“You’re Socially Distant and Trying Not to Be Emotionally Distant.” Physicians’ Perspectives of Communication and Therapeutic Relationships in the ICU During the COVID-19 Pandemic: A Qualitative Study

OBJECTIVES:

To: 1) characterize how COVID-19–related policies influence patient-clinician communication and relationships in the ICU, with attention to race and ethnicity as factors and 2) identify interventions that may facilitate patient-clinician communication.

DESIGN:

We conducted a qualitative study between September 2020 and February 2021 that explored facilitators and barriers to patient-clinician communication and the formation of therapeutic relationships. We used thematic analysis to develop findings describing patient-communication and therapeutic relationships within the ICU early in the COVID-19 pandemic.

SETTING:

We purposively selected hospital dyads from regions in the United States that experienced early and/or large surges of patients hospitalized with COVID-19.

SUBJECTS:

We recruited a national sample of ICU physicians from Veteran Affairs (VA) Health Care Systems and their associated academic affiliate hospitals.

INTERVENTIONS:

None.

MEASUREMENTS AND MAIN RESULTS:

Twenty-four intensivists from seven VA hospitals and six academic-affiliate hospitals participated. Intensivists noted the disproportionate impact of the pandemic on among people holding minoritized racial and ethnic identities, describing how language barriers and restrictive visitation policies exacerbated institutional mistrust and compromised physicians’ ability to develop therapeutic relationships. We also identified several perceived influences on patient-clinician communication and the establishment of therapeutic relationships. Barriers included physicians’ fear of becoming infected with COVID-19 and use of personal protective equipment, which created obstacles to effective physical and verbal interactions. Facilitators included the presence of on-site interpreters, use of web-based technology to interact with family members outside the ICU, and designation of a care team member or specialist service to provide routine updates to families.

CONCLUSIONS:

The COVID-19 pandemic has threatened patient-clinician communication and the development of therapeutic relationships in the ICU, particularly among people holding minoritized racial and ethnic identities and their families. We identified several facilitators to improve patient-clinician communication as perceived by intensivists that may help improve trust and foster therapeutic alliances.

KEY POINTS

Question: How do hospital policies implemented during the COVID-19 pandemic influence patient-clinician communication in the ICU; are there interventions that could facilitate the development of therapeutic relationships?

Findings: Intensivists noted the disproportionate influence of the pandemic on trust and communication between minoritized populations and their clinicians, describing how language barriers and visitor restrictions compromised their ability to develop therapeutic relationships. We identified several influences on patient-clinician communication and therapeutic relationships as perceived by intensivists.

Meaning: We identified several modifiable strategies to improve trust and therapeutic relationships in the ICU during the COVID-19 pandemic, as perceived by intensivists.

As of May 19, 2022, the United States surpassed 1 million deaths due to COVID-19—a pandemic that has disproportionately impacted racial and ethnic minoritized populations (1, 2). The strain placed on healthcare systems and workers by COVID-19 is unprecedented in the modern day and has led to moral distress and increased burnout among clinicians (3–7). Hospital policies and procedures aimed at protecting clinicians and patients (e.g., social distancing and visitor restrictions) may have unintended consequences that negatively influence patients, their loved ones, and clinicians alike (8–11).

The pandemic disrupted typical methods of communication in the ICU, where family presence and participation in rounds are encouraged (12). Quality of communication during healthcare encounters is important to establish trust and therapeutic alliances between clinicians and their patients (13). Degradations in any component of communication can impact multiple aspects of the patient-clinician relationship, including trust and satisfaction, potentially worsening patient outcomes (14). Currently, little is known about how the pandemic influenced the patient-clinician therapeutic relationship, particularly among people holding minoritized racial and ethnic identities who have been disproportionately impacted by COVID-19. Therefore, we sought to: 1) characterize how COVID-19–related policies influence patient-clinician communication and relationships in the ICU, particularly among minoritized populations and 2) identify potential interventions that may facilitate development of therapeutic relationships.

METHODS

Participant Selection

For this qualitative study, we purposively sampled seven hospital dyads (Veteran Affairs [VA] hospital and its academic affiliate) from the Pacific Northwest, Northeast, Southeast, and Southwest United States in early 2020. We sought to explore similarities and differences between VA and academic affiliate hospitals responses to the pandemic, including polices, staffing, and patients served, and how these influenced patient-clinician communication and therapeutic relationships in the ICU setting. We selected 12 hospitals in areas that experienced early and/or large surges of patients hospitalized with COVID-19, using a combination of key informant and snowball sampling to recruit 2–3 ICU physicians from each hospital via email or clinician referral. All participants were provided an information sheet and completed verbal informed consent prior to the interview. Our study was granted a waiver of consent documentation and was approved by the VA Portland Health Care System/Oregon Health and Science University Institutional Review Board (IRB) (VA No. 4627) on June 6, 2020. Every procedure was followed in accordance with the ethical standards of the VA Portland Health Care System/Oregon Health & Science University IRB (VA No. 4627) and with the Helsinki Declaration of 1975. We report details of our methods using the Strengthening the Reporting of Observational Studies in Epidemiology reporting guideline (15).

Data Collection

Data collection occurred between September 2020 and February 2021. We used a semi-structured interview guide focused on two main domains: 1) patient-clinician relationships and communication and 2) the psychologic impact related to the overall strain of the pandemic (Supplemental Table 1, http://links.lww.com/CCX/B134). The focus of the present paper is on the patient-clinician relationship and communication.

Analysis

The multidisciplinary research team, including intensivists (J.C., C.G.S., K.C.V.), sociologists (S.E.G., A.T.), and a clinical psychologist (S.N.), iteratively revised the interview guide during two pilot interviews not included in analyses. S.N. and/or J.C. conducted individual interviews (45–60-min each) over a secure web-based platform. Interviews were digitally recorded, transcribed, verified for accuracy, and de-identified.

We used the Framework Method (16) for analysis of the qualitative data. J.C. and S.N. independently coded the first three transcripts to create a preliminary codebook based on deductive and inductive codes. Deductive codes were derived from questions in the interview guide. The codebook was discussed with S.E.G., who coded the next three transcripts with S.N. to further refine the codebook. S.N. and S.E.G. coded four additional study transcripts together, iteratively refining the codebook and recoding previously reviewed transcripts as necessary to incorporate newly identified themes. They independently coded the remaining transcripts, meeting weekly to review data, collapse themes, and reach agreement on discrepancies. They created framework matrices to aid in final data interpretation. Throughout this process, the multidisciplinary research team iteratively reviewed the codebook and helped perform analytic triangulation. We used an audit trail for tracking of decisions related to the codebook and analyses, using ATLAS.ti 8 (Berlin, Germany) to organize data.

RESULTS

We invited 35 attending ICU physicians, including three ICU directors (also frontline clinicians), by email; 24 (67%) agreed to participate, including two ICU directors. Over half (63%) identified as White non-Hispanic and 54% practiced at an academic institution (Table 1). Figure 1 displays the geographic distribution of participants. Participants are referred to by a number and hospital type (A = Academic, V = VA, B = Both), not connected to any identifying information.

TABLE 1.

Participant Characteristics

Participant Characteristic	n (%)
Gender identity:
Female	11 (46)
Male	13 (54)
Racial/ethnic identity:
White non-Hispanic	15 (63)
Asian	5 (21)
West Asian	2 (8)
Multiple races	1 (4)
Latino	2 (8)
Type of institution where clinician practiced
Veteran Affairs only	8 (33)
Academic hospital	13 (54)
Both	3 (13)

Figure 1.

Regional distribution of participants.

We identified two domains and six themes and a few perceived differences between physician experiences in VA and academic hospitals (Supplemental Table 1, http://links.lww.com/CCX/B134).

Domain 1: Threats to the Formation of Therapeutic Relationships

Theme 1: Visitation Restrictions Dehumanized Patients, Limited Trust, and Complicated Decision-Making

Physicians described not knowing the patients or families as well because of the family’s inability to be present, which led to a loss of the human side of medicine. Prior to COVID-19, physicians noted that family was often present during rounds, allowing clinicians to update families in real time (Supplemental Table 1, quote 1, http://links.lww.com/CCX/B134). The absence of in-person interactions made it challenging for the family to see a loved one’s decline and understand prognosis. One participant explained: “If you can’t have eye-to-eye contact, you lose the human connection… some of my colleagues had families that were not trusting of the diagnosis and were not willing to face reality and that their family member is dying” (SID15-A, Supplemental Table 1, quotes 2–4, http://links.lww.com/CCX/B134). Already time-strained physicians now had to manage communication with families via daily calls or other time-consuming mechanisms. One physician reported “It’s fallen on the nursing staff, there’s not additional time or resources for [contacting families]...” (SID22-B). Methods of communication (phones, other technologies) also contributed to dehumanization of patients. One physician stated: “Almost all of our conversations with families were by phone or by video. It was extremely hard for us; it was extremely hard for patients and families. It was really gut-wrenching and harmful to our souls, because it’s not the kind of care that we want to provide our patients and their families” (SID10-V, Supplemental Table 1, quotes 5–7, http://links.lww.com/CCX/B134).

This theme was particularly strong when physicians described their communication with “families marginalized by the intersection of race, ethnicity, language, and/or socioeconomic status.” Physicians noted the disproportionate effect that COVID-19 had on individuals from groups that have been socially or economically marginalized (Supplemental Table 1, quotes 8–10, http://links.lww.com/CCX/B134). One physician noted, “At one point everybody in the COVID ward had Spanish names, or all the families were Spanish-speaking only…Honestly it just speaks to how we’re not living in a post-racial world where everyone has equal opportunities, and people can just not work because they have families to take care of, bills they need to pay, and those needs are just disproportionately distributed across society” (SID18-A).

Institutional mistrust, coupled with family concerns about discrimination and racism, were heightened by visitation restrictions. One participant stated, “Particularly when we’re asking people to make end of life decisions, which requires an immense amount of trust. And they might even not be able to meet us… it is hard to cultivate trust when there is a discordance in ethnicity and understandable historical context for why people don’t trust” (SID23-A). Another physician described struggling with care concerns in the context of racism, explaining, “I was brought to my knees by some of these families and left completely speechless in trying to have goals-of-care conversations…I was not taught how to traverse incredible racism accusations or resource allocation accusations” (SID6-A). Participants noted challenges with technology that were especially pronounced among lower socioeconomic and racially minoritized families including lack of access to internet or video-capable devices and difficulty using technology.

Finally, several participants highlighted cultural considerations that were difficult to facilitate due to restrictive visitation policies. For example, one physician recounted how honoring the request of a Navajo patient’s family for a large family group present for goals-of-care discussion was not permissible due to visitor restrictions. The physician explained: “I realized after talking to the family that they hadn’t really developed a trusting relationship yet… there’s a historical context of the relationship with Western medicine that the Navajo nation has, that is just really hard to cut through. You really have to earn it, so we worked really hard on it…I felt like it was really important…[to] establish that they had somebody who was fighting for their mom” (SID17-A).

Theme 2: Fear of Infection and Personal Protective Equipment Use Created Barriers

Physicians reported that illness severity—including need for mechanical ventilation, patient intubation, and number of precautions needed to mitigate viral spread—made communication challenging (Supplemental Table 1, quote 11, http://links.lww.com/CCX/B134). One clinician stated, “You’re socially distant trying not to be emotionally distant, and it’s a hard bridge to actually overrun.” (SID5-V). COVID-19 was perceived as highly infectious; some clinicians did not enter rooms as frequently as they would have liked due to fear of infection (Supplemental Table 1, quotes 12–13, http://links.lww.com/CCX/B134), which interfered with communication and rapport-building opportunities. When clinicians did enter rooms, Personal Protective Equipment represented a physical barrier that made verbal and nonverbal communication challenging and reduced human touch. One clinician noted, “Recognizing body language and being able to use physical touch to provide sympathy or empathy, all of that was basically thrown out the window and it really was a traumatic experience for a lot of people not being able to use those things” (SID15-A).

Theme 3: End-of-Life Discussions and Grief Processes Were Disrupted

Participants reported that goals-of-care discussions were altered significantly by family’s inability to be physically present. Physicians felt end-of-life (EOL) conversations were often more uncomfortable compared with pre-pandemic times due to inability to communicate in-person with family. One physician posited, “How do you have end-of-life conversations and not be able to hold their hand or touch their shoulder or hand them Kleenex? And how do you explain to someone that their loved one is dying and they’re not going to be able to come in and see them and say goodbye? That was really the hardest part” (SID10-V). Discomfort stemmed from feeling as if clinicians were making EOL decisions unilaterally. “Decisions that would have been more shared with family members, had they been there throughout, would fall more on our shoulders to make in the moment. There was more of a burden of deciding when was time to stop offering more and more to support keeping them alive” (SID9-V, Supplemental Table 1, quotes 14–15, http://links.lww.com/CCX/B134). Physicians also described the tragedy of not having families there to grieve throughout different stages of critical illness, including death (Supplemental Table 1, quotes 16–17, http://links.lww.com/CCX/B134).

Domain 2: Novel Facilitators to Enhance Communication and Therapeutic Relationships

In addition to the challenges, participants identified several facilitators to enhance communication between the care team, patients, and their families.

Theme 4: Participants Adapted and Modified Technology, Often on Their Own Initiative, to Enhance Communication

Across all institutions, technologies were rapidly introduced and modified to enhance communication in light of visitation restrictions. Participants from non-VA hospitals described having more resources to quickly scale up communication technologies like web-based video conferencing and use of in-room video cameras/tablets (Supplemental Table 1, quotes 18–20, http://links.lww.com/CCX/B134). One clinician explained, “There was definitely investment in telemedicine and getting Zoom capacity up; we as a committee decided that we needed to do at least weekly Zoom family meetings, rather than just doing a daily touch base by phone” (SID2-A). Especially early in the pandemic, these efforts were trial and error, and sometimes staff would use their own devices to enable a patient and family to see one another (Supplemental Table 1, quote 21, http://links.lww.com/CCX/B134). Despite challenges, these technologies were largely viewed as helpful but still inferior to in-person interactions.

Theme 5: Team Members Beyond the Primary Service Were Valued in a New Light

Several clinicians noted that palliative care clinicians were essential to managing the workload of communicating with families (Supplemental Table 1, quote 28, http://links.lww.com/CCX/B134). One physician noted, “Our palliative care colleagues were really critical for our own well-being and that of our patients and their families… And were constantly dialoguing with us to make sure we were doing okay and with patients and families to make sure they felt supported” (SID10-V). In other cases, trainees or nursing managed communication and family updates (Supplemental Table 1, quotes 23–24, http://links.lww.com/CCX/B134). Some institutions developed communication teams responsible for communicating with families, often using standardized language based on smart phrases from the electronic health record (her). “We very quickly mobilized an excess of 40 medical students who called patient family members every day…we came up with, in… our EHR, smart phrases to facilitate communication that was fast, like check-boxy formatted notes” (SID7-B). Other physicians wanted to maintain the responsibility of direct family communication. As one clinician said, “I sort of viewed conversations about the EOL and critical illness as my purview as an intensivist” (SID23-A, Supplemental Table 1, quote 25, http://links.lww.com/CCX/B134).

Theme 6: Availability of Translator Services or a Cultural Mediator Was Essential Yet Not Always Available for Patients Who Did Not Speak English As Their First Language

We identified several facilitators for communicating with individuals who did not speak English, or for whom English was their second language. First, many institutions used interpreters, when available, which had its complexities, especially for languages other than Spanish or when used virtually (Supplemental Table 1, quotes 26–27, http://links.lww.com/CCX/B134). In the absence of available interpreters, medical providers who were fluent in Spanish often stood in for their colleagues: “I’m a native speaker, so I took the responsibility of updating all the families that only spoke Spanish.” (SID3-A). One institution had “cultural mediators” who helped identify cultural beliefs about healthcare and medical decision-making, which helped facilitate communications (Supplemental Table 1, quote 28, http://links.lww.com/CCX/B134). Finally, physicians spent additional time communicating with families where they perceived mistrust or a language barrier.

We identified a few main perceived differences between physician experience at VA and non-VA hospitals. Specifically, participants from non-VA hospitals more often reported seeing patients from racially and ethnically diverse populations (Supplemental Table 1, quote 29, http://links.lww.com/CCX/B134). Participants who worked at both VA hospitals and academic affiliates reported having fewer staff at VA who were responsible for a lighter patient load yet more overall responsibilities (Supplemental Table 1, quotes 30–31, http://links.lww.com/CCX/B134). Finally, those reported having a greater ability to quickly scale up use of technology within the ICU at the beginning of the pandemic (Supplemental Table 1, quote 32, http://links.lww.com/CCX/B134).

DISCUSSION

This qualitative study of intensivists from a national U.S. sample of VA and non-VA hospitals during the COVID-19 pandemic provides an in-depth understanding of perceived challenges to patient-clinician communication and the establishment of therapeutic relationships within the ICU. Specifically, we identified ways in which the pandemic threatened trust and disrupted communication and therapeutic alliances between intensivists, critically ill patients, and their families, particularly those minoritized by race, ethnicity, or language barriers. We identified how visitor restriction policies created barriers to communication and trust, and how the technological solutions to barriers resulted in unintended consequences, including the potential dehumanization of patients and exacerbation of existing disparities impacting families without reliable internet access. Finally, we identified several strategies used to overcome these challenges, which can inform efforts to enhance therapeutic relationships and trust in the setting of the ongoing COVID-19 pandemic.

Although several studies have found that individuals who identify as African American, Black or Latino have suffered more cases, hospitalizations, and death due to COVID-19 compared with White patients (17), little is known about the influence of the pandemic on the development of therapeutic relationships among these populations in the ICU. We identified strategies to improve therapeutic relationships in the ICU, including the use of easily accessible translator services as part of efforts to provide patient-centered care (18). Language-concordant care has been shown to improve patient-centered and clinical outcomes, highlighting the importance of having a diverse and multilingual healthcare team (19).

Additionally, visitor restrictions placed a high burden on clinicians and had many unanticipated consequences that were exacerbated among families from racially diverse or low socioeconomic backgrounds. One study examining the impact of restrictive visitation policies in Canadian ICUs found communication quality between clinicians and patients was severely diminished by disrupting the flow of information, making grief more challenging, and possibly contributing to moral distress of physicians (20). Importantly, another study found that family members, when using appropriate Personal Protective Equipment, are not significant contributors to hospital spread of COVID-19 infection and prioritizing entry of essential family and caregivers over more general hospital visitors is essential to improve patient care (21). Given the detrimental effect on clinicians and families alike, less restrictive visitor policies would likely ease some of the barriers to communication and therapeutic relationship development across ICU patients and their families.

There is an emerging evidence base describing the development and implementation of technologies to enhance therapeutic relationships and communication during COVID-19, using a combination of patient engagement tools, advanced care planning communication guides (22), electronic health record-based messaging (23, 24), and mobile applications (25), to enhance communication and social connectedness. Another study exploring the use of inpatient telemedicine found that its use in the ICU to communicate was acceptable from the patient perspective, but in-person communication was still preferred (20). Similar to prior evidence, we identified barriers to using technology including lack of internet access or smart devices and lack of knowledge about using technologies (26). For example, in 2021, only 65% of adults who identify as Latino and 57% of those earning less than $30,000 per year had at home broadband access, compared with 80% among those who identify as White (27). While broadband access is expanding, preferences for communication should be assessed when choosing alternative modes with each family. Finally, even in the presence of enhanced technology, lack of adequate clinician to communicate with families, remains a barrier, especially in the hospital settings that do not include trainees or specialty services such as palliative care.

Effective patient-clinician communication and involvement of family in communication and care are central components of high quality in the ICU. To have high-quality communication, clinicians need to treat patients with respect by showing commitment to solving their health issues, be knowledgeable and thorough, involve patients and families in decisions, exhibit empathy and positive relationships (e.g., being “on the same page”), and pay attention to cues about relationships (28–30). To enhance communication, our participants reported benefits of including other healthcare team members or having a designated person speak with patients’ families. This method allows physicians to concentrate on taking care of patients directly while others manage the needs and questions of families. Taking a multidisciplinary approach to family communication is essential and consistent with the Society of Critical Care Medicine Guidelines for Family-Centered Care (12).

This study has several limitations. First, we interviewed a convenience, snowball sample of ICU physicians who practiced at VA or academic hospitals and who predominately identified as White, which may reduce generalizability of findings to a broader sample of physicians and/or those who practice at community-based hospitals. Second, our sample was predominantly White; we were limited in our ability to develop an analysis that considered the race and ethnicity of physicians as factors in their perceptions of relationship, nor did our data provide insights into how prior experience serving minoritized patients and their families influenced their experiences as clinicians during the pandemic. These are important areas for future research to learn how to address disparities in care occurring during crises disproportionately impacting minoritized patients. We also did not include patient, family, or other staff perspectives, which are important to consider in future research.

In conclusion, the COVID-19 pandemic has threatened patient-clinician communication and the development of therapeutic relationships in the ICU, particularly among people holding minoritized racial and ethnic identities. We identified several facilitators to improve Patient centered communication that may help improve trust and foster therapeutic alliances between patients, families, and clinicians in the ICU setting.

ACKNOWLEDGMENTS

We thank Tara Thomas, BA, Danielle Apodaca, MA, and Felipe Rubim, MA, VA Portland Healthcare System.