This case series describes 200 patients with laboratory-confirmed COVID-19 infection who died between March 17 and April 16, 2020. No patients had history of international travel, and the majority of patients had no known exposure to COVID-19 positive contacts, making exposure most likely to be community-based. The median age in our case series was 73 years (IQR 62–82), consistent with reports that older persons are at higher risk of dying from COVID-19 infection.10,14 In our population, 58·5% of patients were male, consistent with other reports that males have higher risk of death than females as they may manifest more severe disease.15
In our case series, 178 out of 200 (89%) patients were Black Americans. The COVID-19 pandemic has highlighted the disparities pervasive in our healthcare system. Data from the Centers for Disease Control and Prevention describe disproportionate burden of COVID-19 illness and death among racial minorities.16 In NYC, the COVID-10 fatality rate in Black Americans is double that of white patients.8 The excessive burden of COVID-19 on racial and ethnic minorities is multifactorial and includes, but is not limited to, vast inequalities in socioeconomic status, education, physical environment, social support networks, and access to health care.
More than two-thirds of our patients were residents of the East Flatbush neighborhood of Brooklyn. East Flatbush, along with the surrounding neighborhoods also represented in this study, is a predominantly Black and foreign-born community. 17 Residents of these communities are more likely to live in multigenerational homes, commute using public transportation, and are less likely to receive paid sick leave – all factors which may contribute to increased rates of SARS-CoV-2 transmission. 18 The median household income for East Flatbush is $48,000, compared to $77,000 for the borough of Manhattan, with 42% of East Flatbush residents making less than $25,000 per year. This financial burden translates to limited access to health care and poorer overall health. East Flatbush ranks 12th out of 59 NYC neighborhoods for percentage of adults without health insurance (15%).18 The COVID-19 pandemic has demonstrated that, during times of crises, health disparities are exacerbated and serve to weaken already disadvantaged and vulnerable populations.
More than half of our patients had three or more comorbidities, with hypertension, diabetes, and obesity being the most common. This is consistent with previous data showing that Black Americans experience higher rates of comorbidities compared to white Americans.19 Only 33 patients in our study had a history of asthma and chronic obstructive pulmonary disease, and similar underrepresentation of both diseases has been reported elsewhere. Several factors have been postulated,20 including poor recognition of chronic respiratory disease due to overlapping of symptoms with COVID-19 and potential protection against COVID-19 by chronic respiratory disease. In addition, therapies used by these patients may reduce risk of infection or development of symptoms leading to diagnosis of COVID-19.
Dyspnea was the most common presenting symptom (92·5%) – a prevalence not seen in recent data from the US and China.10,21,22 This likely highlights the level of severity of illness in our patients at presentation. Similar to other studies, more than 50% of our patients had cough and fever.10,14 Interestingly, approximately 10% of our patients presented in diabetic ketoacidosis. It is unclear if this is directly related to pathogenesis of COVID-19 or non-adherence to medications: about 50% of our patients had more than 4 days of COVID-19 symptoms prior to admission, and it is possible that patients, while sick at home, may not have continued their maintenance medications.
AKI is well-documented in viral infections, such as severe acute respiratory syndrome (SARS) and Middle East respiratory syndrome-related coronavirus (MERS-CoV), with estimates ranging between 5–15% of cases, and AKI and has been identified as a risk factor for increased mortality.23 Available data from China24,25 did not identify high rates of AKI, and only small portions of patients in these studies expired. The high rate of AKI (80%) in our population may implicate AKI as a potential indicator for the severity of COVID-19 infection.
Cytokine release syndrome, a state characterized by worsening respiratory failure and markedly elevated inflammatory markers such as CRP, LDH, and ferritin, is a frequent consequence of COVID-19 infection, similar to SARS and MERS-CoV.26 Our findings of elevated CRP, LDH, and ferritin in almost all of our patients at baseline indicate the severity of their presentation, and raise a consideration of these laboratory values as markers of prognosis in COVID-19. CRS often results in mechanical ventilation, and 70 of our 200 patients received mechanical ventilation during their admission, nearly half of whom were intubated on presentation. CRS is thought to peak approximately 8 days after symptom onset.26 Our patients had a median duration of symptoms – from time to symptom onset to expiration – of 8·42 days (IQR 5·57–12·72), and inflammatory markers indicative of CRS were elevated in all patients with available data at expiration. The role of CRS management in COVID-19 remains unclear.
Severe COVID-19 infection has been associated with increased risk of thromboembolic events. Data show an estimated 16·7%27 to 25%28 of patients experience venous thromboembolism during their admission. D-dimer and fibrinogen were elevated in 74·6% and 93·8% of our patients with available data, consistent with reports from France.27 Most of our patients received thromboprophylaxis, and 11 out of 200 patients received therapeutic anticoagulation for new-onset thromboembolic events. Further data are needed to identify the benefit of full-dose anticoagulation therapy in patients with COVID-19 compared to standard thromboprophylaxis.
Only one of our patients received prone positioning, an intervention shown to reduce mortality if initiated early in patients with severe acute respiratory distress syndrome.29 At the time of data collection, prone positioning was infrequently performed at UHB, and teams have since been established to coordinate prone positioning for COVID-19 patients in severe respiratory distress.
As of late April 2020, there are no approved agents for the treatment of COVID-19 infection. Supportive management remains the standard of care, as in other viral pneumonias. Based on preliminary investigational data30,31 showing potential benefit, two-thirds of our patient population received hydroxychloroquine during their stay. Few of our patients received corticosteroids, likely given the inconclusive data to support their use in COVID-19 available during the study period. Additionally, a small group of patients received tocilizumab, an interleukin-6 receptor antagonist, which has been considered in COVID-19 patients in an effort to reduce the inflammation associated with CRS. The true benefit of these agents has yet to be determined.
In non-COVID-19 related therapy, 55·5% of patients received 72 hours or more of antibiotic therapy. Very few patients were found to have culture-confirmed infectious processes other than COVID-19 during their stay, and even fewer had positive respiratory cultures. It remains unknown the exact rate of bacterial coinfection in patients with COVID-19.
Our study is not without limitations. This descriptive observational study was retrospective in nature and lacked a control group. In an effort to reduce selection bias, patients underwent consecutive enrollment, and were only included in analysis if they had laboratory-confirmed diagnosis of COVID-19. Also, inconsistent collection of laboratory values such as CRP, LDH, ferritin, D-dimer, and fibrinogen limited the number of patients in whom these values could be evaluated. Since the data collection period, these values have become part of the standard COVID-19 workup at UHB, both at baseline and trended throughout admission. Finally, lack of robust data such as randomized trials on the novel coronavirus presents difficulty in comparing our observations with those of others, particularly given differences between our patient population and that of other countries reporting on COVID-19.
To our knowledge, this is the largest description of patients who died due to COVID-19 at this time. UHB serves a majority Black American population, with a large portion of patients of Caribbean descent, who have multiple comorbidities. Considering the racial, economic, and healthcare disparities of Brooklyn, multiple systematically-based factors – modifiable and non-modifiable – may contribute to our patients’ baseline characteristics and severity of illness on presentation, which ultimately resulted in their mortality.
Available literature on COVID-19 continues to evolve. Description of the demographics and presentation characteristics of patients who have expired due to COVID-19 identifies populations at high risk of severe illness and ultimate mortality, and assists providers in triaging and caring for COVID-19 positive patients.