Abstract
LeBlanc, Crossett, Bennett, Detweiler, and Carr (2005) described an outpatient model for conducting intensive toilet training with young children with autism using a modified Azrin and Foxx, protocol. In this article, we summarize the use of the protocol in an outpatient setting and the outcomes achieved with a large sample of children with autism spectrum disorders. Thirty archival clinical records were coded for several variables related to treatment implementation and outcome. The majority of participants achieved full continence in an average of approximately 2 weeks. Details on the typical implementation and course of treatment are presented to provide a profile for future practitioners.
Numerous investigations have demonstrated that the intensive behavioral treatment package developed by Azrin and Foxx (Azrin & Foxx, 1971; Foxx & Azrin, 1973) for primary urinary incontinence is effective for individuals with developmental disabilities. The original treatment package, which was first developed for adults with intellectual disability, includes the following components: scheduled toileting, reinforcement of in-toilet urination and dry pant checks, increased fluid intake, and overcorrection (restitution, positive practice) for accidents. Modifications to the treatment package have included attempts to render the original package less intensive (e.g., Luiselli, Reisman, Helfen, & Pemberton, 1979), to evaluate its effectiveness without the overcorrection component (e.g., Post & Kirkpatrick, 2004), and to improve its effectiveness with transfer-of-stimulus-control procedures (e.g., Taylor, Cipani, & Clardy, 1994).
Using surveys, researchers have found that the majority of parents of children with autism report toileting and urination problems (Whiteley, 2004; Williams, Oliver, Allard, & Sears, 2003), and these concerns have stimulated research specifically with this population. For example, several investigations (Cicero & Pfadt, 2002; Luiselli, 1996, 1997; Post & Kirkpatrick, 2004) have evaluated versions of the Azrin and Foxx (1971) treatment package with children with autism and other pervasive developmental disorders. In the most recent empirical investigation of a version of this treatment package, LeBlanc, Crossett, Bennett, Detweiler, and Carr (2005) reported outcomes for three children with autism who received services at an outpatient clinic.
The study by LeBlanc et al. (2005) included three children who had already been exposed to less-intensive toilet training methods (e.g., scheduled toileting) without success. The protocol was implemented by clinic staff and parents on the first day in the clinic and by parents and school staff on all subsequent days at home and school. The treatment components included positive (e.g., preferred edibles and toys) and negative (i.e., escape from the toilet) reinforcement of successful voids, communication (request) training, increased fluids, an intensive and progressive sitting schedule, use of a urine alarm to signal accidents and interrupt the urine stream via a startle response, and positive practice for accidents. Although the participants had not benefited from prior interventions and the protocol was primarily implemented by family and school staff, the LeBlanc et al. protocol was effective in producing fully independent toileting repertoires for two of the three children and full continence with minimal prompts for the third child who did not consistently initiate toileting events. Furthermore, all treatment components except intermittent praise for successful toileting were removed by the end of the evaluation, and continence was maintained at a 1-month follow-up assessment.
Despite the success of the Azrin and Foxx (1971) treatment package and its recent iterations, our experience has been that many practitioners do not use it, opting instead for less-intensive approaches with questionable efficacy. It could be, as Kroeger and Sorensen-Burnworth (2009) indicated, that iterations of the Azrin and Foxx method protocol are simply too onerous. However, we also suspect that it is not only the effort of implementing the treatment package that may deter practitioners but also lack of experience implementing it and, thus, uncertainty about what to expect during implementation. In an effort to better inform practitioners, the purpose of the present article is to summarize the process and outcomes during implementation of the LeBlanc et al. (2005) protocol at a university-based outpatient clinic for children with autism spectrum disorders. Archival data from 30 records were coded and aggregated to provide a profile of the typical implementation of the protocol and its resulting outcomes.
Method
Description of Therapeutic Services
Participants were clients of a small, university-based outpatient psychology training clinic serving children aged 2 to 12 years diagnosed with autism, pervasive developmental disorder not otherwise specified (PDD-NOS), or another developmental disability. Case records from 30 clients served over a 7-year period were drawn from the clinic’s intensive toilet training service for the present analysis. The 7-year period was selected because that was the time frame during which the authors worked together at the clinic; a total of 30 clients presented for toilet training during that time. The intensive toilet training procedure was based on the aforementioned protocol described by LeBlanc et al. (2005; see Table 1 for a description of each intervention component). The intervention, which was overseen by a faculty-supervised graduate student in behavior analysis or clinical psychology, began with an intensive day at the clinic (typically a Friday). Graduate students gradually transferred implementation of the procedures to a parent or caregiver during the course of the first treatment day. The graduate students then provided phone consultation during the following weekend when the intervention was continued at home as well as on-site training to teachers and school staff who typically implemented the procedures at school beginning the following Monday. All therapeutic support was eventually faded as clients progressed through the procedure. The toilet training services provided to the participants in this analysis were overseen by a total of 20 graduate students.
Components of the Intensive Training Protocol (LeBlanc et al., 2005).
The mean age of the 30 participants was 5.5 years (range = 3.7–10.4 years) and 23 (76.7%) were male. Information on prior diagnosis was available for all 30 participants. Of those 30 cases, 25 (83.3%) had a diagnosis of autism, 1 (3.3%) of them had a diagnosis of PDD-NOS, and the remaining 4 (13.3%) participants had a non-PDD developmental disability. The majority of the participants (20, 66.7%) were referred for services by a parent and 1 (3.3%) was referred by a physician; no information on referral source was available for the remaining 9 (30%) participants.
Coding Procedures and Categories
Data for the present analysis were obtained by the second and fifth authors from the archived data sheets, progress notes, and termination summaries located in the case files of clients who received the intensive toilet training procedure. Although the data presented were obtained from 30 different cases, the total sample size in each analysis differs as information was sometimes missing from client files. All files were coded for the following information.
Participant demographics
Each case file was coded for the child’s age, gender, diagnosis, referral source, and communication modality.
Target behaviors
Frequency and latency data were recorded for the following target behaviors during the intensive treatment day: successes, accidents, accident/success conversions, and initiations. Successes were scored when an individual voided in the toilet, and accidents were scored when an individual voided anywhere outside of the toilet. Accident/success conversions were scored when an individual had an accident (i.e., any amount of urine produced away from the toilet) but successfully completed the void in the toilet after being taken to the bathroom. Initiations were scored when an individual independently requested to void using his or her trained communicative response (e.g., vocal, picture exchange).
Intervention components
The different types of programmed consequence for successful voids during differential reinforcement were coded as escape from the toilet, access to a TV, social praise, delivery of toys/edibles, and other. The modality of the communicative response used during communication training was coded as picture exchange, vocal, or other. Information from the intensive treatment day was coded for the duration of the intensive treatment day, the use of increased fluids, and the individuals who were trained in the protocol (e.g., parents). Data also were recorded on the use of the urine alarm and the latency to the discontinuation of the urine alarm after the intensive treatment day. Finally, the number of instances of positive practice occurring on the intensive treatment day was derived from the data on the number of accidents. Positive practice consisted of quickly escorting the participant to the toilet after stating “No wet pants” in a firm voice tone. After removing his or her pants, sitting for 1 min, standing, and replacing the clothing, the participant was rushed back to the site of the accident for four repetitions of the procedure (full positive practice) unless a void occurred in the toilet during the 1-min sit (partial positive practice). Full positive practice occurred after an instance of an accident, and partial positive practice occurred after any accident/success conversions.
Level of continence
The level of continence each child achieved prior to discharge was identified from each case file and coded as one of three levels or unknown. Full continence was defined as consistent and independent initiation of successful voids without any accidents. Schedule-dependent continence was defined as successful voids upon being prompted or taken to the toilet (approximately every 3–4 hr) without consistent independent requests to void. Partial continence was defined as inconsistent successful voids, either prompted or independent, with periodic accidents. Unknown was coded when insufficient data were available from the client’s home and school environments.
Results
For the 23 cases with specific information recorded about communication modality, 13 were trained using only a vocal modality (56.5%), 7 were trained using only a picture exchange (30.4%), and another 3 participants (13%) were trained using both vocal and picture-exchange responses. Whereas data on at least 1 day of treatment were available for all participants, data on the final level of continence achieved at posttraining follow-up were available for only 20 of the 30 cases. The majority of these 20 participants achieved full continence (14, 70%), while several participants reached either a schedule-dependent (3, 15%) or a partial-continence (3, 15%) status. Those participants who achieved full continence achieved this in a mean of 14.4 days.
Table 2 depicts the variety of programmed consequences used to differentially reinforce successful voids for each participant with detailed enough information in the case file for coding this variable (n = 26). The majority of these 26 participants (21, 80.8%) received at least three types of reinforcer (e.g., escape from the toilet, social, toys/edibles), and virtually all of them (24, 92.3%) received at least two types of reinforcer. The mean duration of services during the intensive treatment day was available for 26 of the 30 participants and was 381 min (SD = 51.1), approximately 6.4 hr. Information was available on level of fluid intake for 28 of the cases, and increased fluids were provided to 27 of the 28 (96.4%) participants. Information was available on use of a urine alarm for 28 cases, and an alarm was used with 26 (92.9%) of those participants. For the majority of participants (n = 28), positive practice was programmed to occur after accidents but not after accident/success conversions. On average, participants encountered full positive practice an average of 3.5 times during the intensive treatment day (SD = 3.2) and partial positive practice an average of 2.1 times (SD = 2.4) during that day. However, some participants never encountered full positive practice because all of their accidents converted to successes on the toilet during the 1-min sit.
Programmed Consequences Used During Differential Reinforcement.
No information available.
The top panel of Figure 1 depicts the number of successes, accidents, and accident/success conversions for each participant during the intensive treatment day. The mean latency to the first success was 92.4 min (SD = 73.5), and the mean latency to the first accident was 153.8 min (SD = 105.2). The data on accident/success conversions showed that an average of 69.1% (SD = 88.6) of accidents converted to successes. The bottom panel of Figure 1 depicts the number of initiations during the intensive treatment day (M = 2.1) for each participant. The mean latency to the first initiation was 158.9 min (SD = 95.8).
The number of successes, successes/accidents, and accidents for each participant during the intensive treatment day (top panel). Number of initiations for each participant during the intensive treatment day (bottom panel).
Information regarding the sitting schedule on the intensive treatment day was available for 24 participants. The mean beginning schedule was 9.0 min (SD = 2.1) on the toilet and 8.3 min (SD = 6.9) off the toilet, whereas the mean ending schedule was 5.2 min (SD = 1.0) on the toilet and 45.4 min (SD = 18.2) off the toilet. The beginning sit schedule for 18 (75%) of the participants was identical to the LeBlanc et al. (2005) protocol with 10 min on the toilet and 5 min off, whereas 6 (25%) of the participants began with a schedule of 5 min on the toilet and time off ranging from 10 to 30 min. At the end of the intensive treatment day, 9 (37.5%) participants were on at least a 60-min time-off schedule, whereas 11 (45.8%) participants were on at least a 30-min time-off schedule. Overall, for all coded participants, the mean time-off values steadily increased across the duration of the implementation of the protocol.
Discussion
Despite multiple demonstrations of the effectiveness of intensive behavioral toilet training packages, it has been suggested that they might be too effortful (Kroeger & Sorensen-Burnworth, 2009), potentially limiting the availability of such services to families. The present analysis was conducted to illustrate the common implementation experiences and outcomes achieved during intensive behavioral toileting when implemented by family members and school personnel under practitioner supervision. All of the participants were able to achieve improved functionality in continence across multiple settings, ranging from infrequent accidents to complete independence.
If practitioners were to implement the package described herein with similar individuals, they might reasonably expect to achieve a few to many successes and at least one accident in the first several hours of the intensive treatment day. They also should expect parents and teachers to be capable of implementing this package following training that includes pretreatment preparation along with instructions, modeling, and performance feedback during implementation. Even when implementation is imperfect, as we must assume was the case with some of our participants at least some of the time, positive outcomes can be achieved.
A concern that some practitioners may have about intensive toilet training is the inclusion of positive practice as an intervention component. However, if positive practice is implemented using the procedures described here for accident/success conversions, the majority of children should not experience many instances of full positive practice. In the event that a practitioner attempts to dismantle this package by removing the positive practice component, it is important to keep in mind that accidents are always negatively reinforced by immediate bladder relief and that programmed positive reinforcers for successes may be insufficient to produce differential reinforcement without added consequences for accidents.
In conclusion, the Azrin and Foxx (1971) treatment package has been an invaluable tool for improving continence for many decades. Positive outcomes can be achieved under everyday conditions, and they can be achieved more rapidly and with greater success using this approach compared with nonintensive procedures (LeBlanc et al., 2005). The present analysis of archival data illustrates what might be expected regarding treatment process and outcome with children with autism spectrum disorders. Given the increasing availability of behavioral treatment services for this population, families should be in a position to expect their behavioral treatment providers to produce reasonable and relatively rapid improvements in their children’s continence. We hope that the information from our past clinical efforts proves useful in achieving that goal.
Footnotes
Acknowledgements
We thank the many graduate and undergraduate therapists who provided services to these families as part of their practicum training at Western Michigan University.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.