Barriers to Timely Diagnosis of Eosinophilic Gastrointestinal Diseases

Abstract

Although eosinophilic gastrointestinal diseases (EGIDs), including eosinophilic esophagitis (EoE), have been described over the past 2–3 decades, barriers to diagnosis and treatment are common and compounded by issues related to social determinants of health, race, ethnicity, and access to care. These barriers contribute to delays in diagnosis resulting in persistent inflammation in the gastrointestinal tract that can have significant consequences, including fibrostenotic complications in adults, failure to thrive in children, and decreased quality of life in all affected patients. In this commentary, we summarize gaps in knowledge regarding the epidemiology of EGIDs, highlight barriers to diagnosis, discuss potential approaches based on best practices in other atopic and chronic gastrointestinal diseases, and provide recommendations for reducing barriers to timely diagnosis of EGIDs in underserved populations.

Introduction

Eosinophilic gastrointestinal diseases (EGIDs) are a heterogeneous group of chronic diseases characterized by symptoms related to gastrointestinal (GI) dysfunction and excessive eosinophil accumulation in the affected GI organ(s). First described as a distinct clinicopathologic entity in the 1990’s (1, 2), eosinophilic esophagitis (EoE) is the most common EGID, with an estimated prevalence of 57 cases/100,000 persons (3), presenting with symptoms of dysphagia and food impaction in teens and adults or with less specific symptoms, such as abdominal pain, vomiting, and failure to thrive, in younger children (4). Non-esophageal EGIDs, including eosinophilic gastritis (EoG), duodenitis (EoD), and colitis (EoC), are rarer, with as few as 6 cases/100,000 persons for eosinophilic gastritis (EoG), 8 cases/100,000 for eosinophilic gastroenteritis (with more recent nomenclature defined as eosinophilic enteritis and duodenitis - EoN/EoD), and 3 cases/100,000 for eosinophilic colitis (EoC), (5, 6) and generally present with vomiting, abdominal pain, and diarrhea (5, 7). Since a reliable peripheral biomarker has yet to be identified, the diagnosis of EGID continues to rely on endoscopic assessment and mucosal biopsies (8).

Current data suggest a delay in EoE diagnosis of 1.2–3.5 years in children and 3.0–8.0 years in adults (9). Similarly, the diagnosis of EoG/EoD is delayed by 3.6 years (10). Left untreated, EGIDs may result in complications including growth disturbances, fibrostenosis and obstruction, gastrointestinal bleeding, anemia, hypoproteinemia and hypoalbuminemia (9, 11–15). While recent studies have defined the breadth of clinical phenotypes, as well as potential genetic underpinnings (16–19), there still remains great uncertainty when it comes to identifying who is at risk for developing these complications. Thus, early diagnosis and institution of therapy is critical for improved outcomes.

Despite development of diagnostic criteria for EoE, there is growing recognition of barriers to diagnosis and initiation of treatment. For non-esophageal EGIDs, consensus diagnostic criteria were only recently published for children (20), and are still in development for adults. Generally, the diagnostic criteria for EGIDs require symptoms of esophageal, gastric, or intestinal dysfunction, eosinophil counts exceeding a histologic threshold, and exclusion of other causes of tissue eosinophilia. Table 1 lists recently suggested histological thresholds for children with non-esophageal EGIDs. Identified barriers in diagnosis and management include persistent lack of awareness of disease characteristics, lack of access to subspecialty care, and barriers to care resulting in loss to follow up (21–23). As with other diseases, issues related to social determinants of health also contribute to delays in diagnosis. Here, we evaluate how (1) patient demographics, (2) clinical presentation, (3) social determinants of health, and (4) limitations of current diagnostic approaches may contribute to diagnostic delay; and discuss strategies to overcome these challenges.

Table 1.

Suggested histological thresholds for peak eosinophil counts for diagnosis of non-esophageal EGIDs in children(20)

EGID	Consensus threshold (peak eosinophil count/HPF)
Eosinophilic gastritis	≥30
Eosinophilic enteritis	≥50 (duodenum) ≥60 (terminal ileum)
Eosinophilic colitis	≥100 (cecum, ascending colon) ≥80 (transverse, descending colon) ≥60 (rectum, sigmoid colon)

Descriptive epidemiologic evidence for EGIDs

The descriptive epidemiology of EGIDs is best assessed with population-based data sources providing details on both the demographic characteristics of EGID cases and the source population from which cases were identified. Identifying cases within these population-based sources can be challenging as many data sources rely upon case definitions that optimize specificity over sensitivity, with reliance on billing codes or other administrative documentation for determining case assignment (24). This issue was recently discussed in a systematic review of the demographic distribution of EGIDs (25).

To date, much of the descriptive epidemiologic evidence has focused on EoE, with only two publications describing the sex, race, and age distribution of prevalent cases of non-esophageal EGIDs (5, 26). These studies confirm that non-esophageal EGIDs are rare conditions, but offer little insight into their racial, ethnic, age, or sex distribution. For race and ethnicity, substantial missing data in these administrative data sources precludes examination of the distribution of non-esophageal EGID cases as compared to the source population represented in the dataset (26). For sex, the data presented indicate no clear evidence of sex differences in the distribution of non-esophageal EGIDs. Since the cases identified in these studies represent prevalent cases and not incident cases, the age distribution is indicative of the distribution of cases at a single point in time. This limits inferences on age as a contributing factor in disease diagnosis.

For EoE, numerous population-based studies drawn from large administrative data sources, have described a male sex predominance (3, 27–29). These studies also reported that the racial distribution among patients diagnosed with EoE skews towards a higher proportion of White individuals (27, 28). Still, as with the non-esophageal EGIDs, there are challenges in making any inferences about EoE and race. The existing data, similar to that for non-esophageal EGIDs, are relatively crude in their assessment of race and ethnicity, with data either missing outright, or lack clarity in how the data were collected (e.g. self-reported race and ethnicity or assigned race and ethnicity) (25). Increasingly, the research community has become aware of the importance in distinguishing race and ethnicity from ancestry, with the former likely representing the “lived” experience, as opposed to true biological difference driven by genotypic variability (30, 31).

An assessment of genetic ancestry, as compared to self-reported race, identified self-reported race as a stronger predictor of allergic sensitization, potentially underscoring a greater contribution of environmental factors to disease development (32). Given the strong co-occurrence of allergic disease in patients with EGIDs, and likely shared pathophysiology for disease development for many patients, these findings underscore the need to collect self-reported race in EGID studies, but also critically appraise how social constructs and inequity can contribute to observed differences in disease incidence.

Demographic and comorbid factors influencing time to diagnosis of EGIDs

Diagnosis of EGIDs remains challenging, with many patients remaining undiagnosed or diagnosed very late in the course of their illness. This may be due to demographic features (e.g. age, sex and disparities due to race or ethnicity), cultural, or socioeconomic factors. Alternatively, phenotypic differences including non-specific symptom presentation and allergic comorbidities may also influence diagnostic delay.

Age

The gap between onset of symptoms and diagnosis in patients with EoE is age-dependent. Older patients have a longer time gap (9, 33), which cannot be explained by detection of symptoms alone. Indeed, most adults and adolescents present with overt esophageal symptoms (e.g., dysphagia and food impaction); whereas, symptoms in very young children may be limited to feeding difficulties (34). Therefore, additional factors likely contribute to diagnostic delay. In the context of prolonged disease, adults may be more likely to employ adaptive eating behaviors (e.g., lubrication of foods, excessive chewing) to mitigate dysphagia and prevent food impactions (35). Age was also found to be a predictor of diagnostic delay in patients with EoG/EoD (10); however, the factors associated with diagnostic delay in adults are still unknown.

Sex

Despite variability in reporting among studies (33, 36, 37), it appears that time to EoE diagnosis is longer in males vs. females. Data from a retrospective multicenter cross-sectional study of patients with EoE in the United States (US) (n=793) revealed variability in phenotypic presentation by sex to a small degree (36). Male patients were more likely to have esophageal strictures on endoscopy at presentation and longer symptom duration. In a single-center retrospective study of adults with EoE (37), males reported symptoms of fibrotenosis (e.g., dysphagia, history of food impaction) at diagnosis, while females more commonly reported symptoms of inflammation (e.g., chest pain, heartburn). A Swiss registry of 266 adult patients with EoE (38) similarly found adult males experienced longer diagnostic delay (62 vs. 36 months), greater disease activity, and a higher frequency of esophageal subepithelial lamina propria fibrosis than females.

Race

Previous demographic studies of EoE suggest an association between White race and diagnostic delay despite a more typical clinical presentation. In a large, US multi-site, single-visit registry of children and adults with EoE (n=705) (33), White patients (n=620) experienced longer diagnostic delay despite presenting with more typical features of EoE [e.g., dysphagia and endoscopic features of EoE (i.e., rings and furrows)] (36). In contrast, Black patients (n=50) were more likely to have non-specific symptoms such as heartburn and regurgitation. In the same study, comorbid atopic dermatitis was more common in racially-minoritized groups and was associated with a shorter time to diagnosis; potentially explaining race-associated differences (33). Importantly, atopic dermatitis was also more common in younger patients; therefore, EoE may have been underdiagnosed in older non-atopic individuals from racially-minoritized backgrounds. Reports of race-related differences in diagnostic delay should be interpreted with caution, as most existing epidemiologic studies of EoE are not population-based investigations of diverse communities, and are, thus, subject to detection and inclusion biases.

Comorbid allergic disease

The presence of allergic comorbidities can influence the timing of diagnosis for patients with EGIDs. In EoE, a history of food allergy or atopic dermatitis was associated with a significantly shorter time to diagnosis (median time from symptom onset to diagnosis 1.1 vs. 2.0 years in those with vs. without history of food allergy, and 1.0 vs. 2.0 years in EoE patients with vs. without history of atopic dermatitis) (33). Surprisingly, the opposite was found in a US population-based study of patients with EoG/EoD, where atopic comorbidities were among the predictors of diagnostic delay. In that study, atopic comorbidity factors that independently associated with >2-year diagnostic delay were allergic rhinitis/sinusitis (OR = 1.88, P < 0.001), asthma (OR = 1.45, P < 0.001), urticaria (OR = 1.55, P = 0.003), and allergic conjunctivitis (OR = 1.42, P = 0.021) (10). However, many patients with comorbid allergic diseases were not followed by an allergist before diagnosis, suggesting allergic symptoms may not have been considered during the diagnostic evaluation.

Provider practice patterns influencing time to diagnosis of EGIDs

Wide variability in clinical presentation due to demographic and clinical characteristics can contribute to gaps in provider practice patterns, leading to misclassification and under-diagnosis of EGIDs. In a study of patients with EoE in the United Kingdom, diagnostic failure or delays (mean 4 years, range 4 months-30 years) were caused by a variety of factors including delayed request for endoscopy in patients with dysphagia, poor recognition of EoE endoscopic features, and failed recognition or recording of eosinophil numbers on esophageal biopsies (39). These gaps in provider practice patterns were also reported in patients with EoG/EoD in a recent population-based study of 4,108 patients identified in a US-based administrative claims database (10). A 3.6-year time gap from symptom presentation to diagnosis was found. Factors contributing to diagnostic delay included delayed referral to a gastroenterologist, delayed upper endoscopy, and lack of biopsy procurement and/or histopathologic evaluation. Further complicating recognition of non-EoE EGIDs is the absence of consensus guidelines for clinicopathologic diagnosis in adult patients, and the very recent publication of consensus guidelines in children (20). Moreover, unlike EoE, where symptoms are usually chronic, patients with non-EoE EGIDs may experience a relapsing and remitting course (40).

Geographic and socioeconomic barriers to EGID diagnosis

Geographic barriers

In 2014, Jensen et al. examined the geographic distribution of EoE cases in the US using a national pathology database (41). Among 14,381 cases of esophageal eosinophilia, >97% of which were adults, EoE was more prevalent in areas of lower population density. This finding led to the hypothesis that unique environmental factors in rural areas (e.g., agricultural products or coarse particulate matter) may contribute to EoE. A few years later, McGowan et al. examined the prevalence and geographic distribution of EoE in the US 2012 pediatric Medicaid population (42). In this cross-sectional study of 18,452,886 children, living in a rural area was protective for being diagnosed with EoE (OR 0.68; 95% CI 0.59–0.78, p<0.001). This finding conflicted with the findings by Jensen et al. (41), prompting investigation of whether the observed protective effect of rural residence in the pediatric population was attributed to detection bias from decreased access to care (43).

Geographic information on pediatric gastroenterology providers in the US in 2012 was obtained from the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition database (43). The distance from these providers to individuals enrolled in Medicaid in 2012 was calculated and included in the previous EoE multivariable logistic regression models. After accounting for distance to provider, living in a rural area was no longer protective, and children who lived farther from a pediatric gastroenterologist were less likely be diagnosed with EoE (22). These data strongly suggest under-recognition and under-diagnosis of pediatric EoE in areas with lack of access to specialty care. Further research is needed to assess whether these diagnostic disparities are present in those with commercial health insurance or in adults.

Financial barriers

In addition to geographic barriers to specialty care, financial barriers to EGID diagnosis have been identified. In a case-control study of children seen at a tertiary medical center, individuals with EoE were more affluent and more educated than control populations (44). However, after further analyses, race and ethnicity were found to be significant confounders of this relationship because White individuals were more likely to live in affluent areas and be diagnosed with EoE. In 2015, Kim et al examined a population-based cohort in Southern California (86% adult cases) and found EoE was more prevalent among households with higher annual income (45). This study, however, did not assess differences by gender, race, or ethnicity, and the authors hypothesized that health-seeking behaviors could explain their socioeconomic findings.

Most recently, in the same US 2012 pediatric Medicaid population described above, residences with higher neighborhood-level poverty were found to be protective for EoE diagnosis, despite adjusting for other potential confounders such as urban-rural status, gender, race, or ethnicity (42). These results suggested that either there are unique environmental factors in impoverished environments that are protective for the diagnosis of EoE, or that this finding is due to under-diagnosis or under-recognition of EoE. When further adjusting for distance to care, this protective effect persisted, suggesting that there are unique factors associated with poverty that are distinct from decreased access to care due to geographic barriers (22). Some potential factors include the inability to take off work for appointments, access to specialty providers who accept Medicaid insurance (46), and distrust in medical providers.

In 2020, Adkins et al. performed a population-based, cross-sectional survey of US adults to determine the prevalence of dysphagia and factors that influenced health care-seeking behaviors for this condition (47). Using the validated PROMIS questionnaires to screen for dysphagia, they observed that 4,998 of the eligible 31,139 participants (16.1%) reported this symptom. Among those patients, only 51.1% had sought medical care, despite 92.3% reporting dysphagia symptoms within the previous week. Factors associated with seeking medical care included older age, male sex, having a place the subject usually goes to when sick or needing health advice, having health insurance, having other medical comorbidities, and more severe symptoms of dysphagia. These data suggest that many individuals with dysphagia are not seeking medical evaluation, and that this may be impacted by socioeconomic factors. In addition to the development of less-expensive and less-invasive techniques for diagnosing and monitoring EoE, outreach efforts are clearly needed to improve awareness and education about dysphagia and EoE.

Barriers to diagnosis: Summary and implications

As described above, barriers to diagnosis of EGIDs fall into two main categories: 1) under-recognition of EGIDs by healthcare providers and patients and 2) lack of access to diagnostic testing. Factors that contribute to under-recognition of EGIDs include non-specific symptoms at disease presentation, inadequate familiarity with EGIDs among patients and health care providers, and misconceptions about the epidemiology of the disease. This last factor may contribute to healthcare disparities in EGIDs, as it can affect whether the diagnosis is considered in a patient deemed to be “low risk”. A lack of access to diagnostic testing, is similarly uneven in its effect on different population groups, disproportionately affecting patients with financial or logistical constraints (e.g., inadequate insurance, lack of childcare, inability to take time off from work, language barriers) and those without access to subspecialty care due to distance and/or transportation issues. Finally, additional cultural and socioeconomic factors (e.g., mistrust of the healthcare system) may prevent some patients from seeking care altogether (48–50).

Each of these factors can contribute to a delayed EGID diagnosis. This delay can lead to ongoing inflammation resulting in complications (e.g. esophageal stricture (13, 51), failure to thrive (52). Equally important, uncontrolled symptoms may lead to behavior modification, anxiety (33), social isolation (53) and or severe dietary restriction. These consequences may be more pronounced in patients with limited food choices or poor access to dietary counseling due to socioeconomic factors.

Next steps: What have we learned from management of other diseases?

While barriers to EGID diagnosis are important to identify and address, future directions need to focus on disparities in management.

Access to medical food and pharmacological therapies

As reported among patients with atopy and other chronic gastrointestinal diseases, healthcare-related disparities affect patients with EGIDs beyond diagnosis. Black and Hispanic patients with food allergies have higher hospital admission rates and greater disease severity (54). Additionally, Black children may have significantly decreased access to allergen-free foods than White children (55). Given similarities among disease states and barriers to care, best practices in food allergy can be applied to address disparities in EGIDs. The Food Equality Initiative in Kansas and Missouri addresses food insecurity in individuals with conditions that require specialized diets (e.g. food allergy, celiac disease) (56). Over a 3-year period, this program provided food to over 100 patients in 3 counties (56). Obstacles for obtaining medical foods for patients with EGIDs also persist (57). Even though amino acid-based formulas are medically prescribed, food allergies and EGIDs are not included in federal legislation mandating insurance coverage. Thus, even after an EGID diagnosis, access to medical food is a barrier to care. In addition, potential disparities in access to pharmacological therapies for EGIDs, including biologics, will need to be addressed in the future because of variability in medical insurance coverage and cost.

Multidisciplinary approach

Low socioeconomic status is a well-established social determinant of health disproportionately associated with adverse healthcare outcomes. Leibel et al. investigated the effect of a multidisciplinary approach to severe asthma in children with low socioeconomic status (58). The intervention consisted of validation of diagnosis, assessment of comorbidities, and optimization of treatment with follow-up care every three months. Those with Medicaid were eligible for at-home visits, consisting of screening and mitigation of environmental triggers. Emergency department visits decreased by 75% and duration of hospitalization decreased by 73%. A similar intervention could be useful in EGIDs (e.g., home visits by a dietician). To provide optimal care for EGIDs, care must be coordinated between primary care physicians, allergists, gastroenterologists, and dieticians. Individuals with low SES and chronic disease are disproportionally subject to transportation barriers in healthcare (59), and consolidating care may improve follow up. In rural patients with inflammatory bowel disease, videoconference follow-up appointments were studied in the Scottish Highlands. Eighty-eight patients had 229 appointments via videoconference, saving patients over 70,000 minutes of travel time and over $30 per appointment in travel costs (60). While telehealth may disadvantage some because it requires internet/computer access (Table 2), consolidating care through technology decreases cost and increases convenience.

Table 2.

Barriers to diagnosis and potential solutions

General	Specific Issues	Potential Solutions
Lack of disease recognition	Non-specific symptoms Inadequate awareness among health practitioners Biased estimates in epidemiologic studies Lack of uniform diagnostic criteria for non-EoE EGIDs	Educational outreach Research studies with data sources inclusive of sociodemographic determinants Development of consensus guidelines for diagnosis of non-EoE EGIDs for all age groups
Limited access to diagnostic procedures	Lack of insurance coverage Long distance to specialist care Logistical restraints (e.g., childcare, transportation) Language barriers Lack of internet/computer access	Partnership with primary care physicians Transportation vouchers to address distance to specialist care Additional after hours or weekends openings to accommodate logistical restraints Use of in-person interpreters if available Use of telehealth appointments and videoconferences when accessible and technologically feasible Development of easy to implement clinical guides to screen for EGIDs in the primary care setting Development and validation of less-invasive, affordable tests to diagnose EGIDs in the specialty care setting
Other cultural and socioeconomic factors	Mistrust of health care system	Community outreach

Conclusions

In summary, there are multiple factors that delay EGID diagnosis, many of which disproportionately affect underrepresented populations. Similar barriers have been overcome in other disease states using innovative strategies. Further research is needed to assess the benefits of such interventions in EGIDs. To facilitate better recognition and assessment of EGIDs, pragmatic clinical guidelines should be implemented in the primary care setting and less invasive and affordable diagnostic techniques are needed in the specialty care setting. Overcoming barriers to diagnosis and management is integral in improving the care of all patients with EGIDs equitably.

Abbreviations used:

EGID

eosinophilic gastrointestinal disease

EoC

eosinophilic colitis

EoD

eosinophilic duodenitis

EoE

eosinophilic esophagitis

EoG

eosinophilic gastritis

GI

gastrointestinal

US

United States