Provider and leader perspectives on eating disorder screening and the importance of a clinical pathway in the Veterans Health Administration: A qualitative study

Abstract

We explored United States provider and leader perspectives on eating disorder screening, implementation, and clinical pathways in the Veterans Health Administration (VA) using qualitative interviews with ten medical providers and seven leaders from VA’s National Program Offices (N=17). Providers included two frontline primary care medical support staff, two nurses, two primary care providers, one dietitian, one women’s health provider, one health psychologist and one weight management program provider. We asked about the utility of screening for eating disorders and potential implementation challenges. The tool was in development at the time the qualitative interviews were conducted so that feedback could be taken into consideration. Rapid qualitative analysis was used to identify themes. Three themes were identified. First, a gap was identified in eating disorders screening, with acknowledgement that there is not a uniform way to systemically screen patients. Second, most agreed that a brief tool such as the one we developed could be helpful in a large healthcare system, especially a tool with language that was mindful of stigma issues. Third, there was agreement that a screening tool would need to be part of a more extensive clinical pathway that included diagnosis and treatment within the larger healthcare system. While participants supported the need for an eating disorder screening tool, their perspectives varied on how such a tool would be implemented in a national healthcare system. Specific recommendations were made for ensuring that there was a clear clinical pathway from the tool to treatment referral to best serve veterans with eating disorders.

Introduction

Eating disorders are prevalent psychiatric disorders, with at least 30 million people of all ages and genders having an eating disorder in the United States (Hudson et al., 2007). Eating disorders are also a major public health burden and are associated with a range of significant health problems, including elevated mortality and suicide rates (Ágh et al., 2016; Button et al., 2010; Rosling et al., 2011). Despite elevated rates and associated adverse outcomes, most healthcare systems do not systematically screen patients for eating disorders (Maguen et al., 2018). Consequently, eating disorders may go undetected, resulting in insufficient or delayed treatment and, in the worst case, premature death.

There are several barriers to screening for eating disorders in large healthcare systems. First, existing eating disorder screening tools do not perform well in diverse populations. For example, studies have found that the SCOFF (the acronym SCOFF comes from the first letter of words in each of the questions: Sick, Control, One, Fat and Food), the most commonly used eating disorder screener had limited sensitivity, specificity and generalizability across gender, age, weight status and eating disorder diagnoses (Botella et al., 2013; Kutz et al., 2020; Liu et al., 2015; Maguen et al., 2018; Mond et al., 2008; Parker et al., 2005), and performed worst in overweight, older men (Liu et al., 2015), and women over age 30 years (Liu et al., 2015), who comprise a large percentage of healthcare users. Importantly, the SCOFF has not been validated in a veteran population (Kutz et al., 2020). Furthermore, there is evidence that existing screeners underestimate eating disorders among marginalized groups and create bias in diagnosis of patients (Alexander et al., 2024). These gaps have been exacerbated by inclusion of a wider range of eating disorders in the Diagnostic and Statistical Manual of Mental Disorders-Fifth Edition (DSM-5), given that many of the existing screening tools were developed prior to new diagnoses, which capture a larger and more diverse clinical population (American Psychiatric Association, 2013). More specifically, eating disorder patients were historically characterized as predominately young, White, underweight, or healthy-weight girls and women with anorexia or bulimia nervosa (Strother et al., 2012). The DSM-5 inclusion of Binge Eating Disorder (BED), an eating disorder strongly associated with overweight/obesity and present in a greater percentage of men, improved recognition that eating disorders cut across weight status, gender and age (Higgins et al., 2013).

Given that veterans in the Veterans Health Administration (VHA), the largest US healthcare system, are diverse and typically older with a greater proportion of men than the populations in which screening tools for eating disorders were developed, existing screeners may not adequately capture eating disorders in this population. There may also be unique issues that make detection of eating disorders more challenging in VHA. For example, 78% of veterans with overweight or obesity reported binge eating behaviors (Higgins et al., 2013). Existing screeners were developed prior to inclusion of binge eating in DSM-5. Additionally, prevalence rates among more recent war-era veterans may be high (26–31%; Masheb et al., 2021; Vaught et al., 2021), which highlights the need for a tool than can detect disordered eating in a veteran population. Rates of eating disorders among veterans with other mental health diagnoses are also elevated. More specifically, eating disorders are significantly more common in veterans with depression, posttraumatic stress disorder, and alcohol and/or drug use disorders than in veterans without these mental health disorders (Maguen et al., 2012). Consequently, there is another missed opportunity for screening those with existing mental health issues.

Given the dangers of not detecting those with eating disorders in VHA and other healthcare systems, it seems critical to examine screening gaps and potential solutions to fill these gaps. Consequently, our study aimed to interview both providers and leaders in the largest US healthcare system to understand perspectives on eating disorders screening tools, share a five-item screening tool developed by our team called BRief Eating Disorder Screener (BREDS) for veterans (see Masheb et al., 2025 for more information) that has now been published, and investigate views on implementing such a tool within a large national healthcare system.

1. Methods

Participants.

We prioritized recruiting individuals with high levels of “information power,” (Malterud et al., 2016), more specifically, those with extensive relevant knowledge and experience. For providers, researchers recruited two frontline primary care medical support staff, two nurses, two primary care providers, one dietitian, one women’s health provider, one health psychologist and one weight management program provider. Five providers were based at the VA Connecticut and five at the San Francisco VA – the exact balance was due to intentional sampling. National leaders in primary care (trained MD primary care provider), women’s health (PhD non-clinician, MD primary care provider), weight management (PhD psychologist), and mental health and eating disorder treatment (PhD psychologists, MSW social worker) were also recruited. Two providers and two leaders had specific expertise in eating disorders, and the remaining interviewees had exposure to this population by virtue of their roles. All providers and leaders were recruited through an email describing the study’s purpose and interview duration, and all of those recruited agreed to participate. Participants included ten medical providers from the VA Connecticut and the San Francisco VA, and seven leaders from VA’s National Program Offices (N=17). This sample of 17 providers and leaders held adequate information power to reach meaning saturation (Hennink et al., 2017) for our exploration of eating disorder screening implementation.

Procedures.

We conducted 30–45-minute one-on-one in-person or phone interviews, which covered three main topics. First, we explored providers’ and leaders’ professional experiences screening for and treating eating disorders. Screening is for problematic or maladaptive eating, which can help identify individuals who can benefit from more thorough assessment to diagnose an eating disorder and subsequently receive treatment. Second, we solicited general feedback and recommendations for implementing a screening tool for eating disorders in VA, providing them with a set of seven draft screening items during the interview that covered seven eating disorder domains (distress, restriction, purging, loss of control, night eating, self-evaluation, and weight) to prompt discussion. The development of items for this screener was guided by statistical modeling frameworks and interviewer feedback that included patient-centered and culturally attuned adaptations. Given that the screener was in development, we did not solicit specific feedback about the items. Finally, we asked for their general impressions about eating disorders as a health priority within VA. Interviews were conducted by one of the principal investigators (SM or RM), observed by the qualitative analyst (JH or JS), recorded on secure VA-encrypted computers, and then transcribed. All procedures were approved by the University of California, San Francisco Institutional Review Board, the Human Research Protection Program at the San Francisco VA Health Care System, and the Institutional Review Board at the VA Connecticut Healthcare System.

Data Analysis.

Rapid qualitative data analysis was chosen for this project as it is an action-oriented approach to qualitative data analysis used to inform practice (Hamilton & Finley, 2019). The process entails summarizing key points from the transcripts and using matrices to systematically explore relevant themes. The analysis team included six interdisciplinary members with varied eating disorder and weight expertise, led by the study’s principal investigators (RM and SM). The team included two psychologists (RM and SM), one dietitian (LM), one qualitative analyst (JS), and two research coordinators (JH, SS) all of whom were trained by a rapid qualitative methods expert (AH), who also provided guidance throughout the analysis. Four analysts (JS, SS, LM, and JH) performed a rapid qualitative analysis (Hamilton, 2013; Nevedal et al., 2021; Palinkas et al., 2019), independently summarizing interview transcripts into structured note templates (see supplement for transcript summary template) using the interview guide’s domains, including direct quotes under each domain. The four analysts summarized the same three transcripts and participated in a consensus-forming discussion. The summary template was then refined to ensure analysts were well-aligned. All analysts used the refined template to summarize two more transcripts, followed by another consensus-forming discussion. The template was finalized and used to summarize all 17 transcripts. The templated summaries were then consolidated into a matrix (see supplement for transcript summary template).

The research team used the summary matrix to outline emergent themes. All six research team members, including the principal investigators and four analysts, reviewed the summary matrix and independently generated an initial list of themes. The entire research team then participated in a validation and consensus-forming activity that solidified a final set of themes. During this meeting, team members were well-aligned in interpreting the summary matrix, with each of the final themes meeting a threshold for inclusion by appearing in a minimum of five out of six individual summaries. The interview transcripts were revisited to validate final themes and select additional representative quotes.

2. Results

We organized our findings into three themes describing: (1) the gap in eating disorders screening, (2) the utility of an eating disorders screening tool, and (3) establishment of a clinical pathway. Each is described in greater detail below, with a section describing implementation implications following each theme.

(1). The gap: Current eating disorder screening landscape.

Our interviewees shared that it is not common practice across VA to screen for eating disorders in a structured way. One provider acknowledged, “we don’t do a good job screening, like we just don’t” (Provider 2) and another reflected, “it would probably be better if I had some specific idea of the questions to even ask, but I wouldn’t really even have a great sense of where to start” (Provider 3). Another provider discussed not knowing how to bring up the topic of eating disorders: “patients will not—are embarrassed by it, I think. Or other reasons why they don’t want to share it. And so, I think it takes direct questioning, which many of us don’t do regularly” (Provider 6).

Some specialized clinics use existing tools that are not validated in a veteran population (e.g., SCOFF). One leader described the current situation, reiterating the need for an eating disorder screener: for “patients with potential disordered eating, eating disorder symptoms, eating disorders, we don’t make any blanket recommendations because of the issues with having a valid and reliable measure with our population of veterans. So it certainly, it 100% would fill a gap” (Leader 4).

Most commonly, veterans with concerns about eating disorders are referred to mental health or other available resources for additional assessment where a self-report diagnostic measure could be helpful. One primary care provider “would do a warm handoff to the Integrated Mental Health... they’re the ones who have more experience in the structured mental health interviews” (Provider 8). One of the main issues with these more intensive screeners, as described by one leader, is that “it just goes on and on.”

Finally, most providers reported uncertainty about how best to care for patients with eating disorders, like one who said:

“I’m not sure if I’m contradicting or helping or hurting. For the few, the patients that are right now coming to mind, it was hard. It is and it continues to be hard, because I continue to manage them. . . And then I am left with—how do I manage their other typical chronic diseases in light of this eating disorder? And that’s another area of struggle” (Provider 6).

Implementation Implication

The majority of interviewees expressed that implementing an eating disorder screener would fill a gap but would also be a change from current practice. Many leaders described experiences with making change to practice within a large health system as being extremely challenging. As described by one leader, “the biggest thing would be, you know, change is hard. So if you’re changing the way that things have always been, there’s in some ways a little bit of an uphill battle” (Leader 6). To facilitate screening implementation, leaders and providers expressed that evidence demonstrating there is a gap that needs filling would be key. Evidence would help justify the investment in more screening and then more treatment for eating disorders. One clinical leader who is responsible for spearheading screening efforts said, “We need this, we want this. So we are, we have been waiting for the day where the data is enough, where we can say: Yes, here you go” (Leader 6). Backing evidence around the validity of a screener with epidemiologic evidence demonstrating the need to screen more would significantly strengthen the case for widespread implementation across VA.

(2). Filling the gap: Utility of proposed eating disorder screening tool.

Generally, providers and leaders had a positive reaction to the proposed screener, reporting that it was comprehensive and “did not pathologize” (Leader 4) veterans. One leader with expertise in eating disorder diagnosis and treatment mentioned that the screener “is different, in a good way,” stating further, “I don’t think we’re asking those questions. We’re thinking more about the, you know, kind of binging, purging, restricting” (Leader 6). A clinician said, “I think always having the language for how you can ask would be very useful” (Provider 3).

Emphasis on designing the briefest screener possible was a near universal theme. As said by one leader, “the shortest validated measure that you can come up with would be most ideal” (Leader 3).

Implementation Implications

Participants highlighted that existing screener burdens and competing health priorities among veterans make any new screening initiative challenging in the context of the VA. One provider stated, “unless it’s a project or something particular we’re looking for a lot of, um, the check-in staff will not ask anything extra because it is already a timely process” (Provider 5). Or as stated by another provider:

“I think if you tried to get everybody in primary care to use it, nobody would remember to use it because they wouldn’t think it was going to capture enough people to be worth their time. Not their time, like it’s so valuable, but just they have so many more urgent medical problems” (Provider 7).

One strategy to promote implementation suggested by several leaders was identifying specific clinical contexts or high-risk cohorts of veterans where it would be most impactful to screen for possible eating disorders. One leader was quite adamant that, “I would only want to be screening those who are considered a high-risk cohort” (Leader 2). This suggestion highlights the need for more evidence, specifically identifying which patient cohorts might be considered high-risk for an eating disorder. Some potential specific cohorts mentioned by participants included veterans seeking weight management, those with complex trauma, those going through big transitions (e.g., pregnancy), people with intersectional identities, those with a history of eating disorders, those seeking SUD treatment, and particular women veterans. A minority of providers and leaders advocated for universal screening at new patient enrollment.

Another implementation gap identified by providers and leaders was need for more awareness, education and training for providers who would be involved in the screening process. As described by one leader:

“I think there’s a reluctance, and maybe even a fear, like an apprehension of working with eating disorders and not feeling comfortable knowing what to do. So I think it would be harder to implement consistently in folks who don’t have kind of a background, at least to some extent, in eating disorders” (Leader 6).

A provider described the need to raise awareness to promote buy-in among providers: “the case can be made to re-educate so that it does gain importance, but I think just when you bring it up, most people will think this is not a geriatric male issue and we see a lot of men” (Provider 6). Training and education efforts should incorporate the evidence supporting a case for large-scale implementation.

(3). What’s missing: A clinical pathway for veterans who screen positive.

Key to a successful screening process is a clear pathway and understanding of what to do and where to send veterans who screen positive for a possible eating disorder. As described by one provider, “when you think of the stepped model of care of any disease, you have to have those higher steps in place. Otherwise, I just feel like...you open up the can of worms with screening and you don’t have the downstream resources in place” (Provider 1).

Almost every provider and leader identified this ‘next step’ as critical to screener uptake and buy-in among clinicians. One provider expressed:

“Although we don’t see it very often, the times that we do it’s sort of like, okay, what’s—like what are the next steps? Like who’s a, who’s an attending that can supervise this case? Who feels comfortable supervising the case? What are some—what are the best ways for us to support the patient with the most evidence-based treatment for whatever their presenting concern is” (Provider 4).

Participants expressed that any successful implementation initiative must include a procedure for necessary additional assessments that might leverage various team members and, where appropriate, downstream care options. Importantly, the VHA currently has multidisciplinary eating disorder treatment teams that can provide consultation, assessment, and treatment. While current treatment plans may include medications and psychotherapy, more intensive inpatient programs that might be required for some veterans are not available and would require referral to community care.

Implementation Implications

Leaders highlighted that every facility is different and has different resources that could be leveraged to care for veterans identified as having an eating disorder. The clinical pathway would need to be flexible and adapt to a variety of different healthcare environments.

One risk identified by leaders was that resources for eating disorder assessment and treatment may exist but are likely not prepared to receive a large influx of newly identified patients. As explained by one leader, “if you guys are screening for something, you need to have care available for the folks that have a positive screen. And you need to make sure that they get care” (Leader 7). Part of an implementation strategy would need to include ensuring available resources have capacity to receive and care for the additional eating disorder patients that would likely be identified during a new screening initiative.

Providers and leaders agreed that eating disorder screening and care is an opportunity to create connections with existing clinical resources like dietitians and weight management program coordinators. As one provider described, “our other outpatient dietitians, we’ve kind of seen bits and pieces of this over the years, that I think most of us feel comfortable at least—you know, addressing or bridging some of the dietary interventions” (Provider 7). Or as stated by a leader, “We’ve got another team member who can help us out there. So I’m sure there’s even more than that” (Leader 5).

3. Discussion

The VA providers and leaders we interviewed recognized a gap in current systemwide screening, as well as the utility of a standardized eating disorder screen. They appreciated that the screener we developed and shared is comprehensive, non-stigmatizing, and developed for a diverse population and multiple eating disorders. Providers and leaders supported screening of high-risk individuals, indicating that thoughtful implementation would include establishment of a clear clinical pathway consisting of targeted screening, diagnosis, and referral for treatment. This pathway already exists for conditions like posttraumatic stress disorder, depression and alcohol use disorder, where VHA screening have been implemented, and pathways to diagnosis and treatment are clear.

While supportive of a clinical pathway forward, leaders indicated that continued data demonstrating rates of eating disorders and clinical need would be a critical piece to justify implementation of a screener. This type of data collection requires appropriate tools, including both a screener and a diagnostic tool. Now that we have a better eating disorder screener for diverse individuals, there is a need to develop better diagnostic tools as well and ensure that we are using these tools to identify veterans at risk. Given that our prior studies have found that among Iraq and Afghanistan war era veterans, about one-third of women and one-fifth of men reported symptoms consistent with a DSM-5 eating disorder diagnosis (Masheb et al., 2021), there is clearly a need for both screening and treatment in VA. Furthermore, we have found that veterans with eating disorders reported that their eating disorders were overlooked or conflated with overweight, which can delay detection and treatment (Masheb et al., 2024). Veterans also have expressed openness to screening and referral to eating disorder treatment, highlighting a need for both (Masheb et al., 2024). In line with these needs, the VA established the National Eating Disorder Team, which provides training and ongoing consultation for outpatient, multidisciplinary teams that provide medical, mental health and nutrition services, with at least 88 national teams currently offering these services. Given that some infrastructure exists, a validated screener will help with detection of high-risk patients who are eligible for these services and may otherwise go undetected. Additionally, greater access to learning health systems using electronic health record data would be useful for large systems like the VHA, to assess prevalence, pathways, and risks associated with unidentified individuals who need care.

Another critical part of implementation is identifying the best places to screen high risk veterans. For example, mental health clinics could be appropriate, given high rates of comorbidity (Kowalewska et al., 2024; Sharifian et al., 2024); weight management clinics, given that many veterans with eating disorders are referred to these clinics (Masheb et al., 2024); and women’s clinics, given that women have higher rates of some eating disorders than men (Klimek-Johnson et al., 2024), although recent research has found no gender differences in some eating disorders such as binge eating disorder (Klimek-Johnson et al., 2024). Importantly, screening implementation should carefully consider how to capture the most high-risk veterans and those who are not traditionally considered to have eating disorders, including men and those with intersecting identities.

Identifying VA leaders who can help guide screening implementation as well as assist with education of providers is also critical. Providers in clinics where high-risk patients receive care may need additional training and guidance about screeners and how to discuss screening and eating disorders with veterans. For example, as demonstrated in our study, providers identified some reluctance to discuss eating disorders with veterans because they may not have the language or training to do so. Veterans want to be asked about and evaluated for eating disorders, but also want providers to do so in a sensitive and non-stigmatizing way (Masheb et al., 2024). This creates an opportunity for providers to learn and for an important gap to be filled, ultimately benefiting providers and veterans alike.

Several limitations of this study should be noted, including that we interviewed providers from multiple clinics at two VA medical centers. Future studies can assess providers throughout the US, as there may be regional differences in provider beliefs about eating disorder screening and implementation. However, we did interview national leaders in multiple key areas, which provided important perspectives. We also recognize that although the VA system is the largest national healthcare system in the US, our findings do not represent all healthcare systems but can provide some guidance and insight into utility and implementation of eating disorder screeners in hospitals. Additionally, although we tried to sample a diverse group, provider and leader perspectives may vary by level of experience with eating disorders. There may also be sources of bias, such as demand characteristics, although it is important to note that the goal of qualitative research is to provide a snapshot rather than to be representative of the groups interviewed. Finally, although additional training of providers may be needed for eating disorder screening implementation and treatment referral, further research is needed about the types and targets of training that would be most useful.

We found that providers and leaders recognized a gap in systematic eating disorder screening and agreed that the screening tool that we developed could be helpful for identification of veterans at risk of eating disorders. Implementation efforts will require careful planning, identification of clinics where high risk veterans may be found, training of providers, and clarification of a clinical pathway. The screening tool is the first step in identifying veterans that may otherwise be overlooked, and it can assist in detection and referral to treatment of veterans with eating disorders.

Public Significance Statement.

Although eating disorders are prevalent mental health disorders impacting diverse individuals and are a major US public health issue, there are no systematic procedures to screen and treat eating disorders in most US healthcare systems. Providers and leaders in the Veterans Health Administration, the largest US healthcare system, acknowledged that eating disorder screening is a gap and that a clinical pathway is needed to best serve veterans with eating disorder issues.

Data Sharing Policy:

The data that support the findings of this study are available from the corresponding author upon reasonable request. The data are not publicly available due to privacy or ethical restrictions.