Introduction

Since December 2019, infection with SARS-CoV-2, the virus that causes COVID-19, has led to one of the world’s most serious infectious disease pandemics. While most individuals affected by SARS-CoV-2 exhibit only mild symptoms, many individuals require hospitalization, with the most acute patients needing intensive care unit (ICU) admission, mechanical ventilation and extracorporeal membrane oxygenation (ECMO) treatment [1]. Infection is the second most reported cause of death for those on dialysis [2], with annual death rates from pneumonia and sepsis substantially higher than that of the general population [3]. ESRD patients have a compromised and poorly regulated immune system, which may increase susceptibility to bacterial and viral infections, including the SARS-CoV-2 virus [4].

In a retrospective study of 1,344 adult COVID-19 infected patients discharged alive from the emergency department or hospital between March 3 and May 15, 2020 in a large New York City health system, Kingery et al. [5] examined the risks for rehospitalization within 30 days and overall (combined hospital and non-hospital) death within 30 days post-discharge among individuals post-discharge for COVID-19 infection with ESRD (N = 75) compared to those without. After adjusting for numerous demographic and comorbid conditions, including “socio-demographics (age, sex, race, housing status), length of stay, comorbidities (tobacco use, body mass index [BMI], cardiovascular disease, chronic kidney disease, diabetes, chronic lung disease, HIV, active cancer, and other immunocompromised state), symptoms (fever, cough, diarrhea, nausea/vomiting, myalgia, dyspnea), and chest imaging at presentation; clinical course (new-onset myocardial infarction/arrhythmia/heart failure, dialysis, vasopressor requirement, hypoxia upon presentation, intubation/extubation, and tracheostomy) and infection parameters (PCR viral load, respiratory viral pathogen panel, and blood culture)”, the authors found a significantly higher hazard of rehospitalization within 30 days (adjusted hazard ratio (HR) 2.94, 95% CI 1.78–4.84, p < .0001) among individuals post-discharge with ESRD compared to those without, but no difference in the hazard of overall death within 30 days post-discharge for these two groups (adjusted HR 1.22, 95% CI 0.29–5.12, p = 0.789). In a larger national cohort study by Verna et al [6] of 29,659 patients hospitalized with COVID-19 at one of 297 hospitals across 40 US states between February 15 and June 9, 2020, patients with ESRD (N = 1,336) were also found to have a significantly higher odds of 30-day hospital readmission (adjusted odds ratio (OR) 2.27, 95% CI 1.81–2.86) and a higher, but not statistically significant odds of death during hospital readmission (adjusted OR 1.43, 95% CI 0.63–3.24). Covariates adjusted for in their statistical model included age, sex, length of index hospital stay, census region, insurance type, and the comorbidities chronic kidney disease, ESRD, collagen vascular disease (CVD), hypertension, obesity, cerebrovascular disease, pulmonary disease, diabetes, smoking, cardiac disease, respiratory disease, and sepsis/systemic inflammatory response syndrome. Interestingly, this study also found a difference in the risk of 30-day readmission by geographical area, with odds higher in the United States Northeast compared with either the West or the South.

Huang et al. [7] conducted a retrospective cohort study of 2,180 patients discharged from 15 Kaiser Permanente Southern California medical centers between April 1 and July 31, 2020. After adjusting for age, sex, race, BMI, Charlson Comorbidity Index (a tool used to predict long-term mortality in patients with multiple comorbidities), days from diagnosis to index hospital admission, mechanical ventilation, oxygen therapy, length of stay and discharge disposition for index hospitalization, and the comorbidities chronic pulmonary disease, diabetes, hypertension kidney disease, and chronic heart failure, the authors found that patients with kidney disease had a higher, but not statistically significant, odds of rehospitalization (adjusted OR 1.35, 95% CI 0.81–2.26, p = 0.25). A large meta-analysis of 28 national and international studies of 68,236 patients hospitalized for COVID-19 conducted by Akbari et al [8] found that patients with kidney disease had a significantly higher hazard of rehospitalization within 30 days (adjusted OR 2.52, 95% CI 1.23–2.85, p < .0001). Lastly, in a study of 222,154 Medicare beneficiaries on dialysis with 436,745 live acute-care hospital discharges from 7871 Medicare-certified dialysis facilities between January 1 and October 31, 2020, Wu et al [9] found a significantly higher hazard of rehospitalization (adjusted HR 1.59, 95% CI 1.53–1.65, p < 0.001) for patients with a diagnosis of COVID-19.

It has been documented that ESRD patients in the United States had a high unplanned hospital 30-day readmission rate prior to the COVID-19 pandemic. The United States Renal Data System Annual Data Report lists national year 2020 30-day rehospitalization rates of 39.8% for ESRD patients on hemodialysis and 38.5% for ESRD patients on peritoneal dialysis. A study by Donnelly et al [10] of patients admitted with COVID-19 to the Veterans Affairs Healthcare System found that risk for adverse clinical outcomes following a hospitalization for COVID-19 is highest in the 7–10 days following discharge. For some diseases such as heart failure, hospital readmission rates during the pandemic may have been lower than historical norms due to the impacted hospital system and the desire to reduce patient exposure to other hospitalized COVID-19 patients.

In our study, we examined the post-hospital discharge outcomes of ESRD patients admitted for COVID-19 in a diverse Western United States population. We compared outcomes post-hospitalization for COVID-19 in ESRD patients to that of individuals without ESRD, at 5 University of California academic medical centers. Our objective was to study if ESRD patients hospitalized with COVID-19 had a significantly higher adjusted risk of hospital, non-hospital, or overall death within 180 days or of 30-day hospital readmission, compared with those who did not have this disease.

Methods

We used the University of California COVID Research Database (UC CORDS) which is a large, harmonized database provided by the UC Health Data Warehouse that includes patients from the 5 tertiary care, academic hospitals of the University of California (located in Davis, Irvine, Los Angeles, San Diego and San Francisco). Written informed consent was not required to conduct this study. It was granted an exemption for human subjects protection by the UC Davis Institutional Review Board (protocol# 1604619–1).The authors had access to information that could potentially identify individual participants during or after data collection (the data were accessed for research purposes on 12/22/2023), since the detailed pseudonymized patient data used for this study were potentially re-identifiable. This retrospective longitudinal study included adults with and without ESRD, who were 18 years or older when first hospitalized for COVID-19 (qualifying or index hospital admission), and who tested positive for COVID-19 by polymerase chain reaction nasal swab within 30 days prior to or at any time during their hospital stay. We included all adult patients with COVID-19 admitted between January 6, 2020 and August 3, 2023, and discharged alive by August 4, 2023, with 3 exceptions: individuals were excluded from the study if, for their index hospital admission, they were transferred into the University of California hospital system from a hospital outside this system (N = 1,053), they were transferred out of the University of California hospital system (N = 503), or they were admitted to an inpatient obstetric service (N = 259). The study end date was August 4, 2023. Individuals were classified as having ESRD if their index hospital admission included the International Classification of Diseases, 10^th revision, Clinical Modification (ICD-10-CM) code N18.6. We excluded individuals with an active renal transplant from the ESRD group but included individuals with failed kidney transplants who were back on long-term dialysis.

Demographic and baseline characteristics included age in years at the index hospital admission, sex, race and ethnicity. Self-reported data on race and ethnicity were combined and categorized as White, Hispanic, Black, Asian, and Other/Unknown. The date of the index hospital admission was grouped into hospitalization periods for each of the COVID-19 waves which occurred during the study period. Comorbid conditions identified in individuals with/without ESRD prior to or on the date of their qualifying hospital admission are listed in Table 1. Individuals were classified as having or not having these comorbid conditions by utilizing ICD-10-CM codes.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Demographic and clinical characteristics of individuals with/without ESRD Post-Hospital Discharge for COVID-19 (N = 11,406).

https://doi.org/10.1371/journal.pone.0332203.t001

We included area deprivation index (ADI) as a factor in our analyses. The ADI for a given neighborhood (census block group) provides a measure of relative socioeconomic disadvantage [11,12] that encompasses income, education, employment, and housing against the mean values in the State (ADIs are state specific). The ADI ranges from 1 to 10, with 1 and 10 signifying a neighborhood that is least and most socioeconomically disadvantaged, respectively. Individuals in the UC CORDS database were assigned an ADI based upon their residence zip code. In our study, 5.3% of those with ESRD and 10.2% of those without ESRD were missing an ADI. Medications administered most likely for COVID-19 treatment to patients during their qualifying hospital stay were also queried, as was information collected on the in-hospital use of high acuity treatments. The number of COVID-19 vaccines recorded for patients in the UC system prior to index hospital admission was queried, however in our study, no record of COVID-19 vaccination was found for 81.5% of those with ESRD and 80.1% of those without ESRD.

We utilized simple and multiple Cox proportional hazards regression models to analyze the association of ESRD status with each of the study outcomes. In addition to unadjusted models, two adjusted Cox proportional hazards regression models were fitted for each of the outcomes hospital, non-hospital, and overall death within 180 days post-discharge, and for 30-day hospital readmission. Model 1 included the demographic variables sex, age per 10 years increase, and race/ethnicity, and also a variable for index hospitalization period. Model 2 included the above demographic variables and the variable for index hospitalization period, plus any comorbid conditions associated at the p = 0.15 level with both the outcome and the study exposure, and which caused a ≥ 10% change in the unadjusted Wald chi-square HR when added as a factor to the unadjusted Cox proportional hazards regression model. For the outcomes hospital, non-hospital, and overall death within 180 days post-discharge model 2 included the comorbid conditions anemia, cancer, cardiac peripheral vascular disease, cardiac arrhythmias, major cardiac disease, cardiac valvular disease, coagulopathy, and hypertension, while for the outcome 30-day hospital readmission model 2 included the comorbid conditions anemia, major cardiac disease, and coagulopathy. We checked the proportional hazard assumption for each of our multivariate Cox proportional hazards regression models by plotting the scaled Schoenfeld residuals, and we did not find any violations of this assumption. Additionally, forward stepwise multivariate Cox proportional hazards regression models were utilized to determine independent risk factors significantly associated with hospital and non-hospital death within 180 days post-discharge for all individuals, and with 30-day readmission, separately among individuals with and without ESRD. Risk factors with Wald chi-square HRs significant at the p = 0.05 level were retained in each of the multivariate models. In all analyses, we used two-sided tests and alpha = 0.05 without adjusting for multiple comparisons.

Results

After applying the study inclusion and exclusion criteria, a total of 11,406 individuals were discharged alive after being hospitalized for COVID-19 between January 6, 2020 to August 3, 2023. Of these, 713 (6.3%) had ESRD and 10,693 (93.7%) did not. Of the 11,406 individuals, 1,344 (11.8%) were readmitted to the hospital within 30 days post-discharge, 144 (20.2%) in the ESRD group and 1,200 (11.2%) in the non-ESRD group. Additionally, of the 11,406 individuals, 389 (3.4%) died within 180 days post-discharge, 35 (4.9%) in the ESRD group and 354 (3.3%) in the non-ESRD group. Of the 35 individuals with ESRD who died within 180 days post-discharge, 16 (45.7%) died in-hospital, while 112 (31.6%) of the 354 individuals without ESRD who died within 180 days post-discharge did so in-hospital.

The demographic and clinical characteristics for individuals with/without ESRD prior to hospital admission as well as the medications and high acuity treatments administered to them in-hospital are provided in Table 1. Individuals with ESRD were more likely than those who did not have this disease to live in a more socioeconomically disadvantaged census tract, to self-identify as either Hispanic (48.9% vs 36.7%) or Black (12.1% vs 8.0%), to be obese (38.1% vs 28.0%) to have more comorbid conditions, to receive an oral blood thinner (22.4% vs 16.0%), to receive mechanical ventilation (17.7% vs 11.3%), to have one or more complications during their qualifying hospital stay, and for that stay to be ≥ 7 days (59.5% vs 48.2%). In contrast, individuals with ESRD were less likely than those without ESRD to receive dexamethasone during their qualifying hospital stay (40.3% vs 47.6%).

The baseline characteristics of those who died in-hospital/outside of hospital within 180 days post-discharge, and the medications and treatments administered to them during their index in-hospital stay, are provided in Table 2. Of the 11,406 patients who were discharged alive, 128 (1.1%) died in-hospital within 180 days. Those who died in-hospital were older (median age 68 vs 60), more likely to be smokers (29.7% vs 20.1%), to have more comorbid conditions, and to receive Remdesivir (55.5% vs 43.9%), dexamethasone (52.3% vs 47.1%), and an oral blood thinner (32.0% vs 16.3%), to be on a ventilator (24.2% vs 11.5%), to have more complications, to be admitted to an ICU (27.3% vs 17.4%), and to remain in the hospital ≥7 days (71.1% vs 48.7%) during their qualifying hospital stay. Of the 11,406 patients who were discharged alive, 261 (2.3%) died outside of the hospital within 180 days. Those who died outside of the hospital were older (median age 70 vs 60), more likely to self-identify as White (43.3% vs 29.8%), to be smokers (26.1% vs 20.1%), to have more comorbid conditions, to receive an oral blood thinner (28.4% vs 16.2), to have more complications, and to remain in the hospital ≥7 days (64.0% vs 48.6%) during their qualifying hospital stay. In contrast, these individuals were less likely than those without ESRD to self-identify as Hispanic (25.7% vs 37.7%). The percentages of those who died in-hospital/outside of hospital within 180 days post-hospital discharge with selected demographic and clinical risk factors are shown in Fig 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Demographic and clinical characteristics of individuals who died in-hospital/outside of hospital 180 days post-hospital discharge (N = 11,406).

https://doi.org/10.1371/journal.pone.0332203.t002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Percentage of individuals who died in-hospital/outside of hospital within 180 days post-hospital discharge with selected demographic and clinical risk factors (N = 11,406).

Notes: IHC = index hospital complication, Transplantation = solid organ transplantation, excluding kidney, Stroke = intracerebral bleeding cerebral infraction or stroke. CVD = collagen vascular disease, Pulmonary disease includes asthma and chronic obstructive pulmonary disease.

https://doi.org/10.1371/journal.pone.0332203.g001

The baseline characteristics of those with/without ESRD, and the medications and treatments administered to them in the hospital, by readmission status 30 days post-discharge, are provided in Table 3. Of the 713 patients with ESRD who were discharged alive, 144 (20.2%) were readmitted to the hospital within 30 days. These individuals were more likely to be female (50.0% vs 42.4%), to self-identify as Hispanic (52.8% vs 48.0%) or Black (15.3% vs 11.2%), to be smokers (34.7% vs 19.2%), to be obese (48.6% vs 35.5%), to have more comorbid conditions, and to receive an SSRI (18.1% vs 14.1%), to have more in-hospital complications, and to remain in the hospital ≥7 days (63.9% vs 58.3%) during their qualifying hospital stay. Of the 10,693 patients without ESRD who were discharged alive, 1,200 (11.2%) were readmitted to the hospital within 30 days. These individuals were more likely to live in an area with a high ADI, to be smokers (25.3% vs 19.4%), to have more comorbid conditions, and to have more in-hospital complications and remain in the hospital ≥7 days (52.8% vs 47.7%) during their qualifying hospital stay. These individuals were also less likely to have received dexamethasone during that hospital stay (42.1% vs 48.3%). The percentages of individuals with/without ESRD who were readmitted within 30 days post-hospital discharge with selected demographic and clinical risk factors is shown in Fig 2. This figure shows that individuals with ESRD who were readmitted to the hospital had a much higher prevalence of many comorbid conditions than those without ESRD.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Demographic and clinical characteristics of individuals with/without ESRD by readmission status 30 days post-hospital discharge (N = 11,406).

https://doi.org/10.1371/journal.pone.0332203.t003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Percentage of individuals with/without ESRD readmitted within 30 days post-hospital discharge with selected demographic and clinical risk factors (N = 11,406).

Notes: IHC = index hospital complication, Transplantation = solid organ transplantation, excluding kidney, Stroke = intracerebral bleeding cerebral infraction or stroke. CVD = collagen vascular disease, Pulmonary disease includes asthma and chronic obstructive pulmonary disease.

https://doi.org/10.1371/journal.pone.0332203.g002

Unadjusted and multivariable-adjusted HRs for hospital, non-hospital, and overall death within 180 days post-discharge, and for 30-day hospital readmission among those with and without ESRD are provided in Table 4. After determining the factors to be included in our adjusted model 2 analyses, we found that hospital death and overall death within 180 days post-discharge were significantly higher for individuals with ESRD in the unadjusted analyses but not in the analyses adjusted for demographic variables, hospitalization period, and comorbid conditions (adjusted HR 1.36, 95% CI 0.78–2.37 and adjusted HR 1.05, 95% CI 0.73–1.51, respectively). We did not find ESRD status to be associated with non-hospital death within 180 days post-discharge in either the unadjusted or the adjusted analyses (adjusted HR 0.89, 95% CI 0.55–1.44). Finally, we found that compared to those without kidney failure, those with ESRD had a significantly higher hazard of 30-day readmission post-discharge in both the unadjusted and the adjusted analyses (adjusted HR 1.36, 95% CI 1.14–1.63, p = 0.001).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Hazard ratios for outcomes post-hospital discharge among individuals with/without ESRD (N = 11,406) (the group without ESRD is the referent).

https://doi.org/10.1371/journal.pone.0332203.t004

The analysis of risk factors for hospital death within 180 days post-discharge is shown in Table 5, with multivariate risk factors for this outcome shown in Fig 3. We found cancer to be the factor most associated with a significantly increased risk of hospital death within 180 days post-discharge (adjusted HR = 3.12, 95% CI 2.16–4.49, p < 0.0001), followed by major cardiac disease (adjusted HR = 2.59, 95% CI 1.43–4.7, p = 0.0017), anemia (adjusted HR = 1.97, 95% CI 1.36–2.87, p = 0.0004), mechanical ventilation (adjusted HR = 1.96, 95% CI 1.28–2.98, p = 0.0018), high hospital length of stay (adjusted HR = 1.74, 95% CI 1.15–2.62, p = 0.0082), the hospital complications septic shock (adjusted HR = 1.67, 95% CI 1.13–2.45, p = 0.0093), and acute MI (adjusted HR = 1.55, 95% CI 1.03–2.34, p = 0.0364), and the administration of oral blood thinners during the index hospitalization (adjusted HR = 1.51, 95% CI 1.02–2.24, p = 0.0409). The analysis of risk factors for non-hospital death within 180 days post-discharge is shown in Table 6 with multivariate risk factors for this outcome shown in Fig 4. We found being 80 or more years of age to be the factor most associated with a significantly increased risk of non-hospital death within 180 days post-discharge (adjusted HR = 4.81, 95% CI 2.46–9.41, p < 0.0001), followed by cancer (adjusted HR = 3.93, 95% CI 3.02–5.11, p < 0.0001), being age 50–59 years (adjusted HR = 3.38, 95% CI 1.69–6.74, p = 0.0005), being age 60–69 years (adjusted HR = 3.02, 95% CI 1.53–5.96, p = 0.0015), being age 40–49 years (adjusted HR = 2.48, 95% CI 1.14–5.37, p = 0.0215), being age 70–79 years (adjusted HR = 2.45, 95% CI 1.21–4.93, p = 0.0125), the hospital complications acute myocardial infarction (adjusted HR = 1.79, 95% CI 1.34–2.39, p < 0.0001) and lung embolism (adjusted HR = 1.57, 95% CI 1.07–2.29, p = 0.0211), cerebrovascular disease (adjusted HR = 1.53, 95% CI 1.15–2.03, p = 0.0031), anemia (adjusted HR = 1.52, 95% CI 1.16–1.99, p = 0.0026), liver disease (adjusted HR = 1.48, 95% CI 1.12–1.96, p = 0.0064), coagulopathy (adjusted HR = 1.45, 95% CI 1.11–1.89, p = 0.0065), high hospital length of stay (adjusted HR = 1.38, 95% CI 1.06–1.80, p = 0.0173), and the hospital complication septic shock (adjusted HR = 1.36, 95% CI 1.05–1.77, p = 0.0207).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Univariate and multivariate Cox proportional hazards regression analyses of risk factors associated with hospital death within 180 days post-hospital discharge (N = 11,406, Number of Hospital Deaths = 128).

https://doi.org/10.1371/journal.pone.0332203.t005

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 6. Univariate and multivariate Cox proportional hazards regression analyses of risk factors associated with non-hospital death within 180 days post-hospital discharge (N = 11,406, Number of Non-Hospital Deaths = 261).

https://doi.org/10.1371/journal.pone.0332203.t006

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Forest plot showing the multivariate risk factors for hospital death within 180 days post-hospital discharge (N = 11,406).

HC = hospital complication.

https://doi.org/10.1371/journal.pone.0332203.g003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Forest plot showing the multivariate risk factors for non-hospital death within 180 days post-hospital discharge (N = 11,406).

HC = hospital complication, referent group for age category is 18–39 years.

https://doi.org/10.1371/journal.pone.0332203.g004

The analysis of risk factors for 30-day readmission post-discharge is shown in Table 7, with multivariate risk factors for this outcome shown in Fig 5. We found cancer to be the factor most associated with a significantly increased risk of 30-day readmission post-discharge (adjusted HR = 2.03, 95% CI 1.79–2.29, p < 0.0001), followed by the hospital complication deep vein thrombosis (adjusted HR = 1.75, 95% CI 1.15–2.66, p = 0.0094), major cardiac disease (adjusted HR = 1.36, 95% CI 1.18–1.58, p < 0.0001), coagulopathy (adjusted HR = 1.36, 95% CI 1.20–1.54, p < 0.0001), kidney failure (adjusted HR = 1.30, 95% CI 1.07–1.58, p= 0.0070), the hospital complication septic shock (adjusted HR=1.28, 95% CI 1.13-1.44, p < 0.0001), being a current or a past smoker (adjusted HR = 1.27, 95% CI 1.12–1.45, p = 0.0003), the administration of SSRIs during the index hospital stay (adjusted HR = 1.23, 95% CI 1.06–1.44, p = 0.0073), anemia (adjusted HR = 1.23, 95% CI 1.08–1.40, p = 0.0021), valvular cardiac disease (adjusted HR = 1.21, 95% CI 1.04–1.40, p = 0.0141), high hospital LOS (adjusted HR = 1.19, 95% CI 1.06–1.35, p = 0.0035), hypertension (adjusted HR = 1.18, 95% CI 1.02–1.37, p = 0.0264), and ADI (adjusted HR = 1.03, 95% CI 1.01–1.05, p = 0.0045). Additionally, we found the following factors to be associated with a significantly reduced risk of 30-day readmission: initial hospitalization during the period Jan-May 2020 (adjusted HR = 0.35, 95% CI 0.21–0.60, p = 0.0001), being 80 years of age or older (adjusted HR = 0.56, 95% CI 0.45–0.70, p < 0.0001), being 70−79 years of age (adjusted HR = 0.56, 95% CI 0.45–0.69, p < 0.0001), initial hospitalization during the period Jun-Aug 2020 (adjusted HR = 0.61, 95% CI 0.44–0.85, p = 0.0029), high ICU length of stay (adjusted HR = 0.67, 95% CI 0.51–0.88, p = 0.0040), being age 50−59 years (adjusted HR = 0.70, 95% CI 0.58–0.85, p = 0.0004), being age 60−69 years (adjusted HR = 0.71, 95% CI 0.58–0.85, p = 0.0003), the administration of Remdesivir during the index hospital stay (adjusted HR = 0.74, 95% CI 0.65–0.83, p < 0.0001), and being age 40−49 years (adjusted HR = 0.76, 95% CI 0.61–0.94, p = 0.0126).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 7. Univariate and multivariate Cox proportional hazards regression analyses of risk factors associated with hospital readmission within 30 days post-hospital discharge (N = 10,281, Number of Hospital Readmissions = 1,344).

https://doi.org/10.1371/journal.pone.0332203.t007

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Forest plot showing the multivariate risk factors for hospital readmission within 30 days post-hospital discharge (N = 10,281).

HC = hospital complication, IAP = index admission period, referent groups for index admission period, age category are Feb-May 2021, 18-39 years, respectively.

https://doi.org/10.1371/journal.pone.0332203.g005

Discussion

Using a large database that spans the five academic medical centers of the University of California, we sought to determine if ESRD was independently associated with a higher risk for hospital, non-hospital, and overall death within 180 days post-discharge, and 30-day hospital readmission for those diagnosed with COVID-19. Individuals receiving hemodialysis for ESRD are known to have a high 30-day readmission rate, with the reason for readmission not necessarily related to the reason for the index hospitalization [13]. A similar pattern exists for COVID-19 related hospitalizations. Our retrospective study spanned almost the entire period of the COVID-19 pandemic. As a result, the dominant strain of SARS-CoV-2 in the community, and that which caused hospitalizations during the pandemic, likely changed over the course of the study. However, including a factor for index hospitalization period did not substantially affect our results. Individuals with ESRD have many of the comorbid conditions linked with worse outcomes from COVID-19 and are more frequently readmitted post-hospitalization. It is not unexpected therefore, that as was found by Kingery et al and Verna et al who studied patients with ESRD, and Akbari et al who studied patients with kidney disease, we found a diagnosis of ESRD, upon adjusting for demographic variables, comorbid conditions, and other covariates, to be independently associated with a significantly higher risk of 30-day readmission for individuals hospitalized with COVID-19. The association which we found (adjusted HR 1.36, 95% CI 1.14–1.63, p = 0.001) however, was lower in magnitude than that found by Kingery et al (adjusted HR 2.94, 95% CI 1.78–4.84, p < 0.0001), Verna et al (adjusted OR 2.27, 95% CI 1.81–2.86), and Akbari et al (adjusted OR 2.52, 95% CI 1.23–2.85, p < 0.0001). The magnitude of the association between 30-day readmission and presence of ESRD which we found was also nearly identical to the association found by Huang et al between 30-day readmission and presence of renal disease (N = 1,280, adjusted OR 1.35, 95% CI 0.81–2.26, p = 0.25).

Our second finding was that the presence of ESRD was independently associated with a higher risk of in-hospital and overall death within 180 days (adjusted HR 1.36, 95% CI 0.78–2.37, and adjusted HR 1.05, 95% CI 0.73–1.51, respectively) although we did not find these risks to be statistically significant. This may be possibly due to the immune system compromise of individuals with ESRD, and the resulting reduction of cytokine storms and systemic inflammation [14]. Our finding for the risk of overall death within 180 days post-discharge was also lower than that found by Kingery et al for the risk of overall death within 30 days (adjusted HR 1.22, 95% CI 0.29–5.12, p = 0.789), and Verna et al for the risk of death during hospital readmission (adjusted OR 1.43, 95% CI 0.63–3.24), perhaps because our study spanned the period January 2020 through August 2023, and the other two studies were conducted early in year 2020, when a higher mortality rate from COVID-19 was observed.

Our third finding was that the presence of ESRD was independently associated with a lower risk of non-hospital death within 180 days (adjusted HR 0.89, 95% CI 0.55–1.44 for non-hospital death), although again, we did not find this risk to be statistically significant.

When we examined the associations between our risk factors and our study outcomes, we found that the comorbid conditions cancer and anemia, high hospital length of stay, and the hospital complications septic shock and acute myocardial infarction were significantly associated with increased hazard of both hospital and non-hospital death within 180 days post-discharge. Furthermore, each of the first four of these risk factors were also significantly associated with an increased hazard of 30-day readmission post-discharge. Interestingly, we found that the comorbid conditions valvular cardiac disease and hypertension, the hospital complication deep vein thrombosis, smoking, in-hospital use of SSRIs, and higher ADI were factors significantly associated with an increased hazard of 30-day hospital readmission, but that these factors were not significantly associated with an increased hazard of either hospital or non-hospital death within 180 days. Similarly, high ICU length of stay, in-hospital use of Remdesivir, and index hospitalization in the period January-August 2020 were factors significantly associated with a reduced hazard of 30-day hospital readmission, but not with a significantly reduced hazard of either hospital or non-hospital death within 180 days. We found that being 40 or more years of age was significantly associated with a reduced hazard of 30-day hospital readmission, but in contrast, was also significantly associated with an increased hazard of non-hospital death within 180 days.

The strengths of this study include its large cohort size, its diverse racial groups, 3.5-year long duration, and the geographical and population range covered by the 5 academic medical centers in California from which we enrolled patients. We also included many comorbid conditions and complications, and we were able to capture key therapeutics used to treat COVID-19 patients during the period covered by our study. Like all studies of this kind, the retrospective nature of our study is a limitation. Additional limitations include that comorbidity ascertainment was performed using diagnostic codes, which vary in their degree of accuracy and completeness, and which we were not able to validate with chart review due to the pseudonymized nature of the dataset. Given the nature of the UC CORDS database, it was also not possible to study some patient-level variables, including blood type, familial history, and genetic factors, each of which might have contributed to residual confounding for the associations which we studied. Additionally, since this study spans 3.5 years of the pandemic, patient care management with increased experience with COVID-19 care may have changed clinical practices over time [15], possibly resulting in aggregation bias. Also, some patients with COVID-19 admitted to one of the University of California medical centers during our study period may have had false negative or no recorded COVID-19 test results, possibly resulting in selection bias. The wider availability and use of COVID-19 vaccines over the study period could also conceivably have contributed to aggregation bias. Vaccination against SARS-CoV-2, which reduces severe disease, hospitalizations, and death, was introduced in California in late December 2020. We sought to identify hospitalized patients who had received 1 or more vaccinations and only found 2,265 such patients (19.9%). Our database identified patients who had a vaccination recorded within (or information transferred to) the UC health system, whereas population vaccination was primarily through community vaccination clinics, pharmacy chains, and dialysis clinics. Therefore, our database likely undercounted the number of vaccinated patients. It is also possible that the vaccines worked effectively, reducing hospitalizations for COVID-19 [16]. Nonetheless, adjusting for limited information about COVID-19 vaccination status in the analysis did not affect our results.

In conclusion, we found that after adjusting for demographic factors, hospitalization period, patient comorbidities, and other covariates, ESRD status was not found to significantly increase the risk for hospital, non-hospital, or overall death within 180 days post-discharge (adjusted hazard ratio (HR) 1.36, 95% CI 0.78–2.37, p = 0.278; adjusted HR 0.89, 95% CI 0.55–1.44, p = 0.634; and adjusted HR 1.05, 95% CI 0.73–1.51, p = 0.795, respectively), but was found to significantly increase the risk of 30-day hospital readmission (adjusted HR 1.36, 95% CI 1.14–1.63, p = 0.001). We recommend that caution be exercised when discharging ESRD patients, especially those with a history of major cardiac disease, anemia, or coagulopathy. We also suggest that the etiology of long-haul COVID-19 and of ESRD patients post hospitalization for COVID-19 be further studied to ascertain what extent long-haul COVID-19 impacts clinical outcomes for these patients.