Limited English Proficiency, Cardiovascular Risk Factors, Cardiovascular Disease, and In-Hospital COVID-19 Outcomes

ABSTRACT

Objectives: Cardiovascular risk factors and history of cardiovascular disease are associated with greater morbidity and mortality in patients hospitalized with COVID-19. Limited English proficiency (LEP) has also been associated with worse outcomes in this setting, including requiring intensive care unit (ICU) level of care and in-hospital death. Whether non–English-language preference (NELP) modifies the association between cardiovascular risk factors or disease and outcomes in patients hospitalized with COVID-19 is unknown.

Study Design: Retrospective cohort study of adult patients admitted to a large New England health system between March 1 and December 31, 2020, who tested positive for COVID-19. NELP was defined as having a preferred language that was not English noted in the electronic health record.

Methods: Cardiovascular risk factors, history of cardiovascular disease, and NELP were related to the primary composite clinical outcome—death or ICU admission—using multivariable binary logistic regression adjusted for demographic and clinical characteristics. Interaction terms for NELP and model covariates were evaluated.

Results: Of 3582 patients hospitalized with COVID-19, 1024 (28.6%) had NELP; 812 (79.3%) of the patients with NELP received interpreter services. Death or ICU admission occurred in 794 (22.2%) of the hospitalized patients. NELP was not significantly associated with the primary composite outcome in unadjusted or adjusted analyses. In the adjusted analyses, only male gender, coronary artery disease, pulmonary circulatory disease, and liver disease significantly predicted the primary outcome. NELP did not modify the effect of these associations.

Conclusions: NELP was not significantly associated with odds of death or ICU admission, nor did it modify the association between cardiovascular risk factors or history of cardiovascular disease and this composite outcome. Because most patients with NELP received interpreter services, these findings may support the role of such services in ensuring equitable outcomes.

Am J Manag Care. 2024;30(6):251-256. https://doi.org/10.37765/ajmc.2024.89560

Takeaway Points

This article examines the impact of limited English proficiency on inpatient outcomes including death or intensive care unit (ICU) admission during the first year of the COVID-19 pandemic using data from a Rhode Island health care system.

• Limited English proficiency was not significantly associated with the composite of death or ICU admission.
• Limited English proficiency did not modify the association between cardiovascular risk factors or history of cardiovascular disease and this composite outcome.
• These findings highlight the importance of mitigating disparities in patient outcomes in vulnerable populations.

Health care inequities and unmet needs of patients with limited English proficiency (LEP) and non–English-language preference (NELP) have been highlighted by the COVID-19 pandemic.¹ As with cardiovascular risk factors and cardiovascular disease,^2-9 LEP has been associated with worse outcomes in patients with COVID-19, including those who have been hospitalized.^10-12 Specifically, patients with LEP have higher COVID-19 incidence, increased COVID-19 severity, and more frequent COVID-19–related complications and mortality.^13-15 The reasons for these associations are unclear but may include greater barriers to receiving medical care as an outpatient,^16,17 leading to more severe illness on presentation,^17-19 as well as provider efforts to limit exposure to COVID-19 during admission, resulting in less access to in-person interpreters and further communication difficulties secondary to masking.²⁰ Patients with LEP also have a greater incidence of cardiovascular risk factors and cardiovascular disease,^18,21-23 and it remains plausible that the association between LEP and poorer outcome in the setting of COVID-19 results from the greater burden of cardiovascular risk factors and/or disease in patients with LEP. Finally, it is possible that LEP and cardiovascular risk factor/disease burden both contribute to outcomes, either alone or synergistically. The aim of this study was to evaluate whether NELP modifies the association between traditional cardiovascular risk factors and/or history of cardiovascular disease and outcomes (death or intensive care unit [ICU] admission) in patients hospitalized with COVID-19.

METHODS

Study Design and Setting

This is a retrospective cohort study of adult patients admitted to Lifespan’s Rhode Island Hospital, Miriam Hospital, or Newport Hospital from March 1 through December 31, 2020, who tested positive for COVID-19.

Patients

All consecutive patients 18 years and older who were admitted to Lifespan’s Rhode Island Hospital, Miriam Hospital, or Newport Hospital from March 1 through December 31, 2020, and tested positive for COVID-19 were included. Comorbidities including cardiovascular risk factors, known cardiovascular disease, and COVID-19 status were determined using International Statistical Classification of Diseases, Tenth Revision (ICD-10) diagnostic codes (eAppendix [available at ajmc.com]) or were extracted from preexisting fields in the electronic health record. Cardiovascular risk factors included hypertension, hyperlipidemia, smoking status, diabetes, and obesity. Cardiovascular disease included history of coronary artery disease, percutaneous coronary intervention, myocardial infarction, heart failure, atrial fibrillation or flutter, transient ischemic attack, or stroke. Patients with missing information regarding preferred language (n = 2) were removed from the analysis.

Variables: Exposure and Outcomes

Exposure. Patients were considered to have NELP if they did not speak English as their primary language.²⁴ NELP was ascertained from the Epic electronic health record by identifying patients whose stated preferred language was not English. US Census Bureau definitions for race were employed; patients selecting more than 1 race were classified as being of other race.

Outcomes. The primary outcome was a composite of death or ICU admission. These outcomes were combined into a composite due to concerns over competing risks.^25,26 Secondary outcomes included the composite’s individual component end points (death, ICU admission), length of stay, need for ventilation, need for vasopressors, and discharge disposition. ICU admission included patients admitted directly to the ICU on admission and those transferred to the ICU during their hospital stay. ICD-10 codes used to define secondary outcomes appear in the eAppendix.

Data Source

Baseline demographic, medication, laboratory, clinical, and outcome data were obtained electronically from Epic’s electronic health record.

Statistical Analysis

Continuous variables are presented as the mean with SD or the median with IQR, depending upon their distribution. Continuous variables were compared using the Student t test or Wilcoxon rank sum test, as appropriate. Categorical variables are presented as frequencies and percentages and were compared using the χ² test. NELP was related to the primary composite clinical outcome using multivariable logistic regression adjusted for cardiovascular risk factors and history of cardiovascular disease. Demographics and clinical characteristics listed in Table 1 that were plausibly related to the primary outcome and for which P was less than or equal to .15 were selected a priori for inclusion in multivariable models. Two multivariable binary logistic regression analyses were constructed. The full model related NELP to the primary composite outcome after adjusting for age, gender, income, insurance type (Medicare, Medicaid, private, uninsured), cardiovascular risk factors (hypertension, hyperlipidemia, smoking status, diabetes), coronary artery disease, and other chronic medical conditions (dialysis use, pulmonary circulatory disease, liver disease, dementia, sleep apnea). The reduced model excluded income and insurance status due to concerns about collinearity between these demographic factors and preferred language. Race and ethnicity were excluded from both models because of similar concerns about collinearity between these demographic factors and preferred language. Interaction terms for NELP and each cardiovascular risk factor/disease state were entered into each model to assess whether NELP modified the relationships between cardiovascular risk factors/history of cardiovascular disease and the primary outcome. All analyses were performed with SAS 9.4 (SAS Institute Inc).

RESULTS

Patients

Of 3582 adult patients hospitalized with COVID-19, 1024 (28.6%) had NELP. Of patients with NELP, 80% had a preferred language of Spanish (eAppendix Table 1). The majority of patients with NELP received interpreter services during their hospitalization (n = 812; 79.3%). Patient characteristics according to NELP appear in Table 1. Overall mean (SD) age was 63.7 (20.3) years. Among all patients, 48.0% were women, 58.9% were White, 12.5% were Black, and 26.5% were Hispanic. Cardiovascular risk factors were common, including hypertension (68.2%), hyperlipidemia (48.7%), and diabetes (37.8%), as were cardiovascular diseases, including coronary artery disease (30.7%), heart failure (19.5%), and atrial fibrillation or flutter (18.3%).

Patients with NELP were younger (P < .001), were more likely to be male (P < .01), and had lower income (P < .001) than those with English-language preference. Patients with NELP had a lower burden of cardiovascular risk factors and cardiovascular diseases, including hypertension (P < .05), hyperlipidemia (P < .05), smoking (P < .001), coronary artery disease (P = .001), heart failure (P < .001), cardiac dysrhythmia including atrial fibrillation or flutter (P < .001), pulmonary embolism and deep vein thrombosis (P < .05), and stroke or transient ischemic attack (P < .01). However, patients with NELP were more likely to have diabetes (P < .001). Other medical comorbidities were less prevalent among those with NELP, including chronic obstructive pulmonary disease (P < .001), depression and anxiety (P < .001), dementia (P < .001), and cancer (P < .001).

Outcomes

The primary composite outcome, death or ICU admission, occurred in 794 (22.2%) of the hospitalized patients. NELP was not significantly associated with the primary composite outcome in unadjusted or adjusted analyses. No significant difference was seen in our primary composite outcome between patients with NELP who received interpreter services and those who did not.

In a model including age, gender, income, insurance, cardiovascular risk factors, cardiovascular disease, and other medical comorbidities (Figure), male gender, coronary artery disease, pulmonary circulatory disease, and liver disease were significantly associated with higher odds of the primary composite outcome. In a sensitivity analysis, when income and insurance were removed from the model to avoid introducing multicollinearity with NELP, results were similar, although heart failure also became a significant predictor (P < .05). Significant interactions were not identified between NELP and any of the included cardiovascular risk factors or history of cardiovascular disease states when evaluating the primary composite outcome (male: P = .8; hypertension: P = .24; hyperlipidemia: P = .95; smoker: P = .29; diabetes: P = .41; coronary artery disease: P = .71; and heart failure: P = .59).

Unadjusted secondary outcomes appear in Table 2. NELP was not associated with unadjusted rates of in-hospital death; however, patients with NELP had significantly increased odds vs patients with English-language preference of requiring admission to an ICU (19.2% vs 16.0%; P = .02), mechanical ventilation (10.3% vs 6.9%; P < .001), or vasopressors (10.1% vs 7.5%; P = .008). Nevertheless, patients with NELP had increased odds of being discharged to home rather than to a skilled nursing facility or to hospice (P < .001). NELP was not associated with increased length of stay in the hospital (P = .324).

DISCUSSION

In a retrospective analysis of patients hospitalized with COVID-19 at a large New England health system, NELP was not significantly associated with the risk of composite death or ICU admission, nor did it modify the association between cardiovascular risk factors or cardiovascular disease and this outcome. Further, among patients with NELP, no significant difference was seen in death or ICU admission based on interpreter use. In unadjusted analyses, NELP was associated with a higher likelihood of ICU admission and need for mechanical ventilation or vasopressors but a lower likelihood of discharge to a location other than home.

The prevalence of NELP observed in our study (28.6%) is similar to rates of LEP cited in other studies of COVID-19.^10,11 Two prior articles have evaluated the association between LEP and in-hospital COVID-19 outcomes. In a retrospective cohort study, Chua et al found that LEP was not independently associated with ICU death or ICU length of stay but was associated with increased hospital length of stay (mean difference, 4.12 days; 95% CI, 1.72-6.53; P < .001).¹⁰ Gonzalez et al found that among patients 55 years and older, living in areas with relatively high LEP predicted life-threatening complications including intubation and death.¹¹ We observed no difference in our composite outcome, ICU admission or death, according to NELP. This finding may suggest that support in place at the time of hospitalization (ie, telephonic interpreter services) was adequate to mitigate any potential increased risk of adverse outcome. Alternatively, it is possible that our null result may be due to heterogeneity of the population characterized as having NELP or heterogeneity in the population defined as having COVID-19. We included patients admitted for treatment of COVID-19 symptoms and patients admitted for other medical problems but who tested positive for COVID-19 during admission. We observed no difference in our composite outcome of ICU admission or death according to use of interpreter services. We did not quantify the degree of interpreter use while inpatient, so significant heterogeneity in this population may explain the null result.

In our study based on unadjusted analyses, NELP was significantly associated with the need for an ICU stay and ventilator and vasopressor use, similar to the results noted by Gonzalez et al.¹¹ This observation may indicate that patients with NELP have increased disease severity on admission, as has been observed in other studies,¹⁷ or that patients with NELP may be at increased odds of deteriorating while hospitalized, possibly due to complications associated with communication barriers. Differences observed in rates of intubation, death, and ICU admission may also represent cultural differences in end-of-life care and decision-making. Barwise et al found that patients with LEP were more likely to be “full code” and less likely to transition to “do not resuscitate” or “comfort measures only” and were therefore more likely to receive mechanical ventilation.²⁵

Despite patients with NELP having higher rates of ICU admission, they were at increased odds of being discharged to home rather than to a skilled nursing facility or to hospice. This may be due to referral bias or to higher rates of uninsurance or Medicaid insurance, resulting in access to fewer facilities. There may also be a cultural difference whereby patients or their families prefer to care for their loved ones at home. Finally, patients with NELP were healthier overall, with fewer comorbidities than patients with English as a preferred language. Therefore, it is possible that despite high rates of ICU admission, they may have been less in need of support at the time of discharge.

In our study, male gender and prevalent coronary artery disease were significantly associated with death or ICU admission. Ssentongo et al also found cardiovascular disease to be significantly associated with death,⁷ whereas Gonzalez et al found male sex to be associated with intubation and death.¹¹

Limitations

There are noteworthy limitations to our study. We utilized ICD-10 codes to define clinical outcomes; resultant undercoding or miscoding may have introduced inaccuracies in our data. NELP was ascertained using preferred language in the electronic health record. Although this definition has been used in prior literature to define LEP,^10,18 it may not accurately reflect patients’ preferred language or need for interpreter services. We were not able to quantify utilization of interpreter services, an intervention that might buffer an association between NELP and outcome,^26,27 and therefore characterized this exposure in a dichotomous fashion. If a graded relationship between interpreter use and outcome had been present, we would not have been able to discern this. Further, we were unable to identify the reason that 20.9% of patients with NELP did not receive interpreter services, as this information was not available in the medical record. Our sample size precluded analysis of patient subgroups with NELP according to preferred language. We were also not able to better characterize the patient population classified as being of other race because this information was not available in the data set. Finally, this study focused on data during the first wave of the COVID-19 pandemic, preceding vaccinations and newer therapeutic interventions; therefore, the generalizability of this study to current COVID-19 hospitalizations is unknown.

CONCLUSIONS

Among patients with COVID-19 hospitalized at a large health system, NELP was not associated with the likelihood of death or ICU admission, nor did it modify the association between cardiovascular risk factors or history of cardiovascular disease and these outcomes. Given the high rate at which interpreter services were utilized, our findings may suggest that use of such services during a pandemic is protective against adverse outcomes among patients with NELP. Future studies should explore whether a graded relationship exists between interpreter services utilization and patient outcomes and whether the modality employed (telephone, video, or in person) is related to these outcomes. Further, discerning the reason that interpreter services are not utilized for patients with NELP should also be explored.

Author Affiliations: Division of Cardiology (JDA), Department of Medicine (JLB, CF), Warren Alpert Medical School of Brown University, Providence, RI; statistical consultant (KFK), Kansas City, MO; Center for Health Policy and Health Services Research, Henry Ford Health (DRN), Detroit, MI; Lifespan Cardiovascular Institute (JDA), Providence, RI; Division of Cardiology, Department of Medicine, Michigan State University College of Human Medicine (HDA), Detroit, MI; Heart and Vascular Institute, Henry Ford Health (HDA), Detroit, MI.

Source of Funding: None.

Author Disclosures: Dr Abbott has consulted for Abbott (percutaneous coronary intervention), Medtronic (bifurcation percutaneous coronary intervention), Penumbra (thrombectomy), and Recor (clinical trial adjudication renal denervation); is employed by Lifespan Physician Group; has received grants from Boston Scientific (steering committee, site principal investigator [PI] on AGENT IDE trial [institutional]), Med Alliance (site PI on Selutions4ISR trial [institutional]), and Shockwave (site PI on EMPOWER CAD trial [institutional]); attended as faculty 2023 meetings/conferences of Cardiovascular Research Technologies, Society for Cardiovascular Angiography & Interventions, American Heart Association, and Transcatheter Cardiovascular Therapeutics; and receives royalties as an UpToDate contributor. Dr Aronow is a consultant for Recor and Silk Road Medical. The remaining authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (JLB, KFK, DRN, JDA, HDA); acquisition of data (JLB, CF, HDA); analysis and interpretation of data (JLB, KFK, HDA); drafting of the manuscript (JLB, CF, JDA, HDA); critical revision of the manuscript for important intellectual content (CF, DRN, JDA, HDA); statistical analysis (KFK, HDA); and supervision (DRN, HDA).

Address Correspondence to: Julia L. Berkowitz, MD, Warren Alpert Medical School of Brown University, 593 Eddy St, Providence, RI 02903. Email: jberkowitz1@lifespan.org.

REFERENCES

1. Ortega P, Martínez G, Diamond L. Language and health equity during COVID-19: lessons and opportunities. J Health Care Poor Underserved. 2020;31(4):1530-1535. doi:10.1353/hpu.2020.0114

2. Xiong TY, Huang FY, Liu Q, et al. Hypertension is a risk factor for adverse outcomes in patients with coronavirus disease 2019: a cohort study. Ann Med. 2020;52(7):361-366. doi:10.1080/07853890.2020.1802059

3. Parveen R, Sehar N, Bajpai R, Agarwal NB. Association of diabetes and hypertension with disease severity in covid-19 patients: a systematic literature review and exploratory meta-analysis. Diabetes Res Clin Pract. 2020;166:108295. doi:10.1016/j.diabres.2020.108295

4. Li X, Guan B, Su T, et al. Impact of cardiovascular disease and cardiac injury on in-hospital mortality in patients with COVID-19: a systematic review and meta-analysis. Heart. 2020;106(15):1142-1147. doi:10.1136/heartjnl-2020-317062

5. Fried MW, Crawford JM, Mospan AR, et al. Patient characteristics and outcomes of 11 721 patients with coronavirus disease 2019 (COVID-19) hospitalized across the United States. Clin Infect Dis. 2021;72(10):e558-e565. doi:10.1093/cid/ciaa1268

6. Cheng X, Cai G, Wen X, et al. Clinical characteristics and fatal outcomes of hypertension in patients with severe COVID-19. Aging (Albany NY). 2020;12(23):23436-23449. doi:10.18632/aging.104019

7. Ssentongo P, Ssentongo AE, Heilbrunn ES, Ba DM, Chinchilli VM. Association of cardiovascular disease and 10 other pre-existing comorbidities with COVID-19 mortality: a systematic review and meta-analysis. PLoS One. 2020;15(8):e0238215. doi:10.1371/journal.pone.0238215

8. Bauer AZ, Gore R, Sama SR, et al. Hypertension, medications, and risk of severe COVID-19: a Massachusetts community-based observational study. J Clin Hypertens (Greenwich). 2021;23(1):21-27. doi:10.1111/jch.14101

9. Izcovich A, Ragusa MA, Tortosa F, et al. Prognostic factors for severity and mortality in patients infected with COVID-19: a systematic review. PLoS One. 2020;15(11):e0241955. doi:10.1371/journal.pone.0241955

10. Chua IS, Shi SM, Jia Z, et al. Differences in end-of-life care between COVID-19 inpatient decedents with English proficiency and limited English proficiency. J Palliat Med. 2022;25(11):1629-1638. doi:10.1089/jpm.2021.0541

11. Gonzalez CJ, Hogan CJ, Rajan M, et al. Predictors of life-threatening complications in relatively lower-risk patients hospitalized with COVID-19. PLoS One. 2022;17(2):e0263995. doi:10.1371/journal.pone.0263995

12. Wang PG, Brisbon NM, Hubbell H, et al. Is the gap closing? comparison of sociodemographic disparities in COVID-19 hospitalizations and outcomes between two temporal waves of admissions. J Racial Ethn Health Disparities. 2023;10(2):593-602. doi:10.1007/s40615-022-01249-y

13. Karmakar M, Lantz PM, Tipirneni R. Association of social and demographic factors with COVID-19 incidence and death rates in the US. JAMA Netw Open. 2021;4(1):e2036462. doi:10.1001/jamanetworkopen.2020.36462

14. Liu A, Patel A, Pierce A, Fowler R. Variations in presentation and management of COVID-19 inpatients by race and ethnicity in a large Texas metroplex. Disaster Med Public Health Prep. 2021;17:e21. doi:10.1017/dmp.2021.224

15. Samuel LJ, Gaskin DJ, Trujillo AJ, Szanton SL, Samuel A, Slade E. Race, ethnicity, poverty and the social determinants of the coronavirus divide: U.S. county-level disparities and risk factors. BMC Public Health. 2021;21(1):1250. doi:10.1186/s12889-021-11205-w

16. Timmins CL. The impact of language barriers on the health care of Latinos in the United States: a review of the literature and guidelines for practice. J Midwifery Womens Health. 2002;47(2):80-96. doi:10.1016/s1526-9523(02)00218-0

17. Velasco F, Yang DM, Zhang M, et al. Association of healthcare access with intensive care unit utilization and mortality in patients of Hispanic ethnicity hospitalized with COVID-19. J Hosp Med. 2021;16(11):659-666. doi:10.12788/jhm.3717

18. Biswas S, Dinh D, Lucas M, et al. Impact of limited English proficiency on presentation and clinical outcomes of patients undergoing primary percutaneous coronary intervention. Eur Heart J Qual Care Clin Outcomes. 2020;6(4):254-262. doi:10.1093/ehjqcco/qcz061

19. Joseph NP, Reid NJ, Som A, et al. Racial and ethnic disparities in disease severity on admission chest radiographs among patients admitted with confirmed coronavirus disease 2019: a retrospective cohort study. Radiology. 2020;297(3):E303-E312. doi:10.1148/radiol.2020202602

20. Hampton T, Crunkhorn R, Lowe N, et al. The negative impact of wearing personal protective equipment on communication during coronavirus disease 2019. J Laryngol Otol. 2020;134(7):577-581. doi:10.1017/s0022215120001437

21. Kim EJ, Kim T, Paasche-Orlow MK, Rose AJ, Hanchate AD. Disparities in hypertension associated with limited English proficiency. J Gen Intern Med. 2017;32(6):632-639. doi:10.1007/s11606-017-3999-9

22. Eamranond PP, Legedza AT, Diez-Roux AV, et al. Association between language and risk factor levels among Hispanic adults with hypertension, hypercholesterolemia, or diabetes. Am Heart J. 2009;157(1):53-59. doi:10.1016/j.ahj.2008.08.015

23. Latif Z, Makuvire T, Feder SL, et al. Challenges facing heart failure patients with limited English proficiency: a qualitative analysis leveraging interpreters’ perspectives. JACC Heart Fail. 2022;10(6):430-438. doi:10.1016/j.jchf.2022.02.011

24. Commonly asked questions and answers regarding limited English proficient (LEP) individuals. LEP.gov. Updated April 2011. Accessed July 12, 2022. https://www.lep.gov/sites/lep/files/media/document/2020-03/042511_QA_LEP_General_0.pdf

25. Barwise A, Jaramillo C, Novotny P, et al. Differences in code status and end-of-life decision making in patients with limited English proficiency in the intensive care unit. Mayo Clin Proc. 2018;93(9):1271-1281. doi:10.1016/j.mayocp.2018.04.021

26. Karliner LS, Jacobs EA, Chen AH, Mutha S. Do professional interpreters improve clinical care for patients with limited English proficiency? a systematic review of the literature. Health Serv Res. 2007;42(2):727-754. doi:10.1111/j.1475-6773.2006.00629.x

27. Flores G. The impact of medical interpreter services on the quality of health care: a systematic review. Med Care Res Rev. 2005;62(3):255-299. doi:10.1177/1077558705275416