Impact of Palliative Care Unit Admission on Symptom Control Evaluated by the Edmonton Symptom Assessment System

Vol. 30 No. 4 October 2005 Journal of Pain and Symptom Management 367 Original Article Impact of Palliative Care Unit Admission on Symptom Control Evaluated by the Edmonton Symptom Assessment System Caterina Modonesi, MD, Emanuela Scarpi, BSc, Marco Maltoni, MD, Stefania Derni, MD, Laura Fabbri, MD, Francesca Martini, MD, Elisabetta Sansoni, MD, and Dino Amadori, MD Palliative Care Unit (C.M., M.M., S.D., L.F., F.M., E.S.), Department of Medical Oncology, Forlimpopoli Hospital; Romagnolo Oncology Institute (E.S.); and Department of Medical Oncology (D.A.), Morgagni-Pierantoni Hospital, Forlì, Italy Abstract The aim of the present study was to evaluate the impact of palliative care on patients symptoms, using the Edmonton Symptom Assessment System (ESAS) to measure symptom intensity at the time of admission and variations registered during the first 7 days hospitalization. Three hundred fourteen patients were admitted to the unit during its first year of activity. Of these, 162 patients (51.6%) completed, 62 (19.7%) partially completed, and 90 (28.7%) did not complete the ESAS. The mean (6SD) value of the Symptom Distress Score (SDS) (sum of the values of the different symptoms) for the 162 evaluable patients on Day 1 was 33.93 (616.24). On Day 7 the mean was 28.14 (615.11) (ANOVA for repeated measurements, P! 0.0001). ESAS values for patients with moderate-severe symptom intensity (average values Day 1--Day 7 and P value, ANOVA for repeated measurements) were as follows: pain (7.12--4.23, P! 0.0001), fatigue (7.46--5.68, P! 0.0001), nausea (7.12--1.96, P! 0.0001), depression (7.26--5.28, P! 0.0001), anxiety (7.13--5.14, P! 0.0001), drowsiness (7.42--6.40, P 5 0.002), anorexia (7.33--4.33, P! 0.0001), wellbeing (6.83--3.85, P! 0.0001), and dyspnea (7.08--3.86, P! 0.0001). These data seem to indicate that the patients who benefit most from inpatient palliative care are those with the most complex symptomatology. J Pain Symptom Manage 2005;30:367--373. Ó 2005 U.S. Cancer Pain Relief Committee. Published by Elsevier Inc. All rights reserved. Key Words Edmonton Symptom Assessment System, symptom control, palliative care Introduction Good symptom control is one of the most important elements for delivering effective Address reprint requests to: Marco Maltoni, MD, Palliative Care Unit, Forlimpopoli Hospital, Via Duca d Aosta 33, 47034 Forlimpopoli (FC), Italy. Accepted for publication: April 13, 2005. Ó 2005 U.S. Cancer Pain Relief Committee Published by Elsevier Inc. All rights reserved. palliative care. To ensure quality care, it is vital that palliative care programs document the effectiveness of their interventions aimed at reducing physical and psychological distress. This can be achieved by measuring the intensity of the patient s distress over time. Several evaluation instruments have been created. Some multidimensional questionnaires are extremely complex and ask too 0885-3924/05/$--see front matter doi:10.1016/j.jpainsymman.2005.04.007

368 Modonesi et al. Vol. 30 No. 4 October 2005 much of seriously ill patients in terms of time and concentration, whereas other approaches, such as the Visual Analogue Scale (VAS) or Numerical Rating Scale (NRS), are simple and effective. 1 Patient compliance in answering multi-item instruments is excellent up to one month before death, but the use of more simple methods requiring minimum effort and concentration is advisable in the last few weeks of life. 2,3 The Edmonton Symptom Assessment System (ESAS) is a simple, validated, and reliable multi-item instrument developed to measure different symptoms in palliative care patients. 4--7 It has been demonstrated that ESAS is a useful audit tool for assessing patterns of palliative symptom control and that it allows for institutional comparisons. 8,9 Other studies have also shown the simplicity, validity, and applicability of ESAS as a useful instrument in the last weeks of life. 8--11 ESAS has several advantages in that patients directly report the intensity of their symptoms, allowing for an overall evaluation of individual patient symptomatology. It also facilitates the graphic representation of symptom patterns, which is useful for medical and nursing staff. 4,12,13 The aim of the present study was to evaluate the impact of palliative care on patients symptoms, using ESAS to measure symptom intensity at the time of admission and variations registered during the first 7 days hospitalization. following nine items: pain, fatigue, nausea, depression, anxiety, drowsiness, anorexia, wellbeing, and dyspnea. 1,14,15 A higher score for a symptom indicated a greater intensity. All patients consecutively admitted to the PCU during its first year of activity were enrolled onto the study (314 individuals). Only those patients who completed the ESAS during the first 7 days of admission were evaluated for the present study. The study was examined and approved by the Ethics Committee of the Local Health and Social Services (Azienda USL, Forlì). Patients gave their informed consent to be evaluated for the frequency and intensity of their symptoms. The total Symptom Distress Score (SDS) (range 0--90) was calculated by summing the separate NRS scores for each single day (ESAS global score). Mean values and relative standard error were calculated for each symptom and for total score. A repeated measures analysis of variance sought to determine if there were significant differences in the way each of the mean ESAS scores changed over time. Survival times were measured from the date of admission and death, and were analyzed using the Kaplan- Meier product-limit method. All analyses were performed using SAS Statistical software, 16 and P values of! 0.05 were considered to be statistically significant. Methods The present prospective, open-label study was carried out in the Palliative Care Unit of the Oncology Department in Forlimpopoli Hospital from October 2002 to October 2003. A recent version of ESAS was used in which clinical symptoms assessed with a Numerical Rating Scale (NRS) replaced the previous Visual Analogue Scale (VAS) evaluation. 12,13 The ESAS questionnaire was completed by patients, either by themselves or with the help of nursing staff, daily for the first 7 days of hospitalization and once a week thereafter. All the evaluations were made by the patients, who wrote the appropriate numbers in the form themselves or dictated them to the nurse. The ESAS questionnaire, visualized graphically as an NRS (0--10), was used to evaluate the Results Three hundred fourteen patients were admitted to the PCU during its first year of activity. The ESAS questionnaire was completed by 162 (51.6%) patients for the first 7 days hospitalization, partially completed by 62 (19.7%) patients and not completed by 90 (28.7%) patients. The reasons for incomplete or missing data were early death or terminal phase in 28 (8.9%) patients, alterations in cognitive function in 104 (33.1%) patients, refusal in 8 (2.5%) patients, and unknown in 12 (3.8%) patients (Table 1). Only the patients who completed the ESAS during the first 7 days of admission were evaluated for the study. Seventy-one (43.8%) were women and 91 (56.2%) were men, with a median age of 67 years (range 26--94 years). The most frequent

Vol. 30 No. 4 October 2005 Symptom Control in Palliative Care 369 Table 1 Patient Compliance in Completing ESAS Questionnaire Causes of Non-Compliance n (%) Early Death or Terminal Phase Cognitive Alterations Patient Refusal Unknown Completed a 162 (51.6) Incomplete b 62 (19.7) 4 (1.3) 48 (15.3) 3 (0.9) 7 (2.2) Not completed c 90 (28.7) 24 (7.6) 56 (17.8) 5 (1.6) 5 (1.6) Total 314 (100) 28 (8.9) 104 (33.1) 8 (2.5) 12 (3.8) a Patients for whom ESAS information was available for the entire first week of hospitalization. b Patients for whom ESAS information was available only for some days of the first week of hospitalization. c Patients whose physical condition prevented them from completing the questionnaire. primary tumors were gastrointestinal tract tumors (30.2%), and lung (18.5%) and breast cancer (11.1%) (Table 2). The SDS calculated for the 7-day observation are reported in Table 3. Mean values decrease constantly during hospitalization (ANOVA for repeated measurements P! 0.0001). Comparing values on Days 1 (mean 33.93 6 16.24 SD) and 7 (28.14 6 15.11 SD), a 17.1% reduction in the average scores was observed (P! 0.0001). The mean symptom values (6SD) at the time of admission are reported in Table 4. A statistically significant reduction involved all but depression and drowsiness symptoms. Patients were divided into two groups on the basis of symptom intensity at admission: mild (! 5) and moderate/severe (5--10). There was no great change observed for the group with mild symptom intensity over the 7-day evaluation, with the exception of drowsiness (mean values Day 1: 2.50 6 1.05, Day 7: 2.98 6 2.46, P 5 0.006), which increased considerably. A Table 2 Patient Characteristics n % Patients evaluated 162 Sex Male 91 56.2 Female 71 43.8 Median age in years (range) 67 (26--94) Primary cancer site Gastrointestinal 49 30.2 Lung 30 18.5 Breast 18 11.1 Gynecological 11 6.8 Prostatic 10 6.2 Hematological 10 6.2 Urogenital 9 5.6 Head and neck 8 4.9 Unknown 4 2.5 Miscellaneous 13 8.0 reduction in symptom intensity was observed, in the latter group, during the first few days of admission. Variations in moderate/severe symptom intensities during the first 7 days of admission are shown in Table 5. A statistically significant decrease in intensity between Day 1 and Day 7 was observed for all symptoms. A secondary aim of the study was to evaluate the prevalence of symptoms at admission. Upon entering the hospice, pain was present (O 0) in 67% of patients (severe intensity in 40%: score 7--10), fatigue in 88% (severe intensity in 55%), nausea in 29% (severe intensity in 26%), depression in 71% (severe intensity in 40%), anxiety in 74% (severe intensity in 41%), drowsiness in 79% (severe intensity in 44%), anorexia in 70% (severe intensity in 53%), poor sensation of well-being in 62% (severe intensity in 35%), and dyspnea in 47% (severe intensity in 34%). Of the 162 patients for whom data on symptom intensity were available for the first 7 days, 143 had died by November 30, 2003. Median survival for this group was 46 days (95% CI: 36--53 days). We also calculated the Table 3 Symptom Distress Score (SDS) Over Time for 162 Patients Days of Hospitalization Mean SDS [0--90] (6SD) 1 33.93 (16.24) 2 31.42 (15.45) 3 30.95 (15.54) 4 30.57 (16.87) 5 29.94 (16.39) 6 29.72 (15.37) 7 28.14 (15.11) SD 5 standard deviation. ANOVA P value! 0.0001.

370 Modonesi et al. Vol. 30 No. 4 October 2005 Table 4 Mean Values of Symptom Intensity Over Time (162 patients) Symptom Day 1 (6 SD) Day 7 (6 SD) P a Pain 3.85 (3.42) 2.73 (2.66)!0.0001 Fatigue 5.46 (3.15) 4.67 (3.07) 0.003 Nausea 1.57 (2.68) 0.91 (1.81) 0.001 Depression 4.17 (3.46) 3.73 (3.26) 0.082 Anxiety 4.33 (3.36) 3.73 (3.18) 0.015 Drowsiness 4.38 (3.40) 4.77 (3.11) 0.130 Anorexia 4.52 (3.63) 3.01 (3.27)!0.0001 Well-being 3.39 (3.27) 2.82 (2.84) 0.023 Dyspnea 2.17 (3.03) 1.61 (2.49) 0.006 SD 5 standard deviation. a ANOVA P value. 30-day survival probability as a function of the symptoms present on the day of admission. Patients who had a #30-day survival (Figure 1) showed a higher intensity of all symptoms (significant for fatigue, drowsiness, dyspnea, and anorexia), apart from pain and anxiety, whereas the intensity of depression with respect to survival was similar to that of the group with a O30-day survival. The SDS at baseline was highest for patients with the shortest survival compared to those with a survival time of more than one month (37.5 6 16.5 and 33.1616.4 respectively, P 5 0.05). There is a trend for higher SDS value in relation to shorter survival, but not a statistically significant correlation. This is plausible because SDS is the sum of the means of different symptoms, some of which are not correlated with survival. Transforming our scale of decimal values into equivalent values of hundreds permitted us to compare our SDS values with those of the Edmonton Palliative Care Unit reported by Bruera (Table 6). 4 An absolute reduction of 40 Table 5 Mean Values of Symptom Intensity Over Time in Patients with ESAS 5--10 at Baseline Symptom No. of Patients Day 1 (6SD) Day 7 (6SD) Pain 74 7.12 (1.77) 4.23 (2.60)!0.0001 Fatigue 104 7.46 (1.70) 5.68 (2.76)!0.0001 Nausea 24 7.12 (1.83) 1.96 (2.65)!0.0001 Depression 78 7.26 (1.88) 5.28 (2.99)!0.0001 Anxiety 83 7.13 (1.83) 5.14 (2.92)!0.0001 Drowsiness 80 7.42 (1.66) 6.40 (2.66) 0.002 Anorexia 90 7.33 (1.86) 4.33 (3.23)!0.0001 Well-being 66 6.83 (1.60) 3.85 (3.08)!0.0001 Dyspnea 37 7.08 (1.75) 3.86 (3.14)!0.0001 SD 5 standard deviation. a ANOVA P value. P a in the mean score was observed in our case series in the first five days of admission, while Bruera reported a score reduction of 48, showing comparable symptom management. Discussion In the present study, we verified the impact that PCU admission has on the symptomatology of far-advanced cancer patients. The Edmonton Symptom Assessment System (ESAS) is a validated and reliable instrument created to measure the intensity of 9 different symptoms in palliative care patients. 4--7,12,13 It permits hospice staff to make a rapid assessment and interpretation of the intensity of these symptoms. Over time, the control of some symptoms, such as nausea and pain, has become systematic, thanks to the diffusion of international guidelines for their correct management. 17--19 Other symptoms, such as anorexia and wellbeing, have been studied less systematically. It is difficult to estimate the success that palliative care might have in improving these symptoms, as the currently used treatments appear to offer only limited relief. New therapeutic strategies will perhaps be possible, once pathogenetic mechanisms (impact of release of growth factors and cytokines) are better understood. 20,21 In the present study, the reduction in mean values on Day 7 with respect to Day 1 was statistically significant for all symptoms, apart from drowsiness and depression. Overall, the symptom that fairs worst, in agreement with results reported by Rees et al., 20 is drowsiness, with a higher, albeit not significant, mean value on Day 7 than on admission day. This may be due to pharmacological therapies (antiemetics, analgesics), but is also connected with the fact of being confined to bed and to the far-advanced stage of disease of most of the patients admitted to the PCU. The management strategy for the side effects of opioids and progressive tolerance towards these drugs may have led to a reduction in drowsiness after the seventh day of hospitalization, which, however, was not evaluated in the present study. The limited symptomatic improvement in the depression of patients was probably linked to the difficulty of coming to terms with their

Vol. 30 No. 4 October 2005 Symptom Control in Palliative Care 371 10 9 8 mean NRS score 7 6 5 4 3 2 1 0 pain fatigue nausea depression anxiety drowsiness anorexia poor sensation of well-being dyspnea OS <=30 days OS >30 days P<0.05 Fig. 1. ESAS value at admission and relation to survival. illness and of containing emotional and existential distress about the terminal stage of the disease. As the physical symptoms improved, it became clear that there was not such a good improvement in psychological status. It is likely that, once the most distressing symptom is under control, the patient focuses on the least distressing symptom. As reported by other authors, there was a close correlation between the variation in anxiety and depression values over time. 22 The best results were generally achieved for pain and anorexia. As reported by other authors, 22,23 pain is the symptom for which the most effective treatments are available. However, it also gratifying to see that satisfactory results were obtained for anorexia, which is often difficult to control. Considering the different symptoms and using 5 (! 5 vs. $5) as the cut-off for serious Table 6 Mean Symptom Distress Scores (SDS) (Total ESAS Scores) by Hospital Day Day Forlì Edmonton 1 339 410 2 314 377 3 309 367 4 306 374 5 299 362 symptom, a temporal effect can be observed, with a decrease in mean scores in both subgroups, which was, however, more evident in the $5 group as time passed. In the group with mild symptom intensity, an important, albeit not clinically significant, increase in drowsiness was observed, which was probably correlated with pharmacological treatments administered during hospitalization. The other symptoms with a! 5 intensity did not show a great variation in the 7 days of evaluation as, considered mild, they were not treated pharmacologically. With regard the psychophysical symptoms considered in the present study, we observed that the more severe the symptom, the greater is the influence of symptom control and the greater the reduction of distress. This further confirms that an inpatient palliative care setting has a significant impact on the clinical symptoms of far-advanced cancer patients. 24 In their initial publication describing the use of ESAS, Bruera et al. presented the mean total SDS for the first 5 days of admission. 4 In the present study we compared our SDS values with those of the Edmonton Palliative Care Unit. The higher SDS on Day 1 reported by Bruera et al. (4) is, in part, a reflection of differences in entry criteria between the two

372 Modonesi et al. Vol. 30 No. 4 October 2005 inpatient units, as only highly symptomatic patients were admitted to the Edmonton Care Unit, whereas our PCU also accepts patients suffering from the side effects of radiochemotherapy or patients whose families are no longer able to care for them at home. The absolute reduction in the mean SDS value of the Edmonton Unit (48) compared to ours (40) would seem to indicate comparable symptom management. In accordance with the concept of simultaneous care, 25 we maintain that palliative care should be initiated while patients are still receiving acute care and that it should become part of the active cancer care model in hospitals. 24 In this way, patients would benefit from the experience of palliative care not only when they are considered as off therapy. It has been hypothesized that Specialist Palliative Care Units and Hospices could represent two different levels of palliative intensity assistance. ESAS could permit the comparison of different case-mix of patient populations undergoing palliative care. It could also be a useful tool to identify the most suitable care setting for individuals and to evaluate the impact of our intervention on the psychophysical conditions of patients. Acknowledgments The authors would like to thank the staff of Palliative Care Unit for actively participating in data collection and Gráinne Tierney for editing the manuscript. References 1. Costello P, Wiseman J, Douglas I, et al. Assessing hospice inpatients with pain using numerical rating scales. Palliat Med 2001;15:257--258. 2. Jordhoy MS, Kaasa S, Fayers P, et al. Challenges in palliative care research; recruitment, attrition and compliance: experience from a randomized controlled trial. Palliat Med 1999;13:299--310. 3. Stromgren AS, Goldschmidt D, Groenvold M, et al. Self-assessment in cancer patients referred to palliative care: a study of feasibility and symptom epidemiology. Cancer 2002;94:512--520. 4. Bruera E, Kuehn N, Miller MJ, et al. The Edmonton Symptom Assessment System (ESAS): a simple method for the assessment of palliative care patients. J Palliat Care 1991;7:6--9. 5. Chang VT, Hwang SS, Feuerman M. Validation of the Edmonton Symptom Assessment Scale. Cancer 2000;88:2164--2171. 6. Nekolaichuk CL, Bruera E, Spachynski K, et al. A comparison of patient and proxy symptom assessments in advanced cancer patients. Palliat Med 1999;13:311--323. 7. Nekolaichuk CL, Maguire TO, Suarez-- Almazor M, et al. Assessing the reliability of patient, nurse, and family caregiver symptom rating in hospitalized advanced cancer patients. J Clin Oncol 1999;17:3621--3630. 8. Dudgeon DJ, Harlos M, Clinch JJ. The Edmonton Symptom Assessment Scale (ESAS) as an audit tool. J Palliat Care 1999;15:14--19. 9. Bruera E, Watanabe S. New developments in the assessment of pain in cancer patients. Support Care Cancer 1994;2:312--318. 10. Brescia FJ, Portenoy RK, Ryan M, et al. Pain, opioid use, and survival in hospitalized patients with advanced cancer. J Clin Oncol 1992;10:149--155. 11. Cleeland CS, Mendoza TR, Wang XS, et al. Assessing symptom distress in cancer patients: the M.D. Anderson Symptom Inventory. Cancer 2000; 89:1634--1646. 12. http://www.albertapalliative.net/apn/cgc/a4 SymptonAssessmentScale.html Access verified on March 18, 2005. 13. http://www.palliative.org/pc/clinicalinfo/ AssessmentTools/esas.pdf Access verified on March 18, 2005. 14. Jensen MP, Karoly P, Braver S. The measurement of clinical pain intensity: a comparison of six methods. Pain 1986;27:117--126. 15. De Conno F, Caraceni A, Gamba A, et al. Pain measurement in cancer patients: a comparison of six methods. Pain 1994;57:161--166. 16. SAS Institute Inc. SAS/STAT User s Guide, version 8.02, vol. 1. Cary, NC: SAS Institute, 1989. 17. Breitbart W, Rosenfeld B, Pessin H, et al. Depression, hopelessness and desire for hastened death in terminally ill patients with cancer. JAMA 2000;284:2907--2911. 18. Jenkins CA, Schulz M, Hanson J, Bruera E. Demographic, symptom, and medication profiles of cancer patients seen by a palliative care consult team in a tertiary referral hospital. J Pain Symptom Manage 2000;19:174--184. 19. Mercadante S, Fulfaro F, Casuccio A. The impact of home palliative care on symptoms in advanced cancer patients. Support Care Cancer 2000;8:307--310. 20. Rees E, Hardy J, Ling J, et al. The use of the Edmonton Symptom Assessment Scale (ESAS) within a palliative care unit in the UK. Palliat Med 1998;12:75--82.

Vol. 30 No. 4 October 2005 Symptom Control in Palliative Care 373 21. Strang P. The effect of megestrol acetate on anorexia, weight loss and cachexia in cancer and AIDS patients. [review]. Anticancer Res 1997; 17(1B):657--662. 22. Heedman PA, Strang P. Symptom assessment in advanced palliative home care for cancer patients using the ESAS: clinical aspects. Anticancer Res 2001;21(6A):4077--4082. 23. Potter J, Hami F, Bryan T, Quigley C. Symptoms in 400 patients referred to palliative care services: prevalence and patterns. Palliat Med 2003;17: 310--314. 24. Elsayem A, Swint K, Fisch MJ, et al. Palliative care inpatient service in a comprehensive cancer center: clinical and financial outcomes. J Clin Oncol 2004;22:2008--2014. 25. Meyers FJ, Linder J. Simultaneous care: disease treatment and palliative care throughout illness. J Clin Oncol 2003;21:1412--1415.