Characteristics and Palliative Care Needs of COVID-19 Patients Receiving Comfort-Directed Care.

The first known coronavirus disease 2019 (COVID‐19) related hospitalization in New York City was reported at Columbia University Irving Medical Center/NewYork‐Presbyterian Hospital. Since then, the rapid increase in the number of patients with COVID‐19 associated with acute respiratory distress syndrome (ARDS) and high rates of mortality2, 3 has highlighted the critical need for high‐quality end‐of‐life care. On March 31, 2020, an eight‐bed Palliative Care Unit (PCU) was established at our institution for patients with COVID‐19 whose surrogates opted to not initiate or continue life‐sustaining therapies. To our knowledge, this is the first report describing COVID‐19 patients receiving comfort‐directed care.

This case series aims to describe the characteristics and palliative care needs in patients admitted to the PCU at Columbia University Irving Medical Center/NewYork‐Presbyterian Hospital to inform other clinicians caring for this population at the end of life.

METHODS

Deceased patients with confirmed severe acute respiratory syndrome coronavirus 2 infection by polymerase chain reaction testing of a nasopharyngeal sample, admitted to the PCU at Columbia University Irving Medical Center/NewYork‐Presbyterian Hospital between March 31, 2020, and April 10, 2020, were included.

Before data collection, a waiver was obtained from the Columbia University institutional review board. Deidentified patient data were collected from the electronic medical record Epic Hyperspace and analyzed using Microsoft Excel. Laboratory testing was reviewed at PCU admission. Patient outcome data were evaluated at time of death. Due to the descriptive nature of this case series, no analysis for statistical significance was performed.

RESULTS

A total of 30 patients were included in this case series (mean age = 84.5 years; 53% male) (Table 1). Most patients were of Hispanic origin (20 [66.7%]), followed by white (4 [13.3%]). All 30 patients had comorbidities before hospital admission, with 70% of patients having more than one comorbidity. Twenty‐four patients (80%) had metabolic abnormalities, with hypernatremia observed in 17 patients (57%). Before PCU admission, all 30 patients developed ARDS, with 29 (97%) requiring supplemental oxygen.

Table 1

Baseline Characteristics of 30 Patients With COVID‐19 at Presentation to the PCU

Baseline Characteristics	No. (%) of Patients a
Demographics
Age, mean (range), y	84.5 (71‐97)
Sex
Male	14 (46.7)
Female	16 (53.3)
Ethnicity/race
Hispanic, Latino, or Spanish origin	20 (66.7)
White	4 (13.3)
Black or African American	2 (6.7)
Declined to answer	2 (6.7)
Preadmission comorbidities
Dementia	14 (46.7)
Asthma	2 (6.7)
Coronary artery disease	8 (26.7)
Hypertension	24 (80.0)
Chronic obstructive pulmonary disease	5 (16.7)
Congestive heart failure	4 (13.3)
Diabetes	15 (50.0)
Cancer	7 (23.3)
Prostate	4 (13.3)
Breast	1 (3.3)
Bladder	1 (3.3)
Lung/breast/ovarian	1 (3.3)
Obstructive sleep apnea	0 (0.0)
Chronic kidney disease	12 (40.0)
End‐stage kidney disease	1 (3.3)
History of solid organ transplant	0 (0.0)
Immunosuppression b	2 (6.7)
Cirrhosis	0 (0.0)
Total with >1 comorbidity	21 (70.0)
Admission symptoms
Cough	17 (56.7)
Shortness of breath	24 (80.0)
Fever c	15 (50.0)
Admission laboratory measures
Hypernatremia	17 (56.7)
Other metabolic abnormalities	21 (70.0)
Bacterial coinfection d	5 (16.7)
COVID‐19 manifestation before PCU admission
Cardiomyopathy	1 (3.3)
Acute kidney injury	8 (26.7)
Acute respiratory distress syndrome e	30 (100.0)
Use of nasal cannula	19 (63.3)
Use of nonrebreather	29 (96.7)
Use of high‐flow oxygen therapy >15 L/min	0 (0.0)
Use of noninvasive positive pressure ventilation	0 (0.0)
Received mechanical ventilation f	0 (0.0)
Use of antibiotics	18 (60.0)
Use of vasopressors	3 (10.0)

Abbreviations: COVID‐19, coronavirus disease 2019; PCU, Palliative Care Unit.

Unless otherwise indicated.

Defined as outpatient prescription of greater than 10 mg/d of prednisone or an equivalent, use of chemotherapy, or use of nonsteroidal immunosuppressive agents for solid organ transplant or for an autoimmune disease.

Defined as a temperature of greater than 100°F.

Two patients had positive blood cultures and three patients had positive urine cultures.

Definition and severity according to the Berlin Criteria.

Patients on mechanical ventilation who were palliatively extubated were not transferred to the PCU, and therefore, not included in this study.

On admission to the PCU, the most common symptom observed was dyspnea (30 [100%]), followed by delirium (22 [73%]), pain (10 [33%]), and anxiety (10 [33%]) (Table 2). Intravenous morphine (23 [77%]) and hydromorphone (11 [37%]) were the most commonly used medications. A total of 62 visits and calls were made by chaplains and social workers to provide spiritual and psychosocial support (eg, offering end‐of‐life prayers to patients and assisting family members with funeral planning).

Table 2

End‐of‐Life Care Needs and Time Course

Variable	No. (%) of Patients a
Symptoms observed
Dyspnea	30 (100.0)
Delirium	22 (73.3)
Pain	10 (33.3)
Anxiety	10 (33.3)
Nausea/vomiting	4 (13.3)
Myoclonus	4 (13.3)
Medications used
Morphine	23 (76.7)
Hydromorphone	11 (36.7)
Haloperidol	11 (36.7)
Lorazepam	21 (70.0)
Morphine IV equivalent infusion rate, median (range), mg/h	2 (1‐6)
Morphine IV equivalent bolus, median (range), mg	3.3 (1.3‐10)
Support provided
From social worker	24 (80.0)
Total No. of calls/visits made	51
From chaplaincy	11 (36.7)
No. of calls/visits made	11
From family/friends visiting	5 (16.7)
End‐of‐life decision making and time to death
ED to DNR/DNI order, mean (range), h	20.0 (0‐163)
DNR/DNI to death, mean (range), h	100.5 (21‐219)
PCU admission to death, mean (range), h	34.6 (3‐104)

Abbreviations: DNI, do not intubate; DNR, do not resuscitate; ED, emergency department; IV, intravenous; PCU, Palliative Care Unit.

Unless otherwise indicated.

The average length of stay in the PCU was 34.6 hours or 1.4 days.

DISCUSSION

We report the characteristics and palliative care needs of patients with severe COVID‐19 infection who have forgone life‐sustaining treatments and received comfort‐directed care.

Consistent with other studies,4, 5 we observed an older age group with high rates of comorbidities. Given the high proportion of patients with metabolic abnormalities on hospital admission, further study is needed to explore the potential association between severity of metabolic disarray and its impact on patient outcomes.

Dyspnea and delirium were the most commonly observed symptoms in dying patients with COVID‐19. Relatively low doses of morphine, hydromorphone, and lorazepam were needed for symptom control. The present study also highlights the crucial role of social workers and chaplains in providing psychosocial and spiritual support to patients and families, especially given the limited degree of contact most family members had with their loved one.

The limitations of this study include the small sample size from one hospital center, ethnic and racial makeup of the population given the location in New York City, as well as possible selection bias by admitting moribund patients who are more imminently dying due to limited bed availability in the PCU. Nonetheless, this study can be instructive to other institutions to understand and prepare for the palliative care needs in patients dying from COVID‐19.