Healthcare Disparities and Outcomes of Cancer Patients in a Community Setting from a COVID-19 Epicenter

There have been numerous studies demonstrating how cancer patients are at an increased risk of mortality. Within New York City, our community hospital emerged as an epicenter of the first wave of the pandemic in the spring of 2020 and serves a unique population that is predominately uninsured, of a lower income, and racially/ethnically diverse. In this single institution retrospective study, the authors seek to investigate COVID-19 diagnosis, severity and mortality in patients with an active cancer diagnosis. Demographic, clinical characteristics, treatment, SARS-CoV-2 laboratory results, and outcomes were evaluated. In our community hospital during the first wave of the COVID-19 pandemic in the United States, patients with active cancer diagnosis appear to be at increased risk for mortality (30%) and severe events (50%) due to the SARS-CoV-2 infection compared to the general population. A higher proportion of active cancer patients with Medicaid insurance, Hispanic ethnicity, other race, and male sex had complications and death from COVID-19 infection. The pandemic has highlighted the health inequities that exist in vulnerable patient populations and underserved communities such as ours.


Introduction
The novel coronavirus disease 2019 (COVID-19) pandemic due to SARS-CoV-2 virus has resulted in over 5 million deaths worldwide with over 4 million cases diagnosed in New York State alone [1]. New York City (NYC) had its first confirmed case diagnosed on 29 February 2020 and in the subsequent months the city's hospitals were inundated with infected patients [2]. Within New York City, Elmhurst Hospital Center (EHC) emerged as an epicenter [3]. At the peak on 23 March 2020 there was an 83.1% test positivity rate at EHC [4]. EHC is a 545 bed public hospital and serves one of most racially and ethnically diverse populations in the United States with a catchment zone of roughly 1.2 million residents of Queens [5]. The patients seen at EHC are predominantly uninsured or with medical coverage provided to people with a limited income or people over the age of 65 or with disabilities, such as Medicaid or Medicare, respectively. Several studies have noted how socioeconomic and racial disparities contribute to poorer outcomes [6][7][8][9]. Given the disproportionate burden of morbidity and mortality on communities with a significant proportion of racial and ethnic minority groups, studies that adequately represent these high risk populations are paramount.
In NYC, there was an enormous shift in delivery of cancer care during the height of the COVID outbreak. Non-emergent surgeries and some diagnostic procedures were postponed or redirected elsewhere, as operating rooms were converted to accommodate acutely ill patients. Radiation was transitioned to more short course/hypofractionated regimens and certain populations were encouraged to delay or even completely omit treatment [10,11]. On the other hand, radiation was also utilized for some patients that otherwise would have been managed surgically due to surgical staff reassignment and concern for intraoperative risk [12]. Chemotherapy was switched from intravenous infusions to oral chemotherapy regimens and immunotherapy was altered on a case-by-case basis [13,14]. Any indirect outcomes of these modifications will need to be evaluated in future studies. While the data are still evolving, cancer patients appear to be a vulnerable population to the SARS-CoV-2 virus [15][16][17][18][19][20]. There have not been many studies that have evaluated outcomes in patients with COVID and an active cancer diagnosis or recent treatment.
In the current study, the authors seek to determine COVID-related outcomes in patients with an active cancer diagnosis at Elmhurst Hospital Center (EHC) in the first wave of the pandemic. There have been studies evaluating outcomes in cancer patients in academic centers [2,[21][22][23][24], but the high community infection rates and diverse population in our public health system in NYC provide a unique window into the effects of the disease within one of the hardest-hit communities to date.

Inclusion
A retrospective analysis of patients at Elmhurst Hospital Center with active cancer between January 2020 and June 2020 was performed via chart review (BMJ, VWO, ZS). Inclusion criteria included: diagnosis via pathology report and/or receipt of oncologic treatment (including surgery, radiation therapy or systemic therapy) during the period of the COVID pandemic at Elmhurst Hospital, defined for these purposes as 1 January-30 June 2020. Patients with benign tumors were excluded. We obtained data from the electronic medical record (EMR) at EHC on demographic, clinical characteristics, cancer diagnosis, staging, pathology, treatment, COVID laboratory results, and outcomes. Patients without a COVID laboratory results available in the EMR were subsequently excluded from statistical analysis. This study was reviewed and approved by our institutional review board (IRB) review and informed consent was waived.

Definitions
Active cancer was defined as those with recent diagnosis of cancer (via confirmed pathology) or those undergoing active cancer-directed therapy, including radiation, surgery or chemotherapy within the data collection period (benign tumors excluded). COVID diagnosis was determined by laboratory evidence of SARS-CoV-2 using RT-PCR testing on nasopharyngeal or oropharyngeal swabs. COVID-confirmed mortality was defined as death within 30 days of confirmed laboratory evidence for SARS-CoV-2 using RT-PCR testing on nasopharyngeal swabs. Severe events were defined as patients with laboratoryconfirmed SARS-CoV-2 requiring hospitalization, oxygen supplementation, mechanical ventilation, or death within 30 days of diagnosis.

Statistics
Statistical analysis was performed using an R Studio. Demographics, clinical presentation, and outcomes were analyzed using descriptive statistics, Chi square, Fischer's exact and standard t-tests. The prevalence of COVID diagnosis and rates of severe events and death were the primary outcomes of this study. Predictors of incidence and prognostic factors for COVID mortality and severity were determined using univariate logistic regression.
Univariate logistic regression models were constructed, where the dependent variable was a binary outcome. Odds ratios and their associated 95% confidence intervals were constructed, where the null hypothesis was rejected for p < 0.05. Statistical analyses were conducted using R Studio Version 1.1.383; the "aod" and "ggplot2" packages were utilized. Multivariate analyses were not performed due to the low incidence of COVID-related severe events and mortality.
In the overall cohort who took a COVID test, 63 patients had surgery during this time period, 18 received radiation therapy, and 63 received cytotoxic chemotherapy. Intent of therapy was curative in 76 patients and palliative in 35 patients; 13 patients had a cancer diagnosis but had not received any cancer directed therapy at time of data collection. In the entire cohort of 266 active cancer patients, 177 patients did not have any delay, interruption or change in treatment, while 89 (33.5%) patients did experience a treatment modification. Sixty-one treatments were delayed, 38 had treatment interruption, 16 had treatment change. Four patients were transferred to a different facility for a portion of cancer care.
A majority of the COVID-positive patients had solid (80%) versus hematologic (20%) malignancies. Five out of sixteen (31%) patients with solid cancer died of COVID and one out of four (25%) patients with hematologic cancer died. There were 3 patients who had surgery, 5 who received radiation, and 15 received cytotoxic chemotherapy. Of the 6 COVID-confirmed deaths, 3 received radiation (50%), and 5 (66.7%) received cytotoxic chemotherapy. Refer to Table 1 for demographic and clinical factors. There were no significant associations between demographic and clinical factors with COVID diagnosis, severity, or mortality. Although, there was a trend towards statistical significance for patients with hypertension and COVID diagnosis (p = 0.05, see Table 2). Intent of therapy was curative in majority of COVID-positive (85%), severe-COVID (80%), and COVIDconfirmed deaths (83%).

Treatment and Mortality
There were 10 patients hospitalized for COVID with 1 patient requiring intensive care unit admission. Of the 20 patients with confirmed COVID, 6 were treated with medication (i.e., hydroxychloroquine, azithromycin), 12 with supportive care alone, and 8 required supplemental oxygen. The 30-day overall mortality rate for non-COVID patients was 6.6% and 30% for COVID patients. In the entire initial cohort (n = 266), there were 6 confirmed COVID-related deaths (2.3%). Of those who tested positive, 10 had severe events (50%) and 6 had confirmed COVID-related deaths (30%). There were no significant predictors of COVID diagnosis, severity, or mortality (see Table 3).

Discussion
Our single institutional experience shows the high case fatality rates in patients with active cancer treatment and COVID-19 diagnosis. Among the cancer patients with laboratoryconfirmed COVID, a third of patients died and half had severe events requiring admission, supplemental oxygen, and/or mechanical ventilation. This study contributes to the evolving literature on the impact of COVID in patients with an active cancer diagnosis. The strengths of this study include strict inclusion criteria of an active cancer diagnosis, a non-COVID comparison group, and a community setting. Moreover, this study provides COVID outcomes in the context of an urban, diverse, public health system in an epicenter of the first wave of the COVID-19 pandemic in the United States.
Interestingly, other tertiary care centers in NYC reported a lower case fatality rate (CFR) of 11-12% [2,[21][22][23][24]. The average COVID mortality for all cancer patients based on available data in Table 4 is 19.4% and 22.6% when including studies with hematologic cancers. The exact reasons for these differences in CFR are unclear, but it could reflect health care inequities that exist in NYC and throughout the world. Institutions with a lower CFR were more likely to be private academic centers that serve generally different patient populations than the New York Health and Hospital health system. Poor surge capacity and lack of resources (e.g., personal protective equipment, ventilators, nonrebreathers etc.) at EHC, particularly in March and April, may explain our high case-fatality rate. Notably, all the COVID-related deaths in our cohort occurred between 20 March-28 April 2020. It is possible that mortality rates have improved over time for subsequent waves of the pandemic due to inadequate resources, public awareness of COVID-19, widespread vaccination, and improved medical management of COVID-19. In a study by Chavez-MacGregor et al. [25], mortality and hospitalization rates of cancer patients with COVID decreased towards the end of 2020. A recent European registry analysis (On COVID) of 2634 patients with cancer diagnosed with COVID, also showed that mortality has improved in subsequent waves of the pandemic, which is reassuring and temporal trends in COVID outcomes should continue to be evaluated [26].  The disproportionate burden of morbidity and mortality from COVID in racial and ethnic minorities has been well-documented and should elicit further investigation [6][7][8][9]27,29,30,43,48]. According to the New York State Department of Health, approximately 34% and 28% of COVID fatalities are Hispanic or Black, despite comprising 29% and 22% of the population, respectively [49]. This is compounded with the known propensity to worse cancer outcomes based on race and insurance status, as well as underrepresentation of minorities in cancer clinical trials [50][51][52]. In this study, it is notable that 60% of severe COVID-related events and 83% of COVID deaths in our cohort were of Hispanic ethnicity, despite representing 54% of our entire population of active cancer patients and 42.6% of active cancer patients with a COVID laboratory test. Additionally, 88% of severe events and 83% of deaths had Medicaid insurance compared to 65% of the COVID-negative cohort. We did not find a statistically significant association with race, ethnicity, or insurance status with COVID incidence, severity, or mortality, which could be related to the small sample size.
In view of the high COVID-related mortality and severity rates in our cohort of patients with active cancer during the first wave of the pandemic in NYC, a recent cancer diagnosis or active cancer treatment is likely associated with worse outcomes. A majority of previous studies included patients with current or historical diagnosis of cancer, whereas our study and a few others have focused on patients with an active cancer diagnosis [20,25,27]. In one of the largest retrospective case-control analyses to date, Wang and colleagues [27] found an increased risk of COVID-19 infection, hospitalization, and death in patients with recent cancer diagnosis as compared to patients without cancer. Another cohort study found patients with recent cancer treatment (within 3 months of COVID diagnosis) and COVID had significantly higher risk of adverse outcomes than patients with no recent cancer treatment and in patients without cancer [25]. In fact, cancer patients without recent treatment had outcomes similar to patients without cancer [25]. Several other studies have previously noted receipt of anticancer therapy within 4 weeks of diagnosis of COVID was associated with adverse outcomes [20,30,37,53,54]. It is notable that patients who died from COVID in our cohort received anticancer therapy with the median time interval of 2 weeks of diagnosis (See Supplementary Table S1 for more details). There were no significant predictors for COVID diagnosis, severity, or mortality in our cohort of active cancer patients. However, there was a trend towards significance for hypertension (p = 0.05), which has been previously demonstrated in larger case series [20,39,55]. It is clear that patients with cancer represent a heterogeneous group and particular attention to COVID patients with recent diagnosis and treatment is warranted.
There are obvious limitations to this study as a small, single institution cohort, but this is important nonetheless given the continued outbreaks of COVID throughout the world and because of the relative lack of data that exists in regards to COVID outcomes in patients with active cancer diagnosis in racially and ethnically diverse patient populations. The high mortality rate among patients with confirmed COVID diagnosis may be in part due to limited testing of patients; only 42% of our initial cohort had a COVID test. Moreover, during the early portion of the time studying SARS-CoV-2 RT-PCR, testing was reserved for patients with moderate to severe symptoms, which may contribute to the elevated case fatality rate in our population. Therefore, the elevated mortality rate found in this study may be in part due to preferential testing in patients sufficiently ill to require SARS-CoV-2 testing. It is also possible that data collection for testing and hospitalization did not capture all patients, as many were redirected to other hospitals during the height of the pandemic or were tested in outside clinics. Long-term oncologic outcomes will also be important to evaluate in this population with comparison to patients from previous years. Future studies should be dedicated to evaluating both COVID-related and oncologic long-term outcomes in these vulnerable patients, especially in the context of profound changes in cancer care delivery during the pandemic [56,57].
EHC serves an incredibly diverse and predominantly poor, uninsured, underserved population that faces a multitude of economic, cultural, and language barriers. In diverse and historically underserved communities such as Queens, the first wave of the pandemic exposed the vulnerability of these communities and the healthcare inequities that persist in the United States. In view of new highly infectious variants associated with lower vaccine effectiveness, such as B.1.617.2 (delta) and B.1.1.529 (omicron), COVID continues to pose a significant threat to vulnerable populations [58,59]. Resource allocation for public and safety-net hospitals and public health policies to mitigate disparate COVID-related outcomes in underserved communities is necessary.

Conclusions
In this retrospective cohort study of a diverse community hospital during the first wave of the pandemic, we found that patients with an active cancer diagnosis and COVID had high rates of severe events (50%) and mortality (30%). A disproportionate number of active cancer patients with Medicaid insurance, Hispanic ethnicity, other race, and male sex had adverse COVID-related complications or death. There were no significant predictors of COVID diagnosis, severity, or mortality.

Informed Consent Statement:
In accordance with Mount Sinai's Federal Wide Assurances (FWA#00005656, FWA#0000565), this study was deemed exempt by IRB due to no or minimum risk to subjects and retrospective nature of this study.

Data Availability Statement:
Data are available on request due to restrictions (i.e., HIPPA).

Conflicts of Interest:
The authors declare no conflict of interest.