A Qualitative Analysis of Underexplored Barriers to Physicians Seeking Mental Health Care in Medical License Applications

Visual Abstract.

ABSTRACT

Background

Despite increasing mental health diagnoses among physician trainees, fewer than one-third seek help due to stigma and fears of professional consequences. While national organizations have successfully advocated for licensing application reforms, trainees continue to cite licensure as a primary barrier, suggesting that additional deterrents may persist within application structure and content.

Objective

To identify and quantify potential barriers to seeking mental health care within initial medical licensing applications beyond current recommendations regarding impairment-based questions and supportive language.

Methods

This exploratory qualitative study analyzed initial allopathic medical license applications from all 50 US states and Washington, DC at 2 time points: in 2021 and between 2024 and 2025. Using inductive content analysis, authors identified emergent themes representing potential barriers and applied a coding scheme to quantify changes over time.

Results

Five potential barriers were identified: application inaccessibility, unclear health question wording, scrutiny of training gaps, stigmatizing question grouping (mental health questions co-located with illegal/immoral activities), and uncertain privacy protections. From 2021 to 2025, improvements occurred in 4 domains: unclear language (11 to 6 states), training gap scrutiny (35 to 29 states), stigmatizing grouping (27 to 20 states), and privacy concerns (38 to 28 states). However, application accessibility worsened substantially (29 to 18 publicly accessible applications).

Conclusions

Five distinct potential barriers persist in state medical licensing applications. Longitudinal analysis from 2021 to 2025 showed improvements in 4 domains, though application accessibility worsened substantially.

Introduction

Despite increasing prevalence of mental health diagnoses among physician trainees, fewer than one-third seek help, hindered by stigma, confidentiality concerns, and profound fears of professional consequences.^1-3 The consequences of this are tremendous, with suicide as the leading cause of death among residents in the United States, a tragedy that is often preventable with appropriate mental health support.⁴ A key driver of this reluctance to seek care is the fear of professional repercussions, which has historically stemmed from the medical licensing process itself. For decades, medical licensing applications commonly asked probing questions about mental health in ways that are inconsistent with expert recommendations and the Americans with Disabilities Act (ADA).⁵ Research has shown that in states with ADA-noncompliant questions, physicians are more reluctant to seek help for mental health conditions.⁶

In response, many national organizations have advocated for revisions to reduce these barriers. For example, in 2018 the Federation of State Medical Boards (FSMB), an umbrella organization that provides guidance to state medical boards, published recommendations encouraging state boards to limit mental health questions to those concerning current impairment, provide safe haven non-reporting options, and use supportive language.⁷ While progress in aligning with these recommendations has occurred, physician trainees continue to cite licensure as a primary barrier to care, suggesting that additional, underexplored deterrents may persist within the structure and content of the applications themselves.^8-11

This study aims to explore these additional barriers. We conducted a longitudinal qualitative analysis to first identify and then quantify these barriers, focusing on content, structure, and process of initial medical licensing applications in 2021 and again between 2024 and 2025. These findings will provide graduate medical education leadership with specific data to support advocacy efforts with state boards and to better advise trainees navigating the licensure process.

KEY POINTS

What Is Known

Physician trainees often avoid seeking mental health care due to stigma and fear of professional consequences, with licensing applications historically serving as a major barrier.

What Is New

This study examining US state licensure materials identified 5 persistent barriers in licensing applications. There were improvements in language clarity, training gap scrutiny, stigmatizing grouping of questions, and privacy concerns between 2021 and 2025, but a significant decline in application accessibility.

Bottom Line

Despite progress in reducing stigmatizing elements, licensing applications still pose potential barriers to mental health care, underscoring the need for continued reform.

Methods

This exploratory qualitative study aimed to uncover underlying barriers within initial allopathic medical license applications using constructivist principles. This approach assumes that such barriers may be constructed through the interplay between the application’s content and structure and the physician’s interpretation. We employed an inductive approach to allow for the discovery of emergent structural barriers not anticipated by existing deductive frameworks.

We acquired applications from all 50 US states and Washington, DC (hereafter “states”) at 2 time points: first in 2021 (“2021 applications” hereafter) and again between August 2024 and May 2025 (“2025 applications” hereafter), creating a corpus of approximately 850 pages of text at each timepoint. Some states use the Uniform Application for Licensure as a base template but require state-specific addenda with additional questions. Therefore, we separately analyzed the Uniform Application in addition to analyzing the state applications. An inductive content analysis was first applied to the 2021 dataset to identify emergent themes representing potential barriers to care. To focus on novel insights, this analysis excluded issues already well-documented by the FSMB recommendations: questions about history/treatment rather than current impairment, lack of safe haven non-reporting options, and lack of supportive language. Three authors (E.B., K.H., D.S.) independently reviewed a subset of 20 applications and used iterative coding to develop themes, reaching negotiated agreement through group discussions. This process informed a final coding scheme (Table), which 2 authors (K.H., L.H.) then deductively applied to the entire 2021 dataset, with discrepancies resolved through consultation with a third author (E.B. or D.S.) as needed. For the second phase of the study, the established coding scheme was deductively applied to the 2025 applications to quantify any changes. Throughout coding, authors engaged in reflexivity to acknowledge their perspectives as physician advocates, deliberately attending to ambiguous language that risk-averse applicants might interpret as barriers to disclosure. Finally, we mapped the prevalence of these barriers geographically to visualize distribution and changes between 2021 and 2025.

Table.

Description of Inductively Identified Potential Barriers to Help-Seeking and Respective Coding Criteria

Identified Potential Barrier	Coding Criteria/Examples
Inaccessibility of applications: Applications are not made freely and publicly available, requiring applicants to register with personal information to view questions	Coded “0” (barrier present): Application was not fully accessible from a public website and required registration. Coded “1” (barrier absent): Application was completely accessible, such as via a direct hyperlink to a PDF or the ability to click through the full application online without providing personal information.
Lack of clarity in health questions: The wording of questions related to health conditions is ambiguous, confusing, or uses contradictory phrasing, making it difficult for an applicant to know what information must be disclosed	Coded “0” (barrier present): Questions about mental or physical health were vaguely worded or contained contradictory estimators (eg, “could impair” or “may impair”). Example: Do you have any other condition that may in any way impair or limit your ability to practice medicine safely? Coded “1” (barrier absent): Questions were clearly worded, focused only on current impairment, or no questions about mental health were asked. Example: Do you currently have a medical condition that impairs your ability to practice medicine in a competent, ethical, and professional manner?
Scrutiny of gaps in training: Applications require an explanation for any period not spent in training or work, which may include leaves of absence for medical reasons	Coded “0” (barrier present): Application required an explanation for gaps in training or work for any reason. Example: All of your time from high school (not medical school graduation) to the date of application must be accounted for on the Chronology of Activities page... This requirement must be met or your entire application will be returned. Coded “1” (barrier absent): Application asked no questions about gaps in training or work.
Stigmatizing question grouping: Questions related to an applicant’s health history are placed in application sections otherwise exclusively focused on illegal or immoral activities (eg, felonies, misdemeanors, professional misconduct)	Coded “0” (barrier present): Questions about mental health were co-located in application sections that otherwise exclusively asked about illegal or immoral activities. Coded “1” (barrier absent): Questions about mental health were located in a separate section or were not asked.
Uncertainty of privacy protections: Applications provide unclear guidance on the confidentiality of personal health information or require applicants to assent to the release of this information to third parties or as part of the public record	Coded “0” (barrier present): Application stated that submitted records could become part of the public record, required assent to release information to third parties, or was unclear about how information would be kept private. Example: The contents of licensing files are generally considered public records. If you believe that the additional information you are attaching to explain a “yes” answer should be considered confidential, state that in the attachment. A request for confidentiality may or may not be granted. Coded “1” (barrier absent): Application stated that personal health information would be kept confidential and would not be released without the applicant’s consent.

The project was exempt from institutional review board review as it did not involve human participants.

Results

Our analysis identified 5 additional possible barriers within the medical licensing applications that may deter physician trainees from seeking mental health care: accessibility, clarity, gaps in training, question grouping, and privacy. These findings, supported by specific examples and coding criteria detailed in the Table, reveal a complex landscape of potential barriers.

The first barrier, accessibility of applications, highlights the challenge physicians face in predicting what mental health–related questions are included in licensing applications. Some states’ applications are not freely available on public websites, instead requiring personal information to be registered before viewing them. This inaccessibility exacerbates uncertainty, which may heighten reluctance to seek care, fueled by a historical context in which intrusive questions were not merely academic but could trigger board appearances, costly legal defense, and professional sanctions, even in the absence of practice impairment.¹² In 2021, 29 state applications were publicly accessible. By 2025, only 18 full applications were publicly accessible, representing a substantial decrease from 2021.

The second barrier is the lack of clarity in the wording of questions related to mental health diagnoses and care. Ambiguous language complicates the decision-making process for applicants who must grapple with determining what information is relevant to disclose, potentially discouraging them from engaging with mental health services. In 2021, 11 applications used language deemed unclear or confusing. By 2025, the number of states with unclear language was 6. This was due in part to some states eliminating questions related to mental health and instead focusing only on fitness for duty.

The third barrier stemmed from the requirement for explanations for gaps in training. Having leaves of absence as a source of future scrutiny may dissuade physician trainees from taking medical leave for fear of negative career impacts. In 2021, 35 applications required explanations for any training gap. By 2025, this persisted in only 29 applications.

The fourth barrier concerns the contextual placement, or question grouping, of mental health inquiries. These questions were sometimes co-located in sections otherwise focused exclusively on immoral, unethical, or illegal activities. This grouping risks reinforcing negative stereotypes about the professional competency of people with mental health diagnoses. In 2021, 27 applications co-located mental health questions in this stigmatizing manner. By 2025, only 20 boards continued this practice.

Finally, the issue of privacy emerged as a potential barrier. The potential for reported personal health information (even in the absence of impairment) to be disclosed to third parties or to become public may be a substantial concern for physician trainees, which may lead them to forgo treatment to avoid this risk. In 2021, 38 states had unclear or no guidance on privacy, or stated that information could be released. By 2025, only 28 applications continued this practice.

The geographic variability and changes in these barriers are illustrated in the Figure.

Figure.

Geographic Distribution and Change in 5 Potential Barriers to Mental Health Care Seeking in Medical License Applications, 2021 vs 2025

Note: The figure compares the presence of 5 potential barriers within state licensing applications between 2021 and 2025. The barriers shown are accessibility (whether applications are publicly available), clarity (ambiguity of health questions), gaps in training (scrutiny of leaves of absence), question grouping (stigmatizing placement of health questions), and privacy (lack of confidentiality protections). In each map, states colored in black were identified as having the barrier present; states in gray were identified as having the barrier absent.

Discussion

This longitudinal study of initial medical licensing applications identified 5 persistent structural and content-based barriers that may deter physician trainees from seeking mental health care: inaccessibility, lack of clarity, scrutiny of training gaps, stigmatizing question grouping, and uncertain privacy. Our analysis from 2021 to 2025 reveals that while progress has been made in most areas, many of these potentially harmful practices remain widespread and vary significantly between states.

These findings broaden the understanding of how licensure processes can contribute to the underutilization of mental health resources by trainees. Prior research establishes that fear of professional consequences is a primary driver of the treatment gap, with physicians in states with more intrusive licensure questions showing greater reluctance to seek care.^1,6 Recent studies have documented substantial progress in state adoption of FSMB recommendations, yet trainees continue to cite licensure as a barrier.^1,5,11 This study may help explain this persistence by characterizing specific structural attributes of the licensure process—such as inaccessibility and ambiguity—that may serve as potential loci for these fears. For example, our novel insight that stigma may be reinforced by placing mental health inquiries alongside questions about felonies suggests that simple structural changes to application design could have a positive impact. This work moves beyond the content of the questions to include the process and context in which they are asked.

The origins of these barriers are likely varied. For instance, decreasing accessibility may be an unintentional consequence of boards transitioning from paper to online application portals, a problem that could be remedied by providing publicly accessible PDF examples. Other barriers, such as policies on public disclosure of licensure information, may be rooted in state law and fall outside a board’s immediate control. Even in such cases, identifying these issues creates important opportunities for targeted advocacy.

Our findings also complement the important recommendations made by the FSMB and other advocacy groups. While the FSMB has focused on high-level critical principles like using impairment-based language, our analysis adds granularity. Issues like application accessibility and question grouping are not explicitly covered in current national recommendations but represent potential barriers with relatively easy solutions. Addressing these procedural barriers is an important next step to fully realize the goal of creating a licensure system that does not contribute to physicians’ hesitation to seek care.

It is essential to acknowledge the exploratory nature of this research. Our study identifies potential barriers based on a qualitative analysis of documents, but it does not empirically determine if these barriers directly cause trainees to avoid care. Future research, using surveys or focus groups, would help validate the real-world impact of these application features on trainee behavior. Furthermore, this analysis is based only on the elements of the application that were accessible to us. Other components of the licensure process, such as secondary requests for information, reference forms, or board interviews, were outside the scope of this study and could offer further areas for improvement. Importantly, licensure reform addresses only one barrier to care-seeking. Physician trainees continue to face stigma and a culture that normalizes inadequate self-care, fear of professional consequences beyond licensure (such as fear of workplace or professional advancement repercussions), confidentiality concerns, and structural access challenges including scheduling inflexibility and cost.¹

Despite these limitations, our findings have immediate relevance for the graduate medical education community. Program directors, deans, and mentors can use this information to better advise trainees navigating the complex and heterogeneous landscape of state licensure. The results also provide a clear, evidence-based roadmap for state medical boards and policymakers to continue refining applications to support, rather than inadvertently punish, the health and well-being of the physician workforce.

Conclusions

Our exploratory study found that 5 distinct types of potential barriers to mental health care—accessibility of applications; clarity in language of questions; questions about training gaps; question grouping in sections otherwise focused on illegal, immoral, or unethical behaviors; and privacy of medical records—persist in state medical licensing applications. While our longitudinal analysis from 2021 to 2025 showed notable improvements in 4 of the 5 domains, these barriers persist alongside a significant decrease in application accessibility.

Author Notes

* Dr Barrett and Ms Hicks served as co-first authors and contributed equally to this work.

Funding: The authors report no external funding source for this study.

Conflict of interest: The authors declare they have no competing interests.