The direct and indirect effects of COVID‐19 pandemic in a real‐life hematological setting

Abstract Background Clinical outcomes of novel coronavirus 2019 disease (COVID‐19) in onco‐hematological patients are unknown. When compared to non‐immunocompromised patients, onco‐hematological patients seem to have higher mortality rates. Aims We describe the characteristics and outcomes of a consecutive cohort of 24 onco‐hematological patients with COVID‐19 during the first month of the pandemic. We also describe variations in healthcare resource utilization within our hematology department. Methods and Results Data from patients between the first month of the pandemic were retrospectively collected. Clinical and logistic data were also collected and compared with the average values from the prior 3 months of activity. Prevalence of COVID‐19 in our hematological population was 0.4%. Baseline characteristics were as follows: male sex: 83%, lymphoid diseases: 46%, median age: 69 (22‐82) years. Median follow‐up in survivors was 14 (9‐28) days and inpatient mortality rate was 46%. Average time to moderate/severe respiratory insufficiency and death were 3 (1‐10) and 10 (3‐18) days, respectively. Only 1 out of every 12 patients who developed moderate to severe respiratory insufficiency recovered. Upon univariate analysis, the following factors were associated with higher mortality: age ≥ 70 years (P = .01) and D‐dimer ≥900 mcg/L (P = .04). With respect to indirect effects during the COVID‐19 pandemic, and when compared with the prior 3 months of activity, inpatient mortality (excluding patients with COVID‐19 included in the study) increased by 56%. This was associated with a more frequent use of vasoactive drugs (+300%) and advanced respiratory support (+133%) in the hematology ward. In the outpatient setting, there was a reduction in initial visits (−55%) and chemotherapy sessions (−19%). A significant increase in phone visits was reported (+581%). Conclusion COVID‐19 pandemic is associated with elevated mortality in hematological patients. Negative indirect effects are also evident within this setting.

Conclusion: COVID-19 pandemic is associated with elevated mortality in hematological patients. Negative indirect effects are also evident within this setting.

| INTRODUCTION
The novel coronavirus SARS-CoV2 recently emerged as a global threat. 1 High hospitalization (around 20% of diagnosed cases) and mortality rates (2%-3%) of patients with COVID-19 have led to an unprecedented burden on healthcare systems. 2,3 Indeed, the impact of COVID-19 has even prevented healthcare professionals from continuing normal utilization practices of healthcare resources for all other health matters unrelated to the pandemic. Evidence of such can be seen in emergency departments and other vital services such as onco-hematology departments. 4,5 Recent reports from Chinese colleagues showed how symptomatic COVID-19 is perhaps associated with higher mortality in the oncological population, with a range of 5% to 35%. [6][7][8] In our study, we describe how this pandemic directly affects our hematological patients. Furthermore, we analyze the indirect effects of such pandemic on our hematology department in terms of inpatient and outpatient activity of a population without COVID-19 infection. With these data, we aim to further our understanding of the direct and indirect impacts of COVID-19 on oncohematological patients and departments to facilitate the conception of effective contingency plans for future viral outbreaks.

| Data sources and statistics
The Institut Català d'Oncologia-Hospital Duran i Reynals (Barcelona, Spain) is a tertiary oncology referral center for adult patients, with a hospitalization ward of 26 beds dedicated to hematologic patients, attending to a population of almost 1.5 million people. In the first part of the study, we analyzed the direct effects of COVID-19 on a consecutive series of onco-hematological patients. We retrospectively collected clinical data from patients between March 13, 2020 (the first day of government-imposed restrictions related to COVID-19 in Spain) and April 12, 2020. Data cut-off date was April 19, 2020. Data were collected via a retrospective chart review. All patients were ≥18 years old and had an onco-hematological malignancy. COVID-19 diagnosis and response were confirmed in accordance with World Health Organization criteria. 9 Hematological disease type and status were defined per international standard criteria. 10,11 In the second part of the study, we described the indirect effects of COVID-19 on inpatient and outpatient activity in our department. Clinical and logistic data from both types of activity were compared with average values calculated from the prior 3 months (between December 13, 2019, and March 12, 2020).
The primary endpoint of the study was to describe the all-cause in-hospital mortality rate in patients with hematologic neoplasms and a confirmed diagnosis of COVID19 infection, defined as the proportion of patients with COVID-19 who died to the overall cohort. Secondary endpoints included: description of patient characteristics, average time from diagnosis to moderate/severe respiratory insufficiency (defined as requirement of oxygen support other than nasal cannula, 3 L/m or pO 2 /FiO 2 < 300 mmHg), and average time from diagnosis to death. Descriptive data were reported as counts and percentages. The proportion of deaths was compared among baseline characteristics using chi-squared test or Fisher's exact tests depending on the number of events. The following variables were considered:

| Structural reorganization due to COVID-19
Our institution is a tertiary oncology referral hospital for adult patients, comprising four clinical departments (Medical Hematology, Medical Oncology, Palliative Care, and Radiotherapy) and two services (Radiology and Pharmacy). Each department has its own ward and outpatient visitor area; however, outpatient therapy space is shared.
Furthermore, the hospital has its own infectious disease specialists, and endocrinology and nutrition staff. For all other services, including the intensive care unit, the hospital collaborates with Hospital Universitari de Bellvitge (Barcelona, Spain). Due to the COVID-19 pandemic, our center and activities underwent structural reorganization per regional, national, and international recommendations. 13 Palliative care and the emergency units of the hospital were converted into COVID-19 wards. In order to prevent nosocomial COVID19 outbreaks in a high-risk ward, nasopharyngeal swabs for COVID19 were performed in all patients 24 to 48 hours before their scheduled admission. Due to the restrictive triage protocol, no change in room pressure was implemented. No family members were authorized to stay with patients during their hospital stay. Additionally, physicians were separated into two different teams for each ward service (HCT, myeloid disease, and lymphoid disease) to minimize the loss of workforce due to a possible contagion. Outpatient staff was physically confined to specific areas. Teleworking was encouraged for all activities not requiring physical presence at the hospital. Prior agreement with the ICU to have a referral intensive care specialist in the hematology ward was made. This strategy was implemented with the objective to delay transfers of critical patients and assume more clinically complex cases, defined as those patients who require vasoactive drug support and/or present with rapid acute respiratory insufficiency and need noninvasive ventilation with high-flow oxygen therapy.
Patients admitted to the onco-hematological ward had the same pharmacological treatments available as those admitted to the ICU, including tocilizumab and high-dose corticosteroids.

| Cohort description of patients with COVID-19
Taking into consideration patients followed-up in the last 2 years in our hematology department (n = 6779), the expected prevalence of COVID-19 in our oncological population was 0.4%. This figure is similar to that reported in the general population of the same geographic area (0.5%). 16 Of the 25 patients identified, one was excluded from survival analysis due to a non-oncological disease (sickle cell disease).
Upon analysis, only one patient was still hospitalized. Details concerning patient characteristics and baseline conditions, COVID-19 disease presentation, treatment, and outcomes can be found in Table 1.
Median age of patients was 69 years ; the majority of patients were male (83%); and lymphoid malignancies were the most common hematological disease (46%). Only two patients were neutropenic upon diagnosis. Median follow-up in survivors was 14 days (7-29).
Mortality rate was 46%, higher than in the general population (10%). 16 Eight of 24 patients (33%) required advanced respiratory support (more than nasal cannula 3 L/min). Average time from diagnosis to moderate to severe respiratory insufficiency was 3 days (1-10).
One patient required endotracheal intubation and ICU admission; two received non-invasive mechanical ventilation in the hematology ward and emergency department, respectively. Five patients received high flux nasal cannula in the ward.

| Indirect effects of COVID-19 on hematological outpatient and inpatient activity
During the first month of pandemic restrictions, monthly inpatient activity of non-COVID-19 patients was affected ( Table 2). The total number of admissions dropped by 35% and mortality rate of these patients increased by 56% (Figure 1). We observed an increased number of intensive care procedures performed in the hematology ward (use of vasoactive drugs +300%). Critically ill patients received on-site sub-intensive care. Finally, the number of autologous and allogeneic period. The total number of first visits dropped by 55% (Figure 2). In   studies. Considering that the median age of our study population was 69 years, this could have likely contributed to poor survival. 17 Additionally, the presence of comorbidities is another well-known risk factor. 18 The majority of our patients had at least one chronic cardiorespiratory disease. Indeed, cancer also represents a severe comorbidity due to the disease itself, as well as the toxic effects of oncological treatments. Other reports described a higher COVID-19-related mortality in this population, ranging from 20% to 40%. [6][7][8]19 In a recent large study, hematological malignancy was associated with a higher risk of severe events in a COVID-19 population. 19 The use of different types of therapy could also influence this Nonetheless, such a high mortality rate appears superior to that of any other common seasonal respiratory virus (eg, influenza viruses, RSV), which usually does not exceed 10% to 20%, including in severely immunocompromised patients. 20,21 This may reflect the higher virulence of SARS-CoV-2 as observed in the general population.
We believe that the high mortality rate cannot be attributed exclusively to the COVID19, especially considering that 20/24 patients (83%) had an active underlying hematological neoplasm.
However, in all patients who died, the leading cause of death was respiratory insufficiency, which makes us believe that COVID-19 had a critical role on the poor patients' outcomes.
Our study also confirms elevated age and D-dimer levels as risk factors associated with higher mortality. 17 Female sex was not reported as a risk factor in other studies. However, these results were obtained from a small cohort of patients and should be confirmed in larger series. Apart from patient and virus-related risk factors, it is possible that logistic factors could also play a role in COVID-19 mortality. Even if this does not stand to be our case, patients requiring ICU admission were prioritized according to ethical regional or national guidelines. We can speculate that this prioritization might have disadvantaged onco-hematological patients worldwide.
Finally, we suggest that onco-hematological patients could deteriorate faster than the general population. This is true in terms of time from diagnosis to respiratory deterioration and death. A similar observation was reported elsewhere. 8 Table 2). The drop was especially significant in the myeloid division. It is possible that this effect could be related to societal pressure to avoid hospital visits during this period and to fear that infection may be more probable in such an environment. Indeed, hospitals are considered high-risk spots for contagion. We hypothesize that due to more unspecific symptoms, peo-

ACKNOWLEDGMENTS
All patients and their families. All health care workers and scientists helping our society against the COVID-19 outbreak. We thank CERCA programme/Generalitat de Catalunya for institutional support.