Forgiveness,Depressive Symptoms,and Communication at the End of Life: A Study with Family Members of Hospice Patients

Abstract

Introduction:

Forgiveness has begun to receive empirical attention in end-of-life contexts, but primarily among patients. This study examined forgiveness issues and communication priorities among family members of hospice patients.

Methods:

Surveys were distributed to family members of home-care patients in a large not-for-profit hospice in the Great Lakes region of the United States. Family members wrote what they would like to say to patients before they died. They also rated the importance of several expressions (love, gratitude, giving and seeking forgiveness, saying farewell) and the extent to which they had already expressed these messages. Participants rated their depressive symptoms and the quantity of unresolved offenses committed by themselves and the patient.

Results:

Of 147 surveys returned by participants, 142 were usable. In comparison with forgiveness-related communications, expressions of love, gratitude, and farewell were more consistently rated important; yet many participants rated forgiveness (giving and seeking) as extremely important. If forgiveness was rated important but had not been fully expressed, participants reported more depressive symptoms (p<0.05). Also, unresolved offenses (by participants or patients) correlated positively with depressive symptoms (p<0.05).

Conclusion

: Although not as commonly endorsed as expressions of love and gratitude, forgiveness-related communications are seen as extremely important by many family members of hospice patients. If family members see forgiveness (granting or seeking) as important but have not completed the process, these unresolved issues are associated with depressive symptoms. This study suggests that unresolved offenses and forgiveness issues warrant assessment and clinical attention within families receiving hospice care.

Introduction

After suffering interpersonal hurts, individuals face decisions about forgiveness. When people forgive, they reduce or let go of bitter, resentful feelings and desires for revenge,¹ perhaps replacing negative feelings with positive ones.² Similarly, when individuals have committed offenses, they must decide whether to forgive themselves and seek forgiveness from others. An ability to forgive has been linked with better mental health,³ including lower depression levels.^4,5 Interventions to facilitate forgiveness have reduced depressive symptoms among many groups, including elderly women⁶ and people with histories of divorce,⁷ emotional abuse,⁸ and incest.⁹ The present study extended this work by examining forgiveness-related issues, including associations with depression, among family members of hospice patients.

Researchers in end-of-life care have begun to examine forgiveness. For example, a large study to assess quality of life among patients with human immunodeficiency virus (HIV) showed empirical support for a forgiveness subscale.¹⁰ A grounded theory study showed that for cancer patients (some receiving hospice care), forgiving others for past hurts (some cancer-related, some not) helped resolve tension between personal values and angry, bitter feelings.¹¹ Forgiveness themes have also been introduced into empirically validated palliative care interventions. One study found that 4 weeks of forgiveness therapy, in comparison with a wait-list control, led to reduced anger and increased forgiveness, hope, and quality of life among elderly, terminally ill cancer patients.¹² Another line of research includes forgiveness in a life completion intervention, which has shown effectiveness in a three-arm randomized controlled trial with seriously ill, hospice-eligible patients.^13,14

Forgiveness can involve multiple levels. Forgiveness at the individual or intrapersonal level involves changes in personal feelings and attitudes, whereas relational or interpersonal manifestations of forgiveness often involve communication between two or more persons.^15,16 To date, a disproportionate amount of research has emphasized the individual level.¹⁷ To deepen understanding of the interpersonal processes surrounding forgiveness, more research focused on relationship quality and interpersonal communication is needed. End-of-life contexts could provide a fruitful area in which to study this interpersonal dimension of forgiveness. Knowing that time for communication is limited, patients, family members, and other loved ones may feel pressed to decide what is most important to express before the patient dies.^18,19 Early case studies and anecdotal accounts suggest that forgiveness expressions could provide healing and life closure for patients and family members.^20,21

To our knowledge, forgiveness communications have not previously been examined among family members of hospice patients. However, such communications have received some empirical attention from the patient perspective. For example, a study of cancer patients in hospice home care showed that many rated forgiveness as unimportant or not applicable, whereas almost all rated love and gratitude as very important.²² These findings clarified that expressions related to forgiveness are not always seen as important by patients facing the end of life, perhaps because they have no important unresolved issues or because they do not value forgiveness. We expected to see similar patterns among family members, with love and gratitude endorsed more frequently than forgiveness.

A recent study of patients with advanced cancer (Prince-Paul, Zyzanski, and Exline, forthcoming)²³ assessed for unresolved forgiveness issues (i.e., seeing forgiveness of self or others as important but not completely worked through). A substantial minority reported unresolved forgiveness issues, including forgiveness toward others (15%), asking for forgiveness (18%), and/or self-forgiveness (18%).²³ Importantly, the presence of unresolved forgiveness issues was associated with greater distress, anger, and anxiety and lower scores on life completion. These findings suggest that forgiveness issues are common and can be clinically significant among patients with life-threatening illness.

Although not directly addressing the topic of forgiveness, prior research also suggests that family conflict is frequent and consequential in end-of-life contexts.^21–24 In one study of frail, low-income, older adults with advanced chronic disease, family conflict was present in 55% of the deaths.²⁴ Another study showed that conflict was common in families of patients dying from lung cancer,²⁵ and these conflicts predicted greater complicated grief after the death.²⁶ A large 1999 survey on end-of-life issues showed that more than 85% of patients, family members, physicians, and other care providers reported that it was very important for patients to resolve unfinished business with family or friends before they die.²⁷

The present study built on this prior work on conflict and forgiveness at the end-of-life by examining associations with depressive symptoms, which have shown negative associations with forgiveness in prior research. A main prediction was that perception of ongoing or active family conflicts, which we termed unresolved offenses, would be associated with greater depressive symptoms among family members of hospice patients.

Objectives and hypotheses

This study included four main hypotheses, all related to the overarching aim of examining the relevance of forgiveness for family members of hospice patients:

Hypothesis 1: Because of their link with unresolved offenses, forgiveness-related expressions will not be rated as consistently important as expressions of love and gratitude.

Hypothesis 2: Family members who assign some importance to forgiveness will report more unresolved offenses (by themselves or the patient) compared with those who see forgiveness as unimportant.

Hypothesis 3: Those who rate forgiveness (seeking or granting) as having some importance but who have not completely forgiven (or sought forgiveness) will report more depressive symptoms than: (a) those who have completed the process and (b) those who rate forgiveness as unimportant.

Hypothesis 4: The quantity of unresolved offenses reported (committed by either participants or patients) will correlate positively with participant depressive symptoms.

Methods

Design and analysis overview

The study was survey-based and used Likert scales. Content coding was done on one open-ended question that assessed what participants most wanted to say to patients. Analyses used quantitative methods (descriptive statistics, group comparisons, and correlations).

Sampling procedure

Participants were family members of patients enrolled in home hospice care through a large Midwestern not-for-profit hospice organization. Power analysis determined that a sample size of 123 would detect small correlations (r²=0.25) at a power level of 0.80. After being trained by the researchers, all nine home care teams distributed packets (one per household) to family members of patients during regular home visits. Packets included consent forms, the survey, a name/address form for compensation ($10 gift card), and a postage-paid envelope.

Measures

Family members answered demographic questions and items about the patient's type and length of illness, pain (1 item) and communicativeness (6 items). Participants rated their own depressive symptomatology on the 10-item Center for Epidemiological Studies Depression inventory (CES-D),²⁸ which has shown (a) good predictive accuracy (κ=0.97) compared with the 20-item version, (b) acceptable estimates of test-retest reliability (r=0.71), and (c) good internal consistency as assessed by Cronbach's alpha (α=0.92).²⁹

The survey included one open-ended question for content coding: “If you had a last chance to say something to your family member before he or she dies, what might you want to say?” followed by a text box.

Drawing from prior conceptual²⁰ and empirical²³ work on end-of-life communication, we developed Likert items to evaluate five types of communication: love, gratitude, forgiveness (seeking and expressing), and farewell. Participants read the prompt, “How important would it be for you to say something like this to your family member before s/he dies?” followed by these items: express love (“I love you”), express gratitude (“Thank you”), express forgiveness to your loved one (“I forgive you”), ask for forgiveness from your family member (“Please forgive me”), and say farewell (“Goodbye”). They rated the importance assigned to each message from 0 (not at all) to 5 (extremely). They also rated the extent to which they had already communicated each message from 0 (not at all) to 5 (totally).

Two items were developed to assess the presence of unresolved offenses. Participants read, “Can you think of things you have done (or failed to do) that might be causing some problems or hurt feelings now between you and your family member?” and “Can you think of things your family member has done (or failed to do) that might be causing some problems or hurt feelings now between you and your family member?” Responses were rated from 0 (no, none) to 5 (yes, many).

Results

Sample characteristics and descriptive statistics

From 298 packets distributed, 147 surveys (49%) were returned with consent forms. Five cases were deleted: two because the patient completed the survey and three because the patient had died, leaving 142. Table 1 lists demographics. Table 2 lists patient-related variables as reported by family members: age, diagnosis, length of illness and hospice care, pain ratings, and levels of alertness/communication.

Table 1.

Participant Demographics: Family Members of Hospice Patients

	N	%	M	(SD)	Median
Participant (family member) age in years	–	–	60.3	(13.8)	60.0
Gender
Male	36	26	–	–	–
Female	104	74	–	–	–
Marital status
Single	20	14	–	–	–
Married	97	69	–	–	–
Living w/romantic partner	1	1	–	–	–
Widowed	5	4	–	–	–
Divorced	18	13	–	–	–
Ethnicity
Caucasian	116	82	–	–	–
African American	20	14	–	–	–
Latino/Hispanic	1	1	–	–	–
Asian/Pacific Islander	2	1	–	–	–
Middle Eastern	2	1	–	–	–
Native American	2	1	–	–	–
Income levels
Under $15,000	17	13	–	–	–
$15,000–$24,999	25	20	–	–	–
$25,000–$49,999	41	32	–	–	–
$50,000–$74,999	22	17	–	–	–
$75,000–$99,999	12	9	–	–	–
$100,000–$149,999	4	3	–	–	–
$150,000–$199,999	4	3	–	–	–
$200,000 and up	2	2	–	–	–

Table 2.

Patient-Related Variables (as Reported by Family Members) ^a

	N	%	M	(SD)	Median
Patient age in years	–	–	77.3	(12.7)	79.0
Patient's relationship to participant
Husband	32	23	–	–	–
Wife	27	19	–	–	–
Mother	39	28	–	–	–
Father	22	16	–	–	–
Sister	4	3	–	–	–
Daughter	1	1	–	–	–
Son	2	1	–	–	–
Mother-in-law	3	2	–	–	–
Father-in-law	1	1	–	–	–
Grandmother	1	1	–	–	–
Grandfather	2	1	–	–	–
Aunt	3	2	–	–	–
Uncle	1	1	–	–	–
Other relationship	2	1	–	–	–
Patient diagnosis^b
Cancer	61	43	–	–	–
Dementia/Alzheimer's	45	32	–	–	–
Heart disease	33	23	–	–	–
Lung disease	25	18	–	–	–
Kidney disease	5	4	–	–	–
Multiple system/organ atrophy	4	3	–	–	–
Diabetes	4	3	–	–	–
Stroke	3	2	–	–	–
Autoimmune disease	2	1	–	–	–
Length of illness (years)	–	–	4.5	(5.9)	2.0
Hospice length of stay (months)	–	–	6.3	(9.0)	3.0
Patient's alertness and ability to communicate (0–5 scale)	–	–	3.4	(1.2)	3.5
Rating of patient's pain (0–10 scale)	–	–	3.6	(2.9)	3.0

All participants were family members of hospice patients. The patient-related data in this table are based on reports by the family members.

Totals exceed 100% because some participants reported multiple diagnoses for patients. Diagnoses were listed only if they were endorsed by two or more participants.

Open-ended responses

Our first aim was to evaluate what family members would like to say to their loved ones before they died, using an open-ended question (“If you had a last chance to say something to your family member…”). The first author read responses (n=127; 89% of sample) to identify common themes and create categories for content coding. Two research assistants, blind to hypotheses, coded responses (yes/no for each theme). Disagreements were resolved through discussion and consultation with the first author. The two raters showed good reliability (κs from 0.87 to 1.0; 99% agreement; 15 disagreements from 1524 decisions). Table 3 summarizes the coding results. In addition to love, gratitude, forgiveness, and farewell (the themes used in our Likert items), other themes included praise or affirmation, reassurance, afterlife reunions, and expressions of how the patient would be missed.

Table 3.

What Participants Wanted to Say to Patients: Open-Ended Responses by Category (in Decreasing Order of Frequency)

Category	Sample statement	N	% ^a
Love	I love you.	86	68
Praise/Affirmation	You are my role model.	39	31
Gratitude/Thanks	Thanks so much for almost 60 wonderful years.	28	22
Will miss you	I will miss you terribly.	21	16
Reassurance	Don't worry, sweetheart, about anything or anyone.	9	7
Hanging on	Please don't leave me.	2	2
Letting go	If you feel ready, you can go home.	11	9
See you again	See you when I get there.	11	9
Forgiveness	All is forgiven and forgotten.	2	2
Apology	I'm sorry for any time I offended you in the past.	2	2
Seek forgiveness	I ask you to forgive me.	1	1
Offense by family member	Why didn't you ever let me go away to college?	1	1
Goodbye/farewell	Farewell, my love.	1	1

These percentages are based on the number of participants who wrote something for this open-ended question (n=127; 89% of sample).

Participants were much more likely to endorse themes of love or gratitude than themes related to forgiveness (all ts≥5.21, ps<0.001), supporting Hypothesis 1. Only two participants referred to forgiving the patient; one mentioned an offense by the family member. Only one participant reported directly asking for forgiveness, although two others apologized.

Likert items: Importance ratings

What did the five Likert items suggest about the importance assigned to forgiveness expressions? As Table 4 shows, Hypothesis 1 received further support. Participants were much more likely to rate love and gratitude as extremely important than to rate forgiveness expressions as extremely important (all ts≥9.34, ps<0.001). (Mean comparisons are not reported because of nonnormal distributions.) Forgiveness items were distributed bimodally with responses clustering around the scale endpoints. Most participants reported that expressing forgiveness was either “not important at all” (43%) or “extremely important” (37%). Similarly, for seeking forgiveness, predominant responses were “not important at all” (40%) or “extremely important” (38%). This bimodal pattern was not found on other items, all of which were rated extremely important by most participants (love: 86%; gratitude: 80%; farewell: 68%). Similar patterns were shown on the items asking whether participants had already made these expressions (bottom of Table 4), although many participants had not yet said farewell.

Table 4.

Importance and Completion of Five Types of Messages: Ratings of Likert Items by Family Members

Importance of saying this	0 Not at all	1	2	3	4	5 Extremely
Love (“I love you”)	1 (1%)	5 (4%)	0 (0%)	9 (6%)	5 (4%)	119 (86%)
Gratitude (“Thank you”)	2 (2%)	5 (4%)	1 (1%)	8 (6%)	11 (8%)	110 (80%)
Express forgiveness (“I forgive you”)	56 (43%)	9 (7%)	3 (2%)	8 (6%)	6 (5%)	49 (37%)
Ask for forgiveness (“Please forgive me”)	52 (39%)	10 (8%)	5 (4%)	9 (7%)	6 (4%)	50 (38%)
Farewell (“Goodbye”)	13 (10%)	3 (2%)	2 (2%)	11 (8%)	14 (10%)	90 (68%)

Already said this	0 No, not at all	1	2	3	4	5 Yes, totally
Love (“I love you”)	4 (3%)	1 (1%)	5 (4%)	5 (4%)	6 (4%)	117 (85%)
Gratitude (“Thank you”)	8 (6%)	7 (5%)	8 (6%)	11 (8%)	15 (11%)	89 (64%)
Express forgiveness (“I forgive you”)	59 (46%)	10 (8%)	6 (5%)	12 (9%)	2 (2%)	39 (30%)
Ask for forgiveness (“Please forgive me”)	59 (46%)	13 (10%)	6 (5%)	9 (7%)	6 (5%)	36 (28%)
Farewell (“Goodbye”)	66 (49%)	10 (7%)	4 (3%)	11 (8%)	4 (3%)	40 (30%)

Unresolved offenses

One-third (33%) of participants reported at least one unresolved offense by the patient; 46% reported at least one unresolved offense they had committed themselves. These offense variables showed positive skews (1.6 for offense by patient, 1.3 for offense by participant), which were reduced through square-root transformations before analyzing.

Consistent with Hypothesis 2, those who assigned some importance to expressing forgiveness (n=66) reported more unresolved offenses by the patient (transformed M=0.6, standard deviation [SD]=0.8) than those who rated forgiveness unimportant (transformed M=0.3, SD=0.6; n=51), t (115)=2.26, p<0.05. Similarly, those who assigned some importance to seeking forgiveness (n=77) reported marginally more unresolved offenses by themselves (transformed M=0.7, SD=0.8) than those who saw seeking forgiveness as unimportant (transformed M=0.4, SD=0.7; n=52), t (127)=1.89, p=0.06.

Depressive symptoms

On average, participants reported moderate levels of depressive symptoms (M=12.3, SD=6.3, α=.83, range 0 to 30). In support of Hypothesis 3, planned comparisons confirmed that among those who assigned some importance to seeking forgiveness, depressive symptoms were greater for those who had not totally sought forgiveness (M=14.5, SD=5.6; n=44) than for those who had (M=11.0, SD=5.9; n=35), t (77)=2.68, p<0.01, or those who saw seeking forgiveness as unimportant (M=10.9, SD=6.0; n=52), t (94)=2.98, p<0.01. A similar pattern was found for expressing forgiveness, offering further support for Hypothesis 3: Among those who assigned some importance to expressing forgiveness, those who had not completely expressed forgiveness (n=38) reported more depressive symptoms (M=14.6, SD=5.2) than those who had completely expressed forgiveness (M=11.7, SD=6.4; n=36), t (72)=2.11, p<0.05, or those who saw expressing forgiveness as unimportant (M=10.7, SD=6.0; n=56), t (92)=3.20, p<0.01.

We also examined Hypothesis 3 in another way: Using a cut score of 11 on the CES-D (indicating clinical depression), 57% of participants were depressed. Those who saw forgiveness-seeking as important but not fully expressed had greater odds of clinical depression (75%) than those who saw forgiveness-seeking as unimportant (48%), t (94)=2.78, p<0.01, or had completely sought forgiveness (49%), t (77)=2.48, p<0.05. Similarly, participants who saw expressing forgiveness as important but had not fully done so had greater odds of depression (76%) than those who saw forgiveness as unimportant (46%), t (92)=2.99, p<0.01, or who had completely expressed it (53%), t (72)=2.16, p<0.05), These results supported Hypothesis 3.

Finally, there was a positive correlation between depressive symptoms and the number of unresolved offenses (offenses by patients, r [126]=.19, p<0.05; offenses by participants, r [139]=0.29, p<0.001). These results supported Hypothesis 4.

Discussion

This study builds on prior work on forgiveness at the end of life by focusing on family members of hospice patients. The aim was to begin to assess the relevance of forgiveness-related issues within this population. We examined several indicators of relevance: importance ratings by participants and associations with depressive symptoms.

Importance of expressing and seeking forgiveness

As predicted, forgiveness-related expressions were clearly not the most important messages that participants wanted to communicate to their dying family members. Messages of love and gratitude were more frequently endorsed than forgiveness-related communications (supporting Hypothesis 1), as shown by both open-ended responses and quantitative ratings.

Yet it might not be accurate to rely too heavily on these averaged ratings of the importance of forgiveness. The variables on expressing and seeking forgiveness showed strong bimodal patterns. Many participants endorsed forgiveness as “extremely important”; however, a roughly equal number of participants rated them as “not important at all.” As proposed (Hypothesis 2), those who rated forgiveness communications as unimportant reported relatively few unresolved offenses. People may also find it socially undesirable to mention or endorse forgiveness issues because such issues imply the presence of interpersonal offenses. Love and gratitude, which are more uniformly positive, may seem more socially acceptable to report.

Links with depressive symptoms

Those who saw forgiveness-related expressions as having importance but had not totally forgiven or asked for forgiveness reported more depressive symptoms than those who saw forgiveness as unimportant or had completely expressed it (supporting Hypothesis 3). In clinical terms, then, it seems important to consider the importance family members assign to forgiveness as well as whether forgiveness has been completed. When people see forgiveness communications as relevant but have not completed them, they report distress—perhaps due to a sense of “unfinished business.” But we remain cautious not to draw causal inferences from these cross-sectional, nonexperimental data. Depressive symptoms could impede individuals from expressing important messages to dying family members, or a third variable (e.g., stress, fear, severe physical pain of the patient) might explain the association.

These results also suggested that unresolved offenses by either family members or patients are important to assess in end-of-life contexts. The more unresolved offenses participants reported, the greater their depressive symptoms (supporting Hypothesis 4). These findings, though preliminary, suggest that helping family members work through unresolved conflicts with hospice patients could help reduce their distress as they prepare for the loved one's death.

Limitations

This study focused on one predictor of whether participants saw forgiveness as important: the quantity of unresolved offenses. Clearly, other factors could affect these importance ratings. For example, people can differ in their interpretations of the word “forgiveness” and in their opinions about whether forgiveness is desirable. One who believes that forgiveness requires minimizing or forgetting serious offenses, for instance, may not see forgiveness as a viable option. Others might rate forgiveness as unimportant because they harbor anger that they see as beneficial or necessary. Consistent with this reasoning, forgiveness interventions often spend considerable time validating and processing anger before the topic of forgiveness is raised.^30,31

This study used a survey that was given only once. In future work, it would be useful to use longitudinal designs, particularly those that follow family members into bereavement. Several measures were designed for purposes of this study and should be subjected to formal tests of reliability and validity. Also, because this study was conducted within the context of hospice home care, the findings may not generalize to other populations. The study was not designed to provide a complete picture of the family system; instead, only one family member completed the survey. In most cases, the survey was completed by someone close to the patient, because those who were completely estranged from the patient would typically not be present in the home to receive the survey. Thus, the numbers might represent an underestimate, at least in terms of the number of unresolved offenses.

Conclusion

This study suggests the importance of addressing relational conflicts and forgiveness issues when working with family members of hospice patients—but only after assessing its importance to the family members. Although communication is at the core of end-of-life care, interpersonal hurts and offenses can be challenging to address. By drawing from research on forgiveness, interdisciplinary team members, patients, and family members can gain knowledge to facilitate effective communication and emotional healing in end-of-life contexts.

Footnotes

Acknowledgments

This study was supported by a grant from the Fetzer Institute.

Author Disclosure Statement

No competing financial interests exist.

References

Enright

, Freedman

, Rique

. The psychology of interpersonal forgiveness. Enright

, North

. Exploring Forgiveness. Madison: University of Wisconsin Press, 1998; 46–63.

Worthington

Jr. , Scherer

. Forgiveness is an emotion-focused coping strategy that can reduce health risks and promote health resilience: Theory, review, and hypotheses. Psych Health, 2004; 19:385–405.

Toussaint

, Webb

. Theoretical and empirical connections between forgiveness, mental health, and well-being. Worthington

Jr. Handbook of Forgiveness. New York: Routledge, 2005; 349–362.

Kendler

, Liu XQ Gardner

, McCullough

, Larson

, Prescott

. Dimensions of religiosity and their relationship to lifetime psychiatric and substance use disorders. Am J Psychiatry, 2003; 160:496–503.

Brown

. Measuring individual differences in the tendency to forgive: Construct validity and links with depression. Pers Soc Psychol Bull, 2003; 29:1–13.

Hebl

, Enright

. Forgiveness as a psychotherapeutic goal with elderly females. Psychother Theor Res Pract Train, 1993; 30:658–667.

Rye

, Pargament

, Pan

, Yingling

, Shogren

, Ito

. Can group interventions facilitate forgiveness of an ex-spouse? A randomized clinical trial. J Consult Clin Psychol, 2005; 73:880–892.

Reed

, Enright

. The effects of forgiveness therapy on depression, anxiety, and posttraumatic stress for women after spousal emotional abuse. J Consult Clin Psychol, 2006; 74:920–929.

Freedman

, Enright

. Forgiveness as an intervention goal with incest survivors. J Consult Clin Psychol, 1996; 64:983–992.

10.

WHO-QOL HIV Group: Preliminary development of the World Health Organisation's Quality of Life HIV instrument (WHO-QOL HIV): Analysis of the pilot version. Soc Sci Med, 2003; 57:1259–1275.

11.

Mickley

, Cowles

. Ameliorating the tension: Use of forgiveness for healing. Oncol Nurs Forum, 2001; 28:31–37.

12.

Hansen

, Enright

, Baskin

, Klatt

. A palliative care intervention in forgiveness therapy for elderly terminally ill cancer patients. J Palliat Care, 2009; 25:51–60.

13.

Steinhauser

, Alexander

, Byock

, George

, Tulsky

. Seriously ill patients' discussions of preparation and life completion: An intervention to assist with transition at the end of life. Palliat Support Care, 2009; 7:393–404.

14.

Steinhauser

, Alexander

, Byock

, George

, Olsen

, Tulsky

. Do preparation and life completion discussions improve functioning and quality of life in seriously ill patients? Pilot randomized controlled trial. J Palliat Med, 2008; 11:1234–1240.

15.

Exline

, Baumeister

. Expressing forgiveness and repentance: Benefits and barriers. McCullough

, Pargament

, Thoresen

. Forgiveness: Theory, Research, and Practice. New York: Guilford, 2000; 133–155.

16.

Bright

, Exline

. Forgiveness at four levels: Intrapersonal, relational, organizational and collective-group. Cameron

, Spreitzer

. Handbook of Positive Organizational Scholarship. New York: Oxford, 2011; 244–259.

17.

Worthington

Jr.

Handbook of Forgiveness. New York: Routledge, 2005.

18.

Prince-Paul

. Understanding the meaning of social well-being at the end of life. Oncol Nurs Forum, 2008; 35:365–71.

19.

Prince-Paul

, Exline

. Personal relationships and communication messages at the end of life. Nurs Clin North Am, 2010; 45:449–463.

20.

Byock

. Dying Well. New York: Riverhead Books, 1998.

21.

Baker

. Facilitating forgiveness and peaceful closure: The therapeutic value of psychosocial intervention in end-of-life care. J Soc Work End Life Palliat Care, 2005; 1:83–96.

22.

Prince-Paul

. Relationships among communicative acts, social well-being, and spiritual well-being on the quality of life at the end of life in patients with cancer enrolled in hospice. J Palliat Med, 2008; 11:20–25.

23.

Prince-Paul

, Zyzanski

, Exline

. The ARCom: A screening instrument to assess personal relationships, communication in serious illness. Forthcoming.

24.

Kramer

, Boelk

, Auer

. Family conflict at the end of life: Lessons learned in a model program for vulnerable older adults. J Palliat Med, 2006; 791–801.

25.

Kramer

, Kavanaugh

, Trentham-Dietz

, Walsh

, Yonker

. Predictors of family conflict at the end of life: The experiences of spouses and adult children of persons with lung cancer. The Gerontologist, 2010; 50:215–225.

26.

Kramer

, Kavanaugh

, Trentham-Dietz

, Walsh

, Yonker

. Complicated grief symptoms in caregivers of persons with lung cancer: The roles of family conflict, intrapsychic strains, hospice utilization. Omega, 2010-2011; 62:201–220.

27.

Steinhauser

, Christakis

, Clipp

, McNeilly

, McIntyre

, Tulsky

. Factors considered important at the end of life by patients, family, physicians, and other care providers. JAMA, 2000; 284:2476–2482.

28.

Andresen

, Malmgren

, Carter

, Patrick

. Screening for depression in well older adults: Evaluation of a short form of the CES-D (Center for Epidemiologic Studies Depression Scale) Am J Prev Med, 1994; 10:77–84.

29.

Rivera

, McMillan

. Predictors of depression symptoms in hospice caregivers. J Hosp Palliat Nurs, 2010; 12:345–357.

30.

Enright

, Fitzgibbons

. Helping clients forgive: An empirical guide for resolving anger, restoring hope. Washington, DC: American Psychological Association Books, 2000.

31.

Worthington

. Steps to REACH forgiveness and to reconcile. Boston, MA: Pearson Custom Publishing, 2008.