- Email: [email protected]
Symptom Expression in the Last Seven Days of Life Among Cancer Patients Admitted to Acute Palliative Care Units
Accepted Manuscript Symptom Expression in the Last Seven Days of Life Among Cancer Patients Admitted to Acute Palliative Care Units David Hui, MD, MSc Renata dos Santos, MD Gary B. Chisholm, MS Eduardo Bruera, MD PII:
S0885-3924(14)00466-7
DOI:
10.1016/j.jpainsymman.2014.09.003
Reference:
JPS 8753
To appear in:
Journal of Pain and Symptom Management
Received Date: 21 June 2014 Revised Date:
28 August 2014
Accepted Date: 12 September 2014
Please cite this article as: Hui D, Santos Rd, Chisholm GB, Bruera E, Symptom Expression in the Last Seven Days of Life Among Cancer Patients Admitted to Acute Palliative Care Units, Journal of Pain and Symptom Management (2014), doi: 10.1016/j.jpainsymman.2014.09.003. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Original Article
14-00326R1
Symptom Expression in the Last Seven Days of Life Among Cancer Patients Admitted to
RI PT
Acute Palliative Care Units
David Hui, MD, MSc, Renata dos Santos, MD, Gary B. Chisholm, MS, and Eduardo Bruera,
SC
MD
M AN U
Department of Palliative Care and Rehabilitation Medicine (D.H., E.B.), Department of Biostatistics (G.B.C), and Division of Bioinformatics and Computational Biology (G.B.C.), M. D. Anderson Cancer Center, Houston, Texas, USA; and Department of Palliative Care (R.d.S.),
Address correspondence to: David Hui, MD, MSc
TE D
Barretos Cancer Hospital, Barretos, Brazil
EP
Department of Palliative Care and Rehabilitation Medicine, Unit 1414 University of Texas M. D. Anderson Cancer Center
AC C
1515 Holcombe Boulevard Houston, TX 77030, USA
E-mail: [email protected]
ACCEPTED MANUSCRIPT
Abstract Context. The symptom burden in the last week of life of patients with advanced cancer has not been well characterized.
RI PT
Objectives. To examine the frequency, intensity, and predictors for symptoms in the last seven days of life among patients who were able to communicate and died in an acute palliative care unit (APCU).
SC
Methods. We systematically documented the Edmonton Symptom Assessment System (ESAS) daily and 15 symptoms twice daily on consecutive advanced cancer patients admitted to
M AN U
APCUs at M.D. Anderson Cancer Center (USA) and Barretos Cancer Hospital (Brazil) from admission to death or discharge in 2010/2011. We examined the frequency and intensity of the symptoms from death backwards.
Results. A total of 203 of 357 patients died. The proportion of patients able to
TE D
communicate decreased from 80% to 39% over the last seven days of life. ESAS anorexia (P=0.001 in longitudinal analyses), drowsiness (P<0.0001), fatigue (P<0.0001), poor well-being (P=0.01) and dyspnea (P<0.0001) increased in intensity closer to death. In contrast, ESAS
EP
depression (P=0.008) decreased over time. Dysphagia to solids (P=0.01) and liquids (P=0.005), and urinary incontinence (P=0.0002) also were present in an increasing proportion of patients in
AC C
the last few days of life. In multivariate analyses, female sex, non-Hispanic race and lung cancer were significantly associated with higher ESAS symptom expression (odds ratio>1). Conclusion. Despite intensive management in APCUs, some cancer patients continue to experience high symptom burden as they approached death.
2
ACCEPTED MANUSCRIPT
Key Words: anorexia, deglutition disorders, death, depression, dyspnea, neoplasms, pain, palliative care, symptom assessment, urinary incontinence
RI PT
Running title: Symptom Expression in the Last Days of Life
AC C
EP
TE D
M AN U
SC
Accepted for publication: September 12, 2014.
3
ACCEPTED MANUSCRIPT
Introduction Acute palliative care units (APCUs) are specialized inpatient care settings that provide intensive symptom control and transition of care for patients with advanced diseases and severe
RI PT
distress (1, 2). APCUs are becoming increasingly available in cancer centers (3). Care is
delivered by an interprofessional team to address the physical, psychosocial, and spiritual needs of patients and their families. Several studies examining disease trajectory have shown that
SC
cancer patients develop increasing symptom burden and decreasing function over the last few months to weeks of life (4-7). However, there has been a paucity of studies examining the
M AN U
symptom burden in dying cancer patients (i.e., days of life expectancy) (8-11), and no studies in APCUs where patients receive intensive management focusing on quality of life. A better characterization of symptom intensity in the last few days of life could help clinicians to better understand patients’ needs, to improve their quality of life and to optimize the
TE D
quality of end-of-life care. We previously documented the clinical signs associated with impending death in a cohort of cancer patients (12-14). In the present study, we examined the frequency and intensity of various symptoms, and predictors of symptom expression in patients
Methods
EP
with advanced cancer admitted to APCUs.
AC C
Study Setting and Criteria
This study was part of the “Investigating the Process of Dying” study, a prospective observational study that systematically documents the signs and symptoms in the last days of life. The methodology of this study has been reported in detail elsewhere (14, 15). Briefly, we enrolled consecutive adult patients with advanced cancer who were admitted to the APCUs at M. D. Anderson Cancer Center (MDACC, Houston, TX, USA) between April 5 and July 6, 2010,
4
ACCEPTED MANUSCRIPT
and at Barretos Cancer Center (BCH, Barretos, Brazil) between January 27 and June 1, 2011. This study was approved by the Institutional Review Boards at both sites with waiver of informed consent for patient participation. All clinicians who participated in this study provided
RI PT
written informed consent prior to patient enrollment. Data Collection
We identified target symptoms for this study by conducting a literature review of
SC
published articles (8, 9, 16, 17) and educational materials (18), and discussing relevant variables with participating palliative care clinicians. The symptoms were collected from patients who
M AN U
were able to communicate verbally and interact with our staff. We documented the Edmonton Symptom Assessment System (ESAS) daily. The ESAS is a validated numeric rating scale that assesses the average intensity of 10 symptoms (pain, fatigue, nausea, depression, anxiety, drowsiness, shortness of breath, anorexia, poor well-being and insomnia) over the past 24 hours
TE D
from 0 (absent) to 10 (worst possible) (19-22). We collected ESAS data from patients directly as long as they were able to communicate and answer the study assessments, regardless of whether they were delirious or not because symptom expression was the main outcome of interest. The
EP
ESAS data also were collected from caregivers when patients had impaired communication capacity, but these data were not used because the current study focuses on patient-reported
AC C
outcomes only.
Data collected also included the presence or absence of 15 other symptoms every 12 hours, including epistaxis, cough, dysphagia to solids, dysphagia to liquids, upper gastrointestinal (GI) bleed, lower GI bleed, constipation, diarrhea, fecal incontinence, urinary incontinence, hematuria, vaginal bleed (female only), vaginal discharge (female only), myoclonus and sweating. Standardized data collection forms were completed by full-time
5
ACCEPTED MANUSCRIPT
palliative care nurses who were familiar with end-of-life care every 12-hour shift from admission to discharge or death independent of prior assessments. These symptoms were documented by bedside nurses instead of reported directly by patients because they can be observed objectively.
RI PT
As documented previously, extensive training and monitoring of palliative care nurses was provided before and during the data collection period (12). In Brazil, all data collection forms
conferences weekly to discuss issues related to data collection. Statistical Analysis
SC
were translated into Portuguese and then back-translated. The two study sites held video
M AN U
Our pre-planned sample size was a combined total of 200 deaths at the two study sites to examine the physical signs of impending death, as reported elsewhere (12). This analysis was planned based on the combined data a priori. This sample size also provided adequate power for our secondary multiple regression analysis, which requires 10 patients for each independent
TE D
variable included in the model.
We summarized the baseline demographics with descriptive statistics. We determined the frequency of each symptom from death backwards for all patients who died in the APCUs. We
EP
plotted the average ESAS intensity, the proportion of patients with ESAS scores ≥4, and the
of life.
AC C
proportion of patients with the other 15 symptoms mentioned previously over the last seven days
We calculated the statistical significance for longitudinal change using the generalized linear mixed model (GLMM) for average ESAS intensity, and the generalized estimating equation (GEE) for the proportion of patients with ESAS scores ≥4. We also employed a multivariate GEE model incorporating day prior to death, age, sex, race, cancer, and marital status to identify factors associated with ESAS intensity ≥4. We accounted for correlation within
6
ACCEPTED MANUSCRIPT
patient in the GLMM and GEE analyses using an autogressive correlation structure with a lag of one. Overall survival was calculated from the time of APCU admission. We collected
RI PT
survival data from electronic health records. The Statistical Analysis System (SAS version 9.2, SAS Institute Inc., Cary, NC) was used for statistical analysis. A P-value of <0.05 was considered significant.
SC
Results Patient Characteristics
M AN U
The patient characteristics have been reported elsewhere (12). Briefly, 52 of 151 and 151 of 206 patients died in the APCUs at MDACC and BCH, respectively. Among the 203 APCU deaths, the average age was 58 years (range 18-88), 99 (49%) were female, 158 (78%) were Hispanic, and 63 (33%) had gastrointestinal malignancies. The proportion of patients who were
TE D
able to communicate and to provide patient-reported outcomes decreased over time (Fig. 1). Frequency and Intensity of ESAS Symptoms
Fig. 2A shows that, among responding patients, the average intensity of ESAS anorexia
EP
(P=0.001 for longitudinal trend), drowsiness (P<0.0001), fatigue (P<0.0001), poor well-being (P=0.01) and dyspnea (P<0.0001) was moderate to severe seven days before death, and
AC C
increased as death approached. In contrast, depression (P=0.008) had lower intensity and decreased over time. Anxiety (P=0.09), insomnia (P=0.31), pain (P=0.15) and nausea (P=0.53) did not change significantly over the last seven days. The proportion of patients with ESAS scores ≥4 of 10 are shown in Fig. 2B, and the findings were similar. Moderate/severe drowsiness (P=0.03 for longitudinal trend), fatigue (P=0.006), poor well-being (P=0.19), and dyspnea (P=0.006) were present in an increasing
7
ACCEPTED MANUSCRIPT
majority of patients. The proportion of patients with moderate to severe pain (P=0.42), anorexia (P=0.16), anxiety (P=0.30), depression (P=0.78) insomnia (P=0.35) and nausea (P=0.07) remained similar over time.
RI PT
Frequency of Other Symptoms
A number of symptoms were present in 30-60% of communicating patients in the last few days of life, including dysphagia to solids (P=0.01 for longitudinal trend), dysphagia to
SC
liquids (P=0.005), urinary incontinence (P=0.0002) and constipation (P=0.66) (Fig. 3). Fecal incontinence (P=0.08), cough (P=0.77), sweating (P=0.06) and myoclonus (P=0.17) were
M AN U
experienced by 10-30% of patients. Other symptoms such as epistaxis (P=0.08), hematuria (P=0.07), upper GI bleed (P=0.008), lower GI bleed (P=0.60), diarrhea (P=0.39), vaginal bleed (P=0.80) and vaginal discharge (P=0.10) were detected in less than 10% of patients. Factors Associated With Moderate and High ESAS Intensity
TE D
On multivariate analysis (Table 1), female sex was associated with higher expression of nausea, drowsiness and anorexia; lung cancer was associated with a higher level of dyspnea; and non-Hispanic race was associated with higher expression of fatigue, anxiety, drowsiness, poor
Discussion
EP
well-being, dyspnea and insomnia.
AC C
We documented the trajectory of 25 symptoms in patients admitted to APCUs in the last seven days of life. Among patients who were able to communicate their concerns, anorexia, drowsiness, fatigue, poor well-being, dyspnea, dysphagia and incontinence increased over the last days of life. In contrast, depression decreased with time. Female sex, non-Hispanic race and diagnosis of lung cancer were associated with higher symptom expression.
8
ACCEPTED MANUSCRIPT
he proportion of patients who were able to communicate their symptoms decreased over time, although a small proportion retained their ability to do so even close to death. Our findings are generally consistent with another study by Morita et al. (23). However, the proportion of
RI PT
patients with impaired communication capacity was higher in that study, which could be
explained by differences in patient population and methods of assessment. Taken together, our findings suggest that it is possible to document patient-reported outcomes even in the last days of
SC
life. This is particularly important because caregivers’ and health care professionals’ proxy assessments are often less reliable (24).
M AN U
This is the first study to serially and systematically profile the symptoms in patients admitted to APCUs over the last seven days of life. Interestingly, the scant literature on symptoms in the last days of life focused primarily on the last 48 hours (9, 25, 26) and in other care settings (e.g., hospices, nursing homes) (8, 27, 28). We found that the frequency of many
TE D
symptoms was high and increased over time. The high symptom burden is likely because 1) patients admitted to APCUs were selected for their level of distress and refractory symptoms (29, 30) and 2) the end of life is associated with decreased function and various irreversible
EP
processes, such as post-obstructive pneumonia and bowel obstruction. In the APCUs, patients all underwent comprehensive symptom assessments and intensive
AC C
symptom management on a daily basis by highly trained palliative care specialists and interprofessional teams using well-defined pharmacological and non-pharmacological interventions. Notably, pain and nausea were relatively well controlled in the last days of life, likely related to the relatively effective treatment options available. Depression also improved significantly—this may be related to the intensive psychological interventions provided by the members of the APCU interprofessional team, such as chaplains, child life care specialists,
9
ACCEPTED MANUSCRIPT
psychologists, social workers, nurses and physicians, to support the patients and their families. Patients also may naturally feel less depressed at they enter the final stage of their lives, having come to the resolution that death is inevitable, consistent with another longitudinal study that
RI PT
examined the ESAS in cancer patients at home (4). However, we cannot rule out significant depression in patients who were unable to communicate in the last days of life, which also could contribute to our finding. Further studies are needed to examine the psychological aspects in the
SC
last days of life. Our findings suggest that a number of symptoms, such as dyspnea, fatigue and drowsiness, may be unavoidable in a proportion of patients even with intensive state-of-the-art
M AN U
supportive care. At the same time, it calls for more palliative care research in this population and setting to relieve suffering in the most distressed.
Our results are consistent with those of a study on ESAS symptom trajectories in the last few months of life among cancer patients in the home setting (4). However, limited data were
TE D
presented on the last week of life in that study, making our current findings unique and complementary. Although several studies have examined symptom prevalence in the last 48 hours of life, only one study has documented the symptom intensity using the ESAS visual
EP
analogue scale among nursing home residents (25). Among the 253 conscious patients, anorexia, drowsiness, and poor well-being were rated high in intensity, which is consistent with our study.
AC C
However, dyspnea was reported to be mild (median=10/100) whereas it was moderate to severe in our study. Dyspnea is one of the most distressing and under-studied symptoms in cancer patients (31, 32). Research is urgently needed to examine novel interventions to alleviate this devastating symptom in patients at the end of life. In contrast, pain, nausea and depression were generally well controlled in the last days of life, likely because of the effective therapeutic options available.
10
ACCEPTED MANUSCRIPT
In addition to the ESAS symptoms, our study revealed that dysphagia and urinary/fecal incontinence were more prevalent as patients approached death. This is not surprising given their deterioration in neurological and motor function. Constipation remained common at the end
RI PT
of life, despite the fact that laxatives were often administered in the APCUs. The high frequency of these distressing symptoms supports the need for routine symptom screening and monitoring, which would allow us to initiate appropriate palliative measures promptly. These may include a
SC
urinary catheter or adult diapers for urinary incontinence, and change of medication route for patients with dysphagia. In contrast, other symptoms such as myoclonus, sweating and cough
M AN U
were less frequent.
We found that patients with lung cancer generally had higher levels of dyspnea and anxiety compared with patients with other malignancies at the end of life. This is consistent with other studies demonstrating the high symptom burden in lung cancer patients, with dyspnea
TE D
being the most distressing concern (4, 31, 33). We also found that Hispanic race was associated with lower ESAS symptom scores, which may potentially be explained by differences in cultures and symptom expression (34). Consistent with another study, we found that females had higher
EP
nausea and drowsiness (4).
This study has several limitations. First, we only included two institutions with a
AC C
relatively small patient population. Larger studies are required to validate our findings. Second, we did not examine other symptoms such as xerostomia and bloating. Future research should consider assessing these symptoms. Third, the symptoms were collected by palliative care nurses instead of physicians because they have more time to observe patients. Our nurses were highly trained in symptom assessment, and all received standardized training to ensure high data quality. Fourth, some of the symptoms in this study were observed by nurses (e.g., cough,
11
ACCEPTED MANUSCRIPT
incontinence) instead of reported by patients. Finally, findings from this study are only applicable to cancer patients admitted to APCUs. Because symptom expression may differ by disease and setting (28), further studies are needed to characterize the symptom burden in other
RI PT
populations in the last days of life.
We documented the frequency, intensity, and predictors of various symptoms in cancer patients admitted to APCUs, and identified increasing symptom burden in imminently dying
SC
cancer patients. Lung cancer, female sex and Hispanic race were associated with higher
symptom expression. With the recent interest in patient-reported outcomes and quality of care
M AN U
measures, our study provides information on which symptoms can and cannot be palliated effectively with state-of-the-art supportive care. Further research is urgently needed to improve the quality of life for patients in the last days of life. Disclosure and Acknowledgments
TE D
This work was supported in part by National Institutes of Health grants RO1NR01016201A1, RO1CA122292-01, and RO1CA124481-01 (Dr. Bruera). This study is also supported by the M. D. Anderson Cancer Center Support Grant (CA 016672) and an institutional start-up grant
EP
#18075582 (Dr. Hui). All authors report no relevant financial conflicts of interest The authors thank all the patients, clinical nurses and physicians who participated in this
AC C
study and provided valuable data. References 1.
Hui D, Elsayem A, Palla S, et al. Discharge outcomes and survival of patients with advanced cancer admitted to an acute palliative care unit at a comprehensive cancer center. J Palliat Med 2010;13:49-57.
12
ACCEPTED MANUSCRIPT
2.
Hui D, Elsayem A, Li Z, et al. Antineoplastic therapy use in patients with advanced cancer admitted to an acute palliative care unit at a comprehensive cancer center: a simultaneous care model. Cancer 2010;116:2036-2043. Hui D, Elsayem A, De la Cruz M, et al. Availability and integration of palliative care at
RI PT
3.
US cancer centers. JAMA 2010;303:1054-1061. 4.
Seow H, Barbera L, Sutradhar R, et al. Trajectory of performance status and symptom
SC
scores for patients with cancer during the last six months of life. J Clin Oncol 2011;29:1151-1158.
Mercadante S, Casuccio A, Fulfaro F. The course of symptom frequency and intensity in
M AN U
5.
advanced cancer patients followed at home. J Pain Symptom Manage 2000;20:104-112. 6.
Ventafridda V, Ripamonti C, De Conno F, Tamburini M, Cassileth BR. Symptom prevalence and control during cancer patients' last days of life. J Palliat Care 1990;6:7-11. Morris JN, Suissa S, Sherwood S, Wright SM, Greer D. Last days: a study of the quality
TE D
7.
of life of terminally ill cancer patients. J Chronic Dis 1986;39:47-62. 8.
Morita T, Tsunoda J, Inoue S, Chihara S. Contributing factors to physical symptoms in
EP
terminally-ill cancer patients. J Pain Symptom Manage 1999;18:338-346. Lichter I, Hunt E. The last 48 hours of life. J Palliat Care 1990;6:7-15.
10.
Steindal SA, Ranhoff AH, Bredal IS, Sorbye LW, Lerdal A. Last three days of life in the
AC C
9.
hospital: a comparison of symptoms, signs and treatments in the young old and the oldest old patients using the Resident Assessment Instrument for Palliative Care. Int J Older People Nurs 2012;8:199-206.
13
ACCEPTED MANUSCRIPT
11.
Hui D, Nooruddin Z, Didwaniya N, et al. Concepts and definitions for "actively dying," "end of life," "terminally ill," "terminal care," and "transition of care": a systematic review. J Pain Symptom Manage 2014;47:77-89. Hui D, dos Santos R, Swati B, et al. Clinical signs of impending death in cancer patients.
RI PT
12.
Oncologist 2014;19:681-687. 13.
Bruera S, Chisholm G, Santos RD, et al. Variations in vital signs in the last days of life in
SC
patients with advanced cancer. J Pain Symptom Manage 2014 April 14. [Epub ahead of print].
Hui D, Dos Santos R, Chisholm G, et al. Bedside clinical signs associated with
M AN U
14.
impending death in patients with advanced cancer: preliminary findings. Cancer 2014, in press. 15.
Hui D, Dos Santos R, Chisholm G, et al. Clinical Signs of Impending Death in Cancer
TE D
Patients. Oncologist 2014;19:681-687. NOTE: DUPLICATE OF REF 12 16.
Ferris FD. Last hours of living. Clin Geriatr Med 2004;20:641-667, vi.
17.
Plonk WM, Jr., Arnold RM. Terminal care: the last weeks of life. J Palliat Med
Emanuel LL, Ferris FD, von Gunten CF: Module 6: Last hours of living. EPEC-O: Education in Palliative and End-of-life Care for Oncology. Chicago, IL: The EPEC
AC C
18.
EP
2005;8:1042-1054.
Project, 2005. Available from http://www.cancer.gov/cancertopics/cancerlibrary/epeco/selfstudy/module-6-pdf. Accessed June 12, 2012.
14
ACCEPTED MANUSCRIPT
19.
Bruera E, Kuehn N, Miller MJ, Selmser P, Macmillan K. The Edmonton Symptom Assessment System (ESAS): a simple method for the assessment of palliative care patients. J Palliat Care 1991;7:6-9. Chang VT, Hwang SS, Feuerman M. Validation of the Edmonton Symptom Assessment Scale. Cancer 2000;88:2164-2171.
21.
Richardson LA, Jones GW. A review of the reliability and validity of the Edmonton
SC
Symptom Assessment System. Curr Oncol 2009;16:55. 22.
RI PT
20.
Nekolaichuk C, Watanabe S, Beaumont C. The Edmonton Symptom Assessment System:
M AN U
a 15-year retrospective review of validation studies (1991--2006). Palliat Med 2008;22:111-122. 23.
Morita T, Tei Y, Inoue S. Impaired communication capacity and agitated delirium in the final week of terminally ill cancer patients: prevalence and identification of research
24.
TE D
focus. J Pain Symptom Manage 2003;26:827-834.
Hui D, Morgado M, Vidal M, et al. Dyspnea in hospitalized advanced cancer patients: subjective and physiologic correlates. J Palliat Med 2013;16:274-280. Brandt HE, Ooms ME, Deliens L, van der Wal G, Ribbe MW. The last two days of life of
EP
25.
nursing home patients--a nationwide study on causes of death and burdensome symptoms
26.
AC C
in The Netherlands. Palliat Med 2006;20:533-540. Hall P, Schroder C, Weaver L. The last 48 hours of life in long-term care: a focused chart audit. J Am Geriatr Soc 2002;50:501-506. 27.
Conill C, Verger E, Henriquez I, et al. Symptom prevalence in the last week of life. J Pain Symptom Manage 1997;14:328-331.
15
ACCEPTED MANUSCRIPT
28.
Veerbeek L, van Zuylen L, Swart SJ, van der Maas PJ, van der Heide A. The last 3 days of life in three different care settings in The Netherlands. Support Care Cancer 2007;15:1117-1123. Bruera E, Neumann C, Brenneis C, Quan H. Frequency of symptom distress and poor
RI PT
29.
prognostic indicators in palliative cancer patients admitted to a tertiary palliative care unit, hospices, and acute care hospitals. J Palliat Care 2000;16:16-21.
Bruera E, Hui D. Palliative care units: the best option for the most distressed. [letter].
SC
30.
Arch Intern Med 2011;171:1601; author reply 1601-1602.
Tishelman C, Petersson LM, Degner LF, Sprangers MA. Symptom prevalence, intensity,
M AN U
31.
and distress in patients with inoperable lung cancer in relation to time of death. J Clin Oncol 2007;25:5381-5389. 32.
Hui D, Parsons HA, Damani S, et al. Quantity, design, and scope of the palliative
33.
TE D
oncology literature. Oncologist 2011;16:694-703.
Temel JS, Jackson VA, Billings JA, et al. Phase II study: Integrated palliative care in newly diagnosed advanced non-small-cell lung cancer patients. J Clin Oncol
Smith AK, Sudore RL, Perez-Stable EJ. Palliative care for Latino patients and their families: whenever we prayed, she wept. JAMA 2009;301:1047-1057, E1041.
AC C
34.
EP
2007;25:2377-2382.
16
ACCEPTED MANUSCRIPT
Table 1. Patient Characteristics Associated With ESAS ≥4 in the Last Seven Days of Life Pain
Fatigue
Odds Ratioa (95% CI) 0.99 (0.971.02)
Odds Ratioa (95% CI) 0.97 (0.951.00)
Depression DrowsinessAnorexia Poor well being Odds Odds Odds Odds Odds Odds Ratioa Ratioa Ratioa Ratioa Ratioa Ratioa (95% CI) (95% CI) (95% CI) (95% CI) (95% CI) (95% CI) 1.00 0.99 0.98 1.00 0.98 0.99 (0.97-1.03) (0.97(0.96-1.01) (0.98-1.03) (0.96(0.97-1.01) 1.02) 1.01)
0.77 (0.391.53) 1.00
1.08 (0.492.40) 1.00
0.81 0.34 (0.13-0.89) (0.371.75) 1.00 1.00
1.00 0.47 0.36 (0.49-2.07) (0.18-0.71) (0.230.97) 1.00 1.00 1.00
0.53 (0.181.57) 0.77 (0.311.94) 0.51 (0.161.69) 0.80 (0.312.04) 1.00
1.10 (0.274.58) 0.71 (0.271.84) 0.30 (0.081.04) 0.80 (0.272.35) 1.00
0.91 0.94 (0.14-5.81) (0.283.16) 3.67 0.61 (0.74(0.2218.13) 1.68) 2.09 0.41 (0.20(0.1121.30) 1.55) 1.94 1.01 (0.35(0.3610.82) 2.81) 1.00 1.00
1.43 0.47 0.53 (0.37-5.52) (0.13-1.68) (0.161.71) 1.44 1.13 0.96 (0.43-4.85) (0.40-3.18) (0.352.66) 3.38 0.90 0.69 (0.72(0.23-3.51) (0.182.59) 15.86) 1.74 0.60 0.83 (0.48-6.33) (0.21-1.66) (0.282.42) 1.00 1.00 1.00
0.90 (0.491.66) 1.00
0.94 (0.461.92) 1.00
0.64 1.05 (0.31-1.32) (0.561.98) 1.00 1.00
0.55 1.18 1.14 (0.29-1.03) (0.64-2.18) (0.612.15) 1.00 1.00 1.00
0.16 (0.060.41) 1.00
1.30 0.14 (0.51-3.31) (0.060.31) 1.00 1.00
0.57 0.66 0.35 (0.30-1.07) (0.17-0.74) (0.361.23) 1.00 1.00 1.00
GI
GU
Other
Respiratory Marital Status Married
Not married Race Hispanic race 0.80 (0.441.46) Non-Hispanic 1.00
AC C
EP
CI = confidence interval; GI = gastrointestinal; GU = genitourinary. AU: 1) WHAT DOES SUPERSCRIPT A REPRESENT? 2) WHAT DOES THE BOLDING SIGNIFY?
Insomnia Odds Ratioa (95% CI) 0.98 (0.961.00)
RI PT
Female Cancer Breast
Shortness of breath Odds Ratioa (95% CI) 0.98 (0.96-1.00)
0.91 (0.42-1.95)
0.86 (0.39-1.90)
1.00
1.00
2.02 (0.46-8.91)
SC
Sex Male
Anxiety
0.34 (0.09-1.27)
1.07 (0.35-3.30)
0.13 (0.04-0.40)
0.50 (0.13-1.96)
0.26 (0.07-0.98)
1.17 (0.37-3.71)
0.44 (0.14-1.39)
1.00
1.00
0.56 (0.27-1.18)
1.33 (0.69-2.56)
1.00
1.00
0.41 (0.20-0.85)
0.39 (0.20-0.77)
1.00
1.00
M AN U
Age
Nausea
TE D
Sign
1.33 (0.662.70) 1.00 1.90 (0.507.26) 2.05 (0.715.91) 1.92 (0.467.97) 1.69 (0.604.78) 1.00
1.24 (0.612.53) 1.00 0.17 (0.080.35) 1.00
ACCEPTED MANUSCRIPT
Figure Legends Fig. 1. The proportion of patients who were able to communicate decreased near death. The total
on each day varied because of the different lengths of admission.
RI PT
number of study patients on each day is plotted from death backwards. The number of patients
Fig. 2. Changes in Edmonton Symptom Assessment System (ESAS) in the last 7 days of life. (A)
SC
The average ESAS intensity was plotted from death backwards for patients who were able to communicate. The P-values for longitudinal trend using a generalized linear mixed model are:
M AN U
anorexia (P=0.001), drowsiness (P<0.0001), fatigue (P<0.0001), poor well-being (P=0.01), dyspnea (P<0.0001), depression (P=0.008), anxiety (P=0.09), insomnia (P=0.31), pain (P=0.15) and nausea (P=0.53); (B) The proportion of patients with ESAS ≥4/10 was plotted from death backwards. The P-values for longitudinal trend using generalized estimating equation are:
TE D
drowsiness (P=0.03), fatigue (P=0.006), poor well-being (P=0.19), dyspnea (P=0.006), pain (P=0.42), anorexia (P=0.16), anxiety (P=0.30), depression (P=0.78), insomnia (P=0.35) and
EP
nausea (P=0.07).
Fig. 3. Symptom profile over time. The proportion of patients with constipation, cough,
AC C
dysphagia to liquids, dysphagia to solids, fecal incontinence, myoclonus, sweating and urinary incontinence was plotted from death backwards. The frequencies of epistaxis, upper GI bleed, lower GI bleed, diarrhea, hematuria, vaginal bleed, and vaginal discharge were consistently below 10% and were not included in this diagram.
18
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
S0885-3924(14)00466-7
DOI:
10.1016/j.jpainsymman.2014.09.003
Reference:
JPS 8753
To appear in:
Journal of Pain and Symptom Management
Received Date: 21 June 2014 Revised Date:
28 August 2014
Accepted Date: 12 September 2014
Please cite this article as: Hui D, Santos Rd, Chisholm GB, Bruera E, Symptom Expression in the Last Seven Days of Life Among Cancer Patients Admitted to Acute Palliative Care Units, Journal of Pain and Symptom Management (2014), doi: 10.1016/j.jpainsymman.2014.09.003. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Original Article
14-00326R1
Symptom Expression in the Last Seven Days of Life Among Cancer Patients Admitted to
RI PT
Acute Palliative Care Units
David Hui, MD, MSc, Renata dos Santos, MD, Gary B. Chisholm, MS, and Eduardo Bruera,
SC
MD
M AN U
Department of Palliative Care and Rehabilitation Medicine (D.H., E.B.), Department of Biostatistics (G.B.C), and Division of Bioinformatics and Computational Biology (G.B.C.), M. D. Anderson Cancer Center, Houston, Texas, USA; and Department of Palliative Care (R.d.S.),
Address correspondence to: David Hui, MD, MSc
TE D
Barretos Cancer Hospital, Barretos, Brazil
EP
Department of Palliative Care and Rehabilitation Medicine, Unit 1414 University of Texas M. D. Anderson Cancer Center
AC C
1515 Holcombe Boulevard Houston, TX 77030, USA
E-mail: [email protected]
ACCEPTED MANUSCRIPT
Abstract Context. The symptom burden in the last week of life of patients with advanced cancer has not been well characterized.
RI PT
Objectives. To examine the frequency, intensity, and predictors for symptoms in the last seven days of life among patients who were able to communicate and died in an acute palliative care unit (APCU).
SC
Methods. We systematically documented the Edmonton Symptom Assessment System (ESAS) daily and 15 symptoms twice daily on consecutive advanced cancer patients admitted to
M AN U
APCUs at M.D. Anderson Cancer Center (USA) and Barretos Cancer Hospital (Brazil) from admission to death or discharge in 2010/2011. We examined the frequency and intensity of the symptoms from death backwards.
Results. A total of 203 of 357 patients died. The proportion of patients able to
TE D
communicate decreased from 80% to 39% over the last seven days of life. ESAS anorexia (P=0.001 in longitudinal analyses), drowsiness (P<0.0001), fatigue (P<0.0001), poor well-being (P=0.01) and dyspnea (P<0.0001) increased in intensity closer to death. In contrast, ESAS
EP
depression (P=0.008) decreased over time. Dysphagia to solids (P=0.01) and liquids (P=0.005), and urinary incontinence (P=0.0002) also were present in an increasing proportion of patients in
AC C
the last few days of life. In multivariate analyses, female sex, non-Hispanic race and lung cancer were significantly associated with higher ESAS symptom expression (odds ratio>1). Conclusion. Despite intensive management in APCUs, some cancer patients continue to experience high symptom burden as they approached death.
2
ACCEPTED MANUSCRIPT
Key Words: anorexia, deglutition disorders, death, depression, dyspnea, neoplasms, pain, palliative care, symptom assessment, urinary incontinence
RI PT
Running title: Symptom Expression in the Last Days of Life
AC C
EP
TE D
M AN U
SC
Accepted for publication: September 12, 2014.
3
ACCEPTED MANUSCRIPT
Introduction Acute palliative care units (APCUs) are specialized inpatient care settings that provide intensive symptom control and transition of care for patients with advanced diseases and severe
RI PT
distress (1, 2). APCUs are becoming increasingly available in cancer centers (3). Care is
delivered by an interprofessional team to address the physical, psychosocial, and spiritual needs of patients and their families. Several studies examining disease trajectory have shown that
SC
cancer patients develop increasing symptom burden and decreasing function over the last few months to weeks of life (4-7). However, there has been a paucity of studies examining the
M AN U
symptom burden in dying cancer patients (i.e., days of life expectancy) (8-11), and no studies in APCUs where patients receive intensive management focusing on quality of life. A better characterization of symptom intensity in the last few days of life could help clinicians to better understand patients’ needs, to improve their quality of life and to optimize the
TE D
quality of end-of-life care. We previously documented the clinical signs associated with impending death in a cohort of cancer patients (12-14). In the present study, we examined the frequency and intensity of various symptoms, and predictors of symptom expression in patients
Methods
EP
with advanced cancer admitted to APCUs.
AC C
Study Setting and Criteria
This study was part of the “Investigating the Process of Dying” study, a prospective observational study that systematically documents the signs and symptoms in the last days of life. The methodology of this study has been reported in detail elsewhere (14, 15). Briefly, we enrolled consecutive adult patients with advanced cancer who were admitted to the APCUs at M. D. Anderson Cancer Center (MDACC, Houston, TX, USA) between April 5 and July 6, 2010,
4
ACCEPTED MANUSCRIPT
and at Barretos Cancer Center (BCH, Barretos, Brazil) between January 27 and June 1, 2011. This study was approved by the Institutional Review Boards at both sites with waiver of informed consent for patient participation. All clinicians who participated in this study provided
RI PT
written informed consent prior to patient enrollment. Data Collection
We identified target symptoms for this study by conducting a literature review of
SC
published articles (8, 9, 16, 17) and educational materials (18), and discussing relevant variables with participating palliative care clinicians. The symptoms were collected from patients who
M AN U
were able to communicate verbally and interact with our staff. We documented the Edmonton Symptom Assessment System (ESAS) daily. The ESAS is a validated numeric rating scale that assesses the average intensity of 10 symptoms (pain, fatigue, nausea, depression, anxiety, drowsiness, shortness of breath, anorexia, poor well-being and insomnia) over the past 24 hours
TE D
from 0 (absent) to 10 (worst possible) (19-22). We collected ESAS data from patients directly as long as they were able to communicate and answer the study assessments, regardless of whether they were delirious or not because symptom expression was the main outcome of interest. The
EP
ESAS data also were collected from caregivers when patients had impaired communication capacity, but these data were not used because the current study focuses on patient-reported
AC C
outcomes only.
Data collected also included the presence or absence of 15 other symptoms every 12 hours, including epistaxis, cough, dysphagia to solids, dysphagia to liquids, upper gastrointestinal (GI) bleed, lower GI bleed, constipation, diarrhea, fecal incontinence, urinary incontinence, hematuria, vaginal bleed (female only), vaginal discharge (female only), myoclonus and sweating. Standardized data collection forms were completed by full-time
5
ACCEPTED MANUSCRIPT
palliative care nurses who were familiar with end-of-life care every 12-hour shift from admission to discharge or death independent of prior assessments. These symptoms were documented by bedside nurses instead of reported directly by patients because they can be observed objectively.
RI PT
As documented previously, extensive training and monitoring of palliative care nurses was provided before and during the data collection period (12). In Brazil, all data collection forms
conferences weekly to discuss issues related to data collection. Statistical Analysis
SC
were translated into Portuguese and then back-translated. The two study sites held video
M AN U
Our pre-planned sample size was a combined total of 200 deaths at the two study sites to examine the physical signs of impending death, as reported elsewhere (12). This analysis was planned based on the combined data a priori. This sample size also provided adequate power for our secondary multiple regression analysis, which requires 10 patients for each independent
TE D
variable included in the model.
We summarized the baseline demographics with descriptive statistics. We determined the frequency of each symptom from death backwards for all patients who died in the APCUs. We
EP
plotted the average ESAS intensity, the proportion of patients with ESAS scores ≥4, and the
of life.
AC C
proportion of patients with the other 15 symptoms mentioned previously over the last seven days
We calculated the statistical significance for longitudinal change using the generalized linear mixed model (GLMM) for average ESAS intensity, and the generalized estimating equation (GEE) for the proportion of patients with ESAS scores ≥4. We also employed a multivariate GEE model incorporating day prior to death, age, sex, race, cancer, and marital status to identify factors associated with ESAS intensity ≥4. We accounted for correlation within
6
ACCEPTED MANUSCRIPT
patient in the GLMM and GEE analyses using an autogressive correlation structure with a lag of one. Overall survival was calculated from the time of APCU admission. We collected
RI PT
survival data from electronic health records. The Statistical Analysis System (SAS version 9.2, SAS Institute Inc., Cary, NC) was used for statistical analysis. A P-value of <0.05 was considered significant.
SC
Results Patient Characteristics
M AN U
The patient characteristics have been reported elsewhere (12). Briefly, 52 of 151 and 151 of 206 patients died in the APCUs at MDACC and BCH, respectively. Among the 203 APCU deaths, the average age was 58 years (range 18-88), 99 (49%) were female, 158 (78%) were Hispanic, and 63 (33%) had gastrointestinal malignancies. The proportion of patients who were
TE D
able to communicate and to provide patient-reported outcomes decreased over time (Fig. 1). Frequency and Intensity of ESAS Symptoms
Fig. 2A shows that, among responding patients, the average intensity of ESAS anorexia
EP
(P=0.001 for longitudinal trend), drowsiness (P<0.0001), fatigue (P<0.0001), poor well-being (P=0.01) and dyspnea (P<0.0001) was moderate to severe seven days before death, and
AC C
increased as death approached. In contrast, depression (P=0.008) had lower intensity and decreased over time. Anxiety (P=0.09), insomnia (P=0.31), pain (P=0.15) and nausea (P=0.53) did not change significantly over the last seven days. The proportion of patients with ESAS scores ≥4 of 10 are shown in Fig. 2B, and the findings were similar. Moderate/severe drowsiness (P=0.03 for longitudinal trend), fatigue (P=0.006), poor well-being (P=0.19), and dyspnea (P=0.006) were present in an increasing
7
ACCEPTED MANUSCRIPT
majority of patients. The proportion of patients with moderate to severe pain (P=0.42), anorexia (P=0.16), anxiety (P=0.30), depression (P=0.78) insomnia (P=0.35) and nausea (P=0.07) remained similar over time.
RI PT
Frequency of Other Symptoms
A number of symptoms were present in 30-60% of communicating patients in the last few days of life, including dysphagia to solids (P=0.01 for longitudinal trend), dysphagia to
SC
liquids (P=0.005), urinary incontinence (P=0.0002) and constipation (P=0.66) (Fig. 3). Fecal incontinence (P=0.08), cough (P=0.77), sweating (P=0.06) and myoclonus (P=0.17) were
M AN U
experienced by 10-30% of patients. Other symptoms such as epistaxis (P=0.08), hematuria (P=0.07), upper GI bleed (P=0.008), lower GI bleed (P=0.60), diarrhea (P=0.39), vaginal bleed (P=0.80) and vaginal discharge (P=0.10) were detected in less than 10% of patients. Factors Associated With Moderate and High ESAS Intensity
TE D
On multivariate analysis (Table 1), female sex was associated with higher expression of nausea, drowsiness and anorexia; lung cancer was associated with a higher level of dyspnea; and non-Hispanic race was associated with higher expression of fatigue, anxiety, drowsiness, poor
Discussion
EP
well-being, dyspnea and insomnia.
AC C
We documented the trajectory of 25 symptoms in patients admitted to APCUs in the last seven days of life. Among patients who were able to communicate their concerns, anorexia, drowsiness, fatigue, poor well-being, dyspnea, dysphagia and incontinence increased over the last days of life. In contrast, depression decreased with time. Female sex, non-Hispanic race and diagnosis of lung cancer were associated with higher symptom expression.
8
ACCEPTED MANUSCRIPT
he proportion of patients who were able to communicate their symptoms decreased over time, although a small proportion retained their ability to do so even close to death. Our findings are generally consistent with another study by Morita et al. (23). However, the proportion of
RI PT
patients with impaired communication capacity was higher in that study, which could be
explained by differences in patient population and methods of assessment. Taken together, our findings suggest that it is possible to document patient-reported outcomes even in the last days of
SC
life. This is particularly important because caregivers’ and health care professionals’ proxy assessments are often less reliable (24).
M AN U
This is the first study to serially and systematically profile the symptoms in patients admitted to APCUs over the last seven days of life. Interestingly, the scant literature on symptoms in the last days of life focused primarily on the last 48 hours (9, 25, 26) and in other care settings (e.g., hospices, nursing homes) (8, 27, 28). We found that the frequency of many
TE D
symptoms was high and increased over time. The high symptom burden is likely because 1) patients admitted to APCUs were selected for their level of distress and refractory symptoms (29, 30) and 2) the end of life is associated with decreased function and various irreversible
EP
processes, such as post-obstructive pneumonia and bowel obstruction. In the APCUs, patients all underwent comprehensive symptom assessments and intensive
AC C
symptom management on a daily basis by highly trained palliative care specialists and interprofessional teams using well-defined pharmacological and non-pharmacological interventions. Notably, pain and nausea were relatively well controlled in the last days of life, likely related to the relatively effective treatment options available. Depression also improved significantly—this may be related to the intensive psychological interventions provided by the members of the APCU interprofessional team, such as chaplains, child life care specialists,
9
ACCEPTED MANUSCRIPT
psychologists, social workers, nurses and physicians, to support the patients and their families. Patients also may naturally feel less depressed at they enter the final stage of their lives, having come to the resolution that death is inevitable, consistent with another longitudinal study that
RI PT
examined the ESAS in cancer patients at home (4). However, we cannot rule out significant depression in patients who were unable to communicate in the last days of life, which also could contribute to our finding. Further studies are needed to examine the psychological aspects in the
SC
last days of life. Our findings suggest that a number of symptoms, such as dyspnea, fatigue and drowsiness, may be unavoidable in a proportion of patients even with intensive state-of-the-art
M AN U
supportive care. At the same time, it calls for more palliative care research in this population and setting to relieve suffering in the most distressed.
Our results are consistent with those of a study on ESAS symptom trajectories in the last few months of life among cancer patients in the home setting (4). However, limited data were
TE D
presented on the last week of life in that study, making our current findings unique and complementary. Although several studies have examined symptom prevalence in the last 48 hours of life, only one study has documented the symptom intensity using the ESAS visual
EP
analogue scale among nursing home residents (25). Among the 253 conscious patients, anorexia, drowsiness, and poor well-being were rated high in intensity, which is consistent with our study.
AC C
However, dyspnea was reported to be mild (median=10/100) whereas it was moderate to severe in our study. Dyspnea is one of the most distressing and under-studied symptoms in cancer patients (31, 32). Research is urgently needed to examine novel interventions to alleviate this devastating symptom in patients at the end of life. In contrast, pain, nausea and depression were generally well controlled in the last days of life, likely because of the effective therapeutic options available.
10
ACCEPTED MANUSCRIPT
In addition to the ESAS symptoms, our study revealed that dysphagia and urinary/fecal incontinence were more prevalent as patients approached death. This is not surprising given their deterioration in neurological and motor function. Constipation remained common at the end
RI PT
of life, despite the fact that laxatives were often administered in the APCUs. The high frequency of these distressing symptoms supports the need for routine symptom screening and monitoring, which would allow us to initiate appropriate palliative measures promptly. These may include a
SC
urinary catheter or adult diapers for urinary incontinence, and change of medication route for patients with dysphagia. In contrast, other symptoms such as myoclonus, sweating and cough
M AN U
were less frequent.
We found that patients with lung cancer generally had higher levels of dyspnea and anxiety compared with patients with other malignancies at the end of life. This is consistent with other studies demonstrating the high symptom burden in lung cancer patients, with dyspnea
TE D
being the most distressing concern (4, 31, 33). We also found that Hispanic race was associated with lower ESAS symptom scores, which may potentially be explained by differences in cultures and symptom expression (34). Consistent with another study, we found that females had higher
EP
nausea and drowsiness (4).
This study has several limitations. First, we only included two institutions with a
AC C
relatively small patient population. Larger studies are required to validate our findings. Second, we did not examine other symptoms such as xerostomia and bloating. Future research should consider assessing these symptoms. Third, the symptoms were collected by palliative care nurses instead of physicians because they have more time to observe patients. Our nurses were highly trained in symptom assessment, and all received standardized training to ensure high data quality. Fourth, some of the symptoms in this study were observed by nurses (e.g., cough,
11
ACCEPTED MANUSCRIPT
incontinence) instead of reported by patients. Finally, findings from this study are only applicable to cancer patients admitted to APCUs. Because symptom expression may differ by disease and setting (28), further studies are needed to characterize the symptom burden in other
RI PT
populations in the last days of life.
We documented the frequency, intensity, and predictors of various symptoms in cancer patients admitted to APCUs, and identified increasing symptom burden in imminently dying
SC
cancer patients. Lung cancer, female sex and Hispanic race were associated with higher
symptom expression. With the recent interest in patient-reported outcomes and quality of care
M AN U
measures, our study provides information on which symptoms can and cannot be palliated effectively with state-of-the-art supportive care. Further research is urgently needed to improve the quality of life for patients in the last days of life. Disclosure and Acknowledgments
TE D
This work was supported in part by National Institutes of Health grants RO1NR01016201A1, RO1CA122292-01, and RO1CA124481-01 (Dr. Bruera). This study is also supported by the M. D. Anderson Cancer Center Support Grant (CA 016672) and an institutional start-up grant
EP
#18075582 (Dr. Hui). All authors report no relevant financial conflicts of interest The authors thank all the patients, clinical nurses and physicians who participated in this
AC C
study and provided valuable data. References 1.
Hui D, Elsayem A, Palla S, et al. Discharge outcomes and survival of patients with advanced cancer admitted to an acute palliative care unit at a comprehensive cancer center. J Palliat Med 2010;13:49-57.
12
ACCEPTED MANUSCRIPT
2.
Hui D, Elsayem A, Li Z, et al. Antineoplastic therapy use in patients with advanced cancer admitted to an acute palliative care unit at a comprehensive cancer center: a simultaneous care model. Cancer 2010;116:2036-2043. Hui D, Elsayem A, De la Cruz M, et al. Availability and integration of palliative care at
RI PT
3.
US cancer centers. JAMA 2010;303:1054-1061. 4.
Seow H, Barbera L, Sutradhar R, et al. Trajectory of performance status and symptom
SC
scores for patients with cancer during the last six months of life. J Clin Oncol 2011;29:1151-1158.
Mercadante S, Casuccio A, Fulfaro F. The course of symptom frequency and intensity in
M AN U
5.
advanced cancer patients followed at home. J Pain Symptom Manage 2000;20:104-112. 6.
Ventafridda V, Ripamonti C, De Conno F, Tamburini M, Cassileth BR. Symptom prevalence and control during cancer patients' last days of life. J Palliat Care 1990;6:7-11. Morris JN, Suissa S, Sherwood S, Wright SM, Greer D. Last days: a study of the quality
TE D
7.
of life of terminally ill cancer patients. J Chronic Dis 1986;39:47-62. 8.
Morita T, Tsunoda J, Inoue S, Chihara S. Contributing factors to physical symptoms in
EP
terminally-ill cancer patients. J Pain Symptom Manage 1999;18:338-346. Lichter I, Hunt E. The last 48 hours of life. J Palliat Care 1990;6:7-15.
10.
Steindal SA, Ranhoff AH, Bredal IS, Sorbye LW, Lerdal A. Last three days of life in the
AC C
9.
hospital: a comparison of symptoms, signs and treatments in the young old and the oldest old patients using the Resident Assessment Instrument for Palliative Care. Int J Older People Nurs 2012;8:199-206.
13
ACCEPTED MANUSCRIPT
11.
Hui D, Nooruddin Z, Didwaniya N, et al. Concepts and definitions for "actively dying," "end of life," "terminally ill," "terminal care," and "transition of care": a systematic review. J Pain Symptom Manage 2014;47:77-89. Hui D, dos Santos R, Swati B, et al. Clinical signs of impending death in cancer patients.
RI PT
12.
Oncologist 2014;19:681-687. 13.
Bruera S, Chisholm G, Santos RD, et al. Variations in vital signs in the last days of life in
SC
patients with advanced cancer. J Pain Symptom Manage 2014 April 14. [Epub ahead of print].
Hui D, Dos Santos R, Chisholm G, et al. Bedside clinical signs associated with
M AN U
14.
impending death in patients with advanced cancer: preliminary findings. Cancer 2014, in press. 15.
Hui D, Dos Santos R, Chisholm G, et al. Clinical Signs of Impending Death in Cancer
TE D
Patients. Oncologist 2014;19:681-687. NOTE: DUPLICATE OF REF 12 16.
Ferris FD. Last hours of living. Clin Geriatr Med 2004;20:641-667, vi.
17.
Plonk WM, Jr., Arnold RM. Terminal care: the last weeks of life. J Palliat Med
Emanuel LL, Ferris FD, von Gunten CF: Module 6: Last hours of living. EPEC-O: Education in Palliative and End-of-life Care for Oncology. Chicago, IL: The EPEC
AC C
18.
EP
2005;8:1042-1054.
Project, 2005. Available from http://www.cancer.gov/cancertopics/cancerlibrary/epeco/selfstudy/module-6-pdf. Accessed June 12, 2012.
14
ACCEPTED MANUSCRIPT
19.
Bruera E, Kuehn N, Miller MJ, Selmser P, Macmillan K. The Edmonton Symptom Assessment System (ESAS): a simple method for the assessment of palliative care patients. J Palliat Care 1991;7:6-9. Chang VT, Hwang SS, Feuerman M. Validation of the Edmonton Symptom Assessment Scale. Cancer 2000;88:2164-2171.
21.
Richardson LA, Jones GW. A review of the reliability and validity of the Edmonton
SC
Symptom Assessment System. Curr Oncol 2009;16:55. 22.
RI PT
20.
Nekolaichuk C, Watanabe S, Beaumont C. The Edmonton Symptom Assessment System:
M AN U
a 15-year retrospective review of validation studies (1991--2006). Palliat Med 2008;22:111-122. 23.
Morita T, Tei Y, Inoue S. Impaired communication capacity and agitated delirium in the final week of terminally ill cancer patients: prevalence and identification of research
24.
TE D
focus. J Pain Symptom Manage 2003;26:827-834.
Hui D, Morgado M, Vidal M, et al. Dyspnea in hospitalized advanced cancer patients: subjective and physiologic correlates. J Palliat Med 2013;16:274-280. Brandt HE, Ooms ME, Deliens L, van der Wal G, Ribbe MW. The last two days of life of
EP
25.
nursing home patients--a nationwide study on causes of death and burdensome symptoms
26.
AC C
in The Netherlands. Palliat Med 2006;20:533-540. Hall P, Schroder C, Weaver L. The last 48 hours of life in long-term care: a focused chart audit. J Am Geriatr Soc 2002;50:501-506. 27.
Conill C, Verger E, Henriquez I, et al. Symptom prevalence in the last week of life. J Pain Symptom Manage 1997;14:328-331.
15
ACCEPTED MANUSCRIPT
28.
Veerbeek L, van Zuylen L, Swart SJ, van der Maas PJ, van der Heide A. The last 3 days of life in three different care settings in The Netherlands. Support Care Cancer 2007;15:1117-1123. Bruera E, Neumann C, Brenneis C, Quan H. Frequency of symptom distress and poor
RI PT
29.
prognostic indicators in palliative cancer patients admitted to a tertiary palliative care unit, hospices, and acute care hospitals. J Palliat Care 2000;16:16-21.
Bruera E, Hui D. Palliative care units: the best option for the most distressed. [letter].
SC
30.
Arch Intern Med 2011;171:1601; author reply 1601-1602.
Tishelman C, Petersson LM, Degner LF, Sprangers MA. Symptom prevalence, intensity,
M AN U
31.
and distress in patients with inoperable lung cancer in relation to time of death. J Clin Oncol 2007;25:5381-5389. 32.
Hui D, Parsons HA, Damani S, et al. Quantity, design, and scope of the palliative
33.
TE D
oncology literature. Oncologist 2011;16:694-703.
Temel JS, Jackson VA, Billings JA, et al. Phase II study: Integrated palliative care in newly diagnosed advanced non-small-cell lung cancer patients. J Clin Oncol
Smith AK, Sudore RL, Perez-Stable EJ. Palliative care for Latino patients and their families: whenever we prayed, she wept. JAMA 2009;301:1047-1057, E1041.
AC C
34.
EP
2007;25:2377-2382.
16
ACCEPTED MANUSCRIPT
Table 1. Patient Characteristics Associated With ESAS ≥4 in the Last Seven Days of Life Pain
Fatigue
Odds Ratioa (95% CI) 0.99 (0.971.02)
Odds Ratioa (95% CI) 0.97 (0.951.00)
Depression DrowsinessAnorexia Poor well being Odds Odds Odds Odds Odds Odds Ratioa Ratioa Ratioa Ratioa Ratioa Ratioa (95% CI) (95% CI) (95% CI) (95% CI) (95% CI) (95% CI) 1.00 0.99 0.98 1.00 0.98 0.99 (0.97-1.03) (0.97(0.96-1.01) (0.98-1.03) (0.96(0.97-1.01) 1.02) 1.01)
0.77 (0.391.53) 1.00
1.08 (0.492.40) 1.00
0.81 0.34 (0.13-0.89) (0.371.75) 1.00 1.00
1.00 0.47 0.36 (0.49-2.07) (0.18-0.71) (0.230.97) 1.00 1.00 1.00
0.53 (0.181.57) 0.77 (0.311.94) 0.51 (0.161.69) 0.80 (0.312.04) 1.00
1.10 (0.274.58) 0.71 (0.271.84) 0.30 (0.081.04) 0.80 (0.272.35) 1.00
0.91 0.94 (0.14-5.81) (0.283.16) 3.67 0.61 (0.74(0.2218.13) 1.68) 2.09 0.41 (0.20(0.1121.30) 1.55) 1.94 1.01 (0.35(0.3610.82) 2.81) 1.00 1.00
1.43 0.47 0.53 (0.37-5.52) (0.13-1.68) (0.161.71) 1.44 1.13 0.96 (0.43-4.85) (0.40-3.18) (0.352.66) 3.38 0.90 0.69 (0.72(0.23-3.51) (0.182.59) 15.86) 1.74 0.60 0.83 (0.48-6.33) (0.21-1.66) (0.282.42) 1.00 1.00 1.00
0.90 (0.491.66) 1.00
0.94 (0.461.92) 1.00
0.64 1.05 (0.31-1.32) (0.561.98) 1.00 1.00
0.55 1.18 1.14 (0.29-1.03) (0.64-2.18) (0.612.15) 1.00 1.00 1.00
0.16 (0.060.41) 1.00
1.30 0.14 (0.51-3.31) (0.060.31) 1.00 1.00
0.57 0.66 0.35 (0.30-1.07) (0.17-0.74) (0.361.23) 1.00 1.00 1.00
GI
GU
Other
Respiratory Marital Status Married
Not married Race Hispanic race 0.80 (0.441.46) Non-Hispanic 1.00
AC C
EP
CI = confidence interval; GI = gastrointestinal; GU = genitourinary. AU: 1) WHAT DOES SUPERSCRIPT A REPRESENT? 2) WHAT DOES THE BOLDING SIGNIFY?
Insomnia Odds Ratioa (95% CI) 0.98 (0.961.00)
RI PT
Female Cancer Breast
Shortness of breath Odds Ratioa (95% CI) 0.98 (0.96-1.00)
0.91 (0.42-1.95)
0.86 (0.39-1.90)
1.00
1.00
2.02 (0.46-8.91)
SC
Sex Male
Anxiety
0.34 (0.09-1.27)
1.07 (0.35-3.30)
0.13 (0.04-0.40)
0.50 (0.13-1.96)
0.26 (0.07-0.98)
1.17 (0.37-3.71)
0.44 (0.14-1.39)
1.00
1.00
0.56 (0.27-1.18)
1.33 (0.69-2.56)
1.00
1.00
0.41 (0.20-0.85)
0.39 (0.20-0.77)
1.00
1.00
M AN U
Age
Nausea
TE D
Sign
1.33 (0.662.70) 1.00 1.90 (0.507.26) 2.05 (0.715.91) 1.92 (0.467.97) 1.69 (0.604.78) 1.00
1.24 (0.612.53) 1.00 0.17 (0.080.35) 1.00
ACCEPTED MANUSCRIPT
Figure Legends Fig. 1. The proportion of patients who were able to communicate decreased near death. The total
on each day varied because of the different lengths of admission.
RI PT
number of study patients on each day is plotted from death backwards. The number of patients
Fig. 2. Changes in Edmonton Symptom Assessment System (ESAS) in the last 7 days of life. (A)
SC
The average ESAS intensity was plotted from death backwards for patients who were able to communicate. The P-values for longitudinal trend using a generalized linear mixed model are:
M AN U
anorexia (P=0.001), drowsiness (P<0.0001), fatigue (P<0.0001), poor well-being (P=0.01), dyspnea (P<0.0001), depression (P=0.008), anxiety (P=0.09), insomnia (P=0.31), pain (P=0.15) and nausea (P=0.53); (B) The proportion of patients with ESAS ≥4/10 was plotted from death backwards. The P-values for longitudinal trend using generalized estimating equation are:
TE D
drowsiness (P=0.03), fatigue (P=0.006), poor well-being (P=0.19), dyspnea (P=0.006), pain (P=0.42), anorexia (P=0.16), anxiety (P=0.30), depression (P=0.78), insomnia (P=0.35) and
EP
nausea (P=0.07).
Fig. 3. Symptom profile over time. The proportion of patients with constipation, cough,
AC C
dysphagia to liquids, dysphagia to solids, fecal incontinence, myoclonus, sweating and urinary incontinence was plotted from death backwards. The frequencies of epistaxis, upper GI bleed, lower GI bleed, diarrhea, hematuria, vaginal bleed, and vaginal discharge were consistently below 10% and were not included in this diagram.
18
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT