Abstract
Background
Parent coaching is an evidence-based practice for young autistic children, but it is underutilized in lower-resourced community settings like the Medicaid system. Clinicians often struggle to implement parent coaching with low-income and marginalized families, but little is known about which factors influence clinician decision-making processes about providing parent coaching to this population.
Method
This qualitative analysis used the framework method and thematic analysis. We used the Exploration, Preparation, Implementation, and Sustainment (EPIS) framework to identify factors in the clinical decision-making process that community providers use when offering parent coaching to families of Medicaid-enrolled autistic children. Interviews with 13 providers and a focus group with 13 providers were analyzed.
Results
The following themes emerged: (a) Policies drive provider task priorities and affect competing demands; (b) Providers are more likely to use parent coaching when agency leaders monitor parent coaching benchmarks, though this is rarely done; (c) Logistical factors like scheduling and treatment location affect perceived feasibility of using parent coaching; (d) Previous experience or coursework in parent coaching and/or family systems supports the quality of parent coaching implementation; (e) Provider perceptions of “parent readiness” are initially indicated by overt expressions of parent interest.
Conclusions
In the absence of outer-context and inner-context policies, providers have more decision-making power to offer parent coaching based on their own judgments and preferences, which may result in fewer families being offered parent coaching and increased bias related to which families are offered this service. State-, agency-, and clinician-level recommendations are provided for increasing equitable provision of this evidence-based practice for autism.
Plain Language Summary
We explored how providers make clinical decisions when offering parent coaching to Medicaid-enrolled autistic children. The data came from 13 individual interviews and a focus group with 13 providers. We found that there were five main themes that providers discussed: (a) Providers prioritize using their time to follow requirements made by state or agency policies, and are less likely to prioritize other things that are not required by these policies; (b) Providers are more likely to offer and use parent coaching if their agency leaders monitor how often they do parent coaching; (c) Logistical factors like scheduling and treatment location affect how likely the provider is to use parent coaching with a family; (d) Providers with previous experience or classwork about parent coaching and/or working with families feel more comfortable to use parent coaching with families; (e) Providers think parents are more “ready” to do parent coaching sessions if parents show they are interested with specific actions such as directly asking providers for advice about how to work with their child. These results suggest that policies must prioritize this evidence-based practice to ensure that providers are implementing it equitably to all families who are appropriate for parent coaching. We provide recommendations for clinicians and agency leaders/administrators to improve the implementation of parent coaching in lower-resourced settings.
Background
Autism is a neurodevelopmental condition characterized by challenges with social communication and the presence of restricted and repetitive behaviors (American Psychiatric Association, 2013). Most autistic children in the United States access services through publicly funded systems such as the Part C early intervention system, public schools, or community mental health systems. These community settings are frequently under-resourced, with limited available service providers, limited financial resources, and long waitlists for families to access services (Stahmer et al., 2019).
Parent coaching is an evidence-based practice (EBP) that can help caregivers support their autistic children's development in a variety of domains, including social communication (Ingersoll et al., 2016; Kasari et al., 2010), disruptive behavior (Bearss, Johnson, et al., 2015), and adaptive functioning (Scahill et al., 2016). Not only is parent coaching considered a best practice in working with young autistic children (Steinbrenner et al., 2020), but it also has numerous benefits from an implementation science lens. Particularly in medically underserved areas (e.g., rural communities), short-term parent coaching interventions enable fewer clinicians to reach more families in shorter periods of time. Further, because parents can implement the intervention even when the provider is not present (Rocha et al., 2007), children may receive a greater intervention dosage relative to clinician-delivered treatment. Moreover, parent coaching may be especially helpful for families of low-income backgrounds, who often receive fewer hours of early intervention services than higher-income families (Aranbarri et al., 2021).
Despite these potential benefits, parent coaching is infrequently delivered in community settings. One recent study in the Medicaid system demonstrated that nearly half of the children had not received a single session of parent coaching over the course of 6 months, and only 2.7% of children received it at a frequency consistent with evidence-based models (Straiton et al., 2021b); qualitative data also illustrated that the quality of parent coaching may be inconsistent with best practices. Indeed, clinicians report numerous barriers at the family-, provider-, and agency levels that hinder their use of parent coaching with low-income families, such as perceived limited engagement/interest from caregivers, limited training, and competing demands on providers’ time (Straiton et al., 2021a).
Factors that affect clinical decision-making in healthcare settings fall across three main domains: client factors, provider factors, and practice setting factors. Client-level factors that influence clinical decision-making for child behavioral health services often include limited financial resources in the family, child behavior concerns, and level of parent involvement in services (Furlong et al., 2018). Clinician perceptions of family factors (e.g., multigenerational households) heavily influence whether clinicians determine that parent coaching is an appropriate fit for families of autistic children (Tomczuk et al., 2022). Provider-level factors including provider skill level, competency in delivering the intervention or practice, and previous experiences with making similar decisions can also influence clinician decisions to provide the EBP (M. Smith et al., 2008). Moreover, constraints of a provider's practice setting, such as large caseloads, time constraints, and limited financial resources may also affect clinical decision-making (Furlong et al., 2018).
In the present study, we explored factors related to providers’ decision-making process to offer and deliver parent coaching to Medicaid-enrolled autistic children in the community mental health system in Michigan. We used the Exploration, Preparation, Implementation, and Sustainment (EPIS) framework (Aarons et al., 2011) to guide our analysis. The EPIS framework posits that the implementation of an EBP is influenced by various factors across implementation domains, including the inner context (e.g., organizational climate), outer context (e.g., service environment/policies), innovation factors (e.g., how well an innovation or EBP works for various presenting problems for clients), and bridging factors (e.g., purveyors or trainers who support providers in implementing the EBP).
Method
Study Context
The Michigan Medicaid Autism Benefit, which provides applied behavior analysis (ABA) services to Medicaid-enrolled youth diagnosed with autism spectrum disorder (ASD), is a low-resourced and highly stressed system at the intersection of two publicly funded service systems: the community mental health system and the Medicaid system. In this setting, ABA agencies contract with regional community mental health systems to deliver ABA services to Medicaid-enrolled autistic youth. Data were collected as part of a larger, community-partnered mixed methods project examining the use of parent coaching within this system (Straiton et al., 2021a, 2021b). Administrators were interested in partnering with research teams to improve provider use of parent coaching to autistic youth.
Other analyses using data from this project focused on describing how parent coaching interventions for autistic children are delivered in the Medicaid Autism Benefit (Straiton et al., 2021b), clinician perspectives on barriers and facilitators to parent coaching (Straiton et al., 2021a), and parent perspectives on parent coaching (Casagrande, 2021) in this system. One of these analyses utilized the 13 provider interview transcripts used in the current analysis; but focused on barriers and facilitators to using parent coaching in the system. The current analysis focused on factors that affect clinical decision-making about parent coaching and included additional qualitative data from a focus group with 13 providers that had not yet been analyzed.
Present Analysis
In the Medicaid Autism Benefit, parent coaching (also sometimes referred to as parent/family training) was historically an optional service in this system. About 6 months before the study start date, state administrators issued guidelines that all Medicaid Autism Benefit clients should receive at least one parent coaching session per quarter. However, administrators were having difficulty getting clinicians to follow these guidelines. The present analysis focused on which factors affect clinician decisions to provide parent coaching to their autistic clients in this setting.
Survey and Interview Guide
This study was approved by the Michigan State University Institutional Review Board. Participants gave informed consent and were compensated with a $5 Amazon gift card for completing an online survey and an additional $20 Amazon gift card for completing an optional follow-up interview/focus group. The semistructured interview guide included questions about how providers conceptualize and define parent coaching, barriers and facilitators to parent coaching, competing demands on clinician time, agency climate and level of agency support for implementing parent coaching, and past pre-service and in-service training experiences related to parent coaching. The interview guide was created based on feedback from our community partner, an administrator for the Medicaid Autism Benefit. It was pilot tested with research team members prior to use. While we used the EPIS framework to develop deductive codes in our analysis, it was not used in the development of the interview guide.
Interviewer
The first author (D.S.) conducted the interviews and focus group. At the time of interviewing, she was a graduate student in clinical psychology at an R1 institution in the United States with a bachelor's degree in psychology and educational studies, and extensive qualitative research experience.
Setting
Participants were clinicians who deliver ABA services to Medicaid-enrolled autistic children in the Michigan Medicaid Autism Benefit. This was a convenience sample of providers who elected to complete optional interviews or a focus group after completing a survey about parent coaching practices. Thirteen individual interviews were conducted via synchronous videoconferencing. The focus group (n = 13 providers) was held at a school district building before a community presentation about best practices in parent coaching. At the focus group, the interviewer, her academic adviser, other research team members for the community-partnered project, and participants were all present.
Data Collection
The interviewer audio recorded all interviews and the focus group and took detailed field notes. The interviews/focus group lasted approximately 1 h. Data saturation was met after all interviews/focus groups were completed, such that no additional themes came up in the last interview that had not already been discussed. There were no repeat interviews. Due to provider time limitations and high turnover rates in the Medicaid system, transcripts were not returned to participants for comment or correction, nor did participants provide feedback on our findings.
Participants
We used the following inclusion criteria: providers needed to provide or supervise ABA services for clients in the Medicaid Autism Benefit and needed to meet minimum degree/training requirements to bill for parent coaching in the Medicaid Autism Benefit. We planned to exclude any providers who did not speak in English; none were excluded.
Providers were predominantly white (69%), non-Hispanic (62%), and identified as women (73%). Many providers (50%) were Board-Certified Behavior Analysts (BCBAs). See Table 1 for all demographic data. Based on data from the full survey sample (N = 97), most providers worked primarily with school-aged children: 33% worked with children 3–5 years old, 51.5% worked with children 6–10 years old, and 15.5% worked with children older than 10 years old (15.5%). No providers worked with children under 3 years old.
Demographics of ABA Providers
Note. BCaBA = Board-Certified Assistant Behavior Analyst; BCBA = Board-Certified Behavior Analyst.
Qualitative Approach
We used the framework method to structure our qualitative analysis (Gale et al., 2013; J. Smith & Firth, 2011) and thematic analysis as our methodological orientation (Braun & Clarke, 2012). Coding was conducted using MAXQDA 2020 ( MAXQDA 2020 ). To characterize factors that affected clinical decision-making, we applied deductive codes from the EPIS framework (Aarons et al., 2011) derived from a list of EPIS factors and their definitions published in a recent systematic review (Moullin et al., 2019). We coded for outer context factors (e.g., federal and state Medicaid policy), inner-context factors (e.g., agency characteristics), innovation or EBP factors (e.g., fit of the EBP within the setting), and bridging factors (e.g., characteristics of external consultants helping to support implementation). It was difficult to characterize the stage of implementation (i.e., Exploration, Preparation, Implementation, and Sustainment) for this project, as clinicians were already implementing the EBP (though at a very low frequency). Furthermore, the system had not yet determined whether they would use any implementation strategies to support parent coaching use in the future. For this reason, we decided it was most helpful to document EPIS factors that were involved in the clinical decision-making process, rather than focusing on any specific stage of implementation.
In addition to the 16 deductive codes, one additional inductive code was used: logistical barriers (e.g., rural areas were difficult to access for in-home parent coaching services during winter). The coders felt that these logistical factors were not always better explained by existing deductive codes. See Table 2 in Moullin et al., 2019 for all codes and definitions.
Summary of Themes
Coding Process
We took a consensus-based approach to coding, with the first and second authors reviewing the entirety of the dataset, resolving any disagreements by consensus. This approach positions consensus as a methodologically rigorous agreement goal, because the process of qualitative analysis is an inherently interpretive enterprise in and of itself (Mayan, 2016, p. 107; Saldaña, 2015). The first author read through all transcripts and created memos about initial impressions of the data. Once the codebook was finalized, the first author then read each transcript and applied codes line by line, creating memos about any potential theme ideas and how codes related to other codes within the same transcript. Once all transcripts were coded by the first author, the second author audited the first author's coding (Saldaña, 2015), noting any questions and disagreements about the codes that were applied. The second coder was able to view the first author's coding and memos while auditing.
Developing Themes
Consistent with the framework method, coded excerpts were exported and charted in a spreadsheet, such that excerpts with the same codes could be viewed simultaneously. The first and second author independently reviewed the charted data and developed an initial set of themes based on big-picture ideas that emerged from grouping coded excerpts together. The coders then met to discuss and compare their preliminary themes. Initial themes were highly consistent across the first and second authors. Both coders refined the final themes and then worked together to compile illustrative quotations.
Reflexivity
Consistent with best practices in qualitative research, both coders reflected on their own positionality as researchers and clinicians throughout the coding and writing process. This was evidenced in memoing, the audit process, consensus discussions, and written comments on drafts of the manuscript. Coder 1 (D.S.) identifies as a neurotypical, cisgender Afro-Latina and multiracial woman (Dominican and white) from an upper-middle-class family. Coder 2 (K.F.) identifies as a neurotypical, cisgender white woman from an upper-middle-class family. Both coders were masters-level clinicians pursuing a PhD in clinical psychology, with a clinical focus on parent coaching interventions for autistic youth.
Results
We developed five themes that span the system, agency, provider, and family levels. These themes mapped onto EPIS constructs, including the outer context (service environment/policies, client characteristics) and inner context (organizational characteristics, leadership, fidelity monitoring and support, individual provider characteristics). Themes are outlined in Table 2 and detailed below.
Theme 1. Policies Drive Provider Task Priorities and Affect Competing Demands
In our qualitative analysis, we found that outer-context (e.g., state level) policies drive provider task priorities and affect competing demands on their time. This finding involves interactions across EPIS constructs, including outer context (service environment/policies) and inner context—leadership (competing priorities). In Michigan, state policy mandates that 10% of all direct ABA service hours provided through the Medicaid Autism Benefit be directly supervised by a Board-Certified Behavior Analyst or Qualified Behavioral Health Provider (master's-level clinician with some coursework in ABA). Providers consistently reported that being in compliance with their required supervision hours was their top priority to maintain their contracts within the Medicaid Autism Benefit. Parent coaching, by contrast, is not mandated by state policy, and this was reflected in service provision. One participant described: I know with some other agencies that I have worked with, they do make it a requirement and they have a billing requirement where you have to get in so many billables to meet your salary. […] The agency I work for right now, our big requirement is that we are staying in [Medicaid] compliance as far as doing regular assessments, getting in our regular supervision, but because parent training doesn't really have a requirement attached to it, that kind of makes it where that difference comes into play. Obviously, they would like it to occur and want us to get in as many services as we can to make it more effective, but it doesn't really [have] a requirement – either it happens, or it doesn't happen. – Participant EG0104
In the focus group, one provider noted that an outer-context change might be necessary to shift provider priorities and increase the implementation of parent coaching at a broad scale: “It's almost like policy needs to change or something.”
Theme 2. Providers Are More Likely to Use Parent Coaching When Agency Leaders Monitor Parent Coaching Benchmarks, Though This Is Rarely Done
Although parent coaching is not mandated by state policy or other outer-context factors, some providers talked about how agency-level policies can encourage the provision of parent coaching. This finding involves interactions across EPIS constructs, including inner context—quality and fidelity monitoring/support (monitoring benchmarks) and inner context—leadership (leadership practices of the agency leader). This occurred in the absence of outer context—service environment/policies that focused on minimum requirements for parent coaching sessions. One provider explained: We track currently about five different goals. One of them is family training per month, and [agency leaders] are very efficiently tracking whether or not I can dock at least one parent training once per month with each of my clients. And I can have more than that, but… there's literally the monthly report every week with goals for the four-week period and it tells you your percentage of parent training for the month and where you’re expected to be. And you know, we have corporate summit each month and it's a big discussion. They’re on board with the parent training aspect. – Participant LW0609
Another provider described that there was a minimum agency-wide requirement, such that “it really is not optional” to provide parent training, but noted that the frequency varied: “Each family is different, and they choose the amount of parent training they would like”—Participant RR0518. Providers highlighted that when agency leaders consistently monitor parent coaching benchmarks, it increases the use of parent coaching. This monitoring could take the form of a formal minimum session requirement, or it could be more informal, such as an agency norm to include parent coaching in treatment plans at the agency. One provider stated: I think as far as [support for parent coaching] goes… [the agency is] very positive, very supportive of family training, and to almost the extent of requiring it to occur with every family. So, at this point, [it's] with the discretion of the BCBA [Board Certified Behavior Analyst] but it's becoming more and more the standard of family training goals being included in the plan. – Participant ZD0402
However, perhaps because of a lack of outer-level context policies around parent coaching, very few providers reported that their agencies had formal monitoring requirements in place, with most providers commenting that parent coaching was not prioritized due to competing demands that had formal compliance requirements (as seen in Theme 1).
Theme 3. Logistical Factors Like Scheduling and Treatment Location Affect Perceived Feasibility of Using Parent Coaching
This inductive theme focused on logistical challenges related to the use of parent training in this Medicaid setting. Parent coaching was more likely to occur when providers could schedule sessions immediately before or after other services or supervision, with participants explaining that it was helpful to “just piggyback at the tail-end of that [other service appointment] so the families don't feel like they have to juggle a bunch of different appointments”—Participant EF0216. Relatedly, providers noted that parent coaching is easier to do when working one-on-one with families in their homes, as the caregiver is already physically there, as compared to clinic settings: “For my clients that I have that are center-based, that's pretty difficult”—Participant EF0216.
In the focus group, three providers described logistical barriers such as having limited time to schedule appointments in the early evening when caregivers were more likely to attend, typically due to caregiver work schedules. When asked what support they would need to do more family training, they replied: [Person 1] “Cloning myself;” [Person 2] “More time in the day;” [Person 3] “Making between 3:00 to 6:00 a whole lot longer.” Given the incompatibility between many clinics’ operating hours and families’ routines, many single parents and working families may be unable to access regular family training sessions without taking time off from work.
Theme 4. Previous Experience or Coursework in Parent Coaching and/or Family Systems Supports the Quality of Parent Coaching Implementation
This theme focused on the EPIS construct of the inner context—individual characteristics (previous provider training experiences). Many ABA providers noted that they received little-to-no formal training in working with caregivers, and instead tried to learn these skills through informal means: “I just kind of taught myself, kind of googled ABA parent training. And then from some different blogs from some different BCBAs [Board-Certified Behavior Analysts], I pieced together what I would like to do, but what actually happens is completely different”—participant EF0216. Unsurprisingly, providers noted that having previous training or coursework in parent coaching or family dynamics was helpful during parent coaching sessions. Providers in the focus group discussed: [Person 1:] [Other providers will] say, like, “How did you know how to talk to that parent?” or “Do you just wait for them to say something?” I guess it's from my counseling degree? I don't even know. It's not to say other people don't have that natural ability.
[Person 2:] But I think when you get the social work and counseling degrees, that's so focused on and beat into your head, it like comes very natural.
[Person 3:] And there are actually some behavioral approaches, like motivational interviewing, that we could train in more. Because, you know, there are behavioral approaches.
Indeed, providers with a background in mental health fields (e.g., social work, counseling) felt more competent at working with caregivers. Yet most providers did not have any formal training in family systems, mental health, or parent coaching, and felt this was a barrier to successfully working with parents.
Theme 5. Provider Perceptions of “Parent Readiness” Are Initially Indicated by Overt Expressions of Parent Interest
This theme focused on provider perceptions of the EPIS construct of the outer context—client characteristics. Providers described feeling that some parents were “ready” for parent coaching, while others were not. Provider perceptions of parent readiness were often determined by clear and overt expressions of interest in parent coaching, such as attending appointments regularly without missing sessions, closely observing direct ABA sessions without being asked, being attentive and undistracted during sessions, following through on provider recommendations, and making direct requests to the provider for psychoeducation or advice (see Table 3). Our analysis suggests that these readiness behaviors are an integral part of providers’ clinical decision-making process, such that providers would deem parent coaching to be a less appropriate fit for a family if the parents did not demonstrate enough of these readiness behaviors. Additionally, some hesitation around offering parent coaching was also noted, with providers mentioning the desire to have sufficient time to build rapport prior to offering the service. Many providers described feeling that families become defensive or uncomfortable about having a clinician critique or adjust their parenting practices.
Illustrative Quotations Describing Provider Perceptions of Caregiver “Readiness” for Parent Coaching
Conclusions
Drawing from the five major themes from provider interviews and the focus group (see Table 3), we developed a working model of the clinical decision-making process for offering parent coaching in lower-resourced community settings (Figure 1). This model is informed by the EPIS framework and illustrates the multi-level factors that affect providers’ decisions to offer parent coaching services. In the absence of outer-context and inner-context policies, providers have more leeway in decision-making around delivering parent coaching, which may result in fewer families being offered the service overall and may introduce bias with respect to which families are offered this service. While some provider judgments may be informed by best practices in the literature, legitimate logistical barriers, or client preferences, their judgments are also made based on previous provider experiences and personal beliefs that impact who receives parent coaching. Indeed, a study of speech language pathologists found that while 90% of participants reported that their previous clinical experiences were important in making clinical decisions, far fewer reported that their graduate practica (39%) and coursework (34%) were important in those decisions (Stronach & Schmedding, 2019). This suggests that providers may rely on their personal experience over EBP guidelines in their decision-making process.
Fishbone Diagram Describing Barriers at Multiple Levels That Can Impact Clinical Decisions for Offering Parent Coaching in Lower-Resourced Community Settings
At the highest level, the strong incentive provided by outer context policies affects providers’ propensity to engage in different types of service delivery. When asked about task prioritization, providers in our sample described that mandated tasks consistently took precedence over optional services. In the absence of outer context policies around parent coaching, the agency implementation climate has the potential to facilitate provider use of parent coaching. Our data provide preliminary evidence that when agencies systematically monitor parent coaching hours, providers may offer it more consistently to their families. Although few agencies were monitoring benchmarks, such an approach is consistent with the “audit and feedback” implementation strategy, which has been shown to facilitate clinician behavior change (Jamtvedt et al., 2006). Future research should examine how the implementation of an audit and feedback system related to providing parent coaching supports increased implementation.
In the absence of outer- and inner-context policies incentivizing parent coaching, provider-level factors are emphasized in decision-making around offering parent coaching services. Providers’ sense of self-efficacy around parent coaching—often informed by their preservice training background in best practices about parent engagement and parent coaching models—affects their propensity to offer parent coaching. In line with this, providers discussed wanting more training in clinical skills to support parents (e.g., cognitive-behavioral therapy techniques). This suggests an underlying need for ABA providers to receive training in basic clinical skills for engaging parents, such as reflective listening and collaborative goal setting. However, it also reflects a need for increased education around the function, purpose, and boundaries of parent coaching; directly addressing parent mental health needs is outside the scope of practice for behavior analysts serving autistic youth. Instead, providers should feel comfortable providing resource support and referrals for other family needs.
Furthermore, providers’ training background may affect the quality of parent coaching they provide. Some preliminary analyses from another qualitative project from within the Michigan Medicaid Autism Benefit suggest that caregivers are more interested in parent coaching if high-quality coaching behaviors (e.g., caregiver practice with coach feedback) are implemented, as opposed to psychoeducation alone (Casagrande, 2021). Providers who are skilled in parent engagement and/or training caregivers on how to use intervention strategies may receive more positive feedback from caregivers, as the caregivers may be more likely to experience a strong caregiver-professional alliance and higher caregiver satisfaction with the training process. On the other hand, families who receive poor-quality parent coaching from providers with underdeveloped skills might disengage from or lose interest in the service. Thus, high-quality coaching may indirectly affect provider perceptions about which families are interested in, or willing to, participate in parent coaching.
Lastly, provider perceptions of family-level factors also affect the implementation of parent coaching services in the Medicaid system. Providers in our sample informally assessed caregiver readiness and used this assessment to determine when and to whom to offer parent coaching. Results suggest that providers use these readiness behaviors to assess EBP fit, which is an EPIS innovation factor that describes the extent to which an intervention meets the needs of the population served and/or the service context (Aarons et al., 2017). Clinician views of EBP fit, which are affected by factors at the inner and outer contexts of implementation, heavily influence their clinical decision-making (Palinkas et al., 2018). Our analysis suggests that this readiness assessment is mainly focused on overt expressions of caregiver interest.
Our interpretation of Theme 5 is that the clinician assessment of caregiver (lack of) readiness behaviors is likely to be systematically affected by oppressive social systems, particularly in lower-resourced contexts. For example, providers noted that missing sessions, limited availability and attentiveness, lack of follow-through, and a more hectic home environment (e.g., interruptions from other household members; Table 3) indicate lack of interest in parent coaching. However, for many families, attendance at sessions might mean missed income or losing a job with inflexible hours. Additionally, in the United States, multigenerational and extended family households are especially common among many minoritized cultural groups and in low-income families (Cross, 2018; Pilkauskas et al., 2020). Moreover, various systemic barriers and stressors experienced by low-income families who qualify for Medicaid (e.g., financial strain, shift work hours) likely decrease their ability to attend sessions regularly, follow through on recommendations, and provide undivided attention during sessions.
Our results suggest that providers interpreted the lack of overt expressions of caregiver interest as being driven by the internal motivation of the caregivers, rather than considering how systemic disadvantage might contribute to parent behaviors. Providers often use these potentially biased interpretations to make judgments about the fit of parent coaching for low-income families; if they determine that fit is poor, they are then less likely to offer parent coaching—even if it is an appropriate fit for the client's intervention needs. This finding is consistent with a recent study investigating provider perceptions of family fit for parent coaching for autistic children in publicly funded early intervention settings (Tomczuk et al., 2022). Tomczuk et al. (2022) demonstrated that providers held beliefs that exacerbated disparities in access to parent coaching, with many providers believing that low-income families are more difficult to coach due to poverty, multigenerational households, and cultural/linguistic differences. Given the consistency of these findings across our two samples, we echo the call from Tomczuk and colleagues to deploy equity-focused implementation strategies to improve the use of parent coaching with minoritized and low-income families. For many providers, additional training and support is needed to address this area of potential bias; we provide provider recommendations in Table 4. Likewise, in Table 5, we suggest some policy- and agency-level implementation strategies and recommendations to improve the use of parent coaching, informed by implementation strategies from the ERIC project (Powell et al., 2015; see Table 5).
Clinician-Focused Recommendations to Improve Family-Centered Care and Provide Equitable Access to Parent Coaching
Agency and Policy-Level Recommendations to Improve Family-Centered Care and Provide Equitable Access to Parent Coaching, Informed by the ERIC Project (Powell et al., 2015)
To our knowledge, this is the first investigation of clinical decision-making practices about offering parent coaching interventions for autism. Our analysis has a strong foundation in the EPIS implementation science framework and utilized rigorous qualitative methods. However, there are several limitations. Though our sample size of 26 providers is rather large for a qualitative analysis, all providers were working in a single state and discussed parent coaching specifically within the Medicaid Autism Benefit in Michigan. Thus, it is possible that results may not generalize to other lower-resourced community settings or the Medicaid system in other states. Perspectives of providers working in lower-resourced settings in low- and middle-income countries are also likely to differ from those reported here. Future studies should investigate to what extent this working model of clinical decision-making for offering parent coaching (Figure 1) is relevant in other lower-resourced contexts in the United States and globally. We also did not include interview questions that asked clinicians to delineate which factors in Figure 1 took precedence in their clinical decisions, or whether certain factors were more important than others. Future studies should directly ask clinicians to describe the order in which these factors affect their decisions to offer parent coaching, perhaps working off of our initial list in Figure 1. Additionally, it is possible that our sample may not be representative because it was a convenience sample. Though we offered the follow-up interviews/focus groups to all 97 providers who completed the provider survey for the larger community-partnered project, 27% (n = 26) agreed to participate in the follow-up interviews/focus groups.
Nevertheless, it appears that our sample was indeed representative of the 97 providers in the larger project in terms of demographics and self-reported level of parent coaching extensiveness (a measure of frequency and quality of parent coaching use); see Table 1 in (Straiton et al., 2021a). However, the full sample in this project was a convenience sample of those who opted to participate in a survey about parent coaching, which introduces the threat of response bias. Lastly, this study specifically focused on provider perspectives. While analyzing family perspectives and administrator perspectives was beyond the scope of this analysis, future work should explore these perspectives. Preliminary analyses from qualitative data on families within the Michigan Medicaid Autism Benefit indicate that families are very willing to make the time commitment necessary for parent coaching if it includes an active learning component (e.g., in vivo feedback from a provider about the caregiver's use of intervention strategies), despite time constraints and other barriers facing this population (Casagrande, 2021).
Overall, this analysis suggests that outer-context factors (e.g., state prioritization of other services) and inner-context factors (e.g., agency leaders monitoring parent coaching benchmarks for all providers) affect the provision of parent coaching for autistic children served within low-resourced settings. Our working model suggests that the likelihood of biased decision-making increases when state and agency policies do not prioritize or support parent coaching provision, leaving providers to determine when and to whom to offer parent coaching. Administrators at the state- and agency levels should instate policies that encourage providers to offer parent coaching to all families if they want the use of this EBP to increase in an equitable manner within lower-resourced settings. At the same time, major gaps in provider training must be addressed to help support them in engaging caregivers and providing high-quality parent coaching.
Footnotes
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
Diondra Straiton's and Kyle Frost's time spent working on this manuscript was supported in part by the National Institute of Mental Health (F31MH127814, PI: Straiton) and the Eunice Kennedy Shriver National Institute of Child Health and Human Development (F31HD103209, PI: Frost) of the National Institutes of Health.
