Abstract
BACKGROUND:
Critical care guidelines recommend that clinicians provide emotional support to families of critically ill patients during family meetings. Little is known about how family member race impacts how emotions are expressed and supported in meetings.
RESEARCH QUESTION:
Are there differences in family members’ expression of emotion and clinicians’ provision of empathy in ICU family meetings involving Black and White family members?
STUDY DESIGN AND METHODS:
We conducted a directed content analysis of 40 audio-recorded meetings, matching 20 meetings with Black families and 20 meetings with White families on key characteristics including meeting length. Meetings included Black or White family members of patients receiving prolonged mechanical ventilation and critical care clinicians. We used an established coding scheme to code family expressions of negative emotion and clinician empathic statements. Two analysts, unaware of patient and family race, independently coded all meetings. After coding completion, we unmasked the data, compared code frequencies by family race, and calculated the percent of expressions of negative emotions followed by an empathic response.
RESULTS:
Family members were mostly middle-aged (mean [SD], 50 [10] years) and women (n = 29 [73%]). Physicians leading meetings were mostly young (mean [SD], 38 [7] years), men (n = 27 [68%]), and White (n = 36 [90%]); none were Black. Black families expressed fewer negative emotions than White families (median, 1 [interquartile range, 0-3] vs 4 [interquartile range, 2.5-7.5] expressions of emotions per meeting; P < .001). When families expressed negative emotions, clinicians infrequently responded with empathy to all families and did so less frequently with Black families compared with White families (15% vs 30% of family expressions; P = .099).
INTERPRETATION:
Our results indicate that racial differences exist in family expression of emotion and potentially in clinician empathic responses in ICU family meetings. Empathic communication warrants improvement, with greater improvement needed with Black families. Future interventions may be needed to enhance clinician provision of equitable emotional support.
Keywords: African American, Black, critical care, emotional support, empathy, family conference, family meeting, ICU, intensive care unit, racial differences, racial disparities, racial equity, White
Critically ill patients with severe respiratory failure often endure substantial morbidity in the ICU, and up to one-half die in hospital.1,2 Family members of these patients commonly experience distressing emotions,3-5 including fear, sadness, and anger, that can negatively impact their psychological health and ability to make surrogate decisions for critically ill loved ones.4-7 Recognizing this, critical care guidelines recommend that clinicians provide not only medical information (treatment options, prognosis) to families, but also emotional support, especially during family meetings.6,8 Verbal expressions of empathy, such as acknowledging emotions expressed by family members, are an important way for clinicians to provide emotional support. When families receive emotional support from ICU clinicians, they report less psychological distress, greater satisfaction with care, greater trust in clinicians, and higher-quality surrogate decision-making.9-12
Despite the importance of emotional support, little is known about how ICU clinicians provide such support to family members and if it differs by family member race.13 Racial disparities have been described in communication quality, trust, and the quality of death and dying in the ICU,14-17 and a growing outpatient evidence base suggests that clinicians provide less empathy to Black patients as compared with White patients.18,19 Our goal was to conduct an exploratory study to quantify how Black and White families express emotion to clinicians and how clinicians provide empathy to Black and White families during audio-recorded ICU family meetings.
Study Design and Methods
Design and Setting
We conducted a directed content analysis of 40 unique family meetings that were recorded as part of a multicenter, patient-level randomized clinical trial.20 The trial enrolled critically ill patients, their surrogate decision-makers (henceforth referred to as family members), and their ICU physicians.21 The trial occurred in medical and surgical ICUs at 4 medical centers in North Carolina, Pennsylvania, and Washington. Site institutional review boards approved the trial (Duke Coordinating Center Institutional Review Board Identifier: 00021965) (e-Appendix 1).
Population
Eligible patients were 18 years of age or older and had received mechanical ventilation for ≥ 10 days with no expectation of death or liberation from ventilation within the next 24 hours. One primary family member who self-identified as the most involved participant in surrogate decision-making was enrolled per patient. Each patient’s ICU physician was enrolled. Study participants provided written informed consent.
Context of Family Meetings
The parent trial tested a decision support tool for family members about prolonged mechanical ventilation. The tool provided prognostic estimates and descriptions of treatment goals without any coaching about discussing emotions with the medical team. Families randomized to the intervention viewed the tool, and then all families in both trial arms attended a scheduled meeting with the patient’s enrolled physician, with additional clinicians joining as able. Notably, these meetings were unscripted, because the trial did not prescribe an approach or script to families or clinicians for the meetings.
Data Collection
Meetings were audio recorded if all participants consented. Study staff observed meetings (as nonparticipants) and documented participants, length, study site, and topics discussed.
Sample for This Analysis
We included all 20 meetings with Black families who agreed to recording in the parent trial, and we randomly selected 20 meetings recorded with White families, matching on trial arm, study site, meeting length, and family characteristics that may influence family-clinician communication including age, gender, and relationship to the patient. Meetings occurred between 2013 and 2016.
Data Coding and Content Analysis
We conducted a directed content analysis, applying an established coding scheme to identify family expression of negative emotions and clinician empathic statements. The scheme incorporates a well-accepted framework used to teach clinicians ways to express empathy in serious illness communication training courses and has been applied reliably in prior studies.22-26 Two coders independently coded all family meetings by listening to recordings and reading transcripts. The coders were unaware of patient and family race and were instructed not to guess race when coding. Both were White, PhD-trained sociologists with extensive qualitative research experience (W. W., D. P.). Where discrepancies in coding emerged, the broader interdisciplinary research team discussed code application until consensus was reached. This team included a South Asian critical care physician (D. C. A.), a White critical care physician (M. E. M.), a Black sociologist (D. B. E.), a White sociologist (J. S.), and a White psychologist with experience with the coding scheme (K. I. P.).
Expression of Negative Emotions:
We identified verbal expressions of negative emotions (eg, fear, sadness, anger) by families.22 We classified them as direct expressions (eg, “I’m scared”) or indirect expressions that referenced an emotion without explicitly naming it (eg, “But do you have little concern that it’s going to be meningitis?” representing worry). For indirect expressions, a family member’s tone, pitch, and pace helped to identify underlying emotions.23 If a family member repeated an emotion, we coded it as a separate expression if a pause occurred in between (offering clinicians the opportunity to express empathy before the repetition).
Empathic Statements:
We defined empathic statements as verbal expressions by clinicians that acknowledged family emotions (eg, “You can’t help but be overwhelmed. This is sudden. He’s young.”). We identified these using the mnemonic NURSE: naming, understanding, respecting, supporting, or exploring the family’s feelings (Table 1).23-26 We coded empathic statements as empathic responses if they followed family expression of negative emotion; empathic statements not offered in direct response to family expression of negative emotion were coded as independent empathic statements. We counted an empathic response containing multiple empathic statements as a single empathic response.
TABLE 1 ].
Definition and Examples of Empathy Classified by the Naming, Understanding, Respecting, Supporting, or Exploring Mnemonic
| Type of Empathic Statement |
Definition | Sample Quotations | |
|---|---|---|---|
| Empathic Response | Independent Empathic Statement | ||
| Name | Name the family’s emotion | Family: “It was frightening when we first heard the word [tracheostomy].” Clinician: “When you first hear the word, it is. And it’s very overwhelming.” (Family meeting 25; site 3; White) |
Clinician: “I know it’s frustrating to come in and not look like there’s any progress.” (Family meeting 20; site 1; Black) |
| Understand | Express understanding of the family’s emotional response | Family: “I know you all, I know you and your team are doing everything you all can. Like I told you this morning, but it’s, it’s getting frustrating for me.” Clinician: “If this was my wife I would be frustrated too, right? Cause you want things to get better much faster.” (Family meeting 38; site 2; White) |
Clinician: “It’s really hard seeing her decline. She’s done so many things even as sick as she’s been.” (Family meeting 21; site 3; White) |
| Respect | Respect or praise the family’s expert role in the patient’s care | Family: “And I know soon the tracheotomy talk is gonna start. No offense, you can sit there and try to tell me it’s not, but I’ve got a feeling that eventually it’s gonna come if we don’t get her off the vent and I know she does not want that, but if it can be a temporary tracheotomy, I’ll probably do it, you know. But I’m starting to go against her wishes. I’ll handle that later down the road with her, you know.” Clinician: “Well, our hope is that none of that will be necessary, like we talked about, but I am mindful and I hear and listen and understand your concerns.” (Family meeting 38; site 2; White) |
Clinician: “He’s had a lot of support from you all. I know you’re there every night.” (Family meeting 2; site 1; White) |
| Support | Express support for the family’s needs | Family: “It makes me feel bad cause we kept pushing him and pushing him to like you know, be active, like, ‘Let’s get up, let’s walk around, let’s eat this, let’s…’ you know, not knowing that he really couldn’t.” Clinician: “It’s not like you did anything wrong. This is not anything that you guys did. This happens. You guys were doing what you were supposed to do, which is take care of your brother and love him and that’s what you were doing. So, this is not your fault or anything like that.” (Family meeting 40; site 4; Black) |
Clinician: “Our first job is to take care of him [the patient] and second job is to take care of you.” (Family meeting 1; site 1; White) |
| Explore | Ask the family to expand on their feelings or concerns | Family: “Well, I’ve told other people before that, you know, when he was first intubated the first time during this admission that he made me promise not to let anybody turn anything off. At least I think that’s what he said. But you know, he was just opening his eyes a little while ago again, so I guess my, you know, I’ve heard everything everybody’s been saying although I will say there are some mixed messages throughout this whole thing.” Clinician: “Tell me what mixed messages you’re getting.” (Family meeting 9; site 1; White) |
Clinician: “What worries you the most?” (Family meeting 10; site 1; Black) |
Quantitative Analysis
On completion of coding, we summed the codes for expressed emotion, empathic responses, and independent empathic statements and calculated total numbers, median number per meeting, and proportion of meetings with at least 1 code for each of the 3 code types. We also calculated our primary outcome: the percent of expressed emotions followed by an empathic response. We then unmasked the analytic team to family race and compared these outcomes in meetings with Black and White families.
We present data as frequency (percentage) for categorical variables and mean (SD) or median (interquartile range [IQR]) for continuous variables. We conducted exploratory statistical analyses comparing Black and White families for various outcomes, without and with adjustment for matching factors such as meeting length, study site, and trial arm to reduce the remaining imbalance between Black and matched White families. We used a generalized linear mixed-effects model with binomial distribution for expression-level outcomes (presence of empathic response, direct or indirect expressed emotions). For meeting-level outcomes, we used a generalized linear mixed-effects model with Poisson (number of expressed emotions per meeting, independent empathic statements per meeting) or binomial (presence of expressed emotion in meeting) distribution for count and binary data, respectively. As a check on the matching procedure, we used a linear mixed-effects model to compare meeting length between Black and White families. To account for variations between expression data within meetings and meetings by the same physician, we included meeting and physician as random effects for expression-level outcomes and physician as a random effect for meeting-level outcomes. Using these models with our sample size is supported by literature.27 We examined model assumptions and goodness of fit of each model using residuals and a calibration plot of fitted vs observed values and did not observe considerable deviation (low root mean squared error value). We considered 2-tailed P values of < .05 to be statistically significant. Given the exploratory nature of these analyses, we did not adjust the significance level for multiple comparisons.28 We performed all analyses using R software version 4.3.2 (R Foundation for Statistical Computing).
Results
The 40 patients had a mean (SD) age of 49 (19) years and more than one-half were men (n = 23 [57.5%]). Other than race, patient characteristics were similar among meetings with Black and White families (Table 2).29 The 40 primary family members had a mean (SD) age of 50 (10) years and most were women (n = 29 [72.5%]). Apart from race, characteristics were similar between Black and White family members (Table 2). A median of 2 family members (IQR, 1-2 family members) and 2 clinicians (IQR, 2-3 clinicians) participated in meetings (Table 3). Meetings with Black and White families were similar lengths (median, 21 minutes [IQR,12-32.5 minutes] vs 24.5 minutes [IQR,12.5-29.5 minutes]; P = .772) (e-Fig 1, e-Table 1). One-half of the meetings occurred at 1 study site. Topics commonly discussed were patient diagnosis, treatments, prognosis, and treatment goals and were similar between groups (e-Table 2). Meetings were led by enrolled ICU physicians, who had a mean (SD) age of 38 (7) years and predominantly were men (n = 27 [67.5%]) and White (n = 36 [90%]); none were Black (Table 3).
TABLE 2 ].
Participant Characteristics by Family Race
| Variable | Total (N = 40) | Black (n = 20) | White (n = 20) |
|---|---|---|---|
| Patient characteristics a | |||
| Age, y | 49 (19) | 47 (18) | 52 (21) |
| Women | 17 (42.5%) | 9 (45%) | 8 (40%) |
| Race | |||
| White | 19 (47.5%) | 0 | 19 (95%) |
| Black | 20 (50%) | 20 (100%) | 0 |
| Multiracial | 1 (2.5%) | 0 | 1 (5%) |
| APACHE II score | 23.5 (17.5-27.5) | 25 (16-32.5) | 21.5 (18-26.5) |
| Charlson Comorbidity Index score | 3 (1.5-5) | 3 (1-5) | 3.5 (2-6) |
| Randomized to intervention arm | 21 (52.5%) | 12 (60%) | 9 (45%) |
| Primary family member characteristics b | |||
| Age, y | 50 (10) | 48 (10) | 53 (9) |
| Women | 29 (72.5%) | 15 (75%) | 14 (70%) |
| Race | |||
| White | 20 (50%) | 0 | 20 (100%) |
| Black | 20 (50%) | 20 (100%) | 0 |
| Relationship to patient | |||
| Spouse or partner | 14 (35%) | 6 (30%) | 8 (40%) |
| Parent | 11 (27.5%) | 6 (30%) | 5 (25%) |
| Child | 9 (22.5%) | 5 (25%) | 4 (20%) |
| Sibling | 6 (15%) | 3 (15%) | 3 (15%) |
| Health literacyc | 5 (4-6.5) | 5.5 (3.5-8) | 5 (4-6) |
Data are presented as No. (%), mean (SD), or median (interquartile range). APACHE = Acute Physiology and Chronic Health Evaluation.
Abstracted from the electronic health record.
Family members completed enrollment questionnaires that captured demographic information and health literacy.
Self-assessed 3-item instrument with range of 3 through 15, with lower scores representing greater health literacy.29
TABLE 3 ].
Family Meeting Characteristics by Family Race
| Characteristic | Total (N = 40) | Black (n = 20) | White (n = 20) |
|---|---|---|---|
| No. of family members present | 2 (1-2) | 2 (1-3) | 2 (1-2) |
| No. of clinicians present | 2 (2-3) | 2 (1.5-3) | 2 (2-3) |
| Enrolled physician age, ya,b | 38 (7) | 39 (7) | 37 (8) |
| Enrolled physician sex, femalea | 13 (32.5%) | 6 (30%) | 7 (35%) |
| Enrolled physician racea | |||
| White | 36 (90%) | 16 (80%) | 20 (100%) |
| Black | 0 | 0 | 0 |
| Asian | 3 (7.5%) | 3 (15%) | 0 |
| Other | 1 (2.5%) | 1 (5%) | 0 |
| Length of meeting, min | 23 (12.5-30.5) | 21 (12-32.5) | 24.5 (12.5-29.5) |
| Site | |||
| 1 | 20 (50%) | 10 (50%) | 10 (50%) |
| 2 | 10 (25%) | 5 (25%) | 5 (25%) |
| 3 | 9 (22.5%) | 4 (20%) | 5 (25%) |
| 4 | 1 (2.5%) | 1 (5%) | 0 (0%) |
Data are presented as No. (%), mean (SD), or median (interquartile range).
Physicians completed enrollment questionnaires that captured demographic information. A total of 31 physicians led the 40 meetings: 23 physicians led a single meeting, 7 physicians led 2 meetings, and 1 physician led 3 meetings.
Three physicians did not report their age.
Expressions of Negative Emotions
The total sum of negative emotions expressed in meetings with Black families was less than in meetings with White families (n = 40 vs n = 100) (Fig 1). Within the first 10 minutes of meetings, Black families expressed fewer negative emotions than White families (e-Fig 2). The median number of emotions expressed per meeting was 1 emotion (IQR, 0-3 emotions) in meetings with Black families and 4 emotions (IQR, 2.5-7.5 emotions) in those with White families (Table 4). This difference was statistically significant (P < .001) (e-Table 1). Sixty-five percent of meetings with Black families contained at least 1 expressed emotion compared with 95% of meetings with White families, and this difference was statistically significant (P = .020). Most emotions in both groups were expressed indirectly (n = 25 [62.5%] Black families vs n = 60 [60%] White families; P = .601).
Figure 1 –
A, B, Expressions of negative emotions by families and empathic responses by clinicians by family race: counts of expressions and responses (A) and response percentages (B). We identified verbal expressions of negative emotions (eg, sadness, fear, anger, worry) by families during ICU family meetings. When ICU clinician responses included an acknowledgment of family emotions, we categorized them as empathic responses. A, Counts represent all instances of expression of negative emotion and empathic responses in 20 family meetings with White family members and 20 family meetings with Black family members. Meetings were matched on key characteristics including meeting length. B, Response percentages were calculated using expressions of negative emotion as the denominator. We found that Black family members expressed fewer negative emotions to clinicians than White family members (n = 40 vs n = 100). When families expressed negative emotions, we found that clinicians infrequently responded with empathy to all families and did so less frequently to Black as compared with White family members (15% vs 30%).
TABLE 4 ].
Expressions of Negative Emotion, Empathic Responses, and Independent Empathic Statements by Family Race
| Variable | Total | Black | White |
|---|---|---|---|
| Family expressions of negative emotions | 140 | 40 | 100 |
| Direct expressions | 55 (39.3%) | 15 (37.5%) | 40 (40%) |
| Indirect expressions | 85 (60.7%) | 25 (62.5%) | 60 (60%) |
| Negative emotions expressed per meeting | 2.5 (1-5.5) | 1 (0-3) | 4 (2.5-7.5) |
| Meetings with at least 1 negative emotion expressed | 32 (80%) | 13 (65%) | 19 (95%) |
| Clinician empathic responses | 36 (26%) | 6 (15%) | 30 (30%) |
| Empathic responses to direct expressions | 19 (34.5%) | 3 (20%) | 16 (40%) |
| Empathic responses to indirect expressions | 17 (20%) | 3 (12%) | 14 (23.3%) |
| Meetings with at least 1 empathic response, among meetings with at least 1 negative emotion expressed | 18 (56%) | 5 (39%) | 13 (68%) |
| Clinician independent empathic statements | 326 | 163 | 163 |
| Independent empathic statements per meeting | 7.5 (5-12) | 8.5 (5-12) | 7 (5-11) |
| Meetings with at least 1 independent empathic statement | 39 (97.5%) | 19 (95%) | 20 (100%) |
Data are presented as No. (%) or median (interquartile range).
Empathic Responses
Clinicians responded with empathy to only 26% (n = 36) of negative emotions expressed in all meetings (Table 4). Clinicians responded with empathy to 15% (n = 6) of emotions expressed by Black families and 30% (n = 30) of those expressed by White families (Fig 1). This difference did not reach statistical significance (P = .099) (e-Table 1). Overall, we found that clinicians responded to direct expression of emotions more often than indirect ones (34.5% vs 20%), a statistically significant difference in univariate (P < .001) but not multivariate (P = .112) models. In meetings with Black families, the proportion of empathic responses remained lower than those with White families regardless of whether the emotion was expressed directly or indirectly (direct: n = 3 [20%] vs n = 16 [40%]; P = .502; indirect: n = 3 [12%] vs n = 14 [23.3%]; P = .711). Understand statements were the most common empathic response type overall and for each racial group (e-Table 3).
Independent Empathic Statements
Clinicians expressed a median of 7.5 empathic statements (IQR, 5-12 empathic statements) independent of family expression of negative emotions per meeting overall. This pattern was consistent when interacting with Black and White families (median, 8.5 independent empathic statements [IQR, 5-12 independent empathic statements] vs 7 independent empathic statements [IQR, 5-11 independent empathic statements]; P = .747) (Table 4, e-Table 1). We found that Black families received more name, understand, and explore types of independent empathic statements, whereas White families received more respect and support statement types (e-Table 3).
Discussion
In this exploratory directed content analysis of audio-recorded ICU family meetings, we found preliminary evidence for racial differences in family members’ expression of negative emotions. Black families expressed fewer negative emotions to clinicians than White families did. Our results also raise the possibility that clinician empathic responses may differ by family member race. When families expressed negative emotion, clinicians responded with empathy infrequently to all families and to 15% of those expressed by Black families compared with 30% expressed by White families. This difference did not reach statistical significance and should be interpreted cautiously, given the small sample size and multiple comparisons. We also found that clinicians consistently provided empathic statements to both Black and White families at times when families did not express negative emotion.
Expressions of Negative Emotions
Our finding that Black families expressed less negative emotion than White families to ICU clinicians is consistent with studies showing that Black patients express fewer emotions than White patients to outpatient clinicians.18,19 Although our study did not examine explanations for this, several possibilities are worth considering. Some Black families may feel less inclined to express negative emotions because they may have learned to expect indifferent and unsupportive responses based on personal and historical experiences.30,31 Expressing emotions requires families to trust that revealing their vulnerabilities likely will benefit them or the patient. While expressing emotion could strengthen relationships with clinicians, families also face the risk that their emotions may go unnoticed or unsupported, potentially leading to heightened psychological distress or mistrust. Additionally, some Black families may experience more nonverbal behaviors from clinicians that signal a lack of interest in their emotions.32 These may occur during meetings or in prior bedside interactions, which can impact trust and therapeutic alliance.33 Cultural differences also may contribute to differential expression. Black families may place more emphasis on faith, hope, and being positive than White families.3,30 Consequently, clinicians may have few opportunities to demonstrate their active listening and accurate understanding of Black families through empathic responses. These could then become key moments in a family meeting that, when handled well, may change the dynamic of a relationship, building trust where it may be lacking.
We consider this finding exploratory, and larger studies are needed to confirm it. It is possible some Black families may express negative emotion in ways our coding scheme did not capture. Unmeasured differences also may be influencing our results. Meetings were scheduled as part of a trial, and some meetings may be more aligned with changes in patient condition, which could influence expression of emotion. Additionally, we lack information about family-clinician communication before enrollment. Families of patients with greater severity of illness may have already communicated with clinicians and expressed or processed emotions and may be less likely to express emotion if no new clinical information is shared in the meetings we analyzed.
Empathic Responses
We observed clinician empathic responses at a rate one-half as frequent for Black families compared with White families, raising a possibility of racial differences. We interpret this hypothesis-generating result with caution because this difference fell short of statistical significance, the sample size was small, and we did not adjust for multiple comparisons given our exploratory approach. Yet, we were likely underpowered to detect a difference, and such a difference would be consistent with prior research in other disciplines demonstrating a racial empathy gap.34-36 Recordings have shown that outpatient clinicians respond less frequently to expressions of emotion by Black patients compared with White patients.18,19 The ICU environment also likely amplifies unconscious bias, a key proposed mechanism of a racial empathy gap. The time pressures, uncertainties, high cognitive demand, and lack of prior relationships typical of the ICU increase the likelihood that well-meaning clinicians may unconsciously classify patients and families as within or outside their own group, including racial category.37 Out-group membership, such as a White clinician interacting with a Black family (common in our study and in the United States38), can diminish psychological bonds and human connectedness. Reduced connectedness may hinder clinician recognition of family emotions and subsequent empathic responses. Current practices may be different than when these meetings occurred, given the increasing focus on equitable communication in medical training. However, a simulation study from 2022 found that the communication practices of physicians under cognitive stress continue to be influenced by unconscious racial bias.39
This finding warrants a larger study with preplanned hypotheses to better understand any potential difference. If a racial difference in empathic responses exists, it has the potential to contribute to known racial disparities in communication quality, trust, and the quality of death and dying in the ICU.14-17 Without adequate emotional support, intense emotions disrupt medical information processing, hindering family ability to assess the burdens and benefits of treatment options.7,40,41 Conversely, when ICU clinicians provide emotional support, families report greater trust in their clinicians and enhanced surrogate decision-making that has the potential to increase patient goal-concordant care.9,12 Moreover, families express greater satisfaction with care and report fewer symptoms of psychological distress.11,12 Evidence-based approaches that mitigate unconscious biases and foster empathy, such as perspective taking,34,42,43 may be promising models for interventions aiming to enhance clinician empathic responses equitably.
Independent Empathic Statements
Our study also found that clinicians regularly made empathic statements to all families during moments when families did not express negative emotion, a novel finding that adds complexity to our understanding of potential racial differences in ICU clinician communication behaviors. Prior studies of racial differences in patient expression of emotion and physician responses have not focused on measuring these kinds of statements.18,19 We interpret this finding as evidence that well-meaning clinicians generally recognize how distressing the ICU experience can be for families and that they are trying to support families during a difficult time.
What remains unclear is the importance of these statements to families and whether the importance differs by race, particularly if physicians fall short in responding with empathy in moments when Black families express emotion. Responding with empathy when a family expresses emotion may help families feel more heard and understood.44 An empathic response demonstrates the careful listening and accurate understanding required to first recognize emotions, whereas independent empathic statements could be impersonal and memorized by rote. To respond to emotion with empathy in a racially equitable manner may require intentional perspective taking for racially discordant families as compared with the default social connectedness that may exist for racially concordant families. The ICU environment creates barriers to this, which may explain in part how a racial difference might exist in empathic responses but not in independent empathic statements. Conversely, it is possible that any empathic statement, even impersonal ones memorized by rote, still may be important to families. A study of ICU family meetings with predominantly White families found that more clinician empathic statements were associated with greater family satisfaction, and 70% of statements were not made in direct response to family expression of emotion.11 Unpacking this complexity requires future research; ideally, racially diverse families could be surveyed after family meetings to compare their perceptions of clinician empathy with clinician behaviors captured on recordings.
Strengths and Limitations
This study has several strengths, including listening to recordings along with reading transcripts, an interdisciplinary analytic team, and multicenter data. It also has limitations. First, the sample size was relatively small. Larger studies are needed to better understand our exploratory findings. Second, we could not assess the role of nonverbal expression of emotions by families or nonverbal expression of empathy by clinicians. Prior studies suggest that racial differences are present in nonverbal expressions of empathy by clinicians.32 Relatedly, it is possible some empathic statements coded as independent were actually offered by clinicians in response to family nonverbal expression of emotion. Whether this differs by family race is unknown. One study found that physicians had a harder time accurately identifying nonverbal emotional cues from patients from racial and ethnic minority groups compared with White patients.45 Video recordings could examine this in future research. Third, our study focused on family expression of negative emotions. Future investigations should examine variation in expression of positive emotions, hope, and faith, along with clinician responses, as the quality of communication in these areas may impact therapeutic alliance and trust.46 Fourth, Black families in this sample or the parent trial may not be representative of the general population. Finally, the absence of Black physicians in this sample precludes assessment of the influence of race-concordance.
Interpretation
Black families expressed fewer negative emotions to clinicians than White families did during ICU family meetings. When emotions were expressed, clinicians infrequently responded with empathy to all families and potentially did so less frequently to Black families compared with White families. Clinicians also regularly offered empathy to all families during moments when families did not express emotions. These exploratory findings warrant larger studies. Ultimately, the responsibility lies with clinicians to create environments where families feel welcome expressing emotion and to respond with empathy when they do. Future interventions may be needed to enhance clinician provision of equitable emotional support.
Supplementary Material
Take-Home Points.
Study Question:
Does family expression of emotions and clinicians’ empathic responses in ICU family meetings differ between Black and White families?
Results:
In this directed content analysis of 40 audio-recorded ICU family meetings, 20 Black families expressed significantly fewer negative emotions to clinicians than 20 White families in meetings matched on length, and clinicians responded less often with empathy to emotions expressed by Black families compared with those expressed by White families (15% vs 30%), a nonstatistically significant difference.
Interpretation:
Our results indicate that racial differences exist in how families express emotion and potentially in how clinicians respond empathically, suggesting that interventions may be needed to enhance clinician provision of equitable emotional support.
Acknowledgments
Role of sponsors:
The funder had no role in the design and conduct of the study; collection, management, analysis, or interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication.
Funding/Support
This study was funded by the American Thoracic Society [2021 Unrestricted Research Grant in Critical Care (M. E. M. and D. C. A.) and the National Institutes of Health [Grants R01HL109823 (C. E. C.), K24HL148314 (D. B. W.), K23HL164968 (D. C. A.), National Palliative Care Research Center 2023 Kornfeld Scholars Program (M. E. M.), and NCATS UL1TR001881 (T. K. N.)].
Financial/Nonfinancial Disclosures
The authors have reported to CHEST the following: C. L. H. reports financial support from the American Lung Association and the Centers for Disease Control and Prevention, and board membership of American Thoracic Society. T. K. N. reports financial support from the Agency for Healthcare Research and Quality. None declared (M. E. M., D. B. E., W. W., D. P., J. S., K. I. P., C. E. C., E. K. K., K. S. J., S. K., S. S. C., D. B. W., D. C. A.).
ABBREVIATIONS:
- IQR
interquartile range
Footnotes
Additional information: The e-Appendix, e-Figures, and e-Tables are available online under “Supplementary Data.”
Preliminary findings of this work were presented as an abstract at the American Thoracic Society International Conference, May 22, 2023, Washington, DC.
Data sharing statement:
Deidentified participant data can be obtained for any purpose by researchers whose proposed use of the data has been approved by the first or senior authors and Duke University Health System Institutional Review Board and after approval of a protocol and signed data access agreement by contacting Deepshikha.ashana@duke.edu.
References
- 1.Bellani G, Laffey JG, Pham T, et al. Epidemiology, patterns of care, and mortality for patients with acute respiratory distress syndrome in intensive care units in 50 countries. JAMA. 2016;315(8):788–800. [DOI] [PubMed] [Google Scholar]
- 2.Santus P, Radovanovic D, Saderi L, et al. Severity of respiratory failure at admission and in-hospital mortality in patients with COVID-19: a prospective observational multicentre study. BMJ Open. 2020;10(10):e043651. [Google Scholar]
- 3.Braun UK, Beyth RJ, Ford ME, McCullough LB. Voices of African American, Caucasian, and Hispanic surrogates on the burdens of end-of-life decision making. J Gen Intern Med. 2008;23(3):267–274. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Nelson JE, Hanson LC, Keller KL, et al. The voice of surrogate decision-makers. Family responses to prognostic information in chronic critical illness. Am J Respir Crit Care Med. 2017;196(7):864–872. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Schenker Y, Crowley-Matoka M, Dohan D, Tiver GA, Arnold RM, White DB. I don’t want to be the one saying ‘we should just let him die’: intrapersonal tensions experienced by surrogate decision makers in the ICU. J Gen Intern Med. 2012;27(12):1657–1665. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kon AA, Davidson JE, Morrison W, Danis M, White DB. Shared decision-making in intensive care units. Executive summary of the American College of Critical Care Medicine and American Thoracic Society policy statement. Am J Respir Crit Care Med. 2016;193(12):1334–1336. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Power TE, Swartzman LC, Robinson JW. Cognitive-emotional decision making (CEDM): a framework of patient medical decision making. Patient Educ Couns. 2011;83(2):163–169. [DOI] [PubMed] [Google Scholar]
- 8.Davidson JE, Aslakson RA, Long AC, et al. Guidelines for family-centered care in the neonatal, pediatric, and adult ICU. Crit Care Med. 2017;45(1):103–128. [DOI] [PubMed] [Google Scholar]
- 9.Lincoln TE, Buddadhumaruk P, Arnold RM, et al. Association between shared decision-making during family meetings and surrogates’ trust in their ICU physician. Chest. 2023;163(5):1214–1224. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Nembhard IM, David G, Ezzeddine I, Betts D, Radin J. A systematic review of research on empathy in health care. Health Serv Res. 2023;58(2):250–263. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Selph RB, Shiang J, Engelberg R, Curtis JR, White DB. Empathy and life support decisions in intensive care units. J Gen Intern Med. 2008;23(9):1311–1317. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Torke AM, Callahan CM, Sachs GA, et al. Communication quality predicts psychological well-being and satisfaction in family surrogates of hospitalized older adults: an observational study. J Gen Intern Med. 2018;33(3):298–304. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Ashana DC, Welsh W, Preiss D, et al. Racial differences in shared decision-making about critical illness. JAMA Intern Med. 2024;184(4):424–432. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Johnson KS. Racial and ethnic disparities in palliative care. J Palliat Med. 2013;16(11):1329–1334. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Lee JJ, Long AC, Curtis JR, Engelberg RA. The influence of race/ethnicity and education on family ratings of the quality of dying in the ICU. J Pain Symptom Manage. 2016;51(1):9–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Shen MJ, Peterson EB, Costas-Muñiz R, et al. The effects of race and racial concordance on patient-physician communication: a systematic review of the literature. J Racial Ethn Health Disparities. 2018;5(1):117–140. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Smith AK, Davis RB, Krakauer EL. Differences in the quality of the patient-physician relationship among terminally ill African-American and white patients: impact on advance care planning and treatment preferences. J Gen Intern Med. 2007;22(11):1579–1582. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Park J, Beach MC, Han D, Moore RD, Korthuis PT, Saha S. Racial disparities in clinician responses to patient emotions. Patient Educ Couns. 2020;103(9):1736–1744. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Park J, Saha S, Han D, et al. Emotional communication in HIV care: an observational study of patients’ expressed emotions and clinician response. AIDS Behav. 2019;23(10):2816–2828. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Improving decision making for patients with prolonged mechanical ventilation ClinicalTrials.gov identifier: NCT01751061. Updated May 2, 2019. http://clinicaltrials.gov/ct2/show/NCT01751061
- 21.Cox CE, White DB, Hough CL, et al. Effects of a personalized web-based decision aid for surrogate decision makers of patients with prolonged mechanical ventilation: a randomized clinical trial. Ann Intern Med. 2019;170(5):285–297. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Suchman AL, Markakis K, Beckman HB, Frankel R. A model of empathic communication in the medical interview. JAMA. 1997;277(8):678–682. [PubMed] [Google Scholar]
- 23.Pollak KI, Arnold RM, Jeffreys AS, et al. Oncologist communication about emotion during visits with patients with advanced cancer. J Clin Oncol. 2007;25(36):5748–5752. [DOI] [PubMed] [Google Scholar]
- 24.October TW, Dizon ZB, Arnold RM, Rosenberg AR. Characteristics of physician empathetic statements during pediatric intensive care conferences with family members: a qualitative study. JAMA Netw Open. 2018;1(3):e180351. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Back AL, Arnold RM, Baile WF, Tulsky JA, Fryer-Edwards K. Approaching difficult communication tasks in oncology. CA Cancer J Clin. 2005;55(3):164–177. [DOI] [PubMed] [Google Scholar]
- 26.VitalTalk. Responding to emotion: articulating empathy using NURSE statements. VitalTalk website. https://www.vitaltalk.org/guides/responding-to-emotion-respecting/ [Google Scholar]
- 27.Wiley RW, Rapp B. Statistical analysis in Small-N Designs: using linear mixed-effects modeling for evaluating intervention effectiveness. Aphasiology. 2019;33(1):1–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Althouse AD. Adjust for multiple comparisons? It’s not that simple. Ann Thorac Surg. 2016;101(5):1644–1645. [DOI] [PubMed] [Google Scholar]
- 29.Rosenberg A, Arnold RM, Schenker Y. Holding hope for patients with serious illness. JAMA. 2021;326(13):1259–1260. [DOI] [PubMed] [Google Scholar]
- 30.Chew LD, Bradley KA, Boyko EJ. Brief questions to identify patients with inadequate health literacy. Fam Med. 2004;36(8):588–594. [PubMed] [Google Scholar]
- 31.Chandler D. The underutilization of health services in the black community: an examination of causes and effects. Journal of Black Studies. 2010;40(5):915–931. [Google Scholar]
- 32.Wilson TK, Gentzler AL. Emotion regulation and coping with racial stressors among African mericans across the lifespan. Developmental Review. 2021;61:100967. [Google Scholar]
- 33.Lorié Á, Reinero DA, Phillips M, Zhang L, Riess H. Culture and nonverbal expressions of empathy in clinical settings: a systematic review. Patient Educ Couns. 2017;100(3):411–424. [DOI] [PubMed] [Google Scholar]
- 34.Parente VM, Reid HW, Robles J, et al. Racial and ethnic differences in communication quality during family-centered rounds. Pediatrics. 2022;150(6):e2021055227. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Drwecki BB, Moore CF, Ward SE, Prkachin KM. Reducing racial disparities in pain treatment: the role of empathy and perspective-taking. Pain. 2011;152(5):1001–1006. [DOI] [PubMed] [Google Scholar]
- 36.Forgiarini M, Gallucci M, Maravita A. Racism and the empathy for pain on our skin. Front Psychol. 2011;2:108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Trawalter S, Hoffman KM, Waytz A. Racial bias in perceptions of others’ pain. PLoS One. 2012;7(11):e48546. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Institute of Medicine Committee on U, Eliminating R, Ethnic Disparities in Health C. In: Smedley BD, Stith AY, Nelson AR, eds. Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. National Academies Press; 2003. [Google Scholar]
- 39.Lane-Fall MB, Miano TA, Aysola J, Augoustides JGT. Diversity in the emerging critical care workforce: analysis of demographic trends in critical care fellows from 2004 to 2014. Crit Care Med. 2017;45(5):822–827. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Gonzalez CM, Ark TK, Fisher MR, et al. Racial implicit bias and communication among physicians in a simulated environment. JAMA Netw Open. 2024;7(3):e242181. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Westendorp J, Stouthard J, Meijers MC, et al. The power of clinician-expressed empathy to increase information recall in advanced breast cancer care: an observational study in clinical care, exploring the mediating role of anxiety. Patient Educ Couns. 2021;104(5):1109–1115. [DOI] [PubMed] [Google Scholar]
- 42.Visser LNC, Tollenaar MS, van Doornen LJP, de Haes H, Smets EMA. Does silence speak louder than words? The impact of oncologists’ emotion-oriented communication on analogue patients’ information recall and emotional stress. Patient Educ Couns. 2019;102(1):43–52. [DOI] [PubMed] [Google Scholar]
- 43.Chapman EN, Kaatz A, Carnes M. Physicians and implicit bias: how doctors may unwittingly perpetuate health care disparities. J Gen Intern Med. 2013;28(11):1504–1510. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Todd AR, Bodenhausen GV, Richeson JA, Galinsky AD. Perspective taking combats automatic expressions of racial bias. J Pers Soc Psychol. 2011;100(6):1027–1042. [DOI] [PubMed] [Google Scholar]
- 45.Pollak KI, Davenport CA, Duck V, et al. Discriminatory and valuing communication behaviors in cardiology encounters. Patient Educ Couns. 2024;123:108224. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Coelho KR, Galan C. Physician cross-cultural nonverbal communication skills, patient satisfaction and health outcomes in the physician-patient relationship. Int J Family Med. 2012;2012:376907. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Deidentified participant data can be obtained for any purpose by researchers whose proposed use of the data has been approved by the first or senior authors and Duke University Health System Institutional Review Board and after approval of a protocol and signed data access agreement by contacting Deepshikha.ashana@duke.edu.