Quality of Life Trajectory in Patients with Advanced Cancer during the Last Year of Life

Abstract

Introduction:

Due to the high mortality of cancer a large number of patients pass a preterminal phase of their illness. Within this phase medical care aims at maintaining patients' quality of life (QOL) and reducing symptom burden. Our study investigated the patient-reported severity of QOL impairments during the last year of life, with a special focus on their course at the end of life.

Methods:

All patients with cancer receiving palliative care at Natters State Hospital (Austria) were considered as eligible for the study. QOL data were collected with the EORTC QLQ-C30 questionnaire as part of computerized patient-reported outcome monitoring (ePROM) within clinical routine. QOL was investigated longitudinally in regard to its course toward death as well as to changes in determinants of global QOL.

Results:

Eighty-five patients participated in the ePROM (255 assessments in total). Regarding trajectories, physical, role and cognitive functioning, fatigue and global QOL worsened sharply during the last 3 months of life. A steady decline was found for emotional functioning, pain, appetite loss and taste alterations. The impact of role functioning, sleep disturbances, and taste alterations on global QOL increased within the last 3 months of life.

Conclusion:

Our results indicate that most aspects of QOL are considerably impaired in patients with advanced cancer. Furthermore, they highlight the importance of assessing QOL in general and taste alterations in particular within palliative care.

Introduction

Between 2004 and 2008 the number of people newly diagnosed with cancer increased from 2.9 to 3.2 million in Europe, whereas the number of cancer-related deaths per year remained stable at around 1.7 million.¹ Such figures reveal that despite the availability of advanced oncologic treatment options and increasing survival rates, a high proportion of patients diagnosed with cancer eventually die from it. This means that a large number of patients passes a preterminal phase of their illness, often within hospital palliative care settings.² For this preterminal phase maintaining patients' quality of life and reducing symptom burden are major treatment goals.

Compared to the large number of cancer patients affected, the number of studies focusing on physical symptoms, psychosocial burden, and quality of life (QOL) in patients at the end of life appears insufficient. Available evidence from the literature shows severe QOL impairments across multiple domains and suggests a trajectory of QOL in patients with advanced cancer characterized by a terminal drop, i.e., rapid deterioration during the last weeks of life.^3–7

Hwang et al.⁴ reported a steady decline of QOL in the last 6 months prior death with an increasing deterioration in the last 2 months of life. In this study, overall QOL decreased nearly 4 times faster during the last 2 months of life compared with the preceding 4 months. Lundh Hagelin et al.⁵ compared patients during the last 15 days of life with patients with a survival time above 120 days. They found severe impairments across multiple QOL domains. A time to death below 2 weeks was associated with a significantly worse rating of physical, emotional, and cognitive functioning as well as with global QOL. A significant increase in symptom burden was also found for fatigue and dyspnea. However, patients facing death within the following 2 weeks did not differ from patients living longer than 120 days with regard to role functioning, social functioning, pain, nausea/vomiting, sleep disturbances, appetite loss, diarrhea, constipation, and financial impact of disease.

Global QOL in preterminal patients' was found to be related to fatigue, appetite loss, physical, role and emotional functioning, dyspnea, insomnia. and the number of comorbidities.^5,6,8,9 But even if satisfactory overall QOL scores are reported by patients within 2 weeks before death, especially physical symptoms like pain, fatigue and dyspnea, physical well-being, and existential issues (e.g., autonomy, burden on others, self-esteem, hope) should be paid more attention to. Moreover, routine QOL measurement by means of patient-reported outcome (PRO) assessments is of special importance in terminally ill cancer patients, as ratings by physicians tend to underestimate the symptom severity.¹⁰

Objectives

Our study contributes to knowledge on QOL in preterminal cancer patients by focusing on trajectories and relative importance of various aspects of QOL. In detail, it adresses the following research questions:

To what degree do patients with cancer experience QOL impairments at the end of life?

To what degree does deterioration in QOL accelerate toward the end of life?

Do determinants of global QOL change toward the end of life?

Methods

Sample

Patients with cancer at the Department of Internal Medicine of Natters State Hospital were considered for enrolment in our retrospective analysis. Data collection for this study was done within ongoing computerized patient-reported outcome monitoring (ePROM) as part of clinical routine. Inclusion criteria for ePROM were: any cancer diagnosis, expected survival time of more than 3 months at enrolment, no severe cognitive impairments, and German speaking. In our retrospective analysis of QOL trajectories during the last year of life we used data from patients who received medical care with palliative intent and who had died by the time of the analysis.

Within ePROM patients were approached by a psychology intern during their inpatient stay. After providing initial informed consent for participation in the QOL-monitoring a tablet-PC presenting the EORTC QLQ-C30 on the screen was handed over to the patients along with instructions for the completion of the questionnaire. For patients not being able to fill in the questionnaire themselves, the assessment was done as an interview. Assessments were done weekly all along hospitalisation, no assessments were done while patients were at home. Patients participated in ePROM as long as they were able and willing to report on their QOL.

As software tool for electronic data capture a program called Computer-based Health Evaluation System (CHES^11,12) was used. To increase data quality, hospital records on patients' death were supplemented with data from the Tyrolean Tumour Registry.¹³ The study was approved by the Ethics committee of Innsbruck Medical University.

EORTC QLQ-C30

As a measure of functioning, global QOL and symptoms all patients completed the EORTC QLQ-C30. The EORTC QLQ-C30¹⁴ is an internationally validated and widely used cancer-targeted QOL-instrument. It comprises five functioning scales (physical functioning, social functioning, role functioning, emotional functioning, cognitive functioning), a scale for global QOL and nine symptom scales (fatigue, nausea/vomiting, pain, dyspnea, sleep disturbances, appetite loss, constipation, diarrhea, and financial impact).

In addition, two items concerning taste alteration were taken from the EORTC Quality of Life Group item bank (“Have you had problems with your sense of taste?” and “Did food and drink taste different from usual?”). The two items on taste were summed to generate a novel subscale called the taste alterations subscale.¹⁵ The score was transformed to range from 0 to 100 points, with higher values indicating more severe taste alterations.

Statistical analyses

For detailed description of the distribution of QOL scores at the end of life (time to death less than 90 days) percentiles are given. As the number of assessments varied across patients, we provide weighted means, standard deviations and percentiles.

Investigation of the course of QOL toward the end of life was based on mixed linear models. These models allow the timing and number of assessments to differ across patients which is important to handle attrition. Furthermore, they appropriately adjust variance estimates for the correlation of repeated observations from the same patient. In our model we used a random baseline term and accounted for correlations between assessment points (assuming a first order autoregressive model, AR1).

For analysis of the presence of a nonlinear QOL trajectory (i.e., a stronger decline of QOL at the end of life than in the time period before) regression slopes before and after the cut point of 90 days to death were compared by means of an interaction between time to death and time interval (</> 90 days to death). Thus, the mixed linear model contained the following terms:

Constant term (fixed)

Main effect (fixed): Time to death

Interaction effect (fixed): Time to death * Time interval (</> 90 days to death)

Constant term (random)

In case of a nonsignificant interaction, overall change was investigated in a model comprising only time to death as a main effect. As measure of change we calculated monthly deterioration rates (change in score points per 30 days) based on regression coefficients. Deterioration rate was compared for the period before and within 90 days to death by means of the interaction effect. In these longitudinal analyses we included only patients with at least two assessment points, and at most 12 assessments per patient.

Determinants of global QOL at the end of life (time to death less than 30 days) were compared to earlier determinants present at an earlier stage of the disease trajectory (time to death longer than 90 days). To investigate change of determinants of global QOL, correlations between global QOL and other scales of the EORTC QLQ-C30 were compared for both periods. To test for statistical significance of differences between correlation coefficients we used Fisher Z transformation.

Results

Patient characteristics

Between April 2007 and August 2009 a total of 85 patients met inclusion criteria and were included in the study (mean age 66.0, standard deviation [SD] 11.5). The total number of assessments was 255 (3.0 assessments per patient on average; median 2.0). 28 patients were assessed once, 20 patients twice, 11 patients three times, 12 patients four times, and 14 patients more than four times. 144 assessments took place within 90 days before death, 111 assessments were between 365 and 91 days to death.

Across all assessment points mean time to death was 110 days (SD 100). Most common diagnoses were lung cancer (25.9%), hematologic malignancies (15.3%), and breast cancer (11.8%). Most assessments were done at time points when patients received neither chemotherapy nor radiotherapy (54.5%). Details on sociodemographic and clinical patient characteristics are shown in Table 1.

Table 1.

Descriptive Statistics for Sociodemographic and Clinical Data at Baseline (n = 85)

Age (years)	Mean (SD)	68.0 (10.8)
Time since diagnosis (years)	Mean (SD)	2.3 (4.5)
Gender	Women	61.2%
	Men	38.8%
Marital status	Single	12.7%
	Partnership, marriage	41.8%
	Divorced, separated	17.7%
	Widowed	27.8%
Housing situation	Living alone	48.1%
	Living in partnership and/or with children	43.0%
	Living with family of origin	7.6%
	Other	1.3%
Education	Compulsory school or less	19.0%
	Apprenticeship, professional school	73.4%
	A-level	5.1%
	University degree	2.5%
Employment status	Full time	5.1%
	Part time	1.3%
	Homemaker	7.7%
	Retired	79.5%
	Other	6.4%
Tumor type	Lung cancer	25.9%
	Hematologica malignancy	15.3%
	Breast cancer	11.8%
	Cancer with unknown primum	10.6%
	Gastroesophageal cancer	10.6%
	Colorectal cancer	5.9%
	Gynecological cancer	5.9%
	Pancreatic cancer	2.4%
	Laryngeal and pharyngeal cancer	2.4%
	other	9.4%
Metastases in patients with solid tumours (at last assessment)	Yes	81.9%
	No	18.1%
Current treatment phase^a	Supportive care	54.5%
	Chemotherapy	35.7%
	Radiotherapy	7.8%
	Radio- and chemotherapy	2.0%
Time to death (days)^a	Mean (SD)	109.8 (99.6)

Numbers refer to total number of assessments (n = 255).

SD, standard deviation.

QOL in patients at the end of life

QOL data assessed within 90 days prior to death was available from 65 patients (153 assessments in total). We found severe impairments with regard to physical functioning (mean score: 19.6 points), role functioning (9.4) and global QOL (31.9). Symptom burden was highest for fatigue (83.9), dyspnoea (65.5), appetite loss (61.2), pain (59.7), sleep disturbances (56.7), and taste alterations (52.0). For further details see Table 2.

Table 2.

Quality of Life Scores within Ninety Days Prior to Death (65 patients, 144 assessments)

			Percentiles
EORTC QLQ-C30	Mean	SD	10th	25th	50th	75th	90th
Functioning Scales ^a
Physical Functioning	19.6	14.0	2.0	6.7	20.0	26.7	40.0
Role Functioning	9.4	21.2	0.0	0.0	0.0	8.3	33.3
Social Functioning	82.0	31.3	30.0	66.7	100.0	100.0	100.0
Emotional Functioning	47.2	22.5	16.7	31.2	50.0	61.8	72.5
Cognitive Functioning	47.7	25.7	10.3	33.3	44.4	66.7	83.3
Global QOL	31.9	16.8	13.3	18.8	33.3	41.7	50.0
Symptom Scales ^b
Fatigue	83.9	17.2	61.1	77.8	88.9	100.0	100.0
Nausea/vomiting	30.0	28.1	0.0	0.0	25.0	47.2	70.0
Pain	59.7	31.5	15.0	33.3	62.5	90.3	100.0
Dyspnea	65.5	32.5	6.7	50.0	66.7	100.0	100.0
Sleep disturbances	56.7	33.5	0.0	33.3	58.3	86.1	100.0
Appetite loss	61.2	35.8	0.0	33.3	66.7	100.0	100.0
Constipation	32.6	35.9	0.0	0.0	16.7	58.3	100.0
Diarrhea	14.3	21.0	0.0	0.0	0.0	33.3	50.0
Financial impact	5.4	16.7	0.0	0.0	0.0	0.0	23.3
Taste alterations	52.0	34.0	0.0	23.6	55.6	83.3	100.0

High scores indicate low impairment.

High scores indicate high impairment.

SD, standard deviation; QOL, quality of life.

QOL trajectory toward the end of life

Changes in QOL over time were analyzed for the period of 1 year prior to death. Since there was no association of time to death with age, gender, or cancer type (solid tumor versus hematologic malignancy) we did not adjust for these variables.

For physical, role and cognitive functioning, as well as for global QOL and fatigue deterioration accelerated during the last 90 days of life. This means that these scales showed a nonlinear trajectory with a terminal drop at the end of life (see Figure 1 for an example).

FIG. 1.

Course of global quality of life (QOL) within 1 year prior to death: the circles reflect individual data points, whereas the line is the regression line for the respective time interval; high scores indicate good quality of life.

A steady linear decline was found for emotional functioning, nausea/vomiting, pain, dyspnea, appetite loss, financial impact, and taste alterations (see Figure 2 for an example). Social functioning, sleep disturbances, constipation and diarrhea did not change significantly during the study period. For further details see Table 3.

FIG. 2.

Course of taste alterations within 1 year prior to death: the circles reflect individual data points, whereas the line is the regression line for the respective time interval; high scores indicate high symptom severity.

Table 3.

Quality of Life Trajectory within One Year Prior to Death (57 patients, 222 assessments)

	Change per 30 days		Regression weights (CI 95%)			Overall change rate (Main effect)		Difference between change rates (Interaction effect)
EORTC QLQ-C30	90–365 days to death	Within 90 days to death	Main effect	Interaction	R ²	F	p	F	p
Functioning scales
Physical Functioning	−1.17	−6.87	−0.23 (−0.32, −0.14)	0.19 (0.08, 0.30)	0.209	41.58	0.000	11.47	0.001
Role Functioning	0.42	−4.93	−0.16 (−0.30, −0.03)	0.18 (0.01, 0.35)	0.013	2.28	0.134	4.45	0.038
Social Functioning	0.24	−3.47	−0.12 (−0.24, 0.01)	0.12 (−0.04, 0.29)	0.007	1.31	0.254	2.22	0.137
Emotional Functioning	−1.91	−2.92	−0.10 (−0.22, 0.03)	0.03 (−0.13, 0.20)	0.025	13.35	0.000	0.17	0.683
Cognitive Functioning	0.77	−7.19	−0.24 (−0.39, −0.09)	0.27 (0.07, 0.46)	0.031	3.24	0.075	7.14	0.009
Global QOL	−0.76	−4.60	−0.15 (−0.22, −0.08)	0.13 (0.04, 0.22)	0.092	26.07	0.000	7.96	0.006
Symptom Scales
Fatigue	0.61	4.81	0.16 (0.05, 0.27)	−0.14 (−0.28, 0.01)	0.058	11.37	0.001	3.99	0.048
Nausea/vomiting	2.75	1.46	0.05 (−0.08, 0.18)	0.04 (−0.12, 0.20)	0.064	18.76	0.000	0.28	0.600
Pain	2.08	2.15	0.07 (−0.09, 0.23)	0.00 (−0.21, 0.20)	0.031	7.88	0.006	0.00	0.982
Dyspneoa	1.35	4.99	0.17 (0.01, 0.33)	−0.12 (−0.33, 0.09)	0.022	8.35	0.005	1.29	0.259
Sleep disturbances	1.12	2.25	0.08 (−0.12, 0.27)	−0.04 (−0.29, 0.21)	0.020	2.46	0.120	0.09	0.767
Appetite loss	0.95	8.45	0.28 (0.07, 0.49)	−0.25 (−0.52, 0.02)	0.081	8.81	0.004	3.29	0.072
Constipation	1.10	−1.13	−0.04 (−0.21, 0.13)	0.07 (−0.14, 0.29)	0.003	0.55	0.458	0.47	0.496
Diarrhea	−0.49	3.58	0.12 (−0.01, 0.25)	−0.14 (−0.30, 0.03)	0.017	0.85	0.360	2.66	0.106
Financial impact	−1.52	0.26	0.01 (−0.06, 0.08)	−0.06 (−0.15, 0.03)	0.045	8.93	0.004	1.60	0.211
Taste alterations	2.01	1.43	0.05 (−0.14, 0.23)	0.02 (−0.22, 0.26)	0.029	4.71	0.032	0.03	0.871

CI, confidence interval; QOL, quality of life.

Changes in determinants of global QOL at the end of life

To analyze changes in determinants of global QOL toward the end of life, we calculated correlations of the global QOL scale with all other subscales of the QLQ-C30 and transformed correlation coefficients to Fisher Z values. This was done for the assessments between 365 to 90 days to death and for the period 30 days prior to death. Strength of the association with global QOL increased significantly for role Functioning (0.39 to 0.62), sleep disturbances (−0.33 to −0.58) and taste alterations (−0.37 to −0.60). For physical functioning the association with global QOL became significantly weaker (0.58 to 0.27). For further details see Table 4.

Table 4.

Determinants of Quality of Life for the Period 365–90 Days to Death and for the Period Less Than Thirty Days to Death (73 patients, 177 assessments)

	Correlations with global QOL 90–365 days to death		Less than 30 days to death
EORTC QLQ-C30	Pearson's r	Fishers's Z	Pearson's r	Fishers's Z	p
Functioning Scales
Physical Functioning	0.58	0.66	0.27	0.28	0.017
Role Functioning	0.39	0.41	0.62	0.73	0.046
Social Functioning	−0.06	−0.06	0.11	0.11	0.286
Emotional Functioning	0.58	0.66	0.56	0.63	0.860
Cognitive Functioning	0.44	0.47	0.57	0.65	0.239
Symptom Scales
Fatigue	−0.64	−0.76	−0.69	−0.85	0.535
Nausea/vomiting	−0.11	−0.11	−0.27	−0.28	0.311
Pain	−0.34	−0.35	−0.34	−0.35	0.966
Dyspneoa	−0.32	−0.33	−0.37	−0.39	0.701
Sleep disturbances	−0.33	−0.34	−0.58	−0.66	0.039
Appetite loss	−0.27	−0.28	−0.49	−0.54	0.091
Constipation	−0.11	−0.11	−0.20	−0.20	0.584
Diarrhea	−0.20	−0.20	−0.05	−0.05	0.337
Financial impact	−0.07	−0.07	−0.17	−0.17	0.523
Taste alterations	−0.37	−0.39	−0.60	−0.69	0.047

QOL, Quality of life.

Discussion

Knowledge of QOL in patients with cancer at the end of life is of importance. As a result of considerable cancer mortality, a large number of patients are affected. Our results highlight that determinants of global QOL shift toward the end of life possibly reflecting a change in the relative importance of symptoms. In addition, our results show that as expected most aspects of QOL are substantially impaired in patients with advanced cancer. Physical, role and cognitive functioning, as well as global QOL and fatigue showed a rapid decline during the last 3 months of life. Symptom burden with regard to emotional functioning, nausea/vomiting, pain, dyspnea, appetite loss, and taste alterations increased at a steady rate throughout the last year of life. Sleep disturbances were found to be stable at a high impairment level. These results are mostly consistent with the literature.^3–7,16 An overview of results from similar studies is given in Table 5.

Table 5.

Quality of Life Scores from Various Patient Groups with Advanced Cancer and from General Population

Study	Sahlberg-Blom et al. 2001	Lundh-Hagelin et al. 2006		Ahnler-Elmqvist et al. 2008		Beijer et al. 2008	Johnson et al. 2009	Giesinger et al. 2011	Schwarz and Hinz 2000
Sample	Survival < 30 days	Patients with advanced cancer	Survival < 90 days	AHC	CC	Preterminal cancer patients	Advanced cancer patients	Survival < 90 days	German general population: Total	German general population: > 60 yrs
EORTC QLQ-C30	n = 47	n = 278	n = 194	n = 119	n = 178	n = 98	n = 977	n = 65	n = 2028	n = 666
Functioning Scales
Physical	15	27	23	29	51	45	76	19.6	90	80
Role	13	14	13	15	37	35	68	9.4	88	79
Social	42	45	42	45	58	70	80	82.0	91	86
Emotional	62	60	57	50	63	67	78	47.2	79	78
Cognitive	65	60	58	58	73	76	81	47.7	91	85
Global QOL	27	31	27	35	44	49	66	31.9	71	61
Symptom Scales
Fatigue	79	80	81	77	56	62	37	83.9	17	27
Nausea/vomiting	23	31	33	32	24	19	8	30.0	3	3
Pain	42	49	52	53	41	36	25	59.7	15	26
Dyspnea	51	51	55	56	40	39	18	65.5	8	17
Sleep disturbances	29	37	38	36	33	21	24	56.7	16	28
Appetite loss	54	59	61	53	41	46	17	61.2	5	8
Constipation	26	35	35	40	31	22	12	32.6	4	7
Diarrhoea	19	24	25	18	14	8	13	14.3	3	3
Financial impact	14	17	15	25	24	8	11	5.4	6	10

AHC, advanced home care; CC, conventional care; QOL, quality of life.

Interestingly, social functioning in our patient group was close to scores from general population,¹⁷ whereas Lundh Hagelin et al.⁵ found considerably lower social functioning in their sample of patients at the end of life. This discrepancy is not limited to the last days of life as in both studies social functioning did not decline significantly over time. We suppose that this finding may be explained partly by the fact that most patients treated at our institution lived in rural or semirural areas with good infrastructure. Tight social and family networks are more common in such areas and infrastructure may be a relevant determinant of patients' integration in their social environment.

As mentioned above the QOL trajectory in our study was characterized by an accelerating decline of physical, role and cognitive functioning, and global QOL and fatigue. Similar to this, Elmqvist et al.⁷ found the QOL trajectory to be characterized by a rapid deterioration of physical, cognitive, and social functioning and fatigue at the end of life.

Hwang et al.⁴ investigated deterioration patterns by means of the FACT-G. Emotional well-being declined at a steady rate, whereas social well-being stayed unchanged over time. Deterioration rate increased by a factor of 3 for physical well-being, by a factor of 4 for overall QOL, and by a factor of 5.5 for functional well-being. These results are very much in concordance with findings from our study. With regard to determinants of global QOL we found the strongest overall associations with global QOL for fatigue and emotional functioning. During the last 30 days of life highest correlations were found for fatigue, role functioning and taste alterations. These results are comparable to those of the study by Beijer et al.⁶ who also reported fatigue to be the most important determinant of global QOL. More detailed analyses revealed that the determinants of global QOL changed significantly toward death. Physical functioning became less important to patients' global QOL, whereas the impact of taste alterations, role functioning, and sleep disturbances became stronger. Role functioning as measured by the QLQ-C30 covers work, hobbies, and recreational or other daily activities. The increase of the association of global QOL with role functioning may indicate that patients still being able to participate in such activities at the end of life, take a considerable benefit from it. But also fundamental needs and pleasures like enjoying food and sleeping well may gain importance as physical capabilities decrease and consequently other sources of pleasure (e.g., recreational activities, hobbies) diminish. Note that the level of role functioning decreased rapidly toward death whereas its association with global QOL increased.

Taking a different methodological approach, Strömgren et al.¹⁶ asked patients with advanced cancer to prioritize the five most distressing symptoms. Pain, fatigue, physical function, appetite reduction, nausea/vomiting, dyspnea, and depression were the symptoms most often mentioned. When investigating changes in symptom prioritization over time, the authors found that pain lost importance whereas fatigue was prioritized higher. Based on correlation analysis and not on symptom prioritization, we could not replicate this shift in the importance of pain and fatigue. However, symptoms with a high priority to patients (i.e., fatigue, physical functioning, and depression) were found to be associated with global QOL in our study.

In our study, we tried to extend the range of assessed patients by performing assessments as interviews in case of patients being physically not capable of using the tablet-PC. Also, in case of very bad physical condition patients did not drop out of the monitoring but were again assessed as soon as their health status improved sufficiently. Interestingly, a recent study on the impact of such bias found that it may be of minor importance in patients with advanced cancer.¹⁸ A further limitation of our statistical analysis is the somewhat arbitrary cut-off point at 3 months that we used to investigate nonlinear QOL trajectories. It is assumed that the phenomenon of a terminal drop starts for different scales at different cutoff points, however, larger data sets are needed to model nonlinear trajectories more accurately. Sample size also did not allow comparing QOL trajectories across patient sub-groups defined by, e.g., diagnosis, gender, current treatment phase, or age. Another limitation is the heterogeneity of our sample comprising patients with solid tumours as well as patients with hematologic malignancies. Also, collecting data as part of routine PRO monitoring at the hospital may introduce bias as patients in a bad health status are likely to be assessed more often than patients in a good health status due to frequency of hospital visits. Bias in the opposite direction may arise from the fact that patients in a very bad condition could not be assessed. This is likely to result in underestimation of true variance.

Nevertheless, our results highlight the importance of the assessment of QOL in general and taste alterations in particular within palliative care. Taste alterations are an important issue not only because of their high severity and prevalence, but also because of the high impact on global QOL, especially during the last days of life. As taste alterations are known to be related not only to the cancer itself,¹⁹ but also to medical interventions like chemotherapy^15,20,21 or radiotherapy^22–23 this symptom should be taken into account within medical decision-making in palliative care. However, our results are rather contrary to the study by Strömgren et al.,¹⁶ who found based on patients' symptom prioritization that taste alterations are only of minor importance.

In addition, our study showed that routine electronic QOL data capture is feasible in a palliative care setting, and even patients close to death are not necessarily prevented from reporting on their QOL.⁵ Routine standardized self-reports on physical and psychosocial symptoms may contribute to symptom management and improve patient–physician communication, as has been suggested for other hospital settings.²⁴ Surely, assessing the patient requires a prior appraisal of patients' capability to participate, as it poses a certain burden. Using a shorter questionnaire may help to meet this issue, but naturally limits collected information. Therefore, we preferred the QLQ-C30 over its palliative care version the QLQ-PAL-15, as this short-form provides only a subset of the QLQ-C30 scales. In opposition to Lundh Hagelin et al.,⁵ we claim that the measurement range of the EORTC QLQ-C30 is adequate for preterminal cancer patients. For most scales, we found only small floor or ceiling effects, especially for the symptom scales they appear to be less pronounced than in patient groups with a better health status. This is also supported by a study of Jocham et al.²⁵ on reliability and validity of the QLQ-C30 in palliative care patients with cancer.

In conclusion, our study showed that until 3 months prior to death various aspects of QOL remained rather stable, although on a considerable level of impairment. The shift in determinants of global QOL shown in our study is important to provision of supportive care in this patient group. Specifically, the increasing importance of taste alterations and sleep disturbances could be met by adequate medication or provision of self-care and symptom management strategies, e.g., education sheets concerning food and utensils, dietary intervention or adapted oral hygiene. However, certain symptoms like fatigue or appetite loss may be difficult to manage and rather be a natural phenomenon at the end of life.

Since in cancer patients at the end of life symptom management and patients' QOL is a major focus, a standardized assessment of QOL should be an integral part of clinical routine.

Footnotes

Acknowledgment

The project was partly funded by the “Jubiläumsfond” of the Austrian National Bank (#12207) and “Qualitätsförderprogramm Tirol.” The work of Lisa M. Wintner was funded by a scholarship from the Leopold-Franzens-University in Innsbruck, Austria. We would like to thank Marianna Pircher for her contribution to data collection.

Author Disclosure Statement

No competing financial interests exist.

References

Ferlay J Parkin

, Steliarova-Foucher

. Estimates of cancer incidence and mortality in Europe in 2008. Eur J Cancer, 2010; 46:765–781.

Murray

, Fiset

, Young

, Kryworuchko

. Where the dying live: A systematic review of determinants of place of end-of-life cancer care. Oncol Nurs Forum, 2009; 36:69–77.

Sahlberg-Blom

, Ternestedt

, Johansson

. Is good ‘quality of life’ possible at the end of life? An explorative study of the experiences of a group of cancer patients in two different care cultures. J Clin Nurs, 2001; 10:550–562.

Hwang

, Chang

, Fairclough

, Cogswell

, Kasimis

. Longitudinal quality of life in advanced cancer patients: Pilot study results from a VA medical cancer center. J Pain Symptom Manage, 2003; 25:225–235.

Lundh Hagelin

, Seiger

, Furst

. Quality of life in terminal care—With special reference to age, gender and marital status. Support Care Cancer, 2006; 14:320–328.

Beijer

, Kempen

, Pijls-Johannesma

, de Graeff

, Dagnelie

. Determinants of overall quality of life in preterminal cancer patients. Int J Cancer, 2008; 123:232–235.

Elmqvist

, Jordhoy

, Bjordal

, Kaasa

, Jannert

. Health-related quality of life during the last three months of life in patients with advanced cancer. Support Care Cancer, 2009; 17:191–198.

Higginson

, Costantini

. Dying with cancer, living well with advanced cancer. Eur J Cancer, 2008; 44:1414–1424.

Johnsen

, Petersen

, Pedersen

, Groenvold

. Symptoms and problems in a nationally representative sample of advanced cancer patients. Palliat Med, 2009; 23:491–501.

10.

Nekolaichuk

, Bruera

, Spachynski

, MacEachern

, Hanson

, Maguire

. A comparison of patient and proxy symptom assessments in advanced cancer patients. Palliat Med, 1999; 13:311–323.

11.

ESD Inc.: E.S.D., Computer-Based Health Evaluation System (CHES). 2010 Innsbruck, Austria.

12.

Erharter

, Giesinger

, Kemmler

, Schauer-Maurer

, Stockhammer

, Muigg

, Hutterer

, Rumpold

, Sperner-Unterweger

, Holzner

. Implementation of computer-based quality-of-life monitoring in brain tumor outpatients in routine clinical practice. J Pain Symptom Manage, 2010; 39:219–229.

13.

Oberaigner

, Stuhlinger

. Record linkage in the Cancer Registry of Tyrol, Austria. Methods Inf Med, 2005; 44:626–630.

14.

Aaronson

, Ahmedzai

, Bergman

et al. The European Organization for Research and Treatment of Cancer QLQ-C30: a quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst, 1993; 85,5:365–76.

15.

Zabernigg

, Gamper

, Giesinger

, Bullinger

, Cull

, Duez

, Filiberti

, Flechtner

, Fleishman

, de Haes

et al. Taste alterations in cancer patients receiving chemotherapy: A neglected side effect? Oncologist, 2010; 15:913–920.

16.

Stromgren

, Sjogren

, Goldschmidt

, Petersen

, Pedersen

, Groenvold

. Symptom priority and course of symptomatology in specialized palliative care. J Pain Symptom Manage, 2006; 31:199–206.

17.

Schwarz

, Hinz

. Reference data for the quality of life questionnaire EORTC QLQ-C30 in the general German population. Eur J Cancer, 2001; 37:1345–1351.

18.

Petersen

, Pedersen

, Groenvold

. Does nonparticipation in studies of advanced cancer lead to biased quality-of-life scores? J Palliat Med, 2009; 12:1023–1028.

19.

Johnson

. Alterations in taste sensation: A case presentation of a patient with end-stage pancreatic cancer. Cancer Nurs, 2001; 24:149–155.

20.

Comeau

, Epstein

, Migas

. Taste and smell dysfunction in patients receiving chemotherapy: A review of current knowledge. Support Care Cancer, 2001; 9:575–580.

21.

Bernhardson

, Tishelman

, Rutqvist

. Self-reported taste and smell changes during cancer chemotherapy. Support Care Cancer, 2008; 16:275–283.

22.

Ripamonti

, Fulfaro

. Taste alterations in cancer patients. J Pain Symptom Manage, 1998; 16:349–351.

23.

Ruo Redda

, Allis

. Radiotherapy-induced taste impairment. Cancer Treat Rev, 2006; 32:541–547.

24.

Luckett

, Butow

, King

. Improving patient outcomes through the routine use of patient-reported data in cancer clinics: Future directions. Psychooncology, 2009; 18:1129–1138.

25.

Jocham

, Dassen

, Widdershoven

, Halfens

. Quality-of-life assessment in a palliative care setting in Germany: An outcome evaluation. Int J Palliat Nurs, 2009; 15:338–345.