Primary Care Pediatricians’ Perspectives on Autism Care

We conducted a series of semistructured interviews with pediatric PCPs across 8 different clinic sites at a large midwestern academic medical center. We chose to recruit only PCPs within this system for purposes of planning an intervention and to control for variability in the accessibility of ASD referral and treatment services. The study was deemed exempt from review by the University of Michigan institutional review board.

We recruited 28 pediatric PCPs to take part in semistructured interviews to explore experiences and needs around identification and management of ASD in toddlers. All 65 pediatric PCPs at our institution were eligible and invited to take part via e-mail and announcements at faculty meetings. PCPs were asked to contact the study team if interested in sharing their “experiences with identification, referrals and care coordination of young children with suspected autism.” As is recommended in qualitative research,²⁴  we purposively sampled to capture diversity in PCP gender, years in practice, and clinic characteristics. Twenty-nine PCPs expressed interest, and all were invited to complete interviews. One interested party was unable to participate because of persistent scheduling conflicts. Thematic saturation was achieved after approximately one-half of the interviews, but we continued to collect data to examine new perspectives from interested PCPs.

PCPs were interviewed via password-protected Zoom sessions. After providing informed consent for audio-recording, participating PCPs were asked a series of semistructured questions to elicit opinions about current screening and referral procedures, experiences obtaining diagnostic and treatment services for patients, perceived needs for effective communication with parents, the medical decision-making process when concern for ASD is identified, emotional experiences discussing ASD concerns with stressed families, and perceived capacity for ASD management in primary care (see interview guide in Supplemental Table 5). Interviews were transcribed via Scribie. Participants received a $30 gift card for participating.

Transcriptions were reviewed by a multidisciplinary team of researchers with backgrounds in developmental behavioral pediatrics (N.H., J.R.), general pediatrics (S.D.), and speech-language pathology (C.C.). Using grounded theory methods, an approach that allows the thematic structure to arise from the data rather than superimposing a predetermined theoretical model,²⁵  researchers iteratively analyzed transcripts for recurrent themes. Meetings were held biweekly for 3 months to discuss findings, refine interview questions, generate a final list of themes with representative quotes, and determine when thematic saturation was reached. As is recommended for validity in qualitative research,²⁴  results were triangulated with groups of PCPs and feedback was solicited.

Characteristics of participants and their clinic sites are summarized in Table 1. Participants were 89% women, ranging in clinical experience from less than 1 year to >27 years with representation from 8 diverse clinic sites.

We identified 7 themes relating to PCPs’ cognitive, emotional, and environmental experience of caring for patients with autism, which reflected the trajectory of autism management and related medical decision-making. During initial ASD identification, themes included the “the parent lens” (theme 1), and “the ‘A’ word” (theme 2). During the diagnostic and referral process, several roadblocks were described, falling into themes of the “autism labyrinth” (theme 3), “provider disempowerment” (theme 4), and “parent activation as a source of inequities” (theme 5). Finally, when looking at treatment, follow-up, and long-term management, themes included “the ‘black box’” (theme 6) and “provider emotional investment” (theme 7).

Providers in our sample reported using standardized developmental screening instruments to identify initial concerns, as well as a combination of parent-initiated versus provider-initiated conversation, which is an approach consistent with guidelines. In navigating these early conversations, providers described several factors, outlined as themes 1 and 2 in Table 2 with example quotations, which complicated their diagnostic decision-making and communication with families.

One factor perceived equally as a barrier and facilitator to early ASD recognition and referral was the unique experience and background of the family, or “the parent lens.” PCPs described parents having different perceptions of their children based on several factors including previous exposure to other typically developing children/peers, exposure to family and/or friends with autism, cultural background, fear of stigma, and inconsistent or inaccurate perceptions of ASD. When describing perceived barriers to early identification of autism, providers noted that some parents with less exposure to children do not notice problems or express any concern until the PCP raises the issue. Conversely, providers acknowledged that other parents, who have more experience or education around childhood development, recognize potential delays early and are the ones to bring concerns to the pediatrician’s attention. The parent lens emerged as an important theme that seems to be considered by providers when discussing initial concerns and providing follow-up recommendations to patients.

Another challenge PCPs described is when to introduce the possibility of autism, or the A word as some providers phrased it, into conversation. Although some providers circumvented the specific discussion, instead focusing on the general areas of delay or social a typicality, others made a point of using direct language and introducing the A word early in their conversation with families. Whether a PCP chose to use the A word seemed to be the result of both personal style and perceived family hesitancy. PCPs often felt confident in their identification of ASD concerns and need for referral, but communication around these concerns was complicated, and providers had differing opinions as to whether a direct or indirect approach is appropriate across contexts.

PCPs generally reported significant frustration with the diagnostic process, citing long wait times and the disparity between the advice given (ie, early intervention is important) and the way the system is set up (ie, intricate diagnostic processes and long waitlists). The complicated system of services and referrals is labyrinthian in design, making it hard for medical providers and patient families to navigate their way to a diagnosis. Themes 3, 4, and 5, summarized with associated illustrative quotations in Table 3, describe some of the challenges experienced by providers and families around the time of diagnosis.

Providers expressed dissatisfaction and frustration with the complicated process that patients must go through to obtain an autism diagnosis that is accepted by their health care plan. Providers acknowledged the system was equally complicated from both the provider and patient perspectives. They also described the apathy that seems to eventually result for some families when they face unreasonably long wait times and thus a perceived lack of urgency from the health care system. The autism labyrinth seemed to impose a high cognitive load on PCPs, right at a crucial decision-making point, which seemed to influence how PCPs reacted to ASD concerns. Some PCPs even avoiding making referrals for formal diagnostic evaluation when a concern is first identified because of the complicated system and the frustration they know it can cause families.

Another cause of frustration for PCPs seemed to be the imposed limitations of being a general practitioner. Despite a high degree of confidence in identifying ASD symptoms during well child encounters, PCPs expressed a sense that the established system of care prevented them from making a diagnosis, even in the most straightforward cases, and so hindered their ability to help their patients make progress. Watching and supporting patient families while they waited to see a subspecialist led to feelings of helplessness for PCPs. Providers expressed a general lack of knowledge on how to assist families in navigating autism care, despite asserting that autism is a lifelong condition that should be managed within the primary medical home just as any other chronic condition.

Providers identified meaningful disparities between ASD care for children whose parents were highly concerned and motivated to seek a formal diagnosis versus those who did not agree with the PCP assessment or faced competing stressors that made it difficult to follow through with referrals. Providers described needing to put in more energy and practice more patience when parents resisted diagnosis, which in some cases led to feelings of futility. One provider described “parental buy-in or their connection to the diagnosis being a barrier in those early discussions.” Part of the underlying challenge seemed irrevocably linked to the fact that increased activation was perceived by providers to lead to earlier diagnosis and treatment initiation, and improved outcomes. Numerous providers expressed concerns that it was the families who already faced more significant barriers to care, or who came from more disadvantaged backgrounds, that were more likely to be slow to activate.

Providers expressed some uncertainty when making referrals, not only for diagnostic evaluation as discussed previously, but also for follow-up and treatment services. Despite feelings of frustration and confusion, providers expressed an overwhelming sense of emotional investment in their patients. Themes 6 and 7 are summarized alongside illustrative quotes in Table 4.

The black box is a term that multiple PCPs used to express a lack of knowledge and general confusion about how systems of care work, such as the Michigan insurance system, the autism evaluation referral network, and the educational and community resources that are meant to support children with ASD. Not surprisingly, confusion seemed greatest when placing referrals to providers outside of their health system; however, even when making internal referrals, providers expressed a lack of confidence that they were sending patients to the right place.

Throughout every interview, providers expressed great emotional investment in their patients. Providers consistently conveyed empathy and a general sense of accountability for patient experience and outcomes. Despite general frustrations, feelings of helplessness over external referrals and an overloaded system, and challenges with delivering complicated news to families, providers described an overall sense of hope for children in whom they have identified concerns for autism because they have witnessed the success of effective therapy.

This study provides an in-depth look at the experiences of general pediatricians at a large academic medical center providing ASD care within the primary medical home. We describe 7 themes that summarize the cognitive, emotional, and environmental challenges and considerations facing PCPs when caring for patients with autism and their families. Our study contributes to the literature by providing new insight into the emotional experience of PCPs when providing autism care and navigating barriers, as well as a greater appreciation of the complexities surrounding the medical decision-making process.

The themes described in our study were interesting in that they reflected shared experiences among participants, despite diverse community practice locations and a broad range in years of experience. Nearly all providers acknowledged the parent lens and described the challenges of being able to accurately assess a child’s social-emotional abilities, while understanding the parent’s unique interpretation of the child’s development, during limited appointment times. This theme tied closely to the second theme, the A word, in that the decision to use the word autism in those early discussions seemed partly dependent on the PCP’s consideration of the parent lens. In cases in which parents acknowledged ASD as a concern, providers felt comfortable being more direct when discussing autism as a diagnostic possibility, regardless of potential stigma around the term. Conversely, PCPs who did not use the A word worried they could lose family engagement or trust if autism was mentioned before parents were ready, in which cases PCPs felt it important to ease families into acceptance of possible autism over a longer period.

The delicacy of these initial conversations suggests a high cognitive load for pediatricians when identifying and discussing developmental concerns with families. PCPs must find ways to deliver clear messages to families that incorporate medical knowledge, clinical experience, and personal understanding of the child, while being sensitive to parent feelings and recognizing the ambiguities inherent to the ASD screening process.²⁶  Previous research describes that providing an autism diagnosis can be complicated in that there are both negative and positive impacts, with the positive implications of diagnosis generally outweighing the negative ones.²⁷  Our study adds to existing research by delving into some specifics around how PCPs internalize this complexity, as well as the added complexity of difficult-to-navigate systems of care, and allow it to shape their clinical approach.

PCPs perceived disparities between ASD care for children whose parents were more readily activated to pursue diagnosis and treatment versus those who were slower to activate. Based on PCPs’ reflections, parent narratives of a child’s behavior seemed to directly affect initial buy-in and family motivation to pursue evaluation. Given the complexities of a labyrinthian system and the dependence on a series of black box referrals, parent activation was described as an essential component of timely access to ASD assessments and services.

ASD systems of care consist of multiple siloed entities such as third-party payers, birth-to-3 agencies, private diagnostic and treatment clinics, and educational systems.²⁸  When PCPs describe such systems as black boxes into which they cannot see to help patients navigate care, provider disempowerment and inequities in patient outcomes are likely to result. Although Michigan has its own complicated labyrinth of ASD care (eg, varied and complex insurance requirements for treatment authorization; special education law that does not accept medical ASD diagnoses), similar complexities exist in other states. Therefore, the themes described here have relevance for pediatric providers aiming to address barriers to ASD care by demystifying local black boxes and labyrinths in their communities. In particular, the cognitive and emotional experiences of PCPs facing similar challenges with ASD care are likely generalizable, and could be addressed through simplified identification, referral, and treatment pathways.

As has been observed,²⁹  the forced abrupt shift to telehealth during the pandemic provided a unique opportunity to reassess how autism evaluations can be done in a more inclusive, flexible, and individualized manner. Additionally, several examples exist for successful implementation of tiered autism screening and evaluation systems in primary care settings.^30,31  Such models emphasize increased PCP training around autism diagnosis, treatment, and related communication with families, as well as the need to build community partnerships, which may help to increase transparency when referring to external sources. Research demonstrates the importance of family-implemented interventions,³²  which could be optimized in settings that provide a sense of community, such as joint therapy appointments and pediatric offices. Future research could examine whether providing ASD supports in the medical home or community helps physicians feel less anxiety using the A word and leads to increased parent activation, which PCPs in this study identified as a source of inequity.

Our study was limited to the perspectives of medical providers, and, thus, cannot comment directly about the patient-perceived experience. Because PCPs were all employed by a single academic medical center, findings may not be directly applicable to other nonacademic medical practices. Though we did engage in purposive sampling, the majority of participants were self-selected so our study population may be biased toward a group of physicians who felt passionate about this topic and/or frustrated by the existing system.

Existing literature and results from this study suggest a need for targeted improvements in primary care to: (1) increase PCP capacity and confidence in communicating concerns of ASD with families; (2) refer for timely diagnostic evaluations in trusted and easy-to-navigate care systems; and (3) increase family motivation and capacity to follow up on ASD evaluation and treatment services.

We thank Lynn T. and Stuart C. White for financial support for this project.

ASD

autism spectrum disorder

PCP

primary care physician