Peace, Equanimity, and Acceptance in the cancer experience (PEACE)
Validation of a scale to assess acceptance and struggle with terminal illness
Dr. Mack was supported in part by the Glaser Pediatric Research Network and the American Society of Clinical Oncology Career Development Award.
Fax: (617) 632-2270
Abstract
BACKGROUND.
The role of emotional acceptance of a terminal illness in end-of-life (EOL) care is not known. The authors developed a measure of peaceful acceptance at the EOL, and evaluated the role of peaceful acceptance in EOL decision-making and care.
METHODS.
The authors developed the Peace, Equanimity, and Acceptance in the Cancer Experience (PEACE) questionnaire to measure the extent to which patients with advanced cancer have a sense of peaceful acceptance of their terminal illness. The scale was administered to 160 patients with advanced cancer along with measures of other attributes that hypothetically are related to acceptance, including cognitive acceptance of terminal illness. EOL outcomes in 56 patients who died during the study also were examined.
RESULTS.
The 12-item PEACE questionnaire had 2 subscales: a 7-item Struggle With Illness subscale (Cronbach α = .81) and a 5-item Peaceful Acceptance subscale (α = .78). Both subscales were associated with patients' self-reported peacefulness (correlation coefficient [r] = 0.66 for acceptance [P <.0001]; r = −0.37 for struggle [P < .0001]). Struggle With Illness scores were associated with cognitive terminal illness acknowledgment (mean scores, 14.9 vs 12.4 for patients who were not aware that their illness was terminal; P = .001) and with some aspects of advance care planning (living will or healthcare proxy: mean scores, 13.9 vs 11.5; P = .02). In addition, among patients who had died, the use of a feeding tube at the EOL was associated inversely with Peaceful Acceptance (P = .015).
CONCLUSIONS.
The current study indicated that the PEACE questionnaire is a valid and reliable measure of peaceful acceptance and struggle with illness. Scores were associated with some choices for EOL care among patients with advanced cancer. Cancer 2008. © 2008 American Cancer Society.
Knowledge about prognosis is a prerequisite for end-of-life (EOL) decision making. Previous literature has demonstrated that patients who have unrealistically positive views of their prognosis tend to choose invasive measures at the EOL instead of care directed at comfort.1-3 This literature supports the idea that patients are the best individuals to make decisions for their care based on their personal values when they have accurate knowledge about prognosis. Therefore, prognosis communication has become a focus of palliative care.4-12
Previous studies have focused on cognitive understanding of prognosis—are patients aware that they are terminally ill?1, 2, 13 However, in addition to patients' cognitive appreciation of their prognosis is their emotional reaction to their terminal illness. The existential crisis evoked by knowledge of a terminal illness may result in considerable suffering14, 15 as patients examine their lives in light of illness and impending death. Patients confront losses on many levels, with altered life roles, lost aspirations, and awareness that their loved ones also are suffering.14 Some patients meet what Erikson16 referred to as the final stage of human development (typically occurring in late adulthood)—the crisis of confronting one's own mortality—with integrity. These individuals possess “an informed and detached concern with life itself in the face of death.”16 Such patients wouldappear better able to integrate these experiences by finding meaning at the EOL17, 18 and are able to maintain their sense of dignity.19 By contrast, other patients despair when confronted with their own mortality, which may manifest itself in feelings of fear and foreboding, injustice, anger, and rage at their illness and condition and the sense of a foreshortened future. We developed the Peace, Equanimity, and Acceptance in the Cancer Experience (PEACE) scale to assess these dimensions of peaceful acceptance and struggle with terminal illness.
We previously reported that patients with advanced cancer who were cognitively aware of their terminal status and who considered themselves “at peace” had lower rates of psychological distress and higher rates of advance care planning than patients who were not peacefully aware.20 This finding suggests that the emotional experience of coming to accept one's terminal illness may play an important role in EOL decision making and care.
Therefore, we sought to measure the construct of peace in greater depth with a focus on the existential issues mentioned above in the context of advanced cancer. We hypothesized that attainment of peace might be influenced by sociodemographic and disease characteristics and by intrapersonal, factors such as mental health, coping patterns, and spirituality. The PEACE scale was administered to 160 patients with advanced cancer, and PEACE scores were analyzed in relation to the patient's cognitive acceptance of terminal illness as well as these personal and disease attributes. In addition, we hypothesized that a sense of peace, in turn, may affect EOL care outcomes. We therefore evaluated relations between PEACE and EOL care attributes, such as location of death and use of invasive measures at the EOL, among 56 participants who died during the course of the study.
MATERIALS AND METHODS
Patients were recruited between August 2002 and May 2007 to the Coping With Cancer Study, a multi-institutional investigation of advanced cancer patients and their primary (unpaid) caregivers. Participating sites include Yale Cancer Center (New Haven, Conn), Veterans' Affairs Connecticut Healthcare System Comprehensive Cancer Clinics (West Haven, Conn), Memorial Sloan-Kettering Cancer Center (New York, NY), Simmons Comprehensive Cancer Care Center (Dallas, Tex), and Parkland Hospital Palliative Care Service (Dallas, Tex).
Patients who were diagnosed with cancer at participating sites were eligible to participate if they had advanced cancer (expected prognosis <1 year) with failure of first-line chemotherapy, if they were aged ≥20 years, if they were able to identify an unpaid caregiver, and if they had adequate stamina to complete the interview. Patient-caregiver dyads were excluded if either the patient or the caregiver met criteria for dementia or delirium by neurobehavioral cognitive status examination21 or if either the patient or the caregiver was unable to speak English or Spanish. All study participants provided written informed consent.
Patients were asked to participate in a baseline interview for which a $25 incentive was provided. Research staff from Yale University trained the interviewers. Interviews were conducted in English or Spanish and took approximately 45 minutes to complete. Chart review was used at enrollment and after death to obtain additional clinical information, such as comorbidity at enrollment, location of death, and use of invasive measures at the EOL. In addition, a postmortem questionnaire was administered within 2 weeks of death to a caregiver or nurse who had cared for the patient during his or her last week of life.
The PEACE scale was developed during the course of the Coping With Cancer Study and, thus, was not administered to the entire cohort. Of 338 patients who participated in the Coping With Cancer Study, the PEACE scale was administered to 160 patients beginning in March 2004. Therefore, in this report, we present results for the patients who completed the PEACE scale; information on the full Coping With Cancer Study cohort is available elsewhere.22
The institutional review boards of participating institutions approved study procedures. The measures described below were assessed in the patient's baseline interview.
The PEACE Questionnaire
The PEACE questionnaire was based on the researchers' experience conducting interviews with terminally ill patients along with clinical observations and a review of the relevant literature. The PEACE scale is a 12-item questionnaire that was developed to assess the patient's sense of acceptance, calmness, and peace as well as their sense of struggle or desperation concerning their illness. Initially, there 38 potential items were considered for inclusion in the PEACE scale, but the scale was reduced to the subset of 12 items with item-total correlations ≥0.30 when all 38 items were analyzed collectively. The 12 items that were retained are presented in Table 1. Questions focused on the extent to which patients were able to accept the cancer diagnosis and feel a sense of inner peace and on the extent to which patients struggled with the illness, such as by feeling angry about it or by feeling that the illness was unfair. Possible responses were 1, not at all; 2, to a slight extent; 3, to some extent; and 4, to a large extent.
| Subscale | Circle the number for the answer that best Describes how you are feeling now: | Item-total correlation | |||
|---|---|---|---|---|---|
| Not at all | To a slight extent | To some extent | To a large extent | ||
| Peaceful Acceptance of Illness Subscale | |||||
| 1. To what extent are you able to accept your diagnosis of cancer? | 1 | 2 | 3 | 4 | .55 |
| 2. To what extent would you say you have a sense of inner peace and harmony? | 1 | 2 | 3 | 4 | .66 |
| 3. To what extent do you feel that you have made peace with your illness? | 1 | 2 | 3 | 4 | .60 |
| 4. Do you feel well loved now? | 1 | 2 | 3 | 4 | .33 |
| 5. To what extent do you feel a sense of inner calm and tranquility? | 1 | 2 | 3 | 4 | .62 |
| Struggle With Illness Subscale | |||||
| 1. To what extent do changes in your physical appearance upset you? | 1 | 2 | 3 | 4 | .51 |
| 2. To what extent does worry about your illness make it difficult for you to live from day to day? | 1 | 2 | 3 | 4 | .60 |
| 3. To what extent do you feel that it is unfair for you to get cancer now? | 1 | 2 | 3 | 4 | .33 |
| 4. To what extent do you feel that your life, as you know it, is now over? | 1 | 2 | 3 | 4 | .45 |
| 5. To what extent do you feel angry because of your illness? | 1 | 2 | 3 | 4 | .64 |
| 6. To what extent do you think your illness has beaten you down? | 1 | 2 | 3 | 4 | .69 |
| 7. To what extent do you feel ashamed of, or embarrassed by, your current condition? | 1 | 2 | 3 | 4 | .63 |
Psychometric Properties
The 12-item PEACE scale had adequate internal consistency (Cronbach α = .85). However, analysis of the factor structure and corresponding scree plot indicated the presence of 2 subscales. The Peaceful Acceptance of Illness subscale (Cronbach α = .78) contains 5 items that measure the extent to which patients are able to accept the cancer diagnosis, have a sense of inner peace and harmony, have made peace with their illness, believe that they are well loved, and feel a sense of inner calm and tranquility. Item-total correlations for the Peaceful Acceptance of Illness subscale ranged from 0.33 to 0.66 (see Table 1 for individual item-total correlations). The 7-item Struggle With Illness subscale (Cronbach α = .81) asks patients to characterize the extent to which they feel that changes in appearance are upsetting, that worry about the illness makes it difficult to live from day to day, that it is unfair to have cancer, that life as they know it is over, and that the illness has beaten them down; and it evaluates anger, shame, and embarrassment about their condition. For the Struggle With Illness subscale, item-total correlations ranged from 0.33 to 0.69 (Table 1).
Concurrent validity was assessed by examining relations between the subscales and independent variables, including sociodemographic and disease attributes, intrapersonal characteristics, cognitive terminal illness acceptance, and EOL care attributes. Variables that were used in the analyses are described below.
Sociodemographic and Disease Attributes
Demographic information
Patients were asked to provide information on their sex, age, race/ethnicity, marital status, income, health insurance status, religion, and highest grade completed in school.
Symptom burden and comorbidity
Symptom burden was measured by using the Physical Symptom subscale of the McGill Quality of Life Questionnaire.23, 24 The Karnofsky score25 was used to describe the patients' current state of health, ranging from 100 (normal) to 0 (death), and the Zubrod score was completed as a summary measure of current health status based on 4 categories: activity, pain, food intake, and nausea. The research interviewer determined these scores at the time of enrollment. The patient's medical record, in consultation with the referring clinician, was used to determine the Charlson Index of Comorbidity score, a measure of baseline health status, in which higher scores indicate a greater burden of comorbid conditions.26
Intrapersonal Attributes
Psychiatric illness
The Structured Clinical Interview for the Diagnostic and Statistical Manual27, 28 was used to diagnose current major depressive disorder, generalized anxiety disorder, panic disorder, and posttraumatic stress disorder (PTSD). Screening questions were followed by the full modules for those patients who screened positive.
Spirituality
Spirituality was assessed by asking patients the following questions: “To what extent do you consider yourself a spiritual person?” The possible responses were “very spiritual,” “moderately spiritual,” “slightly spiritual,” and “not spiritual at all.” Response categories were dichotomized for analysis, so that patients who considered themselves very spiritual or moderately spiritual were classified as “spiritual,” whereas patients who considered themselves slightly spiritual or not at all spiritual were classified as “not spiritual.”
Coping
The 15-item Brief COPE instrument29 was administered to determine positive coping strategies, such as “trying to come up with a strategy regarding what to do” or “getting comfort and understanding from someone,” and negative coping strategies, such as “refusing to believe that it has happened” or “using alcohol or other drugs to help.” The items were combined to create 2 summative scores for positive and negative coping strategies.
Cognitive terminal illness acceptance
Patients were asked to “describe your current health status,” with response options of “relatively healthy,” “seriously but not terminally ill,” and “seriously and terminally ill.” This question was developed in response to interviews with terminally ill geriatric patients, because cognitive acceptance or acknowledgment of terminal illness was associated previously with the completion of do-not-resuscitate orders,20 the receipt of palliative care, and physician prognostic disclosure.30 Patients who described themselves as “seriously and terminally ill” were considered to have cognitive acceptance of the terminal illness; those who chose the other 2 descriptions were categorized as not cognitively accepting their terminal prognosis.
Advance care planning
Patients were asked whether they had completed a do-not-resuscitate order, a living will, or a durable power of attorney for healthcare.
Outcomes of EOL Care
For patients who had died (N = 56), a postmortem assessment, which included chart review and caregiver interviews, was performed. The location of death was recorded as occurring in the intensive care unit, hospital (nonintensive care), nursing home, inpatient hospice, or at home. We also recorded interventions that were used in the last week of life, including any care in the intensive care unit and the use of a ventilator, a feeding tube, or chemotherapy. Finally, caregivers were asked to report, “in your opinion, how would you rate the overall quality of the patient's death/last week of life?” with response options ranging from 0 (worst possible) to 10 (best possible.)
Statistical Methods
For binary independent variables (terminal illness acceptance, advance care planning, location of death, psychiatric illness, spirituality), relations were assessed with nonpaired t tests comparing mean scores on the Peaceful Acceptance and Struggle With Illness subscales. When independent variables were continuous (such as age, education, symptom burden, and functional status), relations were assessed by using linear regression or the Pearson correlation coefficient (r). When independent variables were ordinal (income and comorbidity), the Spearman correlation coefficient was used to describe correlations.
RESULTS
Patient Characteristics
Half of participating patients were women (Table 2), and the mean patient age was 63 years (standard deviation [SD], 12.5 years). Non-Hispanic white patients comprised 88% of the patients studied, with 5% Black and 7% Hispanic patients. Most patients were married (68%) and insured (82%). Most had some postsecondary education, with a mean ± SD of 13.2 ± 3.6 years completed in school.
| Characteristic | No. of patients (%)* |
|---|---|
| Women, N = 153 | 77 (50) |
| Mean age ± SD, y | 63±12.5 |
| Race/ethnicity, N = 153 | |
| White | 134 (88) |
| Black | 7 (5) |
| Asian | 1 (1) |
| Hispanic | 11 (7) |
| Married, N = 151 | 103 (68) |
| Income, N = 143 | |
| <$11,000 | 9 (6) |
| $11,000–20,999 | 16 (11) |
| $21,000–30,999 | 15 (10) |
| $31,000–50,999 | 31 (21) |
| $51,000–99,999 | 22 (15) |
| ≥$100,000 | 21 (14) |
| Do not know | 29 (19) |
| Health insurance, N = 152 | 124 (82) |
| Religion, N = 154 | |
| Catholic | 93 (60) |
| Protestant | 35 (23) |
| Other | 20 (13) |
| None | 6 (4) |
| Highest grade completed in school, mean±SD | 13.2±3.6 |
- SD indicates standard deviation.
- * N=160 patients except where indicated.
Mean patient follow-up from the date of enrollment was 179 days. Fifty-six patients had died and had postmortem assessments available at the time this report was being prepared. These patients lived a mean ± SD of 77 ± 68 days after study enrollment. The remaining 104 patients, who either were either alive or had died but did not have postmortem assessments available at the time this report was being prepared, were a mean ± SD of 234 ± 94 days from enrollment.
The Peaceful Acceptance and Struggle With Illness subscale scores had no relation with sex, race/ethnicity, marital status, income, religious affiliation, or education. Older patients tended to have lower scores on the Struggle With Illness subscale (scores decreased by 0.09 per increasing year of age; P = .01) and higher Peaceful Acceptance scores (scores increased by 0.04 per increasing year of age; P = .04). Patients with health insurance had lower Struggle With Illness scores than uninsured patients (15.2 vs 13.2; P = .05).
Relations Between Peace and Struggle With Illness
Responses to the single item “to what extent do you believe a sense of inner peace and harmony?” were associated closely with the total Peaceful Acceptance score (r = 0.66; P < .0001) and were associated inversely with the Struggle With Illness score (r = −0.37; P < .0001). Similarly, the total Peaceful Acceptance score was correlated inversely with the Struggle With Illness score (r = −0.54; P < .0001).
EOL Awareness and Planning
We observed no difference in Peaceful Acceptance scores between patients who were aware that they had terminal illnesses and those who were not (P = .32) (Table 3). Patients who were aware that they had a terminal illness, however, tended to score higher for Struggle With Illness (14.9 vs 12.4; P = .001) than those who were unaware. Struggle With Illness scores were higher among patients who had a living will, healthcare proxy, or durable power of attorney (13.9 vs 11.5; P = .02), even after adjustment for cognitive acceptance of terminal illness (adjusted means, 14 vs 11.8; P = .04).
| Characteristic | Peaceful acceptance subscale | Struggle with illness subscale | ||||
|---|---|---|---|---|---|---|
| Mean score (SD) | P | Mean score (SD) | P | |||
| Patients with characteristic | Patients without characteristic | Patients with characteristic | Patients without characteristic | |||
| EOL awareness | ||||||
| Cognitive terminal illness acceptance | 16.5 (2.7) | 16.9 (3.0) | .32 | 14.9 (4.4) | 12.4 (5.1) | .001 |
| EOL care planning | ||||||
| DNR order | 16.8 (2.7) | 16.4 (3.0) | .43 | 14.2 (4.9) | 13.1 (4.9) | .19 |
| Living will, healthcare proxy, or durable power of attorney | 16.5 (2.9) | 17.6 (2.2) | .09 | 13.9 (4.8) | 11.5 (4.7) | .02 |
- SD indicates standard deviation; EOL, end of life; DNR, do not resuscitate.
Psychiatric Illness
Thirteen patients met diagnostic criteria for major depression. Depressed patients had higher Struggle With Illness scores than patients who did not meet criteria for depression (P < .0001) (Table 4). Similarly, Struggle With Illness scores were higher in patients who had a current diagnosis of anxiety disorder (N = 4; P = .0004), PTSD (N = 8; P < .0001), or any major Axis I psychiatric disorder (N = 18; P < .0001). Peaceful Acceptance scores were lower in patients with current depression (P = .005), anxiety (P < .0001), PTSD (P < .0001), or any major Axis I psychiatric disorder (P < .0001).
| Variable | Peaceful acceptance subscale | Struggle with illness subscale | ||||
|---|---|---|---|---|---|---|
| Mean score (SD) | P | Mean score (SD) | P | |||
| Patients with psychiatric illness | Patients without psychiatric illness | Patients with psychiatric illness | Patients without psychiatric illness | |||
| Current major depression | 14.62 (3.15) | 16.92 (2.74) | .005 | 19.39 (3.64) | 13.08 (4.67) | <.0001 |
| Current anxiety disorder | 11.25 (2.99) | 16.90 (2.70) | <.0001 | 22.00 (2.16) | 13.33 (4.79) | .0004 |
| Current PTSD | 12.38 (2.83) | 16.99 (2.63) | <.0001 | 20.25 (3.20) | 13.23 (4.74) | <.0001 |
| Any current major psychiatric disorder | 13.61 (3.27) | 17.14 (2.51) | <.0001 | 19.61 (3.52) | 12.81 (4.50) | <.0001 |
- SD indicates standard deviation; PTSD, posttraumatic stress disorder.
Coping
Patients who used more negative coping strategies tended to have higher Struggle With Illness scores (r = 0.24; P = .003) and lower Peaceful Acceptance scores (r = −0.23; P = .004). The use of positive coping techniques was not associated significantly with struggle (r = 0.10; P = .23) or acceptance (r = 0.14; P = .08).
Spirituality
Most patients considered themselves either moderately spiritual or very spiritual (N = 112 of 144 patients; 78%) (Table 5). Patients who considered themselves either moderately spiritual or very spiritual tended to have higher Peaceful Acceptance scores (mean score ± SD, 17 ± 2.59; P = .049) and lower Struggle With Illness scores (mean score ± SD, 13.05 ± 4.79; P = .014).
| Subscale | Mean (SD) | P | |
|---|---|---|---|
| No. of patients who consider themselves very or moderately spiritual | No. of patients who consider themselves slightly spiritual or not spiritual at all | ||
| Peaceful Acceptance score | 17 (2.59) | 15.91 (3.31) | .049 |
| Struggle With Illness score | 13.05 (4.79) | 15.44 (4.70) | .014 |
- SD indicates standard deviation.
Symptom Burden and Comorbidity
Patients who reported higher symptom burdens had lower Peaceful Acceptance scores (r = −0.38; P < .0001) and higher Struggle With Illness scores (r = 0.53; P < .0001) (Table 6). Patients with poorer functional status also tended to have higher Struggle With Illness scores (Karnofsky: r = −0.32 [P < .0001]; Zubrod score: r = 0.35 [P < .0001]) and lower Peaceful Acceptance scores (Karnofsky: r = 0.17 [P = .03]; Zubrod score: r = −0.28 [P = .0005]).
| Measure | Correlation coefficient (p) | |
|---|---|---|
| Peaceful acceptance score | Struggle with illness score | |
| McGill Symptom Burden | −0.38 (<.0001) | 0.53 (<.0001) |
| Karnofsky score | 0.17 (0.03) | −0.32 (<.0001) |
| Charlson Comorbidity Index | 0.02 (0.82) | −0.16 (0.06) |
| Zubrod score | −0.28 (<.001) | 0.35 (<.0001) |
EOL Outcomes
Among 56 patients who had died at the time this report was prepared, 5 patients had a feeding tube in the last week of life. Patients with feeding tubes had lower Peaceful Acceptance scores (mean ± SD, 13.6 ± 3.3) than patients without feeding tubes (16.7 ± 2.6; P = .015). The use of a feeding tube was not associated statistically with the Struggle With Illness score (P = .69). Two patients received chemotherapy in the last week of life; chemotherapy use was not associated with either acceptance (P = .12) or struggle (P = .56). None of the patients studied were on a ventilator in the last week of life.
Among the patients who died, 43 patients died at home, 3 patients died in an inpatient hospice, and 9 patients died in the hospital, including 4 patients who died in the intensive care unit. Location of death, however, was not associated statistically with Peaceful Acceptance scores (P = .43) or Struggle With Illness scores (P = .19). Caregiver perceptions of the quality of the patient's death also were not related to peaceful acceptance (r = 0.12; P = .38) or struggle with illness (r = 0.09; P = .52). Peaceful Acceptance and Struggle With Illness scores also were not associated with proximity to death, as determined by the number of days between PEACE scale completion and death (Peaceful Acceptance: r = 0.09 [P = .48]; Struggle With Illness: r = −0.19 [P = .17]).
DISCUSSION
Cognitive acceptance of terminal illness plays a fundamental role in EOL decision making.1-3 We sought to examine acceptance of terminal illness as a broader construct, encompassing emotional adjustment to terminal illness. We created a valid and reliable scale for use in adult patients with terminal illness and observed that adjustment in our population had 2 major dimensions—peaceful acceptance and struggle with illness. Together, these constructs describe important aspects of the EOL experience for patients with cancer.
The resulting subscales of Peaceful Acceptance and Struggle With Illness had good internal reliability and criterion validity, the latter established by associations with patients' reports of coping, spirituality, and peace and harmony. In addition, the Peaceful Acceptance score was not simply reflective of cognitive acceptance of terminal illness, because the 2 were not associated significantly. Rather, the subscales appear to provide unique information concerning the EOL period for patients with advanced cancer.
The Struggle With Illness subscale scores tended to be higher in younger patients. This finding is consistent with Neugarten's notion of “off-time” life events resulting in more difficult adjustment31 and possibly the need to resolve Erikson's final life crisis16 before reaching the age-appropriate time of confronting death; that is, late adulthood. Patients with higher symptom burdens, more comorbid conditions, and poorer functional status also had higher Struggle With Illness scores. Along with the association with the physical burdens of illness, Struggle With Illness scores also were higher among patients with psychiatric conditions, including depression, anxiety, and PTSD, and were lower among patients who considered themselves highly spiritual. We believe that this construct captures physical and existential dimensions of distress associated with terminal illness and, as such, may be an important target for clinical interventions aimed at improving well being at the EOL.
It is noteworthy that struggle with illness may have some positive outcomes, including EOL care planning. Patients who had higher scores for struggle were more likely to have a living will, healthcare proxy, or durable power of attorney for healthcare. Although we cannot determine causation, we do hypothesize that patients who are struggling actively with issues around terminal illness may wish to make specific plans to ensure that their wishes are upheld. Those with higher Struggle With Illness scores also had worse Karnofsky scores, which suggests that they also were struggling physically with their illness. This may have heightened their appreciation that time was short and, thus, that they needed to plan for appropriate care before their death. Rather than being a purely negative aspect of the terminal illness experience, struggle with illness may be one aspect of emotional processing that allows patients to grapple with important issues around EOL care. This possibility deserves further exploration in future studies.
Similar to the Struggle With Illness subscale, the Peaceful Acceptance subscale was associated with psychiatric illness, with a lesser degree of peaceful acceptance in those patients who had depression, anxiety, PTSD, or any major psychiatric illness. These associations underscore the importance of psychiatric care for patients with significant psychiatric illness near the EOL. For patients with significant psychiatric comorbidity, the evolution of personal acceptance could depend on such treatment.
The Peaceful Acceptance subscale had other associations in common with the Struggle With Illness subscale, including spirituality and symptom burden. However, peaceful acceptance appeared to be a clinically important domain with some features that were distinct from struggle with illness. Perhaps most important, peaceful acceptance was associated with decreased use of feeding tubes at the EOL, with higher peaceful acceptance scores observed in patients who did not use feeding tubes in the last week of life. The use of feeding tubes was only 1 of 3 indicators of aggressive care that we studied. The use of chemotherapy and a ventilator in the last week of life were rare occurrences in our sample, however, suggesting that our study was underpowered to detect the effects for these other 2 outcomes. Follow-up of our full cohort through the time of death may allow us to determine whether acceptance is associated with these other important aspects of care. Nonetheless, although other studies have demonstrated that cognitive understanding of prognosis affects choices for care, we believe that this is the first study to find EOL care outcomes that relate to emotional processing of prognostic information.
This was a cross-sectional study; we administered the scale once to newly enrolled patients. Therefore, we can only speculate about which associations may indicate causation. For example, we have suggested that significant psychiatric illness may impact the ability of patients to reach a level of peaceful acceptance in a terminal illness. However, it is also possible that depression or anxiety could be exacerbated by emotional struggle with a terminal illness or that, in measuring psychiatric illness and struggle, we are simply measuring 2 features of a single construct. Whether treatment of psychiatric illness can help patients to resolve struggle with illness and to reach greater emotional acceptance is an important future question.
In addition, future research should consider the trajectory of struggle and acceptance over time. Our cross-sectional data revealed no association between proximity to death and either struggle or acceptance, suggesting that these constructs could represent stable traits rather than evolving emotional states. This possibility, if borne out in longitudinal studies, could be consistent with Erikson's work on the final stage of human development, in which the approach to death is met with either ego integrity or despair, depending on the personality of the individual encountering death in late adulthood.16 Alternatively, it is possible that acceptance and struggle could change over time in a manner analogous to the process of grieving.32 To our knowledge, little is known regarding how the stage theory of grief applies to the personal emotional processing of patients who are living with life-threatening illnesses. In this vein, our scales may be able to provide a window into the EOL period as a lived experience. If acceptance and struggle with terminal illness change over time, then they also may be targets for interventions to improve EOL care.
Additional limitations to this study also should be considered. Of the 160 patients who completed the PEACE questionnaire, 56 patients died and had postmortem assessments available at the time this report was being prepared. Follow-up of surviving patients to their time of death is planned, and examination of the relations between PEACE scores and outcomes of care in the full population will be important. In addition, following patients to death will allow us to determine how successfully we captured patients in their last year of life and, in consequence, how well our design allowed us to assess the EOL period for the full population under study. Our sample size also was too small for reliable multivariate analyses in the patients who had died. Complex interactions between psychiatric illness, spirituality, and care preferences are possible, however, and deserve attention in future studies. Finally, most participants were white and had health insurance, and >50% of patients were Catholic. Racial, ethnic, and cultural differences in EOL care preferences33-37 make future study in a more diverse population important.
We evaluated patients' sense of peace and equanimity in the cancer experience and observed that the separate dimensions of peaceful acceptance and struggle with illness marked the EOL experience. Cognitive death acceptance does not fully describe the process of coming to terms with death. Patients with psychiatric illness in particular may experience greater struggle and less peaceful acceptance of a terminal illness. In addition, peaceful acceptance is associated with decreased use of feeding tubes at the EOL, suggesting that emotional processing of a terminal diagnosis may have an impact on patients' choices for care. Acceptance and struggle as measured by the PEACE scale are clinically meaningful aspects of coming to terms with a terminal illness beyond the acceptance that takes place on a purely cognitive level. These constructs have direct relevance to EOL experience and decision making.
