Abstract
Background:
The COVID-19 pandemic has spurred unprecedented need for specialty palliative care. The Palliative Care Quality Collaborative (PCQC) provides unique infrastructure for rapid data collection and analysis.
Objectives:
To capture and describe real-time, real-world experiences of specialty palliative care professionals caring for patients with COVID-19 through a rapid reporting tool and registry.
Setting:
Palliative care clinicians consulted for patients either positive for COVID-19, under investigation for COVID-19, or recovered from COVID-19.
Design:
The PCQC created a 13-item COVID-19 case report form (CRF), modeled after the PCQC core dataset for specialty palliative care quality measurement. Twelve items offered discrete answer choices and one was open-ended. The CRF was publicized widely (e.g., social media, e-mail list serves) and completed through a link on the PCQC website.
Results:
Three hundred six reports (298 adult, 8 pediatric) were submitted between April 6, 2020 and October 7, 2020. The majority of patients (83%) were 50 years or older; 25% were 80 or older, and 78% were COVID-19 positive. Male gender identity was significantly more prevalent than female (58% vs. 40%, p < 0.002). The most common comorbidity was cardiovascular disease (23%). Of adult hospital-based patients, 69% were full code before palliative care consultation versus 28% after (p < 0.05). All pediatric patients were full code before and after palliative care consult. Qualitative themes were strained communication with patients, family visitation challenges, communication barriers between clinicians and families, rapid changes in palliative care medical management, community care options difficult to find, lack of testing in community-based settings, and guardianship and legal challenges.
Conclusion:
Preliminary data from the first 306 patients reported to the PCQC COVID-19 Registry describe palliative care use concentrated among older and higher risk patients and challenges to the provision of palliative care during this pandemic.
Introduction
The serious and uncertain nature of the severe acute respiratory syndrome (SARS-CoV-2) coronavirus (COVID-19) pandemic has spurred an unprecedented involvement of specialty palliative care in routine care. This increased integration has spanned the trajectory of viral illness, including patients at risk, suffering from, or in recovery from the novel coronavirus. Inherent to palliative care delivery is a balance between practice guided by a core set of standards and quality measures, 1 and care that is tailored to the needs of individual patients and families and responsive to the specific requests of referring clinicians. 2 To date, our summative understanding of the varied experiences in delivering palliative care to COVID-19 patients remains limited.3–5
Describing the quality of care delivery during routine clinical practice is the mission of a specialty-level quality registry. Recently, the Palliative Care Quality Collaborative (PCQC) was created as the first coordinated organization- and patient-level quality registry for specialty palliative care. 6 The PCQC integrated the efforts of several national palliative care Registries, including the National Palliative Care Registry (NPCR), the Palliative Care Quality Network (PCQN), and the Global Palliative Care Quality Alliance (GPCQA). Guided by the American Academy of Hospice and Palliative Medicine (AAHPM) with support from the Gordon and Betty Moore Foundation and Cambia Health Foundation, the PCQC was created to be the unified “front door” for collaborative quality measurement and improvement activities in the field.
To capture real-time, real-world experiences of specialty palliative care professionals in caring for patients with COVID-19, the PCQC rapidly created a case report form (CRF) and registry dedicated to novel coronavirus experiences. In this study, we report and describe the first 306 cases reported during the first six months of data collection.
Materials and Methods
Overall
We collected de-identified, clinician-reported COVID-19 cases starting April 6, 2020 and herein report the consecutive cases submitted through October 7, 2020. The study was deemed exempt by the Duke Health Institutional Review Board.
Case report design and distribution
We created a COVID-19 CRF modeled after the PCQC core dataset for specialty palliative care quality measurement (www.palliativequality.org). The CRF requests manually entered demographic and clinical information from interprofessional specialty palliative care clinicians consulted for the management of adult or pediatric patients with COVID-19, persons under investigation for COVID-19, or patients who have resolved COVID-19 infection. We revised the CRF based on iterative feedback relating to clarity and comprehensiveness provided by palliative care physicians on the PCQC Board of Directors alongside PCQC staff. The final version consists of 12 quantitative questions with discrete answer choices, and 1 qualitative open-ended question regarding lessons learned, challenges faced, and barriers encountered in providing palliative care to the patient. Respondents accessed the secure registry (ArborMetrix, Ann Arbor, MI) through an open link on the PCQC website. For inpatient care, respondents reflected on an entire episode of care, which may comprise several clinical visits from admission through discharge or death. For nonhospital reports, respondents reported on individual ambulatory visits. The CRF was launched for use by any specialty palliative care professional without cost on April 6, 2020.
Invitations to specialty palliative care teams to participate in this registry were distributed through multiple networks, including the membership electronic mailing lists of the PCQC, AAHPM, PCQN, GPCQA, and AAP (American Academy of Pediatrics), and advertised on social media platforms. Membership in the PCQC was not a prerequisite for a site to upload data. The Center to Advance Palliative Care (CAPC) and National Palliative Care Research Center (NPCRC) invited respondents through Twitter and e-mail list serves.
Statistical analysis
We consecutively analyzed all submitted case reports with at least one data field completed from April 6, 2020 through October 7, 2020. Descriptive statistics, two-tailed chi square, and simple thematic induction of qualitative responses are presented (SAS, Cary, NC).
Qualitative analysis
Qualitative data were collected through an open-ended question that asked respondents “What challenges, lessons learned, or ethical barriers did you encounter in caring for this patient, and did you use telehealth/remote communications with this patient or family/caregiver?” Respondents were asked not to include any patient identifiable information in their response. Qualitative data were analyzed by two team members (R.T. and K.M.) using a manual process inductive coding and a flat coding frame. 7 Codebooks were independently developed, then resultant themes compared and contrasted, and differences were discussed in an iterative manner until consensus was reached regarding final themes and exemplar quotes.
Results
Demographic and clinical characteristics
Patient characteristics are presented in Tables 1 and 2. We analyzed 306 CRFs of which 299 were submitted from the United States, 2 from international sites, and 5 with unknown locations. One hundred seventy (55.6%) cases originated from the U.S. South, 60 (19.6%) from the U.S. Midwest, 43 (14%) from the U.S. Northeast, and 26 (8.5%) from the U.S. West. The most common administrative home of palliative care teams was health system (51.6%) and hospital (42.8%). Among patients, 298 were adults and 8 were children or infants. The vast majority of patients were aged 50 years or older (82%), with 25% aged 80 years or older. Most (78.1%) reported cases were for confirmed COVID-19 cases. Men were significantly more prevalent than women (p < 0.002). The most common underlying serious illness was cardiovascular disease (23.1%), followed by metabolic/endocrine (11.1%), cancer solid tumor (9.8%), no illness/previously well (9.8%), dementia (9.2%), pulmonary (7.8%), and renal disease (7.8%). At the time of consultation for adult patients, 69% were full code versus 28% after the consultation (p = <0.05). All pediatric patients were full code before and after palliative care consult. Among cases where a survival outcome is known, approximately half (50.7%) of patients in the overall cohort who received a consultation died.
Demographic and Clinical Characteristics: Adult Patients
AND, allow natural death; CKD, chronic kidney disease; DM, diabetes mellitus; DNAR, do not attempt resuscitation; DNI, no not intubate; DNR, do-not-resuscitate order; PUD, peptic ulcer disease.
Demographic and Clinical Characteristics: Pediatric Patients
Palliative care consultation characteristics
Consultation characteristics for adult patients are presented in Table 3 (N = 298). In adults, almost all consultations (97.3%) were received from inpatient specialty medical services, most commonly Critical Care (65.4%) or Hospital Medicine (25.2%). The two most common reasons for consultation, for adult patients, were “Decision Making including goals of care” (34.0%) and “Providing Support to Patient and Family” (25.2%).
Consultation Characteristics Adult Patients
APN, advance practice nurse; CBPC, community based palliative care; ESRD, end stage renal disease; GOC, goals of care; LPN, license practical nurse; MICU, medical intensive care unit; OT, occupational therapist; PA, physician assistant; PICU, pediatric intensive care unit; PT, physical therapist; RN, registered nurse; SICU, surgical intensive care unit; TICU, trauma intensive care unit.
Regarding palliative care disciplines involved, in adults ∼80% of consultations included a physician. Around one-fourth of consultations included an advanced practice nurse (27.2%) or social worker (25.8%) and 38.2% included a chaplain. For at least 60% of the consultations with adult patients, palliative care teams (1) assisted with discussing goals of care (GOC) with the health care proxy or surrogate decision maker, (2) supported shared decision making about use of life-sustaining treatments, or (3) provided support to the patient and family. For adult patients, palliative care teams also assisted with symptom management of shortness of breath (19.1%), anxiety (8.4%), and pain (7.4%).
Eight pediatric cases were reported, three of which included qualitative data (Table 4). Seven cases were from the Southern United States and one was from the U.S. West. All pediatric consults (100%) included providing support to the patient and family, 37.5% were to provide support to staff or colleagues, and 25% were to support decision making, including GOC. Two (25%) of the eight children died. Qualitative responses revealed that one was a premature infant delivered emergently when the mother suffered an apparent code event. The mother was under suspicion for COVID-19, but was ultimately found to be negative. The infant was pronounced dead after neonatal resuscitation; the mother ultimately also died. Memory making materials were completed (handprints, photographs) with bedside caregivers wearing appropriate personal protective equipment (PPE). No family was allowed to visit due to COVID-19. The other qualitative comments regarding children indicated that one was tested following a potential exposure by a health care worker and the other patient was transferred for cardiac transplant evaluation elsewhere.
Consultation Characteristics Pediatric Patients
Themes from open-ended responses
We asked respondents: “What challenges, lessons learned, or ethical barriers did you encounter in caring for this patient, and did you use telehealth/remote communications with this patient or family/caregiver?” There were 50 unique responses to this prompt. Most of the responses were focused on challenges. Seven emergent themes related to clinical challenges and associated exemplar quotes are presented.
Strained communication with patients
Respondents shared that the pandemic has created challenges for both family and clinicians in communicating with patients. With the addition of PPE, respondents noted that it is difficult to express compassion while your face is covered. It can be logistically difficult to have GOC conversations through PPE with hard-of-hearing patients. Although telemedicine can be helpful, challenges persist. Others shared that patients may become too disoriented to communicate over the phone or video, so providers should not delay arranging video communication opportunities (provider-patient and patient-family).
“Very difficult to have thorough GOC conversation with [patient] through PPE (PAPR), as [patient] was elderly and hard of hearing, and nearly aphasic (due to previous medical condition)”
“Difficult to express compassionate caring just with voice, eyes, and upper face (!), and with minimal touch. However, [patient] and family both appeared very grateful for communication and attempts to normalize situation as much as possible.”
“Challenges include that the patient was a resident in an assisted living facility …Her old age made it difficult for telemedicine, however, we made use of it once just prior to diagnosis to discuss advance care planning in light of COVID.”
“I used an iPad to allow final visit [between patient and family]. Staff RN held room phone to patient ear; unknown what he heard or didn't.”
“I learned not to wait too long before arranging for video communication. Shortly after the video call, the [patient] became unresponsive and remained this way.”
Family visitation challenges
Respondents commented that facilities frequently disallowed family visitation or that family members could visit, but chose not to due to risk of COVID-19 exposure. Lack of family at bedside led to concern of additional emotional distress for both the patient and family, poor understanding of perceived quality of care, and impedance in the family's ability to comprehend the severity of illness. Patients often died without family members at the bedside, greatly exacerbating the long-term impact of these communication challenges on families and staff.
“Family could have visited but would have had to be quarantined for 14 days. Elderly wife and all children chose to not visit.”
“As no visitors allowed, had phone conference with daughter. Mother with end-stage dementia, 20% weight loss in last several months was stable on 4L oxygen. Daughter supported emotionally changed code. That night, mother unexpectedly worsened rapidly and died. Too fast to arrange last visit, sadly. Enrolled daughter in bereavement program.”
“Family was ‘allowed’ to be at bedside but fearful due to their own health issues and did not come in. Family not able to witness patient condition until >2 weeks into the illness when video capabilities became available.”
“Inability for family to be present and fully understand severity of illness.”
Communication barriers between clinicians and families
Respondents frequently cited that visitation challenges led to significant communication barriers between staff and family members. Providers frequently utilized video and phone calls to communicate with family members. Language barriers and the need for a translator for difficult conversations were an added challenge via telemedicine.
“Group phone calls were used to communicate with family and colleagues in the ICU for this consult. Challenges included a language barrier and the use of a translator for difficult conversations.”
“Communicating with Spanish speaking son in Mexico, located in an area not accessible by cell phone.”
“GOC discussion by phone with family we never met face to face…Family not allowed to visit.”
“Telehealth was used within the hospital setting. Generally, our palliative team has multiple members interact with each patient/family unit. In this setting, we only had our NP involved [due to] limited screen availability. Our MSW and Chaplain plan to call via telephone only to provide support to the family.”
Rapid changes in palliative care medical management
Respondents cited concerns about bedside staff limiting the amount of time they spent in the room and missing signs of distress they would normally identify, leading to changes in how medications were ordered. COVID-19 staffing changes impacted patient care in other ways. Lack of rapid COVID-19 testing early in the pandemic limited options for mechanical respiratory support due to safety concerns.
“I learned to order scheduled Morphine instead of PRN because the staff was trying to limit the number of times they were in and out of the room. It was frustrating that the staff could not see the [patient] unless they were in the room, so signs of distress may have been missed.”
“Unavailability of neuropsychiatric testing due to COVID regulations and cutbacks on staffing.”
“Challenging to not be able to fully support patients with respiratory support (BiPAP, CPAP) without ruling out COVID-19 even if low risk. Forced providers to intubate as unable to use BiPAP. May lead to future issues.”
Community care options difficult to find
To add to the difficulty of providing care during a pandemic, respondents shared that it is hard for patients with COVID-19 to meet facility acceptance criteria (after discharge), hospice programs may lack staff and PPE, and that finding paid caregivers for COVID-19-positive patients is difficult and more expensive.
“Finding caregivers for COVID-positive patients costs more.”
“Discharge planning/Placement was difficult given need for 2–3 (depending on accepting facility criteria) COVID negative [tests] prior to accepting patient.”
“He required 24-hour care and it was a challenge to find a new caregiver who could be provided with proper PPE. It also took some time to find a hospice with adequate staff and PPE.”
“The family did not feel that the facility and hospice were able to provide good support to the patient overall and it was hard to not be present at the time of death. He died about 1 week after his positive COVID test.”
Lack of testing in community-based settings
Two reports we received from community-based palliative care programs reported lack of access to COVID-19 testing as a significant challenge at the beginning of the pandemic. These quotes did not persist through the pandemic after the time that widespread testing became available.
“No tests available, so no diagnosis. Referred to hospice for COPD. Patient ‘recovered,’ but dying of lung damage. Family continues to struggle with accepting diagnosis and ongoing poor prognosis, due to lack of testing.”
“Challenges in getting testing for COVID-19 in the outpatient setting”
Guardianship and legal challenges
Two respondents shared experiences regarding delays in patient care due to slow responsiveness by the court system, in one case with concern that the patient would die before the courts responded.
“Delays in setting limits in care while awaiting expansion of court-appointed guardianship.”
“[Patient] has a significant psychiatric history and was conserved by LPS, and all consents (particularly Remdisivir and convalescent plasma) had to go to court for a decision. It was very frustrating dealing with the tons of paperwork and bureaucracy that is very entrenched in its ways, with everyone at LPS already resigned to the possibility that [patient] will likely die or get better before the case is event presented to the courts.”
Discussion
Specialty palliative care teams have served many roles in the care for patients with COVID-19 and their families. Among the preliminary findings drawn from 306 case reports submitted to the PCQC, we found a preponderance of older, male patients with underlying cardiovascular disease being the most common underlying serious illness. Surprisingly, nearly 10% of palliative care consults were for patients with no underlying serious illness before COVID-19. Palliative care teams were most frequently consulted to address GOC and support families. Among requests for symptom management assistance, shortness of breath was the most common, consistent with the clinical presentation of COVID-19. Palliative care teams were also called to provide support to colleagues and staff. Shared decision making supported by palliative care teams led to changes in care plans, modifications in resuscitation status preferences, and enabling of patient-family, family-clinician communication despite limitations in family visiting. Challenges related to palliative care delivery during the COVID-19 era included strained communication with patients resulting from masking due to PPE, significant visitation barriers for family, rapid transitions to tele-communication with family members, and limited discharge options in the community.
Our results highlight the unique characteristics of the population referred to palliative care. For example, we compared these preliminary reports of 306 COVID patients receiving palliative care consultations with the COVID-19–Associated Hospitalization Surveillance Network (COVID-NET), which collects population-based surveillance data from 14 states across 1482 hospitalized patients. 8 Compared to COVID-NET, more (82% vs. 75%) of the population reported to the PCQC registry were older than 50 years, perhaps due to greater clinician concerns for the outcomes of older patients resulting in referral to specialty palliative care. Further, COVID-NET reported an 80% incidence of shortness of breath at admission; symptoms at admission were not collected in the PCQC COVID-19 Registry, and palliative care teams were only consulted specifically for this symptom in 19% of patients. In the United States, mortality rates for hospitalized patients have ranged between 17% and 21%.9,10 In our sample, 50.7% of patients reported to the Registry did not survive to discharge, again reflecting a likely higher risk of mortality among those referred to palliative care. This limited comparison suggests greater utilization of specialty palliative care among older patients at higher risk of death, with referring teams requesting help primarily with GOC and family support, rather than for symptom management.
These preliminary data on a small sample of COVID-19 patients referred to palliative care specialist services and reported to the PCQC COVID-19 Registry suggest an essential role for palliative care services, especially among older and higher risk patients. While registry aggregation of case report data is a key element in assessing the role and impact of palliative care services during the pandemic, findings must be complemented with other reported experiences. Similar to our findings, the role of palliative care specialists in the assessment and management of patient and caregiver distress with COVID-19 has been reported from multiple single centers.3–5,11 To our knowledge this is the first report representing multiple geographies. This report highlights the timely need to rapidly collect data to inform care guidelines and policy makers. During the pandemic, palliative care specialists have provided significant value to health systems, including specialist team-run telephone support lines for families and clinicians, 11 support for colleagues in beleaguered emergency departments, 12 creation of digital health resources to collect and share clinical best practices for management of palliative needs (capc.org/covid-19/), and deployment of novel palliative care/psychiatry joint consult liaison services. 13 Our registry demonstrates that routine, de-identified, concise (<5 minutes per report) data collection is feasible in busy palliative care clinical environments. Although this is, to our knowledge, the largest registry report of specialty palliative care experiences with COVID-19, we encourage palliative care clinicians working in high prevalence areas to participate in the PCQC COVID-19 registry (https://www.palliativequality.org/covid-19-case-report) to ensure a large and adequately representative sample for geographic distribution, program type, and patient populations. We recognize that real-time reporting in the midst of a surge is not feasible and encourage teams simply to collect data in real-time and have a nonclinical member of the team submit them at a later time. When fully operational, the PCQC registry will provide seamless integration with the electronic health record, which will simplify this and routine data collection and reporting.
There are several key limitations. First, the self-report and nonrepresentative nature of the case reports introduces selection bias among experiences in the registry (the busiest programs seeing the highest volume of cases have not yet submitted reports, e.g., in New York City). Our numbers are too small and nonrepresentative to permit general conclusions. Second, the analysis of the qualitative data was limited to two authors using inductive coding and a flat coding frame. Nonetheless, the comments provided offer helpful insights and strategies for caring for patients with COVID-19. Third, the vast majority of case reports reflect experiences with adult patients, limiting conclusions regarding care in the perinatal and pediatric settings. Finally, to increase feasibility of data collection, respondents were not required to complete specific items before submitting data, and race/ethnicity was added after the first 42 cases. This resulted in missing data, most notably patient race/ethnicity, limiting our ability to draw conclusions regarding racial and ethnic disparities in this cohort.
Conclusion
Preliminary data from the first 306 patients with COVID-19 treated by hospital palliative care teams and reported to the new PCQC COVID-19 Registry suggest feasibility of data collection and show that specialty palliative care teams were most often called to care for older and higher risk patients. Palliative care teams were most commonly consulted to support their colleagues in ensuring shared decision making about life-sustaining treatments, including CPR, and to facilitate communication and support for family members unable to visit. Challenges identified included the difficulty of communicating with patients through masks and eye shields, an immediate and complete reliance on telemedicine to inform and support families, and lack of safe discharge options for end-of-life or convalescent care. Given the widespread utilization of palliative care specialist services during the pandemic, we encourage our colleagues from across the country to continue to submit their data in an effort to build a representative and diverse sample that could further help improve care.
Footnotes
Acknowledgments
The PCQC acknowledges the tireless and dedicated palliative care professionals who generously contributed to the COVID19 registry and the Gordon and Betty Moore Foundation and Cambia Health Foundation for supporting the resources required.
Funding Information
Funding was provided by grant support to the Palliative Care Quality Collaborative from the Gordon and Betty Moore Foundation and the Cambia Health Foundation.
Author Disclosure Statement
Mr. Mueller and Ms. McKenna are paid employees of the Palliative Care Quality Collaborative. All other authors are volunteer members of the PCQC Board of Directors. All authors have received salary or travel support through the Gordon and Betty Moore Foundation or Cambia Health Foundation during the course of grant-funded work in developing and launching the PCQC. Authors have no other relevant disclosures.