Delayed Diagnosis and Racial Bias in Children with Genetic Conditions

Abstract

As more therapeutics for genetic conditions become available, the need for timely and equitable genetic diagnosis has become urgent. Using clinical cases, we consider the health system-, provider-, and patient-level factors that contribute to the delayed diagnosis of genetic conditions in pediatric patients from minority populations, leading to health disparities between racial groups. We then provide suggestions to address these factors, with the aim of improving minority health and access to genetic care for all children.

Introduction

Genetic diagnoses that are identified early in the course of disease can lead to more timely treatment, limit unnecessary medical interventions, and improve both patients’ and family members’ quality of life (Reiff et al., 2012; ACMG Board of Directors, 2015; Rosell et al., 2016). Similar to other medical subspecialties, however, access to medical genetics care and appropriate genetic testing is unequally distributed among patients of different racial backgrounds and geographic locations (Bohnhoff et al., 2019; Penon-Portmann et al., 2019). These health care disparities may contribute to delayed genetic diagnosis in non-white patients. These potential disparities in medical genetics care persist against a historical backdrop of eugenics and discrimination based on social constructs, which may further complicate families’ willingness to obtain and receive care (Holtzman & Rothstein, 1992; Epstein, 2003). As more therapies for genetic conditions become available, the need for front-line clinicians to recognize features of genetic disease and ensure that all patients have equitable access to timely genetic diagnosis and management has become increasingly important.

Health disparities can delay genetic diagnosis at multiple timepoints. Prior reviews have suggested that racial and socioeconomic disparities may affect rates of referral to genetic specialists, attendance of appointments, and both the application and diagnostic yield of genetic testing (Fraiman & Wojcik, 2020). However, specific case examples in medical genetics and actionable recommendations for improvement have not yet been suggested. In this case series, we apply Kilbourne’s model of potential determinants of health disparities within the health care system to illustrative cases within medical genetics (Kilbourne et al., 2006). For investigations of health disparities, Kilbourne et al. (2006) have proposed a framework that considers contributing factors at one of three levels: the health system, provider, and patient (Kilbourne et al., 2006). Health system disparities refer to the role of health care organizations, medical education, funding, and delivery. Provider factors include personal experience, attitudes, implicit biases, and clinical constraints of individual clinicians. Patient factors include both a patient’s race and ethnicity, as well as their culture, beliefs, preferences, education, and access to resources. Taken together, all of these factors influence the delivery of equitable health care and patient experience, broadly within the medical system, and as highlighted in this report, more specifically within medical genetics care.

Here, using three clinical cases, we consider the health system, provider and patient factors that contributed to delayed genetic diagnoses in pediatric patients from minority populations. We then provide suggestions for addressing these factors to improve children’s access to genetic care.

Materials and Methods

Editorial Policies and Ethical Considerations

This study is exempt from IRB approval as it includes only three cases. A medical geneticist who participated in the care of each participant provided de-identified clinical information for each case with permission from the participant’s family.

Results

Patient 1: Health System Factors

An African American boy repeatedly presented to his pediatrician with parental concern for progressive skin darkening over the course of two years. His pediatrician assumed this finding was normal in individuals with darker skin. He later presented with hypoglycemic seizure, prompting a genetics evaluation, and was ultimately diagnosed with a neurodegenerative disease, X-linked adrenoleukodystrophy (X-ALD) (OMIM ID: 300100).

Patient 2: Provider Factors

A two-year-old African American girl was referred by her pediatrician to a neurologist for evaluation of proportionate, non-progressive microcephaly. The neurologist ordered a brain MRI which demonstrated a structurally normal brain, prompting referral to a medical geneticist for further evaluation. She presented to clinic with her mother who was 42 years old at the time. On exam, she made limited eye contact, was nonverbal, and had repetitive stereotyped behaviors. She additionally had multiple dysmorphic features consistent with Down syndrome that had not previously been documented in her medical record. A clinical diagnosis of Down syndrome (OMIM ID: 190685) was made at that visit, later confirmed by karyotype analysis, with concern for concurrent autism spectrum disorder.

Patient 3: Patient Factors

A Filipino boy presented to his pediatrician at four months old with motor delay and hypotonia. He was referred to a biochemical geneticist and a laboratory evaluation was completed, which was non-diagnostic. Liver biopsy was completed at 2 years old and suggestive of a glycogen storage disease. Sequencing of AGL and GBE1 were unrevealing. He continued to receive care from a biochemical geneticist until 7 years of age, when he was lost to follow-up. Similarly, he had been following with specialists in neurology, gastroenterology and physical therapy but stopped seeing all of these clinicians by 10 years old.

At 13 years old, he presented to the emergency room with pneumonia and was found to be severely underweight, with BMI 4% and weight at the 50%ile for a 7-year-old. Biochemical genetics was consulted and recommended whole exome sequencing, which revealed a pathogenic variant in PHKA2, consistent with glycogen storage disease type IXa (GSD IXa) (OMIM ID: 306000), a treatable inborn error of metabolism.

Discussion

These cases provide an introduction to the factors that contribute to inequities in care for children with genetic conditions and allow us to consider the impact that racial discrimination can have on access to medical genetics care. While we are limited by the qualitative nature of this analysis, we note that there was a delay in diagnosis for each child and that improved partnerships between patient’s families and clinicians are needed to overcome these health system, provider, and patient barriers to equitable medical genetics care.

In the case of Patient 1, a timely diagnosis of X-ALD may have in part been delayed due to the pediatrician’s insufficient exposure to educational and resource materials on adrenal insufficiency in non-white children. X-ALD is a progressive endocrinologic and neurologic disease, an early sign of which is hyperpigmentation due to adrenal insufficiency (Shapiro et al., 2000). Commonly used medical texts and reference manuals emphasize adrenal-related hyperpigmentation in white individuals and often describe it as “bronzing,” which is not applicable to non-white individuals (Chemaitilly et al., 2012; Nieman et al., 2020). Images of light skin tones are overrepresented in medical school textbooks across the country, when compared to both medium and dark skin tones, emphasizing that this is an issue across medicine (Louie & Wilkes, 2018). Although expanded libraries of photographs of diverse individuals have been developed in both medical genetics (Kruszka et al., 2017; Dowsett et al., 2019; National Human Genome Research Institute, 2021) and dermatology (Connelly & Bikowski, 2010; Lugo-Somolinos et al., 2011), these images have not historically been included in more popular and commonly used educational resources.

In recent years, X-ALD has been added to the Recommended Universal Screening Panel for newborn screening (Health Resources and Services Administration, 2018). Although universal screening may improve diagnostic rates of X-ALD in all children, additional disparities remain. Newborn screening for X-ALD relies on a two-tier analysis of very long chain fatty acids and genetic sequencing (Maternal and Child Health Bureau, 2015). Genetic testing in individuals from non-European ancestries is associated with a decreased positive detection rate and increased rate of inconclusive results when compared to white individuals (Landry & Rehm, 2018), at times limiting the sensitivity of genetic testing in these populations. This is typically due to decreased participation in health research by African Americans when compared to their white counterparts (Webb et al., 2018), attributable to the troubled past that minority populations have with medical research (Cobb, 1973). Therefore, children with a positive biochemical screening test may have non-diagnostic molecular testing, leading to further uncertainty for some non-white patients and their families. Inclusion of individuals from diverse backgrounds in future research and genetic databases is crucial to improving genetic variant curation and testing until it is of equal clinical value for all children. While large-scale efforts such as Project Baby Bear (Dimmock et al., 2021) have broadened the use of genome sequencing in infants, the underrepresentation of participants from non-European ancestries in biomedical research may continue to limit the application of these tests.

The experience of Patient 2 illustrates a possible provider-level bias in diagnosis. The clinical features of Down syndrome were not noted by multiple clinicians caring for Patient 2, leading to a delayed referral to medical genetics. Down syndrome is a common aneuploidy that often results in multisystemic manifestations and is diagnosed prenatally in 95% of cases (Hobson-Rohrer & Samson-Fang, 2013; Antonarakis et al., 2020). Among postnatal diagnoses, 98.3% are made within the first year of life (Heuterman et al., 2004). Diagnosis of Down syndrome facilitates appropriate management, such as an echocardiogram and complete blood count in the neonatal period, different from routine newborn care (Bull & Committee on Genetics, 2011).

Individuals with Down syndrome of African descent are significantly less likely to have classic physical findings, including brachycephaly, ear anomalies, clinodactyly, sandal gap, and abundant neck skin (Kruszka et al., 2017), potentially making diagnosis in this population more challenging and contributing to the delay in diagnosis. Dysmorphic features, multiple congenital anomalies, unexplained cognitive impairment, and a history of advanced maternal age, however, should all raise concern for an underlying genetic etiology (Solomon & Muenke, 2012). In this case, despite not recognizing Down syndrome in an African American patient, her dysmorphic features and developmental delay should have prompted genetic testing (Manning et al, 2010; Miller et al., 2010), especially in the setting of advanced maternal age.

The diagnostic delay may be explained, at least in part, by implicit or unconscious bias. Implicit physician bias, most often pro-white, can lead to inferior care recommendations for minority patients (Green et al., 2007). Awareness, acknowledgement, and discussion of one’s bias, however, is beneficial as it has been demonstrated to counteract this effect and allow for more equal care of patients from different backgrounds (Green et al., 2007). Importantly, this patient lives in a medically underserved area, possibly also contributing to the delay in diagnosis, and highlighting that these factors do not occur in isolation. As the majority of individuals with Down syndrome are diagnosed prenatally, it is also important to consider the duration of prenatal care mothers of patients presenting for genetics referrals might receive.

Patient 3, who received a diagnosis of GSD IXa at 13 years old, nearly 11 years after the average age of diagnosis of 2.6 years (Liang et al., 2020), highlights a patient factor that contributes to delayed diagnosis: loss to follow-up. Across pediatrics, referrals are less likely to be both scheduled and attended when families are non-white, use public insurance, or come from lower income homes (Bohnhoff et al., 2019). This effect is exaggerated for referrals to medical subspecialties, such as genetics, versus surgical fields (Bohnhoff et al., 2019). Poor geographic access to pediatric subspecialty care is associated with living in rural communities and in areas where a high proportion of the population lives below the federal poverty level (Mayer, 2008; Penon-Portmann et al., 2019). Increasing telemedicine capabilities for subspecialty care, and eliminating the barrier of travel, may be valuable for these families.

Similar to other specialties, lack of insurance can be a barrier to care for many patients who seek access to medical genetics care. In the experience of Patient 3, a contributing factor to the family’s loss to follow-up was a lack of insurance and fear of large medical bills. Fortunately, during his hospitalization at 13 years old, he was able to become insured. Across the United States, insurance coverage for medically indicated genetic testing varies considerably, both by state and the type of testing covered (i.e., non-invasive prenatal testing vs. whole exome sequencing) (Babu et al., 2020). Further, determining whether a genetic test will be covered by a patient’s insurance can be incredibly difficult, as discussed by Hooker (2020), potentially impeding testing and posing another barrier to adequate genetic care for our patients. Clearer and more streamlined reimbursement models that cover all medically indicated genetic tests are needed.

Additionally, referring physicians play an essential role in educating families about the role of geneticists in their child’s evaluation and treatment. In this case, the actionable and progressive nature of glycogen storage disease may not have been clearly conveyed at the initial genetics evaluation, possibly related to differences in counseling provided to patients of different racial backgrounds. Among adult women with a family history of breast cancer, African American women were less likely to undergo genetic counseling for BRCA1/2 testing when referral was offered by their primary care provider. This remained true when the likelihood of harboring a pathogenic variant, socioeconomic status, and perceived personal risks were accounted for, suggesting that a patient’s race influenced the level of genetic-based care they received (Armstrong et al., 2005). In pediatrics, eliciting parental understanding and concerns when recommending a genetics evaluation and discussing their potential reservations is especially important when a genetic diagnosis may lead to changes in management.

While not unique to genetics, patient distrust may also be heightened by patients’ perceptions of genetic testing. Studies have found that African Americans are less willing to undergo predictive genetic testing due to concerns and fears that the government would use their results to label them as inferior beings and further racial discrimination (Peters et al., 2004; Singer et al., 2004). This finding persists even in cases where African American participants showed greater preference for genetic testing compared to white participants (Singer et al., 2004). Though both the American Society for Human Genetics and the American College of Medical Genetics and Genomics have made formal statements denouncing white supremacy and differential treatment of patients based on race (Gregg, 2020; Wynshaw-Boris, 2020), there is still additional work to be done.

These systemic issues are further compounded by biased representation within clinical medicine, including in the medical genetics workforce. In 2019, 90% of genetic counselors (National Society of Genetic Counselors, 2019) and 79% of medical geneticists (Jenkins et al., 2021) identified as white (compared to 56% of the U.S. physician workforce identifying as white) (Association of American Medical Colleges, 2019). When patient and provider race is concordant, African American patients are more likely to engage in preventive care, especially when invasive interventions are recommended (Alsan et al., 2019). Within genetics, concordant patient and provider race may also help to ameliorate additional factors touched on above, such as loss to follow-up, for example, or a patient’s willingness to undergo testing.

Given the qualitative nature of this case series, the observations associating race and delayed diagnosis cannot yet be empirically substantiated. Future research should investigate the rates of referral to and attendance of medical genetics clinical visits, stratified by demographic factors, including race, insurance status, and proximity to the treating medical facility. For those patients seen by medical geneticists, the types of testing recommended, as well as yield of molecular diagnoses by testing modality should be assessed. Awaiting empirical proof of systemic bias, however, is not reason to delay implementation of strategies to diminish health care disparities within medical genetics now. We propose strategies that can be deployed within the field of medical genetics to recognize and improve upon systemic biases that have been well-recognized in other areas of medicine (Table 1).

Table 1.

Suggestions to address the health system, provider, and patient factors that contribute to delays in genetic diagnoses for pediatric patients.

Health System Factors Include examples of physical exam findings in individuals of different ethnicities and skin tones in reference texts, manuals, and databases. Incorporate virtual clinics into the clinic model and optimize telemedicine capabilities for patients that may be geographically isolated from specialty clinics, such as genetics clinics. Provide financial support for medically indicated genetic testing for all patients. Develop pathways for enrollment in clinical trials and federally sponsored testing and research programs that are accessible to patients from racial minority, non-English speaking, low literacy, and/or low socioeconomic groups. Hire a work force that is representative of the patient population served as representation matters and impacts the patient-provider relationship. Provider Factors Consider genetic conditions when developing differential diagnoses, especially in children with dysmorphic features, multiple congenital anomalies, and/or intellectual disability of unclear etiology. Acknowledge that implicit biases may be present and aim to mitigate the impact this can have on engagement with families and discussions about diagnoses, management, and research opportunities. Participate in cultural humility training that is now available, if not required, at many institutions. Patient Factors Educate patients about the role of medical genetics in their evaluation and the reason for referral, explaining it is more than just research or an academic pursuit. Combat distrust by acknowledging the troubling history of genetics in medicine and addressing patient worries about this when expressed. Elicit patient concerns about genetics referrals and discuss any reservations they may have. Invite patients to bring family members and/or trusted colleagues to these, at times, difficult conversations with implications for others.

A shorter diagnostic odyssey can provide relief for families and allow for better coping, as well as informed family planning for the future (Reiff et al., 2012; ACMG Board of Directors, 2015; Rosell, 2016; Michaels-Igbokwe, 2021). As more genetic-based therapies become available, molecular diagnosis will become increasingly important. Health system-, provider-, and patient-level factors may all contribute to delays in genetic diagnosis but developing awareness of entrenched biases can begin to dismantle potential health care disparities within medical genetics.

Data Availability Statement:

Data sharing not applicable to this article as no datasets were generated or analyzed during the current study.