The keyword search returned 923 initial results, 784 of which were academic journal articles. When articles not available in English were set aside, 729 articles remained. Duplicates occurring across the three databases were removed, leaving 556 articles. Following the examination of abstracts for the above-mentioned criteria as well as full-text review of apparently applicable articles, a total of 22 studies were identified as meeting the criteria for inclusion. Four additional articles were identified through review of the reference lists of these articles and corresponding abstract and/or full-text review. All 26 identified articles were categorized according to the type(s) of barrier described. Readers should note that the barrier categories named in this review were arrived at inductively and do not always reflect the exact terms that primary source authors used to describe each barrier. Further, few studies employed experimental methods to evaluate modifications, and primary source authors did not refer to modifications as “solutions.” To facilitate readers’ identification of information with relevance to their clients or applied research questions, results are organized by the order in which these barriers may arise during a toilet training case.

Four categories of barriers to daytime urine continence were identified: persistent problem behavior occurring in the toileting setting, excessive urine retention (i.e., withholding urine even while sitting on the toilet), recurrent accidents (i.e., ongoing accidents despite experiences with reinforcement for continent voids), and problems with self-initiations to toilet. Table 2 shows the range of topographies identified within each barrier category. For example, the barrier category “problems with self-initiations” includes both excessive self-initiations and absence of self-initiations. Table 3 shows the number of studies addressing each barrier category ordered by date of publication. Sixteen studies corresponded to multiple barrier categories. Eleven studies described problem behavior that occurred during toilet training, and 10 of those studies described modifications related to the problem behavior that occurred. Withholding urination during training was observed in 13 studies and addressed with a procedural modification in 11 of those identified studies. Recurrent accidents were observed in 17 studies and addressed in 10 of these studies. Five of the 11 studies that described problems with self-initiations attempted solutions to that barrier.

Table 3 Literature indicating barriers to toilet training in children with disabilities

Results are summarized with an introductory description of each barrier followed by a description of the potential solutions found in the studies included in this review. Results are also displayed in Table 4.

Table 4 Solutions for barriers to toilet training described in researchHow Can Practitioners Modify BATT for Children Who Engage in Problem Behavior?

Past research demonstrates that problem behavior may occur during BATT. Some studies have described generalized problem behavior that may occur across contexts such as aggressive responses or flopping, and others have described problem behavior that seems specific to the toileting environment. For example, Dalrymple and Ruble (1992) shared parent reports of disruptive behavior such as urinating out of the toilet, smearing feces, clogging the toilet with excessive amounts of toilet paper, and repeated unnecessary flushing. Strategies used to address problem behavior occurring during BATT are described below.

Modify or Eliminate Potentially Aversive Treatment Components

Some elements of BATT may constitute aversive stimulation for some individuals and thus occasion problem behavior. For example, longer-duration toilet sits increase the likelihood that a practitioner will observe and reinforce continent urination; however, problem behavior may also occur during extended sits due to either aversive stimulation or lack of access to alternative reinforcing activities. Overcorrection is a behavior-change strategy in which the child is required to engage in an effortful response that repairs an antecedent undesirable behavior (Cooper et al., 2020). Restitutional overcorrection, included in the Azrin and Foxx (1971) protocol, consists of requiring the participant to return the environment to an undisturbed or improved state (e.g., cleaning the floor after an accident). Positive practice overcorrection, included in Foxx and Azrin’s (1973b) procedure, consists of the participant repeating a desirable behavior multiple times following the undesirable behavior (e.g., walking to the bathroom and pulling their pants down and up multiple times following an accident). Overcorrection may occasion problem behavior, rendering BATT procedures that include overcorrection components impractical (Cicero & Pfadt, 2002).

Discontinue Overcorrection

Two studies eliminated overcorrection to reduce problem behavior. Doan and Toussaint (2016) noted distress in one participant during overcorrection in their evaluation of parent-implemented BATT. Overcorrection was discontinued by the parent to avoid further distress. Hagopian et al. (1993) observed self-injurious behavior (SIB) during toilet sits. Investigators discontinued overcorrection and implemented a differential reinforcement of other behavior contingency as well as reinforcement of continent urination. Investigators also reduced the duration of required toilet sits and blocked instances of SIB. Although these modifications to BATT resulted in decreases in SIB, decreases were also observed in continent voids. In other words, without overcorrection, the training package described by Hagopian et al. was safer but less effective for toilet training.

Modify Overcorrection

Two studies reported successful resolution of problem behavior when overcorrection was simplified rather than eliminated. Brown and Peace (2011) stated concerns related to aggressive and sexualized behavior (i.e., masturbation) occurring during changing of clothing and following directions in baseline. They implemented a simplified positive practice procedure consisting of a neutral verbal statement describing the accident and a single trip to the changing area where the participant was prompted to remove soiled clothing, wash, and dress. Investigators also modified the typical BATT toileting schedule such that toilet visits occurred only every 60–90 min. Near-zero rates of aggression and sexualized behavior were reported. Other examples of simplified procedures are evident in the literature: Richmond (1983) required participants to simply wash off and change clothes following accidents. Likewise, Cocchiola et al. (2012) stated, “You wet your pants. You need to change,” before guiding participants to the bathroom and assisting them in changing clothes following accidents without additional consequences. LeBlanc et al. (2005) stated, “No wet pants,” and required participants to complete a 1-min toilet sit immediately following accidents. Full positive practice was avoided if the participant urinated in the toilet during that toilet sit. Hanney et al. (2013) suggested that, if positive practice is implemented in this way (i.e., with an option to avoid repeated toilet visits contingent upon continent urination), children will experience fewer instances of full positive practice. Comparative studies have yet to evaluate the efficacy of partial positive practice.

Reduce the Duration of Toilet Sits

Three studies reported successful resolution of problem behavior when toilet sits were abbreviated. Like Hagopian et al. (1993), Post and Kirkpatrick (2004) reduced the prescribed duration of toilet sits in an effort to decrease problem behavior. At first, the investigators prescribed 20-min toilet sits following the observation that the participant urinated soon after shorter toilet sits had ended. However, the participant engaged in hitting and pinching during 20-min toilet sits. Investigators verbally redirected aggression and reduced the toilet sit duration to 10 min. Aggression decreased, and 80% of urination occurred in the toilet across a 3-day period, at which point investigators further reduced the duration of toilet sits to a maximum of 5 min without urination. Even a small reduction in toilet sit duration may be effective: Doan and Toussaint (2016) reduced sits from 5 to 3 min, and that change correlated with decreases in a participant’s crying behavior. This individualized modification was informed by the observation that voids tended to occur within the first 3 min of sitting on the toilet.

Modify Access to Stimuli

Fully independent toileting requires unassisted access to a toilet, toilet paper, soap, and water as well as dressing and undressing. In the training setting, access to these stimuli and activities may interfere with skill acquisition. On the other hand, providing access to preferred stimuli unrelated to training may enrich the environment and reduce interfering behavior. Studies describing these manipulations are shared below.

Limit Access to Stimuli that Occasion Problem Behavior

As noted above, Brown and Peace (2011) described masturbation that occurred during post-accident clothing changes, which can become problematic in public settings. In addition to limiting overcorrection, Brown and Peace omitted activities historically correlated with this behavior (e.g., unnecessary washing of the genital area) while increasing and reinforcing the participant’s independence in his toileting routine (i.e., differential reinforcement of alternative behavior; DRA), potentially contributing to reported decreases in masturbation. The specific contributions of each component in this training package are unknown; however, this strategy may also be useful for preventing the sort of interfering behavior described by Dalrymple and Ruble (1992; e.g., clogging the toilet may be avoided by limiting client access to excessive toilet paper).

Provide Noncontingent Access to Preferred Stimuli

Two studies reported attempts to resolve problem behavior with noncontingent access to preferred stimuli. Dowdy et al. (2020) provided noncontingent access to a cell phone that was demonstrated to compete with toilet lid destruction during routine bathroom visits. Results indicated a decrease in destruction from an average of 1.5 responses per minute to an average of 0.1 responses per minute. Although this behavior was targeted in part because it interfered with toileting routines, data on continent voids were not reported. Lomas Mevers et al. (2018) described noncontingent access to preferred items for three participants who engaged in problem behavior. It should be noted that these items were identified via preference assessment as being less preferred than the reinforcer delivered for continence behaviors. All three participants were noted to be continent during a follow-up assessment, but data were not reported for problem behavior.

Implement Functional Assessment and Differential Reinforcement

Research indicates the use of functional assessment to identify reinforcers for interfering problem behavior such as property destruction, but urination away from the toilet can also acquire a socially mediated function. In these cases, inappropriate urination or socially maintained urination may be apt terms because the term accident implies a lack of physical control or skill. Discerning whether urination away from the toilet is evidence of a skill deficit or a history of differential consequences requires assessment of the conditions under which accidents occur. Socially maintained urination may be suspected when accidents are temporally related to previously established escape-maintained topographies of behavior (e.g., vocal refusal, destroying work materials) or when children reliably pull down their pants prior to urinating away from the toilet. In either case, functional assessment remains a useful tool for informing differential reinforcement contingencies to increase functional communication and continent urination. Although differential reinforcement is embedded in BATT (i.e., self-initiations and continent urination are differentially reinforced), some studies detail the use of additional strategies to address problem behavior.

Assessment of Interfering Problem Behavior

Two studies reported assessment-informed solutions for problem behavior. Flood and Luiselli (2016) described an intervention targeting crying and yelling during bathroom visits and avoidance of community bathrooms. Functional assessment results suggested that crying and yelling were maintained by negative reinforcement in the form of escaping or avoiding the presence of other people near the bathroom. Investigators provided tokens contingent on “talking nicely” (i.e., the absence of crying and yelling in the bathroom) while therapist proximity was systematically increased. Tokens could be exchanged for access to video games. This contingency was successful in increasing toleration of others near the bathroom. Likewise, Dowdy et al. (2020; described above) implemented functional assessment of toilet lid destruction as well as a competing stimulus assessment to inform their noncontingent reinforcement intervention for this automatically maintained behavior.

Assessment of Accidents

Two studies identified social consequences as potential maintaining variables for accidents. Ricciardi and Luiselli (2003) conducted functional assessment interviews and observations before concluding that urination away from the toilet was maintained by escape from demands and access to attention for a participant who also demonstrated independent toileting. During the intervention, investigators provided differential attention for continent urination as well as escape extinction (i.e., accidents were no longer followed by escape from demands). They required the participant to wear a diaper during escape extinction so that urination would not require an immediate trip to the bathroom or change of clothes. Accidents resolved, and diapers were discontinued. Antecedent manipulations may also be helpful in these situations: Foxx and Garito (2007) reduced overall instructional time from 6 hr per day to 2.5 hr per day after observing that urinary and bowel movement accidents appeared to be maintained by escape from demands.

How Can Practitioners Modify BATT for Children Who Withhold Urination on the Toilet?

Behavior analytic toilet training typically includes frequent, extended-duration toilet sits and increased fluid consumption across long (e.g., 8-hr) training periods. Whereas these procedures often increase the negative reinforcing value of urinating in the toilet, studies have included some children who hold their urine or who produce very small amounts of urine while withholding proportionally larger amounts, increasing the likelihood of accidents later in the day. Unlike other toileting responses (e.g., sitting, wiping), urination cannot be manually prompted by the trainer. When BATT does not result in continent voids and resulting opportunities for reinforcement, research indicates the use of individualized strategies exemplified in the following studies.

Transfer Stimulus Control to the Target Conditions for Toileting

A case of faulty stimulus control may be suspected when children withhold urination and urinate only at night or when wearing a diaper outside of BATT sessions. Two potential solutions have been described for establishing urination discriminated by the presence of the toilet and associated bathroom stimuli.

Employ Shaping and Fading Techniques

Luiselli (1996a) designed a procedure to transfer stimulus control over urination from the diaper to the toilet. He recommended that children wear training pants except during training sessions (scheduled every 60–90 min), when they should wear a diaper and sit on the toilet. After 3–5 min of sitting on the toilet, the diaper should be removed, and reinforcement delivered if urination has occurred. Luiselli recommended that stimulus fading, consisting of cutting a 1-in hole in the diaper, should begin after the child wets the diaper in 85%–100% of training sessions for 1 week. The hole should be made progressively larger until continent urination occurs consistently.

Three studies implemented this diaper fading procedure when children did not urinate in the toilet. Luiselli (1996b) initiated diaper fading with a participant who only urinated in disposable diapers, despite repeated caregiver attempts to prompt continent urination. After the participant was reliably urinating on the toilet while wearing a diaper, she began to sit on the toilet independently before the altered diaper could be applied during training sessions. In this case, simply reinforcing urination that occurred while the participant was sitting on the toilet while wearing a diaper appeared to be sufficient to transfer stimulus control. Lomas Mevers et al. (2018) reported that a similar diaper fading procedure was associated with continence at a follow-up assessment within 6 months of training for two participants in their study who experienced this modification. Diaper fading requires that voids regularly occur during daytime hours (or whenever training is possible). Therefore, it may be necessary to shape daytime urination prior to implementing this modification. Smith et al. (2000) demonstrated this process by reinforcing urination in the diaper at certain times of day. After several days, Smith et al. introduced a chair, which the participant sat upon in a diaper to eliminate before moving to the toilet and undergoing successful diaper fading.

Stimulus control manipulations may also be useful if the child does not wear diapers. Despite long scheduled toilet sits (upwards of 15 min), Taylor et al. (1994) observed that their participant typically had an accident within 3 min of dressing after a toilet sit. The investigators hypothesized that underpants were functioning as a discriminative stimulus for urination and suspected that the tactile differences between clothing and the toilet may have inhibited urination. To transfer stimulus control from underpants to the toilet, Taylor et al. removed the participant’s pants and underpants 5 min after delivering liquids. The participant sat on the toilet unclothed until 10 min had passed or he had urinated in the toilet, whichever occurred first. The duration of time per session that the participant was undressed was faded based on the occurrence of continent urination. Zero accidents were reported at 10-month follow-up observations, and unprompted continent urination was reported to occur approximately once per hour. Taylor et al. suggested switching to underpants made from an unfamiliar material during prescribed no-underpants times if nudity is not acceptable or feasible in a given setting. The investigators noted both that the duration of nudity decreased during each consecutive session and that the overall duration of nudity was not more than the participant’s estimated time out of clothing while cleaning up accidents prior to intervention.

Once urination reliably occurs in the toilet, it is possible that shaping may be applied to increase the duration of continent voids. Lomas Mevers et al. (2018) attempted to use a percentile schedule of reinforcement, in which the criterion for reinforcement of voids is recalculated with each successive void to shape differentially longer voids. However, only partial continence was reported at the end of the study (65% of voids occurred in the toilet) for the participant who received this modification. Croteau et al. (2022) describe a relatively simpler method for increasing urine output: when investigators observed that immediate reinforcer delivery terminated the flow of urine, they modified their procedure to wait until the urine stream was complete before delivering reinforcers.

Provide Water Prompts

One study described a procedure to elicit continent urination as a prerequisite to delivering reinforcement and establishing stimulus control over elimination in the toilet. To address a decrease in continent urination for their participant, Hagopian et al. (1993) poured a small amount of lukewarm water over the boy’s genitals for 3–5 s while he was seated on the toilet. When this procedure was added to BATT along with contingencies designed to reduce SIB (described under the previous barrier), continent urination increased. Hagopian et al. hypothesized that water prompting may have elicited urination, thus providing opportunities for reinforcement. Because the water prompt was not faded, it is unknown whether urine continence was achieved under naturally occurring stimulus conditions.

Apply a Negative Reinforcement Contingency for Toilet Sits

Four studies in this review detailed the use of negative reinforcement contingencies.

In one example, when a child would not use the toilet at school in the same way he did at home, Heyward (1988) initially designed a shaping and fading procedure to transfer stimulus control over continent urination from proximity of the child’s mother to proximity of a school classroom assistant. However, the child stopped urinating on the toilet as soon as his mother was no longer visible. Heyward then introduced a negative reinforcement procedure. The child was given 1.5 pints of liquid to drink at home without access to his home toilet. Upon arriving to school, the child sat on the toilet until he urinated. Within 1 day, the latency to urination decreased from 2.75 hr to less than 30 min. The child’s toileting schedule was later adjusted to match his classmates, and low latencies to urination maintained for at least 4 years.

In a second empirical demonstration of negative reinforcement techniques, Luiselli (2007) described a case study in which a child had never urinated in a toilet despite a history of scheduled 3-min toilet sits. First, the investigator presented a series of positive reinforcement contingencies illustrated on cards. The child continued to withhold urine on the toilet. Luiselli then introduced a negative reinforcement contingency by replacing the 3-min scheduled sit with sitting either until urination occurred or until 20 min elapsed, whichever came first. If the child did not urinate within 20 min, he was allowed to leave the toilet for 10 min and then return to sit for another 20 min until he urinated. This strategy was similar to the negative reinforcement procedure designed by Azrin and Foxx (1971) except that Luiselli ceased scheduled sits after one successful continent void each day.

Increase Opportunities for Positive Reinforcement of Continent Voids

Two studies portrayed increases in positive reinforcement contingencies to resolve urine retention. Luiselli (1997) described a toileting baseline that involved a negative reinforcement contingency applied to two daily toilet visits which was unsuccessful because the participant never urinated in the school toilet despite having used a toilet at home. The intervention included one training session per school day to minimize toileting demands. In contrast to Luiselli (2007), the training session was scheduled at the end of each day with the notion that the participant would be more likely to urinate with a full bladder. A positive reinforcement contingency (drinking water from a preferred container) was added to the negative reinforcement contingency, and the participant began to urinate on the toilet and experience both contingencies.

Post and Kirkpatrick (2004) initially elected to prompt toileting opportunities according to the natural schedule of incontinent voids observed in baseline. However, they observed that their participant urinated right after his training pants were reapplied following a sit without urination. Investigators implemented a regular 30-min sit schedule with 20-min prescribed sits, thus providing more opportunities for reinforcement of continent voids across the day. The sit schedule was systematically faded after 3 consecutive days with 80% continent voids. Later, a negative reinforcement contingency was added to each sit.

How Can Practitioners Modify BATT for Children Who Have Ongoing Accidents after Reinforcement of Continent Urination?

Sometimes accidents occur despite frequent successful continent voids with contingent reinforcement and other elements of BATT. When accounting for the contrast between behavior and the known effects of reinforcement, it is possible that the reinforcer available for producing urine in the toilet is not as powerful as the reinforcement history for accidents or that the reinforcer is not applied contingently. However, assuming that reinforcer efficacy has been confirmed and procedural integrity is intact, research indicates the following modifications to BATT procedures under conditions of ongoing, or recurrent, accidents.

Enhance the Salience of Accidents

Recurrent accidents may signal a failure to establish continent urination as a discriminated operant. Thus, some researchers have explored methods to facilitate discrimination by enhancing the salience of accidents as detailed below.

Review Programmed Consequences

Although ethical precautions are warranted on a case-by-case basis, the addition of effortful requirements following accidents may result in behavior change. Lomas Mevers et al. (2018) added positive practice overcorrection as a remedial strategy for children who continued to have accidents after an initial BATT package without overcorrection. Of the 11 participants who met criteria for positive practice, five were reported to be continent at follow-up. Lomas Mevers et al. suggested that these mixed results might indicate either that positive practice was not sufficiently aversive to serve as a punisher or that positive practice might have functioned as a reinforcer for some participants.

Use a Urine Alarm or Implement Dry Pants Checks

A urine alarm alerts both the trainer and the child to the occurrence of accidents. For the trainer, the urine alarm permits immediate detection of accidents and immediate delivery of consequences. For the child, the alarm may increase awareness of the urge to urinate as well as occasion the engagement of muscles that control urine flow (Friman, 2010). Modern urine alarm technology includes wireless alerts for both learner and therapist as well as disposable sensors (e.g., Mruzek et al., 2019). A urine alarm was used in three studies that met criteria for this review (LeBlanc et al., 2005; Lomas Mevers et al., 2018; Taylor et al., 1994). Urine alarms have been proven efficacious across participants with varying skill levels (e.g., Lancioni & Markus, 1999), as a standalone treatment (Friman & Vollmer, 1995), and as components of comprehensive treatment packages (e.g., Azrin & Foxx 1971; LeBlanc et al., 2005). However, the literature also includes studies in which BATT variations with urine alarms were not effective (e.g., Didden et al., 2001; Lomas Mevers et al., 2018). In a review of the use of urine alarms, Levato et al. (2016) cited a lack of understanding of the behavioral mechanism that mediates improvement in toileting as well as insufficient overall evidence of the efficacy and effectiveness of urine alarms for treating incontinence.

Practitioners may incorporate a low-tech alternative to urine alarms such as a dry pants check. For example, Smith (1979) reported positive results when dry pants checks were implemented with fluid loading, scheduled toileting, and reinforcement of continent voids. Although six studies meeting criteria for inclusion in this review included dry pants checks as a treatment component, none provided within-subject comparisons of BATT with and without dry pants checks.

Exchange Diapers for Underpants

It is reasonable to conclude that the effects of accidents (e.g., wet pants) are more salient for children in underpants compared to an absorbent diaper. The discomfort of wet pants contingent on accidents may sometimes function as an automatic punisher, decreasing the likelihood of future accidents. Nine studies meeting criteria for inclusion in this review illustrate support for removal of diapers as a component of BATT. Cagliani et al. (2021) exchanged diapers for underpants as a component of a toileting baseline phase, along with a 90-min toileting schedule, graduated guidance following accidents, and differential reinforcement for continent voids. Following these simple manipulations, one participant’s accidents decreased to zero. It should be noted that the three remaining participants in this study required additional intervention components to achieve continence. Research beyond the scope of this review supports the removal of diapers for nondisabled children and adults with disabilities (e.g., Greer et al., 2016; Tarbox et al., 2004; Simon & Thompson, 2006).

A fading method may be an alternative to outright withdrawal of diapers. Ricciardi and Luiselli (2003) observed an increase in accidents for a participant whose incontinence appeared to be maintained by socially mediated consequences. Experimenters revised their procedure such that the participant wore a diaper throughout each day except for a 30-min interval that was randomly scheduled, during which he wore underpants. After the participant refrained from accidents during this 30-min interval for multiple days, the duration of the underpants interval was systematically increased.

Revise the Toileting Schedule to Provide More Opportunities for Reinforcement of Continent Voids

Four studies detailed increases in prescribed toilet sits as a method of decreasing accidents. For example, Richmond (1983) described accidents that persisted when children sat on the toilet each hour and received praise for continent urination. When the inter-sit interval (i.e., programmed duration of time away from the toilet) was reduced to 15 min during BATT, the investigator observed an immediate decrease in toileting accidents. The sit schedule was successfully faded over a period of 4 weeks to one sit every 2 hr. The denser toileting schedule might have increased the likelihood that urination would occur and be reinforced during toilet sits. However, simple correction for accidents (i.e., reprimands and instructions) was added at the same time as the schedule manipulation, so it is unclear whether a decrease in the inter-sit interval alone would be effective in this case.

Two recent publications provide guidance for data-based decision-making related to the sit schedule. Perez et al. (2020) implemented a denser sit schedule for two participants who did not become continent with package consisting of exchanging diapers for underpants, a 3-min toilet sit scheduled every 30 min, a dry pants check, and differential reinforcement for dry pants and in-toilet urination. The denser schedule (toilet sits every 15 min) was initially implemented during the time when accidents were observed most frequently and later extended across the day. A toddler potty chair was introduced to minimize time spent making the transition as a function of more frequent bathroom visits. One participant conformed to the denser sit schedule after the potty chair was implemented and their percentage of continent voids increased, but the second participant required a more comprehensive training solution. Cocchiola et al. (2012) moved one participant from a 30-min sit schedule to a 15-min sit schedule from 11:15 am to 12:30 pm as a BATT modification after observing a higher rate of accidents during this specific window of time. The 30-min sit schedule was in place during the remaining training hours each day, and the participant met the mastery criterion of 100% continent voids on a 2-hr schedule after 32 training days.

When a toileting schedule is revised to include shorter inter-sit intervals, researchers and practitioners must plan to efficiently fade the sit schedule so that the child may resume their regular activities. Fading criteria and duration varied across the included studies that modified sit schedules. For example, Richmond (1983) faded the toilet sit schedule over a month-long period. LeBlanc et al. (2005) faded a toilet sit schedule from prescribed sits every 5 min to zero scheduled sits over a period ranging from 9-23 days across three participants. It should be note that in the LeBlanc et al. study, one participant had not urinated as expected on the first day. Her sit schedule was held in place rather than faded until she successfully voided on the toilet. In a different approach to fading, Cicero and Pfadt (2002) removed their 30-min toilet sit schedule the first day after a participant’s unprompted initiation was followed by continent urination. Practitioners may consider conducting intermittent probes of the terminal schedule to potentially shorten the duration of training and the labor associated with schedule thinning (e.g., Perez et al., 2020).

Finally, adding a supplemental toilet sit is an alternative to setting one inter-sit interval for the entire training day. Cagliani et al. (2021) observed ongoing morning accidents despite the participant’s experiences with positive reinforcement, so they scheduled an extra bathroom trip to occur in the morning without requiring more frequent toileting visits throughout the day. The percentage of voids occurring in the toilet increased to 100%.

Monitor Fluid Intake

Although it is important for children to be well-hydrated, the additional fluids offered during toilet training may result in increased accidents. Both Cagliani et al. (2021) and Doan and Toussaint (2016) observed increases in accidents correlated with the provision of additional fluids. Doan and Toussaint reported eventual decreases in accidents, and Cagliani et al. discontinued additional fluids when accidents did not resolve after several sessions. Increased hydration occurring outside of the training setting may also need to be considered: Chung (2007) observed that accidents occurred after a participant swallowed excessive amounts of water while swimming.

Use a Urine Target

Sometimes voids occur on an ongoing basis near (but not in) the toilet. This type of toileting accident may be related to either a skill deficit or access to reinforcers associated with urinating onto the seat or floor around the toilet. Siegel (1977) placed a visual target in the toilet to reduce accidents with three participants. Prior to the intervention, participants regularly urinated on the seats, walls, and floors of toilet stalls. During treatment sessions, Siegel prescribed increased fluid intake for participants and a visual target was placed in the toilet. Participants began to direct their urine streams toward the target, resulting in fewer accidents. When the target was removed, accidents increased. The target was replaced, and accidents decreased again. In the final phase, participants were instructed to hit the target, resulting in near-zero rates of accidents. Siegel reported that participants appeared to enjoy trying to hit the target.

How Can Practitioners Modify BATT with Children Who Self-Initiate Too Rarely or Too Often?

It would be ideal for well-hydrated children to learn to identify the urge to urinate (i.e., proprioceptive stimuli), initiate an independent toilet request, and proceed to the toilet to void. Multiple studies have included prompts for participants to request the toilet as a component of BATT (e.g., Cagliani et al., 2021; Cicero & Pfadt, 2002; LeBlanc et al., 2005, Perez et al., 2020). Effects on self-initiations have been mixed. For example, Kroeger and Sorensen (2010) reported that 100% of toileting events were self-initiated for both participants at the end of their investigation. Croteau et al. (2022), by contrast, observed a decrease in self-initiations from 1.5 per day in baseline and initial treatment to 0 by the end of treatment for one participant and little to no self-initiations from the other three participants.

Differences in the outcomes of studies that addressed self-initiations during BATT may be related to varying definitions of self-initiation. Azrin and Foxx (1971) defined self-initiation as independent toilet approach (p. 93). Cicero and Pfadt (2002) redefined self-initiations as the child requesting the toilet in the absence of prompting. Cicero and Pfadt’s definition may be more practical for children with disabilities who require caregiver supervision and must ask to use the toilet across settings (e.g., school, community). It should be noted that six of the seven most recently published studies meeting criteria for inclusion in this review included unique definitions for self-initiations. These definitions encompassed a range of self-initiation topographies, and some but not all definitions included that participants should urinate after requesting and accessing the bathroom. Once established, self-initiations may become problematic if they occur too frequently (i.e., more often than the child needs to empty their bladder).

Teach the Child to Respond Independently to Supplemental Cues

If a client does not demonstrate self-initiations under conditions of a full bladder and a history of contingent access to the bathroom, it may be helpful to introduce supplemental prompts while fading adult prompting. Luiselli (1987) observed a failure to teach self-initiations with a toileting schedule, restitutional overcorrection (washing clothes), and reinforcement for completing a toileting routine. The investigator gave the participant a wristwatch and instructed her to request the bathroom within specified time intervals. Reinforcement was contingent on the emission of toilet requests within the correct interval followed by continent urination. Self-initiated toileting increased from near-zero to approximately 14 instances per week.

Discontinue Prompts to Toilet

LeBlanc et al. (2005) suggested that frequent toilet prompts may inhibit self-initiations due to insufficient motivation (i.e., children are less likely to experience a full bladder). Thus, they implemented a sit schedule that was rapidly faded across days and removed when the participant reached 80% success with a 4-hr inter-sit interval across 2 consecutive days. In total, 11 studies included in this review included progressive sit schedules with fading. As a component of their toilet training package, Lomas Mevers et al. (2018) included a toilet sit schedule that increased time away from the toilet in 5 to 15 min increments. The inter-sit interval was increased or decreased based on continent or incontinent voids. If participants progressed through the schedule up to 60 min away from the toilet without the emergence of self-initiations, Lomas Mevers et al. discontinued the sit schedule and therefore relied upon self-initiations. Investigators reported that this strategy was associated with sustained continence (100% continence on the last day for four participants; 67% continence for one participant), although data on self-initiations were not reported. By contrast, continence decreased for at least one participant when Perez et al. (2020) removed the sit schedule, and only 3 of 11 participants were reported to self-initiate toilet requests at the conclusion of the study. Seven participants continued to require a 120-min prompted sit schedule, and 1 participant required more intensive training beyond the methods employed in the Perez et al. study.

Implement Functional Assessment to Identify Potential Maintaining Reinforcers for Requests

Research indicates that toilet requests may serve multiple functions, indicated when self-initiations to toilet are not consistently followed by voids. For example, Perez et al. (2020) reported correspondence values between self-initiations and subsequent continent urination for 11 participants post-mastery of targeted toileting skills. The values ranged from 50% or lower (3 participants) to 100% (2 participants), suggesting that, in some cases, self-initiations may h