Clinical perspectives and strategies for confronting disparities in social determinants of health for Hispanic bilinguals with aphasia

Abstract

Disparities in social determinants of health (SDOH) such as socioeconomic status and access to quality healthcare present serious barriers to enrollment in clinical rehabilitation programs for individuals who have experienced a stroke, especially for those who identify with a racial-ethnic minority group. Hispanic bilinguals with aphasia (HBWA) are one marginalized group who face even greater enrollment challenges since post-stroke language impairment and limited English proficiency make it difficult to advocate for one’s needs and identify appropriate rehabilitation programs. Given the increasing representation of Hispanic individuals in the U.S. (projected to be 30% of the population in 2050), it is imperative that clinicians counter disparities in stroke care by facilitating access to clinical services for HBWA. However, the Hispanic population remains largely understudied in the stroke and aphasia literature, due in part to reduced opportunities to enroll in large-scale clinical research studies. In this paper we highlight how our team at Boston University has designed and implemented a variety of recruitment practices, assessment modifications, and treatment accommodations to circumvent the known barriers to participation in clinical research experienced by HBWA. Furthermore, we discuss the importance of cultural responsiveness and demonstrate how including principles of sensitivity and humility in clinical trial protocols improves participant enrollment and retention. Although clinical adjustments in this study were developed for use with HBWA, the effectiveness of the procedures suggests they may be useful blueprints for expanding access to research opportunities for various marginalized groups.

1. Introduction

Nearly 795,000 strokes occur in the U.S. each year (Virani et al., 2020) and almost one-third of these incidents result in aphasia (Flowers et al., 2016). Deficits in post-stroke aphasia vary across language modalities but typically involve some combination of reduced ability in speaking, listening, reading, and writing at different levels of linguistic complexity. Individual profiles of aphasia hinder social functioning and quality of life (Lam & Wodchis, 2010), leading to varying degrees of limitations in functional communication and long-term disability (Hilari, 2012). Early and sustained enrollment in clinical interventions is generally associated with better outcomes for post-stroke populations (Duncan et al., 2005) yet rehabilitation services may be underutilized because language deficits make it difficult to navigate healthcare systems and advocate for one’s medical care. Additionally, language difficulties may interact with social determinants of health (SDOH), or “the conditions in which people are born, live, work, and age”, such as medical history, insurance status, immigration status, geographic location, and degree of English language proficiency, to further reduce clinical opportunities for racially, ethnically, and linguistically diverse groups (Ellis & Jacobs, 2021). The intersecting nature of these factors produce health disparities, or “a particular type of health difference that is closely linked with social, economic, and/or environmental disadvantage” and which may directly or indirectly affect a marginalized group’s access to healthcare resources (Office of Disease Prevention and Health Promotion, 2021).

Previous research has documented negative stroke risk factors and long-term outcomes in post-stroke care associated with health disparities in minority groups (Cruz-Flores et al., 2011). Racial-ethnic minorities have more stroke risk factors (Sacco et al., 2001; Rodriguez et al., 2014), face a higher degree of stroke complications (both fatal and non-fatal) (Shen et al., 2004), and may experience a lower quality of life post-injury when compared to white, non-Hispanic (WNH) individuals (Trimble & Morgenstern, 2008). Population growth expectations and the shift towards increasing multiculturalism in the U.S. suggest that racial-ethnic minority groups will constitute an even greater proportion of annual stroke cases in the coming years (Centeno, 2008). To coincide with this rise in stroke incidence among minority groups, speech and language clinicians should expect higher rates of post-stroke communication deficits, especially among older adults in this population (Centeno, 2009). However, demand for clinical speech and language services to serve these populations will undoubtedly be tempered by health disparities restricting access to necessary care. It is imperative then, that clinicians anticipate and identify barriers to post-stroke care and work to increase opportunities for traditionally underrepresented groups in speech and language interventions.

1.1. Stroke in Hispanic individuals

Individuals of Hispanic background are one racial-ethnic group who may experience difficulties accessing stroke care, including speech and language services. In the United States, the term Hispanic generally refers to individuals of any race whose familial origins may be traced to the Caribbean, Central America, South America, or other Spanish speaking regions (U.S. Office of Management and Budget, 1997). Although Latino may be used similarly and interchangeably with Hispanic (Rodriguez et al., 2014), the latter is the preferred intragroup identifier (Taylor et al., 2012) and will be used throughout this manuscript to refer to individuals in the U.S. specifically. The Hispanic population is largely composed of Mexican-American (63%), Puerto Rican (9.2%), and Cuban (3.5%) (Abraído-Lanza et al., 2016) individuals, the majority of which reside in five US states (California, Texas, Florida, New York, and Illinois) (Ennis et al., 2011). Currently, Hispanics comprise 18.4% of the total U.S. population and, as the fastest growing minority group in the nation, are projected to constitute 28% of the total population by the year 2060 (U.S. Census Bureau, 2018). The estimated risk of stroke for Hispanics is higher than that of WNH individuals which a) is driven in part by their higher vascular risk factor prevalence (i.e., higher rates of diabetes, hypertension, reduced physical activity, obesity, and coronary heart disease among other cardiovascular risk factors) (Rodriguez et al., 2014), b) puts Hispanics at high risk for post-stroke cognitive impairment resulting in vascular dementia (Chen & Zissimopoulos, 2018), and c) is associated with higher stroke mortality relative to their WNH counterparts (Cruz-Flores et al., 2011).

Although the number of Hispanic individuals living with post-stroke language impairment remains unknown, trends in stroke incidence suggest this demographic will encompass a larger portion of clinician caseloads in the coming years (Centeno, 2009; Centeno & Harris, 2018). Furthermore, the Hispanic population in the U.S. is largely bilingual, albeit to varying degrees of language proficiency; in fact, 71.1% of Hispanic households reported speaking a language other than English at home in 2019 (U.S. Office of Minority Health, 2021). Therefore, it is likely that Hispanic individuals may require aphasia assessment and treatment in more than one language, leading to an increase in demand for bilingual rehabilitation services that meet their particular needs.

1.2. Hispanic Bilinguals with Aphasia

Bilingual speakers with aphasia present language deficits across their two or more languages with various patterns of overall impairment which may severely limit their ability to communicate in a variety of sociolinguistic environments (Peñaloza et al., 2020). In this paper, we will use the term Hispanic bilinguals with aphasia (HBWA), a subset of Spanish-English bilinguals with aphasia (BWA) who self-identify as Hispanic, to discuss relevant challenges and solutions to bilingual service delivery for this important demographic group.

HBWA may seek assessment and treatment in one or both of their languages, especially if the nature of their language impairment limits their use of the dominant language pre-stroke and therefore causes a shift in typical communication habits (Peñaloza, Barrett, & Kiran, 2020). Providing culturally appropriate assessment and treatment for HBWA may present a series of challenges for speech-language pathologists who may be relatively unfamiliar with bilingual service delivery (Centeno, 2009; Wiener et al., 1995). HBWA may also struggle to find programs if they live in rural areas, need trained interpreters or translators to effectively support clinical enrollment and service delivery, or rely on caregivers for transportation to and from sessions. Additionally, the ongoing COVID-19 pandemic has worsened these existing difficulties and created new obstacles to securing therapy by diminishing in-person speech and language opportunities for all individuals with aphasia (Kong, 2021)

It is well established that in-person language therapy for HBWA is effective and may result in improvement in one or both languages (Kiran & Roberts, 2010; Kiran, Sandberg et al., 2013). Importantly, a recent study providing language therapy for lexical retrieval deficits in HBWA revealed comparable treatment effectiveness and reliability for teletherapy relative to in-person language treatment (Peñaloza et al., 2021). These findings suggest that teletherapy may help improve limited access to rehabilitation services for HBWA increasing their opportunities for better language therapy outcomes. However, although evidence exists to support the effectiveness of bilingual therapy approaches and service delivery modes, limited research has been conducted to identify barriers to HBWA participation in treatment, propose solutions to difficulties with enrollment and retention, and monitor instituted plans of action. Limited evidence based on specific populations results in a lack of best clinical practice guidelines for minority populations with communication disorders (Beveridge & Bak, 2011) which underscores the need of research evidence on the limitations and opportunities for neurorehabilitation practices with HBWA.

1.3. Aims

The goal of this paper is to detail the various barriers to care that HBWA may face as they seek rehabilitation services, assessing factors that challenge successful treatment delivery and describing how clinical protocols can be modified to improve clinical service accessibility. The specific aims are to a) catalogue and describe the various inequities and access disparities facing Hispanic individuals and the corresponding difficulties arising in aphasia care, especially those which have been worsened by the COVID-19 pandemic and b) describe practices instituted by our research team to mitigate the barriers to HBWA participation in clinical rehabilitation research.

2. Disparities Between Hispanic and WNH Individuals in the U.S.

There are fundamental inequities present in SDOH, between WNH individuals and minority groups in the United States. There is a wealth of evidence demonstrating that negative impacts from disparities in SDOH, particularly compounded for minority groups, influence health outcomes at the micro and macro levels (National Academies of Sciences, Engineering, and Medicine, 2017). These disparities are unnecessary, avoidable, and unfair and unjust, as highlighted by the World Health Organization (Vega et al., 2009). Disparities in SDOH and (dis)advantage can be captured by examining factors such as access to economic resources, education, housing, transportation, support services, technology, and healthcare availability and quality, as well as immigration status and language proficiency (see Figure 1 for SDOH for HBWA).

Figure 1:

SDOH contributing to inequities in health outcomes for HBWA, prior to stroke, directly after stroke (acute~0–6 months), and post-stroke. Items in bold are discussed in section 2. Effects of SDOH at all stages are suggested to compound; negative disparities in SDOH are predicted to result in inequities in stroke risk, rehabilitation and potential recurrence.

Investigation into how disparities in SDOH impact certain racial-ethnic minority groups is extensive although researchers often encounter significant barriers in trying to reduce such disparities (Borfitz, 2021). Nonetheless, it is more important than ever to enact such change as the COVID-19 pandemic has accelerated and magnified the effects of existing disparities in SDOH. For instance, as of March 2021, Hispanic individuals accounted for nearly one-fifth of all U.S. coronavirus cases and were more than two times as likely to be hospitalized for severe COVID-19 symptoms when compared to WNH individuals (Centers for Disease Control [CDC], 2022). In this section, we catalog and describe the most persistent socioeconomic, cultural and environmental factors contributing to these disparities in SDOH, and will review their corresponding effects on individuals at risk for chronic post-stroke impairment in the context of the COVID-19 pandemic.

2.1. Socioeconomic Factors

Many poor health outcomes can be directly traced to the complex interaction of SDOH related to socioeconomic status (Vega et al., 2009). Hispanic individuals are twice as likely to live below the poverty line, four times less likely to have finished high school, are the largest population segment to be uninsured, at a rate of nearly two times all other racial-ethnic groups, and are twenty times more likely to have limited English proficiency (LEP) compared to WNH (U.S. Census Bureau, 2015; U.S. Office of Minority Health, 2021; Velasco-Mondragon et al., 2016). These disparities in SDOH compound and have extensive effects on chronic health outcomes, in the form of inequities (Institute of Medicine, 2003).

2.1.1. Stroke Costs and Disparities in Insurance Coverage

Hispanic and other minority group members are at a distinct disadvantage in the United States, in terms of representation and integration in the formal public health structure. High rates of uninsured Hispanic individuals reduce opportunities for both preventative and follow-up care for chronic health conditions (Balluz et al., 2004). Lack of access to healthcare and fear of possible detection by immigration authorities regularly leaves individuals to only pursue assistance only after medical conditions are very severe, both in cases of stroke as well as COVID-19 (Bebinger, 2021; Hacker et al., 2015). Unfortunately, presenting with more severe cases at time of care often yields dire outcomes.

In addition to reduced insurance coverage, comorbid diagnoses of chronic illnesses such as hypertension, diabetes, and obesity are widely known to worsen stroke-related outcomes. Each of these illnesses lowers quality of life for patients and increases risk of experiencing chronic health conditions later in life. Hispanics lead WNH individuals in rates of diabetes (21% of Hispanics over 20 years of age, compared with 13% of WNH individuals) and obesity (45% of Hispanics over 20 years of age, compared to 37% of WNH individuals) (National Center for Health Statistics, 2019), and have high prevalence of hypertension, at 34%, which, though lower than their WNH counterparts (47%), is 24% more poorly managed (Campos & Rodriguez, 2019; CDC, 2015). Between 2005–2050 alone, ischemic strokes in Hispanic-Americans are projected to cost $357 billion health care dollars (Brown et al., 2006). As Hispanic individuals tend to have strokes at younger ages than their WNH counterparts, this projection covers not only initial hospitalization ($86 billion), but also informal care/care provided by family and friends ($60 billion), and lost wages ($79 billion) (Brown et al., 2006).

2.1.2. Limited English Proficiency and Health Literacy

There are nearly 30 million U.S. residents with LEP, and half of these individuals are uninsured (Kaiser Commission on Medicare and the Uninsured, 2012; U.S. Census Bureau, 2020). The formal healthcare system in the U.S. has not been created with a focus on assisting and supporting LEP individuals nor accounting for their health literacy needs. The National Institutes of Health (NIH) define health literacy as the “degree to which individuals have the capacity to obtain, process and understand basic health information and services needed to make appropriate health decisions”; based on this standard, 41% of Hispanic individuals have low health literacy in their second language which decreases their ability to navigate the complex American healthcare system (Key, 2019; Hacker et al., 2015). Low health literacy indexes lower utilization of preventive services and higher rates of hospitalization, further widening health inequities as seen through the higher prevalence of certain chronic diseases (Key, 2019).

This was starkly apparent during the COVID-19 pandemic, where the lack of health information and reduced availability of resources in Spanish contributed to increased rates of infection for Hispanics (Velasquez et al., 2020). Lack of information for LEP individuals also impeded progress in halting or reducing within-community spread of COVID-19 (U.S. Office of Minority Health, 2021) by reducing awareness of pandemic control measures such how and where to receive the COVID-19 vaccine. Similarly, LEP is strongly associated with lack of stroke knowledge, both in identification of signs and knowledge of what to do in case of stroke amongst Spanish-speaking Hispanics (Mozaffarian et al., 2016) when compared to English-speaking Hispanics, non-Hispanic blacks, and WNH.

2.2. Cultural Factors

2.2.1. Determinants of Reduced Health Care Usage

Gaps in health access and low health literacy are not only due to insurance status and LEP, but also due to a lack of meaningful relationships with providers (Key, 2019). Hispanic individuals have reported significantly more perceived discrimination from healthcare providers, compared to white individuals (Blanchard & Lurie, 2004). As a result, Hispanic individuals have been less likely to pursue optimal chronic disease screening or follow post-discharge care instructions (Blanchard & Lurie, 2004). Patients are more likely to seek care from individuals who share their language, cultural and/or ethnic backgrounds, expressing greater satisfaction and compliance when these are patient-concordant, as less discrimination is perceived (Sánchez et al., 2015). Clinicians and health-care providers can play an important role in motivating Hispanic individuals to complete preventive screenings and seek early treatment for illnesses such as diabetes and high blood pressure (Balluz et al., 2004). Unfortunately, there is a limited representation of Hispanic and/or Spanish-speaking physicians in the health care system to aid these vulnerable patients (Sánchez et al., 2015), and therefore many individuals may access care less regularly, especially for chronic health conditions.

Furthermore, it is difficult for many Hispanic patients to travel to doctors and facilities of choice if not geographically accessible. Many Hispanic individuals rely predominantly on public transport (Smart, 2015), are over two times less likely to have access to private transportation compared to WNH individuals, and are more likely to live further from transport hubs, in order to acquire more affordable housing (Ramirez, 2019). These factors, together with perceived cost and unreliability of public transport, diminish individuals’ abilities to travel to providers who may be race- and/or language-concordant. In the case of stroke or chronic illness follow-up care, the lack of reliable access to transport can directly affect patient outcomes.

2.2.2. Immigration Status

Immigration and citizenship status may also constitute barriers to receiving care for Hispanic individuals living in the U.S. (Escarce & Kapur, 2006; Hacker et al., 2015; Pérez-Escamilla et al., 2010). Immigrants who are not U.S. citizens typically lack employment opportunities through which they qualify for insurance (Escarce & Kapur, 2006) nor are they generally eligible to receive care through federally-funded public health programs (Hacker et al., 2015). At the individual level, immigrants are more likely to experience difficulties a) securing transportation to medical appointments, b) affording routine care visits and medications, and c) navigating the health care system to find providers and specialists. Additionally, feelings of shame and fear of stigma in receiving care or of being reported to the immigration authorities and facing deportation prevent many individuals from seeking medical care (Maldonado et al., 2013). These barriers result in noncitizen immigrants avoiding healthcare and waiting until health issues are critical. Often, individuals are left to acquire services in community clinics or hospital outpatient departments (Escarce & Kapur, 2006; Hacker et al., 2015), where there may be less satisfaction in patient-clinician relationships and a decreased continuity of care.

2.3. Environmental Factors

2.3.1. Lack of Digital Access to Providers: COVID-19 and Telehealth

The social distancing measures and government mandates associated with the COVID-19 pandemic have altered the day-to-day functioning of vital institutions and slowed the pace of life globally. In the U.S., disruptions to nonessential outpatient medical services have narrowed the scope of routine medical care (Birkmeyer et al., 2020), making it harder for clinicians to provide in-person consultation visits and diagnostic care. These changes have severely reduced treatment opportunities for those at high risk for coronavirus infection, including adults 65 and older (Giunti et al., 2020), residents of nursing homes or other long-term facilities, and pertinent to this paper, stroke survivors with a history of cardiovascular risk factors. In particular, seeking conventional in-person therapy became more challenging for the Hispanic/Latino population since older adults in this group are already at higher risk for infection and death from COVID-19 relative to WNH individuals (Sáenz & Garcia, 2020). Challenges to service delivery have also been reported in the speech and language rehabilitation field since, in compliance with social distancing guidelines, many clinics suspended outpatient speech and language programs for high-risk individuals including stroke survivors with aphasia.

Clinicians attempted to minimize these pandemic-related effects by implementing alternative methods for speech and language therapy (Kong, 2021). The abrupt cancellation of nearly all in-person services at the beginning of the crisis forced a rapid transition to teletherapy, an appropriate clinical service delivery model for adult neurogenic language disorders (Weidner & Lowman, 2020). Telerehabilitation presented a unique solution to circumvent issues such as lack of transportation and inadequate access to bilingual practitioners for this at-risk population. However, the pandemic has exacerbated pre-existing digital inequalities for individuals who already have less representation in virtual rehabilitation. Also, there is evidence for lower use of telemedicine visits by Spanish-speaking patients as compared to English proficient patients during the pandemic (Eberly et al., 2020; Rodriguez et al., 2021) likely tied to reduced language competence.

2.3.2. Technological Support and Access Limitations

There is evidence for reduced access to internet and personal computers among Hispanic/Latino individuals relative to WNH (Lopez et al, 2013). This digital divide can be thought of as the gap between those who have access to and can engage with technology, compared to those who cannot (Rodriguez et al., 2021). While there are a range of factors that may contribute to this discrepancy, two primary factors are: 1) lack of technology; 2) lack of digital literacy. Thus, for some HBWA, barriers to care may be material: they lack reliable high-speed internet or the appropriate computer technology to meet the needs of teletherapy. Others may be unfamiliar with technology in general and have insufficient support at home to manage the language and motor demands of learning new applications.

Video teletherapy can be particularly beneficial for LEP individuals by providing visual reinforcement for speech processing and providing opportunities for social interaction and engagement among clinicians, patients, and family members—essential for many Hispanic individuals, as detailed in section 2.2. However, a recent study of older adults’ telemedicine readiness determined that 13 million older adults are not ready for video visits due to difficulties with technology (Lam et al., 2020). Another study investigating the use of telephone and video telemedicine visits during COVID-19, found lower use of video versus telephone visits among older Hispanic and Spanish-speaking primary- and specialty-care patients (Rodriguez et al., 2021).

These trends of inequities in telehealth use began early in COVID-19 lockdowns and grew, rather than diminished over the course of the pandemic. Individuals most likely to use video visits were white, enrolled in commercial insurance programs, and living in areas with higher income and broadband internet access. Conversely, patients in areas with the lowest broadband access, lowest median income and lowest educational attainment were less likely to utilize video visits (Rodriguez et al., 2021). This digital divide was further increased by the need for language-specific modifications for telehealth utilization by LEP individuals.

Access for patients with LEP needs to be more user friendly, and differ from that used for English-proficient patients. Teletherapy options may be overwhelming for many HBWA, but if clinicians adjust for these access difficulties, virtual aphasia rehabilitation may emerge as a feasible, long-term option. Our own work in telerehabilitation for HBWA (Peñaloza et al., 2021) supports this notion and suggests that these methods would be critical for expanding services for underrepresented aphasia groups. It is essential to identify effective strategies to overcome the digital divide, in order to facilitate optimal care for HBWA.

3. Facilitating Access to Clinical Services in a Research Context

Although speech-language pathologists may be unable to enact socioeconomic changes to benefit HBWA due to the political and infrastructure-based nature of these issues, they might ease cultural and environmental constraints on clinical enrollment in a variety of ways. Low representation of Hispanic adults receiving clinical care is mirrored in their under-representation in research/clinical trials (Arevalo et al., 2016). In previous qualitative studies about research participation, Hispanic-Americans have cited a variety of reasons for low enrollment including, geographic inaccessibility, time constraints and cost associated with travel, language difficulties, and family considerations or responsibilities (Nodora et al., 2010; Ford et al., 2008). Our own work in this area underscores these challenges and in the remaining sections we demonstrate how incorporating principles of cultural awareness and responsiveness in recruitment practices and implementing telerehabilitation approaches for assessment and treatment delivery may present tangible solutions to these documented obstacles. Each of these factors will be discussed in the context of an ongoing, randomized controlled trial (RCT) focused on the prediction of language treatment outcomes in HBWA (Peñaloza et al., 2020).

3.1. The PROCoM clinical trial

Since 2017, our research team at Boston University has taken a number of steps to recruit HBWA to enroll and participate in a bilingual language clinical trial (Predicting Rehabilitation Outcomes in Bilingual Aphasia Using Computational Modeling (PROCoM; Clinical trial registration number NCT02916524). The primary goal of the ongoing PROCoM RCT (see Peñaloza et al., 2020 for study protocol) is to predict post-treatment language recovery in 48 Spanish-English bilinguals with aphasia using BiLex, a previously validated computational model of bilingual lexical access (Peñaloza, Grasemann, et al., 2019). For each participant, the BiLex model simulates a) pre-stroke lexical-retrieval abilities in both languages, b) post-stroke lexical-retrieval impairment in each language, and c) treatment-induced recovery when therapy is provided in English versus Spanish. By comparing the simulated treatment outcomes in both the treated and the untreated language resulting from each treatment language option, the BiLex model can recommend Spanish or English as the optimal treatment language for maximum therapy gains in both languages. In the study, participants are randomly assigned to a model-prescribed experimental group to receive treatment in the model-selected language or to a model-opposite control group to complete therapy in the model non-prescribed language. Irrespective of treatment language (Spanish or English), the model-prescribed group is expected to show greater linguistic improvement than the control group in both languages post-treatment, thus confirming that the BiLex model can accurately predict the optimal treatment language for HBWA. All procedures discussed below were reviewed and approved by the Institutional Review Board at Boston University (reference number: 4492E).

3.2. Participant Recruitment

During participant recruitment, our research team acknowledged the aforementioned barriers faced by potential HBWA participants and subsequently employed a variety of measures to circumvent these obstacles. First and foremost, HBWA were recruited for the study from three separate locations: one primary site at Boston University, MA and two secondary sites in Austin, TX and San Francisco, CA. Individuals were considered for the study if they met the following inclusion criteria: a) aged 18–85 years, b) medical history demonstrating evidence of a left-hemisphere stroke resulting in a diagnosis of aphasia, c) at least six months post-stroke, and d) any level of bilingual proficiency in Spanish and English. Individuals were not included in the study if they had a previous neurological diagnosis not relating to the stroke (e.g., schizophrenia, dementia) or active medical conditions which would have prevented their participation in weekly clinical sessions.

Trained bilingual clinicians, some of whom shared cultural backgrounds with study participants, gathered information about stroke history and demographics, provided details for informed consent, and discussed enrollment with participants and their families in their language of preference. Research staff employed a community-based recruitment approach by contacting university speech and language clinics across the country, online aphasia community groups, local rehabilitation hospitals, and relevant community centers (e.g. churches and adult enrichment programs) near the research sites to disseminate advertisements and other materials in both Spanish and English. At the primary recruitment site specifically, individuals were often referred for the study by two Hispanic stroke specialists who initially saw participants as medical providers. To alleviate accessibility concerns, clinicians offered to pay transportation costs for in-person visits (i.e. via a ride-sharing application) or visited participants in long-term care facilities or their homes if they were unable to travel and preferred face-to-face sessions. Other participants expressed interest in participation via teletherapy but lacked appropriate technology to engage with the study. Therefore, recruitment efforts also included a laboratory loan program provided from the primary recruitment site through which participants at any site could borrow laptop computers and internet hotspots to support online interaction and task completion with study clinicians.

This latter strategy combined with robust telerehabilitation protocols for assessment and treatment (Peñaloza et al., 2021) directly enabled participants from across nine states in the U.S. (MA, CA, TX, RI, CT, NY, WA, NC, and FL) and one internationally to enroll in the study, regardless of their proximity to a recruitment site. In general, participants were enrolled in the study for about 5–6 months which included pre-treatment assessment, treatment, and post-treatment assessment. The recruitment process, however, was individualized for each participant pending informal assessment of support and materials they might need to effectively enroll. For example, some participants inquired about the study before they were six months post-stroke; in these cases, we maintained communication with individuals and their families until they met this inclusion criterion. For others, we sent materials such as laptops or devised plans with caregivers to support virtual enrollment and circumvent resource-based barriers to joining the study. Some participants who were referred to us by their medical providers did not understand much about bilingual aphasia or how language therapy could support their post-stroke communication goals. In these cases, study staff remained sensitive to cultural differences and attitudes towards receiving and planning for clinical services and dedicated time and resources to educating potential participants about bilingual aphasia and discussing the likely benefits of their participation in the study. By working directly with participants and their families, we assessed the needs of each individual with aphasia and adapted our procedures to facilitate their inclusion and representation in clinical research. These national efforts were instrumental in facilitating access for HBWA with diverse backgrounds and experiences and allowed the RCT to proceed relatively unhindered by the additional restrictions imposed by the COVID-19 pandemic. As of August 2021, 148 participants had been contacted for the study, 72 had completed screening measures, and 54 had enrolled in the study. Figure 2A indicates the geographical origin of these Hispanic participants. Figure 2B highlights their educational attainment (17 of the 40 have 12 years of education or less). Figure 3 reviews geographic enrollment (organized into maps based on the three major enrollment sites) for consented individuals who had completed at least some pre-testing assessments by August 2021. The maps are shaded by income inequality (U.S. Census Bureau, 2021) at the county level to demonstrate that the majority of study participants live in areas of relative socioeconomic disadvantage—an important contributor to health inequity as discussed in section 2.1 above.

Figure 2:

A) Educational Attainment for consented participants in the PROCOM trial; B) Parental/familial origin as determined by parents’ place of birth. Note: Puerto Rico is considered part of the ‘Caribbean’ group. Educational attainment not available for one participant.

Figure 3:

A-C) Participant enrollment via three large recruitment sites. Each yellow dot represents one participant; counties are shaded by income inequality (expressed as Gini indices of inequality where 0 indicates perfect income equality and 1 represents near total inequality) to demonstrate that most participants live in areas of socioeconomic inequity. Note: Additional locations not pictured on these maps include participants from WA (1), FL (1), NY (1), NC (1), and Canada (1). Income inequality values were retrieved from the American Community Survey 2019 5-year estimates (U.S. Census Bureau, 2021).

3.3. Telerehabilitation for Assessment and Treatment Delivery

Telerehabilitation is emerging as a tool for developing and implementing standardized protocols for assessment and treatment for use with BWA (Peñaloza et al., 2021). Telerehabilitation programs may connect HBWA, many of whom demonstrate a high degree of linguistic heterogeneity in their pre-morbid linguistic abilities and post-stroke profiles of impairment (Peñaloza, Barrett, & Kiran, 2019) with bilingual clinicians proficient in both Spanish and English regardless of geographic limitations. This may be especially helpful for participants who are unable to travel due to a) post-stroke complications, b) distance-related travel costs, or c) limited caregiver support (Peñaloza et al., 2021).

As detailed in section 3.2 above, PROCoM participants completed study visits in-person or via teletherapy in accordance with their needs or preferences. To date, more than two-thirds of all participants have opted for remote assessment and/or treatment which suggests that the virtual option has facilitated access for HBWA who a) do not live near our main recruitment sites, or b) are unable to travel for rehabilitation due to post-stroke complications, or c) may not have widely-available bilingual services in their communities.

Prior to the pandemic, to accommodate HBWA without access to appropriate technology, clinicians loaned out laboratory-owned laptop computers to teletherapy participants. However, over time, comprehensive recruitment practices and the COVID-19 pandemic required us to mail or hand-deliver additional laptop computers and internet hotspots to participants as needed. Clinicians managed difficulties with computer proficiency by traveling to participants’ homes to set up and teach participants and caregivers how to use the technology or by providing tech support via FaceTime or WhatsApp videos and pictures to facilitate contactless setup and use. In many cases, these interactions established rapport with HBWA and their families prior to study initiation and encouraged them to consult clinicians as resources throughout enrollment. Furthermore, clinicians were able to observe communication in the home environment and use this information to adapt their own discourse to suit the cultural and sociolinguistic styles of individuals and their caregivers.

3.3.1. Assessment Modifications

During pre and post-treatment assessments, participants met with clinicians to complete a wide array of cognitive-linguistic testing in both Spanish and English (see Peñaloza et al., 2020 for a full list of assessments). Testing was conducted in two-hour sessions, twice per week with English and Spanish being evaluated on different days to minimize cross-language interference effects. During the assessment periods, clinicians remained sensitive to cultural-linguistic differences which could have affected participant performance such as dialectal variation or regional word production. For example, when participants produced unfamiliar words on tests of lexical-retrieval, clinicians noted the production and conferred with other study staff at weekly meetings to determine whether the response should be considered correct or incorrect. When necessary, clinicians also consulted the participant’s caregivers or family members to determine the meaning or a regional word or phrase. These accommodations created a more accurate and inclusive testing environment for participants and allowed us to tailor our procedures to variety of Spanish dialects.

Due to a variety of factors such as aphasia severity, motor impairment, or insufficient computer literacy, many participants required support to complete assessment tasks remotely. In general, clinicians used the screen-sharing feature in the Zoom communication software available at www.zoom.us and electronic versions of assessment tools to administer tasks with pictured stimuli. For example, during the Auditory Word Recognition task on the Western Aphasia Battery-Revised (WAB-R; Kertesz, 2006), administered following our validated protocol for videoconference testing (Dekhtyar et al., 2020), clinicians showed the stimulus book via Zoom and directed participants to select an answer using the remote access feature (i.e., participants selected an answer using a computer mouse or their finger on an iPad). When participants with motor impairment were unable to point using a mouse reliably, clinicians asked a caregiver to provide support and control the mouse on behalf of the participant. If a caregiver was unavailable, clinicians labelled stimuli and permitted a verbal or gestural answer instead of a manual one.

If participants experienced audiovisual difficulties in Zoom during the assessments, clinicians utilized secondary sources to troubleshoot issues and continue the sessions. Strategies included calling participants via cellphone or initiating a video call via FaceTime or WhatsApp to circumvent audio and video disruptions, respectively. Additionally, secondary video sources were useful when some participants found it difficult to adjust a computer or iPad screen so that their actions were visible to the clinician for certain tasks. During commands-based subtests such as Sequential Commands from the WAB-R, caregivers or participants often used smartphone video to improve visibility and aid the clinician in item scoring. These accommodations and solutions to common telerehabilitation difficulties for assessments are summarized in Figure 4.

Figure 4:

Remote testing decision tree to assist clinicians in modifying select assessments for telerehabilitation.

3.3.2. Treatment Modifications

During the treatment phase, all study participants received lexical-retrieval therapy in either English or Spanish for 15 items across 20 sessions (2 hours each, twice per week). The details of this computerized, semantic-feature based treatment have previously been reported elsewhere (see Peñaloza et al., 2020; 2021). In brief, treatment was conducted using Qualtrics surveys available at www.qualtrics.com and followed an identical six step procedure for each item: 1) Naming; 2A) Feature selection; 2B) Feature assignment; 3) Association; 4) Yes/No questions; 5) Naming; and 6) Sentence production. For some participants, independently managing different aspects of the computerized intervention (i.e. moving from screen to screen, reading and comprehending each semantic feature) proved challenging. Clinicians accounted for these difficulties by using similar methods to those developed during the assessment phases. The first few subjects were given the choice to self-direct (i.e. receive a link to the treatment survey and share their own screen with the clinician) while the remaining participants were guided through each session using the screen-sharing feature in Zoom. Under this clinician-controlled method, study staff retained control of the mouse and moved it in response to a participant’s answers. This improved procedures as it allowed participants with motor or digital literacy impairments to a) focus their attention on comprehending each therapy task and b) complete more items in each session. It should be noted that the only difference between these two methods was whether the participant or the clinician controlled the computer mouse during treatment tasks. Regardless of the method employed, clinicians were always present during treatment and were responsible for providing feedback about task accuracy to each participant in the same manner.

A variety of adjustments were also made within Qualtrics to support difficulties stemming from aphasia severity. During steps which required participants to comprehend and manipulate semantic features (i.e. 2A, 2B, and 4), audio support was added to Qualtrics such that every time clinicians clicked on a semantic feature, participants heard the content read out loud. In these same steps, clinicians used the ‘annotate’ feature in Zoom to reorient participants or draw their attention to aspects of semantic features for accuracy and feedback purposes.

4. Discussion

The present work aimed to review the barriers and challenges to stroke care leading to negative health outcomes for Hispanic individuals living in the U.S. Many of the socioeconomic, cultural, and environmental factors underlying these health disparities are compounded for HBWA who face additional difficulties in identifying suitable rehabilitation programs and advocating for their own health care needs. Without intervention and large-scale political, social, and economic reform, Hispanic individuals will continue to experience higher rates of stroke compared to their WNH counterparts and may face fewer opportunities for aphasia care.

4.1. Clinical and Research Implications

Evidence from our ongoing PROCoM RCT at Boston University demonstrates that community-based recruitment practices and telerehabilitation procedures for language therapy can facilitate access to clinical research opportunities for HBWA. The success of this large clinical trial (54 participants enrolled) suggests that recruitment is particularly effective when principles of cultural sensitivity are incorporated into the process such as a) disseminating study materials in both English and Spanish and b) connecting potential participants with bilingual clinicians who have similar cultural-linguistic backgrounds. This option to communicate in a preferred language with clinical personnel who share cultural-linguistic backgrounds with participants is an important predictor of Hispanic satisfaction in care settings.

Following enrollment, clinical researchers and speech-language pathologists may rely on telerehabilitation methods to deliver services directly to HBWA and other individuals with aphasia. Emerging evidence suggests that technology loaning programs and ongoing support from clinicians and caregivers are effective in alleviating concerns associated with digital literacy, computer proficiency, and internet access for the purposes of clinical research. Additionally, clinicians may make a variety of adjustments during assessment and treatment to account for telerehabilitation difficulties associated with differences in motor impairment, aphasia severity, technological disruptions, and caregiver availability. Although evidence supporting the feasibility and validity for telerehabilitation with BWA remains limited, a recent study comparing the effectiveness of in-person and remote administrations for PROCoM procedures revealed no significant differences for the two delivery modalities across a) post-treatment effect sizes reflecting change on primary outcome measures nor b) post-treatment assessment scores for key instruments such as the WAB-R reflecting treatment-related change on secondary outcome measures (Peñaloza et al., 2021). This finding highlights the flexibility of telerehabilitation methods and upholds the validity of modified assessment and treatment protocols.

Although the modifications and accommodations presented in this manuscript were initially developed with HBWA in mind, these methods should not be considered exclusive to this group. We have included a condensed protocol in the Appendix which details the changes we made to recruitment, assessment, and treatment practices with recommendations and justifications for implementation. Clinicians and researchers alike may find this resource useful in expanding our practices to support service delivery for other populations with aphasia.

4.2. Future Directions

To address the inequities in health outcomes between Hispanic and non-Hispanic individuals, future research should include more rigorous and uniform assessments and measurements of relevant SDOH such as income inequality, access to care, English language proficiency, and health and technological literacy (Balluz et al., 2004). Much of the research underlying what is currently known about SDOH and health outcomes is based on individuals of designated priority populations (e.g., people of color, low-income individuals, women, older adults, etc.), but not in subgroups in which factors intersect (e.g., HBWA) (Ndugga & Artiga, 2021). There is a clear need for strategies which increase representation of Hispanic individuals in clinical stroke research (Hayes-Bautista, 2020) such that study cohorts embody a variety of family backgrounds, migration statuses, and language histories as differential health outcomes have been found between different subgroups of Hispanic individuals (Artiga et al., 2015). Finally, future work should investigate the connections between barriers to care experienced by individuals with aphasia and participant satisfaction with clinical methods employed to overcome these obstacles. Qualitative methods to solicit feedback from participants would directly involve them in the process of improving our procedures while ensuring that our procedures grow and evolve to match individuals’ needs.

5. Conclusion

Overall, this work suggests that traditional care and delivery approaches should be reviewed and adjusted to offer equitable opportunities for all stroke survivors without compromising the standard of care which the field has already achieved. Individualized and customized rehabilitation is possible when stakeholders (i.e. clinicians, HBWA, caregivers) collaborate to overcome disparities and other barriers to clinical enrollment and service delivery. In the future, modifications via telerehabilitation and other inclusive practices may improve health outcomes for HBWA and increase the accessibility of services for bilingual aphasia.

Appendix. Clinical guidelines for improving recruitment, assessment, and semantic-feature based treatment procedures for HBWA

Recruitment:

• Adopt a community-driven approach: distribute bilingual recruitment materials/flyers to adult day programs, religious organizations, local and online aphasia support groups, and university speech and language clinics

• Establish relationships with Spanish-speaking clinicians and physicians who share cultural-linguistic backgrounds with HBWA; HBWA-concordant providers may discuss with participants and their care providers the benefits of participating in clinical research as a part of stroke recovery

• Follow-up directly with participants after referrals or initial contact; personal phone calls, emails, and messages (i.e., via WhatsApp or related services) are more effective than mailing letters to individuals’ homes when building rapport

• Consider accessibility when writing budget proposals or grant awards: funds for technology loaning programs or transportation assistance (e.g., covering the cost of public transportation or ride-sharing services) may support highly motivated individuals who might be experiencing materials-based or financial barriers to participation

• Offer in-home opportunities (if permitted under institutional regulations) to accommodate some participant preferences for remaining at home during recovery

• Explore push-in collaborations/contracts with rehab sites to maximize retention by working with individuals who are already traveling to or residing in a facility for specific services (e.g., physical therapy, occupational therapy, etc.)

Assessment:

Virtual Recommendations:

• Review list of planned assessments and identify which instruments will require modifications for use in a virtual setting (e.g., does the test require the participant to manipulate objects on a table in front of them?)

• Create virtual versions of all assessment tools which may include: scanning stimulus booklets, taking pictures of objects typically provided by the examiner (e.g., spoken and written recognition subtests on the WAB-R), labelling items to compensate for motoric difficulties, and updating specific portions of instruments to facilitate online testing (e.g., converting ‘draw a cross with your foot’ to ‘draw a cross with your finger’ on reading commands subtest on the WAB-R)

• Evaluate participant comfort level when navigating features on Zoom (or other videoconference platform) by asking them to: a) toggle on/off the microphone and video camera; b) initiate screen sharing; c) give remote control of their own screen to the examiner; d) take control of the examiner’s screen; e) utilize annotate functions; and f) send and receive items through the chat window

• Inquire whether a caregiver will be available to support a participant during testing if technological difficulties arise; if so, discuss guidelines and expectations with caregivers such as not repeating stimuli or giving hints on any assessment measure

• Change videoconference language settings to Spanish if requested by the participant or their caregivers

General Recommendations:

• Determine whether a participant may use or comprehend a different regional word or phrase than what is printed in the assessment manual/record form, especially during Spanish testing (e.g., on sequential commands on the Spanish WAB, participants may prefer bolígrafo/pluma/lapicero for English pen); anticipating these differences and using the preferred lexical item is important for improving the quality of cognitive-linguistic evaluation while also including principles of cultural sensitivity in clinical and/or research settings

• Note all participant utterances (i.e., spoken and written) which are unfamiliar to the examiner and consult bilingual clinicians, regional dictionaries and other reputable linguistic resources, or the participant’s family/caregivers to determine meaning and possible accuracy in context

• Conduct extensive clinician training on bilingual assessment procedures and develop a written protocol that lists each planned assessment with: a) clearly defined procedures and b) lists of accepted alternative stimuli or responses in both Spanish and English

Treatment:

• Determine in advance if the clinician or the participant will be responsible for controlling the screen/device during virtual treatment delivery; the participant-controlled method may promote competency and feelings of independence which may be motivating during treatment while the clinician-controlled method may allow the participant to better focus on the treatment tasks and complete more items each session

• Encourage participants to ask questions about or request explanations for task stimuli which might be confusing or unfamiliar to them (e.g., semantic features); given dialectal variation, clinicians should clarify when possible in the intervention language (e.g., Spanish or English)

• Utilize the annotation feature in Zoom to reorient participants during the session and/or deliver feedback about task errors

• Explore computerized features which might facilitate task completion without compromising the principles of the intervention (e.g., adding audio support for semantic features)