Not lost in translation: Learning from the non-native English-speaking heart transplant recipient experience

Heart failure (HF) is a disease that impacts populations of people differently and unequally. Racial and ethnic disparities have been well described. Specifically, minority patients experience a higher burden of risk factors for developing HF and reduced access to specialized care, including advanced therapies. Collectively, these disparities ultimately result in worse morbidity and mortality.[1] For patients with stage D HF, heart transplantation remains the preferred treatment for most patients, with the greatest improvements in overall symptoms and survival. Racial and ethnic disparities in outcomes after heart transplantation have previously been reported.[2] Once disparities are identified, the next critical step is to understand modifiable factors that can be targeted to improve care for all.

With this background, Vargas et al. performed a retrospective cohort study of adult patients who underwent heart transplantation from 2005 to 2022 at Columbia University Irving Medical Center.[3] Given the complexities of treatment after a heart transplantation, one potentially modifiable driver of care disparities is inadequate support for non-English speaking recipients. The study sought to explore the association between the primary language spoken and 1- and 5-year post-transplant survival. Several secondary patient-centered and quality of life outcomes were also analyzed. Survival analyses were performed adjusting for a priori clinical and demographic factors that are relevant determinants of HF outcomes.

The study included 1,066 heart transplant recipients, of whom 10% (n = 103) were non-native English speakers. In total, sixteen non-English primary languages were spoken, with Spanish being the most common (73/103, 71%). Compared to native English-speaking recipients, non-native English-speaking recipients were more likely to belong to underserved populations. These populations were less likely to identify as White, be US citizens, or have attended college, and more likely to have public insurance and reside in lower socioeconomic status neighborhoods.

Heart transplant recipients were followed for a mean of 6.2 (SD ±4.9) years. Overall, survival for the entire cohort was excellent, with less than five percent and 20 percent one- and five-year mortality rates. Survival was numerically better for non-native English-speaking individuals compared to native English-speaking individuals at 1 and 5 years in unadjusted and adjusted models; however, this did not achieve statistical significance (adjusted HR 0.24 [95% CI 0.06-1.01] at 1 year and adjusted HR 0.48 [95% CI 0.22-1.03] at 5 years). Similarly, secondary outcomes were numerically, but not statistically, better among non-native English-speaking as compared to native English-speaking heart transplant recipients. One notable exception was a statistically significant 39% decrease in coronary artery vasculopathy rate among non-native English-speaking as compared to native English-speaking heart transplant recipients.

While the authors didn’t explicitly state a hypothesis, the study sought to identify a potentially modifiable mechanism of disparities in heart transplantation. The implied hypothesis is that non-native English-speaking recipients would have worse post-transplant outcomes than native English-speaking recipients. At this single-center study in a resource-rich environment, the opposite was observed – there was a non-statistically significant but potentially clinically relevant, numerically improved 1 and 5-year survival among non-native English-speaking recipients, with similar observations for the patient-centered secondary outcomes. Although the total of over 1,000 patients, including more than 100 non-native English-speaking individuals, is relatively large for a single-center transplant study, the reality is that mortality remains fortunately rare in the early years post-transplantation, with only 2.2% at one year for non-native English-speaking individuals. Thus, a limitation that must be acknowledged and is critical to the interpretation of the results is the rarity of the outcome, given the dependence of statistical significance on sample size and event frequency.[4] While the p-value is just above the cutoff for “statistical significance,” the potential clinical importance of a considerable associated reduction, for example, a 52% reduced hazard of death at 5 years for non-native English-speaking individuals, should be considered to fully contextualize the implications of the findings for the transplant community. Overall, the study rejects the alternative hypothesis that there would be an association between non-native English-speaking recipients and worse heart transplantation outcomes and might support the reverse association in which non-native English-speaking recipients do better after heart transplantation.

The authors highlight several possible reasons why non-native English-speaking individuals did not experience inferior health outcomes as compared to native English speakers. Central to these hypotheses are the various ways social care, particularly language and culturally concordant services, are integrated into health care at Columbia.[5] These include translating educational materials into multiple languages,[6] having a diverse team of multilingual, multidisciplinary providers, and offering 24/7 professional interpreter services. Additionally, the hospital’s location in a neighborhood with a high Hispanic population may reduce barriers related to social determinants of health, such as transportation, and better support the patients it serves compared to health systems outside of underserved communities. Taken together, the research proposes possible evidence-based strategies that heart transplant centers can adopt, inspired by Columbia’s efforts, to improve access and quality of care for all recipients.

Additional limitations, specifically pointed out by the authors, need more context to interpret the results. There is a high likelihood of selection bias in including only transplant recipients, and the findings may not be widely applicable to centers lacking strong social care integration. There is a notable concern about selection bias, since patients who did not receive heart transplants were not included in the analysis. Non-native English-speaking individuals who reach transplantation may represent a more resource-supported group compared to those who never receive a transplant. The true denominator—non-native English speakers who could benefit from a transplant—is probably much larger than those who successfully go through the process. Key observations in the study support this. While the Hispanic population receiving transplants is diverse compared to national averages [7], the non-native English-speaking group was more affected by social determinants of health (e.g., lower neighborhood socioeconomic status), which have been shown before to influence outcomes for patients with advanced heart failure. A higher threshold for getting a transplant may lead to better post-transplant outcomes, as has been demonstrated with LVAD recipients.[8–11]

The study critically aimed to identify potentially modifiable disparities that could support targeted interventions to eliminate care differences. With the counterintuitive findings, the authors have laid the foundation for systematically studying why the traditionally underserved populations experienced better health outcomes. These results underscore the need to go beyond hypotheses about “why” differences in care exist and to focus on more rigorous evaluation of strategies to improve care for all.

Specifically, the fields of quality improvement and implementation science utilize rigorous, evidence-based methodologies that can be enhanced with a health equity lens to understand “why” differences in evidence-based care delivery and outcomes exist and to design programs aimed at improving care for everyone (Figure 1). Most traditional improvement efforts begin by identifying a gap in the delivery of evidence-based care (Step 1) and then examining the people, processes, and decision-making factors that influence care delivery (Step 2) before pinpointing the main barriers to equitable evidence-based care (Step 3).[12] In the case of heart transplantation care among non-native English speakers at Columbia, a modified approach can be equally insightful. The article by Vargas and colleagues highlights that a care gap may not be present and non-native English speakers may actually have improved outcomes (Step 1). Therefore, the following essential steps are to better describe how heart transplant care is delivered, who is involved in that care, and what people, policies, and procedures promote good outcomes (Steps 2 and 3). Findings from positive deviance investigations can offer vital insights into which implementation strategies (Step 4) may be beneficial for other centers seeking to develop an implementation playbook (Step 5) that can improve outcomes for non-native English-speaking heart transplant recipients (Step 6).

Figure 1. Proposed overview of an implementation science approach to developing, implementing, and evaluating interventions to reduce disparities.

The figure highlights six potential steps to undertake to develop, implement, and evaluate interventions to reduce disparities. Central to this is the involvement of stakeholders and consideration of the impact on disparities and equity throughout.

For example, one could imagine that there is extra time spent on reviewing medications to ensure understanding by non-English speaking recipients, which could lead to decreased rejection. If this hypothesis is confirmed through observation, the method used for teaching medications might be adopted and applied more widely, both in other centers and with different populations. Involving stakeholders is key to any systematic study, as their lived experience will reveal why an unexpected, clinically significant improvement happened. One example from the heart transplant field utilizing such methodology is the ongoing Seeking Objectivity in Allocation of Advanced Heart Failure (SOCIAL HF) study, which addresses observed biases in decision-making about advanced therapies. The study uses a robust implementation science-based multicomponent intervention to foster organizational and cultural change, promoting equitable access to advanced therapies, which has proven feasible in preliminary work.[13]

In conclusion, this work by Vargas and colleagues represents a significant step in health disparities research, progressing from mere observation to understanding the mechanisms that can be targeted. The entire field of advanced heart failure, including clinicians, researchers, and policymakers, should prioritize the systematic evaluation of disparities, as initiated by this work, to enhance care for all individuals with heart failure.