|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2002 American Society for Clinical Oncology
Asking the Obvious Questions Regarding Patient BurdenCancer Center Statistics, Mayo Clinic, Rochester, MN IS BREVITY THE SOUL of wit or the loss of important information? The article by Bernhard et al1 in this issue of the Journal of Clinical Oncology attempts to investigate whether a simple, single-item measure of patient burden can provide supplementary information for standard measures of toxicity. The target construct is the patient’s perception of the amount of burden the disease and treatment processes have placed on him. The authors investigate the possibility of using the approach of simple, single linear analog self-assessment (LASA) items to measure the burden perceived by the patient. These measures were incorporated into a randomized, double-blinded, stratified clinical trial of 249 cancer patients who were receiving their first chemotherapy treatment. The primary goal of the trial was to compare two agents (granisetron v metoclopramide), each of which was combined with dexamethasone, as prophylaxis of chemotherapy-induced emesis. The primary end point was the incidence of episodes of emesis or nausea. Patient "burden" or "bother" has long been recognized informally as an important construct in medical literature. Much work exists on this topic, particularly in the nursing and psychology literature.2-8 Cleeland’s recent work on symptom inventories, building on his work in pain and fatigue measures, is also related to the topic of patient burden.9-11 The challenge has always been to construct an efficient and sensitive method of assessing patient burden. The literature is replete with numerous multi-item and multidimensional measures of patient burden. The practical reality of clinical research and practice is such that most of these tools would increase patient burden and therefore be somewhat self-defeating. A practical and simple measure of patient burden would be valuable in a wide range of clinical trials. The novelty of the approach used in this study is the straightforward manner in which the authors attempted to elicit from patients secondary end points that measured the degree of burden associated with the treatment and with emesis-related symptoms. The patients were asked, "on a scale from 0 to 100, how much have you been bothered by nausea/vomiting?" and "on a scale from 0 to 100, how much have you been bothered by any treatment-related difficulties?" The results indicated that these two simple questions provided complementary information to the primary end point of emesis episode incidence. In addition, sensitivity to changes over time was also demonstrated. The correlation with toxicity data indicated that while these two issues were related to routine incidence data, they were not redundant. Greater variability was observed in the overall assessment of treatment burden relative to what was observed for the emesis-related burden. Is less sometimes more? Bernhard et al’s data provide compelling evidence that simple, single-item assessments of patients’ burdens are sensitive and useful. This is the key contribution of the article and is consistent with various initiatives intended to simplify quality-of-life (QOL) assessment in oncology. The use of simple, single-item measures has been studied extensively and validated as an approach for gathering valuable clinical data.12,13 Linear or numerical analogs are a long-accepted approach.14,15 Specifically, in oncology clinical trials, it has been demonstrated that single-item global measures of QOL are, in fact, more sensitive to change with greater variability than more complex, multi-item instruments.16-18 The basic issue is a tradeoff between patient burden and sufficient detail required to indicate a clinically significant change. While a single, simple question cannot provide a detailed description of the QOL components involved, research has demonstrated that, on a 0 to 10 scale with 10 being the best response, most people will usually respond somewhere between 7 and 9. Hence, the presence of a problem appears if a score of 6 or lower is given by a patient. By itself, this does not provide an indication of the precise nature of the perceived deficit in QOL, but it does provide the impetus for further investigation. The simple question also allows for the triaging process to identify patients in need of psychosocial intervention. Should every question in a clinical investigation be derived from a reliable and psychometrically sound measurement instrument? Recently accepted national guidelines indicate that patient pain should be assessed on a reliable and sound instrument.3 Can we just ask patients simple questions without having to go through an extensive validation process? This is a hotly debated topic and is also a barrier to the routine inclusion of QOL measures in oncology clinical trials.16-18 An example of this phenomenon arose in our studies of hot flash activity among cancer patients,19,20 in which women were asked to record their estimates of the number and intensity of their hot flashes. Several reviewers over the years have asked for validity data regarding these simple hot flash diaries, despite the long-established practice of using such diaries to obtain usable clinical information. Reviewers asked whether women could actually perceive when they were having a hot flash. Our reply has been that we would not want to be the person who tries to tell a woman she is not having a hot flash when she says she is having one. The key point here is, of course, if patients say they are burdened, then they are burdened. There is no need for an in-depth psychometric assessment. How do we know that the patient is not lying to us? We do not. But if we allow such skepticism to undermine the investigation of such concepts, then we are assuming that patient/physician communication is undermined by a basic mistrust. Clinical trials are becoming increasingly complex. Thus any approach that is simple becomes appealing. "Asking the obvious" using single-item scales directly addresses the issues that have dogged the implementation of QOL measurement in oncology practice and clinical trials: patient burden and systemic cost. The inherent irony here is that there are no data to suggest the costs associated with brief QOL assessments are any more than costs required by additional laboratory assays or genetic profiling, and yet there is greater acceptance of routinely adding the latter complicating factors to clinical trials.21 Further, the clinical significance of laboratory measures and genetic testing may be no more understood than are QOL results. Assessing clinically significant change is a topic that is just beginning to emerge in the oncology literature. Here again, there is a clash between the need for thoroughness and the desire for simplicity. We are at a crossroad where the clinical oncology community is convinced that QOL assessments are an integral and important part of patient information. Unfortunately, the guidelines for implementation and interpretation of QOL assessments are still lacking. The definition of clinical significance is a classic example of having a thermometer without a calibration scale. How do we tell when the patient has a fever that requires clinical intervention if we do not know how to read the thermometer? Recent work has demonstrated that a change of 50% of the SD of any QOL tool can be considered a clinically significant change.22 This translates to a change of 8 to 10 points on a 100-point LASA item similar to that used in the study by Bernhard et al.1 There is a natural resistance to such a sweeping, simple conclusion, but within the appropriate contextual setting, it is reasonable. Perhaps a consensus on assessing the clinical significance of such simple, global QOL measures can be developed using lessons learned from measuring patient blood pressure. There is an appropriate consensus among clinicians as to what constitutes a clinically significant shift in a blood pressure reading. You can ask 100 clinicians and get the same answer: "it depends." While wonderfully consistent, this is not particularly helpful. Useful information can be obtained by asking three questions: "What order of magnitude of change would be clinically unimportant in all circumstances?" An answer something like "from 140/80 to 141/80" is then obtained. "What change would always be clinically meaningful?" elicits a response akin to "from 140/80 to 180/110." These answers provide a calibration of blood pressure readings, leaving the final question: "What order of change is likely to be minimally important?" One can semifacetiously characterize this process of trichotomizing effects to that of looking for changes in the size of a worm, a duck, or an elephant22 (Sloan and Symonds, unpublished manuscript). Both the worm and the elephant are easy to identify, recognize, and interpret. The middle-effect size (the duck) is the challenge. It is in this context that the 50% of an SD rule comes into play. While some may argue that 50% of the SD may be modified to 55%, 60%, or 40%, the discussion is basically one of the plumage of the waterfowl involved. Is it that simple? It can be, if we have the courage and common sense to interpret the results of such data and recognize the difference between a worm, a duck, and an elephant in the same manner that we make use of blood pressure data. Clinical experience and familiarity are the keys for such information being incorporated into clinical practice. The article by Bernhard et al1 does not directly address the issue of whether a single-item assessment of patient burden is superior to a more complex multi-item measure. Nonetheless, it demonstrates that even a single item can provide valuable information. This speaks directly to the heart of the question, "What, if any, QOL-related data should be gathered routinely in oncology clinical trials?" At present, regulatory agencies and clinicians are struggling with this issue. Is it valid and reliable to ask patients about their subjective interpretations of their clinical experiences? The answer to this question is "Yes." The difficulty lies in how to ask the QOL questions. There is a variety of experience and opinion on this issue. Some maintain that any such investigations must be carried out through the use of validated measurement instruments with decades of development and hundreds of patients’ data to back it up. In reality, few, if any, of these tools ever gain such validation. Even the most longstanding tools, such as the Minnesota Multiphasic Personality Inventory, Mini–Mental State Examination, Functional Assessment of Chronic Illness Therapy, and Medical Outcomes Short Form instruments, are undergoing an evolutionary process. This is neither unusual nor unwarranted. Even a physical end point such as tumor response was revamped recently via the Response Evaluation Criteria in Solid Tumors. QOL measures will continue to appear and evolve in the same manner that new clinical tools, such as prostate-specific antigen and measures for cytostatic anticancer agents, will be implemented. It has been argued that QOL is no different than any other surrogate end point biomarker21 and second only in importance to patient survival.23 Hence, QOL tools should not be feared or questioned any more than tools for measuring any other end point. There is, unfortunately, a natural fear of the unfamiliar regarding QOL measures. This is most likely due to the fact that these measures fall largely outside the scope of expertise of most clinical oncologists. It is then incumbent on the QOL and psychometric fields to provide understandable guidance to the oncology community. Psychometrics simultaneously provides a mechanism for examining a mathematical backdrop of an instrument and hurdles and barriers to simple straightforward approaches. Should every informational need be met by a multi-item, multidimensional, psychometrically sound, valid, and reliable instrument? The obvious answer is "No." Taking this issue of asking the obvious one step further, we could propose to ask all patients participating in clinical trials a series of questions related to the issue of "Was It Worth It?" (WIWI). These WIWI questions would provide information such as whether the patient would recommend the clinical trial or treatment to others, if they would do it again if they had the chance, and, in their opinion, whether the administration and side effects of the treatment were worth the benefit obtained. While adding to the burden of the clinical trial, such measures would provide the most direct and obvious assessment of patient perception/satisfaction related to the treatment given. Besides survival, what better measure is there to use to guide treatment selection? Why not ask the obvious? Indeed. On a scale from 0 to 10, how much did this article bother you, the reader? If this article score represented a burden of at least a 7 out of 10, then intervention is needed, not a validation study as to whether the reader is capable of providing such an assessment. Is such a simple assessment process any less valid than asking a patient to estimate the amount of burden perceived from medical treatment? REFERENCES
1.
Bernhard J, Maibach R, Thurlmann B, et al: Patients’ estimation of overall treatment burden: Why not ask the obvious? J Clin Oncol 20: 65-72, 2002 2. McCorkle R, Benoliel JQ: Symptom distress, current concerns, and mood disturbances after diagnosis of life threatening disease. Soc Sci Med 17: 431-438, 1983 3. Holland J: Psychosocial distress in the patient with cancer: Standards of care and treatment guidelines. Oncology (Huntington) 15: 19-24, 2000 (4 suppl Pt 2) 4. Holland JC: Cancer’s psychological challenges. Sci Am 275: 158-161, 1996[Medline] 5. King CR, Haberman M, Berry DL, et al: Quality of life and the cancer experience: The state-of-the-knowledge. Oncol Nurs Forum 24: 27-41, 1997[Medline] 6. Ferrell BR, Dow KH: Quality of life among long-term cancer survivors. Oncology 11: 565-568, 1997[Medline] 7. Caserta MS, Lund DA, Wright SD: Exploring the Caregiver Burden Inventory (CBI): Further evidence for a multidimensional view of burden. Int J Aging Hum Dev 43: 21-34, 1996[Medline] 8. Schene AH, Tessler RC, Gamache GM: Instruments measuring family or caregiver burden in severe mental illness. Soc Psychiatry Psychiatr Epidemiol 29: 228-240, 1994[CrossRef][Medline] 9. Cleeland CS, Ryan KM: Pain assessment: Global use of the Brief Pain Inventory. Ann Acad Med Singapore 23: 129-138, 1994[Medline] 10. Cleeland CS, Mendoza TR, Wang XS, et al: Assessing symptom distress in cancer patients: The M.D. Anderson Symptom Inventory. Cancer 89: 1634-1646, 2000[CrossRef][Medline] 11. Mendoza TR, Wang XS, Cleeland CS, et al: The rapid assessment of fatigue severity in cancer patients: Use of the brief fatigue inventory. Cancer 85: 1186-1196, 1999[CrossRef][Medline] 12. Hyland ME, Sodergran SC: Development of a new type of global quality of life scale, and comparison of performance and preference for 12 global scales. Qual Life Res 5: 469-480, 1996[CrossRef][Medline] 13. Wewers ME, Lowe NK: A critical review of visual analogue scales in the measurement of clinical phenomena. Res Nurs Health 13: 227-236, 1990[Medline] 14. Grunberg SM, Groshen S, Steingass S, et al: Comparison of conditional quality of life terminology and visual analogue scale measurements. Qual Life Res 5: 65-72, 1996[CrossRef][Medline] 15. Sriwatanakul K, Kelvie W, Lasagna L, et al: Studies with different types of visual analog scales for measurement of pain. Clin Pharmacol Ther 34: 234-239, 1983[Medline] 16. Sloan JA, Loprinzi CL, Kuross SA, et al: Randomized comparison of four tools measuring overall quality of life in patients with advanced cancer. J Clin Oncol 16: 3662-3673, 1998[Abstract] 17. Cella DF: Measuring quality of life in palliative care. Semin Oncol 22: 73-78, 1995 (suppl 3)[Medline] 18. Osoba D: Lessons learned from measuring health-related quality of life in oncology. J Clin Oncol 12: 608-616, 1994[Abstract] 19. Loprinzi CL, Kugler JW, Sloan JA, et al: Venlafaxine in management of hot flashes in survivors of breast cancer: A randomised controlled trial. Lancet 356: 2059-2063, 2000[CrossRef][Medline]
20.
Sloan JA, Loprinzi CL, Novotny PJ, et al: Methodological lessons learned from hot flash studies. J Clin Oncol 19: 4280-4290, 2001 21. Sloan JA, Varricchio C: Quality of life endpoints in prostate chemoprevention trials. Urology 57: 235-240, 2001[CrossRef][Medline] 22. Sloan JA, O’Fallon JR, Suman VJ, et al: Incorporating quality of life measurement in oncology clinical trials. Proc Am Stat Assoc 282-287: 1998 23. Outcomes of cancer treatment for technology assessment and cancer treatment guidelines. J Clin Oncol 14:671-679, 1996
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2002 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
