This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kornblith, A. B. Search for Related Content PubMed PubMed Citation Articles by Kornblith, A. B. Social Bookmarking                            What's this?

Does Palliative Care Palliate?

Dana-Farber Cancer InstituteBoston, MA

IN PALLIATIVE care, the goals of care shift from cure to improved symptom control, functioning, and quality of life, and thus, quality-of-life outcomes have assumed great importance in evaluating the efficacy of these treatments. However, there has been a growing awareness and frustration with interpreting the clinical significance of statistically significant results, as the meaning of many quality-of-life scores is not readily apparent. Nor are normative data often available from either large community samples or relevant cancer patient populations to provide a frame of reference for interpreting patients’ scores. This problem is accentuated by the range of measures used across studies, inhibiting the development of these normative databases. The clinical significance of quality-of-life scores is possibly the most pressing scientific question today in quality-of-life research. Providing satisfactory answers to critically important treatment questions, such as "Does palliative care palliate?", will hinge on our ability to translate quality-of-life findings into readily understandable terms of clinical relevance.

Quality-adjusted life-years analysis was the first effort to determine the clinical significance of quality-of-life scores, in which treatment side effects were statistically combined with disease-free and overall survival. Thus, quantity of time was adjusted by its quality to better understand the total impact of different treatments on patients’ lives. In time without symptoms of disease and toxicity (TWiST) analysis, by the simple subtraction from survival of a month’s time for any symptoms experienced, and all time subsequent to relapse, the clinical significance of quality-of-life ratings became readily understandable.¹ To capture the greater complexity of quality of life, TWiST was expanded to Q-TWiST, in which symptoms were weighted by the value of different health states. Because most of the research to date using Q-TWiST has relied on hypothetical rather than empirically derived weights and with quality-of-life ratings limited to only symptoms, a subset of the full range of quality-of-life areas, the value of this statistical method in translating complex quality-of-life information into clinically relevant terms has remained limited. Further, Q-TWiST’s focus only on time without symptoms might make it insufficiently sensitive to testing for clinically relevant differences between palliative care treatments, with many patients becoming increasingly symptomatic as the disease progresses.

Well-years analysis,² a related quality-adjusted life-years approach, is perhaps more readily applicable to evaluating palliative care treatments because of empirically derived weights of health states applied to four quality-of-life areas, yielding a statistical adjustment to survival. Unfortunately, the complexity of administration of the Quality of Well-Being Scale on which well-years analysis is based has contributed to its rarely being used in cancer clinical trials. An appropriate modification of well-years analysis for the palliative care setting is needed, based on easily administered instruments, in which weights of importance of differing health status levels for each of the major quality-of-life dimensions, derived from a normative sample of palliative care patients, are used to adjust progression-free and overall survival. Alternatively, the quality-of-life scores for each of the quality-of-life dimensions could themselves be used as the individual’s own weights, adjusting for progression-free survival.

With an increasing number of studies being conducted using the same measure, sizable databases are being developed that characterize the distribution of quality-of-life ratings for the different scales, for the different patient populations.³ Large databases of the European Organization for the Research and Treatment of Cancer (EORTC)⁴ and Medical Outcome Study SF-36,^5,6 and potentially for the Functional Assessment of Cancer Therapy⁷ and Functional Living Index-Cancer (FLIC),⁸ provide normative data and thus a framework by which to interpret where a patient’s quality of life lies with respect to other patients of similar diagnosis and disease stage. Thus, for example, it would be possible to identify patients at the extreme ends of the distribution, with quality-of-life scores deviating 1.5 SD or greater from the norm, with an absolute value suggestive of a very high or very low level of functioning. In King’s⁴ analysis of EORTC scores from 14 trials, effect sizes were determined for the different quality-of-life dimensions, specifying the change in EORTC scores required to indicate small, medium, and large changes in patients’ quality of life.

Another approach is the use of criterion validation or known groups methods⁹ to test for significant differences in quality-of-life scores between clinically relevant, known groups. This type of analysis has been conducted by pairing of quality-of-life scores from the EORTC Quality of Life Questionnaire (QLQ)-C30,⁴ FLIC, and the Memorial Symptom Assessment Scale¹⁰ with groups of expected clinical severity, including performance status, toxicity, extent of disease (local v metastatic), and treatment phase (eg, hospitalization v out-patient treatment v follow-up care). For example, the magnitude of differences in EORTC QLQ-C30’s total scores for patients at different performance and disease statuses were examined, from which effect sizes were then determined.⁴ Similarly, in a study that we conducted of ovarian cancer patients’ quality of life, a range of quality-of-life measures (eg, the FLIC, Mental Health Inventory, and Memorial Symptom Assessment Scale) were examined in relation to the Karnofsky performance status.¹⁰ The critical point at which quality-of-life scores significantly decreased across all measures was determined to be between a Karnofsky rating of 80 and 90.

Using convergent validity techniques,¹¹ psychological-state subscales in quality-of-life measures could be paired with psychiatric measures, such as the Hospital Anxiety and Depression Scale¹² and Hamilton Depression Rating Scale,¹³ with validated cutoff scores that identify those most likely to have a psychiatric diagnosis. For example, those patients scoring above cutoff scores on the Hospital Anxiety and Depression Scale would determine the appropriate cutoff score for the quality-of-life psychological-state subscale, indicative of significant distress. In a related analysis, the Wisconsin Brief Pain Inventory’s pain severity rating was examined in relation to the Brief Pain Inventory’s Pain’s Interference with Daily Activities items. The critical point in the pain severity rating indicative of a clinically important level of pain was 5, with ratings at or above that score accompanied by significant decrements in functioning in a range of activities.¹⁴

A different approach altogether is pairing patients’ subjective assessments of what they consider to be a beneficial change in their quality of life with their assessments of their quality of life, using any of the measures.^15-17 By pairing these two types of measures, one can calculate the magnitude of change of a quality-of-life score over time that is required for patients to perceive it as a clinically important improvement or worsening. This method was applied by Doyle et al¹⁸ in last month’s issue of the Journal of Clinical Oncology . Patients were asked to rate the amount of change in various areas of their quality of life from the time of their last assessment to the present, on a Likert-type scale ranging from ‘much worse’ to ‘much better’. Although the ‘minimal clinically important difference’ is defined in terms of ‘a little better’ and ‘a little worse’ in the study by Doyle et al,¹⁸ it is possible that improvements of greater magnitude, such as those expressed as ‘moderately better’ or ‘much better’, would be likely to reflect more clinically significant improvements in functioning and, therefore, the preferred value of interest.

This subjective method of analysis puts the determination of what is a clinically beneficial change squarely in the hands of the patient and may be quite different than those determined by validation methods in relation to changes in functional status. However, with both types of information, we would be able to significantly advance our ability to communicate to oncologists and patients the impact of cancer treatments on patients’ lives; what do patients consider to be important improvements, and how do these values relate to changes in their functional status or quality-adjusted time? Although both functional and subjective approaches would provide valuable information for palliative care patients as they try to make treatment decisions, the former is critical to advancing the meaning of quality-of-life scores.

With routine inclusion of criterion variables of functioning in clinical trials, the clinical significance of each of the major quality-of-life dimensions could be more easily tested, permitting the expression of quality-of-life scores in concrete functional terms. In addition to performance status and toxicity grading criteria, which are already incorporated in trials, employment status or return to one’s pretreatment or prediagnosis activity level, and a brief measure of emotional distress, with cutoff scores suggestive of a psychiatric disorder, would be additional universally occurring factors to consistently include in any quality-of-life assessment. Other measures of symptom severity that have established cutoff levels, such as pain,¹⁴ might also be considered. A criterion measure of social functioning is essential but is not yet available because what is considered to be a healthy level of social functioning is quite complex and highly individualistic. Only the extreme end of the spectrum, social isolation, is effectively anchored. A rating of social functioning would need to be developed and validated, reflecting either the degree of independence in ability to participate in social activities or the degree of resumption of social activities to a precancer diagnosis level.

It should be underscored that the ability to validly detect clinically important differences in palliative care treatments will primarily reside in the selection of measures that are sufficiently sensitive to palliative-care quality-of-life issues and their change in response to treatment over time. Any of the above criterion variables are therefore not meant to replace the more expanded quality-of-life assessment but are suggested as a mechanism for translating these scores into understandable functional terms.

Does palliative care palliate? When the significance of quality-of-life scores becomes more clinically interpretable, it will be possible to more completely answer this question. However, it is also likely that conclusions drawn from these data will vary by one’s role (ie, oncologist v patient) and from patient to patient. A number of studies concerning treatments for early and metastatic disease demonstrate that close to the majority of patients are willing to undergo toxic treatments even with only modest improvements in survival or prolongation of life.^19-21 This finding implies that feeling a little better, the ‘minimally important clinical difference’, would be justification enough for many patients to undergo palliative treatment.

Armed with more interpretable quality-of-life data, generated through the phase III setting, that are based on measures that are sensitive to palliative care issues, we will be positioned to truly evaluate whether palliative care palliates, with the understanding that effective palliation for one person will be inconsequential to another.

REFERENCES

1. Goldhirsch A, Gelber RD, Simes RJ, et al: Costs and benefits of adjuvant therapy in breast cancer: A quality-adjusted survival analysis. J Clin Oncol 7: 36-44, 1989[Abstract]

2. Kaplan RM, Anderson JP, Wu AW, et al: The Quality of Well-Being Scale: Applications in AIDS, cystic fibrosis and arthritis. Med Care 27: S27-S43, 1989 (suppl)[Medline]

3. Lydick E, Epstein RS: Interpretation of quality of life changes. Qual Life Res 2: 221-226, 1993[Medline]

4. King MT: The interpretation of scores from the EORTC quality of life questionnaire QLQ-C30. Qual Life Res 5: 555-567, 1996[Medline]

5. McHorney CA, Ware JE Jr, Raczek AE: The MOS-36-Item Short-Form Health Survey (SF-36): II. Psychometric and clinical tests of validity in measuring physical and mental health constructs. Med Care 31: 247-263, 1993[Medline]

6. Ware JE Jr: SF-36 Health Survey: Manual and Interpretation Guide. Boston, MA, MOS Trust, Inc, 1993

7. Cella DF, Tulsky DS, Gray G, et al: The Functional Assessment of Cancer Therapy Scale: Development and validation of the general measure. J Clin Oncol 11: 570-579, 1993[Abstract/Free Full Text]

8. Clinch JJ: The Functional Living Index-Cancer: Ten years later, in Spilker B (ed): Quality of Life and Pharmacoeconomics in Clinical Trials ( ed 2 ). Philadelphia, PA, Lippincott-Raven, 1996, pp 215-225

9. Nunnally JC: Psychometric Theory Second edition. New York, NY, McGraw-Hill, 1978

10. Kornblith AB, Thaler H, Wong G, et al: Quality of life of women with ovarian cancer. Gyn Oncol 59: 231-242, 1995[Medline]

11. Campbell DT, Fiske DW: Convergent and discriminant validation by the multitrait-multimethod matrix. Psychol Bull 56: 81-105, 1959[Medline]

12. Ibbotson T, Maguire P, Selby P, et al: Screening for anxiety and depression in cancer patients: The effects of disease and treatment. Eur J Cancer 30A: 37-40, 1994

13. Rabkin JG, Klein DF: The clinical measurement of depressive disorders, in Marsella AJ, Hirschfeld RMA, Katz MM (eds): The Measurement of Depression. New York, NY, Guilford Press, 1987, pp 30-83

14. Cleeland CS, Bonica JJ, Ventafridda V: Assessment of pain in cancer, in Foley KM, et al (eds): Advances in Pain Research and Therapy. New York, NY, Raven Press, 1990, pp 47-55

15. Jaeschke R, Singer J, Guyatt GH: Measurement of health status: Ascertaining the minimal clinically important difference. Control Clin Trials 10: 407-415, 1989[Medline]

16. Juniper EF, Guyatt GH, Willan A, et al: Determining a minimal important change in a disease-specific quality of life questionnaire. J Clin Epidemiol 47: 81-87, 1994[Medline]

17. Osoba D, Rodrigues G, Myles J, et al: Interpreting the significance of changes in health-related quality-of-life scores. J Clin Oncol 16: 139-144, 1998[Abstract/Free Full Text]

18. Doyle C, Crump M, Pintilie M, et al: Does palliative chemotherapy palliate? Evaluation of expectations, outcomes and costs in women receiving chemotherapy for advanced ovarian cancer. J Clin Oncol 19: 1266-1274, 2001[Abstract/Free Full Text]

19. McQuellon RP, Muss HB, Hoffman SL, et al: Patient preferences for treatment of metastatic breast cancer: A study of women with early-stage breast cancer. J Clin Oncol 13: 858-868, 1995[Abstract]

20. Ravdin PM, Siminoff IA, Harvey JA: Survey of breast cancer patients concerning their knowledge and expectations of adjuvant therapy. J Clin Oncol 16: 515-521, 1998[Abstract]

21. Lindley C, Vasa S, Sawyer WT, et al: Quality of life and preferences for treatment following systemic adjuvant therapy for early-stage breast cancer. J Clin Oncol 16: 1380-1387, 1998[Abstract/Free Full Text]

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

This article has been cited by other articles:

				B. P. Himelstein, J. M. Hilden, A. M. Boldt, and D. Weissman Pediatric Palliative Care N. Engl. J. Med., April 22, 2004; 350(17): 1752 - 1762. [Full Text] [PDF]

This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kornblith, A. B. Search for Related Content PubMed PubMed Citation Articles by Kornblith, A. B. Social Bookmarking                            What's this?