“I'm Being Forced to Make Decisions I Have Never Had to Make Before”: Oncologists' Experiences of Caring for Seriously Ill Persons With Poor Prognoses and the Dilemmas Created by COVID-19

PURPOSE:

The COVID-19 pandemic has created a new set of problems for clinicians. This study examines the experiences of oncologists providing care to seriously ill persons near the end of life in the context of the COVID-19 pandemic.

METHODS:

Between January 2020 and August 2020, we conducted semistructured, in-depth individual interviews with 22 purposefully sampled oncologists from practices enrolled in the Michigan Oncology Quality Consortium. Deidentified transcripts of the interviews were examined using thematic analysis.

RESULTS:

Our respondents described several novel problems created by the COVID-19 pandemic, including: (1) ethical challenges, (2) the need to manage uncertainty—physically and emotionally—on the part of both patients and oncologists, and (3) the difficulty of integrating technology and communication for seriously ill persons. These problems were made more complex by features of the pandemic: resource scarcity (and the need to fairly allocate poor resources), delays in care, high levels of fear, and the increased importance of advance care planning. Nonabandonment served as a way to cope with increased stress, and the use of telemedicine became an increasingly important medium of communication.

CONCLUSION:

This study offers an in-depth exploration of the problems faced by oncologists as a result of the COVID-19 pandemic and how they navigated them. Optimal decision making for seriously ill persons with cancer during the COVID-19 pandemic must include open acknowledgment of the ethical challenges involved, the emotions experienced by both patients and their oncologists, and the urgent need to integrate technology with compassionate communication in determining patient preferences.

INTRODUCTION

Affecting nearly all aspects of health care and its delivery systems globally, the COVID-19 pandemic raises new and complex moral dilemmas for health care providers.^1-9 Oncologists, like many other clinicians, care for seriously ill persons at disproportionately high risk of complications and increased mortality if infected with the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2).^10-13

We know very little about how oncologists work with patients with poor prognoses—a term we use throughout this paper to refer to patients with incurable cancer near the end of life—to make decisions pertaining to withholding or continuing chemotherapy and other anticancer treatments within the context of the pandemic. In this study, we examine the experiences and attitudes of oncologists providing clinical care to seriously ill persons in the context of the COVID-19 pandemic.

METHODS

Participants

Between January 2020 and August 2020, the authors conducted semistructured, in-depth individual interviews with 22 medical oncologists purposefully selected for diversity from 52 practices enrolled in the Michigan Oncology Quality Consortium, a physician-led collaborative sponsored by Blue Cross Blue Shield of Michigan that seeks to improve the quality of cancer care for all patients regardless of insurance status. All participants routinely cared for seriously ill persons and prescribed cytotoxic chemotherapy within the prior 2 years of the study.

Procedure

Using the American Society of Clinical Oncology Quality Oncology Practice Initiative data for practice sites from 2019, we categorized practices according to low, medium, and high groupings for following best practices for care near the end of life, specifically the measure “chemotherapy administered within the last 2 weeks of life,” in which a lower proportion is considered favorable.¹⁴ We performed internet searches on all potential participants to categorize sex, race, and practice setting (hospital and/or clinic and rural/urban/suburban). Using sex, race, Quality Oncology Practice Initiative grouping, and practice setting, participants were grouped into recruitment groups to ensure sample diversity. Aside from the first two participants enrolled in our study, we randomly selected from these purposeful sampling groups to create lists of 12-15 oncologists to invite via e-mail, fax, and/or phone calls each week to interview, for a total of 120. Information regarding the study was provided during two in-person meetings of the members of the Michigan Oncology Quality Consortium. As oncologists accepted invitations and scheduled interviews, we iteratively adjusted our subsequent invitations to ensure a reasonably balanced representation of individuals using the aforementioned categorizations. We proceeded with interviews until we reached thematic saturation, the point at which no new information or novel themes emerged from the interview data. The study was deemed exempt from full review by the University of Michigan Institutional Review Board.

Data Collection

Participation was voluntary, and all participants provided informed consent. Each interview lasted between 45 and 90 minutes in length. All participants were provided a $50 honorarium.

A semistructured interview guide (Data Supplement, online only) elicited detailed narratives of oncologists' experiences with the use of chemotherapy in patients near the end of life and the oncologists' responses to interventions to reduce such practices. Interviews were digitally recorded, transcribed verbatim, and deidentified before analysis. Themes related to COVID-19 were introduced 3 months into the study after the fourth interview, at which point we added one open-ended question to the existing interview guide, asked toward the end of the interview if the topic did not enter the conversation earlier: “How has the COVID-19 pandemic impacted the care you provide for seriously ill patients with cancer who have advanced stages of disease and poor prognoses?”

Data Analysis

Data were coded and analyzed using thematic analysis, a qualitative method allowing for inductive and deductive reasoning, as described by Braun and Clarke.¹⁵ Authors C.R.P. and E.C., both with graduate training in social sciences and qualitative methods, independently reviewed and thematically coded each transcript using MAXQDA 2020 (VERBI Software, Berlin, Germany). The investigators were diverse in terms of educational training, race, ethnicity, sex, and age. Investigator triangulation, a process that brings together diverse analytic perspectives to form a collaborative approach that maximizes the application of different expertise levels in various subject areas relevant to our study, minimized the possibility of systematic bias and improved trustworthiness. Coding began while interviews were ongoing so that the analysis could inform the conduct of interviews, and vice versa. The coding scheme was developed by reading the first six transcripts line by line and labeling with words from the text (or descriptions of those words) all pieces of text that identified or described ideas related to several code categories created both inductively and a priori based on the research aims of the overall study. The categories included: experiences with the use of chemotherapy near the end of life, factors influencing this practice, interventions to reduce the use of end-of-life chemotherapy, facilitators and barriers to interventions, patient decision making, physician decision making, risk tolerance, decision regret, emotions experienced by oncologists, advanced therapies, financial factors, artificial intelligence, hospice and palliative medicine, COVID-19, and practice-related information. All transcripts were then independently double-coded using the analytic code categories and the coding scheme. To further reduce the risk of systematic bias, coding differences were discussed and reconciled between investigators. The coding scheme allowed us to identify and analyze several patterns within and across the interviews, including elements that appeared frequently or were particularly illustrative or unexpected. Quoted responses included in this article are those that best capture the themes in the data. The results are presented following the Standards for Reporting of Qualitative Research.¹⁶

RESULTS

Table 1 details the demographic and practice characteristics of the interview respondents. A total of 22 oncologists participated between January 2020 and August 2020; another 15 oncologists had expressed interest in participating, but because of COVID-19–related disruptions, either canceled their scheduled interviews or were unable to participate before thematic saturation was attained. Of the interviews, 18 contained narratives related to the emerging COVID-19 pandemic, which might be expected, given that the majority of interviews were conducted after the pandemic is widely recognized to have started in the United States. The topic of the pandemic organically entered nearly every interview after March 10, 2020, when a state of emergency was declared in Michigan due to COVID-19.^17,18

TABLE 1.

Oncologist Characteristics

Several major thematic clusters were identified, one of which related to problems associated with the COVID-19 pandemic. In this article, we report the results pertaining to three themes within this cluster: (1) the ethical challenges faced by oncologists due to the COVID-19 pandemic, (2) the need to manage uncertainty and associated emotions for both patients and oncologists, and (3) the importance, and complexity, of integrating technology and communication for seriously ill persons. Illustrative quotations are presented in Table 2 with numbers for reference.

TABLE 2.

Themes and Illustrative Quotes

Ethical Challenges Faced by Oncologists During the COVID-19 Pandemic

Oncologists in urban, suburban, and rural settings reported a scarcity of resources and an increase in patients foregoing care (Q1-Q4). One described the need to factor in both existing and projected resource scarcity to determine whether her patients could receive chemotherapy (Table 2, Q1). Other forms of resource scarcity presented as delays in care and diagnoses, with oncologists describing the once routine process of setting up testing and visits as “a challenge” (Q2-Q5). Fear of infection caused inappropriate delays in managing therapy complications, surgeries, and radiation treatments and slowed patient decisions involving necessary (and possibly curative) treatments (Q3, Q8). One oncologist detailed it as a cascading stream of events for her patients 6 months into the pandemic, with delays in screenings leading to later presentations of disease, loss of “windows” in which to treat, and further delays in treatments once diagnosed (Q4). Several oncologists described similar experiences with engaging hospice and palliative medicine sooner than anticipated, as patients presented later in their disease course and were quite ill on presentation, and cytotoxic chemotherapy was not deemed beneficial (Q4). Another oncologist noted that delays were not as much of a problem at his institution compared with patients declining to adhere to recommendations for acute care or monitoring due to fear of contagion (Q6).

Fear of the unknown affected both patients and oncologists. Some oncologists were aware of early cohort data on patients with oncologic diagnoses and the increased risk of acute complications in patients with cancer due to the SARS-CoV-2 virus.¹⁰ As one oncologist who later became infected with COVID-19 expressed, “the unknown” is an existential threat to understanding the etiology and sequalae of COVID-19 infection. She noted the “unsettling” feeling of being unable to answer her patients' questions in the face of uncertainties about COVID-19 (Q5).

Additionally, exacerbation of racial inequities and injustices among patient groups was simultaneously becoming salient for many of the oncologists. One oncologist in Detroit, an early epicenter of the COVID-19 pandemic in the United States with high mortality,¹⁸ described an exacerbation of pre-existing inequities and distrust with the medical profession in his community as a result of the pandemic (Q9). This oncologist described an already stressed population undergoing an even greater challenge caused by the COVID-19 pandemic and his role in witnessing the crisis unfold while treating patients and their families. Respondents described a perceived duty to promote equity, especially when patients were non-English speaking or from minority ethnic groups (Q9-11).

The Need to Manage Uncertainty and Emotions for Both Patients and Oncologists

Uncertainty abounded during the initial 6 months of the pandemic, and oncologists witnessed disturbing experiences resulting from restricted visitor policies that prevented family members from being present when a loved one was dying in the hospital setting. Oncologists with patients whose cancer treatments were complicated by COVID-19 infection described uncertainty for patients and physicians, noting that the knowledge of patients testing positive signaled an increased risk of death and accompanying devastation as a prelude to death (Q14, Q17). A professor of medicine noted with angst the improbability of having to make the kinds of unique decisions he was making during the pandemic (Q15). He e-mailed the study team an article⁸ and implored, “We are being asked to make hard decisions on who to give/hold chemotherapy, accept/deny new patients who need chemotherapy. Is this an opportunity for you to design a study around how to make these decisions?”

Many specifically strategized ways to mitigate the possibility of patients dying alone. One community-based oncologist cited experiences with inpatients dying alone during the pandemic as a reason for a more paternalistic approach for one of her elderly patients (Q17).

Oncologists specifically discussed nonabandonment as a coping mechanism in the face of increased stress and adversity, despite incurring risk to themselves and their families. One oncologist shared that her practice overhauled its processes to increase their ability to provide care, a strategy reported by more than half of our respondents. Her practice adopted a rallying cry of “‘COVID didn't cure cancer!’” throughout the care team environment to promote encouragement (Q7). Another oncologist continued to see outpatients in person despite the health system sequestering alcohol-based hand sanitizer and masks from the clinic to bolster hospital supply.

Yet another oncologist discussed the personal impact caused by the death of patients in her practice due to COVID-19, either directly due to infection with SARS-CoV-2 while undergoing cancer care or indirectly due to delayed care and progressive disease (Q18). Several oncologists discussed the need for compassion, honesty, optimism, and reframing hope consistent with palliative care principles, with one reflecting on the value of giving talks to colleagues about this subject. From his perspective, the pandemic enhanced the need to have compassionate, honest conversations and to educate others about reframing hope when curative cytotoxic chemotherapy is no longer beneficial (Q19).

The Importance and Complexity of Integrating Technology and Communication in Care for Seriously Ill Persons

Several oncologists mentioned that the pandemic resulted in their rapid introduction to telemedicine. One described how his clinic used “video visits” to reassure patients (Q21). Another noted that he struggled with telemedicine, in part because his patients have unique issues related to prognostic uncertainty and the use of newer immunotherapies. He opted to keep as many of his patient interactions as in-person as possible (Q22).

Other participants noted how telemedicine improved their communication with patients who were receiving home hospice care, an otherwise poor resource in their geographic area. While discussing one experience, one oncologist noted it as an important lesson in the potential of using telemedicine more frequently for patients near the end of life and avoiding missed opportunities (Q23). She reflected, “Seeing the patient in their home teaches you a lot about that patient.” Another oncologist subspecialized in palliative care reflected on how much he had learned about telemedicine in a compressed time frame about reaching rural patients (Q24). But others in our sample felt that patients with cancer need to be seen in person at times (Q25).

We did hear stories of how telemedicine complicated care, especially when hospital policies restricted family visitation near death. One oncologist described a telephone call where a patient's son was insisting on a second opinion, while the patient was imminently dying in the hospital after stopping chemotherapy treatments (Q26). He noted that without face-to-face in-person interaction, the difficulty of having these complex conversations with the family, and between the patient and the family, was amplified. The need to negotiate differences in preferences for end-of-life care via telephone or a computer made difficult conversations harder. The complexity of such decision making resulted in some oncologists recognizing the need for advance care planning conversations, specifically characterizing advance directives as necessary (Q12, Q13).

DISCUSSION

The fact that our qualitative study coincided with the outbreak of the COVID-19 pandemic allowed us to learn about the many problems faced by oncologists during the first 6 months of the pandemic, building on prior literature.^19,20 The challenges we identified entailed grappling with resource scarcity, delays in care, a duty to promote equity and nonabandonment, high levels of uncertainty and fear, and the importance of appropriate end-of-life care planning. Oncologists' decision making confronted the ethical imperative to shift from a patient-centered paradigm to one with more of a public health focus. This included recognizing the impact of visitor restriction policies on seriously ill patients with poor prognoses, including the possibility of patients dying alone in the hospital. Oncologists also faced more patient deaths as a result of both the COVID-19 infection itself and delayed and different treatment options for cancer due to capacity restrictions, the need to minimize in-person contact, and patients' avoidance of the health care setting. At the same time, they were adapting to seismic changes in practice, such as the widespread urgent adoption of telemedicine, while confronting substantial infection risks for themselves and their families.^19,20

Ethical challenges should be expected in a rapidly developing public health crisis, underscoring the need for interventions and guidelines to improve clinicians' ability to respond and for research to evaluate such interventions.^21-23 Hospice and palliative medicine providers and clinical ethicists, always a limited resource, became even more limited during the pandemic.^24-28 Difficult conversations are part of an oncologist's daily workload, and yet research suggests that responsibility for life-and-death decision making is quite difficult for oncologists who are caring for patients near death.^29-32 Our study calls attention to the pandemic as a teachable moment on the importance of primary palliative care skills training for all physicians on aligning treatment with individual preferences using excellent communication skills and sound symptom management.^33,34 Critical federal policy proposals such as H.R.647, the Palliative Care and Hospice Education Training Act, would have increased nationwide research and training in palliative care prior to the COVID-19 pandemic, but it was stalled for several years in Congress despite bipartisan support.³⁵ Passing such legislation would have also potentially addressed improving advance care planning for seriously ill persons with poor prognoses. The latter is an important subject that has been researched extensively but often found to be problematic due to the necessary complexity of managing serious illness conversations and aligning preferences over time.^36-39

Recent guidance on the use of excellent communication skills in the context of the pandemic and oncologic practice has incorporated invaluable guidance to clinicians from VitalTalk (a 501c3 nonprofit developed by national leaders in palliative care). An interdisciplinary team, including experts in health communication and oncologists confronting the COVID-19 pandemic, created an evidence-based communication guide based on the perspectives of oncologists and patients, an intervention that merits future testing.^20,40 The pandemic has taught us that complex communication using mixed modes of delivery, including telemedicine, must be studied as a means to provide serious news virtually, an approach that is not common among oncologists or palliative care providers.^39,41-46 Technology and communication need to be improved to handle these challenges, combined with expanding existing services, such as clinical ethics and palliative care consultations for complex situations. Preliminary results from the Early Integrated Telehealth Versus In-Person Palliative Care for Patients With Lung Cancer trial (REACH PC) demonstrate that telemedicine has potential to eliminate some challenges associated with more traditional palliative care services for outpatients with cancer.^47,48 Importantly, a telemedicine virtual visit may allow for some version of the traditional home visit with considerable benefits, including eliminating the sometimes expensive and physically taxing efforts of patients with serious illness and their caregivers to travel to the doctor's office, while allowing the physician to reach geographic areas with low access to in-person care faster.

Advances in technology often exacerbate racial and socioeconomic inequities.⁴⁹ Growth in the use of telemedicine requires close ethical examination of unintended consequences, especially during a pandemic that has disproportionately affected Black and other minority communities by exacerbating existing racial inequities.⁵⁰ Telemedicine has great promise for the delivery of cancer care at home^51,52 while also designing systems and policies that are responsive to translational research findings, improved accessibility, and compassionate preference-sensitive communication as recommended by expert consensus.⁵³ Finally, increasing patients' choice of providers with race-concordant and/or gender-concordant diverse providers may be beneficial to reducing barriers to complex conversations for seriously ill persons and needs further evaluation.⁵⁴

As with any study, our research has limitations. One potential limitation is that responses may have a social desirability bias; when speaking with another physician, there may be a performative aspect to answers. We guarded against this by having one nonphysician research analyst conduct six of the interviews. Since our study does not include direct narratives with patients and families or other care team members, we can only report from the perspective of oncologists. Finally, interpretation of interview data can be subject to inherent bias of analysts; as detailed in the description of our methods, we used rigorous and accepted techniques to enhance the trustworthiness of data collection and analysis.^15,16

In conclusion, our study offers an in-depth exploration of the problems faced by oncologists and how they navigated them during the COVID-19 pandemic. Many demonstrated morally admirable behavior and resiliency in the face of adversity, addressing problems with alacrity and ingenuity. Optimal decision making for seriously ill persons with cancer during the COVID-19 pandemic must include open acknowledgment of the ethical challenges involved, the heightened emotions experienced by both patients and their oncologists, and the urgent need for integrating technology with compassionate communication in determining patient preferences.

AUTHOR CONTRIBUTIONS

Conception and design: Chithra R. Perumalswami, Susan D. Goold, Raymond De Vries, Reshma Jagsi

Financial support: Chithra R. Perumalswami, Reshma Jagsi

Administrative support: Chithra R. Perumalswami, Reshma Jagsi

Provision of study materials or patients: Chithra R. Perumalswami, Jennifer J. Griggs

Collection and assembly of data: Chithra R. Perumalswami, Emily Chen, Carly Martin

Data analysis and interpretation: Chithra R. Perumalswami, Emily Chen, Carly Martin, Susan D. Goold, Raymond De Vries, Jennifer J. Griggs, Reshma Jagsi

Manuscript writing: All authors

Final approval of manuscript: All authors

Accountable for all aspects of the work: All authors

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

“I'm Being Forced to Make Decisions I Have Never had to Make Before”: Oncologists' Experiences of Caring for Seriously Ill Persons With Poor Prognoses and the Dilemmas Created by COVID-19

The following represents disclosure information provided by the authors of this manuscript. All relationships are considered compensated unless otherwise noted. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO’s conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/op/authors/author-center.

Open Payments is a public database containing information reported by companies about payments made to US-licensed physicians (Open Payments).