Content of Tele-Palliative Care Consultations with Patients Receiving Dialysis

Abstract

Background:

Little is known about the content of communication in palliative care telehealth conversations in the dialysis population. Understanding the content and process of these conversations may lead to insights about how palliative care improves quality of life.

Methods:

We conducted a qualitative analysis of video recordings obtained during a pilot palliative teleconsultation program. We recruited patients receiving dialysis from five facilities affiliated with an academic medical center. Palliative care clinicians conducted teleconsultation using a wall-mounted screen with a camera mounted on a pole and positioned mid-screen in the line of sight to facilitate direct eye contact. Patients used an iPad that was attached to an IV pole positioned next to the dialysis chair. Conversations were coded using a preexisting framework of themes and content from the Serious Illness Conversation Guide (SICG) and revised Edmonton Symptom Assessment System—Renal.

Results:

We recruited 39 patients to undergo a telepalliative care consultation while receiving dialysis, 34 of whom completed the teleconsultation. Specialty palliative care clinicians (3 physicians and 1 nurse practitioner) conducted 35 visits with 34 patients. Median (interquartile range) duration of conversation was 42 (28–57) minutes. Most frequently discussed content included sources of strength (91%), critical abilities (88%), illness understanding (85%), fears and worries (85%), what family knows (85%), fatigue (77%), and pain (65%). Process features such as summarizing statements (85%) and making a recommendation (82%) were common, whereas connectional silence (56%), and emotion expression (21%) occurred less often.

Conclusions:

Unscripted palliative care conversations in outpatient dialysis units through telemedicine exhibited many domains recommended by the SICG, with less frequent discussion of symptoms. Emotion expression was uncommon for these conversations that occurred in an open setting.

Introduction

Patients with kidney failure or end-stage kidney disease (ESKD) have high morbidity and mortality rates, low quality of life, and experience high rates of intensive care use and hospitalization at the end of life.^1–5 There is growing awareness of the role that palliative care might play in dialysis care.⁶ However, there are few empirical data available about what is discussed during palliative care visits with patients receiving dialysis. In other settings, direct observation of palliative care conversations reveals discussions of symptoms and sources of suffering, prognoses and medical decision-making, communication preferences, and coping strategies.⁷ Although palliative care improves quality of life, and in certain populations is associated with extended survival,⁸ the critical elements of palliative care that lead to improved outcomes are not known, particularly among people receiving dialysis. Understanding the mechanisms of serious illness conversations has been identified as a research priority.⁹

Earlier research has focused on review of medical records to understand the content of palliative care consultations.^10,11 A study of in-person specialist palliative care consultations with 91 patients receiving dialysis found documentation of code status and goals of care discussions improved following the intervention, but it is uncertain what was actually discussed.¹² To date, there is little direct observation of palliative care conversations that would permit investigation of these elements.¹³ We sought to describe the content and processes of palliative care conversations among outpatients receiving dialysis.

Methods

The Institutional Review Board at The University of Vermont approved this study. The study was registered with Clinicaltrials.gov (NCT03744117).

Design

This was a single-arm clinical trial conducted in Vermont. In October 2018, we recruited clinicians from our Palliative Medicine division to conduct palliative care consultations while located at our academic medical center. We recruited patients between November 2018 and January 2020 from five dialysis units that are part of our nonprofit, university-affiliated medical center, starting with the most remote location; details have been prescribed.¹⁴ The target population comprised patients receiving in-center maintenance dialysis for kidney failure at satellite facilities. Informed consent procedures occurred chairside in the dialysis unit and participants provided written consent.

Patients were told the purpose of the study was to learn whether they think it is helpful to have a palliative care consult using telemedicine, what they think of the quality of telemedicine communication, and to understand if the consult has any impact on their health care choices in the future. They were told the main benefit may be talking with a palliative care clinician and the main risk was the potential loss of privacy by having the palliative care consult video recorded or other loss of confidentiality. Patients were given the option during the consent process to choose to do the consultation before or after their scheduled dialysis treatments, or at home if they had Internet service and a communication device.

Intervention

The intervention consisted of a one-time consultation between a palliative care clinician and patient. Consultations were scheduled with dialysis staff input as to ideal start time to avoid disruption with dialysis care (e.g., fistula cannulation or administration of medications). In general, consultations occurred ∼30 minutes after starting dialysis. Before the consultation, the research team provided the clinician with a patient summary that included years on dialysis, etiology of ESKD, and presence of advance directive/COLST on chart. This information was derived from nephrologist, dietician, and social worker notes in the dialysis electronic medical record, to which palliative care clinicians do not otherwise have access. To maximize privacy and audiovisual quality of the video recordings, the clinicians conducted the consultation from the Telepresence Studio in the Vermont Conversation Lab (www.vermontconversationlab.com). The lab is adjacent to the medical center and includes a 65-inch wall-mounted high-definition screen, studio lighting, and a pole-mounted low profile camera placed in front of the screen and adjusted to the clinician's direct line of sight to facilitate patient-participant experience of eye contact.

Patients used an iPad that was mounted on a portable stand (IV pole) and positioned next to the dialysis chair. Zoom was used for the telemedicine platform and the password-protected, medical center network was used for WIFI connection. Dialysis staff or research personnel brought the iPad to the patient at the time of the consultation, opened the televideo app, and returned at the end of the consultation to remove the iPad. Patients were offered headsets to cancel out ambient noise and to enhance privacy. Immediately preceding the consultation, research personnel confirmed the patient wished to proceed with the research study and then introduced the palliative care clinician. Participants received a nominal token of appreciation in the form of a $25 grocery store gift card for their time. Palliative care clinicians conducted unscripted consultations as per their usual practice and documented the conversation in the electronic medical record.

Measures

Patient participants completed a baseline questionnaire on demographics, quality of life, and health care experience. Characteristics included gender identification (woman/man/other), ethnicity (Hispanic/Latino origin or non-Hispanic), race (White, Black, Asian, Native Hawaiian or other Pacific Islander, American Indian or Alaskan Native, other), education (less than high school, high school or GED, one to three years college or technical school, bachelor's degree, master's or doctorate degree), religious affiliation (Christian, Jewish, Muslim, Hindu, Buddhist, Other, no formal affiliation), income (enough income for things you really need like food, clothing etc., in a typical month: all, most, some, none of the time).

Quality of life and health care were measured using the McGill Quality of life Single Item Scale (0–10, 10—excellent quality of life over past week), and the Heard and Understood Question (over the past year how much have you felt heard and understood by doctors, nurses, staff: completely, quite a bit, moderately, slightly, not at all). We also assessed over the past week, how much patients had been bothered by physical problems, emotional problems, uncertainty, and how much they felt at peace (not at all, slightly, moderately, quite a bit, extremely).

Analysis

We developed a preliminary coding scheme to identify content and processes in serious illness conversations as well as symptoms addressed, based on expert recommendations and to include these components in palliative care conversations with patients receiving dialysis.^15,16 This scheme included nine content items derived from the Serious Illness Conversation Guide (SICG).¹⁷ These content domains include illness understanding, prognosis sharing, information preferences, goals, fears/worries, sources of strength, critical abilities, trade-offs, and what family knows. In addition, we coded for the presence of any discussion about 11 symptoms identified by the revised renal Edmonton Symptom Assessment System (ESAS).

With respect to conversation process, we coded for the presence of summarizing statements, recommendations, emotion expression, and connectional silence.^15,18–21 Connectional silence was defined as a pause of at least two seconds in the conversation that is preceded by a moment of gravity, indicated by expression of a distressing emotion or by a topic identified by the speaker as hard to talk about (e.g., new unfavorable prognosis information).²¹ Length of the conversation was defined as time from when clinician first spoke to patient to sign off of the clinician.

Content and processes were coded by two independent trained researchers (K.L.C. and S.S.) by direct observation of the video recordings. One researcher (K.L.C.) was a nephrologist known to the clinician participants and some of the patient participants. The second researcher (S.S.) was not known to the participants but had relevant experience in dialysis. The presence or absence of content or processes was coded for the first 10 conversations. Minor revisions were made to the coding structure and the remaining 24 conversations were coded by both researchers. Sets of two to four conversations were reviewed at a time. Interrater agreement was assessed using kappa statistic; however, given the low prevalence of prognostic sharing, we also present percent agreement as Kappa values may underestimate agreement when prevalence is low²² (Supplementary Table S1).

All conversations were double-coded and any disagreements were resolved by reviewing code definitions derived from SICG literature^15,19 and direct observation of the video recording. If needed, a third researcher (R.G.) was available to discuss code definitions to help resolve coder disagreements. Final review of every third recording was performed to assess for drift in definition to ensure adherence to the codebook. To identify content or processes that were not determined a priori, we reviewed “field notes” made by coders.

One patient had two conversations recorded separated by 40 days owing to a technical problem. The content and processes of these two conversations were coded separately and results were consolidated as one conversation.

Results

We recruited four clinicians (three men, one woman; three MDs and one NP) to participate in video consultations with patients receiving dialysis. Between November 2018 and January 2020, we recruited 39 participants from five satellite dialysis units. Thirty-six completed the baseline survey and were scheduled for a consultation and 34 completed the consultation. All 34 completed the postconsultation survey. Reasons for noncompletion included: feeling unwell on day of televisit and declined to reschedule (n = 1), changed their mind (n = 4).

Among the enrolled participants, 22 were men, 14 were women; mean age for the entire group was 70.8 ± 10.9 years. Thirty-five self-identified as non-Hispanic White. Overall mean (standard deviation) quality of life was rated as 7.35 (±1.47), and more than half reported not being bothered at all by either physical or emotional symptoms in the past week. Forty-one percent (n = 14) felt completely heard and understood by doctors, nurses, and staff over the past year. Baseline characteristics are given in Table 1. No patients had family present during the consultation. Two patients chose to complete the televisit immediately before their dialysis shift in a separate room in the dialysis unit.

Table 1.

Baseline Characteristics of the Participants

	N = 36
Age (years), mean ± SD	70.8 ± 10.9
Female, n (%)	14 (39)
Non-Hispanic White, n (%)	35 (97)
College education, n (%)	14 (39)
Enough income to cover essentials this month, n (%)	13 (36)
Overall quality of life
0–10 scale, 10-highest	7.3 ± 1.5
Bothered by physical symptoms
1—not at all	13
2—slightly	9
3—moderately	7
4—quite a bit	7
5—extremely	0
Bothered by emotional symptoms
1—not at all	21
2—slightly	9
3—moderately	3
4—quite a bit	3
5—extremely	0
Bothered by uncertainty
1—not at all	15
2— slightly	9
3—moderately	9
4—quite a bit	3
5—extremely	0
Felt at peace
1—completely	6
2—quite a bit	16
3—moderately	7
4—slightly	5
5— not at all	1
Missing	1
Heard and understood over the past year by doctors, nurses and staff
1—completely	14
2—quite a bit	14
3—moderately	3
4—slightly	4
5—not at all	1

SD, standard deviation.

Content

Of the nine content items included in the SICG, sources of strength, was discussed most often (31/34 conversations, 91%), followed closely by critical abilities (30/34, 88%), understanding, family, fears and worries (each 29/34, 85%), and goals (28/34, 82%) (Fig. 1). Conversely, prognosis sharing occurred much less frequently (3/34, 9%). As given in Figure 2, the most commonly discussed symptoms were fatigue (26/34, 77%) and pain (22/34, 35%). A greater proportion of participants discussed more content domains of the SICG than the ESAS (Fig. 3). Six to seven content items were typically included throughout the consultations, whereas the number of symptoms discussed varied considerably. Furthermore, one consultation included no symptom discussion at all, yet others discussed 9 to 10 of the 11 possible symptoms obtained from the ESAS. In conversations high in symptom content (i.e., greater than six symptoms discussed) most (8/9) included at least six SICG domains. Whereas conversations high in SICG content (more than six items discussed) were less likely to also be high in symptom content (9/28).

FIG. 1.

Prevalence of SICG content in palliative teleconsultations with patients receiving dialysis. SICG, Serious Illness Conversation Guide.

FIG. 2.

Prevalence of symptom content using the revised ESAS—Renal in palliative teleconsultations with patients receiving dialysis. ESAS, Edmonton Symptom Assessment System.

FIG. 3.

SICG content was discussed more commonly than symptoms from the ESAS in teleconsultations with patients receiving dialysis.

Process features

Conversations universally included an overview of palliative care, examples of topics that could be discussed and assessing patient preferences and priorities around what was offered. Summarizing statements occurred in 85% of conversations. Recommendations were made in 82% of conversations and included advice to document their goals (10%), to discuss goals/wishes with family/health care proxy (25%), to spend time considering critical abilities (13%), and to reconsider choice of health care proxy (13%), and suggestions for symptom management (25%) or function/safety concerns (8%), and other (8%) (Table 2). Emotion was overtly expressed in only seven consultations. Among the 15 total episodes of emotion expression in 7 teleconsultations, sadness was the most frequent (6/15, 40%), followed by frustration (4/15, 27%), feeling overwhelmed (2/15, 13%), anger (2/15, 13%), and anxiety (1/15, 7%). At least one connectional silence occurred in 56% of the teleconsultations.

Table 2.

Illustrative Quotations of Recommendations by Palliative Care Clinicians via Telemedicine to Patients Receiving Maintenance Dialysis

• Talk with family • Critical abilities	“The only thing that I would suggest is, when you do talk to your husband, … maybe say, “Is it ok if we talk about some of the things that are critically important to me?,” so that if you are faced with decision making, you would know what a minimum would be that I would need.”
• Critical abilities	“Can I offer a few suggestions for you to think about? … Some of the things that are helpful for surrogates is to spend some time talking with them about what are some of the things that you hold dear with respect to your function, symptoms, what you would consider to be a minimum standard for a quality of life that is meaningful to you…”
• Talk with family • Critical abilities	“I would encourage you to try to clarify things a little bit with regard to your code status, with things like a breathing tube and CPR … I do think that those are worth thinking about and talking with your wife a little more about...”
• Talk with family	“Would it be ok with you if I had suggested a few ways of talking about that [end of life planning, wishes] that might help her [health care proxy] know how to advocate for you?”
• Talk with family • Document	“One thing I would suggest is that you and Steve and your mom, that you talk with them and share with them some of the things that are important to you, because if they have to make decisions for you, they would be wondering ‘what would [patient] say if she was faced with this decision?’” “I might suggest that your doctor or we provide you with those [health care proxy] forms so that you could write their names down and sign the forms so that it is your voice saying ‘these are the people for me’…”
• Function • Symptoms	“One thing that I would encourage you to do, in addition to talking with your primary care doctor about getting back to the aqua-therapy… I would make sure to talk to your primary care doctor about your anorexia…”
• Symptoms	“I'd recommend something that's got some urea in it to help with the itching a little bit more…also… anything that you could do to stop your skin from getting too dry...”

CPR, cardiopulmonary resuscitation.

The mean consultation time was 43 ± 17 minutes, range of 12–79 minutes.

Discussion

In this study we describe the content and processes of telemedicine-facilitated palliative care consultations in patients receiving outpatient dialysis. To our knowledge this is the first report on directly observed qualities of palliative care communication with patients on dialysis. We identified three main findings. First, we found that participants naturally addressed most domains of the SICG with some notable exceptions. In particular, information preferences and sharing of prognostic information by the clinician was rarely discussed. Second, we identified there was wide variability of attention to symptoms, but among those discussed, fatigue and pain were most prevalent. Third, summarizing statements and recommendations were highly prevalent, but emotion expression was less common.

There is scant literature on what is discussed in palliative care conversations with patients receiving dialysis. An ambulatory kidney palliative care program serving predominantly patients with stage 4–5 chronic kidney disease (CKD) reported through chart review the following prevalence of topics discussed during the first visit: coping with disease (95%), advance care planning (87%), symptom management (55%), prognostic awareness (27%), and information preference (11%).¹¹ Our directly observed findings are consistent with this clinic experience, with the exception of prognosis, which was found in only 9% of our conversations. Prognosis communication has been reported as frequently as 93% of palliative care consultations in the hospitalized (nondialysis) setting.⁷

We coded for the presence of prognostic sharing only if the clinician delivered prognostic information to the patient, as described by the SICG,¹⁵ which may explain why we noted it less frequently than the kidney palliative care and PCCRI studies, which used a more inclusive definition of any discussion about a future health state. An alternative explanation is that prognostic sharing is less common once patients have selected or are receiving maintenance dialysis as opposed to when patients with advanced CKD are considering options in a clinic setting or when seriously ill people are amid acute crisis in the hospital setting.

Symptom burden is high in dialysis patients.²³ It is not known how frequently symptoms are assessed in palliative care consultations with patients receiving dialysis. In the aforementioned kidney palliative care clinic, 49 of 55 advanced CKD patients had completed the Integrated Palliative Care Outcome Scale—Renal survey and 55% of the initial visits included symptom management.¹¹ In the limited literature on palliative care discussions in dialysis, report of whether symptoms are discussed is absent.¹² We found that symptoms were not discussed routinely or in a predictable format. Although our data are derived from the outpatient context, data from the Palliative Care Quality Network indicate that the condition prompting palliative care consultation was less frequently symptom based (pain management in 14%, other symptoms in 10%) than goals of care in 82%.²⁴

Barriers to discussing symptoms in patients receiving dialysis have been previously identified, including provider unawareness of symptoms and difficulty in symptom management.²⁵ In our study, symptoms may not have been discussed frequently because of lower levels of symptoms in a stable outpatient cohort, who were healthy enough to participate in this study. In addition, given the design of the intervention as a single consultation, both patients and providers may have been less likely to engage in review of symptoms and address medication changes if they anticipated there would be no scheduled follow-up.

We observed that emotion expression was uncommon in our study. There are several potential explanations for this finding. The palliative care consultation occurred while the patients were receiving dialysis treatments in a shared space with clinicians who were not known to the patients who may have felt hesitant to express emotion with a stranger and in a public setting. The research protocol prescribed having one palliative care consultation, which may have inhibited patients from expressing knowing there was not going to be a future relationship with the clinician. Lower frequency of emotion expression may also be related to the type of content that was absent, namely, prognosis sharing was infrequently discussed in these conversations.

The mode of conversation through telemedicine theoretically may have diminished the chance of emotion expression if patients felt uncomfortable with the technology, although it is also possible some patients would feel freer to express emotion with the distance created through telemedicine. Finally, emotion expression may have been less common owing to quasi-stable outpatient setting. Previous studies observed emotional distress in 59%–69% of audio-recorded palliative care conversations of conversations with hospitalized, seriously ill patients.^26,27 In contrast, negative emotion was reported in 17% of 415 audio-recorded outpatient oncology visits.²⁸

There are limitations to consider in our study. (1) Conversations occurred on dialysis for most patients and the degree to which they are able to engage may have been limited by alterations in cerebral blood flow caused by the dialysis procedure. This may have reduced the depth of conversation or other characteristics. (2) These conversations occurred through telemedicine as opposed to in-person, which may have altered what was discussed. It is not known whether patients might feel more or less comfortable discussing serious illness with a palliative care clinician through telemedicine as opposed to in-person.

Our previous data suggest that patients are accepting telemedicine as a vehicle for these discussions,¹⁴ but it is possible that some patients might not feel as inclined to participate in a meaningful serious illness conversation through telemedicine. (3) Similarly, our study utilized palliative care clinicians who were not known to the patients. This too may have influenced the content and processes of the conversations, although it is not known if quality of communication is similar. Although qualitative data indicate that patients would like to have serious illness conversations with their nephrologists or dialysis team, these conversations are infrequent, hence the impetus for this study. (4) The context of these conversations occurring as part of routine care, as opposed to in response to a change in clinical status, may have influenced content as well as emotion expression. (5) Finally, televisits were conducted in English, in one rural state with little racial/ethnic heterogeneity, potentially limiting the generalizability of our findings.

Conclusions

We describe the content and processes of palliative care consultations in outpatients receiving dialysis. Direct observation through video recording revealed elements consistent with much of the SICG but variable attention to symptoms. Future studies should investigate the relationship between serious illness conversations in the outpatient and inpatient setting, and what aspects of the conversation facilitates future decision-making among patients receiving dialysis.

Authors' Contributions

K.L.C., M.K.T., R.D.S., T.R., M.A.L., and R.G. contributed to the conception and design of the study, methodology, and interpretation of the data. K.L.C. acquired funding, wrote the original draft and collected the data. K.L.C. and S.S. conducted the formal analysis. All authors revised and edited the article.

Funding Information

Research reported in this publication was supported by the National Palliative Care Research Center Kornfeld Scholar Career Development Award (K.L.C.). R.G. was supported by the Bob and Holly Miller Endowed Chair in Palliative Medicine. Research was also supported by the National Institute of General Medical Sciences of the National Institutes of Health under Award No. P20GM135007: Vermont Center for Cardiovascular and Brain Health (K.L.C.).

Author Disclosure Statement

No competing financial interests exist.

Supplementary Material

Supplementary Table S1