Treatment of Liquid and Medication Refusal in an Adult With Intellectual Disability,Epilepsy,and Challenging Behavior

Abstract

The effectiveness of a behavioral intervention was evaluated in an adult with intellectual disability, epilepsy, and challenging behavior who had ceased oral intake of liquids and medications. The intervention consisted of fading, designed to gradually increase the volume of water accepted; backward chaining, involving bringing a cup to the lips, introducing liquid into the mouth, and ultimately swallowing; and differential reinforcement. A descriptive functional assessment revealed that the most common consequences of intake refusal were access to preferred conversation topics from parents and hospital staff and escape. Baseline measures showed little or no acceptance of liquids or medications. Following fading and differential reinforcement, swallowing was established and increased to 90 ml of water, while medication intake reached 100% success. Treatment gains were maintained in both clinic and home settings. The study replicated the findings of Hagopian et al. (1996) and extended their generality to an adult client and the oral intake of medication. Social validity scores were favorable for both parents.

Theoretical and Research Basis for Treatment

Recent applied behavior-analytic intervention research with adolescents and adults with autism spectrum disorder (ASD) and related neurodevelopmental disorders has demonstrated the effectiveness of reinforcement-based approaches for improving food selectivity without the use of escape extinction procedures (Vanderzell et al., 2025). These findings are consistent with a broader shift within applied behavior analysis toward treatment models that prioritize positive reinforcement, differential reinforcement, and gradual shaping procedures, while also attending to issues of autonomy, assent, and treatment acceptability in clinical practice.

Despite this emerging literature, intervention research specifically targeting liquid and medication refusal in adults with neurodevelopmental disorders remains limited. Feeding and related intake difficulties appear to persist across the lifespan in this population, with evidence suggesting comparable prevalence in adults and children with neurodevelopmental disorders (Gravestock, 2000; Thomessen et al., 1991). Liquid and medication refusal in particular may result in significant health risks, including life-threatening consequences or substantial functional impairment (Taylor, 2023).

Given the complexity and severity of such difficulties within this population, an interdisciplinary approach to treatment is important to ensure client safety and best interests (Taylor, 2023; Tereshko et al., 2021). Prior to conducting a behavioral assessment and intervention for refusal of oral intake, it is essential to ensure that a comprehensive medical evaluation has been performed and that any organic factors contributing to the problem have been addressed or ruled out (Behavior Analyst Certification Board, 2020; Tereshko et al., 2021). For individuals who refuse liquids or medication, the implementation of a surgically implanted feeding tube may be necessary to sustain adequate hydration and health (Krom et al., 2017). Although life-saving, prolonged reliance on enteral intake may be associated with tube dependency and a subsequent reluctance to resume oral intake (Krom et al., 2017). Tube dependency may reduce quality of life by increasing caregiving demands, logistical challenges, and family stress (Nelson et al., 2015). Consequently, an important treatment objective for some individuals may be the restoration of oral intake.

When intake refusal is conceptualized as a behavioral problem, best-practice guidelines recommend evidence-based assessment procedures, such as functional assessment, to identify maintaining reinforcement contingencies (Borrero et al., 2010; Piazza et al., 2003; Reitman & Passeri, 2008). In the pediatric feeding literature, food and liquid refusal have been shown to be reinforced by escape from meals, access to caregiver attention, or access to preferred items, and the identification of these maintaining contingencies has been associated with improved intervention outcomes (Piazza et al., 2003; Saini et al., 2019). Similarly, medication refusal has been documented in a clinical case to be maintained by caregiver attention and escape from the medication administration context (Reitman & Passeri, 2008). Taylor (2023) evaluated a behavioral treatment package targeting medication refusal in two children with ASD. The treatment package included escape extinction (via re-presentation and/or non-removal of the spoon), differential reinforcement, finger prompting, and side-deposit administration protocols. The treatment rapidly increased medication acceptance across multiple forms, flavors, and delivery methods. Both participants met the established treatment goals, and caregivers were trained to implement procedures and maintain treatment gains in the home environment. However, the medications used in the study were limited to chewable, drinkable, and flavored forms. Furthermore, although the treatment components demonstrated effectiveness, several of the escape extinction procedures may be physically intrusive. As a result, the direct application of these procedures may be less feasible or acceptable in adult populations. Other applied studies have examined liquid refusal and difficulties with oral medication intake as separate target behaviors. Furthermore, no empirical research has successfully addressed both targets concurrently in either children or adults, despite the fact that some individuals exhibit refusal of both liquids and oral medication. As pill intake usually requires concurrent oral intake of liquid, the restoration of both target behaviors benefits individuals with neurodevelopmental disorders.

Three studies were identified on behavioral interventions for individuals with neurodevelopmental disorders that have difficulties or refusal to swallow pills. In all studies, fading combined with differential reinforcement was found to be the main treatment components (Babbitt et al., 1991; Beck et al., 2005; Ghuman et al., 2004). Babbitt et al. (1991) taught four children with moderate to profound intellectual disabilities to swallow pills by fading gradually increasing sizes of mock pills (e.g., cake decorations, candy, and capsules filled with cornstarch). The intervention also included differential reinforcement and individualized modifications, such as timeout, use of a syringe, a tongue depressor, and capsule type. This treatment resulted in all children mastering oral intake in the clinic, at home, and at a 1-year follow-up. In a pilot study, Ghuman et al. (2004) used several behavioral components with four children with ASD. These included modeling, prompting, visual aids, fading in increasing sizes of mock pills, and differential reinforcement of acceptance and swallowing. The results showed that two out of four children were able to swallow capsules with their caregivers after 3–5 visits. Beck et al. (2005) utilized a similar behavioral protocol in eight children with ASD and attention deficit hyperactivity disorder (ADHD), resulting in six of the eight children swallowing medication in several settings. In the previously described clinical case study by Reitman and Passeri (2008), an 8-year-old boy with ADHD exhibited refusal to swallow pills. The results of a functional assessment directly informed the development of the intervention. A fading procedure combined with differential reinforcement was implemented, resulting in the successful consumption of tablets after 12 treatment sessions. One limitation of the current evidence base is that only Babbitt et al. demonstrated experimental control using a single-case methodology. Additionally, there is limited information in the published literature on how researchers have individualized treatment procedures, making replication difficult. A related issue concerns the differentiation of pill swallowing as a skill never learned versus a skill that has been lost or is part of a recent onset of intake refusal. For example, the fading of small edibles may not be a viable option for individuals who refuse oral intake. Additionally, some individuals may present with a long-standing aversion to the taste or texture of mock pills. Such issues may increase treatment resistance (for a review of intervention studies on pill swallowing in children with normal intellectual functioning, see Patel et al., 2015).

The behavioral treatment literature on liquid refusal has also applied fading with differential reinforcement as common components, similar to those of medication refusal; however, escape extinction has more commonly been utilized. Smith et al. (2019) found that escape extinction (nonremoval of the cup) in combination with noncontingent access to music was more effective in the treatment of liquid refusal than escape extinction alone for a 4-year-old boy with ASD. Roth et al. (2010) used stimulus fading, reinforcement, and escape extinction to treat food and liquid refusal in a 16-year-old adolescent male with Asperger’s syndrome. After 14 days of treatment, the adolescent expanded his diet from 3 foods to 78 foods and 13 beverages, stopped tube feeding, and maintained these improvements at home and school. Patel et al. (2001) applied stimulus fading, combined with differential reinforcement, and escape extinction. The participant was a 6-year-old boy with a pervasive developmental disorder who was already drinking water. The fading involved mixing water with a nutritional drink mix, and finally milk, which resulted in the child drinking undiluted milk with a nutritional mix. Luiselli et al. (2005) also utilized fading where one liquid (milk) was gradually mixed into another liquid (Pediasure®) combined with differential reinforcement to establish the consumption of milk in a 6-year-old girl with ASD, which resulted in her consuming undiluted milk. In this case, escape extinction was not utilized. Kozlowski et al. (2016) developed a liquid preference assessment that could help treat three children with ASD who refused most liquids. It found that incorporating preferred liquids increased acceptance and consumption in a treatment package consisting of differential reinforcement or a combination of differential reinforcement and extinction (nonremoval of the cup). In these studies, participants consumed some liquids prior to treatment. Three empirical studies document the behavioral intervention of a total refusal of all liquids. Babbitt et al. (2001) presented two cases, a 2-year-old boy with severe developmental delays and an 8-year-old girl with Down syndrome and severe intellectual disability, who both had total liquid refusal. Their treatment consisted of initially fading liquid from a spoon, combined with positive feedback for acceptance and escape extinction and then transferring fading from the spoon to a cup, resulting in successful acceptance of liquid from a cup for both children. Groff et al. (2014) described the case of a 4-year-old neurotypical boy who exhibited refusal of both liquids and solids. A functional analysis indicated that his liquid refusal was maintained by escape, thereby informing the selection of escape extinction as the primary intervention. Because the child engaged in teeth clenching upon presentation of liquids, escape extinction was implemented using a syringe-fading procedure, beginning with 0.1 ml increments, followed by a gradual transition from syringe to cup. This intervention resulted in successful consumption of both liquids and solids, with the authors noting that the child began accepting liquids via syringe after only a few presentations. Hence, an alternative interpretation of these results is that syringe fading may increase acceptance and make escape extinction unnecessary. Schiff et al. (2011) presented a case of a 3-year-old boy with ASD who avoided liquid medication. During treatment, a syringe was used to fade in water in ml increments, increasing concentrations of placebo liquid, then an actual medication, and then finally fading from syringe to a cup. The syringe-to-cup fading protocol successfully resulted in liquid consumption and finally liquid medication intake. Another case study was conducted by Hagopian et al. (1996). Here, a 12-year-old boy with ASD and a moderate intellectual disability who exhibited total refusal of food and liquid intake underwent a behavioral intervention comprised of backward chaining, fading, and differential positive reinforcement. Initially swallowing by instruction was differentially reinforced. Water was then incrementally faded using a syringe in steps of 0.2 ml, 0.5 ml, 1 ml, and 3 ml. Then liquid was faded from a cup, as soon as the boy accepted water in 20 ml probes from a cup; then fading commenced in 10 ml increments and ending in 90 ml. Furthermore, each successful trial resulted in an activity reinforcer, specifically 90 s of access to clipping paper. This intervention resulted in the boy consuming 90 ml of water and juice, which was generalized to his living environment. However, significant avoidant behavior was observed during the syringe fading phase, occurring in 20%–100% of the trials, which diminished once the boy began drinking from a cup.

In sensitive adult individuals with total liquid refusal, it may be feasible to implement small fading steps using a syringe when practical. For instance, a 1-ml syringe allows for fading in ten 0.1-ml increments (0.1–1.0 mL), potentially enhancing acceptance by clients and caregivers. In addition to empirical support, the combination of fading and positive differential reinforcement may have an intuitive appeal and social acceptability when addressing liquid and medication refusal in adults with neurodevelopmental disorders. Furthermore, research indicates that stimulus fading may render escape extinction unnecessary (Luiselli et al., 2005; Volkert et al., 2016) and that positive reinforcement may be an effective and preferred treatment component for escape-maintained problem behavior (DeLeon et al., 2001; Lalli et al., 1999).

Recognizing the empirical support and potential advantages of integrating fading designed to gradually increase the volume of water accepted with differential positive reinforcement, we sought to replicate the work of Hagopian et al. (1996) and extend the procedure to encompass medication intake in an adult with moderate intellectual disability, problem behavior, and epilepsy. The original procedure of Hagopian et al. (1996) was also modified to reduce the likelihood of problem behaviors. This modification involved the use of initially smaller fading steps and an open-door policy (Rajaraman et al., 2021), allowing the client to exit the treatment room at all times.

Case Introduction

Jon, a Caucasian male in his twenties, was diagnosed with moderate intellectual disability and epilepsy. Jon was generally described as being social and communicative, cooperative, and able to participate in scheduled activities. At the time of the study, he resided with his parents and received 10 hr of user-controlled home personal assistance weekly. Additionally, he participated in a daycare program 5 days per week, each lasting 6 hr.

Presenting Complaints

Jon was referred to a neurohabilitation unit due to an increase in problem behavior, consisting of screaming, hitting, and property destruction 3 months prior to his total intake refusal. This was accompanied by an increased frequency of nocturnal epileptic seizures and insomnia. Parents reported that Jon had severe bouts of anxiety and/or confusion, mostly during the evening or when he woke up at night.

He was admitted to a gastrointestinal surgery department due to transient abdominal pain, elevated C-reactive protein (CRP), and liver enzymes, but the evaluation/investigation did not reveal a specific cause. He was then admitted to a neurology ward unit due to medication refusal, severe undernourishment, underweight, and increased epileptic activity, where he received a nutritional emulsion (Kabiven), liquid and medication intravenously. There his CRP and liver values normalized.

Previously, Jon had been able to eat independently, consume liquids, and successfully mastered self-administering up to six medications orally (capsules and tablets) on his own. Upon admission, Jon refused liquids, food, and medications. Food refusal was initially identified as the primary target of assessment and intervention. However, during the course of treatment for liquid refusal, Jon began consuming solid food with his parents. Consequently, the initial assessment included refusal of food in addition to liquids and medications. In agreement with his parents, the subsequent treatment focused on liquid and medication refusal for the remainder of the study.

History

Jon had experienced severe epilepsy since childhood, which had resulted in multiple hospital admissions. The epilepsy was managed with several antiepileptic medications. His seizures frequently occurred during the nighttime and could result in partial or complete unconsciousness, as well as secondary generalized tonic–clonic seizures (GTCS). Following seizures, he was reported to exhibit reduced appetite and liquid intake, although this typically improved after 1–2 hr. He could also display signs of confusion and wandering behavior before sleep or during the nighttime. Nevertheless, medical treatment had resulted in a considerable reduction in seizure frequency approximately 2 years prior to the onset of the current problem.

A medical assessment did not identify any possible somatic causes for the intake refusal, such as metabolic disorders or swallowing difficulties. Historically, Jon had taken tablets and capsules orally without difficulty. However, he was occasionally observed to complain about pain in his abdomen and back. No medical cause was identified for these complaints.

Jon was reported to exhibit challenging behavior, involving vocal aggression, defined as screaming or making verbal threats of physical harm toward another person. It could also include property destruction, defined as damaging or attempting to damage objects in the environment, including breaking furniture. This behavior occurred exclusively in the home environment and mostly during transitions, for example when transitioning from home to the day program. Prior to the current intake refusal, Jon was not reported to have had any long-standing difficulties with the intake of food, liquids, or medication.

Assessment

The behavior analyst (first author) conducted four visits to the hospital ward, each lasting between one and one and a half hours. These visits were used to operationalize the target behavior, conduct a descriptive assessment similar to Borrero et al. (2010), and establish a naturalistic baseline. The behavior analyst also probed and verified that Jon could open his mouth on instruction and/or modeling. In addition, preferred stimuli and activities were identified through an interview with parents, similar to a caregiver-report procedure described by Fisher et al. (1996).

During the descriptive assessment, antecedents and consequences were recorded through direct observation. What was presented and parent or staff responses (attention, escape, tangible or none) were coded if they occurred immediately and no later than 10 s after a refusal or acceptance response. The behavior analyst did not alter any environmental conditions during the assessment. The resulting data were analyzed to identify the consequences most commonly associated with refusal (Borrero et al., 2010).

In common scenarios, Jon was either in his private hospital room or a common area in the ward unit equipped with a table, chairs, and sofas. Observations showed that nursing staff and parents frequently attempted to have Jon drink liquid, take medication, or consume food. Drinks, medications, and food were continuously available and visible, except when Jon moved around. When he was seated, staff presented a glass of liquid under his nose or placed medication in a cup in front of him. They also attempted to place food into his mouth. When presented with these, he would typically stare at, look away, push the items away, or walk away from them. When medication was held near his lips, he consistently avoided oral intake by keeping his mouth closed, clenching his teeth, or turning his head away. He imitated persuasive statements of nursing staff and parents; for example, “it is important to eat and drink!”. Additionally, parents reported that Jon seemed to request specific conversation topics, such as vehicles, fire stations, cartoon characters, and hospital staff, during intake attempts and other interactions. No aggressive behavior was observed during the descriptive assessment or reported by the parents or hospital staff during the hospital stay.

After multiple days and several attempts per day by the hospital staff and parents to help feed Jon, a Percutaneous Endoscopic Gastrostomy (PEG) was performed. Parents reported that Jon had a negative reaction to PEG, displayed symptoms of pain around the area of the feeding tube; he also displayed resistance to the administration of liquid, food, and medication via the PEG.

Case Conceptualization

The behavior analyst’s descriptive assessment indicated that the most common consequence of intake refusal was (a) receiving attention from parents and hospital staff, as previously noted, his preferred type of attention was certain conversation topics and (b) avoiding or escaping servings. During interviews, parents reported signs of stomach aches, both through Jon’s utterances and body language (holding his stomach and uttering “au, au”). Medical examinations did not reveal any medical causes; however, doctors noted that pain could stem from long periods of food abstinence.

The interview indicated that Jon’s most preferred activities included solving puzzles, watching various short videos, and playing a firefighter memory game on his iPad©. The descriptive assessment and preference assessment informed the design of the treatment plan. Preferred conversation topics and activity reinforcers were used as reinforcing consequences during treatment. Further, the behavioral treatment plan included chaining, fading, and differential reinforcement. In addition, the treatment plan was sensitive to the fact that Jon might experience pain during fading of liquid. Therefore, treatment progressed in small steps as described below. A finalized treatment plan was reviewed by an assigned registered clinical nutritionist and neurologist. The behavior analyst presented the treatment plan to Jon’s parents, who read the plan and provided written informed consent on behalf of Jon.

Course of Treatment and Assessment of Progress

Measurement

Target behaviors were operationalized and coded similar to Hagopian et al. (1996), as follows: Acceptance was defined as a drink or pill (capsule or tablet; hereafter pill) surpassing the lips. The six pills comprised five different medications, five used in epilepsy management: one Orfiril Long 300 mg capsule (24.0 × 7.0 mm), two Lyrica 150 mg capsules (17.0 × 6.0 mm), two Fenytoin 100 mg oblong tablets (11.6 × 6.5 mm each), and one round multivitamin tablet measuring 6 mm in diameter. Swallowing was defined as an observable and noticeable movement of the larynx that occurred within 10 s of instruction of liquid/pill acceptance. Owing to the subtleness of Jon’s swallowing (in visits 32–50) this target behavior was verified by placing two fingers gently over Jon’s larynx to detect larynx movement. During the pill intake sessions, Jon was instructed to open his mouth after swallowing to verify the absence of pills.

Expulsion was defined as the ejection of whole or partial liquid, or pill from the mouth. Avoidance/denial was recorded if Jon turned his head or body away from the presented liquid or pill, clenched his teeth or lips, pushed the target items (glass, medicine cup) away, or attempted to block or push the therapist’s arm. As avoidance/denial and expulsion closely followed the absence of acceptance and to facilitate visual analysis, these data were excluded from the final analysis (data available from the first author).

Target behaviors, amount consumed, and fading target were recorded on a customized paper-recording sheet on a trial-by-trial basis. The behavior analyst also documented any other events deemed relevant to the treatment. Swallowing without liquid and successful pill intake trials were plotted as a percentage of successful trials. The liquid consumption was plotted as milliliters consumed during each fading target. If Jon swallowed all the liquid in each fading target, it was recorded as 100% correct and converted to milliliters consumed.

The behavior analyst served as the primary observer and collected data during all the baseline and treatment sessions. Two social educator students, with training in behavior analysis, collected trial data independent of the behavior analyst during 13% of the treatment sessions based on the availability of each observer. Interobserver agreement (IOA) was assessed by comparing agreement and disagreement for each recording code on a trial-by-trial basis, totaling the agreement, dividing by trial count, and multiplying by 100. IOA was 96% on average, with a range of 67%–100%.

Baseline

The baseline assessment was conducted in two distinct settings: (1) during the four visits to the neurology hospital ward, which constituted a naturalistic baseline for the refusal of intake (including food and liquids such as water or juice) and medication, and (2) baseline probes conducted in a treatment room, involving the administration of 20 ml of water in a cup, as detailed below. These probes were administered once by the parents prior to treatment and intermittently by the behavior analyst during the treatment phase.

Treatment

The treatment and probe sessions were conducted in a treatment room measuring 3.5 m by 4.7 m. During these sessions, Jon and the behavior analyst were seated at a table measuring 0.8 m by 1.4 m. The treatment room was equipped with windows, curtains, a desk, two chairs and a computer. In accordance with Jon’s preferences and corroborated by parental reports, curtains remained open. Given Jon’s history of challenging behavior, several precautionary measures were implemented, similar to those described by Wiskirchen et al. (2017). Jon was consistently seated at the short end of the table, near the door, to maintain an adequate distance from the behavior analyst. Padding was applied to the corners of all tables. Additionally, during the initial four visits, prior to each session, the behavior analyst accompanied Jon to familiarize him with the doors available to exit both the treatment room and the clinic, instructing him to use them if desired, in accordance with an “open-door policy” (see Rajaraman et al., 2021). Jon was also reminded that his parents were present in the waiting room. The doors of the treatment room remained closed and unlocked during all sessions. It was planned that if problem behavior occurred more than once during a session, the session was terminated, and the parents assumed responsibility. However, this did not occur.

Parents were instructed to minimize verbal attention on preferred conversation topics and to withhold preferred activities for as long as possible (establishing operations), ideally for a minimum of 2 hr prior to each treatment session. The treatment involved backward chaining, fading, and differential reinforcement. The treatment targeted the establishment of two related behavior chains: Behavior chain 1 was hypothesized to involve bringing the cup to his lips, applying liquid through his lips, and, as the final step, swallowing water (Hagopian et al., 1996). Backward chaining was utilized to establish this behavior chain. Behavior chain 2 involved introducing pills to his mouth, bringing the cup to his lips, applying liquid through his lips, and then swallowing the pills with water as the final step. Behavior chain 2 was established with whole task presentation (Cooper et al., 2020), once behavior chain 1 was mastered with 50 ml of water. Thus, relying on the previously mastered steps from chain 1, Jon was required to complete all the steps in chain 2 to receive reinforcement.

Backward chaining for behavior chain 1 involved attempting to reinforce swallowing when an instruction to swallow was issued. If Jon did not comply, a model prompt was provided at the same time as the instruction. If modeling and instruction proved ineffective, the behavior analyst placed two fingers gently on his larynx while continuing to provide instructions for approximately 10 s up to three times.

As a second link in the chain, accepting water was attempted, with fading in water gradually. During the initial stage, an empty syringe was placed in Jon’s mouth, requiring him to close his lips around the tip of the syringe and swallow. From this point, no prompting was employed. The second stage mirrored the first stage, with the exception that the behavior analyst first immersed the syringe in a cup of water. Subsequent stages involved the same procedure, with a gradual increase in water volume, starting at 0.1 ml and incrementally increasing to 0.6 ml in 0.1 ml steps. This was followed by a single 0.4 ml increase to 1 ml, and finally a 0.5 ml increase to 1.5 ml. Thereafter, a probe for 20 ml was introduced with a cup. Transitioning from syringe to cup occurred after Jon successfully swallowed 20 ml from the probe cup. Fading then continued in 10 ml increments until he successfully consumed 50 ml from a cup. At this stage, medication fading was introduced, and the water volume was increased in 10 ml increments to 90 ml, where it was maintained for the remainder of treatment. The medication consisted of six pills. Initially, all pills were introduced, as Jon had mastered this quantity. Modifications to the medication intake criteria were made following several unsuccessful trials during sessions 80–86. The reinforcement criterion was adjusted to one pill per trial, subsequently increasing to 2–3 pills per trial from session 89 onwards. This modification reflected a demand-fading approach in which response requirements were initially reduced following difficulty with responding and then systematically increased as responding stabilized, consistent with procedures described by Piazza et al. (1996).

Positive reinforcement was contingent on accepting and swallowing water or pills. Reinforcement consisted of approximately 2 min of access to Jon’s preferred activities as well as attention to preferred topics from the behavior analyst. The reinforcement period ended when a stage in the game ended or when Jon looked up from the screen. If acceptance did not occur or if an insufficient quantity was swallowed, reinforcement was withheld. No attention was given to verbal utterances on preferred topics without consumption. However, verbal utterances outside of the preferred topics received attention and were reciprocated regardless of performance, such as commenting on objects in the environment or requesting a toilet break. Escape extinction was not utilized. In the event of non-acceptance, a new trial was presented for up to three times. If Jon did not accept after the third trial, the treatment session ended, and during the medication intake sessions, the parents administered the medication through the PEG. The criterion for introducing a new step was the acceptance and swallowing of the total volume in two consecutive sessions up to the target amount of 50 ml. Subsequently, acceptance and swallowing in a single session were required for sessions involving 60 ml or more from a cup.

The criterion for introducing a new step was based on performance across three treatment phases: (a) swallowing without water, (b) water-only intake, and (c) water combined with pills, depending on the phase of treatment. For (a), swallowing without water, progression required 100% successful swallowing within a session, with mastery defined as two consecutive sessions with 100% success. For (b), water-only phases required swallowing the full volume in two consecutive sessions up to 50 ml and in a single session for 60 ml or more from a cup. For (c), pill phases required 100% successful acceptance and swallowing of the full prescribed pill dose within a session without refusal or expulsion.

The treatment was administered three to four times per week, with one to three sessions per visit, for a total of 38 visits. Each visit ranged from 10 min to 1 hr and 10 min in duration, and as the treatment progressed, both session time and the number of trials gradually decreased. Each session involving swallowing without water or training with a syringe, comprised 7–10 trials. Sessions involving a cup consisted of 1–5 trials.

The extension and generalization of the treatment were conducted as follows: From the 72nd session onwards, the parents were requested to be present during the treatment sessions to observe the procedures. Initially, the parents were seated 1.5 m away from Jon and faced away from him. They were gradually moved closer to the behavior analyst. Eventually, they were seated next to him and faced Jon during sessions. Then, the behavior analyst sat alongside the parent and implemented treatment. Subsequently, the parent took over while the behavior analyst gradually distanced himself. In the event of an increase in failed trials, parents were asked to leave the treatment room for the remainder of the session and for a subsequent visit. Parents were reintroduced after Jon successfully completed two consecutive sessions, with 100% success.

Generalization to Jon’s home environment was conducted as follows: The parents continued to implement the treatment with the behavior analyst present, receiving coaching and feedback as necessary. The presence of the behavior analyst was gradually reduced in approximately 2–3 m increments until he was outside of Jon’s apartment. Follow-up session probes were scheduled at 7, 15, and 65 days. Throughout the treatment process, a registered dietician, psychiatrist, and neurologist maintained contact with Jon and his family members.

Experimental Design

The current study employed a single-case experimental design with repeated baseline probes and systematic treatment phase comparisons. Experimental control was established through reversal to baseline and by observing Jon’s consumption following the fading steps (either the amount of water or the combination of water and pills). The fading-in components adhered to a logic similar to a changing criterion design (Kazdin, 2021), as progressively larger doses of water or water combined with medication were introduced, with positive reinforcement contingent upon successful completion of the specified criterion.

Data were graphed following each session and visually inspected throughout both the baseline and treatment phases. Initially, visual inspection focused on comparing responding to baseline levels. As treatment progressed, ongoing inspection of the graphed data was used to monitor changes in performance and to guide data-based decisions at each step of the treatment process.

Results

As depicted in Figure 1, during the baseline phase in the neurology hospital ward, Jon demonstrated a 0% acceptance of medication and liquid intake. In the initial baseline probe conducted at the clinic, Jon did not accept the water. However, during the second baseline probe in session 10, Jon consumed 4 ml of water from a cup. Attempts to establish swallowing as the final link in the behavior chain were unsuccessful, with only 0%–33% of trials being successful during sessions 14–29. The third baseline probe in session 31 also resulted in 0 ml acceptance. Subsequently, two treatment sessions involving empty-syringe insertion achieved a 30% success rate. The fourth baseline probe during session 34 again indicated 0 ml acceptance of water from a cup.

Figure 1.

Correct swallowing trials and water intake (ml.; top graph) and correct pill swallowing (bottom graph)

When treatment targeted swallowing with an empty syringe, performance reached 100% success after four sessions. Swallowing 0.1 ml of water from a syringe was then successfully established, and Jon’s performance met the criterion across all steps from 0.1 ml to 0.6 ml of water, as well as during steps involving 1 and 1.5 ml.

During the fifth baseline probe session, Jon successfully consumed 20 ml from the cup. Thereafter, the treatment process continued with fading with the cup alone, in 10 ml incremental amounts of water, while mastering each fading step within the criteria range of 20–50 ml. The probe for water and medications showed that Jon still displayed a refusal to orally accept medication and drink water together, consuming 17 of the 20 ml of water and having 33% success in medication intake.

During the process of fading water and medication (lower graph), Jon initially achieved 100% success criteria for medication intake and mastered the intake of 90 ml of water during sessions 68–70. In session 71, one parent was introduced, which coincided with several unsuccessful trials and variable responding in both water and medication intake, with water consumption ranging from 45 to 73 ml and medication swallowing varying between 0% and 73%.

Following a modification from administering all medications per trial to one pill per trial, and after two consecutive successful sessions, medication swallowing increased to 100% and water intake to 90 ml. Subsequently, four sessions involved one or two pills per trial, and the fading step of three pills per trial was successfully implemented for the remainder of the treatment. The parents were then reintroduced, and Jon continued to master oral intake in their presence. Success was maintained at 100% for medication and 90 ml of water for the remainder of the treatment at the clinic as well as during follow-up in Jon’s home environment.

Complicating Factors

Jon never opted to leave the treatment room except for using the bathroom on a few occasions. Jon’s epilepsy, history of problem behavior, and other symptoms presented a challenge during treatment. His history of problem behavior necessitated the design of a safe treatment environment. As problem behavior only occurred at home, it was considered unlikely that it would occur during treatment at the clinic. Jon displayed neither problem behavior nor a precursor to problem behavior during treatment at the clinic, with the exception of one visit.

During this visit, with his father present, Jon displayed aggressive vocalizations in the waiting room and resisted going into the treatment room. The behavior analyst met him there and gave him time to calm down, and he finally joined the treatment. It is unclear why this occurred, but it occurred during a period in which he experienced an increase in his epileptic seizures at night.

As Jon could also have epileptic seizures during treatment, the behavior analyst received instructions from parents and neurologists on how to assist Jon. It involved helping him remain seated, preventing him from falling by standing behind his chair, holding him steady, holding his hands, and speaking to him in a soothing manner. At least one parent waited in the waiting room, which was located 26 m away from the treatment room. A neurologist was always available in the department if there was a need for medical assistance. Decisions to modify treatment were made on a session-by-session basis.

For example, Jon had at least five visits where his health became an issue: fatigue and concentration problems, which his parents reported as the result of nocturnal epileptic activity. In such cases, the sessions were shorter and fewer trials were conducted. During three sessions, Jon experienced a seizure, which necessitated a break of several minutes in treatment. In one session, Jon vomited while drinking the first dose of water. This session was terminated because of health issues, but Jon participated in the session the following day.

During the fading-in of the medication, the number of failed trials increased. For instance, on two occasions, sessions had to be discontinued due to the repeated expulsion of medications, necessitating the administration of the remaining medication via PEG. Consequently, the reinforcement criteria were modified from 2–3 pills per trial, to one pill per trial. Upon successful completion of this stage, the dosage was gradually increased back to 2–3 pills per trial in subsequent sessions, which was considered an acceptable dosage per trial.

Access and Barriers to Care

It is important to note that one or both of Jon’s parents accompanied him to the clinic and remained in the waiting room throughout the intervention period. The family reported that the requirement to transport and accompany Jon to treatment was somewhat burdensome and may therefore be a barrier to care for some families. One family member was on medical leave during this period and indicated that this was necessary to enable consistent participation in the treatment process.

The same clinician conducted all treatment sessions, which occurred three to four times per week over approximately 4 months. Accordingly, the intervention was resource-intensive for both the clinic and the family and may be considered somewhat vulnerable to disruption in the event that one or more key stakeholders were unavailable. Nevertheless, this level of investment was deemed necessary and was informed by prior research.

In addition, the hospital department involved in Jon’s care included physicians, registered dietitians and a Board Certified Behavior Analyst, professionals who may not be readily available in all clinical settings. Finally, the department in which Jon received treatment is based on a neurorehabilitation framework, which emphasizes the application of neurology and behavioral principles in the assessment and treatment of individuals with neurodevelopmental disorders. This evidence-based model may not be accessible across all clinical or hospital settings. These and other contextual barriers should therefore be considered when designing individualized treatment plans.

Follow-Up

Follow-up assessments were conducted at 7, 15, and 65 days post-treatment at Jon’s residence. During these direct observations, Jon consistently mastered his water intake and orally administered medication. His parents reported that Jon no longer required the administration of medication, liquids, or food via the PEG tube.

However, during the 15-day follow-up, they noted a period during which Jon experienced reduced consumption of drinks. This was resolved at the 65-day follow-up.

After the final follow-up assessment, the parents were asked to complete a social validity assessment comprising four questions: three questions utilizing a 7-point Likert-scale and one open-ended question. The three questions pertained to their satisfaction with the results, utility, and acceptability of treatment. The fourth question allowed additional open-ended written feedback. Parents rated the treatment highly (M = 6.8).

Two months after the final follow-up, a telephone conversation with Jon’s parents revealed that his PEG had been surgically removed, as confirmed by a video of Jon sent to the clinic by the parents. One year post-treatment, the parents reported that Jon consumed beverages independently and continued to take his medication orally.

Treatment Implications of the Case

Our results show that a behavioral intervention incorporating fading, chaining, and differential reinforcement can restore oral intake of liquid and medication in an adult with intellectual disability. The treatment initially involved making the last link in the behavior chain, swallowing, contingent on access to preferred conversation topics, and an activity reinforcer. However, the initial step of backward chaining, swallowing by instruction, was unsuccessful, whereas the subsequent step, involving the insertion of the syringe into Jon’s mouth, resulted in a gradual increase in fluid intake that aligned with the fading steps, ultimately leading to drinking water from a cup.

The second behavior chain was established during the final phase of treatment. It involved placing medication in the mouth, lifting the cup to the lips, accepting water into the mouth, and swallowing both the water and medication. These results extend Hagopian et al. (1996) by demonstrating the efficacy of this intervention in restoring not only water intake, but also oral medication intake. As noted by Hagopian et al., establishing the final response in the chain, in conjunction with fading, may be essential in a behavior chain involving drinking from a cup. This case study introduces an alternative to the initial step, specifically, establishing the first response with a syringe (thereby potentially evoking swallowing) and subsequently reinforcing it.

We expand upon Luiselli et al. (2005) and Hagopian et al. (1996) by treating liquid or medication refusal using positive reinforcement and fading without implementing escape extinction. Additionally, this study adds to the current evidence base of syringe fading in the treatment of liquid or medication refusal (Groff et al., 2014; Hagopian et al., 1996; Schiff et al., 2011). Modifications from previous studies were guided by ethical considerations to tailor the intervention to Jon’s specific needs (Bailey & Burch, 2022; Behavior Analyst Certification Board, 2020).

Multiple treatment components were employed to reduce the likelihood of severe problem behavior, including initially small fading steps, an open-door policy, and differential reinforcement of acceptance and swallowing. Prior research indicates that the combination of demand fading with differential reinforcement and extinction can effectively reduce severe destructive behavior and enhance compliance during instructional activities (Piazza et al., 1996). Similarly, acceptance and swallowing was programmed to be differentially reinforced at each fading step. Consequently, our procedure may be characterized as a differential reinforcement of compliance in conjunction with demand fading. Research has also demonstrated that differential positive reinforcement of compliance alone can reduce problem behavior and enhance compliance (Lalli et al., 1999; Slocum & Vollmer, 2015). Therefore, the integration of multiple treatment elements may have contributed to treatment compliance and the absence of problem behavior in the present study. The lower success observed with fading in medication and water may be attributed to the relatively high increase in demand during this treatment phase compared to the initial fading steps. The treatment modification and subsequent stabilization of performance support this hypothesis.

In the current study, no contingency test was conducted and no definitive evidence identified the variables maintaining intake refusal, although descriptive assessment suggested a combination of escape from liquid or medication consumption and access to preferred conversation topics. Jon’s intake refusal may be conceptualized as a restricted operant (Hanley, 2012), as the relevant antecedents and consequences were readily observable during baseline and probe sessions. Consequently, a comprehensive functional analysis may not have been necessary and may have been contraindicated given his medical instability. Therefore, intervention development was guided by direct observation of naturally occurring contingencies rather than a formal functional analysis.

However, in a study by Groff et al. (2014), the treatment of liquid and food refusal in a typically developing 4-year-old toddler was preceded by a functional analysis of attention and escape. Functional analysis revealed that escape was a function of the inappropriate behavior observed in liquid feeding, while attention served as an additional function when the child was presented with solid food. Therefore, future research could explore which type of assessment is most effective in addressing the multiple factors that may contribute to the onset and maintenance of liquid and medication refusal. Additionally, further research is warranted to address recent-onset intake difficulties in adult clients with neurodevelopmental disorders, as many of these issues may be maintained by reinforcement contingencies, and a function-based approach to treatment may offer a viable solution.

A number of limitations warrant consideration. We had only one client, which limits the generalizability of the results. The original baseline data represent food, drink, and medication-feeding attempts. Therefore, the original baseline is somewhat limited in that medication or water was never presented alone due to the original treatment goal of regaining food, liquid, and medication intake. In addition, we had brief reversals and no analysis of the contribution or necessity of the individual treatment components. A potential confound was introduced during Sessions 32–50 when the behavior analyst contacted Jon’s larynx to verify whether swallowing had occurred. Because this contact may also have functioned as a prompt, it is difficult to determine whether responding was controlled by the programmed tactile prompt or by the laryngeal verification itself, limiting interpretation of the results during this phase.

Additionally, IOA was collected in only 13% of sessions due to limited observer availability, which is below commonly recommended levels and raises concerns about data reliability and experimental control. Observer agreement was further challenged by the subtle nature of the swallowing response, as reflected by a lower-bound IOA score of 67% recorded on one occasion during this phase. During the initial baseline probe, Jon consumed a portion of the water, but the reason for this remains unclear. Water was faded in predetermined quantities, first via a syringe and later from fixed amounts in a cup. This raises the question of whether performance would have improved with greater access to water in each trial (see Kazdin, 2021). Nevertheless, the incremental nature of the intervention facilitated the monitoring of treatment effects.

An interdisciplinary team meeting concluded that restoration of oral intake was a feasible objective for Jon. Existing treatment literature has primarily focused on pediatric populations and short-term outcomes, with comparatively limited evidence addressing adults with long-term dependence on enteral intake. As a result, the generalizability of these procedures to adult populations remains limited and may warrant additional empirical investigation (Williams & Seiverling, 2023). Future research may consider modifying the procedure to allow for increased liquid availability at each fading step, thereby amounting to a free-operant arrangement. Additionally, some individuals may benefit from using smaller mock pills of increasing sizes before transitioning to actual medication (e.g., Babbitt et al., 1991; Beck et al., 2005).

In our study, the treatment room was designed to be a calm, distraction-free environment, and the treatment was meticulously structured and incorporated an open-door policy. This may be a relevant factor in treatment design for some adult individuals. We hope that this study will encourage the further development of evidence-based procedures for adults with neurodevelopmental disorders.

Recommendations to Clinicians and Students

Medical causes of intake refusal need to be assessed and addressed before commencing with a behavioral intervention. Input from different members of an interdisciplinary team ensures that adults with liquid and medication refusal are treated responsibly (Tereshko et al., 2021).

Adult individuals who exhibit total refusal of liquid and medication may experience multiple somatic issues, such as gastrointestinal discomfort. Therefore, even if a medical examination has been performed, and medical variables have been ruled out, it is imperative for the practitioner to remain vigilant for indications of pain, particularly during the initial stages of fading liquid. Consequently, and if possible, behavioral treatment for total liquid and medication refusal should rely exclusively on non-edible reinforcers, particularly in the initial stages of treatment. To identify an effective reinforcer to be used during treatment, an evidence-based preference assessment should be conducted.

In the present study, most clinical decisions were data-driven and guided by the client’s performance. Furthermore, the intervention was highly individualized and structured, adhering to a predefined treatment plan, to which the client’s family provided informed consent. The treatment plan was also flexibly adapted to address challenges encountered during the course of treatment. Safety was prioritized throughout the process, with active involvement from the client’s family and an open-door policy.

We recommend these principles in working with adult clients with neurodevelopmental disorders who suffer from similar difficulties with oral intake. To date, no clinical guidelines exist for treating intake refusal in adults with neurodevelopmental disorders. Clinicians are encouraged to consult up-to-date research on comparable conditions, as well as the pediatric feeding literature, when working with adult populations, and appropriately adapt relevant procedures to ensure their responsible application to adult needs.