Personalized Symptom Goals and Patient Global Impression on Clinical Changes in Advanced Cancer Patients (2024)

Participants

A consecutive sample of advanced cancer patients who were admitted to six palliative care units from Italy, Brazil, and Greece was assessed for a period of 12 months (from January 2016 to December 2016). Documentation was translated and reviewed by coordinator center.

All institutions were tertiary care hospitals with access to comprehensive cancer treatments and supportive care. All patients underwent symptom evaluation by a specialist palliative care team. Investigators were coordinated by the principal investigator center to help the data collection process, to provide training to the local research staff, and to organize an electronic database for entering data for analysis.

Inclusion criteria were age ≥18 years with a diagnosis of advanced cancer. Advanced cancer was considered locally advanced, recurrent, or metastatic disease for solid cancers and relapsed or refractory disease for hematologic tumors [1], [12]. Exclusion criteria were an expected survival of ≤2 weeks and a value of ≥13 in the Memorial Delirium Assessment Scale (MDAS). MDAS is used to assess the cognitive status of patients and is a validated tool to quantify the intensity of delirium [13].

Data Collection

Baseline patient characteristics, including age, sex, familiar condition, education level, cancer diagnosis, and Karnofsky performance status, were recorded.

The intensities of 10 common symptoms included in the Edmonton Symptom Assessment Score (ESAS; i.e., pain, fatigue, nausea, depression, anxiety, drowsiness, shortness of breath, appetite, feelings of well‐being, and sleep) were assessed both at admission (T0) and after a week of palliative care (T7). ESAS is a self‐reported tool assessing the intensity of most common psychological and physical symptoms and is a valid and reliable tool for assessing the overall symptom burden, sensible to changes produced by a treatment. ESAS uses a 0 (no symptom) to 10 (worst intensity) point numeric rating scale to examine the intensity of each symptom over the past 24 hours [2], [3]. A screening tool for history of alcohol dependence (CAGE: cut down, annoy, guilt, eye‐opener) was also administered. A positive CAGE score of 2 has been shown to be of prognostic value in the opioid management of cancer pain [14].

At admission, for each of the 10 symptoms, patients were asked about their PSG. The question was as follows: “At what level would you feel comfortable with this symptom?” The same 0–10 numeric rating scale used for ESAS was used [12]. Palliative treatment was started according to patients’ needs and local policy. One week after (T7), ESAS and PSG were measured to detect changes after a palliative care intervention. Patients were considered to have achieved a PSG response if their follow‐up symptom intensity (measured T7) was equal or less than their PSG. PGI of changes after symptom management was measured after a week (T7) according to the following scale: 3 = much better, 2 = better, 1 = a bit better, 0 = the same, −1 = a little worse, −2 = worse, −3 = much worse. The PGI has been used as an anchor for a clinically significant change of a symptom [12]. MCIDs were calculated by PGI of improvement or deterioration at T7 (bit better or a little worse, respectively). The data collection was performed by a palliative care specialist.

Statistical Analysis

The sample size of 800 patients was based on previous studies performed with similar designs and aims, which included 777 patients [10], [11], [12]. The analysis performed an explorative data analysis of the symptoms' scales, ESAS, PSG, and PGI using their distributions, means and standard deviations, medians, and interquartile range. We checked symptoms' changes carrying out paired t tests estimating mean symptoms' changes and their corresponding standard deviation (SD) with 95% confidence intervals, with I type error set at 5%. Likewise, we tested PSG variations, estimating them according to each symptom. PGI was categorized into three classes: deterioration (PGI ≤ −1), no change (PGI = 0), improvement (PGI ≥ 1). We assessed the minimal symptoms' variation impacting patients' well‐being. Furthermore, we calculated the minimal changes associated to PGI grades ranging from −3 up to +3. An analysis of variance (ANOVA) test has been applied to compare PSG and PGI response proportions graphing the pattern detected. ANOVA analysis has been carried out stratifying symptoms' intensity (Mild = 1–3; Moderate = 4–6; Severe = 7–10). The factors associated with PGI scores have been analysed setting a multivariate polytomous logistic regression, setting the reference category at PGI = 0 (no PGI changes). Model fitting has been assessed using a likelihood ratio test. The statistical analysis has been carried out using the statistical software STATA (Release 14; StataCorp, College Station, TX).

Table 1.

Table 2.

PGI

Data regarding PGI are presented in Table ​Table3.3. Patients perceived an MCID (a bit better) with an improvement of symptom intensity ranging from 1.71 to 2.16 points, with some differences among symptoms. A better improvement ranged from 2.64 to 3.51 points, and a much better improvement ranged from 2.64 to 4.60 points. In 278 patients (31.7%), the clinical condition was the same. A minority of patients deteriorated. A little worse ranged from −0.34 to −2.5 points, with some differences among symptoms, with nausea having the highest value (−2.5). Worse ranged from −1.61 to −3.8 points, and much worse ranged from −3 to −7.17 points.

PSG

The distribution of PSG for the different ESAS items recorded at T0 and changes at T7 are shown in Tables ​Tables44 and ​and5.5. PSG ranged between 0.40 (for dyspnea) and 1.84 (asthenia). Most patients had a PSG of ≤3 for all ESAS items as a target at T0. In a rank order: pain 90.1%; asthenia 77.5%; nausea 96.5%; depression 92.6%; anxiety 92.2%; drowsiness 89.8%; dyspnea 97.8%; insomnia 94.1%; appetite 85.7%; and well‐being 86.3%. About 5% of patients targeted their symptoms at ≥4, being >10% for asthenia, appetite, and well‐being.

At T7, a statistically significant increase in the proportion of patients who achieved their PSG was observed (pain, n = 331 [37.8%]; nausea, n = 698 [79.7%]; asthenia, n = 213 [24.3%]; anxiety = 342 [39%]; depression, n = 395 [45.1%]; appetite, n = 374 [42.7%]; dyspnea, n = 632 [72.1%]; drowsiness, n = 463 [52.8%]; insomnia, n = 487 [55.6%]; well‐being, n = 250 [28.5%]; p = .00001). All PSG targets significantly changed after 1 week of palliative care, according to a correspondent decrease in all ESAS items (Pearson's correlation p = .00001).

PGI

As expected, symptom intensity significantly improved after 1 week of palliative care. One week has been invariably reported to be an acceptable period to stabilize patients admitted to a palliative care unit, where symptom management can be more intensive and effective [1]. PGI improved, particularly for pain, possibly as a consequence of a decrease in intensity of all ESAS items. Only a minority of patients felt that their pain had worsened, and in about 30% of patients, the clinical condition did not change. For many symptoms, including nausea, anxiety, depression, dyspnea, appetite, drowsiness, and insomnia, no significant changes were found in the majority of patients. In a previous study, based on PGI scale, globally half of patients had a clinically significant improvement, 32% no significant changes, and 21% a deterioration [12].

The finding of the MCID, evaluated by PGI, suggested that patients require at least two points of a numerical scale from 0 to 10 to be able to distinguish a minimal improvement, more than two points to feel better, and more than three points to stay much better, on average. This in contrast with previous data in which values ranging from 0.1 to 1.2 and 0.1 to 1.8 of MCID for improvement and deterioration, respectively, were found among the different ESAS items [7]. This can be explained by the retrospective design of the study, the use of categories anchored to well‐being rather than raw numbers, the outpatient setting of radiotherapy, and the longer intervals among observations (4–12 weeks). Indeed, different outcomes can be achieved when moving from one category to another of well‐being, that is a multidimensional construct, unable to detect the weight of each symptom [15]. In a previous study, an MCID between 1 and 2 was found for the majority of symptoms. In the sensitivity‐specificity analysis, however, that was less than 80%, the universal cutoff for both improvement and deterioration was ≥1 [12]. As opposed to the present study, in which patients were admitted to a palliative care unit and followed up for a week, this study was performed in an outpatient setting with a 3‐week follow‐up interval. One could wonder if an optimal management requires different magnitude of effect, for example, that patients feel better rather than only a bit better. In this case, a significant PGI would require 2–3 points of differences with baseline data. Of interest, for acute pain, ≥2 points or 33% decrease are considered to be the cutoff for MCID [9], [16], resembling data found in this study.

The factors principally related to PGI improvement have never been explored. The level of symptom intensity and PSG at admission seem to have a role. The higher the symptom intensity, the better the PGI, despite not all patients having achieved their target, as expressed by PSG. This observation can be explained by the best possibilities to improve a symptom when this has an elevated level of intensity, for example, for pain, that was the typical example of a manageable symptom because, different from other symptoms, there are many efficacious therapeutic options. Similarly, higher PSG were also independently associated with a better PGI. This can be explained by the fact that patients experiencing even a high level of symptom intensity are more likely to achieve a better satisfaction. In this study, 5%–10% of patients had higher values of PSG and admission and possibly required only minimal changes of symptom intensity to improve their condition in terms of PGI.

PSG

Most patients admitted to the palliative care unit had as a target a PSG of ≤3 for all ESAS items, as reported in previous trials [10], [11]. Although a relevant percentage of patients achieved the target after 1 week of treatment, a gap between response rates evaluated by achieving PSG and PGI has been found. It was expected that patients’ expectations would be superior to the real personal outcome achieved after a palliative care intervention. This is confirmed by another finding of this study with PSG changing after a week of treatment, accordingly to the improvement in intensity of ESAS items, lowering the targeted intensity, as patients would achieve something more once an improvement has been achieved.

In a previous retrospective study, PSG remained unchanged in outpatients at a follow‐up visit performed 1–6 weeks after the initial consultation [10]. This dynamic aspect underlines the utility of PSG rating the individual improvement rather than an average change. The finding that the higher the symptom intensity at admission, the larger the number of patients having a favorable PSG response is in contrast with previous studies in which patients with a lower baseline symptom intensity were more likely to achieve the PSG response [10], [11]. As mentioned before, differences may rely on the different setting and evaluation times.

Important information was gathered from the multivariate analysis, in which some factors were independently associated with PGI for each symptom examined. The symptom intensity and PSG recorded at T0 were independently associated with a positive PGI for many ESAS items. As mentioned above, PGI may be perceived to be positive much more often in patients starting from high levels of symptom intensity. The clinical response has been reported to be higher among patients with baseline severe intensity [10]. Similarly, patients with higher PSG (that means modest requests or limited expectations) may experience a better impression of improvement. Of interest, about 5%–10% of patients, depending on the symptom examined, have these characteristics.

In contrast, patients with higher Karnofsky level had fewer chances for PGI improvement, particularly for some symptoms like pain, depression, drowsiness, dyspnea, and insomnia. One could argue that patients with a lower Karnofsky status may be more positively impressed after a palliative care treatment or have fewer expectations. This aspect deserves specific studies.

This study has some limitations. In comparison with previous trials examining issues regarding the clinical changes as perceived by patients and PSG, data were gathered from a large number of patients recruited in palliative care units where symptom assessment and therapeutic changes are more frequently performed to achieve an improvement in the clinical condition. Therefore, data are not extendable to outpatients or home‐care settings. In this study, a PGI scale was used to test MCID, that is, the patients’ feeling to perceive a clinical change. This proved to be easy for patients, although other external criteria could be useful and require further investigation.

For Further Reading: Stuart L. Goldberg, Dhakshila Paramanathan, Raya Khoury et al. A Patient‐Reported Outcome Instrument to Assess Symptom Burden and Predict Survival in Patients with Advanced Cancer: Flipping the Paradigm to Improve Timing of Palliative and End‐of‐Life Discussions and Reduce Unwanted Health Care Costs. The Oncologist 2019;24:76–85; first published on September 28, 2018.

Implications for Practice: A seven‐item patient‐reported outcome (PRO) instrument was administered to 1,191 patients with advanced cancers. Patients self‐reporting higher levels of physical and psychological symptom burden had inferior overall survival rates. High individual item symptom PRO responses should serve as a useful trigger to initiate supportive interventions, but when scores indicate global problems, discussions regarding end‐of‐life care might be appropriate.

1. Mercadante S, Adile C, Caruselli A et al. The Palliative‐Supportive Care Unit in a Comprehensive Cancer Center as crossroad for patients' oncological pathway. PLoS One2016;11:e0157300. [PMC free article] [PubMed] [Google Scholar]

2. Hui D, Bruera E.The Edmonton Symptom Assessment System 25 years later: Past, present, and future developments. J Pain Symptom Manage2017;53:630–643. [PMC free article] [PubMed] [Google Scholar]

3. Chang VT, Hwang SS, Feuerman M.Validation of the Edmonton Symptom Assessment Scale. Cancer2000;88:2164–2171. [PubMed] [Google Scholar]

4. Norman GR, Sloan JA, Wyrwich KW.Interpretation of changes in health‐related quality of life: The remarkable universality of half a standard deviation. Med Care2003;41:582–592. [PubMed] [Google Scholar]

5. Lydick E, Epstein R.Interpretation of quality of life changes. Qual Life Res1993;2:221–226. [PubMed] [Google Scholar]

6. Maringwa J, Quintien C, King M et al. Minimal clinically meaningful differences for the EORTC QLQ‐30 and EORTC QLQ‐BN20 scales in brain cancer patients. Ann Oncol2011;22:2107–2112. [PubMed] [Google Scholar]

7. Bedard G, Zeng L, Zhang L et al. Minimal clinically important differences in the Edmonton Symptom Assessment System in patients with advanced cancer. J Pain Symptom Manage2013;46:192–200. [PubMed] [Google Scholar]

8. Lauridsen HH, Hartvigsen J, Manniche C et al. Responsiveness and minimal clinically important difference for pain and disability instruments in low back pain patients. BMC Musculoskelet Disord2006;7:82. [PMC free article] [PubMed] [Google Scholar]

9. Farrar JT, Portenoy RK, Berlin JA et al. Defining the clinically important difference in pain outcome measures. Pain2000;88:287–294. [PubMed] [Google Scholar]

10. Dalal S, Hui D, Nguyen L et al. Achievement of personalized pain goal in cancer patients referred to a supportive care clinic at a comprehensive cancer center. Cancer2012;118:3869–3877. [PMC free article] [PubMed] [Google Scholar]

11. Hui D, Park M, Shamieh O et al. Personalized symptom goals and response in patients with advanced cancer. Cancer2016;122:1774–1781. [PMC free article] [PubMed] [Google Scholar]

12. Hui D, Shamieh O, Paiva CE et al. Minimal clinically important differences in the Edmonton Symptom Assessment Scale in cancer patients: A prospective, multicenter study. Cancer2015;121:3027–3035. [PMC free article] [PubMed] [Google Scholar]

13. Breitbart W, Rosenfeld B, Roth A et al. The Memorial Delirium Assessment Scale. J Pain Symptom Manage1997;13:128–137. [PubMed] [Google Scholar]

14. Parsons HA, Delgado‐Guay MO, El Osta B et al. Alcoholism screening in patients with advanced cancer: Impact on symptom burden and opioid use. J Palliat Med2008;11:964–968. [PMC free article] [PubMed] [Google Scholar]

15. Hui D, Bruera E.Minimal clinically important differences in the Edmonton Symptom Assessment System: The anchor is key. J Pain Symptom Manage2013;45:e4–e5. [PubMed] [Google Scholar]

16. Farrar JT, Pritchett YL, Robinson M et al. The clinical importance of changes in the 0 to 10 numeric rating scale for worst, least, and average pain intensity: Analyses of data from clinical trials of duloxetine in pain disorders. J Pain2010;11:109–118. [PubMed] [Google Scholar]