Racial and Neighborhood-level Disparities in COVID-19 Incidence Among Patients on Hemodialysis in

Journal: Journal of the American Society of Nephrology

Manuscript ID JASN-2020-11-1606.R1

Manuscript Type: Original Article - Clinical Research

Date Submitted by the 24-Feb-2021 Author:

Complete List of Authors: Tummalapalli, Sri Lekha; , Division of Healthcare Delivery Science & Innovation, Department of Population Health Sciences; Rogosin Institute; Weill Cornell Medicine, Division of Nephrology & , Department of Medicine Silberzweig, Jeffrey; Rogosin Institute; Weill Cornell Medicine, Division of Nephrology & Hypertension, Department of Medicine Cukor, Daniel; Rogosin Institute Lin, Jonathan; Rogosin Institute; Weill Cornell Medicine, Division of Nephrology & Hypertension, Department of Medicine Barbar, Tarek; Weill Cornell Medicine, Division of Nephrology & Hypertension, Department of Medicine Liu, Yao; Rogosin Institute Kim, Kwan; Rogosin Institute Parker, Thomas; Rogosin Institute; Weill Cornell Medicine, Department of Biochemistry Levine, Daniel; Rogosin Institute; Weill Cornell Medicine, Department of Biochemistry Ibrahim, Said; Weill Cornell Medicine, Division of Healthcare Delivery Science & Innovation, Department of Population Health Sciences

Keywords: COVID-19, hemodialysis, Disparities, Social determinants of health

Journal of the American Society of Nephrology Page 1 of 34

1 2 3 Authors: Tummalapalli, Sri Lekha; Silberzweig, Jeffrey; Cukor, Daniel; Lin, Jonathan; Barbar, Tarek; Liu, 4 Yao; Kim, Kwan; Parker, Thomas; Levine, Daniel; Ibrahim, Said 5 6 7 Title: Racial and Neighborhood-level Disparities in COVID-19 Incidence Among Patients on Hemodialysis 8 in New York City 9 10 Running title: Racial and Neighborhood Disparities in COVID-19 11 12 Manuscript Type: Original Article - Clinical Research 13 14 15 Manuscript Category: Clinical dialysis 16 17 Funders: National Institute of and Digestive and Kidney Diseases, (Grant / Award Number: 18 'F32DK122627') 19 20 National Kidney Foundation, (Grant / Award Number: 'Young Investigator Grant') 21 22 Financial Disclosure: No Dr. Tummalapalli received consulting fees from Bayer AG unrelated to the 23 submitted work. The remaining authors have nothing to disclose. D. Cukor reports Research Funding 24 from NIH. D. Levine reports Patents and Inventions with The Rogosin Institute. J. Silberzweig reports 25 Consultancy Agreements with Kaneka Pharma, Bayer Pharmaceuticals, Alkahest Biotech; Scientific 26 Advisor or Membership with American Society of Nephrology: COVID-19 Response team, Emergency 27 28 Partnership Initiative. 29 30 Study Group/Organization Name: CUST_STUDY_GROUP/ORGANIZATION_NAME :No data available. 31 32 Study Group Members’ Names: CUST_STUDY_GROUP_MEMBERS :No data available. 33 34 Total number of words: 2775 35 36 37 Abstract: Background The coronavirus disease 2019 (COVID-19) pandemic has 38 disproportionately affected socially disadvantaged populations. Whether disparities in COVID-19 39 incidence related to race/ethnicity and socioeconomic factors exist in the hemodialysis population is 40 unknown. 41 Methods Our study involved patients receiving in-center hemodialysis in New York City. We 42 used a validated index of neighborhood social vulnerability, the Social Vulnerability Index (SVI), which 43 comprises 15 census tract–level indicators organized into four themes: socioeconomic status, 44 household composition and disability, minority status and language, and housing type and 45 46 transportation. We examined the association of race/ethnicity and the SVI with symptomatic COVID-19 47 between March 1, 2020, and August 3, 2020. COVID-19 cases were ascertained using PCR testing. We 48 performed multivariable logistic regression to adjust for demographics, individual-level social factors, 49 dialysis-related medical history, and dialysis facility factors. 50 Results Of the 1378 patients on hemodialysis in the study, 247 (17.9%) developed symptomatic 51 COVID-19. In adjusted analyses, non-Hispanic Black and Hispanic patients had significantly increased 52 odds of COVID-19 compared with non-Hispanic White patients. Census tract–level overall SVI, 53 modeled continuously or in quintiles, was not associated with COVID-19 in unadjusted or adjusted 54 55 analyses. Among non-Hispanic White patients, the socioeconomic status SVI theme, the minority status 56 57 58 59 60 Journal of the American Society of Nephrology Page 2 of 34

1 2 3 and language SVI theme, and housing crowding were significantly associated with COVID-19 in 4 unadjusted analyses. 5 6 Conclusions Among patients on hemodialysis in New York City, there were substantial 7 racial/ethnic disparities in COVID-19 incidence not explained by neighborhood-level social vulnerability. 8 Neighborhood-level socioeconomic status, minority status and language, and housing crowding were 9 positively associated with acquiring COVID-19 among non-Hispanic Whites. Our findings suggest that 10 socially vulnerable patients on dialysis face disparate COVID-19–related exposures, requiring targeted 11 risk-mitigation strategies. 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Journal of the American Society of Nephrology Page 3 of 34 Copyright 2021 by ASN, Published Ahead of Print on 6/3/21, Accepted/Unedited Version.

1 2 Significance Statement 3 4 Coronavirus disease 2019 (COVID-19) has disproportionately affected socially disadvantaged 5 6 populations. Whether racial/ethnic and socioeconomic disparities in COVID-19 incidence exist 7 8 in the hemodialysis population is unknown. The authors examined the association of 9 10 race/ethnicity and a validated neighborhood-level index of social vulnerability (the Social 11 Vulnerability Index) with acquiring symptomatic COVID-19 among patients receiving in-center 12 13 hemodialysis from a dialysis organization in New York City. They found substantial 14 15 racial/ethnic disparities in COVID-19 incidence; Black and Hispanic individuals on hemodialysis 16 17 were more likely than non-Hispanic White patients to acquire COVID-19. Neighborhood-level 18 socioeconomic status, minority status and language, and housing crowding were positively 19 20 associated with COVID-19 acquisition among non-Hispanic white patients but did not explain 21 22 racial/ethnic disparities. These findings indicate that targeted strategies are needed to mitigate 23 excess COVID-19 risk among socially vulnerable patients on dialysis. 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Journal of the American Society of Nephrology Copyright 2021 by ASN, Published Ahead of Print on 6/3/21, Accepted/Unedited Version. Page 4 of 34

1 2 Racial and Neighborhood-Level Disparities in COVID-19 Incidence among Patients on 3 4 Hemodialysis in New York City 5 6 7 Sri Lekha Tummalapalli1,2,3, Jeffrey Silberzweig2,3, Daniel Cukor2, Jonathan T. Lin2,3, Tarek 8 3 2 2 2,4 2,4 1 9 Barbar , Yao Liu , Kwan Kim , Thomas S. Parker , Daniel M. Levine , Said A. Ibrahim 10 11 12 1. Division of Healthcare Delivery Science & Innovation, Department of Population Health 13 14 Sciences, Weill Cornell Medicine, New York, NY 15 16 2. The Rogosin Institute, New York, NY 17 3. Division of Nephrology & Hypertension, Department of Medicine, Weill Cornell 18 19 Medicine, New York, NY 20 21 4. Department of Biochemistry, Weill Cornell Medicine, New York, NY 22 23 24 25 26 27 28 Running Title: Disparities in COVID-19 Incidence in Hemodialysis Patients 29 30 31 32 33 34 35 36 37 38 Corresponding Author: 39 Sri Lekha Tummalapalli 40 Division of Healthcare Delivery Science & Innovation 41 Department of Population Health Sciences 42 th 43 402 East 67 Street 44 New York, NY 10065 45 Ph: 646-962-8001 46 E-mail: [email protected] 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Journal of the American Society of Nephrology Page 5 of 34 Copyright 2021 by ASN, Published Ahead of Print on 6/3/21, Accepted/Unedited Version.

1 2 3 Abstract 4 5 6 Background The coronavirus disease 2019 (COVID-19) pandemic has disproportionately 7 affected socially disadvantaged populations. Whether disparities in COVID-19 incidence related 8 9 to race/ethnicity and socioeconomic factors exist in the hemodialysis population is unknown. 10 11 Methods Our study involved patients receiving in-center hemodialysis in New York City. We 12 13 used a validated index of neighborhood social vulnerability, the Social Vulnerability Index 14 (SVI), which comprises 15 census tract–level indicators organized into four themes: 15 16 socioeconomic status, household composition and disability, minority status and language, and 17 18 housing type and transportation. We examined the association of race/ethnicity and the SVI with 19 20 symptomatic COVID-19 between March 1, 2020, and August 3, 2020. COVID-19 cases were 21 ascertained using PCR testing. We performed multivariable logistic regression to adjust for 22 23 demographics, individual-level social factors, dialysis-related medical history, and dialysis 24 25 facility factors. 26 Results Of the 1378 patients on hemodialysis in the study, 247 (17.9%) developed symptomatic 27 28 COVID-19. In adjusted analyses, non-Hispanic Black and Hispanic patients had significantly 29 30 increased odds of COVID-19 compared with non-Hispanic White patients. Census tract–level 31 32 overall SVI, modeled continuously or in quintiles, was not associated with COVID-19 in 33 unadjusted or adjusted analyses. Among non-Hispanic White patients, the socioeconomic status 34 35 SVI theme, the minority status and language SVI theme, and housing crowding were 36 37 significantly associated with COVID-19 in unadjusted analyses. 38 39 Conclusions Among patients on hemodialysis in New York City, there were substantial 40 racial/ethnic disparities in COVID-19 incidence not explained by neighborhood-level social 41 42 vulnerability. Neighborhood-level socioeconomic status, minority status and language, and 43 44 housing crowding were positively associated with acquiring COVID-19 among non-Hispanic 45 Whites. Our findings suggest that socially vulnerable patients on dialysis face disparate COVID- 46 47 19–related exposures, requiring targeted risk-mitigation strategies. 48 49 50 51 52 53 54 55 56 57 58 59 60 Journal of the American Society of Nephrology Copyright 2021 by ASN, Published Ahead of Print on 6/3/21, Accepted/Unedited Version. Page 6 of 34

1 2 Introduction 3 4 The novel coronavirus disease 2019 (COVID-19) pandemic has disproportionately 5 6 affected socially disadvantaged populations, including Black and Hispanic individuals,1-3 those 7 8 4 5 9 with limited English proficiency, and persons of low socioeconomic status. Individuals with 10 11 have also been disproportionately affected; an analysis of early Medicare claims 12 13 data confirmed that individuals with kidney failure were at 3.5 times the risk of acquiring 14 15 6 16 COVID-19 compared with other Medicare fee-for-service beneficiaries. 17 18 While COVID-19 hospitalization and mortality outcomes among patients on dialysis 19 20 have been reported,7-10 there have been fewer reports characterizing factors associated with 21 22 COVID-19 incidence,11 and in particular, social factors such as race/ethnicity and neighborhood- 23 24 25 level characteristics. One study of a large national sample of patients on hemodialysis found that 26 27 patients who were Non-Hispanic Black, Hispanic, or resided in high poverty or majority Black 28 29 and Hispanic neighborhoods were more likely to have severe acute respiratory syndrome 30 31 12 32 coronavirus 2 (SARS-CoV-2) antibodies. Greater examination of neighborhood-level 33 34 characteristics may reveal drivers of racial/ethnic disparities seen in COVID-19 incidence in the 35 36 dialysis population, with greater viral transmission in neighborhoods with more housing 37 38 13,14 39 crowding and a higher essential workforce. 40 41 The Social Vulnerability Index (SVI) is a composite of census indicators used by the 42 43 Centers for Disease Control and Prevention for emergency preparedness and disaster response. 44 45 County-level SVI is strongly associated with county-level COVID-19 cases in the general 46 47 48 population, but how SVI associates with individual-level COVID-19 cases in the dialysis 49 50 population is unknown.5,15-17 Furthermore, the relationship between race/ethnicity and SVI is 51 52 incompletely characterized. Therefore, we examined racial/ethnic differences in COVID-19 53 54 55 incidence among patients on hemodialysis in New York City during the first wave of the 56 57 58 59 60 Journal of the American Society of Nephrology Page 7 of 34 Copyright 2021 by ASN, Published Ahead of Print on 6/3/21, Accepted/Unedited Version.

1 2 COVID-19 pandemic, and assessed if SVI explained racial/ethnic differences in COVID-19 3 4 incidence. 5 6 7 8 9 Methods 10 11 Study Design and Population 12 13 We performed a retrospective cohort study of prevalent in-center hemodialysis patients at 14 15 16 eight dialysis units within a non-profit nephrology organization in New York City. We included 17 18 patients receiving in-center hemodialysis as of March 1, 2020 and followed them until August 3, 19 20 2020. We excluded patients who had acute kidney injury requiring dialysis (AKI-D) (6 patients), 21 22 those on home dialysis modalities (5 patients), and those with missing SVI values (11 patients). 23 24 25 As a supplemental data source, we obtained publicly available data on COVID case rates by race 26 27 and ZIP Code Tabulation Area (ZCTA) from the New York City Department of Health and 28 29 Mental Hygiene.18 30 31 32 Predictor, Outcome, and Covariates 33 34 Our primary predictor was patient race/ethnicity obtained from the EHR, and classified as 35 36 1) Non-Hispanic white, 2) Non-Hispanic Black, 3) Hispanic, 4) Asian or Pacific Islander, or 5) 37 38 39 Other, Unknown, or Missing. Our co-primary predictor was neighborhood social vulnerability as 40 41 measured by overall SVI, the four SVI themes, and housing crowding. We used census tract- 42 43 level SVIs for New York State calculated from the 2014-2018 American Community Survey 5- 44 45 year estimates.19 The SVI is a composite of 15 census indicators organized into 4 themes: 1) 46 47 48 socioeconomic status (% of persons below poverty, % unemployed, percentile per capita income, 49 50 % with no high school diploma), 2) household composition & disability (% aged 65 or older, % 51 52 aged 17 or younger, % with a disability, % single parent households), 3) minority status & 53 54 55 language (% minority, % who speak English “less than well”), and 4) housing type & 56 57 transportation (% in multiunit housing, % in mobile homes, % households with more people than 58 59 60 Journal of the American Society of Nephrology Copyright 2021 by ASN, Published Ahead of Print on 6/3/21, Accepted/Unedited Version. Page 8 of 34

1 2 rooms [housing crowding], % households with no vehicle, % in group quarters). Each census 3 4 tract is given a percentile ranking for the four SVI themes and an overall ranking, compared with 5 6 other census tracts in New York State. We geocoded patient addresses to census tracts and 7 8 9 ZCTAs using the United States Census Bureau website and linked them to census tract-level 10 11 SVI.20 Overall SVI, the four SVI themes, and housing crowding percentiles were modeled as 12 13 continuous variables. Additionally, overall SVI was divided into quintiles within our study 14 15 16 population. 17 18 Our primary outcome of interest was confirmed or presumed COVID-19. We created a 19 20 COVID-19 tool in REDCap and surveyed our electronic health record weekly using a direct data 21 22 connection and automated scripting to identify and track patients with COVID-19. Patients were 23 24 25 tested for COVID-19 during routine clinical care when there was clinical suspicion based on 26 27 signs and symptoms. Testing was performed on nasopharyngeal swab specimens sent for 28 29 reverse-transcriptase–polymerase-chain-reaction (PCR) assay for SARS-CoV-2. A positive result 30 31 32 from a SARS-CoV-2 PCR nasal swab was verified by chart review and classified as a confirmed 33 34 COVID-19 case. Because NYC was limiting PCR testing early in the pandemic to primarily 35 36 hospitalized patients, patients with clinical signs and symptoms of cough, fever, or respiratory 37 38 39 symptoms were classified as presumed cases based on manual chart review. 40 41 Covariates included patient demographics (age, sex), individual-level social factors 42 43 (employment, marital status, transportation type to dialysis), dialysis-related medical history 44 45 (vintage [length of time on dialysis], cause of kidney failure), and dialysis facility factors (unit 46 47 48 fixed effects and dialysis unit SVI). Individual-level social factors were pulled from structured 49 50 fields in the EHR. Transportation to dialysis was classified as 1) van service (including 51 52 ambulette, ambulance, and Access-A-Ride), 2) private vehicle (including taxi services), or 3) 53 54 55 public transportation (including bus or subway). A missing indicator category was used for 56 57 missing covariates; multiple imputation was not employed because covariates were thought to be 58 59 60 Journal of the American Society of Nephrology Page 9 of 34 Copyright 2021 by ASN, Published Ahead of Print on 6/3/21, Accepted/Unedited Version.

1 2 missing not at random. Missingness was as follows: race/ethnicity (9%), employment (35%), 3 4 marital status (22%), transportation type (44%). 5 6 Statistical Analysis 7 8 9 We first compared the demographics, individual-level social factors, and dialysis-related 10 11 medical history of patients on hemodialysis residing in neighborhoods with high SVI (higher 12 13 than median SVI in our study cohort, reflecting higher social vulnerability) versus low SVI 14 15 16 (lower than median SVI, lower social vulnerability). We then compared characteristics of 17 18 patients on hemodialysis who were COVID-19-positive and not COVID-19-positive. Differences 19 20 in characteristics were assessed using chi-squared tests for categorical variables and Wilcoxon 21 22 rank-sum tests for continuous variables. We also examined the association of ZCTA-level 23 24 25 COVID-19 cumulative incidence (COVID cumulative PCR+ count divided by ZCTA population 26 27 denominator) within the New York City general population with COVID-19 among patients on 28 29 hemodialysis residing in ZCTAs, using unadjusted logistic regression. 30 31 32 We then performed logistic regression to examine the association between patient 33 34 race/ethnicity and overall SVI quintile with COVID-19 positivity, accounting for race/ethnicity 35 36 and SVI interactions. We employed two nested multivariable regression models: Model 1 37 38 39 adjusted for age, sex, race/ethnicity or SVI, individual-level social factors, and dialysis-related 40 41 medical history. Model 2 additionally adjusted for dialysis facility factors. 42 43 To examine if our findings were consistent over time, as a sensitivity analysis we 44 45 stratified our results by three time periods: March 1 – March 21 (pre-stay-at-home order), March 46 47 48 22 – April 14 (pre-mask order), and April 15 – June 7 (before reopening). We also performed a 49 50 time-to-event analysis using Fine and Gray subdistribution hazard models to estimate the 51 52 association of race/ethnicity and overall SVI quintile with the subhazard of COVID-19, 53 54 55 accounting for the competing risk of death and censoring at transplant, dialysis withdrawal, and 56 57 modality change. All analyses were clustered at the level of the dialysis unit. Our study was 58 59 60 Journal of the American Society of Nephrology Copyright 2021 by ASN, Published Ahead of Print on 6/3/21, Accepted/Unedited Version. Page 10 of 34

1 2 approved by the Weill Cornell Medicine Institutional Review Board. Data analyses were 3 4 performed using Stata/IC, version 15.1 (StataCorp) and R version 4.0.2 statistical software. 5 6 7 8 9 Results 10 11 Patient Characteristics 12 13 Of the 1378 patients on hemodialysis in our study cohort, 294 (21.3%) were Non- 14 15 16 Hispanic white, 578 (41.9%) were Non-Hispanic Black, 207 (15.0%) were Hispanic, 174 17 18 (12.6%) were Asian or Pacific Islander, and 125 (9.0%) had other, unknown, or missing 19 20 race/ethnicity. Patients on hemodialysis lived in census tracts with higher social vulnerability 21 22 (Figure 1, mean SVI 68) compared with the general population of New York City (mean 23 24 25 population-weighted SVI 61 in Manhattan, Queens, and Brooklyn). 26 27 Patients who were non-Hispanic Black and Hispanic were younger, more likely to be 28 29 disabled, and had a longer dialysis vintage than patients in other race/ethnic categories 30 31 32 (Supplemental Table 1). Sex, marital status, transportation to dialysis, and primary cause of 33 34 kidney failure also differed by race/ethnicity. Patients living in high SVI neighborhoods were 35 36 younger, more likely to be female (46% vs. 39%), more likely to be non-Hispanic Black (51% 37 38 39 vs. 33%) and Hispanic (17% vs. 13%), less likely to be married (27% vs. 39%), and more likely 40 41 to have diabetes (44% vs. 38%) as the cause of their kidney failure, compared with those living 42 43 in low SVI neighborhoods (Table 1). 44 45 COVID-19 and Patient Characteristics 46 47 48 A total of 247 (17.9%) patients developed symptomatic COVID-19, of whom 230 49 50 (93.1%) were laboratory confirmed positive and 17 (6.9%) were presumed positive. The timing 51 52 of COVID-19 cases among patients on hemodialysis coincided with the New York City-wide 53 54 55 surge (Supplemental Figure 1). Patients with COVID-19 were more likely to be single (30% vs. 56 57 25%), more likely to travel to dialysis using a van service (45% vs. 31%) and more likely to have 58 59 60 Journal of the American Society of Nephrology Page 11 of 34 Copyright 2021 by ASN, Published Ahead of Print on 6/3/21, Accepted/Unedited Version.

1 2 diabetes as their cause of kidney failure (51% vs. 39%) (Table 2). ZCTA-level COVID-19 3 4 cumulative incidence in the New York City general population were strongly associated with 5 6 COVID-19 among patients on hemodialysis (odds ratio [OR] 1.34, 95% confidence interval [CI] 7 8 9 1.14 – 1.58 per percentage increase in PCR+ cases divided by the ZCTA population 10 11 denominator) (Supplemental Figure 2). 12 13 Race/Ethnicity and COVID-19 14 15 16 Non-Hispanic Black (OR 1.68, 95% CI 1.14 – 2.48) and Hispanic patients (OR 2.66, 95% 17 18 CI 1.52 – 4.65) had higher odds of acquiring symptomatic COVID-19, compared with non- 19 20 Hispanic white patients in unadjusted analyses (Table 3). Similar relative risks by race/ethnicity 21 22 were seen in the New York City general population (Supplemental Table 2).18 Adjusting for 23 24 25 ZCTA-level COVID-19 cumulative incidence somewhat attenuated racial/ethnic disparities seen 26 27 among patients on dialysis (non-Hispanic Black aOR 1.47, 95% CI 1.03 – 2.09 and Hispanic 28 29 aOR 2.23, 95% CI 1.39 – 3.58). Non-Hispanic Black and Hispanic patients were more likely to 30 31 32 acquire COVID-19 in multivariable analyses accounting for age, sex, SVI, social factors, and 33 34 dialysis-related medical history. In multivariable analyses additionally accounting for dialysis 35 36 facility factors (unit fixed effects and dialysis unit SVI, Supplemental Table 3), non-Hispanic 37 38 39 Black (aOR 1.76, 95% CI 1.25 – 2.48) and Hispanic patients (aOR 2.66, 95% CI 1.50 – 4.75) 40 41 had increased odds of COVID-19, compared with non-Hispanic white patients. Patients who 42 43 were Asian or Pacific Islander were not at increased risk of incident COVID-19, compared with 44 45 non-Hispanic white patients, in unadjusted or adjusted models. Racial/ethnic disparities were 46 47 48 largely consistent across time periods during the first wave of the pandemic, and in time-to-event 49 50 analyses accounting for the competing risk of death (Supplemental Tables 4 and 5). 51 52 Neighborhood Social Vulnerability and COVID-19 53 54 55 The interaction terms of race/ethnicity and SVI themes were statistically significantly 56 57 associated with COVID-19 (Supplemental Table 6). Because of statistical evidence of an 58 59 60 Journal of the American Society of Nephrology Copyright 2021 by ASN, Published Ahead of Print on 6/3/21, Accepted/Unedited Version. Page 12 of 34

1 2 interaction, we stratified analyses by race/ethnicity when examining the association between 3 4 overall SVI, SVI themes, and housing crowding and COVID-19 (Table 4). Overall SVI, modeled 5 6 either continuously or in quintiles, was not significantly associated with COVID-19 in 7 8 9 unadjusted or adjusted models across racial/ethnic categories. Among non-Hispanic white 10 11 patients, the socioeconomic status SVI theme (OR 1.11, 95% CI 1.02 – 1.20 per 10 percentile 12 13 increase), minority status & language SVI theme (OR 1.22, 95% CI 1.07 – 1.38), and housing 14 15 16 crowding (OR 1.24, 95% CI 1.11-1.37) were significantly associated with COVID-19 in 17 18 unadjusted analyses. In analyses adjusted for demographics, social factors, dialysis-related 19 20 medical history, and dialysis facility factors, overall SVI or SVI themes were not associated with 21 22 COVID-19 across racial/ethnic categories, but housing crowding remained significantly 23 24 25 associated with COVID-19 among non-Hispanic white patients (aOR 1.14, 95% CI 1.03 – 1.26) 26 27 (Supplemental Table 7). 28 29 Discussion 30 31 32 In this cohort study of patients on in-center hemodialysis, we found that non-Hispanic 33 34 Black and Hispanic individuals were substantially more likely to acquire symptomatic COVID- 35 36 19 during the first wave of the pandemic in New York City. Census tract-level socioeconomic 37 38 39 status, minority status & language, and housing crowding were associated with COVID-19 40 41 among non-Hispanic white patients on hemodialysis, but did not explain racial/ethnic disparities. 42 43 Neighborhood-level COVID-19 prevalence was associated with COVID-19 cases among 44 45 patients on hemodialysis, suggesting community transmission. 46 47 48 Our results contribute to a growing literature documenting racial/ethnic disparities in 49 50 COVID-19 across different US geographies and clinical populations. Our results are significant 51 52 because patients on hemodialysis are a highly vulnerable patient population that face unique, 53 54 6,8 55 excess risks of acquiring COVID-19 and high mortality rates. Thus, stark racial/ethnic 56 57 disparities in kidney failure incidence21 and COVID-19 incidence represent a situation of 58 59 60 Journal of the American Society of Nephrology Page 13 of 34 Copyright 2021 by ASN, Published Ahead of Print on 6/3/21, Accepted/Unedited Version.

1 2 multiplicative risk, placing non-Hispanic Black and Hispanic individuals on hemodialysis at high 3 4 risk of acquiring COVID-19 and subsequent adverse outcomes including mortality. Another 5 6 analysis of 2178 patients on dialysis in New York City also showed a high prevalence of 7 8 9 COVID-19 among dialysis patients and higher incidence in Black, Hispanic, and Asian patients, 10 11 contributing to a greater population burden of mortality in these racial/ethnic groups.22 Structural 12 13 and historical racism, such as redlining (race/ethnicity-based discriminatory practices, including 14 15 16 discriminatory lending practices, that place financial and other services out of reach for residents 17 18 of certain areas), result in neighborhood segregation and lack of economic opportunity and likely 19 20 contribute to racial/ethnic disparities in COVID-19 incidence.23-25 21 22 In our study, racial/ethnic disparities in COVID-19 incidence in the hemodialysis 23 24 25 population mirrored trends seen in the general population of New York City, but were only 26 27 partially explained by neighborhood COVID-19 cumulative incidence and not explained by 28 29 census indicators of social vulnerability, suggesting that additional unmeasured social variables 30 31 32 contribute to excess risk. Our findings that race/ethnicity modifies the association between 33 34 neighborhood social vulnerability and COVID-19 suggests that neighborhood-level factors 35 36 contributed to COVID-19 incidence among non-Hispanic white patients, whereas other residual 37 38 39 factors such as unmeasured household exposures accounted for excess COVID-19 cases in the 40 41 Black and Hispanic patients in our study sample. For example, in the general population, Black 42 43 individuals are more likely to live in households with healthcare workers, and Hispanic persons 44 45 are more likely to live in households with essential workers unable to work from home.26 46 47 48 Similarly, patients on hemodialysis who are Black or Hispanic may be more likely to live with 49 50 essential workers unable to work from home, increasing their risk of acquiring COVID-19 from 51 52 household contacts. Limited English proficiency and immigration status are additional 53 54 27 55 unmeasured social factors that contribute to excess risk of acquiring COVID-19. A greater 56 57 understanding of patients’ household composition and other community exposures may be useful 58 59 60 Journal of the American Society of Nephrology Copyright 2021 by ASN, Published Ahead of Print on 6/3/21, Accepted/Unedited Version. Page 14 of 34

1 2 for risk-stratifying patients and providing individualized education about risk mitigation. 3 4 Additionally, exposure to dialysis facility staff and other patients who may be traveling from 5 6 COVID-19 prevalent areas may contribute to racial/ethnic disparities. 7 8 9 Our findings provide evidence that both neighborhood and dialysis-related exposure 10 11 contribute to COVID-19 risk among patients on hemodialysis. Formal guidance has been 12 13 disseminated to dialysis facilities on infection control practices,28,29 but less attention has been 14 15 16 provided to patient-oriented education within dialysis facilities to decrease risk of community 17 18 acquisition. As the COVID-19 pandemic continues, qualitative and survey-based research on the 19 20 perceptions and practices of dialysis patients and their household contacts surrounding social 21 22 distancing, mask wearing, and vaccination will be highly informative. 23 24 25 The lack of association between overall SVI and COVID-19 in our data differs from 26 27 findings from ecological studies in the general population, which show a strong association 28 29 between county-level SVI and positive tests per capita.5,15,16 There are several explanations for 30 31 32 this discrepancy. First, patients on hemodialysis in our cohort disproportionately lived in high 33 34 social vulnerability neighborhoods and face excess risk from dialysis-related exposures, so the 35 36 same relationship between overall SVI and COVID-19 may not apply. Second, certain variables 37 38 39 in the SVI, such as multiunit housing, mobile homes, vehicle ownership, may perform differently 40 41 or not be relevant in New York City in relation to COVID-19, so similar analyses should be 42 43 investigated in other geographies. The magnitude of our findings and interactions between 44 45 race/ethnicity and neighborhood-level social vulnerability may not fully generalize to other 46 47 48 geographies, depending on urbanicity, population demographics, and timing/patterns of the 49 50 COVID-19 pandemic. 51 52 Strengths of our analysis include the substantial size of our cohort and use of patient-level 53 54 55 address data that allowed for granular geocoding. Limitations include missing data for some 56 57 variables, particularly individual-level social factors, which may contribute to misclassification 58 59 60 Journal of the American Society of Nephrology Page 15 of 34 Copyright 2021 by ASN, Published Ahead of Print on 6/3/21, Accepted/Unedited Version.

1 2 bias. Misclassification of race/ethnicity data may have occurred if race/ethnicity was entered by 3 4 dialysis facility staff by presumption and not based on patient self-report. Our results may also be 5 6 subject to misclassification of neighborhood-level social vulnerability if patients moved or there 7 8 9 were lags in updating addresses in the EHR. Importantly, we did not systematically test for 10 11 SARS-CoV-2 using PCR or antibody testing, so our results do not incorporate asymptomatic 12 13 cases, which may be considerable among patients on hemodialysis.30 Lastly, our results may 14 15 16 have been subject to residual confounding and/or were underpowered to detect some associations 17 18 between neighborhood-level characteristics and COVID-19, particularly in stratified analyses. 19 20 In summary, racial/ethnic disparities in COVID-19 incidence among patients on 21 22 hemodialysis largely mirror community transmission patterns, and likely reflect neighborhood 23 24 25 spread to this vulnerable population. Neighborhood-level socioeconomic status, minority status 26 27 & language, and housing crowding were positively associated with acquiring COVID-19 among 28 29 non-Hispanic white patients, but did not explain racial/ethnic disparities. Our findings suggest 30 31 32 that socially vulnerable patients on dialysis face disparate COVID-19-related exposures, calling 33 34 for targeted risk mitigation strategies. 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Journal of the American Society of Nephrology Copyright 2021 by ASN, Published Ahead of Print on 6/3/21, Accepted/Unedited Version. Page 16 of 34

1 2 Author Contributions: SLT, JS, and SAI designed the study; TB, YL, KK, TSP, and DML 3 4 constructed the database and performed data collection; SLT performed statistical analysis; SLT, 5 6 JS, DC, JTL, TSP, DML, and SAI interpreted the results; SLT drafted the paper; JS and SAI 7 8 9 provided supervision; all authors revised the manuscript for important intellectual content and 10 11 approved the final version of the manuscript. 12 13 14 15 16 Acknowledgements: The authors acknowledge Beth Beltran, Marci Rosner, Bjorn Brogle, 17 18 Hilary Marion, Michael Roldan, John Lopez, Mozelle Lafleur, Debora Lidov, Robin Grande, 19 20 Susan Katz, Joshua Zimmeman, Michelle Grant Tate, Ronald Wilson, Natasha Miller, Beth 21 22 Epstein, Diane Morris, Jeanene Bennett-Nazario, Jason Emralino, Rohonie Persaud, Marilyn 23 24 25 Sure, Cathy Reydel, Allen Herman, and Betty-Jane Sloan for their outstanding patient care and 26 27 assistance with data collection. 28 29 30 31 32 Funding: Dr. Tummalapalli is supported by funding from the National Institute of Diabetes and 33 34 Digestive and Kidney Diseases F32DK122627 and the National Kidney Foundation Young 35 36 Investigator Grant. 37 38 39 40 41 Disclosures: Dr. Tummalapalli received consulting fees from Bayer AG unrelated to the 42 43 submitted work. The remaining authors have nothing to disclose. D. Cukor reports Research 44 45 Funding from NIH. D. Levine reports Patents and Inventions with The Rogosin Institute. J. 46 47 48 Silberzweig reports Consultancy Agreements with Kaneka Pharma, Bayer Pharmaceuticals, 49 50 Alkahest Biotech; Scientific Advisor or Membership with American Society of Nephrology: 51 52 COVID-19 Response team, Emergency Partnership Initiative. 53 54 55 56 57 58 59 60 Journal of the American Society of Nephrology Page 17 of 34 Copyright 2021 by ASN, Published Ahead of Print on 6/3/21, Accepted/Unedited Version.

1 2 Figure 1. Census Tract-level Social Vulnerability Index (SVI) percentile among patients on 3 hemodialysis. 4 5 Higher SVI percentile indicates greater neighborhood social vulnerability. Patients on hemodialysis in our 6 cohort did not reside in census tracts pictured in grey. 7 8 9 10 Supplemental Material Table of Contents: 11 12 Supplemental Table 1. Characteristics by race/ethnicity among patients on hemodialysis (n = 13 14 1,378). 15 16 17 Supplemental Figure 1. Timing of COVID-19 cases among patients on hemodialysis and across 18 19 New York City ZIP Code Tabulation Areas. 20 21 Supplemental Figure 2. COVID-19 cumulative cases by PCR testing in the New York City 22 23 general population. 24 25 26 Supplemental Table 2. COVID-19 cumulative cases by PCR testing in the New York City 27 28 general population by race/ethnicity. 29 30 Supplemental Table 3. Dialysis unit Social Vulnerability Index and cumulative COVID-19 31 32 33 cases. 34 35 Supplemental Table 4. Association of Race/Ethnicity and Social Vulnerability Index with 36 37 COVID-19 among patients on hemodialysis, stratified by time period. 38 39 40 Supplemental Table 5. Association of Race/Ethnicity and Social Vulnerability Index with 41 42 COVID-19 among patients on hemodialysis (n = 1378), in a time-to-event analysis accounting 43 44 for the competing risk of death. 45 46 Supplemental Table 6. Association of Race/Ethnicity with COVID-19 among patients on 47 48 49 hemodialysis (n = 1378), accounting for Race/Ethnicity and SVI interactions. 50 51 Supplemental Table 7. Association of Social Vulnerability Index with COVID-19 among 52 53 patients on hemodialysis (n = 1378), stratified by race/ethnicity (adjusted analyses). 54 55 56 57 58 59 60 Journal of the American Society of Nephrology Copyright 2021 by ASN, Published Ahead of Print on 6/3/21, Accepted/Unedited Version. Page 18 of 34

1 2 3 Table 1. Characteristics of patients on in-center hemodialysis by overall Social Vulnerability Index 4 (n = 1,378).* 5 Low SVI* High SVI* (%), n = 6 (%), n = 689 689 p-value 7 8 Demographics 9 Age 10 18-44 7 14 <0.001 11 45-64 37 42 12 65-79 39 34 13 ≥80 17 11 14 15 Sex 16 Male 61 54 0.012 17 Female 39 46 18 Individual-level Social Factors 19 Race/Ethnicity 20 21 Non-Hispanic white 32 11 <0.001 22 Non-Hispanic Black 33 51 23 Hispanic 13 17 24 Asian or Pacific 15 10 25 Islander 26 Other, unknown, or 7 11 27 missing 28 Employment 29 Full-time or part-time 16 8 <0.001 30 31 Retired (age) 21 19 32 Retired (disabled) 19 21 33 Unemployed 8 11 34 Homemaker, medical 4 4 35 leave, or student 36 Missing 32 37 37 Marital Status 38 39 Married 39 27 <0.001 40 Divorced or separated 9 9 41 Widowed 7 7 42 Single 25 33 43 Missing 20 24 44 Transportation type 45 Van service 33 35 0.757 46 47 Private vehicle 20 21 48 Public transportation 2 2 Other, unknown, or 49 45 42 50 missing 51 Dialysis-related Medical History 52 53 Dialysis vintage (years)^ 3.0 [1.2 – 5.9] 3.6 [1.7 – 6.4] 0.0095 54 Primary kidney failure cause 55 Diabetes 38 44 0.023 56 Hypertension 31 26 57 Glomerulonephritis 11 11 58 59 60 Journal of the American Society of Nephrology Page 19 of 34 Copyright 2021 by ASN, Published Ahead of Print on 6/3/21, Accepted/Unedited Version.

1 2 Cystic kidney disease 3 2 3 HIV 1 3 4 Malignancy 2 0.4 5 Post-transplant 5 4 6 Other or unknown 9 10 7 8 9 *High (above median) Social Vulnerability Index (SVI) indicates greater neighborhood social 10 vulnerability. Low (below median) SVI indicates lower neighborhood social vulnerability. ^ 11 Dialysis vintage presented as median [interquartile range] and reported among patients classified as 12 ESRD as of March 1, 2020 (n = 1362). 13 Data are from noninstitutionalized adults residing in the New York City area receiving hemodialysis in 14 one of eight dialysis units in Manhattan, Brooklyn, and Queens. 15 Percentages may not add to 100% due to rounding. P-values presented for chi-squared tests for 16 categorical variables and Wilcoxon rank-sum tests for continuous variables. 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Journal of the American Society of Nephrology Copyright 2021 by ASN, Published Ahead of Print on 6/3/21, Accepted/Unedited Version. Page 20 of 34

1 2 3 Table 2. Patient characteristics by COVID positivity among patients on hemodialysis (n = 1,378). 4 COVID-19-Positive Not COVID-19-Positive 5 (%), n = 247 (%), n = 1,131 p-value 6 7 Demographics 8 Age 9 18-44 8 11 0.205 10 45-64 38 40 11 65-79 42 35 12 ≥80 13 14 13 14 Sex 15 Male 52 58 0.073 16 Female 48 42 17 Individual-level Social Factors 18 Employment 19 Full-time or part-time 10 13 0.122 20 21 Retired (age) 23 19 22 Retired (disabled) 25 19 23 Unemployed 9 10 24 Homemaker, medical 3 4 25 leave, or student 26 Missing 30 36 27 Marital Status 28 Married 33 33 0.002 29 30 Divorced or separated 12 8 31 Widowed 12 6 32 Single 25 30 33 Missing 19 23 34 Transportation type 35 Van service 45 31 0.001 36 Private vehicle 18 21 37 38 Public transportation 2 2 Other, unknown, or 39 36 45 40 missing 41 Dialysis-related Medical History 42 Dialysis vintage (years)^ 3.5 [1.5 – 6.0] 3.2 [1.4 – 6.1] 0.9833 43 44 Primary kidney failure cause 45 Diabetes 51 39 0.046 46 Hypertension 25 29 47 Glomerulonephritis 10 11 48 Cystic kidney disease 2 2 49 HIV 2 2 50 Malignancy 1 1 51 52 Post-transplant 2 5 53 Other or unknown 7 10 54 ^Dialysis vintage presented as median [interquartile range] and reported among patients classified as 55 ESRD as of March 1, 2020 (n = 1362). 56 Percentages may not add to 100% due to rounding. P-values presented for chi-squared tests for 57 categorical variables and Wilcoxon rank-sum tests for continuous variables. 58 59 60 Journal of the American Society of Nephrology Page 21 of 34 Copyright 2021 by ASN, Published Ahead of Print on 6/3/21, Accepted/Unedited Version.

Table 3. Association of Race/Ethnicity and Social Vulnerability Index with COVID-19 among patients on 1 hemodialysis (n = 1378). 2 Unadjusted Model 1 Model 2 3 OR aOR aOR 4 5 Variable (95% CI) p-values (95% CI) p-values (95% CI) p-values 6 Race/Ethnicity Non-Hispanic white (ref.) 7 ------8 (n = 294) 1 1 1 9 Non-Hispanic Black 1.68 1.77 1.76 0.008 0.004 0.001 10 (n = 578) (1.14 – 2.48) (1.20 – 2.62) (1.25 – 2.48) 11 Hispanic 2.66 2.90 2.66 12 0.001 <0.001 0.001 (n = 207) (1.52 – 4.65) (1.72 – 4.88) (1.50 – 4.75) 13 Asian or Pacific Islander 1.32 1.19 1.22 14 0.401 0.658 0.555 15 (n = 174) (0.69 – 2.50) (0.56 – 2.53) (0.63 – 2.36) 16 Other, unknown, or 1.21 1.30 1.37 0.495 0.194 0.106 17 missing (n = 125) (0.70 – 2.06) (0.87 – 1.94) (0.94 – 2.02) 18 Overall Social Vulnerability Index 19 Quintile 1 (ref.) 1 0.499 1 0.920 1 0.801 20 1.34 1.06 1.01 21 Quintile 2 22 (0.80 – 2.23) (0.67 – 1.66) (0.60 – 1.68) 1.32 0.99 1.00 23 Quintile 3 24 (0.86 – 2.02) (0.75 – 1.32) (0.73 – 1.38) 25 1.55 1.16 1.19 Quintile 4 26 (0.96 – 2.53) (0.85 – 1.58) (0.88 – 1.60) 27 1.33 1.05 1.11 Quintile 5 28 (0.76 – 2.31) (0.63 – 1.75) (0.66 – 1.87) 29 30 SVI calculated at the census tract-level. Quintile 1 of SVI represents the lowest level of social vulnerability. Quintile 5 of 31 SVI represents the highest level of social vulnerability. 32 Model 1 adjusted for age, sex, race or SVI, individual-level social factors (employment, marital status, transportation type 33 to dialysis), and dialysis-related medical history (dialysis vintage, primary kidney failure cause). Model 2 additionally 34 adjusted for dialysis facility factors (unit fixed effects and dialysis unit SVI). 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Journal of the American Society of Nephrology Copyright 2021 by ASN, Published Ahead of Print on 6/3/21, Accepted/Unedited Version. Page 22 of 34

Table 4. Association of Social Vulnerability Index with COVID-19 among patients on hemodialysis (n = 1378), stratified by race/ethnicity (unadjusted 1 analyses). 2 Other, 3 Non-Hispanic Non-Hispanic Asian or Pacific Unknown, or 4 5 white Black Hispanic Islander Missing 6 (n = 294) (n = 578) (n = 207) (n = 174) (n = 125) 7 Variable OR (95% CI) p-value OR (95% CI) p-value OR (95% CI) p-value OR (95% CI) p-value OR (95% CI) p-value 8 Overall Social Vulnerability Index† 9 1.08 0.99 1.03 1.12 0.95 SVI 0.201 0.788 0.834 0.186 0.580 10 (0.96 – 1.21) (0.92 – 1.07) (0.81 – 1.30) (0.94 – 1.33) (0.79 – 1.14) 11 Social Vulnerability Index Theme† 12 1.11 0.95 1.06 1.07 0.91 Socioeconomic status 0.013 0.206 0.604 0.275 0.215 13 (1.02 – 1.20) (0.88 – 1.03) (0.86 – 1.30) (0.94 – 1.22) (0.79 – 1.05) 14 Household composition 0.90 1.02 0.94 1.09 0.96 15 0.233 0.580 0.399 0.265 0.599 16 & disability (0.76 – 1.07) (0.96 – 1.08) (0.82 – 1.08) (0.94 – 1.28) (0.83 – 1.11) Minority status & 1.22 0.97 1.11 0.99 1.02 17 0.002 0.523 0.214 0.887 0.928 18 language (1.07 – 1.38) (0.88 – 1.07) (0.94 – 1.31) (0.87 – 1.13) (0.69 – 1.51) 19 Housing type & 1.05 1.01 1.04 1.11 1.07 0.340 0.766 0.594 0.032 0.275 20 transportation (0.95 – 1.16) (0.93 – 1.10) (0.90 – 1.21) (1.01 – 1.22) (0.95 – 1.21) 21 Social Vulnerability Index Component† 22 1.24 0.98 0.96 1.07 1.08 23 Housing Crowding* <0.001 0.719 0.672 0.382 0.690 24 (1.11 – 1.37) (0.89 – 1.08) (0.78 – 1.17) (0.92 – 1.24) (0.74 – 1.57) 25 †Presented as unadjusted odds ratios of COVID-19 positivity per 10 percentile increase in overall SVI, SVI theme, or SVI component. 26 Overall SVI, SVI themes, and SVI components calculated at the census tract-level. 27 Housing crowding percentile defined as the percentage of occupied housing units in a census tract with more people than rooms. 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Journal of the American Society of Nephrology 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 23 of 34 Copyright 2021 by ASN, Published Ahead of Print on 6/3/21, Accepted/Unedited Version.

1 2 3 References: 4 1. Oppel R, Gebelhoff R, Lai K, Wright W, Smith M. The fullest look yet at the racial inequity of 5 coronavirus. people 2020;25:50. 6 7 2. Tirupathi R, Muradova V, Shekhar R, Salim SA, Al-Tawfiq JA, Palabindala V. COVID-19 8 disparity among racial and ethnic minorities in the US: A cross sectional analysis. Travel Med Infect Dis 9 2020;38:101904. 10 3. Rentsch CT, Kidwai-Khan F, Tate JP, et al. Covid-19 by Race and Ethnicity: A National Cohort 11 Study of 6 Million United States Veterans. medRxiv 2020. 12 4. Diamond LC, Jacobs EA, Karliner L. Providing equitable care to patients with limited dominant 13 language proficiency amid the COVID-19 pandemic. Patient education and counseling 2020;103:1451. 14 5. Khazanchi R, Beiter ER, Gondi S, Beckman AL, Bilinski A, Ganguli I. County-Level 15 Association of Social Vulnerability with COVID-19 Cases and Deaths in the USA. Journal of General 16 Internal Medicine 2020:1-4. 17 6. Services. USDoHaH. Advancing American Kidney Health: 2020 Progress Report. 18 https://aspe.hhs.gov/pdf-report/advancing-american-kidney-health-2020-progress-report. 19 7. Valeri AM, Robbins-Juarez SY, Stevens JS, et al. Presentation and Outcomes of Patients with 20 ESKD and COVID-19. J Am Soc Nephrol 2020. 21 8. Ng JH, Hirsch JS, Wanchoo R, et al. Outcomes of patients with end-stage kidney disease 22 hospitalized with COVID-19. Kidney International 2020. 23 9. Naaraayan A, Nimkar A, Hasan A, et al. End-Stage Renal Disease Patients on Chronic 24 Hemodialysis Fare Better With COVID-19: A Retrospective Cohort Study From the New York 25 Metropolitan Region. Cureus 2020;12. 26 27 10. Chan L, Jaladanki SK, Somani S, et al. Outcomes of Patients on Maintenance Dialysis 28 Hospitalized with COVID-19. Clinical Journal of the American Society of Nephrology 2020. 29 11. De Meester J, De Bacquer D, Naesens M, Meijers B, Couttenye MM, De Vriese AS. Incidence, 30 Characteristics, and Outcome of COVID-19 in Adults on Kidney Replacement Therapy: A Regionwide 31 Registry Study. J Am Soc Nephrol 2020. 32 12. Anand S, Montez-Rath M, Han J, et al. Prevalence of SARS-CoV-2 antibodies in a large 33 nationwide sample of patients on dialysis in the USA: a cross-sectional study. The Lancet 2020;396:1335- 34 44. 35 13. KC M, Oral E, Straif-Bourgeois S, Rung AL, Peters ES. The Effect of Area Deprivation on 36 COVID-19 Risk in Louisiana. medRxiv 2020:2020.08.24.20180893. 37 14. Carrion D, Colicino E, Pedretti NF, et al. Assessing capacity to social distance and neighborhood- 38 level health disparities during the COVID-19 pandemic. medRxiv 2020. 39 15. Nayak A, Islam SJ, Mehta A, et al. Impact of Social Vulnerability on COVID-19 Incidence and 40 Outcomes in the United States. medRxiv 2020. 41 16. Karaye IM, Horney JA. The Impact of Social Vulnerability on COVID-19 in the US: An Analysis 42 of Spatially Varying Relationships. American Journal of Preventive Medicine 2020. 43 17. Dasgupta S, Bowen VB, Leidner A, et al. Association Between Social Vulnerability and a 44 County’s Risk for Becoming a COVID-19 Hotspot—United States, June 1–July 25, 2020. Morbidity and 45 46 Mortality Weekly Report 2020;69:1535. 47 18. Health. N. COVID-19: Data. https://www1.nyc.gov/site/doh/covid/covid-19-data-testing.page. 48 19. State. CSVIDN. 49 https://www.atsdr.cdc.gov/placeandhealth/svi/data_documentation_download.html. Accessed on August 50 15, 2020. 51 20. Bureau. USC. https://geocoding.geo.census.gov/geocoder/geographies/addressbatch?form 52 Accessed August 15, 2020. 53 21. United States Renal Data System. 2019 USRDS annual data report: Epidemiology of kidney 54 disease in the United States. National Institutes of Health NIDDK. 55 22. Weiss S, Bhat P, del Pilar Fernandez M, Bhat JG, Coritsidis GN. COVID-19 Infection in ESKD: 56 Findings from a Prospective Disease Surveillance Program at Dialysis Facilities in New York City and 57 Long Island. Journal of the American Society of Nephrology 2020. 58 59 60 Journal of the American Society of Nephrology Copyright 2021 by ASN, Published Ahead of Print on 6/3/21, Accepted/Unedited Version. Page 24 of 34

1 2 3 23. Millett GA, Jones AT, Benkeser D, et al. Assessing differential impacts of COVID-19 on Black 4 communities. Annals of Epidemiology 2020. 5 24. Aaronson D, Hartley DA, Mazumder B. The Effects of the 1930s HOLC'Redlining'Maps. 2020. 6 25. Choi Y, Unwin J. Racial Impact on Infections and Deaths due to COVID-19 in New York City. 7 arXiv preprint arXiv:200704743 2020. 8 26. Selden TM, Berdahl TA. COVID-19 And Racial/Ethnic Disparities In Health Risk, Employment, 9 10 And Household Composition: Study examines potential explanations for racial-ethnic disparities in 11 COVID-19 hospitalizations and mortality. Health Affairs 2020;39:1624-32. 12 27. Chamie G, Marquez C, Crawford E, et al. SARS-CoV-2 Community Transmission 13 disproportionately affects Latinx population during Shelter-in-Place in San Francisco. Clin Infect Dis 14 2020. 15 28. Corbett RW, Blakey S, Nitsch D, et al. Epidemiology of COVID-19 in an urban dialysis center. 16 Journal of the American Society of Nephrology 2020;31:1815-23. 17 29. Watnick S, McNamara E. On the frontline of the COVID-19 outbreak: keeping patients on long- 18 term dialysis safe. Clinical Journal of the American Society of Nephrology 2020;15:710-3. 19 30. Clarke C, Prendecki M, Dhutia A, et al. High Prevalence of Asymptomatic COVID-19 Infection 20 in Hemodialysis Patients Detected Using Serologic Screening. Journal of the American Society of 21 Nephrology 2020;31:1969-75. 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Journal of the American Society of Nephrology Page 25 of 34 Copyright 2021 by ASN, Published Ahead of Print on 6/3/21, Accepted/Unedited Version.

1 2 3 Figure 1. Census Tract-level Social Vulnerability Index (SVI) percentile among patients on 4 hemodialysis. 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 Higher SVI percentile indicates greater neighborhood social vulnerability. Patients on hemodialysis in our 28 cohort did not reside in census tracts pictured in grey. 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Journal of the American Society of Nephrology Page 26 of 34

1 2 3 Supplemental Material Table of Contents: 4 5 6 Supplemental Table 1. Characteristics by race/ethnicity among patients on hemodialysis (n = 7 8 1,378). 9 10 Supplemental Figure 1. Timing of COVID-19 cases among patients on hemodialysis and across 11 12 13 New York City ZIP Code Tabulation Areas. 14 15 Supplemental Figure 2. COVID-19 cumulative cases by PCR testing in the New York City 16 17 general population. 18 19 Supplemental Table 2. COVID-19 cumulative cases by PCR testing in the New York City 20 21 22 general population by race/ethnicity. 23 24 Supplemental Table 3. Dialysis unit Social Vulnerability Index and cumulative COVID-19 25 26 cases. 27 28 29 Supplemental Table 4. Association of Race/Ethnicity and Social Vulnerability Index with 30 31 COVID-19 among patients on hemodialysis, stratified by time period. 32 33 Supplemental Table 5. Association of Race/Ethnicity and Social Vulnerability Index with 34 35 36 COVID-19 among patients on hemodialysis (n = 1378), in a time-to-event analysis accounting 37 38 for the competing risk of death. 39 40 Supplemental Table 6. Association of Race/Ethnicity with COVID-19 among patients on 41 42 hemodialysis (n = 1378), accounting for Race/Ethnicity and SVI interactions. 43 44 45 Supplemental Table 7. Association of Social Vulnerability Index with COVID-19 among 46 47 patients on hemodialysis (n = 1378), stratified by race/ethnicity (adjusted analyses). 48 49 50 51 52 53 54 55 56 57 58 59 60 Journal of the American Society of Nephrology Page 27 of 34

1 2 3 Supplemental Table 1. Characteristics by race/ethnicity among patients on hemodialysis (n = 1,378). 4 Non-Hispanic Non-Hispanic Hispanic Asian or Pacific Other, Unknown, 5 White Black (%), n = 207 Islander or Missing 6 (%), n = 294 (%), n = 578 (%), n = 174 (%), n = 125 p-value 7 8 Demographics 9 Age 10 18-44 5 13 16 4 7 <0.001 11 45-64 23 47 45 33 39 12 13 65-79 41 31 34 49 38 14 ≥80 31 9 4 14 15 15 Sex 16 Male 65 51 60 61 55 0.001 17 18 Female 35 49 40 39 45 19 Individual-level Social Factors 20 Employment 21 Full-time or part- 22 14 13 11 13 7 <0.001 time 23 24 Retired (age) 29 15 13 22 27 25 Retired (disabled) 15 24 21 16 15 26 Unemployed 6 9 14 15 5 27 Homemaker, 28 medical leave, or 4 3 4 5 4 29 student 30 31 Missing 31 35 36 30 42 32 Marital Status 33 Married 47 23 38 56 6 <0.001 34 Divorced or 35 9 10 13 7 2 separated 36 37 Widowed 9 7 3 9 5 38 Single 21 39 32 14 15 39 Missing 15 20 14 14 71 40 Transportation type 41 42 43 44 45 Journal of the American Society of Nephrology 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 28 of 34

1 2 3 Van service 33 35 32 48 14 <0.001 4 5 Private vehicle 16 23 25 25 6 Public 6 2 2 5 2 2 7 transportation 8 Other, unknown, 50 40 38 25 79 9 or missing 10 Dialysis-related Medical History 11 Dialysis vintage 12 2.3 [1.0 – 4.9] 4.0 [1.8 – 7.1] 3.7 [1.7 – 5.4] 3.3 [1.4 – 6.0] 2.4 [1.3 – 5.4] <0.001 13 (years)^ 14 Primary kidney failure cause 15 Diabetes 36 39 43 49 52 <0.001 16 Hypertension 27 33 27 22 22 17 18 Glomerulonephritis 10 12 10 11 9 19 Cystic kidney 4 2 1 2 1 20 disease 21 HIV 0 3 0 0 5 22 Malignancy 3 1 0.5 0 0 23 24 Post-transplant 4 3 7 8 2 25 Other or unknown 15 7 11 6 10 26 ^Dialysis vintage presented as median [interquartile range] and reported among patients classified as ESRD as of March 1, 2020 (n = 1362). 27 Percentages may not add to 100% due to rounding. P-values presented for chi-squared tests for categorical variables and Wilcoxon rank-sum tests 28 for continuous variables. 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Journal of the American Society of Nephrology 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 29 of 34

1 2 3 Supplemental Figure 1. Timing of COVID-19 cases among patients on hemodialysis and across New 4 York City ZIP Code Tabulation Areas. 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 Figure depicts 7-day moving averages of PCR-positive COVID-19 cases. 32 33 34 35 36 Supplemental Figure 2. COVID-19 cumulative cases by PCR testing in the New York City general 37 population. 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 Data presented are the percentage PCR+ cases divided by the population denominator at the modified ZIP 53 54 Code Tabulation Area (ZCTA)-level ascertained until August 25, 2020. 55 56 57 58 59 60 Journal of the American Society of Nephrology Page 30 of 34

1 2 3 Supplemental Table 2. COVID-19 cumulative cases by PCR testing in the New York City general 4 population by race/ethnicity.a 5 COVID-19 Cumulative Cases Relative COVID-19 6 Race/Ethnicity 7 per 100,000 individuals Cumulative Cases 8 White 1111.24 1 (ref.) 9 Black/African- 1763.03 1.59 10 American 11 Hispanic/Latino 1825.3 1.64 12 Asian/Pacific- 13 726.05 0.65 14 Islander 15 COVID-19 cumulative cases adjusted for age at diagnosis and ascertained until August 25, 2020. 16 aAccessed August 25, 2020. https://www1.nyc.gov/site/doh/covid/covid-19-data-testing.page 17 18 19

20 21 22 Supplemental Table 3. Dialysis unit Social Vulnerability Index and cumulative COVID-19 cases. 23 Cumulative Dialysis Unit Dialysis Unit Dialysis Unit SVI (%tile) 24 COVID-19 Cases (%) 25 A 49.3 17 26 B 92.2 15 27 C 19.5 34 28 D 98.0 28 29 30 E 80.3 22 31 F 40.9 14 32 G 71.4 16 33 H 67.0 34 34 COVID-19 cases ascertained until August 3, 2020. 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Journal of the American Society of Nephrology Page 31 of 34

Supplemental Table 4. Association of Race/Ethnicity and Social Vulnerability Index with COVID-19 among patients on hemodialysis, stratified by time 1 period. 2 Time Period 1 Time Period 2 Time Period 3 3 OR (95% CI) OR (95% CI) OR (95% CI) 4 5 Variable (n at-risk = 1378) p-values (n at-risk = 1364) p-values (n at-risk = 1195) p-values 6 Race/Ethnicity 7 Non-Hispanic white (ref.) 1 -- 1 -- 1 -- 8 1.27 1.66 1.65 Non-Hispanic Black 0.718 0.115 0.228 9 (0.34 – 4.75) (0.88 – 3.10) (0.73 – 3.70) 10 2.15 2.64 2.06 Hispanic 0.453 0.007 0.099 11 (0.29 – 15.8) (1.31 – 5.33) (0.87 – 4.84) 12 2.56 1.48 0.78 13 Asian or Pacific Islander 0.176 0.158 0.738 14 (0.66 – 10.0) (0.86 – 2.55) (0.19 – 3.29) Other, unknown, or 1.18 1.20 0.79 15 0.880 0.610 0.631 16 missing (0.14 – 9.86) (0.60 – 2.38) (0.31 – 2.05) 17 Overall Social Vulnerability Index 18 Quintile 1 (ref.) 1 0.785 1 0.336 1 0.829 19 0.97 1.39 1.01 20 Quintile 2 21 (0.18 – 5.29) (0.87 – 2.21) (0.44 – 2.32) 1.99 1.37 0.82 22 Quintile 3 23 (0.58 – 6.82) (0.99 – 1.90) (0.31 – 2.16) 24 1.99 1.33 1.57 Quintile 4 25 (0.24 – 16.2) (0.66 – 2.68) (0.44 – 5.58) 26 0.99 1.28 1.12 Quintile 5 27 (0.13 – 7.58) (0.70 – 2.33) (0.44 – 2.89) 28 29 Results presented are from unadjusted logistic regression models. SVI calculated at the census tract-level. Quintile 1 of SVI represents the lowest level of social 30 vulnerability. Quintile 5 of SVI represents the highest level of social vulnerability. 31 Time periods analyzed were March 1 – March 21 (time period 1, pre-stay-at-home order), March 22 – April 14 (time period 2, pre-mask order), and April 15 – 32 June 7 (time period 3, before reopening), and June 8 – August 3 (time period 4, reopening). Participants diagnosed with COVID-19 were removed from the at-risk 33 patient population during subsequent time periods. There were 14 COVID-19 cases during time period 1, 171 cases during time period 2, and 55 cases during time 34 period 3, and 5 cases during time period 4. Results from time period 4 not presented due to the small number of cases. 35 36 37 38

39 40 41 42 43 44 45 Journal of the American Society of Nephrology 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 32 of 34

Supplemental Table 5. Association of Race/Ethnicity and Social Vulnerability Index with COVID-19 among patients on hemodialysis (n = 1378), in a 1 time-to-event analysis accounting for the competing risk of death. 2 Unadjusted Model 1 Model 2 3 sHR sHR sHR 4 5 Variable (95% CI) p-values (95% CI) p-values (95% CI) p-values 6 Race/Ethnicity Non-Hispanic white (ref.) 7 ------8 (n = 294) 1 1 1 9 Non-Hispanic Black 1.60 1.65 1.65 0.014 0.018 0.021 10 (n = 578) (1.10 – 2.33) (1.09 – 2.49) (1.08 – 2.52) 11 Hispanic 2.43 2.57 2.34 12 <0.001 <0.001 <0.001 (n = 207) (1.60 – 3.69) (1.64 – 4.03) (1.48 – 3.71) 13 Asian or Pacific Islander 1.31 1.18 1.20 14 0.297 0.544 0.528 15 (n = 174) (0.79 – 2.16) (0.70 – 1.99) (0.68 – 2.13) 16 Other, unknown, or 1.18 1.25 1.33 0.562 0.486 0.417 17 missing (n = 125) (0.67 – 2.08) (0.67 – 2.31) (0.67 – 2.68) 18 Overall Social Vulnerability Index 19 Quintile 1 (ref.) 1 0.475 1 0.962 1 0.953 20 1.30 1.08 1.03 21 Quintile 2 22 (0.86 – 1.98) (0.71 – 1.64) (0.67 – 1.58) 1.29 1.01 1.01 23 Quintile 3 24 (0.85 – 1.97) (0.65 – 1.57) (0.65 – 1.57) 25 1.47 1.15 1.15 Quintile 4 26 (0.98 – 2.21) (0.75 – 1.75) (0.74 – 1.79) 27 1.29 1.06 1.10 Quintile 5 28 (0.85 – 1.95) (0.68 – 1.65) (0.69 – 1.74) 29 30 SVI calculated at the census tract-level. Quintile 1 of SVI represents the lowest level of social vulnerability. Quintile 5 of SVI represents the highest level of social 31 vulnerability. 32 From March 1 to August 3, 2020, 118 patients died in total, 17 patients received transplants, 6 patients withdrew dialysis and/or transferred to hospice, and 8 33 patients switched to PD or HHD. Results are from a Fine and Gray subdistribution hazard model estimating the association of race/ethnicity with the subhazard of 34 COVID-19, accounting for the competing risk of death and censoring at transplant, dialysis withdrawal, and modality change. Separately, a Fine and Gray 35 subdistribution hazard model was used to estimate the association of overall SVI quintile with the subhazard of COVID-19, accounting for the competing risk of 36 death and censoring events. 37 Model 1 adjusted for age, sex, race/ethnicity or SVI, individual-level social factors (employment, marital status, transportation type to dialysis), and dialysis- 38 related medical history (dialysis vintage, primary kidney failure cause). Model 2 additionally adjusted for dialysis facility factors (unit fixed effects and dialysis 39 unit SVI). 40 41 42 43 44 45 Journal of the American Society of Nephrology 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 33 of 34

1 2 3 Supplemental Table 6. Association of Race/Ethnicity with COVID-19 among patients on 4 hemodialysis (n = 1378), accounting for Race/Ethnicity and SVI interactions. 5 Adjusted 6 OR 7 8 Variable (95% CI) p-values 9 Race/Ethnicity and Overall SVI 10 Race/Ethnicity 11 Non-Hispanic white (ref.) (n = 294) 1 -- 12 2.78 Non-Hispanic Black (n = 578) 0.008 13 (1.31 – 5.90) 14 3.38 Hispanic (n = 207) 0.358 15 (0.25 – 45.3) 16 0.95 17 Asian or Pacific Islander (n = 174) 0.964 18 (0.12 – 7.64) 2.74 19 Other, unknown, or missing (n = 125) 0.262 20 (0.47 – 15.97) 21 Race/Ethnicity*Overall SVI† 22 1.08 Non-Hispanic white*Overall SVI 0.201 23 (0.96 – 1.21) 24 0.99 25 Non-Hispanic Black*Overall SVI 0.788 (0.92 – 1.07) 26 1.03 27 Hispanic*Overall SVI 0.834 28 (0.81 – 1.07) 1.12 29 Asian or Pacific Islander*Overall SVI 0.186 30 (0.95 – 1.33) 31 0.95 Other, unknown, or missing*Overall SVI 0.576 32 (0.79 – 1.14) 33 Race/Ethnicity and Socioeconomic Status SVI Theme 34 Race/Ethnicity 35 36 Non-Hispanic white (ref.) 1 -- 3.77 37 Non-Hispanic Black <0.001 38 (1.79 – 7.95) 39 2.91 Hispanic 0.286 40 (0.41 – 20.7) 41 1.38 Asian or Pacific Islander 0.670 42 (0.32 – 5.94) 43 3.79 44 Other, unknown, or missing 0.042 45 (1.05 – 13.7) Race/Ethnicity*Socioeconomic Status SVI 46 47 Theme† 48 1.11 Non-Hispanic white*Socioeconomic Status 0.013 49 (1.02 – 1.20) 50 0.95 51 Non-Hispanic Black*Socioeconomic Status 0.206 (0.88 – 1.03) 52 1.06 53 Hispanic*Socioeconomic Status 0.604 54 (0.86 – 1.30) 55 Asian or Pacific Islander*Socioeconomic 1.07 0.275 56 Status (0.95 – 1.22) 57 58 59 60 Journal of the American Society of Nephrology Page 34 of 34

1 2 3 Other, unknown, or missing*Socioeconomic 0.91 0.210 4 Status (0.79 – 1.05) 5 6 Race/Ethnicity and Household Composition & Disability SVI Theme 7 Race/Ethnicity 8 Non-Hispanic white (ref.) 1 -- 9 1.05 Non-Hispanic Black 0.900 10 (0.47 – 2.38) 11 2.39 Hispanic 0.290 12 (0.48 – 12.0) 13 0.69 14 Asian or Pacific Islander 0.608 15 (0.17 – 2.86) 1.06 16 Other, unknown, or missing 0.942 17 (0.24 – 4.75) 18 Race/Ethnicity* Household Composition & Disability SVI Theme† 19 Non-Hispanic white*Household 0.90 0.233 20 Composition & Disability (0.76 – 1.07) 21 Non-Hispanic Black* Household 1.02 22 0.580 23 Composition & Disability (0.96 – 1.08) Hispanic*Household Composition & 0.94 24 0.399 25 Disability (0.82 – 1.08) 26 Asian or Pacific Islander*Household 1.09 0.265 27 Composition & Disability (0.94 – 1.28) 28 Other, unknown, or missing*Household 0.96 0.595 29 Composition & Disability (0.83 – 1.11) 30 31 Race/Ethnicity and Minority Status & Language SVI Theme 32 Race/Ethnicity 33 Non-Hispanic white (ref.) 1 -- 34 7.85 Non-Hispanic Black <0.001 35 (4.08 – 15.1) 36 4.22 37 Hispanic 0.116 (0.70 – 25.4) 38 5.24 39 Asian or Pacific Islander 0.015 40 (1.37 – 20.0) 41 3.86 Other, unknown, or missing 0.462 42 (0.11 – 140.8) 43 Race/Ethnicity*Minority Status & Language SVI Theme† 44 Non-Hispanic white*Minority Status & 1.22 0.002 45 Language (1.07 – 1.38) 46 Non-Hispanic Black*Minority Status & 0.97 47 0.523 48 Language (0.88 – 1.07) 1.11 49 Hispanic*Minority Status & Language 0.214 50 (0.94 – 1.31) 51 Asian or Pacific Islander*Minority Status & 0.99 0.887 52 Language (0.87 – 1.13) 53 Other, unknown, or missing*Minority Status 1.02 54 0.927 & Language (0.69 – 1.50) 55 56 Race/Ethnicity and Housing Type & Transportation SVI Theme 57 58 59 60 Journal of the American Society of Nephrology Page 35 of 34

1 2 3 Race/Ethnicity 4 Non-Hispanic white (ref.) 1 -- 5 2.14 6 Non-Hispanic Black 0.098 7 (0.87 – 5.29) 8 2.79 Hispanic 0.152 9 (0.69 – 11.4) 10 0.90 Asian or Pacific Islander 0.863 11 (0.26 – 3.11) 12 1.05 13 Other, unknown, or missing 0.936 14 (0.35 – 3.15) † 15 Race/Ethnicity*Housing Type & Transportation SVI Theme Non-Hispanic white*Housing Type & 1.05 16 0.340 17 Transportation (0.95 – 1.16) 18 Non-Hispanic Black*Housing Type & 1.01 0.766 19 Transportation (0.93 – 1.10) 20 1.04 21 Hispanic*Housing Type & Transportation 0.594 (0.90 – 1.21) 22 Asian or Pacific Islander*Housing Type & 1.11 23 0.032 24 Transportation (1.01 – 1.22) 25 Other, unknown, or missing*Housing Type 1.07 0.270 26 & Transportation (0.95 – 1.21) 27 Race/Ethnicity and Housing Crowding 28 Race/Ethnicity 29 Non-Hispanic white (ref.) 1 -- 30 8.20 31 Non-Hispanic Black <0.001 32 (4.09 – 16.4) 16.03 33 Hispanic 0.004 34 (2.37 – 108.5) 35 3.39 Asian or Pacific Islander 0.226 36 (0.47 – 24.5) 37 2.99 38 Other, unknown, or missing 0.500 (0.12 – 72.3) 39 † 40 Race/Ethnicity*Housing Crowding 1.24 41 Non-Hispanic white*Housing Crowding <0.001 42 (1.11 – 1.37) 43 0.98 Non-Hispanic Black*Housing Crowding 0.719 44 (0.89 – 1.08) 45 0.96 Hispanic*Housing Crowding 0.672 46 (0.78 – 1.17) 47 Asian or Pacific Islander*Housing 1.07 48 0.382 49 Crowding (0.92 – 1.24) 50 Other, unknown, or missing*Housing 1.08 0.687 51 Crowding (0.74 – 1.57) 52 Overall SVI, SVI themes, and housing crowding calculated at the census tract-level and expressed as 53 percentiles. 54 †Presented are odds ratios of COVID-19 positivity per 10 percentile increase. 55 56 57 58 59 60 Journal of the American Society of Nephrology Page 36 of 34

Supplemental Table 7. Association of Social Vulnerability Index with COVID-19 among patients on hemodialysis (n = 1378), stratified by race/ethnicity 1 (adjusted analyses). 2 Other, 3 Non-Hispanic Non-Hispanic Asian or Pacific Unknown, or 4 5 white Black Hispanic Islander Missing 6 (n = 294) (n = 578) (n = 207) (n = 174) (n = 125) 7 Variable OR (95% CI) p-value OR (95% CI) p-value OR (95% CI) p-value OR (95% CI) p-value OR (95% CI) p-value † 8 Overall Social Vulnerability Index 9 1.00 0.98 1.03 1.10 1.00 SVI 0.994 0.759 0.864 0.364 0.992 10 (0.88 – 1.14) (0.89 – 1.09) (0.76 – 1.40) (0.90 – 1.35) (0.80 – 1.25) 11 Social Vulnerability Index Theme† 12 1.03 0.93 1.06 1.04 0.99 Socioeconomic status 0.520 0.168 0.663 0.634 0.897 13 (0.95 – 1.12) (0.85 – 1.03) (0.82 – 1.37) (0.88 – 1.24) (0.84 – 1.17) 14 Household composition 0.84 1.02 0.95 1.13 1.05 15 0.039 0.603 0.561 0.324 0.417 16 & disability (0.70 – 0.99) (0.95 – 1.09) (0.80 – 1.13) (0.89 – 1.45) (0.93 – 1.18) Minority status & 1.05 0.98 1.15 0.87 1.26 17 0.492 0.722 0.136 0.207 0.419 18 language (0.92 – 1.19) (0.86 – 1.11) (0.96 – 1.37) (0.70 – 1.08) (0.72 – 2.21) 19 Housing type & 1.14 1.03 1.05 1.11 1.05 0.194 0.491 0.633 0.138 0.603 20 transportation (0.94 – 1.37) (0.94 – 1.14) (0.87 – 1.26) (0.97 – 1.28) (0.88 – 1.24) 21 Social Vulnerability Index Component† 22 1.14 0.98 0.94 0.88 1.08 23 Housing Crowding* 0.010 0.708 0.538 0.132 0.715 24 (1.03 – 1.26) (0.86 – 1.11) (0.76 – 1.15) (0.75 – 1.04) (0.73 – 1.59) † 25 Presented as adjusted odds ratios of COVID-19 positivity per 10 percentile increase in overall SVI, SVI theme, or SVI component. 26 Overall SVI, SVI themes, and SVI component calculated at the census tract-level. 27 Housing crowding defined as the percentage of occupied housing units in a census tract with more people than rooms. 28 All analyses adjusted for age, sex, individual-level social factors (employment, marital status, transportation type to dialysis), dialysis-related medical history 29 (dialysis vintage, primary kidney failure cause), and dialysis facility factors (unit fixed effects and dialysis unit SVI). 30 31 32 33 34 35

36

37 38 39 40 41 42 43 44 45 Journal of the American Society of Nephrology 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60