Teaching Practitioners to Use the Bristol Stool Form Scale

Abstract

The Bristol Stool Form Scale (BSFS) is an empirically validated and reliable tool used to measure stool health. Stool health may be especially important to monitor among individuals with limited verbal repertoires, as they may not be able to self-report stool health. Further, individuals with autism spectrum disorder are known to have a higher incidence of gastrointestinal (GI) distress. Participants in the present study were staff responsible for the care of individuals with ASD. These staff were trained to use the BSFS to identify stool type. A series of phases were conducted in which participants were shown various pictures of stool. In phase 1, participants were provided with instructions to assign a number based on the chart scoring system (1–7) to each picture. In phase 2, participants were given access to the BSFS. In phase 3, a trainer-facilitated discussion of each stool type was conducted. In phase 4, a full review of each picture was conducted, and feedback was provided for individual errors. Participants were re-tested at the end of each phase. All participants achieved mastery for BSFS use by phase 4, and of the participants who received a follow-up evaluation, this skill was maintained for up to 6 months from the completion of the training. This training took a maximum of one hour to complete, could be conducted individually or in a group, and required very little additional training to deliver effectively.

Children commonly experience gastrointestinal (GI) problems, including constipation (Mutyala et al., 2020; Nurko, 2000), dyspepsia (indigestion), vomiting, aerophagia (Rasquin et al., 2006), and abdominal bloating (Erickson et al., 2005). Children with autism spectrum disorder (ASD) have an increased likelihood of experiencing GI problems compared to children who do not (Mcelhanon et al., 2014). Ferguson et al. (2019) reported that, given a sample of 340 individuals with ASD, 65% experienced constipation, 47.9% experienced stomach pain, and 29.7% experienced diarrhea within their lifetime. Lewis et al. (2016) reported a significantly lower percentage of reported GI problems among a sample of 949 individuals who did not have ASD or any other related disorders (23.1%), with constipation and abdominal migraines being the most prevalent symptoms.

The factors contributing to GI problems in individuals with ASD may also differ from what is commonly experienced by individuals who do not. For example, individuals with ASD are more likely to engage in self-injurious behavior (SIB; Steenfeldt‑Kristensen et al., 2020), and prescribing medications such as risperidone and clonidine is a common pharmacological strategy for reducing rates of self-injury and other topographies of challenging behavior (Buitelaar & Wilemson-Swinkels, 2000; Zarcone et al., 2001). These medications can cause symptoms such as diarrhea, constipation, and nausea by altering the physiology of the gastrointestinal tract (Philpott et al., 2014). Further, psychotropic medications can affect an individual’s appetite and overall diet (Ferguson et al., 2019; Fromson et al., 2012; Ibrahim et al., 2009), which could also increase the likelihood that an individual would experience GI problems (Ibrahim et al., 2009). Additionally, individuals with ASD are more likely to experience inadequate dietary intake (Bandini et al., 2010), which may also result in GI problems. GI-related symptoms may cause pain or discomfort to an individual and may result in increased rates of challenging behavior in children with ASD (Kennedy et al., 2025; McKeown et al., 2022). Thus, not only can the presence of challenging behavior result in GI problems due to treatment efforts involving pharmacological intervention, but GI problems may also exacerbate challenging behavior.

Several studies have evaluated the relation between various biological variables and challenging behavior, such as allergies and sleep deprivation (Kennedy & Meyer, 1996), menstruation (Carr et al., 2003), and cold symptoms (Carr et al., 2003). Christensen et al. (2009) identified that rates of challenging behavior during behavioral assessments were influenced by the presence or absence of constipation symptoms, identified via a series of X-rays. As symptoms decreased, rates of challenging behavior also decreased.

Any suspected relation between biological variables and challenging behavior should be considered during assessment and treatment (Carr, 1994). This point is highlighted by item 2.12 of the Ethics Code for Behavior Analysts: Considering Medical Needs (Behavior Analyst Certification Board, 2022). However, limited access to relevant medical resources and training could serve as a barrier to identifying these relations. For example, although the use of technology such as an X-ray machine has been demonstrated to be an effective tool for measuring GI-related symptoms, access to these tools may be resource intensive (e.g., associated cost of repeated gastroenterology appointments), and is unlikely to be accessible in settings where evaluating GI symptoms is clinically relevant (e.g., when providing outpatient applied behavior analysis [ABA] services). Practitioners in these settings (e.g., Board Certified Behavior Analyst, BCBA®; Board-Certified Behavior Analyst – Doctorial; BCBA-D®; Registered Behavior Technician, RBT®) may not have the proper qualifications to utilize these tools and are not often equipped to monitor and resolve medical complexities for their clients. However, there are more accessible tools that may be more feasible to teach these individuals to use. This is particularly important given the likelihood they will need to establish data collection systems to facilitate interdisciplinary collaboration for a client experiencing GI problems.

A tool to identify GI-related symptoms that may be feasible to use across a variety of different treatment settings is the Bristol Stool Form Scale (BSFS; Lewis & Heaton, 1997; see Supplemental Material for a depiction of the BSFS). This empirically validated tool was designed to provide clinicians with a relatively unintrusive way of measuring stool health and has been regarded by GI physicians as an effective tool for clearly classifying types of stools (Bird et al., 2022). The tool is used by assigning the numbers 1–7 to the stool based on its shape and consistency. The numbers are on a spectrum where the extremes are constipation (1) and diarrhea (7), and a 3 or 4 represents a healthy stool. Such a tool should be considered for use in behavior analytic practice as it is widely used by medical professionals (Blake et al., 2016) and may be required to operate more comprehensive health monitoring systems.

An example of such a system is the one used by Bird et al. (2022) for stool tracking. Bird et al. developed an electronic data collection system for recording BSFS data, and they implemented this system with six participants with documented histories of GI dysfunction who resided in community-based group homes. They found that their system produced high levels of data collection integrity, yielded high social validity ratings, and GI physicians reported that the data helped them when consulting with the participants. Although the authors described their training as simulated behavioral skills training (BST; Vladescu & Marano, 2022) followed by in vivo observations (Maguire et al., 2022; Reed & Henley, 2015), they did not thoroughly describe the contents of the training in a way that can be easily replicated by other clinicians or researchers. Further, they did not describe how they evaluated the effects of the training, training efficiency, or the effects of different training methodologies, all of which can be important considerations for clinical application.

It is likely that practitioners would need to be trained to use the BSFS in practice. Further, making the training as efficient as possible would increase its feasibility. For example, group training has been demonstrated to be effective and is regarded as resource efficient (Parsons et al., 2012). Previous research has used this method to teach individuals how to implement different variations of a functional analysis (Griffith et al., 2020; Iwata et al., 2000; Lloveras et al., 2022). Other methods that have been used to teach a variety of skills include computer-based instruction (Retzlaff et al., 2020), didactic instruction (Wolfe & Slocum, 2015), prompting with differential reinforcement (Young & Daly, 2016), and BST (see Slane & Lieberman-Betz, 2021, for a review). These methods were often implemented in combination with feedback, which has been repeatedly demonstrated to be effective across a variety of different training contexts and is influenced by several factors, including frequency, timing, and preference (Bacotti et al., 2021; Lechermeier & Fassnacht, 2018; Park et al., 2019; Sleiman et al., 2020). In a meta-analysis of performance feedback in organizational settings, Sleiman et al. (2020) found that feedback effectiveness had a consistent effect on performance in 81.25% of the applications, specifically when combined with other treatment components.

A notable exception to the research demonstrating the necessity of feedback is provided by Graff and Karsten (2012), who demonstrated a positive effect of training using enhanced written instruction without feedback. In their experiment, they trained 11 teachers to implement preference assessments. First, they provided teachers with standard written instructions, which did not produce accurate implementation of the preference assessment procedures. Next, they provided teachers with enhanced written instructions, which included pictures, diagrams, step-by-step examples, and minimal technical jargon. The enhanced written instructions produced accurate implementation of preference assessments for all 11 teachers. The success of this intervention is notable because it decreases the amount of time and resources necessary to train new people. We hypothesized that a similarly minimal amount of training may be sufficient for teaching practitioners to accurately collect data according to the BSFS because this tool already includes pictures and minimal technical jargon.

The purpose of the current study is to evaluate the relative effectiveness of different components of a rapid training progression, including visual aids, trainer-facilitated discussion, and feedback to teach individuals to accurately categorize stools using the BSFS. This training was also evaluated across (a) individuals with and without prior experience using the BSFS and (b) group and individual training formats to determine whether these variables influence the effectiveness of different training components within the progression.

Method

Participants and Setting

Twelve individuals participated. Of these participants, nine were RBTs (7 women and 2 men), and three were BCBAs (2 women and 1 man). All participants were between the ages of 20 and 31. Of these participants, six RBTs and one BCBA had prior (but limited) exposure to the BSFS. All participants worked in a day treatment setting delivering ABA services to children with ASD. At the time of the study, all participants were working with a child for whom clinicians had hypothesized that challenging behavior covaried with the presence or absence of GI-related symptoms.

Training was initially provided by a BCBA and BCBA-D who had prior training and exposure to using the BSFS in clinical practice. To evaluate the generality of this training when delivered by an individual with limited exposure to using the BSFS in practice, an RBT who had received this training from one of the original trainers also delivered the training package to a subset of participants. Please refer to Table 1 for additional information related to participant credentials, groupings, and trainers. The effects of the training were evaluated for each participant in a multiple baseline across participants design.

Table 1.

Participant credentials and BSFS history, and trainer credentials

Participant	Credential	Prior exposure to BSFS	Trainer credential
1	BCBA	No	BCBA
2	RBT	No
3	BCBA	No	BCBA
4	RBT	No
5	RBT	No	BCBA-D
6	RBT	Yes	BCBA
7	RBT	Yes	RBT
8	RBT	Yes
9	RBT	Yes
10	RBT	Yes
11	RBT	Yes
12	BCBA	Yes	RBT

Materials and General Procedure

Each participant had access to technology (e.g., laptop, trainer sharing the PowerPoint via Smartboard) that was compatible with Microsoft Teams, Word, and PowerPoint, as well as a paper or digital datasheet. The trainer had access to a separate answer sheet that also contained a bulleted list of discussion points for relevant training phases. The purpose of these discussion points was to control for stylistic differences across trainers, ensuring the same core points were covered across all participants. Trainers also utilized a PowerPoint containing slides with the BSFS, a breakdown of each type of stool, and various pictures of stool equally representing types 1 through 7. Three variants of the training materials were adjusted to account for the differences in the number of pictures presented in baseline.

The pictures of the stool for the training were sourced from the internet and are public domain. Follow-up data were collected on actual stool immediately after a client had defecated or photos of stool sourced from clients within the ABA clinic (with parental consent). It is important to note that in these cases, the children were not independent in toileting, so monitoring was considered essential for toileting and for monitoring stool health due to potential GI distress, interactions with severe behavior, or both. The type of stool for each picture was determined by having the first, third, and fourth authors independently classify each picture. If the picture was found through a reputable website and the stool type had already been classified, agreement with the website-provided classification was also required to achieve consensus. If an agreement could not be achieved, the picture would be discarded. However, this did not occur for any pictures considered for inclusion in this training.

The training was conducted individually (participants 5, 6, and 12), in dyads (participants 1–2 and 3–4), or in a group of five (participants 7–11). There were no procedural differences between individual or group training other than ensuring adequate space between participants to prevent attempts to copy other participants’ answers and discuss answers to individual questions during the administration of a quiz. On no occasion was such behavior observed.

Baseline and Training Phases

In phase 1, or baseline, individuals were provided with a brief introduction, asked if they require any testing accommodations, and provided the instruction to “rank each picture by assigning it a number between one and seven on your data sheet.” A trainer then made the PowerPoint slides accessible to the participants by sharing their screen in a Microsoft Teams meeting or on a Smartboard. The PowerPoint contained seven, nine, or eleven slides. No other information was provided during this phase.

Phase 2 was identical to phase 1, with two exceptions. Each participant was provided with a digital copy of the BSFS for reference (which remained available for all phases that followed), and 14 slides were presented for categorization according to the BSFS. These slides included the same seven to eleven pictures from phase 1 plus five, seven, or nine new pictures. Repeated pictures were quasi-randomized (as was the case for all remaining phases).

During phase 3, re-testing was preceded by a trainer-facilitated discussion of each stool type accompanied by visual samples (one animated representation taken from the BSFS and one picture of an actual stool) of each stool type. Participants were prompted to ask questions at the end of each discussion point. A total of 21 slides were then presented for categorization. These slides included the same 14 pictures from phase 2 plus seven new pictures.

During phase 4, re-testing was preceded by a full review of each of the 21 slides presented in phase 3. This provided participants with an opportunity to discuss individual errors. After all participants had indicated that they had no further questions, the same 21 slides were presented again.

For all training phases, each participant experienced all phases in a single session that only exceeded an hour on one occasion. The group of five took slightly longer than a full hour due to a higher rate of questions from participants in phases 3 and 4.

Data Collection and Mastery Criteria

Answers to each question were summarized as the percentage correct. This was achieved by first identifying how many “degrees” a participant’s answer differed from the correct answer (i.e., the absolute value of the difference between the participant’s response and the correct answer) and dividing that number by six. That number was then subtracted from one and multiplied by one hundred. For example, if an individual classified a type-4 stool as a 1, the resulting formula would be (1 − [(4 − 1)/6]) *100. If an individual answered correctly, the resulting formula would always be [1 − (0/6)] *100. This method captures the magnitude of the error, which is important when considering stool classification. For example, classifying a type-1 stool (constipation) as a type-7 stool (diarrhea) would be calculated as 0% correct, and a medical professional might regard this as a more detrimental error than classifying a type-3 stool (healthy) as a type-5 stool (also relatively healthy), which would be calculated as 67% correct. An individual was considered to have reached “mastery” if no more than two total answers achieved a score of less than 80%, no more than one answer achieved a score of less than 70%, zero total answers achieved a score of less than 60%, and the total number of errors made within a phase did not exceed 25% of the total number of questions within that phase. These stringent mastery criteria were used because, in practice, these data would be used to inform potentially invasive intervention strategies, and it is necessary to minimize potential harms associated with inaccurate stool classification. For example, if a direct care provider scores a client’s stool as a type-6 or −7, the client may be advised by a medical provider to change their diet (e.g., eat bananas, rice, applesauce, and toast) or take medications (e.g., Imodium). Similarly, if the client’s stool is scored as type-1 or −2, the client may be advised to take medications (e.g., stool softener, laxative, suppository) or receive an enema.

Follow-Up

In situ follow-up data were collected for P3 and P4 at two weeks and one month post-training, respectively. Data were collected for a total of six stools recorded for a client with whom both participants worked daily. When the primary therapist had classified the stool, they were instructed to recruit another individual who had received the same training, or one of the trainers if available, to also make an independent classification. The resulting percentage was calculated using the same method described in the previous section.

Follow-up data were also collected for P7–P9 and P10 at six months and P12 at two months. Because the clients the providers currently worked with did not reliably defecate in the clinic or defecated in a diaper, data were collected for two photos of stool from a client who was receiving services in the same clinic. Because stool photos were used to collect these follow-up data (rather than actual stool), it was not considered “in situ”. The resulting percentage was calculated using the same method described in the previous section.

Results

Figures 1 and 2 show the results across all phases for all participants. Each trial on the x-axis represents a single picture of stool that was presented in the accompanying PowerPoint presentation for that phase. The y-axis displays the percentage correct measure. Figure 3 displays the cumulative results for all participants. The x-axis displays each of the four phases that each participant experienced. The y-axis displays the average percentage correct measure for each participant. This was calculated by summing a participant’s scores on each item scored in a phase and dividing the total by the total number of items for that phase. No participant met mastery criteria prior to phase 4, at which time all participants demonstrated mastery.

Fig. 1.

Percentage correct responding for participants 1, 2, 5, 6, 7, and 12. Participants are displayed according to the baseline condition they experienced

Fig. 2.

Percentage correct responding for participants 3, 4, 8, 9, 10, and 11. Participants are displayed according to the baseline condition they experienced

Fig. 3.

Percentage correct responding for all participants who received the training. The individuals who received the training in a group of five are displayed as closed circles. The horizontal lines indicate the average percentage correct for all participants

The average percentage correct tended to increase across phases for all participants. There appeared to be a large increase in average percentage correct from phase 1 to phase 2, followed by smaller increases with the addition of each subsequent phase. These data violated the assumption of normality of residuals necessary to conduct a repeated-measures ANOVA, so we conducted a nonparametric Friedman test to evaluate whether there were statistical differences in average percentage correct across training phases, which showed a significant effect of training level, χ²(4) = 30.30, p < 0.0001. We conducted a follow-up Dunn’s multiple comparisons test to evaluate exactly which training levels differed, which showed a significant difference between phase 2 compared to phase 3, p = 0.0054; phase 1 compared to phase 4, p < 0.0001; and phase 2 compared to phase 4, p = 0.0017. All other comparisons were not significant (i.e., p > 0.05). This suggests that (a) phases 2–4 improved correct responding relative to phase 1 and (b) adding both training and feedback improved correct responding relative to providing the BSFS in isolation.

Figure 4 shows the follow-up data collected for participants at two weeks (P3), one month (P4), two months (P12), and six months (P7–P9 and P11). The x-axis displays the names of the participants for whom follow-up data were collected. The y-axis displays the average percentage correct measure and performance on each individual probe for each participant. Of the participants for whom we collected follow-up data, six out of seven achieved 100% on all probes conducted. One participant’s (P9) performance differed. This individual received training in a group of five, along with P7, P8, P10, and P11.

Fig. 4.

Follow-up data collected two weeks (P3), four weeks (P4), two months (P12), and six months (P 7–9, P11) from training. Open data points show the percentage correct for each individual probe conducted, and the bars show the total average percentage correct

Figure 5 shows a comparison of performance across phases for individuals who did and did not have prior exposure to the BSFS. The x- and y-axes are identical to Fig. 2. Those with no prior exposure to the BSFS tended to perform worse during phase 1 compared to those who had prior exposure, but there was no pattern of differences between groups across subsequent phases. These data met the assumption of normality based on a Shapiro–Wilk test (prior experience: W = 0.8192, df = 31, p = 0.1413; no prior experience: W = 0.9965, df = 15, p = 0.9880), so we conducted a 2 × 4 (prior experience × training level) mixed-method ANOVA to evaluate if correct responding was related to (a) whether participants did or did not have experience using the BSFS prior to the experiment, (b) participants’ training level within the experiment, and (c) an interaction between prior experience and training level. The ANOVA showed a significant main effect of training level, F(1.382, 13.82) = 64.19, p < 0.0001, and a significant prior experience × training level interaction effect, F(3, 30) = 19.58, p < 0.0001. There was not a significant main effect of prior experience, F(1, 10) = 3.422. We conducted a follow-up Bonferroni’s multiple comparisons test to evaluate the interaction effect further by evaluating which training steps were associated with differences in responding across participants with and without prior experience with the BSFS. The follow-up test showed that correct responding differed significantly during phase 1 between those with prior BSFS experience compared to those without prior BST experience, p = 0.0078, but there were no significant differences based on prior experience during subsequent training levels (for BSFS, p > 0.9999; for BSFS + Training, p > 0.9999; for BSFS + Training + Feedback, p = 0.2304).

Fig. 5.

Comparison of performance across individuals with and without prior exposure to the Bristol Stool Form Scale

An error analysis was also completed to determine whether any pictures produced differentially higher rates of errors relative to other pictures. This analysis was completed in response to anecdotal reports from trainers that certain pictures generated more discussion than others due to discrepancies between the “correct” answer and the one the trainee provided. This was done by calculating an overall percentage correct by averaging all participants’ scores for each of the 21 pictures. If errors across participants for a given picture tended to be more substantial or more reliable (or both), the overall percentage for that item would have been significantly lower. When this calculation was completed for the combined results of phases 3 and 4 (the only phases comprising every picture involved in the training), the percentages ranged from 85 to 100%, and there were a total of six pictures for which the majority of participants made an inaccurate stool classification. When this calculation was completed for the results of phase 4 only, the percentages ranged from 91 to 100%, and there were no pictures for which the majority of participants made an inaccurate stool classification. On the basis of these results, it was concluded that there were no pictures that produced a significantly higher rate of errors.

Discussion

Overall, all participants achieved mastery by phase 4, and of the participants who received a follow-up evaluation, this skill was maintained for up to six months from the completion of the training. This training progression produced desirable performance across individuals with varying levels of exposure to the BSFS, was successfully conducted in a group format, and took no more than 90 min to complete. In situations where even the delivery of an hour-long training in a group format may not be feasible for a clinical team, this training demonstrated that providing access to the BSFS as a visual aid may also improve performance. This is consistent with the results achieved by Bird et al. (2022), who were able to produce desired outcomes by providing the BSFS as a visual aid to participants.

Notably, no participants met mastery criteria during phase 1, even if they had prior experience using the BSFS. This finding can be interpreted in several ways. For example, it may suggest that (a) the participants’ prior experience did not include training to similar mastery criteria, (b) the prior experience did not generalize across time or to our novel exemplars, (c) our training was more salient because it was more recent, or (d) our training classified stool according to the BSFS differently from their previous experience. Regardless, all participants showed improved scores and met mastery criteria by the end of all training phases in this study. Although providing participants with the BSFS in phase 2 improved scores relative to phase 1, no participant met mastery in phases 2 or 3, and scores improved significantly following feedback relative to providing the BSFS alone. Collectively, this demonstrates a probable need for training and feedback to produce mastery.

Delivery of this training individually or in a group (of up to five individuals in the current study) had no demonstrable influence on training outcomes. These results suggest that the use of a group training format can be efficacious outside of a standard BST training progression. The credentials held by the trainer (BCBA-D, BCBA, RBT) also had no demonstrable influence on training outcomes, which supports this training progression as a feasible option even when training resources are limited.

With consideration of limited resources, it may be worth evaluating whether these results could have been obtained through an abbreviated training progression. It may be the case that a more efficient training could be conducted by skipping certain training phases in this experiment, for example, by omitting phase 2 entirely. We did not evaluate this or other similar possibilities because we were interested in evaluating whether an even simpler training procedure (i.e., providing the BSFS in isolation) would be feasible, similar to how Graff and Karsten (2012) produced mastery of skills with enhanced written instructions in isolation.

Despite the overall success of this intervention, there are several limitations that warrant discussion. One such limitation is the lack of in situ follow-up data for all participants. Two barriers to collecting these data were cited. First, several of the participants’ caseloads included clients who defecated regularly, but the integrity of the stool was compromised owing to the client only defecating in a diaper. Future researchers should consider including pictures of stool in diapers if the population they serve consists of individuals who do not reliably defecate in a toilet. To our knowledge, this has not yet been addressed in the research literature. Second, several of the individuals who received the training had transitioned to cases in which stool classification was not clinically indicated or the client did not reliably defecate at the day treatment center where they provided clinical services.

Data collection would be required across settings to analyze a client’s bowel habits fully. For example, if the client is not defecating during their ABA sessions, the clinical team may falsely conclude that they are constipated. However, it may be the case that the client reliably has a type-4 stool once they return home in the evening, which is an indication of healthy bowel activity. To be able to collect data on an individual’s stools across all relevant settings, it would be necessary to provide this training, or a similar version, to all stakeholders (e.g., parents, teachers, babysitters). In situations where an individual may not have time to participate in formal training, it may be feasible to have them collect data on the occurrence or nonoccurrence of defecation, perhaps through the development of a scatterplot data collection method (Touchette et al., 1985). Further, it should be noted that simply providing participants with the BSFS without training tended to produce greater than 80% correct responding, so even providing a modified version of the digital health monitoring system similar to the one developed by Bird et al. (2022), which includes the BSFS, could provide useful information as long as it is interpreted with appropriate caution. Although participant credentials did not appear to have an influence on the rate of skill acquisition in the present study, it would also be valuable to determine if this training is as effective when delivered to individuals who do not have an educational or professional history of exposure to similarly structured training environments.

Even if data are being collected across all settings, GI problems consist of symptoms other than diarrhea and constipation (i.e., some GI problems cannot be measured by the BSFS). Teaching individuals to accurately classify stool is not the final step toward a comprehensive method for identifying a relation between GI problems and challenging behavior. For example, future researchers could consider developing similar training to teach individuals to identify other overt signs that an individual may be experiencing GI problems. Instead of pictures of different stool types, they could develop a series of PowerPoint slides containing pictures or short videos depicting other GI-related symptoms such as stomach distension, excessive belching or flatulence, and vomiting or rumination (Horvath & Perman, 2002). For individuals who have the verbal repertoire to do so, it may also be valuable to teach them how to describe symptoms that are more difficult to observe directly (McKeown et al., 2022), such as abdominal pain and heartburn (Horvath & Perman, 2002).

In addition to the need for more research on establishing a method for identification of GI problems, there are several areas for which this information can be used to guide the course of treatment. One such area is evaluating the influence of those symptoms on challenging behavior. For example, repeated measures of both stool type and severe behavior frequency could be maintained and evaluated contextually (e.g., Carr et al., 2003; Carr et al., 2003; Christensen et al., 2009; Kennedy & Meyer, 1996). This was the aim for all clinical teams that participated in this study. However, we do not presently have data demonstrating that improved accuracy of BSFS reporting led to improved interdisciplinary collaboration or changes in medical care that had a positive impact on the clients’ GI health. Future researchers should consider reporting on successful applications of this method to inform the course of treatment (either with respect to resolving the GI-related symptoms or some undesirable behavior influenced by those symptoms).

Author Contribution

The first, third, and fourth authors established rationale for the study and initial methods. The fifth author provided feedback on the aforementioned items and assisted with finalization of methods. The first, second, and third authors trained participants in this study. All authors contributed to the writing of the manuscript. The first author contributed to the writing of all sections, the second author focused primarily on the introduction, third author: results and statistics, fourth author: data display, and fifth author: overall revision and feedback.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Data Availability

The authors confirm that the data supporting the findings of this study are available within the article and its supplementary materials.

Declarations

Competing Interests

The authors declare that they have no known competing financial or non-financial interests that influenced the content contained within this manuscript.