the process of treatment, determine timelines for treatment effectiveness, support research replications and inform the field about the essential components of treatment. This could help bridge the gap between the empirical literature and practice, for instance, in practitioners’ interpretation and use of the research, as well as that of researchers outside of this specialty area.

The purpose of this paper was to survey the reporting of treatment processes and outcomes within the behaviour-analytic treatment literature for paediatric feeding problems. Outcomes were evaluated in terms of measurement of behaviour and nutrition, including generalisation and maintenance of effects. Further variables included the multidisciplinary involvement, duration of treatment and time to effectiveness. Results inform recommendations for improving reporting in the literature, guidelines for multidisciplinary involvement, and expected goals and progression of treatment.

Method

We conducted a literature search on 15 March 2020 to identify English-language, peer-reviewed treatment articles for paediatric feeding problems (birth to 17 years) published between 2009 and 2020. Keywords searched in PsycInfo and Medline databases included intervention or treatment and fussy eating, picky eating, tube dependency, food refusal, pediatric feeding disorder, avoidant/restrictive food intake disorder (ARFID) and food selectivity. We screened abstracts to identify behaviour-analytic interventions, involving the manipulation of environmental antecedents or consequences. While not exhaustive, examples included escape or attention extinction, differential or noncontingent reinforcement and fading. Articles were removed at this stage if the abstract specified that the intervention was non-behaviour analytic (e.g., family-based therapy, psychotropic medication, hypnosis). If the intervention could not be determined as behaviour-analytic from the abstract (e.g., did not specify procedures, experimental design or stated “behavioural intervention”), the first author then screened the article.

During full-text review, we included single-subject design studies if they evaluated treatment effects within a single-subject experimental design and displayed directly observed measures of mealtime behaviour. Similarly, we included group design studies (randomised and nonrandomised) if they reported the use of behaviour-analytic treatment procedures and reported statistics in relation to directly observed mealtime behaviours. Thus, we excluded publications such as letters to the editor, challenging cases, case reports (without an experimental design) or studies that did not include child outcome data (e.g., parent training data only). We also removed studies if the primary treatment focused on nonfeeding behaviour, such as interventions for challenging behaviour, and translational studies where the purpose was to examine specific behaviours or variables in simulated conditions. We coded each article in terms of setting, participants and key intervention components. Coding definitions can be accessed from the Supplementary Information file. A second coder (second author) also screened abstracts or articles obtained from the initial search to apply inclusion/exclusion criteria and interobserver agreement was 100%. The second author also coded items, applying the coding definitions in 31% of the included articles and agreement was mean 98.2% (range, 94.4% to 100%). The authors resolved disagreement on included articles and items by consensus, prior to the final summary of data.

Results

The search yielded an overall list of 136 unique articles, of which 68 behaviour-analytic studies met inclusion criteria. Figure 1 shows the search process and categories of articles excluded. Almost all studies used single-case designs (96%), with the remaining studies being a randomised controlled trial (RCT (Sharp et al., 2017)) or analysis of retrospective group data (Laud et al., 2009; Taylor et al., 2017). Two studies were both RCT and single-case experimental designs (Peterson et al., 2016, 2019).

Figure 1. Search process.

Population

The mean age of participants was 5.1 years (range 0.5 to 17). Only 27% of studies (n = 18) reported participants without known medical or developmental conditions (i.e., typical development). Studies reported the following conditions: medical (n = 38, 56%), feeding tube (n = 30, 44%), autism spectrum disorder (n = 32, 47%), developmental disabilities (n = 25, 37%) and ARFID (n = 1, 2%). Only 22% (n = 15) of studies reported the duration of the child’s feeding problem at admission or noted that the child had received prior treatment attempts. Studies from intensive behaviour-analytic hospital/medical school programmes were more likely to report prior treatments from other disciplines (Bachmeyer et al., 2019; Sharp et al., 2010; Taylor et al., 2017) and participants with tube feeding, medical conditions and developmental disabilities. Between and within studies, feeding problems varied widely in range, from participants without medical and skill problems already consuming regular texture food (e.g., Ewry & Fryling, 2016; Whelan & Penrod, 2019) to some who had never consumed solids or liquids orally in their lives fully dependent on tube feeding with severe medical, developmental and skill problems (e.g., Taylor et al., 2017; Wilkins et al., 2011).

Components of the literature

In Figure 2, we display specific components we assessed in the literature, pertaining to 1) programme characteristics, 2) goals and outcomes and 3) caregiver implementation and generalisation.

Figure 2. Surveyed components of the literature.

University settings were coded under “clinic”. ID = interdisciplinary, defined as the involvement of at least one other discipline during evaluation or treatment.

Programme characteristics

Figure 2 displays programme characteristics, including location, setting, intensity and involvement of multiple disciplines. The majority of studies were from the United States (97%), with 52% (n = 25) from behaviour-analytic hospital/medical school feeding programmes. Only 28% (n = 19) of studies were conducted in homes or schools (72% [n = 49] were other hospital, clinic or university settings). Some studies reported multiple settings (e.g., inpatient then home, daypatient then telehealth outpatient). Telehealth was reported in three studies, of which two were further interventions following initial intensive treatment (Peterson et al., 2015; Volkert et al., 2014). In terms of treatment intensity, 82% of studies reported more frequent than weekly services. Intensive programmes were frequent (inpatient 3% [n = 2]; day treatment 38% [n = 26]), with a typical duration of 8 weeks. Less intensive services were infrequently reported (10%, n = 7), and five studies were from prior intensive programmes. Thirty-two percent of studies reported overall duration of treatment (M = 51 days, range 5–177). Studies reported longer durations to teach chewing and achieve regular texture (e.g., years of outpatient follow-up, Volkert et al., 2014).

Feeding problems have multiple contributing factors (e.g., medical, skill, environmental), and prior to treatment, evaluation and clearance by multiple disciplines is warranted (Taylor & Taylor, 2021). In our literature search, 49% (n = 33) articles reported having at least one other discipline involved in evaluation, and some reported their consultative roles (e.g., medical and nutritional monitoring, speech pathologists conducting swallow studies). A few studies highlighted a lack of disciplinary involvement, such as not obtaining an oral motor evaluation prior to treatment of packing or not obtaining medical clearance. During treatment, interdisciplinary involvement was rarely reported, in only 13% (n = 9) of studies. Overall, the mean number of involved disciplines was 1.2 (range, 0 to 6), with the most frequent disciplines being physician (28%), speech language therapist (27%) and dietitian/nutritionist (19%). A few studies noted involvement of occupational therapists in selecting seating, adaptive mealtime equipment and supporting fine motor skills related to meals (e.g., Sharp et al., 2015). Two studies noted that a social worker provided emotional support and resource planning to families (e.g., Laud et al., 2009; Sharp et al., 2017). We present relevant cited examples of disciplinary roles in Table 1. It is important to note that our review may underestimate multidisciplinary team involvement, as behaviour-analytic hospital/medical school programmes include evaluation or ongoing consultation as the clinical standard (e.g., Babbitt et al., 1994; Milnes & Piazza, 2013).

Table 1. Interdisciplinary roles and expertise.

Physician/Medical* (Gastroenterologist, Immunologist/Allergist) (see Yeung et al., 2015)

Evaluate medical stability for treatment and provide clearance or required further evaluation prior to treatment. Expertise required in testing, treatment and monitoring for: Timing of endoscopy coordinated with reflux medications and elimination diets/food exposure Anatomical throat and swallowing problems (Laud et al., 2009; Vaz et al., 2012; Wilkins et al., 2011) Emptying, motility and absorption problems and conditions (Vaz et al., 2012) Vomiting and reflux (Levin et al., 2014; Wilkins et al., 2011) Food allergy, eosinophilic conditions (including repeat testing schedules) (Alaimo et al., 2018; Gonzalez et al., 2013) Nutritional deficiencies (e.g., scurvy, pellagra, rickets, xerophthalmia), malnutrition, dehydration and dental health Constipation, diarrhea, bowel management regimens, hydration, diet (Peterson et al., 2015) Medications and any side effects (e.g., appetite, nausea) impacting feeding Conditions such as H. pylori, stomach ulcers, gastritis, gastroenteritis (Laud et al., 2009)

Dietitian

Evaluate and recommend intervention for nutritional deficiencies (Laud et al., 2009; Peterson et al., 2019) and conditions such as malnutrition and dehydration Recommend appropriate food types to use in treatment (Addison et al., 2012; Bachmeyer et al., 2019; Seiverling et al., 2018; Sharp et al., 2010, 2012) After food allergy testing, advise foods/liquids allowed or eliminated and train caregivers on label reading. For children who have not consumed solids or with food allergy or intolerance, inform plan for novel food introduction including order of foods to introduce, amount and time frame (Rivas et al., 2010; Sharp et al., 2017). Recommend formula types and adjustments for food allergies, food intolerances and eosinophilic conditions. Adjust recommendations with required repeat testing and reintroduction schedules/amounts. Coordinate and adjust dietary recommendations during treatment for blood sugar conditions, celiac disease, phenylketonuria, reflux, vomiting, constipation, diarrhea and emptying, motility, and absorption problems and conditions. Provide recommendations (e.g., types, schedule) for medication and supplement refusal and side effects impacting feeding; impact of liquid (e.g., milk, formula, baby bottle) dependence on dental health, constipation and absorption (e.g., iron) Monitor growth (Williams et al., 2019) Adjust tube calories and fluids as oral intake increases (Sharp et al., 2017; Taylor et al., 2017)

Speech Pathologist

Conduct swallow studies (prior to/during treatment), advise related recommendations (Sharp et al., 2010, 2015; Wilkins et al., 2011) If swallow study preparation is required, recommend solid textures and liquid consistencies and volume to reach criteria (Girolami et al., 2007) Provide recommendations/advise testing or treatments for anatomical oral motor and swallowing conditions (Sharp et al., 2010) Recommend consistencies for liquids, e.g., swallowing safety (Wilkins et al., 2011), textures or bolus sizes (Bachmeyer et al., 2009; Kadey et al., 2013; Seiverling et al., 2018; Sharp et al., 2012; Wilkins et al., 2014) Recommend modifications to reduce aspiration risk (Bachmeyer et al., 2013; Sharp et al., 2017; Vaz et al., 2012) Evaluate oral motor skills, e.g., chewing (Alaimo et al., 2018; Peterson et al., 2016; Vaz et al., 2011); therapy alongside behaviour-analytic sessions (Laud et al., 2009; Taylor et al., 2017)

Behaviour Analyst

Conduct a comprehensive thorough intake evaluation (including all areas listed above) Record review from other disciplines, tests and treatment recommendations, comprehensive review of systems* Caregiver forms and interview Direct observation and assessment of skills and mealtimes Determine when above evaluation has cleared the child for appropriate and safe treatment Determine if child may need treatment in a specialised behaviour-analytic hospital/medical school setting such as for intensive medical oversight, for severe problem behaviour or need for timely clinical outcome (e.g., prevent G-tube placement Rubio et al., 2015) Determine if the case and circumstances are in line with the behaviour-analyst’s level of competence or require referral. Seek supervision and consultation as needed. Determine the individualised treatment intensity and resources needed (e.g., adaptive and safety equipment, seating, staff, funding) Determine consultation involvement needed from other disciplines before and during treatment. Monitor for symptoms requiring referral to other disciplines. Determine support needed for caregivers before, during and after treatment. Provide the support and/or refer to others. Determine what foods, textures and bolus sizes to start treatment with, and progress to, in what timeframe Perform systematic, data-based, individualised treatment evaluations and progress monitoring Prioritise goals over time (e.g., based on nutrition, hydration and growth; when to target certain textures or feeding/drinking skills) Perform intensive caregiver training, generalisation and maintenance (Alaimo et al., 2018; Gonzalez et al., 2013; Seiverling et al., 2012)

Note: This table refers to examples identified in the literature search but is not exhaustive. The reader may wish to refer to further additional literature that details more standard interdisciplinary roles and processes (Babbitt et al., 1994; Cohen et al., 2006; González et al., 2018; Ibañez et al., 2020; Kozlowski et al., 2015; Manikam & Perman, 2000; Milnes & Piazza, 2013; Patel, 2013; Piazza et al., 2020; Sharp et al., 2017; Silverman, 2010; Yeung et al., 2015). *A physician’s clearance checklist form is provided in Supplementary Information.

As we selected behaviour-analytic studies, treatment sessions were primarily behaviour-analytic apart from some exceptions (9%, n = 6). These exceptions included studies where speech and/or occupational therapists conducted additional sessions (Laud et al., 2009; Taylor et al., 2017) or where the study had a specific aim to compare ABA with other therapies (Addison et al., 2012; Peterson et al., 2016; Seiverling et al., 2018).

The level of involvement and qualification of the behaviour analyst was often unclear. In 79% of studies [n = 54], “trained therapists” implemented sessions, but the level of behaviour-analyst supervision and programme implementation was not reported. Some studies indicated the qualification level of the behaviour analyst (n = 14; 21%), typically a master’s or doctoral degree and BCBA or psychologist licensure. However, no study provided details regarding years of training and/or experience in paediatric feeding. Feeders were typically individuals with a minimum of a bachelor’s degree, but no details were provided regarding years of experience.

Goals and outcomes

A substantial strength in the literature, all studies reported increased appropriate mealtime behaviour (e.g., acceptance, swallowing/consumption) and/or decreased inappropriate mealtime behaviour (e.g., refusal, negative vocalisations, expulsion, packing), and 99% of studies reported interobserver agreement in at least 20% of sessions. We found an overall lack in reporting of overall admission goals and broader outcomes. Four studies reported the number of treatment goals and percentage of goals met (Borrero et al., 2013; Taylor et al., 2017, 2019; Volkert et al., 2014). Goals typically included the acceptance and consumption of food or drink, low inappropriate mealtime behaviour, to train caregivers, and to generalise progress to other settings or people. Again, our review may underestimate the literature, particularly from behaviour-analytic feeding hospitals/medical schools with established clinical standards for the full range of goals such as skill, independence, texture, volume and cup drinking.

Figure 2 (middle panel) relates to treatment goals: depicting variety, volume, texture, and skill processes and outcomes. A key nutritional measure in paediatric feeding is oral intake (food or liquid grams/mL consumed), but this was only reported in 51% of studies. In 56% of studies, sessions were reported to include all food group items. Seventy-four percent of studies reported the number of foods used, but some studies tended to report using foods across food groups in sessions (e.g., 8–16 foods) rather than the total variety achieved. A few studies listed foods with no observation or definition of consumption (e.g., foods introduced or approached), thus were not counted as variety. For volume, studies reported bolus sizes and bite number more frequently than total volume or calories. For texture, studies reported more detail for the range of lower textures (e.g., puree, wet ground, fork mashed) compared to regular texture (e.g., mashables, hard). For skills, studies did not report the full range of skills and most reported only self-feeding without details. Studies reported increasing self-feeding (n = 11), chewing (n = 4) and cup drinking (n = 5). Only 10% (n = 7) of studies reported weight outcomes with all reporting weight maintenance or gain.

Caregiver implementation and generalisation

Figure 2 (bottom panel) depicts reporting of caregiver training, generalisation and maintenance. Treatment integrity was reported in 47% of studies, including the fidelity of therapist or caregiver implementation. One-third of studies reported using behavioural skills training (BST), but others lacked detail of the training process, including whether caregivers were required to meet a criterion. Studies reported generalisation across caregivers and different foods but less frequently with regard to generalisation across settings or stimuli (e.g., school, community, seating or cutlery).

Only 38% of studies reported follow-up, generally including directly observed data (31% of overall studies). Follow-up durations varied from as brief as 1 week, to notable examples of 12 months across multiple participants (Wilkins et al., 2014). In a few studies, follow-up included further treatment for ongoing skill development (e.g., outpatient sessions up to 3 years; Volkert et al., 2014).

Only 29% of studies reported social validity, which included caregiver treatment satisfaction with outcomes or the acceptability of procedures. Measures included standardised scales (e.g., Intervention Rating Profile; Witt & Martens, 1983) or those developed by authors (e.g., Hoch et al., 1994; Rubio et al., 2020; Ulloa et al., 2019). Knox et al. (2012) conducted intervention within a special school and surveyed teachers. Two studies reported social validity in a more informal manner, such as qualitative information provided by caregivers, or providing caregivers with the opportunity to choose between procedures (Shalev et al., 2018; Tarbox et al., 2010).

Treatment procedures and time to effect

Most studies (n = 50, 74%) specifically reported the use of treatment targeting escape. In conjunction with reinforcement and antecedent treatment packages, studies reported the following escape extinction treatments: nonremoval (n = 45, 66%) (i.e., Hoch et al., 1994; Kerwin et al., 1995), physical guidance (n = 18, 27%) and other (e.g., exit criterion; n = 8, 12%). The large majority of studies using physical guidance (89%, n = 16 out of 18) and nonremoval procedures (71%, n = 32 out of 45) were conducted in behaviour-analytic hospital/medical school feeding programmes. Some studies did not implement escape extinction procedures with the child for the treatment evaluation but had implemented it prior (e.g., for nonself-drinking, Peterson et al., 2015) or on a model (Fu et al., 2015; O’Connor et al., 2020). The description of escape extinction procedures varied across studies (e.g., escape prevention, prolonged presentation, contingency contacting, hand-over-hand assistance) and sometimes lacked procedural details for replication (e.g., spoon distance or placement at lips, exit criterion contingencies).

Time to effect may be defined as the frequency of meal sessions, time or days to reach clinically significant levels of consumption (i.e., clean mouth over 80%) of varied foods across food groups. It was difficult to assess time to effect in this review, due to the gaps in reporting of goals, outcomes and treatment durations. However, studies using escape extinction and high intensity typically showed rapid increases in food consumption in just a few sessions (e.g., Addison et al., 2012; Borrero et al., 2013; LaRue et al., 2011; Woods & Borrero, 2019) and tended to report the use of all food groups (72%) and increasing volume.

Some escape extinction studies discussed the time to effect in relation to other procedures. For example, Peterson et al. (2016) reported effective behaviour-analytic treatment in an average of 73 minutes (range 15–249) compared to ineffective modified sequential oral sensory (SOS) therapy for an average of 1,111 minutes (range 1020–1292). Rivas et al. (2010) reported that escape extinction alone was faster than when it was combined with spoon distance fading (6 versus 93 sessions), and Seiverling et al. (2014) showed antecedent manipulations such as modelling by the feeder to be ineffective (approximately 25 sessions) until a bite requirement was added (initially latency of 54.5 min for one target bite).

Compared to rapid effects with escape extinction, other procedures required longer time, or were less effective, especially when implemented at lower intensity. Other procedures included modelling, simultaneous presentation, reinforcement, shaping and the high-probability sequence (e.g., Bloomfield et al., 2019; Ewry & Fryling, 2016; Penrod et al., 2010; Vandalen & Penrod, 2010). Constraints on the use of escape extinction were not consistently reported but included parental discomfort with extinction (Bloomfield et al., 2019) and implementation difficulties in practice settings (Ewry & Fryling, 2016).

Summary

This survey of the behaviour-analytic literature further asserts the effectiveness of treatment for paediatric feeding disorders, as demonstrated by directly observed behaviour change, data reliability and experimental control. From the information obtained, studies involving high intensity and function-based empirically supported procedures (e.g., escape extinction) reported the shortest time to effect and achieved higher behavioural and nutritional outcomes.

Improvements are needed in reporting of processes and outcomes in the literature, and this will influence future evaluations of treatments. This is not an entirely new finding, as previous reviews have recommended increased documentation of nutritional outcomes, evaluating the contributions of other disciplines, and describing participant characteristics. The current paper extends these reviews by surveying very specific components of treatment processes such as disciplinary involvement and qualifications and outcomes such as goals, skill progression and generalisation. Our attempts to define some components, such as time to effect, were difficult due to variability in reporting outcomes and durations of treatment, but future reviews may address this. Similarly, for variables such as volume, texture, skills and generalisation, we included any reporting of these components without strict criteria for comprehensive information or data.

Recommendations for reporting and practice

We provide recommendations for reporting in the literature in Table 2. In the behaviour-analytic field, it is recognised that intensive programmes have standardised processes with respect to intensity, disciplinary roles, staff training and supervision, procedures, data reliability and treatment goals. However, many articles do not specify these standards, which affect the ability of reviewers to assess the literature (Lukens & Silverman, 2014). It is understood that journal page limits may be a barrier to reporting. Additionally, it is unlikely that the full treatment experience would be reported as studies may be focused on a specific research hypothesis or evaluation of more novel or complex components. Thus, online supplementary material could be a useful strategy or the reference to certain standardised procedures (e.g., programme descriptions in chapters). However, feeding problems are heterogeneous, and processes must be sensitive to individual factors. For example, treatment outcomes for tube dependence may have a larger focus on growth and volume compared to treatment of food selectivity focused on increasing variety.

Table 2. Recommendations for the literature.

Area	Recommendation
Participants	Observable description of feeding problems before treatment including full range of outcome measures (nutritional intake, growth, bolus, variety, volume, texture, skills, independence) Specific duration of feeding problem and supplements (formula, tube) Prescribed medications or supplements (formula, tube) Type and intake of foods and liquids Other feeding-related impacts on the child and family Type and duration of previous treatments and effects Other: socioeconomic status, cultural considerations, adaptive functioning, outside of meal behaviour, ongoing therapies
Study Quality	Design meets either single-case or group quality criteria Demonstrate experimental control Interobserver agreement (IOA) and procedural integrity (with IOA) Report intention to treat data Report effect size
Procedures	Report disciplinary involvement roles in intake, consultation and direct sessions, and qualification level (i.e., type and years of feeding training) of behaviour analyst Report duration of treatment evaluation, admission length and intensity of services Describe all procedures in a replicable manner, specifically escape extinction variations Report entire course of treatment, not just analyses of specific procedure
Outcomes	Define treatment goals and time to effect (consumption across food groups) Include directly observed behaviour change Include nutritional intake and permanent product data: grams of consumption, images of items consumed (e.g., meal plate pictures), list of food group varieties, liquids, supplements and medications consumed; oral and tube calories, volume, fluid and percentage of needs Boluses, variety, volume (per session, meal, day), texture (including range for regular texture), skills, independence (self-scooping, utensils [e.g., finger, spoon, fork], self-drinking from open cup, biting off portions) Growth: Parameters and percentiles for weight for age, height for age and proportion, pre- and post- treatment and change Caregiver training, integrity data and generalisation (including settings outside of home, nonadaptive equipment and seating) Report social validity measure data for the full range of procedures before, during and after treatment
Follow-up	Observational follow-up data Follow-up at a minimum of 6 months from treatment, consideration for long-term follow-up (1 to 2 years or more)

Multidisciplinary involvement and roles

The description of involved professionals requires further clarification in future studies. First of all, intake processes could clearly outline roles and the specific evaluations that were conducted to provide clearance for treatment. Full reporting would inform practitioners about the complexity of feeding problems, and the extent to which interdisciplinary evaluation may be required (Williams & Seiverling, 2022). An interdisciplinary team with expertise in EST for paediatric feeding should conduct a full evaluation for relevant medical and physical variables, for example, food allergies, swallowing, dental health (before chewing is targeted) and bowel management. To this end, we describe disciplinary roles and specific paediatric feeding expertise required in detail in Table 1 and provide a sample medical clearance form (Supplementary Information). Detail in this table has been further supported by original literature as well as additional overviews on more standard team roles (see footnote), given the lack of reporting in treatment studies.

At this point, it is important that we consider behaviour analyst competency alongside interdisciplinary roles (Williams & Seiverling, 2022). Competency and qualification level are important to report in the literature with regard to implementation of required procedures, for varied children (Rubio et al., 2020). Some authors indicate that the treatment of severe cases and implementation of required procedures requires a doctoral-level behaviour analyst, with several years of advanced training within a specialised behaviour-analytic hospital/medical school feeding programme (Rubio et al., 2020; Taylor et al., 2020). Behaviour analyst competency level is also important in considering the involvement of other disciplines in conjunction with case severity. The behaviour analyst is not expected to have the equivalent knowledge of other disciplines but needs sufficient feeding expertise to navigate the requirement for certain evaluations and make appropriate referrals (e.g., swallow study, allergy testing; Piazza et al., 2020; Williams & Seiverling, 2022; Yeung et al., 2015). Behaviour analyst competency in multidisciplinary issues becomes even more critical in less controlled settings without expert teams.

During evaluation, it is vital that the behaviour analyst and interdisciplinary team are also able to identify when treatment needs to take place within a behaviour-analytic hospital/medical school programme setting. This may include children that have issues with breathing, swallowing, severe malnutrition, severe food allergies, vomiting or blood sugar concerns. Older children and those who engage in severe problem behaviour (e.g., aggression, self-injury) may also require additional oversight (Bachmeyer-Lee et al., 2023; Briggs & Greer, 2021).

Once it is determined to proceed to treatment, further multidisciplinary involvement is outlined in Table 1. Based on the literature, it is evident that interdisciplinary involvement during treatment is often consultative only. Examples may include guidance with regard to liquid consistency, food type or texture, or adjustments to tube feeds. Actual treatment sessions conducted by other disciplines are less commonly reported but require further detail for interpretation of effect. Specifically, the number of sessions, specific procedures and equivalent analysis of outcome data with experimental control and reliability.

In treatment, the behaviour analyst should have the necessary competency to implement all required procedures; otherwise, referral to a specialised treatment provider is required (Rubio et al., 2020). Linked to the surveyed research, behaviour analysts should continue to use a least restrictive and timely approach in determining the progression to escape extinction procedures, include components that may reduce the likelihood of a burst in inappropriate mealtime behaviour and incorporate social validity (Knight et al., 2019; Phipps et al., 2022; Piazza et al., 2020; Taylor & Taylor, 2022; Woods & Borrero, 2019).

Process

Behaviour-analytic studies could improve upon fully reporting the entire course of treatment as well as outcomes beyond directly observed behaviour change. Full outcome reporting would again support reviewers of other disciplines to assess the evidence. Research designs such as the consecutive controlled case series (CCCS) should be implemented (Hagopian, 2020; Taylor et al., 2020) and recognised by reviewers. Further research could also involve standardising a definition of time to effect (including both total therapy hours and calendar time) as well as the timeframe required to reach certain stages (e.g., to achieve initial effect, to observe generalisation to targets and caregivers). Visual displays of data could include days noted on the x-axis alongside sessions.

Aligning with the scientist-practitioner model, it is important that practice in the area of paediatric feeding disorders is strongly informed by the literature. However, the current paucity in reporting may make it difficult for practitioners to understand the full process of intervention. Thus, to support practice and future research reporting, we provide a guide for expected treatment progression in Table 3. The child should be consuming new foods from all food groups within a few weeks of treatment (unless limited by allergy or texture [i.e., chewing skill]). There must also be a process and plan for caregiver training, achieving independence (self-feeding and cup drinking), regular texture and eliminating dependence on formula or the feeding tube. We recognise that treatment goals and progression must be individualised and may vary with the severity and topography of the child’s feeding problem (e.g., total refusal versus selectivity, chewing skills). Following assessment, a practitioner must be able to competently determine the most appropriate starting points, procedures and progression for the child.

Table 3. Progress monitoring and goals for intensive treatment.

Food items: Novel nutritional food group items, not including snack/junk foods.
Example standard goals
(applicable to all goal areas that follow e.g., variety, liquids, independence, texture, volume, generalisation)
1. Accept within 10 s (entire bolus past the plane of the lips)
2. Swallow within 30–45 s (mouth clean; no food/liquid more than size of a pea left in the mouth and not due to expulsion/spitting out)
3. Decrease inappropriate mealtime behaviour (e.g., turning head 45 degrees, covering mouth, batting at the utensil/feeder) to <1 per minute
4. Decrease negative vocalisations (e.g., crying, whining, negative statements) to <20% of the meal
5. Train caregivers to implement protocol with procedural integrity (<1 incorrect per minute or >85% correct) (Alaimo et al., 2018; Seiverling et al., 2012)
Achieve in less than 3 months:
Solids
6. Increase variety to 16 new foods (4 from each food group: protein, starch, vegetable, fruit) (Bachmeyer et al., 2013; Peterson et al., 2019; Pizzo et al., 2012; Taylor et al., 2019; Wilkins et al., 2011). Continue to further increase variety once this goal is met.
7. Teach utensil use (e.g., closing lips on spoon) for 75% of bite presentations
8. Teach self-feeding independence to 75% of bite presentations (Rivas et al., 2014; Rubio et al., 2018; Vaz et al., 2011)
Liquids
1. Teach open cup drinking to 75% of drink presentations (closing lips on cup) (Shalev et al., 2018)
2. Increase liquid variety to water and formula/milk (if applicable)
3. Teach self-drinking independence to 75% of open cup presentations (Peterson et al., 2015)
Medication
1.Increase acceptance and swallowing of medication/laxatives/supplements
Individual factors (e.g., chewing skills, volume tolerance, growth, medical/health status, food allergies, swallowing) will determine progression of goals below (achieve in 6 months to a year):
Texture (Volkert et al., 2013, 2014)
1. Assess texture and individualise goals in order of difficulty and teach chewing:
a. Lower textures (Bachmeyer et al., 2013)
i. Naturally lower texture foods (e.g., yoghurt, pureed initial baby foods)
ii. Increase thickness/consistency to table purees
iii. Increase texture (junior, wet ground, fork mashed)
b. Toddler meals/soft saucy combos; soft regular (e.g., mac & cheese, baked beans)
c. Dissolvables
i. Infant/toddler products with increased ages (e.g., rusks, veggie straws, puffs, biscuits)
ii. Easy crackers/chips and dry cereals
iii. Harder (e.g., pretzel, crustini, breadsticks, vitawheat)
d. Dissolvable combinations with spreads/mashables
e. Soft, mashable solid bites (e.g., avocado, banana, cooked carrot, potato, pumpkin, strawberry, kiwi, zucchini, eggplant)
f. Regular texture bites of more difficult textures and foods:
i. Soft (e.g., soft ripe fruits, tofu, soft cooked vegetables, quiche, eggs, muffins, soft cheeses, soft/tender meats [nuggets, meatballs, deli slices])
ii. Regular (e.g., melons, cucumber, capsicum, pear, water chestnut, raw zucchini, blanched green bean)
iii. Hard (e.g., raw fruit [apple] and veg [carrot, celery, snow pea], tougher meats)
g. Mixed regular texture scoop (e.g., yogurt with fruit/granola, uneven chunks in thin soup)
2. Increase texture ratio (amount of lower and regular texture in meals) to age-appropriate
Volume
1. Increase solid bolus (spoon size for lower texture; bite size for regular texture)
2. Increase liquid bolus
3. Increase oral solid volume (Laud et al., 2009; Taylor et al., 2017)
4. Increase oral liquid volume (Peterson et al., 2015)
5. Increase volume to 100% solids orally (Bachmeyer et al., 2013; Volkert et al., 2014)
6. Increase volume to 100% liquids orally (water and formula/milk if applicable)
7. 100% medications/laxatives/supplements orally independently
Independence
1. Full plate meals (rather than single-bite presentations)
2. Teach self-scooping
3. Teaching biting off portions (instead of pre-cut-up bites)
4. Teach full open cup sipping
5. Change from adaptive to typical age-appropriate equipment (e.g., spoon, cup)
6. Eliminate baby bottle use (Alaimo et al., 2018)
7. Evaluate independent access procedures prior to prompting sequences
8. Teach fork use
Generalisation
1. Intersperse drinks and other targets into one meal
2. Generalise protocol (e.g., to natural age-appropriate seating, settings [school, community, childcare, relatives], utensils, plates/bowls, cups, water bottles, family meals) (Alaimo et al., 2018; Gonzalez et al., 2013; Taylor et al., 2019)
3. Simplify protocol; fade reinforcement and prompting

Note: This table refers to examples identified in this literature search, but is not exhaustive. A further example is also available in Milnes and Piazza (2013, p. 402; Fig 22.7).

Translation from research to practice

We recognise the challenges in translating EST for paediatric feeding to practice (Taylor & Taylor, 2021, 2021), given the paucity of specialised behaviour-analytic training programmes. Additionally, given the critical and complex nature of paediatric feeding disorders, extensive training is required (i.e., years) for competency. There may be limited recognition of feeding disorders and limited funding or health insurance coverage for treatment (Jones et al., 2020; Raatz et al., 2021). There may be the absence of a multidisciplinary team with the necessary expertise for feeding evaluation. Specialised behaviour analysts in paediatric feeding have recently begun reporting the translation of intensive EST to home and community settings, but careful considerations of the processes and outcomes reviewed here are warranted (e.g., competency, resources, appropriate intensity, case severity/needs). Coordination with behaviour-analytic paediatric feeding programmes could inform better practice guidelines to foster disciplinary involvement and earlier access to treatment and referral (Forbes & Grover, 2015; Gardiner et al., 2014). These guidelines may outline minimum competencies of professionals, requirements for training and consultation, and support the training and recruitment of behaviour analysts with specific expertise. We hope the resources provided here have utility in promoting improved dissemination and translation of EST for paediatric feeding disorders, professional training, multidisciplinary evaluation clearance and consultation, early intervention and referral, timely progress monitoring, and ultimately patient outcomes.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Supplemental data

Supplemental data for this article can be accessed online at https://doi.org/10.1080/15021149.2023.2259219

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.