Terminally Ill Cancer Patients' Distinct Symptom-Functional Patterns/States Are Differentially Associated with Their Accurate Prognostic Awareness in the Last Six Months of Life

Abstract

Background:

Terminally ill cancer patients' worsening symptom distress and functional impairment may signal disease deterioration, thus facilitating their accurate prognostic awareness (PA). However, the joint roles played by symptom distress and functional impairment in association with cancer patients' accurate PA remain unexplored.

Methods:

We used hierarchical generalized linear modeling to assess associations between our five identified worsening conjoint symptom-functional states and accurate PA in a convenience sample of 317 terminally ill cancer patients over their last six months.

Results:

The majority of our participants (70.1%–76.3%) had accurate PA in their last six months. This proportion did not increase as death approached but varied significantly by the five identified distinct symptom-functional states. Participants in the four worst symptom-functional states (moderate/profound symptom distress with mild/profound functional impairment) had a higher likelihood of accurate PA than those in the best state (mild symptom distress with high functioning). Participants with severe or profound symptom distress (states 3 and 5) had a substantially higher likelihood of accurate PA than those with moderate symptom distress (states 2 and 4).

Conclusion/Clinical Implications:

Terminally ill cancer patients' five distinct conjoint worsening symptom-functional states were differentially associated with their likelihood of accurate PA. Health care professionals should cultivate these patients' accurate PA when they are still free from severe symptom distress and functional impairment, effectively manage symptoms for those suffering from severe/profound symptom distress, and facilitate their psychological-spiritual adjustment to acknowledge their poor prognosis and the accompanying challenges of end-of-life care decisions to maximize quality of life and achieve a good death.

Introduction

Terminally ill cancer patients' accurate prognostic awareness (PA) is a prerequisite for actively engaging them in advance care planning^1,2 to make informed end-of-life (EOL) care decisions, thus improving EOL care quality^3,4 by limiting potentially futile anti-cancer and life-sustaining treatments and promoting care in line with one's values and wishes.^1,3–6 Indeed, accurate PA increases the likelihood of terminally ill cancer patients questioning the value of further chemotherapy^7,8 and decreases patients' preferences for^1,2,5,6,9 and use^1,10,11 of life-sustaining treatments, for example, intensive care unit care and cardiopulmonary resuscitation, thus counteracting the international trend toward increasingly aggressive and costly EOL care.^12,13

However, approximately half of advanced/terminally ill cancer patients in a systematic review did not have accurate PA.¹⁴ This finding was recently verified across 11 countries for advanced cancer patients under palliative care,¹⁵ despite guidelines calling for improved patient–physician prognostic communication in the context of serious and life-limiting illnesses^4,16 and evidence that far more cancer patients desire to know their prognosis than physicians disclose it.^2,17,18 These statistics indicate the need for understanding clinically changeable factors precipitating or precluding accurate PA¹⁹ to facilitate physicians' prognostic disclosure, thereby promoting accurate PA and facilitating value-concordant EOL care.^4,16

Symptom distress and functional impairment, which are common manifestations of terminal illness for cancer patients,²⁰ contribute to increased mortality.^21–23 Evidence shows that patients with advanced cancer may gradually perceive a poor prognosis when their physical well-being deteriorates as death approaches,^18,23,24 but few studies have directly explored the roles played by increasing symptom distress^15,25,26 or functional decline,^7,8,15 let alone conjoint symptom distress and functional impairment, in predisposing cancer patients to become aware of their prognosis. All cited studies^7,8,15,25,26 were cross-sectional, precluding their ability to detect the relationship between increasing disease burden of symptom distress/functional impairment and PA as death approaches. Furthermore, symptom distress and functional impairment do not necessarily deteriorate synchronously at EOL.^27–29 We formerly identified five distinct worsening conjoint symptom-functional states and found that terminally ill cancer patients in any two distinct symptom-functional states had significantly differing risks of subsequent death.³⁰ These symptom-functional states contributed to patients' deteriorating quality of life (QOL), anxiety symptoms, and depressive symptoms, but each state was negatively and uniquely associated with psychological well-being in patients' last year of life.³¹ We extended this line of research to design the current study to longitudinally evaluate how these five distinct symptom-functional states are associated with accurate PA over terminally ill cancer patients' last six months of life.

Methods

Design and sample

Data for this secondary analysis study came from a longitudinal study on the quality of death and dying in a convenience sample of terminally ill cancer patients recruited in 2009–2012 and followed through 2015.³² Sample size was estimated in the original study to achieve the primary goal of understanding whether physician–patient EOL care discussions were associated with concordance between terminally ill cancer patients' preferred and received EOL care. Detailed information on sampling has been reported.^30,32 Briefly, terminally ill adult Taiwanese cancer patients were referred by their primary oncologist who recognized their disease as progressive and unresponsive to curative treatments and identified them as cognitively competent to participate. Data were collected from patients in one-on-one interviews by trained, experienced oncology nurses approximately every two to four weeks when patients were hospitalized or returned for clinical care until they declined participation or died. The study site's ethical committee approved the study (98-0476B). All participants signed written informed consent.

Measures

Symptom-functional states

Physical symptom distress from cancer patients' common symptoms (i.e., pain, dyspnea, nausea/vomiting, anorexia, constipation, and insomnia) was measured using the 13-item Symptom Distress Scale (SDS).³³ Symptom distress is defined as the patient-reported degree of discomfort from specific symptoms. Distress is not differentiated by whether it results from the disease itself or the treatment. Each symptom is rated by patients on a five-point Likert scale ranging from 1 (normal or no distress) to 5 (extreme distress). Scores range from 13 to 65, with higher scores indicating greater symptom distress.

Functional impairment was assessed by the 10-item Enforced Social Dependency Scale (ESDS).³⁴ Enforced social dependence is defined as a state in which patients require help from others in performing activities or roles that under ordinary circumstances they could have performed by themselves. The personal competence subscale includes dependence for activities such as eating, dressing, walking, traveling, bathing, and toileting. The social competence subscale includes three specific role activities (activities in the home, work activities, and social and recreational activities) and a fourth behavior related to communication. ESDS scores range from 10 to 51, with higher scores reflecting greater impairment in personal and social functioning.

We identified five distinct worsening conjoint symptom-functional states³⁰: (1) mild symptom distress with high functioning, (2) moderate symptom distress with mild functional impairment, (3) severe symptom distress with moderate functional impairment, (4) moderate symptom distress with severe functional impairment, and (5) profound symptom distress and functional impairment. We labeled these five symptom-functional states by their cumulatively worsening functional impairment. Detailed information has been reported³⁰ on methods for identifying and examining changes in participants' distinct symptom-functional states between consecutive times.

Prognostic awareness

PA was measured by asking patients whether they knew their prognosis and, if so, whether their disease (1) was curable; (2) might recur in the future, but their life was not currently in danger; and (3) could not be cured, or they would probably die soon.⁹ Patients were recognized as accurately understanding their prognosis only if they chose option 3. Patients were recognized as inaccurately understanding their prognosis if they did not know their prognosis or chose option 1 or 2. We developed this measure based on Taiwanese physicians' common practice of prognostic disclosure as well as the conceptualizations and measures of PA used in a 34-study review of PA.¹⁴

Confounding variables

To determine the associations among the five distinct conjoint symptom-functional states with accurate PA, we controlled for sociodemographics (age,^35,36 gender,³⁶ and educational attainment^35,36) and disease-related characteristics (time since diagnosis^26,37 and comorbidity). Comorbidities were measured using the Deyo-Charlson comorbidity index,³⁸ categorized as 0, 1, 2, or ≥3 comorbidities.

Statistical analysis

To explore longitudinal changes in accurate PA in participants' last six months of life, we categorized time proximity to patient death, that is, the period between death and assessment, as 1–30, 31–90, and 91–180 days—conventional periods for estimating cancer patients' survival. Associations of the identified five distinct symptom-functional states with accurate PA were examined by hierarchical generalized linear modeling (HGLM).³⁹ HGLM uses random intercepts to account for within-subject correlations of repeated observations from each participant.³⁹ HGLM allows different waves of data (unbalanced data) across participants to accommodate variable numbers of follow-up points due to different post-enrollment survival times, different time intervals for data collection, and missing data for the dependent variable, thus eliminating the need to delete observations in analysis. Therefore, HGLM provides a more flexible and powerful approach by maximizing the use of available information from each participant. HGLM also allows for hypothesis testing of differences in the five symptom-functional states' associations with the outcome variable between any pair of states.⁴⁰ We arranged the lagged symptom-functional states in the previous wave of assessment to ensure a clear time sequence of associations with the outcome variable. The regression parameter for each symptom-functional state was exponentiated to transform into adjusted odds ratio (AOR), with 95% confidence interval (CI).

Results

Sample characteristics

Sample characteristics are shown in Table 1. Briefly, among the 380 enrolled participants, 317 participated until their death with repeated assessments to supply sufficient data to comprise the final sample (study completion rate, 83.4%). Participants were predominantly male (57.7%), with a mean (SD) age of 58.73 (12.97) years, married (81.1%), and had a junior high school education or less (59.0%). Participants' most common diagnoses were liver (18.0%), stomach (17.0%), pancreas (15.1%), lung (10.4%), and head and neck (9.2%) cancer, with 62.1% having comorbidities. After enrollment, participants survived 177.59 days (SD, 210.91; median, 97; range, 3–1506). On average 8.55 follow-up assessments (SD, 10.13; median, 5; range, 1–64) were made about 18.72 days (SD, 7.71; median, 15.0; range, 4–84) apart. Participants' last assessment was on average 36.32 days (SD, 61.35; median, 18.0; range, 1–567) before death.

Table 1.

Baseline Demographics and Clinical Characteristics of Participants (N = 317)

Parameter	n	%	Parameter	n	%
Gender			Educational level
Male	183	57.7	Elementary school or below	124	39.1
Female	134	42.3	Junior high school	63	19.9
Marital status			Senior high school	94	29.7
Single	16	5.0	Above senior high school	31	9.8
Married/cohabiting	257	81.1	Missing	5	1.6
Divorced/separated	20	6.3	Cancer site
Widowed	23	7.3	Liver	57	18.0
Missing	1	0.3	Stomach	54	17.0
Metastasis			Pancreas	48	15.1
Yes	237	74.8	Lung	33	10.4
No	79	24.9	Head and neck	29	9.2
Missing	1	0.3	Esophagus	21	6.6
Chronic disease			Colon-rectum	18	5.7
Yes	197	62.1	Skin	13	4.1
No	120	37.9	Breast	12	3.8
			Other	32	10.1
	Mean (SD)	Median		Mean (SD)	Median
Age (years)	58.73 (12.97)	58.00	Post-diagnostic survival (months)	18.92 (34.34)	9.0

SD, standard deviation.

Changes in accurate PA in patients' last six months of life

The five distinct worsening symptom-functional states' levels of symptom distress and functional impairment as well as their prevalence of accurate PA across all assessments are shown in Table 2. As death approached, the proportion of terminally ill cancer patients with accurate PA increased slightly, whereas symptom distress and functional impairment worsened substantially (Table 3). HGLM confirmed no significant improvement in accurate PA as death approached (Table 4).

Table 2.

Levels of Symptom Distress, Functional Impairment, and Proportion with Accurate Prognostic Awareness by Symptom-Functional State

	Symptom-functional state
	1	2	3	4	5
Average state size (%)	26.1	30.3	14.5	15.7	13.4
SDS score, M (SD)	19.02 (4.12)	26.17 (5.38)	36.21 (5.68)	24.89 (5.01)	40.18 (6.99)
ESDS score, M (SD)	16.13 (3.41)	20.47 (3.67)	28.41 (4.45)	34.09 (4.82)	41.89 (4.16)

	M (SD)	Median		Range	Interquartile
SDS^a	26.93 (8.94)	26.00		13–58	20–33
ESDS^a	24.96 (9.84)	23.00		10–51	17–32
Accurate prognostic awareness (%)	53.4	70.3	82.6	65.4	82.2

Scores are for all participants, regardless of symptom-functional state.

ESDS, Enforced Social Dependency Scale; SDS, Symptom Distress Scale.

Table 3.

Changes in Accurate Prognostic Awareness, Symptom Distress, and Functional Impairment in Terminally Ill Cancer Patients' Last Six Months of Life

Outcome	Time proximity to death (days)
Outcome	91–180	31–90	1–30
Accurate prognostic awareness, %	70.1	73.3	76.3
Symptom distress (SDS score)	26.50 (8.55)	29.39 (8.63)	33.78 (9.67)
Functional dependence (ESDS score)	23.90 (8.95)	28.60 (9.23)	34.72 (9.13)

Table 4.

Longitudinal Associations of the Five Symptom-Functional States with the Likelihood of Accurate Prognostic Awareness

	Adjusted odds ratio	95% Confidence interval		p
Lagged symptom-functional state
1	Ref.
2	3.537	1.761	7.102	<0.001
3	15.695	7.315	33.674	<0.001
4	5.396	2.365	12.310	<0.001
5	19.262	8.101	45.797	<0.001
Time proximity to death (days)
≤30	1.777	0.979	3.225	0.059
31–90	1.325	0.905	1.939	0.148
91–180	Ref
Gender
Male	1.485	0.803	2.746	0.206
Female	Ref
Age	0.953	0.926	0.980	0.001
Educational attainment
Below Senior high school	0.668	0.345	1.294	0.231
Senior high school or above	Ref.
Comorbidities^a
0	Ref.
1	0.985	0.462	2.101	0.970
2	2.760	1.119	6.810	0.028
≥3	1.376	0.508	3.729	0.529
Time since cancer diagnosis (months)
≤6	Ref.
7–12	1.895	0.824	4.357	0.132
13–24	2.208	0.928	5.252	0.073
>24	4.891	2.127	11.248	<0.001

Deyo-Charlson comorbidity index.

Ref., reference.

Associations of the five symptom-functional states with accurate PA in patients' last six months of life

Results of our HGLM showed that, after controlling for confounding variables, participants in the four worst symptom-functional states had a significantly higher likelihood of reporting accurate PA in the subsequent assessment than those in the best state of mild symptom distress with high functioning (state 1; Table 4), with AOR (95% CI) ranging from 3.537 (1.761–7.102) to 19.262 (8.101–45.797). Among participants in the four worst symptom-functional states, those in state 5 (profound symptom distress and functional impairment) had a significantly higher likelihood of reporting accurate PA than those in states 2 and 4 (moderate symptom distress with mild and severe functional impairment, respectively) (Table 5), but their likelihood of accurate PA was comparable to that of patients with severe symptom distress and moderate functional impairment (state 3). The likelihood of reporting accurate PA was significantly lower for participants in state 4 than for those in state 3, but comparable for those in state 2.

Table 5.

Differential Associations of Terminally Ill Cancer Patients' Five Distinct Symptom-Functional States with the Likelihood of Accurate Prognostic Awareness

Difference (p-value)	Accurate prognostic awareness
Symptom-functional state	4	3	2
Symptom-functional state	Moderate symptom distress with severe functional impairment	Severe symptom distress with moderate functional impairment	Moderate symptom distress with mild functional impairment
5. Profound symptom distress and functional impairment	<0.001	>0.500	<0.001
4. Moderate symptom distress with severe functional impairment		0.002	0.173
3. Severe symptom distress with moderate functional impairment			<0.001

Discussion

Approximately three-fourths (70.1%–76.3%) of Taiwanese terminally ill cancer patients had accurate PA in their last six months of life. This proportion did not increase as death approached but varied significantly by symptom-functional state. The prevalence of our participants with accurate PA was substantially higher than that reported in a systematic review (49.1% [95% CI, 42.7%–55.5%])¹⁴ of studies published in 1994–2014, post-2014 studies from Canada (55%),³⁷ the United States (16.5%–61.0%),^8,23,41 and a 2018 survey across 11 countries on advanced cancer patients under palliative care (45%).¹⁵ Our higher prevalence of accurate PA may be due to our participants being terminally ill and in their last six months of life rather than primarily subjects with advanced cancer in the aforementioned studies. Cancer patients with a baseline assessment closer to their time of death^23,42 were more likely to acknowledge being terminally ill than those who survived longer after enrollment. Therefore, our participants were more likely to accurately understand their prognosis than advanced cancer patients assessed in previous studies. In addition, physicians tend not to disclose prognosis until curative treatments are no longer effective or when patients suffer profound physical deterioration and when symptom distress accelerates as they transit into the terminal stage.⁴³ However, our participants' perceptions of accurate PA remained largely stable over their last six months of life, consistent with observations from a few studies that measured patients' PA till death.^25,44,45 The persistent lack of accurate PA for one-fourth of terminally ill cancer patients even when death was approaching underscores the importance of understanding factors precipitating or precluding accurate PA, especially for factors amenable to clinical interventions,¹⁹ which our current study aimed for.

Terminally ill cancer patients' accurate PA was associated with their conjoint levels of symptom distress and functional impairment. Our participants in the worst symptom-functional states (states 2–5) had a significantly higher likelihood of accurate PA than those in the best state of mild symptom distress with high functioning (state 1; Table 4). This finding is consistent with previous reports of worse symptom distress^25,26 and functional impairment¹⁵ predisposing cancer patients to develop accurate PA. The novel findings of our study lie in the differential associations between the likelihood of accurate PA and different conjoint levels of symptom distress and functional impairment (shown by the distinct symptom-functional states). Our results indicate that participants in the two states with the most severe symptom distress (states 3 and 5) had a substantially higher likelihood of accurate PA than those with moderate symptom distress (states 2 and 4) (Table 5). However, states 3 and 5 as well as states 2 and 4 had comparable associations with the likelihood of accurate PA, despite substantial differences in functional impairment between each pair of states. These results suggest that the association of conjoint levels of symptom distress and functional impairment with accurate PA depends more on symptom distress than on functional impairment. Exacerbated symptoms may signal deterioration of disease,¹⁸ leading cancer patients to more negatively appraise their illness,⁴⁶ thereby increasing their likelihood of developing accurate PA.

Study strengths and limitations

The major strength of our study was its prospective longitudinal design using repeated assessments of symptom distress, functional impairment, and PA as well as advanced statistics. These statistics allowed us to not only longitudinally characterize conjoint symptom-functional states and changes in accurate PA in terminally ill cancer patients' last six months of life, but also evaluate the differential associations of the five distinct symptom-functional states with accurate PA. However, the generalizability of our findings may be limited by convenience sampling from a single hospital in Taiwan, thus compromising the sample's representation of national and international target populations. Furthermore, it remains unknown whether our findings on the associations of the five distinct symptom-functional states with patients' accurate PA are applicable to the cancer patients who declined to participate (12.2%) or withdrew (5.6%) from the study.³⁰ However, our participants' exceptionally high participation rate and low attrition rate⁴⁷ suggests that our data collection schedule, despite being labor-intensive, was feasible and captured the dynamic nature of terminally ill cancer patients' symptom distress, functional impairment, and PA over their dying process. Unlike randomized controlled trials, our study was observational, limiting the ability to determine a cause–effect relationship between our identified symptom-functional states and accurate PA, despite our arrangement of time-varying symptom-functional states in a distinct time sequence and controlling for several covariates.

We did not record whether participants received palliative/hospice care, the adequacy of symptom management, or prognostic disclosure by health care professionals (either oncologists or palliative care clinicians) at each assessment, despite two-thirds of our participants (68.1%) receiving comfort-oriented care in their last month of life. Therefore, whether the suffering of a substantial proportion (27.9%) of our participants with severe or profound symptom distress (symptom-functional states 3 and 5; Table 2) derived from refractory suffering, even with adequate palliative care, or from inadequate palliative care warrants further investigation. By the same token, whether the lack of significant improvement in accurate PA was due to insufficient (too little, too late) hospice referral or inadequate physician prognostic disclosure (whether or not the patient was under hospice care) deserves further research. We also cannot exclude the possible impact of unmeasured residuals commonly found in observational studies (e.g., patient preferences for prognostic information, family caregivers' attitudes toward supporting patient autonomy and prognostic understanding, and physicians' attitudes toward and practice of prognostic disclosure).

Conclusion and Clinical Implications

Our identified five distinct conjoint symptom-functional states are associated with terminally ill cancer patients' accurate PA. Symptom distress seems to play a more prominent role, than functional impairment, in patients' perception of accurate PA. Health care professionals should cultivate terminally ill cancer patients' accurate PA while they are still free from severe symptom distress and enjoy high functioning, giving them sufficient time to prepare practically and psychologically for their EOL care.^3,16,19 Health care professionals should also tailor EOL care³ to the needs of terminally ill cancer patients not only in the worst conjoint symptom-functional states (e.g., states 3 and 5) but also with accurate PA. Indeed, we have shown³¹ that being in the worse symptom-functional states contributed to terminally ill cancer patients' worsening QOL, anxiety symptoms, and depressive symptoms in their last year of life. The development of accurate PA also does not come without costs; confronting one's mortality may precipitate emotional^8,41,45 and existential/spiritual distress^32,48 as well as worse QOL.^8,45,48 Health care professionals should not only provide effective symptom management to terminally ill cancer patients with severe/profound symptom distress (as in conjoint symptom-functional states 3 and 5) to relieve their suffering (physical and psychological),³¹ but also sensitively support those with the greatest likelihood of accurate PA who may struggle with making highly emotion-laden EOL care decisions and facing their forthcoming death. By doing so, health care professionals may cultivate cancer patients' accurate PA early in their terminal illness trajectory and facilitate their psychological-spiritual adjustment to accepting their poor prognosis and accompanying challenges of EOL care decisions. This adjustment will help them achieve the maximal QOL and a good death that may honor their wishes.⁴

Footnotes

Acknowledgments

Funding sources: The National Health Research Institutes (NHRI-EX107-10704PI), Ministry of Science and Technology (MOST 104-2314-B-182-027-MY3), and Chang Gung Memorial Hospital (BMRP888). No funding sources had any role in designing and conducting the study; collecting, managing, analyzing, and interpreting the data; or preparing, reviewing, or approving the article. The corresponding author has full access to all study data, analyzed the data with Dr. Fur-Hsing Wen, and takes responsibility for the integrity of the data and the accuracy of data analysis.

Author Disclosure Statement

The authors declare no financial or other conflict of interest.

References

Mack

, Weeks

, Wright

, et al.: End-of-life discussions, goal attainment, and distress at the end of life: Predictors and outcomes of receipt of care consistent with preferences. J Clin Oncol, 2010; 28:1203–1208.

Enzinger

, Zhang

, Schrag

, Prigerson

: Outcomes of prognostic disclosure: Associations with prognostic understanding, distress, and relationship with physician among patients with advanced cancer. J Clin Oncol, 2015; 33:3809–3816.

Peppercorn

, Smith

, Helft

, et al.: American Society of Clinical Oncology statement: Toward individualized care for patients with advanced cancer. J Clin Oncol, 2011; 29:755–760.

Institute of Medicine: Dying in America: Improving Quality and Honoring Individual Preferences Near the End of Life. Washington, DC: The National Academies Press, 2014.

Weeks

, Cook

, O'Day

, et al.: Relationship between cancer patients' predictions of prognosis and their treatment preferences. JAMA, 1998; 279:1709–1714.

Ray

, Block

, Friedlander

, et al.: Peaceful awareness in patients with advanced cancer. J Palliat Med, 2006; 9:1359–1368.

Baek

, Kim

, Heo

, et al.: Effect of advanced cancer patients' awareness of disease status on treatment decisional conflicts and satisfaction during palliative chemotherapy: A Korean prospective cohort study. Support Care Cancer, 2012; 20:1309–1316.

El-Jawahri

, Traeger

, Park

, et al.: Associations among prognostic understanding, quality of life, and mood in patients with advanced cancer. Cancer, 2014; 120:278–285.

Tang

, Liu

, Chow

, et al.: Associations between accurate prognostic understanding and end-of-life care preferences and its correlates among Taiwanese terminally ill cancer patients surveyed in 2011–2012. Psychooncology, 2014; 23:780–787.

10.

Temel

, Greer

, Admane

, et al.: Longitudinal perceptions of prognosis and goals of therapy in patients with metastatic non small-cell lung cancer: Results of a randomized study of early palliative care. J Clin Oncol, 2011; 29:2319–2326.

11.

Yun

, Lee

, Kim

, et al.: Impact of awareness of terminal illness and use of palliative care or intensive care unit on the survival of terminally ill patients with cancer: Prospective cohort study. J Clin Oncol, 2011; 29:2474–2480.

12.

Bekelman

, Halpern

, Blankart

, et al.: Comparison of site of death, health care utilization, and hospital expenditures for patients dying with cancer in 7 developed countries. JAMA, 2016; 315:272–283.

13.

Hung

, Liu

, Wen

, et al.: Escalating healthcare expenditures in cancer decedents' last year of life: A decade of evidence from a retrospective population-based cohort study in Taiwan. Oncologist, 2017; 22:460–469.

14.

Chen

, Kuo

, Tang

: Current status of accurate prognostic awareness in advanced/terminally ill cancer patients: Systematic review and meta-regression analysis. Palliat Med, 2017; 31:406–418.

15.

Yennurajalingam

, Rodrigues

, Shamieh

, et al.: Perception of curability among advanced cancer patients: An international collaborative study. Oncologist, 2018; 23:501–506.

16.

Gilligan

, Coyle

, Frankel

, et al.: Patient-clinician communication: American Society of Clinical Oncology Consensus Guideline. J Clin Oncol, 2017; 35:3618–3632.

17.

Fujimori

, Akechi

, Morita

, et al.: Preferences of cancer patients regarding the disclosure of bad news. Psychooncology, 2007; 16:573–581.

18.

Yun

, Kwon

, Lee

, et al.: Experiences and attitudes of patients with terminal cancer and their family caregivers toward the disclosure of terminal illness. J Clin Oncol, 2010; 8:1950–1957.

19.

Tulsky

, Beach

, Butow

, et al.: A research agenda for communication between health care professionals and patients living with serious illness. JAMA Intern Med, 2017; 177:1361–1366.

20.

Seow

, Barbera

, Sutradhar

, et al.: Trajectory of performance status and symptom scores for patients with cancer during the last six months of life. J Clin Oncol, 2011; 29:1151–1158.

21.

Stukenborg

, Blackhall

, Harrison

, et al.: Longitudinal patterns of cancer patient reported outcomes in end of life care predict survival. Support Care Cancer, 2016; 24:2217–2224.

22.

Yamada

, Morita

, Maeda

, et al.: A prospective, multicenter cohort study to validate a simple performance status-based survival prediction system for oncologists. Cancer, 2017; 123:1442–1452.

23.

Liu

, Landrum

, Weeks

, et al.: Physicians' propensity to discuss prognosis is associated with patients' awareness of prognosis for metastatic cancers. J Palliat Med, 2014; 17:673–682.

24.

Kurita

, Siegler

, Reid

, et al.: It is not what you think: Associations between perceived cognitive and physical status and prognostic understanding in patients with advanced cancer. J Pain Symptom Manage, 2018; 56:259–263.

25.

Hinton

: The progress of awareness and acceptance of dying assessed in cancer patients and their caring relatives. Palliat Med, 1999; 13:19–35.

26.

The Italian Group for the Evaluation of Outcomes in Oncology (IGEO): Awareness of disease among Italian cancer patients: Is there a need for further improvement in patient information?. Ann Oncol, 1999; 10:1095–1100.

27.

Sutradhar

, Atzema

, Seow

, et al.: Is performance status associated with symptom scores? A population-based longitudinal study among cancer outpatients. J Palliat Care, 2014; 30:99–107.

28.

Reese

, Blackford

, Sussman

, et al.: Cancer patients' function, symptoms and supportive care needs: A latent class analysis across cultures. Qual Life Res, 2015; 24:135–146.

29.

Buckner

, Wang

, DeWalt

, et al.: Patterns of symptoms and functional impairments in children with cancer. Pediatr Blood Cancer, 2014; 61:1282–1288.

30.

Wen

, Chen

, Chou

, et al.: Distinct patterns of conjoint symptom distress and functional impairment in the last year of life predict terminally ill cancer patients' survival. J Pain Symptom Manage, 2018; 55:1443–1451.

31.

Wen

, Chen

, Chou

, et al.: Quality of life and psychological distress are differentially associated with distinct symptom-functional states in terminally ill cancer patients' last year of life. Psychooncology, 2018; 27:2111–2118.

32.

Tang

, Chang

, Chen

, et al.: Associations of prognostic awareness/acceptance with psychological distress, existential suffering, and quality of life in terminally ill cancer patients' last year of life. Psychooncology, 2016; 25:455–462.

33.

McCorkle

, Young

: Development of a Symptom Distress Scale. Cancer Nurs, 1978; 1:373–378.

34.

Benoliel

, McCorkle

, Young

: Development of a social dependency scale. Res Nurs Health, 1980; 3:3–10.

35.

Applebaum

, Kolva

, Kulikowski

, et al.: Conceptualizing prognostic awareness in advanced cancer: A systematic review. J Health Psychol, 2014; 19:1103–1119.

36.

Papadopoulos

, Vrettos

, Kamposioras

, et al.: Impact of cancer patients' disease awareness on their family members' health-related quality of life: A cross-sectional survey. Psychooncology, 2011; 20:294–301.

37.

Fisher

, Seow

, Cohen

, et al.: Patient characteristics associated with prognostic awareness: A study of a Canadian palliative care population using the InterRAI palliative care instrument. J Pain Symptom Manage, 2015; 49:716–725.

38.

Deyo

, Cherkin

, Ciol

: Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. Clin Epidemiol, 1992; 45:613–619.

39.

Raudenbush

, Bryk

: Hierarchical Linear Models, 2nd ed. Thousand Oaks, CA: Sage Publications, 2002.

40.

Garson

: Introductory guide to HLM with HLM 7 software. In: Garson

(ed): Hierarchical Linear Modeling: Guide and Applications. Los Angeles, CA: Sage, 2013, pp. 55–96.

41.

Shin

, El-Jawahri

, Parkes

, et al.: Quality of life, mood, and prognostic understanding in patients with metastatic breast cancer. J Palliat Med, 2016; 19:863–869.

42.

Lichtenthal

, Nilsson

, Zhang

, et al.: Do rates of mental disorders and existential distress among advanced stage cancer patients increase as death approaches?. Psychooncology, 2009; 18:50–61.

43.

Keating

, Landrum

, Rogers

, et al.: Physician factors associated with discussions about end-of-life care. Cancer, 2010; 116:998–1006.

44.

Fried

, Bradley

, O'Leary

: Changes in prognostic awareness among seriously ill older persons and their caregivers. J Palliat Med, 2006; 9:61–69.

45.

Greer

, Pirl

, Jackson

, et al.: Perceptions of health status and survival in patients with metastatic lung cancer. J Pain Symptom Manage, 2014; 48:548–557.

46.

Sumpio

, Jeon

, Northouse

, Knobf

: Optimism, symptom distress, illness appraisal, and coping in patients with advanced-stage cancer diagnoses undergoing chemotherapy treatment. Oncol Nurs Forum, 2017; 44:384–392.

47.

Perez-Cruz

, Shamieh

, Paiva

, et al.: Factors associated with attrition in a multicenter longitudinal observational study of patients with advanced cancer. J Pain Symptom Manage, 2018; 55:938–945.

48.

Mack

, Nilsson

, Balboni

, et al.: Peace, equanimity, and acceptance in the Cancer experience (PEACE): Validation of a scale to assess acceptance and struggle with terminal illness. Cancer, 2008; 112:2509–2517.