Emotional distress predicts palliative cancer care attitudes: The unique role of anger

Although palliative care can mitigate emotional distress, distressed patients may be less likely to engage in timely palliative care. This study aims to investigate the role of emotional distress in palliative care avoidance by examining the associations of anger, anxiety, and depression with palliative care attitudes.


| BACKGROUND
Individuals with cancer commonly experience emotional distress.
Approximately 50% of patients may experience some form of distress throughout the illness trajectory, including elevated levels of anxiety, depression, anger, or other psychological symptoms. [1][2][3] Not only does this emotional distress directly impact quality of life, but it may also have broader implications for patients' health-related behaviors and cancer outcomes. For example, those experiencing emotional distress may have difficulty managing their care, which may be associated with treatment nonadherence, uncontrolled symptoms, and subsequently, increased healthcare utilization and poorer survival. [3][4][5][6] The impact of emotional distress can be wide-reaching, and this investigation aims to examine whether its influence extends to patients' attitudes toward palliative care.
Palliative care is a form of supportive care for serious illnesses that focuses on improving patients' symptoms and distress rather than curing the underlying disease. 7 Research has shown its efficacy in improving patients' physical and psychological well-being and overall quality of life, yet palliative care is significantly underutilized. 8,9 This underutilization may be partially due to unfamiliarity, misinformation, and negative associations with this form of care. [9][10][11] Emotionally distressed patients may be particularly prone to holding these unfavorable associations, as distress can foster negative interpretations of ambiguous or unfamiliar stimuli like the concept of palliative care. 12 Such interpretations may engender negative emotions around palliative care, like fear or helplessness. Furthermore, those experiencing emotional distress are more likely to find negative emotions aversive, and thus avoid thoughts and actions that they believe could further contribute to negative mood. 13,14 This is coupled with the psychological process of "terror management," which leads individuals to reduce their mortality-related distress through defensiveness and avoidance. Those who experience higher levels of emotional distress may also exhibit more existential distress or death anxiety, and therefore may employ more avoidance around end-of-life care, including palliative care. 15 Thus, emotional distress could instill less favorable attitudes toward palliative care, which is particularly troubling given that palliative care can help alleviate such distress.
Prior research suggests that two commonly examined domains of emotional distress-depression and anxiety-are associated with less favorable attitudes toward palliative care. 10,13,14 However, these studies are limited conceptually in that none incorporated the third widely accepted domain of emotional distress-anger. [16][17][18] People often feel anger in response to frustration and pain, 18,19 common experiences in cancer, and anger can detrimentally affect attitudes, communication, relationships, and care utilization. 3,18 Whereas depression and anxiety tend to foster passive avoidance and withdrawal, anger uniquely motivates individuals to approach problems, often in a conflictual manner that actively angers others and deters them from offering support. 20

| Participants and procedure
The present sample consisted of adults with a history of cancer who were able to read and understand English. We recruited participants through two online methods: (1) ResearchMatch, an NIH-funded recruitment tool 23

| Demographic and health characteristics
Participants responded to questions about their demographic characteristics and health histories. Demographic information included age, gender, race and ethnicity, education, marital status, and financial strain. Health information included cancer type, presence of metastases, time since diagnosis, and other non-cancer health comorbidities.
First, we created raw total scores for each measure by summing the individual item responses. We then converted the raw total scores to PROMIS-generated T-scores that had been centered and calibrated on the United States general population. Additionally, we created a composite emotional distress score computed as the mean of the measures' T-scores, which is supported by prior psychometric work. 26 The measure displayed excellent internal consistency reliability for each of the anger (α = 0.913), anxiety (α = 0.914), and depression (α = 0.909) subscales.

| Palliative care attitudes
Participants completed the 9-item Palliative Care Attitudes Scale (PCAS-9). 22 The PCAS-9 provides brief instructions explaining palliative care, followed by three subscales containing three items each: emotional, cognitive, and behavioral subscales. The emotional subscale asks participants how stressful they would find palliative care (e.g., "how stressful would you find discussing emotions, like feeling sad, scared, or angry?"), the cognitive subscale asks participants how helpful they think palliative care would be (e.g., "do you think a palliative care consultation would help with physical quality of life?"), and the behavioral subscale asks participants how willing they would be to attend palliative care (e.g., "would you be willing to attend the consultation?"). We created subscale and total scores for the PCAS-9 by summing all item responses within each subscale and the entire measure, respectively (the emotional subscale is reverse scored for consistency in direction). Higher scores on each scale or subscale indicated more positive attitudes toward palliative care. The measure had good internal consistency (α = 0.746).

| Statistical analyses
Analyses first included descriptive statistics to characterize the sample's demographic and health characteristics, emotional distress levels, and palliative care attitudes. Next, we examined bivariate Pearson correlations among demographic and health variables, emotional distress, and palliative care attitudes. Finally, for hypothesis testing, we conducted multiple linear regression analyses to examine the association of emotional distress with palliative care attitudes while adjusting for covariates.
We employed two models within the main regression analyses.
The first model examined the impact of total emotional distress using the composite score (one independent variable) on palliative care attitudes (dependent variable). The second model simultaneously examined unique impacts of anger, anxiety, and depression (three separate independent variables) on palliative care attitudes (dependent variable). Both models used PCAS-9 total scores as the dependent variable and adjusted for relevant demographic and clinical covariates that were selected based on prior research 10,11 : age, gender, education level (presence of a Bachelor's degree), race and ethnicity (non-Latino/a white vs. all others), cancer type (breast and colorectal, the two most common types in the sample), presence of metastatic disease, and time since diagnosis.
Additionally, we conducted a set of sensitivity analyses examining the impact of anger, anxiety, and depression on PCAS-9 subscales (emotional, cognitive, and behavioral). Each of the three exploratory models contained anger, anxiety, and depression as predictors, as well as the same covariates used in the main regression analyses. All regression models met statistical assumptions required for analysis, including multicollinearity as assessed via variance inflation factor (threshold: <5). 27 3 | RESULTS      Furthermore, findings revealed a significant relationship between gender and palliative care attitudes, with women viewing palliative care more favorably than men. Prior research has found similar trends in palliative care attitudes, 11,31 and the present findings broaden this understanding by showing that attitudinal gender differences are more cognitive than emotional or behavioral. Specifically, women in this study were more likely to believe palliative care would be helpful in terms of physical quality of life, emotional distress, and survival, which contrasts with some prior findings on the actual benefit received from palliative care in women versus men. 32,33 Some studies suggest that men are more likely to experience quality of life improvement from early palliative care than women, 32,33 creating a paradoxical relationship between gender differences in beliefs versus documented benefits of palliative care.

| Bivariate associations
Although women are more likely to believe palliative care may benefit them, they are less likely to actually receive this benefit, 32,33 and vice versa among men. Given this contrast, more research is needed on how to make palliative care more acceptable to men and more effective for women. 32,33 Moreover, findings showed that patients who were younger, had less education, and were more financially strained also had higher emotional distress, including higher anger, anxiety, and depression.
Prior research has found similar correlations among those who are more anxious and depressed, 34,35 but the evidence connecting patients with elevated anger to these demographic characteristics is limited and variable. While some research within oncology has shown that younger patients experience more anger, other research has found no relationship with age, nor with education or financial strain. 18,36 This study reveals and clarifies relationships between anger and these demographic characteristics of interest.

| Study limitations
This was the first large study to examine the association between anger and palliative care attitudes. Although this sample included a wide range of cancer diagnoses, it lacked demographic diversity. As a cross-sectional study, we make no firm causal inferences. Additionally, the study did not assess participants' previous experience with palliative care, which may influence attitudes. However, prior

| Clinical implications
The results of this study suggest that those experiencing emotional distress, especially anger, may be more reluctant to engage in timely palliative care as a result of more negative palliative care attitudes. 38 Research supports a strong link between individuals' attitudes and subsequent behavior, 38 meaning that those who believe palliative care will be more stressful, less helpful, and see themselves as less likely to accept a palliative care referral will be less likely to ultimately engage with palliative care. Clinicians should be attentive to signs of emotional distress, as this not only indicates patient suffering, but also the potential to underutilize services that may be particularly helpful to them, such as palliative care. Fortunately, research has demonstrated the efficacy of psychosocial interventions to treat distress within oncology and palliative care populations. 39,40 In particular, this study suggests the potential relevance of anger, and clinicians may benefit from training on how best to manage and attend to patients' expressions of this emotion. 29