key: cord-1035271-ltruqtoj
authors: Fisher, Arielle M.; Schlauch, Daniel; Mulloy, Matthew; Dao, Ann; Reyad, Ashraf I.; Correll, Mick; Fromell, Gregg J.; Pittman, James; Bingaman, Adam W.; Sankarapandian, Balamurugan; Allam, Sridhar R.
title: Outcomes of COVID‐19 in hospitalized solid organ transplant recipients compared to a matched cohort of non‐transplant patients at a national healthcare system in the United States
date: 2021-01-18
journal: Clin Transplant
DOI: 10.1111/ctr.14216
sha: 90ebbb0f780373c789701285eb231f5bdc70adf1
doc_id: 1035271
cord_uid: ltruqtoj

Data describing outcomes of solid organ transplant (SOT) recipients with coronavirus disease 2019 (COVID‐19) are variable, and the association between SOT status and mortality remains unclear. In this study, we compare clinical outcomes of SOT recipients hospitalized with COVID‐19 between March 10, and September 1, 2020, to a matched cohort of non‐SOT recipients at a national healthcare system in the United States (US). From a population of 43 461 hospitalized COVID‐19‐positive patients, we created a coarsened exact matched cohort of 4035 patients including 128 SOT recipients and 3907 weighted matched non‐SOT controls. Multiple logistic regression was used to evaluate association between SOT status and clinical outcomes. Among the 4035 patients, median age was 60 years, 61.7% were male, 21.9% were Black/African American, and 50.8% identified as Hispanic/Latino ethnicity. Patients with a history of SOT were more likely to die within the study period when compared to matched non‐SOT recipients (21.9% and 14.9%, respectively; odds ratio [OR] 1.93; 95% confidence interval [CI]: 1.18–3.15). Moreover, SOT status was associated with increased odds of receiving invasive mechanical ventilation (OR [95% CI]: 2.34 [1.51–3.65]), developing acute kidney injury (OR [95% CI]: 2.41 [1.59–3.65]), and receiving vasopressor support during hospitalization (OR [95% CI]: 2.14 [1.31–3.48]).

the course of the pandemic, and the effect prior SOT has on mortality and other clinical outcomes remains unclear.

Single-center studies from New York and Houston reflecting the initial surge of the pandemic in the United States (US) report mortality rates of 24%-36% and 7%, respectively, among prior SOT recipients hospitalized with COVID-19. [3] [4] [5] [6] [7] Pooled mortality rates derived from these studies and other small-sized series [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] are shown to be similar to the observed mortality rate in a large multicenter cohort study from the University of Washington including 482 SOT recipients, which was 20.5%. 18 A more recent observational study on a nationwide SOT registry from France (n = 279) shows a 30-day mortality rate of 22.8% among hospitalized kidney transplant recipients with COVID-19, which is a comparable finding to the US-based large cohort study. 19 Few studies have compared clinical characteristics and outcomes among SOT recipients and matched non-SOT control groups, which is critical to appropriately ascertain risk of mortality in this population. [20] [21] [22] [23] [24] Results from these studies have been conflicting or non-significant regarding association between SOT status and mortality, with three reports concluding risk of mortality is similar in both transplant and non-transplant patients with COVID-19. 20, 21, 23 It is important to note that the source populations and endpoints differ across these studies, which poses a challenge to generalizability of results. However, results from these studies suggest that a large source population from which to sample the matched non-SOT cohort is required to make firm conclusions.

In this study, we aim to better characterize the relationship between SOT status and mortality by comparing outcomes from this population to a large matched non-SOT control group across a multistate community-based healthcare system. This experience is unique given the geographical diversity and scale of this non-tertiary medical system.

This study was supported by HCA Healthcare and institutional review board exempt. The design, analysis, and data interpretations were conducted independently by the investigators. All authors testify to the accuracy and completeness of the data.

The HCA Healthcare system consists of 184 affiliated acute care facilities and over 2000 sites of care in 21 US states. 25 We included consecutive patients with laboratory-confirmed COVID-19 that were hospitalized at an affiliated facility between March 10, and September 1, 2020. Patients were followed until the first of hospital discharge, death, or September 1, 2020-the date on which the data were queried for analysis. A COVID-19-confirmed case was defined as a positive SARS-CoV-2 result on high-throughput sequencing, real-time reverse transcriptase polymerase chain reaction (RT-PCR), or rapid antigen testing of nasopharyngeal swab and other clinical specimens.

Data were collected from the enterprise electronic health record (EHR; Cerner, EPIC, and Meditech) reporting database and compiled in an enterprise data warehouse. We collected detailed data including demographics, coexisting conditions, home medications (including immunosuppressive medications), longitudinal data on vitals and laboratory values, and inpatient medications. Longitudinal data on respiratory support requirements beginning at date of presentation were also collected; patients were categorized based on a modified 5-point clinical scale, adapted from the World Health Organization (WHO) R&D Blueprint group and others to assess clinical improvement. [26] [27] [28] Patients were assigned a score at presentation and subsequently evaluated each day following admission, such that patients received a daily score reflecting level of respiratory support received throughout hospitalization. The modified 5-point scale, hereby referred to as "WHO Index," is as follows: 1, no supplemental oxygen; Table S1 .

The primary exposure was SOT at baseline.

The primary outcome was death within the study period. Key secondary outcomes included intensive care unit (ICU) status, defined as receiving intensive care at any point during hospitalization; receipt of invasive mechanical ventilation, defined as WHO Index 4 during hospitalization; receipt of ECMO, defined as WHO Index 5 during hospitalization; receipt of vasopressor during hospitalization; acute kidney injury (AKI), defined as a ≥0.3 mg/dl increase in serum creatinine from baseline; and acute respiratory distress syndrome (ARDS), defined as PaO2/FiO2 ratio ≤300 mmHg and a score of 3 or higher on the modified 5-point scale. We also collected data on complications including pneumonia, sepsis, and bacteremia based on ICD-10 codes. Additional clinical outcome measures included length of stay, defined as the time between admission until death or discharge from hospital; ICU length of stay, defined as the time between ICU admission until death or discharge from the ICU; length of time from infection to outcome, defined as the time between sample collection for COVID-19 testing to discharge or death; level of respiratory support at most severe, defined as the highest daily score reached in the modified 5-point scale during hospitalization. Cause of death was ascertained by manual chart review for SOT recipients only.

Patients were assigned to the SOT and non-SOT groups using coarsened exact matching (CEM) on the basis of covariates for age, sex, race, ethnicity, body mass index (BMI), hypertension, diabetes mellitus, congestive heart failure, and obesity (defined as BMI ≥30 kg/ m 2 ). We utilized the R implementation of the CEM algorithm in the package "MatchIt" 29 to construct the study population. Baseline patient characteristics were summarized according to SOT status as counts and percentages for categorical variables and median and interquartile range (IQR) for continuous variables. Differences between variables across both groups were assessed by weighted t test for continuous variables and weighted chi-square test for categorical variables.

We used multiple logistic regression to evaluate the association between prior SOT status and death and each of the key secondary outcomes with reporting of coefficients as conditional odds ratios (OR). In addition to the covariates used in the exact matching, values at presentation for absolute neutrophil count, absolute lymphocyte count, D-dimer, and SPO2, as well as level of respiratory support received at presentation (WHO Index 2, 3, and 4) were included in the model. Missing data for baseline characteristics in the multiple logistic regression were imputed with Multivariate Imputation By Chained Equations, using the R package "Mice." 30 Odds ratios, p-values, and 95% confidence intervals were reported for all covariates used in the multiple logistic regression model.

Between March 10, and September 1, 2020, a total of 95 908 patients had a confirmed positive COVID-19 test result within the health system. Of these patients, 45.3% (n = 43 461) were hospitalized, 40.6% (n = 38 944) were emergency department (ED) encounters, 13.2% (12 649) were managed as outpatients, and < 1% (n = 854) of cases were scheduled/same-day surgeries. Only those patients requiring hospitalization were included in our source population (n = 43 461), of which 136 were confirmed SOT recipients. Table S2 includes baseline characteristics, laboratory values and vitals at presentation, and comorbidities for the source population before CEM. Median age was similar between SOT and non-SOT patients in the source population, 60 years vs. 62 years, respectively (p = .33). SOT recipients were more likely to be male, Hispanic/ Latino ethnicity and have comorbid conditions including diabetes, hypertension, and CKD (Table S2) . 

Clinical characteristics in the SOT and non-SOT groups were the same across covariates used in the exact matching (Table 1) . 

In Table 3 , we summarize clinical outcomes across the study population, and In addition to prior SOT, other significant predictors of mortality in the multiple logistic regression model include male sex, age, diabetes mellitus, and values at presentation for absolute neutrophil count and D-dimer ( Figure 2 , Table S4 ). Level of respiratory support received at presentation-namely WHO Index 3 and 4-was also significant predictors of mortality when conditional on other covariates in the model. Similar predictors were observed across multiple logistic regression models for each of the (Tables S5-S9 , Figures S1-S5 ). 

In this study, we report outcomes of COVID-19 among hospitalized SOT recipients compared to non-SOT patients using data from a large multistate, community-based healthcare system in the United However, our main finding of higher mortality among hospitalized SOT recipients is in contrast to results from similar publications. [20] [21] [22] [23] [24] We speculate that differences in inclusion criteria for patient enrollment across studies, such as severity of disease and type of organ transplant, as well as study limitations due to small sample size of SOT cohorts and degree of matching to a control group are contributing to variable findings in these reports.

Several studies each with <50 SOT recipients found no significant association between prior SOT and mortality, with one study detecting a trend toward higher mortality in SOT recipients. 20, 21, 23 These studies were largely limited by small sample sizes and underpowered statistical analyses. Two studies with larger SOT cohorts used propensity score matching to evaluate outcomes between SOT and non-SOT patients; however, findings are limited in generalizability. 22 , 24 Molnar et al included only critically ill patients admitted to the ICU, reporting a mortality rate of 40% among SOT recipients versus 43% for non-SOT patients. 22 We hypothesize that once patients become critically ill from COVID-19, they will experience adverse outcomes regardless of their underlying comorbidities, a possible explanation for the lack of observed effect.

and reported a mortality rate of 8% in both the SOT and non-SOT cohorts. 24 These patients are likely on less intense maintenance immunosuppression than kidney transplant recipients in general, which may explain the lower mortality observed in this group compared to our SOT cohort.

Incidence of AKI in the SOT group in our study is significantly higher at 33.6% vs. 20.2% in the non-SOT group. Higher incidence of AKI in SOT recipients is reported in two other case-control studies. 20 In our assessment of inpatient management, we found usage of remdesivir and dexamethasone to be higher among non-transplant patients, whereas SOT recipients were more likely to receive prednisone. A manual chart review of SOT recipients showed that prednisone was continued at home dosage in a majority of patients. Remdesivir has also been shown to provide clinical benefit in randomized clinical trials. 33 It is possible that higher usage of remdesivir contributed to lower mortality in non-transplant patients; however, evidence has been conflicting regarding potential benefit of this drug. 34 Usage of other therapies that have showed possible clinical benefit in non-controlled studies like convalescent plasma and tocilizumab were similar in both groups.

Our study has several strengths. First, the unique experience provided by data from a community-based healthcare system serving a geographically diverse population of patients across 21 US states and second, our use of CEM to match the comparison group.

We chose CEM over other commonly used methods due to the size of our source population, which offered a broad matching distribution of patients from which to sample the comparison group. This mitigated concerns of loss of sample size due to matching. Further, we sought to match a discrete list of static covariates available at baseline that reflect key risk factors associated with mortality in COVID-19. CEM is expected to produce high performance relative to other matching options in this scenario. 35 This study also has its limitations. First, given this is a large cohort retrospective study, data were collected from EHR systems, which precluded the level of detail possible with a manual medical chart review. Second, we were unable to assess adjustments to immunosuppression regimens during hospitalization and how this may or may not impact mortality. Similarly, our study does not address potential associations between inpatient therapies and outcomes. Additionally, we did not study temporal trends in mortality as treatment algorithms evolved with steroids and remdesivir becoming standard of care for moderate-to-severe COVID-19 during the later months of our study period. Third, we defined CKD using ICD-10 codes rather than estimated glomerular filtration rate values due to lack of longitudinal data, however only those diagnoses codes entered after the transplant date were considered. We chose not to include CKD as a covariate in the logistic regression model given its association with transplantation, but it is possible this contributed to increased mortality among SOT recipients. We acknowledge that data on patients with AKI requiring renal replacement therapy are missing in our study. Lastly, as our healthcare system is primarily inpatient centered, we were not able to study outcomes of COVID-19

for non-hospitalized patients and if SOT recipients have different outcomes in the outpatient setting.

In conclusion, patients with prior SOT that are hospitalized with COVID-19 have a higher risk of mortality compared to patients without a history of SOT, suggesting an independent relationship between SOT status and mortality. SOT patients are also more likely to develop AKI and require invasive mechanical ventilation and vasopressor support during hospitalization. More research is needed to assess the effect of adjustments in immunosuppression regimens and inpatient management on mortality and other clinical outcomes. 

Authors of this manuscript have no conflicts of interest relevant to this study to disclose. 

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

Ann Dao https://orcid.org/0000-0003-3457-1568

Ashraf I. Reyad https://orcid.org/0000-0002-1128-928X

Sridhar R. Allam https://orcid.org/0000-0001-6051-9443

COVID-19 in solid organ transplant recipients: initial report from the US epicenter

Early experience with COVID-19 and solid organ transplantation at a US high-volume transplant center

COVID-19 in liver transplant recipients: an initial experience from the US epicenter

Covid-19 and kidney transplantation

COVID-19 in kidney transplant recipients

A single center observational study of the clinical characteristics and short-term outcome of 20 kidney transplant patients admitted for SARS-CoV2 pneumonia

Characteristics and outcomes of recipients of heart transplant with coronavirus disease

First experience of SARS-CoV-2 infections in solid organ transplant recipients in the Swiss transplant cohort study

Covid-19 in solid organ transplant recipients: a single center experience

COVID-19 in solid organ transplant recipients: a single-center case series from Spain

Should cyclosporine be useful in renal transplant recipients affected by SARS-CoV-2?

COVID-19 in elderly kidney transplant recipients

Clinical outcome in solid organ transplant recipients with COVID-19: a single-center experience

Coronavirus disease-2019 in heart transplant recipients in Southeastern Michigan: a case series

Coronavirus disease 2019 pneumonia in immunosuppressed renal transplant recipients: a summary of 10 confirmed cases in Wuhan, China

COVID-19 in solid organ transplant: a multi-center cohort study

An initial report from the French SOT COVID registry suggests high mortality due to Covid-19 in recipients of kidney transplants

Clinical characteristics and outcomes of COVID-19 in solid organ transplant recipients: a cohort study

COVID-19 in solid organ transplant recipients: no difference in survival compared to general population

Outcomes of critically Ill solid organ transplant patients with COVID-19 in the United States

COVID-19 in solid organ transplantation: a matched retrospective cohort study and evaluation of immunosuppression management

Clinical characteristics, hospitalization and mortality rates of COVID-19 among liver transplant patients in the United States: a multi-center research network study

Top US health systems by total member hospital beds, definitive healthcare

Novel coronavirus COVID-19 therapeutic trial synopsis

A minimal common outcome measure set for COVID-19 clinical research

Treatment of COVID-19 patients with convalescent plasma in Houston

MatchIt: nonparametric preprocessing for parametric causal inference

Mice: multivariate imputation by Chained Equations in R

Association between administration of systemic corticosteroids and mortality among critically Ill patients With COVID-19: a meta-analysis

Drug treatments for covid-19: living systematic review and network meta-analysis

Remdesivir for the #reat-ment of Covid-19 -final report

Remdesivir in adults with severe COVID-19: a randomised, double-blind, placebo-controlled, multicentre trial

Evaluating the utility of coarsened exact matching for pharmacoepidemiology using real and simulated claims data

Outcomes of COVID-19 in hospitalized solid organ transplant recipients compared to a matched cohort of non-transplant patients at a national healthcare system in the United States