Mechanisms of Dignity Therapy: Positive Outcomes in Older Cancer Patients

Abstract

Objective:

To investigate two mechanisms, patient-level narrative richness and provider-patient empathy, of positive outcomes following Dignity Therapy, to focus provider training and intervention delivery.

Background:

Dignity Therapy is a brief reminiscence-based psychotherapeutic intervention designed to help seriously ill patients preserve dignity, reduce distress, and improve quality of life. Identifying mechanisms through which the therapy works can improve training and delivery of this increasingly popular intervention.

Methods:

Outpatients living with cancer (N = 203; M = 65.80 years; SD = 7.45 years; 66% women) from palliative care programs across the United States completed Dignity Therapy with a trained provider. Transcripts of their interview sessions were examined using interactional analyses to determine provider-level empathic communication. Transcripts were also content-analyzed for patient-level narrative richness. Changes in dignity, peaceful awareness of prognosis, and completion of existential tasks from pre- to postintervention were assessed.

Results:

The extent of narrative richness in patients’ Dignity Therapy sessions was positively associated with post-test dignity (t = 3.09, p = 0.002) and completion of existential tasks (t = 2.65, p = 0.009) even when accounting for patient demographics. Providers’ level of empathic communication did not affect patient outcomes at traditional significance levels. Results were not moderated by patients’ symptom severity.

Conclusions:

Dignity Therapy benefits patients most when they richly engage in the process, narrating their life story and describing their legacy with elements of communion, meaning, and purpose. Future research might aim to follow up on forms of empathy or other provider behavior that elicit rich narratives during therapy.

Introduction

Given that the Baby Boomers¹ face a one in three chance of developing cancer by age 70,² cancer care continues to be a significant health issue. Implementing psychosocial interventions to improve older individuals’ ability to cope with serious cancer is more important than ever. In particular, maintaining dignity for older adults with cancer is central to emotional resilience in face of disease.³ Despite that, older oncology patients do not always experience dignity during care.⁴ We support patient dignity as a central value in oncology care, consistent with the Institute of Healthcare Improvement’s Triple Aim that includes improving patient satisfaction.⁵ Dignity comes from Latin for worth or worthiness. In clinical contexts, it refers to “being seen” in terms of all patients being deserving of others’ honor, respect, and esteem.

Dignity Therapy (DT) was developed to address an ongoing challenge: maintaining patients’ dignity throughout the illness process.^6,7 DT is a brief, individualized psychotherapeutic intervention designed to help seriously/terminally ill patients maintain dignity through a guided life review and creation of a generativity document. Recent systematic reviews concluded DT has high acceptability, shows high patient satisfaction and effectiveness, and increases sense of dignity.^8,9 Efficacy for improving dignity has also been demonstrated in a large sample RCT of older palliative patients.¹⁰ We suggest that delineating precise mechanisms by which DT operates may allow refinements in provider training and intervention delivery. Establishing mechanisms also helps distinguish unique components of this intervention from other approaches designed to support cancer patients’ well-being.

Palliative care increasingly serves as a model of compassionate, patient-centered health care, with growing institutional and policy endorsement in recent years.¹¹ It prioritizes humanistic constructs such as empathic provider communication and recognition of patients’ own life experiences (i.e., narrative medicine).¹² As such, we selected humanistic constructs as potential central mechanisms of DT (see Fig. 1: direct effects of empathic communication and narrative richness on DT outcomes constitute mechanisms). The first mechanism is empathic communication, an interpersonal process.^13–18 This involves patients feeling listened to, valued and understood. Research on clinical interactions has demonstrated the positive impact of empathic communication on patient outcomes,^19–22 including in the cancer context.^16,17 As per Neumann and colleagues’ model,²³ empathic communication encourages more patient self-disclosure and deeper engagement in the clinical encounter (i.e., the DT session). In DT, the provider guides the patient to express their life story and legacy (i.e., wishes for family, friends). There is thus a good rationale for expecting providers’ empathic communication during the session to operate as a mechanism (Fig. 1: direct path from empathic communication to DT outcomes) of effective DT.

FIG. 1.

Conceptualized moderated-mediation model.

The second hypothesized mechanism is extent of narrative richness in the life story-legacy shared during DT. Narrative richness includes three indicators empirically linked to well-being: expressing communion (i.e., love and concern for others),²⁴ creating meaning of lived events (i.e., making sense of challenges),²⁵ and delineating one’s sense of purpose (i.e., life motivation).²⁶ Life story theory²⁷ emphasizes variability in richness when individuals tell life stories.²⁸ Sharing rich, organized, affect-laden narratives leads to positive outcomes.²⁹ In contrast, telling a skeletal, emotionally-neutral story is not related to psychosocial outcomes.³⁰ As such, patients’ amount of narrative richness during DT may operate as a mechanism (i.e., Fig. 1: direct path from narrative richness to DT outcomes).

Beyond these proposed mechanisms, we speculatively explored two issues. The first was whether the effects of empathic communication on DT outcomes would be mediated by narrative richness. That is, if the provider is more empathic does this lead to patients showing greater narrative richness in the DT session, and thereby positive DT outcomes. The second was whether patients’ symptom severity at the time of engaging in DT may moderate relations between empathic communication and narrative richness, and thereby therapeutic effects. Early initiation of palliative care improves quality of life,^31–33 reduces symptom burden,^31–33 increases satisfaction with care,^34–36 decreases health care utilization,³¹ and improves survival.^31,33,34 Most DT research has focused on patients with advanced or terminal illness.³⁴ What is the best time in disease progression to engage patients in DT? We hypothesize patients with moderate symptoms may feel able and motivated to engage in the therapy. In contrast, patients with only mild symptoms may lack such motivation.^37,38 Those with severe symptoms may be motivated but too ill to benefit.⁹

As such, we developed three specific aims. The first two assessed mechanisms of positive DT outcomes using regression analyses:

1. examine the direct effect of amount of empathic responding by the DT provider on post-therapy outcomes (i.e., dignity, peaceful awareness [PA] of prognosis, completion of existential tasks).

2. examine the direct effect of amount of patients’ narrative richness during DT sessions on positive post-therapy outcomes.

3. We then explored additional relations using moderated-mediation analyses, assessing:

4. indirect effects of empathic responding on post-therapy outcomes through patients’ narrative richness (i.e., mediation) and whether patient symptom severity moderated relations between provider empathic communication and patients’ narrative richness levels.

Methods

Study overview and design

This study is a secondary analysis of data from 203 older adult cancer patients who received DT as part of a six-site, pre/post-test, randomized, controlled design comparing effects of usual outpatient palliative care to usual care plus DT.^10,39 DT administration followed the standard protocol: DT provider guided patients, in one-on-one approximately 1 hour sessions, through core questions about their life and legacy. This was recorded and transcribed for sharing with family and close others.^5,6 Providers were nurses and chaplains trained in DT. Transcriptions of the sessions were used, in the current research, for rigorous quantitative analyses based on interactional (empathic communication) and content-analytic (narrative richness) coding. Our methods are grounded in the voices of the interviewed patients while resulting in numeric values subjected to quantitative analyses. Our single-group pretest-post-test design was utilized to identify mechanisms of positive DT outcomes through examining direct effects of empathic communication, and of narrative richness, on patient outcomes (i.e., dignity, existential task completion, PA).

Procedures

Sample.

Patients were eligible for inclusion if: (1) diagnosed with cancer, (2) receiving outpatient palliative care, (3) 55 years or older, (4) able to speak/read English, and (5) able to complete study (i.e., Palliative Performance Score >50; PPS)⁴⁰ suggesting mean life expectancy of at least 55 days. Informed consent was obtained from all the patients involved in this study.

Unitizing.

DT transcripts were separated into idea units. Idea units represent shifts in content of the narrative based on initiation of new topics arising due to guiding questions from provider or the patient changing the subject.

Narrative richness coding.

Narrative richness coding examines how DT patients narrate their life story-legacy. It consists of three elements: communion, patients’ care, and concern for important others;^23,41 meaning-making, patients’ sense-making of life’s varied experience;²⁵ and positive sense of purpose, life having an overarching motivation.^26,42 An extensive manual was used to train coders, based on best practices and previous work.^9,43,44 Each DT idea unit in every transcript was reliably double-coded (κ = 0.85) for the presence (1) or absence (0) of each element of narrative richness. The few remaining discrepancies were resolved through consensus. Sum scores involved adding together the total for communion, meaning-making, and positive sense of purpose that were coded as present for each patient.

Empathic communication coding.

Empathic communication coding examines how DT providers convey empathy. It consists of: acknowledgment, recognition of patient’s experience; reflection, interpreting patient’s experience in own words; validation, indicating patient’s experience is reasonable; and empathic self-disclosure (ESD),^17,45,46 responding by sharing personal information. In tandem with the interview audio, DT idea units in transcripts were reliably double-coded (κ = 0.89) for presence (1) or absence (0) of each element of empathy, with discrepancies resolved through consensus. Sum scores were created by adding together the totals of acknowledgment, reflection, validation, and ESD that were coded as present for each patient.

Outcome measures

Dignity impact scale.

The effectiveness of care on patient dignity was assessed through the Dignity Impact Scale (DIS). The DIS has shown differences between patients who did or did not receive DT in an RCT.^47,48 Item wording for the 7-item scale was modified to fit the pretest-post-test design. A sample item is “The care I received during the past month has increased my sense of dignity.” Patients respond on a 5-point scale from (1) “strongly disagree” to (5) “strongly agree.”

Existential tasks.

Items are from the Quality of Life at End of Life, that assesses quality of life at end of life.⁴⁹ About 11 items assess completion of existential tasks (ET), including feelings of life completion and death preparation. The measure assesses concepts such as life integrity, comfort in relations with those left behind, and lack of unfinished business. A sample item is “Despite my illness I have a sense of meaning in life.” Items are rated on 5-point scales: (1) “not at all true for me” to (5) “completely true for me.” Scale shows good reliability (Cronbach’s α > 0.68).

Peaceful awareness.

PA of prognosis was used as a dichotomous indicator. To receive a “1,” patients needed to indicate awareness of their terminal status and also that they felt peaceful more often than not. The peacefulness component used a 6-point scale from (1) “almost never” to (6) “many times a day” regarding extent patients felt deep inner peace. Patients who did not meet these criteria (i.e., reported that they were healthy or not terminally ill and/or answered “2” or lower on the peacefulness scale) were assigned “0” for PA. That is, it was not enough for patients to be peaceful but not aware, or aware but not peaceful, to receive a “1.”

Symptom severity.

The Edmonton Symptom Assessment Scale (ESAS) shows good reliability and validity in a systematic review.⁵⁰ The ESAS-r is widely used to assess symptoms in palliative patients (e.g., pain, fatigue, loss of appetite). Symptoms are rated ranging from (0) no issue with symptom to (10) worst possible experience of symptom.

Analyses

Analyses were conducted using statistical software R. Amount of missing data was small (2.5%), but we utilized multiple imputation to generate multiple completed datasets on which inference was performed and then aggregated using Rubin’s rule. Descriptives for demographic variables and major variables of interest are presented. Next, regression analysis (i.e., linear regression for DIS and ET outcomes; binary logistic regression for PA) adjusting for pre-test were used to address the first two specific aims regarding DT mechanisms (i.e., direct effects of empathic communication, narrative richness on patient outcome variables). For logistic regression, we used the likelihood ratio R² measure (Tables 2 and 3). Moderated-mediation analysis was performed to test Aim 3, using Monte-Carlo Markov Chain-based Bayesian multivariate modeling.

Table 2.

Estimates of Empathic Communication Effects for Each of Three Post-test Outcomes

Patient outcome	Variable	B estimate	SEB	t	df	p	R 2
DIS	Pretest DIS	0.51	0.07	7.41	159	<0.001	0.27
	Empathic communication	0.04	0.02	1.96	195	0.051
ET	Pretest ET	0.50	0.05	9.71	182	<0.001	0.35
	Empathic communication	0.02	0.02	0.91	192	0.36
PA	Pretest PA	2.85	0.43	6.65	191	<0.001	0.25
	Empathic communication	0.001	0.02	0.05	190	0.96

DIS, Dignity Impact Scale; ET, Existential Task (QUAL-E); PA, peaceful awareness.

Aim 2: Narrative richness as a mechanism: direct effect on post-test outcomes.

Table 3.

Estimates of Narrative Richness Effects for Each of Three Post-test Outcomes

Patient outcome	Variable	B estimate	SEB	t	df	p	R 2
DIS	Pretest DIS	0.48	0.07	7.12	161	<0.001	0.29
	Narrative richness	0.14	0.04	3.09	193	0.002
ET	Pretest ET	0.50	0.05	9.83	182	<0.001	0.37
	Narrative richness	0.12	0.05	2.65	192	0.009
PA	Pretest PA	2.91	0.44	6.63	190	<0.001	0.25
	Narrative richness	−0.03	0.03	−0.91	168	−0.36

Aim 3: Symptom Severity as a Moderator and Narrative Richness as Mediator.

DIS, Dignity Impact Scale; ET, Existential Tasks (QUAL-E); PA, Peaceful Awareness.

Results

Sample characteristics

The sample included 203 cancer patients in outpatient palliative care in the United States. Patients were 55–88 years old (M = 65.8, SD = 7.4) and self-identified as 66% female, 85% White, and 92% non-Hispanic. Of 31 patients who did not identify as White, 24 identified as Black, one as Asian, one as Native Hawaiian/Pacific Islander, one as Alaskan native, and four as being of another, not-listed, race. PPS ranged from 60 to 100. Descriptive statistics appear in Table 1.

Table 1.

Demographics

Variable	Mean (SD)	Range
Age	65.8 (7.4)	55–88
Variable	Category	n (%)
Gender	Female	134 (66%)
	Male	69 (34%)
Race (2 missing)	Other	31 (15%)
	White	170 (85%)
Ethnicity (6 missing)	Hispanic or Latino	16 (8%)
	Not Hispanic or Latino	181 (92%)
Marital Status (4 missing)	Married/Partnered	121 (61%)
	Single	78 (39%)
Education (7 missing)	High school or less	28 (14%)
	Some college	70 (36%)
	College	51 (26%)
	Advanced degree	47 (24%)
Income (16 missing)	< $10,000	10 (5%)
	$10,000–$19,999	25 (13%)
	$20,000–$29,999	18 (10%)
	$30,000–$39,999	18 (10%)
	$40,000–$49,999	16 (9%)
	$50,000 or more	100 (53%)

Aim 1: Empathic communication as a mechanism: direct effect on post-test outcomes.

Our first aim was to examine relations of provider’s level of empathic communication during the session to each posttest outcome, adjusting for the pretest, so as to establish whether empathic communication during DT is a mechanism of positive outcomes. All sessions involved some instances of empathic communication (i.e., range = 6–93) with M = 28.0, SD = 13.4. Compared with null models, regression models with pre-test measures of DT outcomes and empathic communication as predictors were highly significant, p < 0.001. Individual effects of regression analyses for each outcome variable show the association between empathic communication and post-test outcomes was not statistically significant at p < 0.05 (Table 2). Empathic communication cannot be interpreted as a mechanism of positive DT outcomes. The relation between empathic communication and DIS trended toward significance at p = 0.051. Findings were similar when including patient age, race, and gender in the models.

The second aim focused on associations between narrative richness and each of the post-test outcomes, adjusting for the pretest (Table 3), to establish whether patients’ level of narrative richness during DT is a mechanism of positive outcomes. The range for narrative richness was 0–32 instances, M = 12.6, SD = 6.4. Compared with null models, regression models with pre-test patient outcome measures and narrative richness as predictors were highly significant, p < 0.001. Individual effects for each outcome variable show narrative richness acted as a mechanism, that is, was positively associated with DIS (t = 3.09, p = 0.002) and completion of ET (t = 2.65 p = 0.009). Effects held when patient age, race, and gender were included in the models.

Aim 3 used separate moderated-mediation models (Fig. 1) to examine the three post-test outcomes.

ESAS scores from 18 to 35 (i.e., average per item score >1 but <4) was used to represent moderate symptom severity with scores below indicating low, and above indicating high severity. Moderated-mediation analyses showed no significant direct or indirect effects of empathic communication on posttest outcomes (i.e., DIS, ET, PA) and no moderation by symptom severity. As such, narrative richness was not a mediator but was a mechanism (i.e., had a direct effect) as shown in Aim 2 analyses, on post-test outcomes DIS and ET.

Discussion

The need for effective psychosocial interventions for coping with cancer remains a priority.³ Patient dignity is of paramount concern⁴ and DT has been demonstrated as an effective intervention now being used across the globe.¹⁰ We focused on identifying mechanisms underlying efficacy of this therapy. Our findings show that the narrative richness of patients’ life stories-legacy told during DT is such a mechanism. That is, the extent to which patients contextualize their illness through telling a rich life story-legacy narrative during DT is related, at post-test, to both improved patient dignity and sense of completion of existential tasks.

We have demonstrated that being able to tell a full, rich life story-legacy during the DT process relates to positive patient outcomes. This richness involves patients looking back on and talking about their life story in a way that emphasizes areas of life satisfaction, including love and caring for others, forging meaning in the face of life challenges and consideration of one’s personal purpose in life.^23–25 Note that patients’ narratives varied widely in the extent of narrative richness they contained: some told more basic, skeletal life story-legacies. Specific to DT, provider training might emphasize interview techniques that could elicit rich, textured patient narratives—focused on sharing elements of communion, meaning, and purpose in life—during DT sessions.

Empathic responding (e.g., provider validation of the patient’s narrative) trended in the direction of improved patient dignity at post-test but was not statistically significant (p = 0.051). In addition, the relation of providers’ empathic communication during DT sessions to patient post-test outcomes was not mediated by narrative richness, and effects were not moderated by patients’ symptom severity. Given that we did find narrative richness to be a mechanism of positive DT outcomes, future research might focus on how empathy or other provider communication variables can be implemented that help elicit rich patient narratives during therapy.

Our findings regarding the importance of patients narrating their stories with richness may be expanded to other psychotherapeutic contexts and aspects of clinical care. This study provides strong evidence for the power of the narrative medicine⁶ approach that characterizes the patient as a person with a story and the provider as a warm, interested listener. As such, our results also contribute to a humanistic, person-centered approach to care that is applicable in health care settings, particularly end-of-life palliative and psycho-oncological care.^51,52 Simple, time-effective approaches such as the Patient Dignity Question are also a strong start for giving patients their own voice during care.^53,54

Limitations of the present study include sample composition that limits generalizability (i.e., predominantly White, well-educated, older adult palliative cancer outpatients). The current findings did hold when adjusting for patients’ age, gender, and race and were unaffected by symptom severity. Recruitment for this study faced issues (at system, individual, and interpersonal levels) commonly seen in recruiting diverse samples for cancer clinical trials.⁵⁵ In addition, recruitment extended into the Covid-19 pandemic. To highlight the importance of recruiting highly diverse samples, for future research, we note that research has identified ways that predominately Black as compared with White individuals differ regarding end-of-life treatment preferences.⁵⁶ Future research is needed to understand the nuances in mechanisms at work during DT delivery to patients from a broader range of age and cultural backgrounds.

Conclusion

Nearly 600,000 Americans die from cancer each year.¹ DT can part of palliative care that supports the dying person and their family. This research delineates narrative richness as a mechanism for enhancing psychosocial outcomes after DT. Future research should investigate both provider and patient characteristics and processes that lead to optimal delivery of DT.

Key Message.

Dignity Therapy (DT) supports positive psychosocial outcomes. Richness of patients’ narratives is a mechanism of this impact: sharing rich narratives during the DT interview promotes a sense of dignity and completion of existential tasks before death. Broader implementation of this therapy is suggested to comfort and support older patients suffering from serious cancer.