Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Symptom Clusters: The New Frontier in Symptom Management Research

Symptom Clusters: The New Frontier in Symptom Management Research Abstract The majority of clinical studies on pain, fatigue, and depression associated with cancer are focused on one symptom. Although this approach has led to some advances in our understanding of a particular symptom, patients rarely present with a single symptom. Therefore, even though research focused on single symptoms needs to continue, it is imperative that symptom management research begins to focus on evaluating multiple symptoms, using cross-sectional and longitudinal study designs. In addition, research needs to focus on evaluating the relationships among multiple symptoms, specific interventions, and patient outcomes. One of the initial challenges in research regarding multiple symptoms is the terminology that should be used to describe the concept (e.g., symptom cluster, symptom constellation). Another significant area related to this aspect of symptom management research is determining the nature of clinically significant clusters of symptoms and their associated prevalence rates. Equally important is the need to determine what types of tools/instruments will provide the most valid and reliable data for the assessment of symptom clusters. Other areas that need to be considered as related to the assessment of symptom clusters include the establishment of cut points for symptom severity that would qualify a symptom for inclusion in a cluster; the focus of the assessment; and the choice of the outcome measures that will be used to judge the effect of a symptom cluster on the patient. In the area of intervention studies for symptom clusters, research will need to build on the limited number of clinical trials with single symptoms. Additional considerations related to research on symptom clusters include the determination of the mechanisms underlying the development of symptom clusters; the timing of the measurements for symptom clusters; and statistical challenges in the evaluation of symptom clusters. Research on symptom clusters in patients with cancer is cutting-edge science and a new frontier in symptom management research, and it needs to be done in tandem with research on single symptoms. Significance of the Problem Patients with cancer experience a variety of symptoms as a result of their disease or as a result of treatments for their disease. These symptoms are a major problem for patients, as well for their family caregivers, because the management of these symptoms is often the responsibility of the patients themselves (1). In addition, unrelieved symptoms can have deleterious effects on patient outcomes (e.g., functional status, mood states, and quality of life [QOL]) (2–5). Despite the magnitude of the problem and the negative consequences of unrelieved symptoms in patients with cancer, a recurrent theme in the research literature is that cancer pain is not assessed adequately (6–7), and when it is assessed it is undertreated (8–9). Additional evidence supports the underassessment of fatigue (10–11) and depression (12–13) in patients with cancer. Given the lack of systematic investigations into the prevalence and characteristics of symptoms associated with cancer treatment in general, and of pain, fatigue, and depression in particular, it is understandable that the number of research studies on effective interventions for pain, fatigue, and depression in cancer patients is miniscule in number. It should be noted that the majority, albeit of an already limited number, of clinical studies on pain, fatigue, or depression associated with cancer is focused on one symptom. Although this approach has led to some advances in our understanding of a particular symptom, patients rarely present with a single symptom. In fact, patients often present with multiple symptoms that may or may not be interrelated. Therefore, although research focused on single symptoms needs to continue, it is imperative that symptom management research begins to focus on evaluating multiple symptoms, using cross-sectional and longitudinal study designs. In addition, research needs to center on evaluating the relationships among multiple symptoms, specific interventions, and patient outcomes. The major goals of the recent State of the Science Conference on Symptom Management in Cancer, with a focus on pain, fatigue, and depression, are to examine the current state of knowledge regarding the occurrence, assessment, and management of pain, depression, and fatigue and to identify directions for future research. To date, symptom management researchers have taken a rather focused and reductionistic approach to the investigation of clinically significant symptoms associated with cancer and cancer treatment. This approach has yielded some important, although limited, information on the prevalence of pain, fatigue, and depression in patients with cancer, and has also yielded effective approaches for the assessment of these symptoms. However, very few randomized clinical trials have been performed to determine effective pharmacologic and nonpharmacologic interventions to manage cancer-related pain, fatigue, and depression. The deficits in symptom management research are enormous, particularly in the areas of mechanisms, developmental perspectives, assessments across the various stages of the natural history of cancer, disease-specific considerations, and effective interventions. Although additional research is warranted using a focused perspective on individual symptoms, future research in symptom management will need to take on a new challenge. As every clinician knows, patients rarely present with only one symptom. Patients with cancer often experience multiple symptoms simultaneously. Several groups of symptom management researchers (4,5,14), including our own group at the University of California, San Francisco, are beginning to explore this area of symptom management research; namely, an evaluation of symptom clusters. In the remainder of this article, I predict some of the key areas that need to be considered in determining future research trajectories on symptom clusters. I feel that I can predict with some degree of certainty that each of the presenters at the recent State of the Science Conference on Symptom Management in Cancer who were charged with addressing some aspect of symptom cluster research will present three resounding and recurring themes: This area of research is extremely important, there are very few data in this area, and more research is warranted to answer critical questions that will improve the care of patients with cancer as well as the QOL of their family caregivers. Terms One of the initial challenges in research regarding multiple symptoms is the terminology that should be used to describe this concept. Terms that could or have been used include symptom clusters,’ ‘symptom constellations,’ and the ‘co-occurrence of symptoms.’ According to the dictionary (15), a “cluster” is defined as “a number of things grouped together.” One of the definitions of a “constellation,” specifically with reference to the discipline of psychology, is given as “a group of related thoughts regarded as a cluster about one central idea” (15). Neither of these dictionary definitions appears to fit the concept of a symptom cluster in symptom management research. Dodd, Miaskowski, and Paul (5) define a “symptom cluster” as three or more concurrent symptoms that are related to each other. This definition specifies two important characteristics of a symptom cluster: it must be three or more symptoms, and the symptoms must be related to each other. However, it should be noted that the meaning of “related to each other” was not specified in the Dodd et al. paper. Several possibilities for what the phrase means can be supposed, including that the symptoms in the cluster share a common mechanism or a common etiology, that measurements of the intensity/severity of the symptoms are correlated with each other, or that the occurrence of the symptom cluster produces different outcomes compared with each of the individual symptoms. Therefore, one of the critical steps in the development of this conceptual approach to symptom management research will be to clarify this concept and determine the critical elements that need to be met to establish a symptom cluster. Nature and the Prevalence of Symptom Clusters Research on symptom clusters is still in its infancy. One of the critically important areas for this aspect of symptom management research is determining the nature of clinically significant symptom clusters and their associated prevalence rates. Undoubtedly, this task will not be an easy one. For example, in a recent study of the symptoms of advanced cancer (16), patients (n = 1000) reported a median of 11 symptoms (range = 1-27) on admission to a palliative care unit. The prevalence rates for the various presenting symptoms that affected over 40% of the patients in this study are listed in Table 1. One can see a large number of potential combinations for a variety of symptom clusters, and it will be imperative that criteria be developed to determine what constitutes a symptom cluster and which symptom clusters are clinically significant. Additional questions that can be asked and answered regarding the prevalence of clinically significant symptom clusters and how these clusters change over time, depending on a wide variety of factors, are as follows: Does the prevalence of clinically significant symptom clusters vary depending on age, gender, ethnicity, socioeconomic status, stage of disease, type of cancer, or type of cancer treatment? Does the prevalence of clinically significant symptom clusters vary over time (e.g., during a course of chemotherapy or radiation therapy, following the completion of cancer treatment)? Preliminary work from our own group has noted that the symptom cluster of pain, fatigue, and sleep disturbance has deleterious effects on patient outcomes (specifically, depressive symptoms, functional status, and QOL). Perhaps one criterion that can be used to determine whether a symptom cluster is clinically significant is whether the symptom cluster has a negative effect on one or more important patient outcomes. Assessment of Symptom Clusters One of the major challenges in the assessment of symptom clusters is to determine what types of tools or instruments will provide the most valid and reliable data. Using a more focused approach to the assessment of a single symptom, it is feasible to use a multidimensional tool, or more than one tool, to evaluate a particular symptom. For example, in the area of pain research, symptom assessment tools often evaluated multiple dimensions of the pain experience, including description, location, severity/intensity, aggravating and relieving factors, interference with various functional activities, and distress (17). The assessment of fatigue and depression has followed a similar approach, and valid and reliable instruments to evaluate multiple dimensions of the symptom experience have been developed (18–21). Researchers will need to determine which dimensions of a symptom are critical for the assessment of a symptom within a symptom cluster. Respondent burden will be a major consideration as we begin to conduct this type of research, particularly when we study the prevalence, severity, and effect of a symptom cluster in a longitudinal fashion. Cutoff Points for Symptoms within Symptom Clusters Another consideration that will need to be grappled with is whether a symptom is a symptom is a symptom. In other words, if a patient reports the presence of a symptom, does that qualify the symptom for entry into the cluster, or does the symptom have to achieve some level of intensity or distress for it to be considered a part of a symptom cluster? Said another way, do cutoff point scores for symptom severity or distress need to be established to qualify a symptom for inclusion in a symptom cluster? Research will need to be done to determine whether different levels of severity or distress of a symptom (e.g., mild, moderate, or severe) change its relationship to other symptoms within a cluster and, more important, whether the level of severity or distress of the symptom within a cluster has a differential effect on patient outcomes. Focus of the Assessment Related to the assessment of a symptom is the need at this time to focus on both an assessment of the prevalence of various symptom clusters as well as on the severity of the symptoms within a specific cluster. The focus of the assessment (i.e., prevalence or severity) will influence the choice of instrument that is used and the sample of patients that will need to be evaluated. Outcomes Measures for Symptoms Clusters Another area that requires exploration is the enumeration of the appropriate outcome measures that need to be used to evaluate the effect of symptom clusters. The two most common outcome measures that are used in clinical trials research of cancer treatments are mortality and disease progression. To date, no studies have been done on the effect of symptom clusters on mortality. However, it remains very reasonable to ask what this effect is. Additional outcome measures could include QOL, functional status, and alterations in mood states. However, one could argue that an alteration in mood (e.g., depression, anxiety) is a symptom, not an outcome. Indeed, future research will need to determine the most appropriate outcome measures for various symptom clusters both in relationship to their effect on the patient as well as in relation to judging the effectiveness of pharmacologic and nonpharmacologic interventions for symptom clusters. Intervention Studies for Symptom Clusters Undoubtedly, the absence of large-scale, randomized clinical trials that have evaluated the effectiveness of various pharmacologic and nonpharmacologic strategies for the management of pain, fatigue, and depression associated with cancer and cancer treatment will be a recurrent theme in this issue. Having worked on the development of the Clinical Practice Guideline on Cancer Pain Management, published by the Agency for Health Care Policy and Research in 1994 (22), and as the current co-chair for the revision of that guideline for the American Pain Society, I can testify to the paucity of clinical research in cancer pain management. In fact, the evidence report prepared by the New England Medical Center's Evidence-Based Practice Center for the Agency for Healthcare Research and Quality (23) stated that “the lack of high-quality evidence required that nearly every key question be answered on the basis of suboptimal or incomplete data” (p. 145). An equally distressing picture is found when one examines the paucity of intervention studies for cancer-related fatigue and depression. Again, testing the effectiveness of interventions for symptom clusters will be challenging. One approach to this challenge would be to design an intervention for one symptom within the cluster and to evaluate whether this intervention has an effect on the other symptoms within the cluster. This approach would be the easiest and perhaps the simplest. However, the dose of the intervention, particularly for cognitive-behavioral interventions, will need to be carefully defined to evaluate the effect of the intervention on the targeted symptom, as well as on the associated symptoms within the cluster. In addition, longitudinal studies will need to be done to evaluate the effect of the intervention over time, as well as the sustainability of the intervention. Another approach to intervention studies for symptom clusters would be to develop a multimodal intervention for the symptom cluster. In the area of pharmacologic interventions, concerns about polypharmacy and drug interactions will be of the utmost importance. In the realm of cognitive-behavioral interventions, the challenge will be to determine the active ingredient or ingredients present within the intervention. Equally important will be the need to develop clinically feasible and clinically useful interventions that will not overburden oncology patients and their family caregivers. Given the paucity of research on effective interventions for pain, fatigue, or depression as single symptoms, the outlook for intervention studies for symptom clusters appears grim indeed. One would think that the easiest way to approach the design of intervention studies for symptom clusters would be to combine effective interventions for single symptoms. However, these clinical trials remain to be done. Additional Considerations Related to Research on Symptom Clusters Although the list of research questions/issues in the area of symptom clusters is already quite voluminous, several additional areas are worth noting (even though this list is not exhaustive, it should stimulate discussion throughout the meeting). Determination of the Mechanisms Underlying the Development of Symptom Clusters Chronic pain is now viewed as a medical condition. Although more research is needed, unrelieved chronic pain associated with cancer and cancer treatment can be classified, based on its mechanisms, into two main types; namely, nociceptive and neuropathic pain. Knowledge of the underlying mechanisms for a specific type of pain is extremely important because it has implications for treatment. In addition, an understanding of the molecular and biochemical mechanisms for cancer pain, for example, has implications for the development of new treatments for a pain problem that is not responsive to current therapies. A similar model needs to be developed and applied to the determination of the underlying mechanisms for symptom clusters. One can hypothesize that a symptom cluster that includes pain, depression, and fatigue may share some common molecular and neurochemical pathways. Undoubtedly, research on the mechanisms underlying symptom clusters will require the development and refinement of animal models to help unravel the mysteries surrounding single or multiple mechanisms for symptom clusters. Timing of Measurements for Symptom Clusters Most of the research studies on the prevalence of pain, fatigue, and depression are descriptive and cross-sectional in nature. Initial studies were done with hospitalized patients, and then as oncology care moved to the outpatient setting, cross-sectional studies were done in that setting. Assessments of the prevalence of symptom clusters should begin with cross-sectional evaluations. However, as research in this area progresses, the timing of the assessments will be a critical variable in any study that evaluates for changes in the types and severity of symptom clusters as patients undergo cancer treatment and recover from therapy, as well as for changes in the effect of a symptom cluster on patient outcomes. Statistical Challenges in the Evaluation of Symptom Clusters Our colleagues in biostatistics will need to develop new models and refine existing models for the evaluation and analysis of symptom clusters. The need to evaluate for changes in symptom clusters over time, changes in the severity of the symptoms within the cluster and their relationship to each other over time, and the effect of a symptom cluster on multiple patient outcomes will present new statistical challenges. Approaches To Advance the Science of Symptom Cluster Research Research on symptom clusters in patients with cancer is cutting-edge science and a new frontier in symptom management research, and it needs to be done in tandem with research on single symptoms. Data from a recent Institute of Medicine report titled Improving Palliative Care for Cancer (24) explains why the science of symptom management for patients with cancer has not progressed at a more rapid rate. Foley and colleagues noted in the Institute of Medicine report (24) that in 1999 the National Cancer Institute (NCI) spent less than 1% of its 2.9 billion dollar budget on research and training in symptom management and palliative care. If the science of symptom management research is to advance faster than a snail's pace, then the NCI needs to invest more dollars to support symptom management research. Unrelieved pain, fatigue, and depression affect a huge number of cancer patients and cancer survivors. These symptoms also have significant negative effects on mood states, functional status, and QOL. The time has come for the NCI to make a substantial financial investment in symptom management and palliative care research and training. Existing funding mechanisms within the NCI need to be used to support research and training in this area, and research dollars need to be set aside from the individual-investigator grant pool to support descriptive and intervention studies in symptom management. In addition, the program project grant mechanism needs to be used to support symptom management research. This level of grant support to established research teams would move the science of symptom management forward exponentially. Particularly in the area of symptom clusters, program project grants integrated around the theme of symptom management could focus a set of integration aims on the evaluation of symptom clusters in heterogeneous patient populations. This approach would allow for comparisons across patient groups in terms of the prevalence of symptom clusters, changes in the prevalence and severity of the clusters over time, and the effect of the clusters on patient outcomes. Given the state of the science in symptom clusters research, the approach of using heterogeneous patient populations is similar to the approach used in a phase I study of a new chemotherapeutic drug. In a phase I study, the safety and maximum tolerated dose of the chemotherapy drug is determined in patients with a variety of cancer diagnoses. Using heterogeneous populations of patients in the initial investigations of symptom clusters will allow investigators to determine whether the same symptom clusters occur across different patient samples. If the same clusters do occur across different patient populations, these data would provide important clues regarding common mechanisms that may underlie the development of a specific symptom cluster. Equally important, if the same symptom clusters do not occur across different patient populations (e.g., patients with different cancer diagnoses, patients with other chronic illnesses), then these data may suggest different underlying mechanisms. It should be noted that to achieve the goal of increased funding of symptom management research within the RO1 and PO1 pools, a critical mass of experts in symptom management research is needed on study sections that review these types of grant applications. Another approach that the NCI could use to advance the science of symptom management research is the cooperative groups. Symptom management studies should be incorporated into cooperative group trials. These studies should go beyond the usual evaluations of the NCI common toxicity criteria or QOL. Although the inclusion of a QOL measure in cooperative group trials is an important initial step, these data are rarely published. However, this situation could be corrected easily with the inclusion of symptom management researchers on the committees that design cooperative group trials. Another avenue that the NCI could consider is the establishment of a cooperative group to do symptom management research. This level of commitment and financial investment by the NCI would substantiate the importance of this work within the “war on cancer.” Another mechanism that could be used to advance the science of symptom management is the network of NCI-designated Comprehensive Cancer Centers. The University of California, San Francisco, Comprehensive Cancer Center placed within its core grant renewal application a developing program in Symptom Management and Palliative Care. I believe that if the university is successful in establishing this program within the University of California, San Francisco, Comprehensive Cancer Center, it will be the first center in the country to achieve this goal. The NCI should encourage and support this approach in all NCI-designated Comprehensive Cancer Centers. One final approach that can be considered is the development of public and private partnerships in symptom management research. The NCI should take the lead in facilitating the development of partnerships among government (particularly the Food and Drug Administration), industry, and symptom management researchers to improve the science of symptom management and, ultimately, the quality of care that cancer patients and their family caregivers receive. Summary The study of symptom clusters in patients with cancer represents a new frontier in symptom management research. Although numerous challenges exist as we approach this area of scientific investigation, the rewards for patients and family caregivers will be substantial. A firm commitment is needed to provide the financial, as well as the infrastructure, support to conduct research on the occurrence of, assessment of, and interventions for pain, fatigue, and depression either as single symptoms or as symptom clusters. Table 1. Prevalence of symptoms in a sample of patients (n = 1000) on admission to a palliative care unit Symptom  Percentage  Pain  84  Easy fatigue  69  Weakness  66  Anorexia  66  Lack of energy  61  Dry mouth  57  Constipation  52  Early satiety  51  Dyspnea  50  >10% weight loss  50  Sleep problems  49  Depression  41  Symptom  Percentage  Pain  84  Easy fatigue  69  Weakness  66  Anorexia  66  Lack of energy  61  Dry mouth  57  Constipation  52  Early satiety  51  Dyspnea  50  >10% weight loss  50  Sleep problems  49  Depression  41  Adapted from reference (16). View Large References 1 Dodd MJ, Miaskowski C. The PRO-SELF© Program: a self-care intervention program for patients receiving cancer treatment. Semin Oncol Nurs  2000; 16: 300–8. Google Scholar 2 Glover J, Miaskowski C, Dibble S, Dodd MJ. Mood states of oncology outpatients: does pain make a difference? J Pain Symptom Manage  1995; 10: 120–8. Google Scholar 3 Burrows M, Dibble SL, Miaskowski C. Type of cancer pain influences patient outcomes. Oncol Nurs Forum  1998; 25: 735–41. Google Scholar 4 Miaskowski C, Lee KA. Pain, fatigue, and sleep disturbances in oncology outpatients receiving radiation therapy for bone metastasis: a pilot study. J Pain Symptom Manage  1999; 17: 320–2. Google Scholar 5 Dodd MJ, Miaskowski C, Paul SM. Symptom clusters and their effect on the functional status of patients with cancer. Oncol Nurs Forum  2001; 28: 465–70. Google Scholar 6 Von Roenn JH, Cleeland CS, Gonin R, Hatfield AK, Panda KJ. Physician attitudes and practice in cancer pain management: a survey from the Eastern Cooperative Oncology Group. Ann Intern Med  1993; 119: 121–6. Google Scholar 7 Ward SE, Goldberg N, Miller-McCauley V, Mueller C, Nolan A, Pawlik-Plank D, et al. Patient-related barriers to management of cancer pain. Pain  1993; 52: 319–24. Google Scholar 8 Cleeland CS, Gonin R, Hatfield AK, Edmonson JH, Blum RH, Stewart JA, et al. Pain and its treatment in outpatients with metastatic cancer. N Engl J Med  1994; 330: 592–6. Google Scholar 9 Cleeland CS. Undertreatment of cancer pain in elderly patients. JAMA  1998; 279: 1877–82. Google Scholar 10 Rieger PT. Assessment and epidemiologic issues related to fatigue. Cancer  2001; 92(6 Suppl): 1732– 6. Google Scholar 11 Stone P, Richardson A, Ream E, Smith AG, Kerr DJ, Kearney N. Cancer-related fatigue: inevitable, unimportant, untreatable? Results of a multi-centre patient survey. Cancer Fatigue Forum. Ann Oncol  2000; 11: 971–5. Google Scholar 12 Breitbart, W. Psycho-oncology: depression, anxiety, delirium. Semin Oncol  1994; 21: 754–69. Google Scholar 13 Breitbart, W. Identifying patients at risk for, and treatment of major psychiatric complications of cancer. Support Care Cancer  1995; 3: 45–60. Google Scholar 14 Kurtz ME, Given B, Kurtz JC, Given CW. The interaction of age, symptoms, and survival status on physical and mental health of patients with cancer and their families. Cancer  1994; 74: 2071–8. Google Scholar 15 Webster's New Twentieth Century Dictionary of the English Language. 2nd ed. New York (NY): Simon and Schuster; 1979 . Google Scholar 16 Walsh D, Donnelly S, Rybicki L. The symptoms of advanced cancer: relationship to age, gender, and performance status in 1000 patients. Support Care Cancer  2000; 8: 175–9. Google Scholar 17 Cleeland CS, Ryan KM. Pain assessment: global use of the Brief Pain Inventory. Ann Acad Med Singapore  1994; 23: 129–38. Google Scholar 18 Piper, BF, Dibble SL, Dodd MJ, Weiss MC, Slaughter RE, Paul SM. The revised Piper Fatigue Scale: psychometric evaluation in women with breast cancer. Oncol Nurs Forum  1998; 25: 677–84. Google Scholar 19 Meek Pm, Nail LM, Barsevik A, Schwartz AL, Stephen S, Whitmer K, et al., Psychometric testing of fatigue instruments for use with cancer patients. Nurs Res  2000; 49: 181–90. Google Scholar 20 Mendoza TR, Wang XS, Cleeland CS, Morrissey M, Johnson BA, Wendt JK, et al. The rapid assessment of fatigue severity in cancer patients: use of the Brief Fatigue Inventory. Cancer  1999; 85: 1186–89. Google Scholar 21 Yellen SB, Cella DF, Webster K, Blendowski C, Kaplan E. Measuring fatigue and other anemia-related symptoms with the Functional Assessment of Cancer Therapy (FACT) measurement system. J Pain Symptom Manage  1997; 13: 63–74. Google Scholar 22 Jacox A, Carr, DB, Payne R, Berde CB, Breitbart B, Cain JM, et al. Management of Cancer Pain.  Clinical Practice Guideline 9. AHCPR Publication 94-0592. Rockville, MD: Agency for Health Care Policy and Research, U.S. Department of Health and Human Services, Public Health Service; March 1994 . Google Scholar 23 Goudas L, Carr DB, Bloch R, Balk E, Ioannidis J, Terrin N, et al. Management of cancer pain.  Evidence Report/Technology Assessment 35 (Prepared by the New England Medical Center Evidenced-based Practice Center under Contract 290-97-0019). AHRQ Publication 02-E002. Rockville, MD: Agency for Healthcare Research and Quality; October 2001 . Google Scholar 24 Foley KH, Gelband H, editors. Improving palliative care for cancer—Summary and recommendations.  Washington (DC): National Academy Press; 2001 . Google Scholar © Oxford University Press http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png JNCI Monographs Oxford University Press

Symptom Clusters: The New Frontier in Symptom Management Research

Miaskowski, Christine; Dodd, Marylin; Lee, Kathryn
JNCI Monographs , Volume 2004 (32) – Jul 1, 2004
Download PDF
5 pages

Loading next page...
 
/lp/oxford-university-press/symptom-clusters-the-new-frontier-in-symptom-management-research-vIUCHPHAUk

References (25)

Publisher
Oxford University Press
Copyright
© Oxford University Press
ISSN
1052-6773
eISSN
1745-6614
DOI
10.1093/jncimonographs/lgh023
pmid
15263036
Publisher site
See Article on Publisher Site

Abstract

Abstract The majority of clinical studies on pain, fatigue, and depression associated with cancer are focused on one symptom. Although this approach has led to some advances in our understanding of a particular symptom, patients rarely present with a single symptom. Therefore, even though research focused on single symptoms needs to continue, it is imperative that symptom management research begins to focus on evaluating multiple symptoms, using cross-sectional and longitudinal study designs. In addition, research needs to focus on evaluating the relationships among multiple symptoms, specific interventions, and patient outcomes. One of the initial challenges in research regarding multiple symptoms is the terminology that should be used to describe the concept (e.g., symptom cluster, symptom constellation). Another significant area related to this aspect of symptom management research is determining the nature of clinically significant clusters of symptoms and their associated prevalence rates. Equally important is the need to determine what types of tools/instruments will provide the most valid and reliable data for the assessment of symptom clusters. Other areas that need to be considered as related to the assessment of symptom clusters include the establishment of cut points for symptom severity that would qualify a symptom for inclusion in a cluster; the focus of the assessment; and the choice of the outcome measures that will be used to judge the effect of a symptom cluster on the patient. In the area of intervention studies for symptom clusters, research will need to build on the limited number of clinical trials with single symptoms. Additional considerations related to research on symptom clusters include the determination of the mechanisms underlying the development of symptom clusters; the timing of the measurements for symptom clusters; and statistical challenges in the evaluation of symptom clusters. Research on symptom clusters in patients with cancer is cutting-edge science and a new frontier in symptom management research, and it needs to be done in tandem with research on single symptoms. Significance of the Problem Patients with cancer experience a variety of symptoms as a result of their disease or as a result of treatments for their disease. These symptoms are a major problem for patients, as well for their family caregivers, because the management of these symptoms is often the responsibility of the patients themselves (1). In addition, unrelieved symptoms can have deleterious effects on patient outcomes (e.g., functional status, mood states, and quality of life [QOL]) (2–5). Despite the magnitude of the problem and the negative consequences of unrelieved symptoms in patients with cancer, a recurrent theme in the research literature is that cancer pain is not assessed adequately (6–7), and when it is assessed it is undertreated (8–9). Additional evidence supports the underassessment of fatigue (10–11) and depression (12–13) in patients with cancer. Given the lack of systematic investigations into the prevalence and characteristics of symptoms associated with cancer treatment in general, and of pain, fatigue, and depression in particular, it is understandable that the number of research studies on effective interventions for pain, fatigue, and depression in cancer patients is miniscule in number. It should be noted that the majority, albeit of an already limited number, of clinical studies on pain, fatigue, or depression associated with cancer is focused on one symptom. Although this approach has led to some advances in our understanding of a particular symptom, patients rarely present with a single symptom. In fact, patients often present with multiple symptoms that may or may not be interrelated. Therefore, although research focused on single symptoms needs to continue, it is imperative that symptom management research begins to focus on evaluating multiple symptoms, using cross-sectional and longitudinal study designs. In addition, research needs to center on evaluating the relationships among multiple symptoms, specific interventions, and patient outcomes. The major goals of the recent State of the Science Conference on Symptom Management in Cancer, with a focus on pain, fatigue, and depression, are to examine the current state of knowledge regarding the occurrence, assessment, and management of pain, depression, and fatigue and to identify directions for future research. To date, symptom management researchers have taken a rather focused and reductionistic approach to the investigation of clinically significant symptoms associated with cancer and cancer treatment. This approach has yielded some important, although limited, information on the prevalence of pain, fatigue, and depression in patients with cancer, and has also yielded effective approaches for the assessment of these symptoms. However, very few randomized clinical trials have been performed to determine effective pharmacologic and nonpharmacologic interventions to manage cancer-related pain, fatigue, and depression. The deficits in symptom management research are enormous, particularly in the areas of mechanisms, developmental perspectives, assessments across the various stages of the natural history of cancer, disease-specific considerations, and effective interventions. Although additional research is warranted using a focused perspective on individual symptoms, future research in symptom management will need to take on a new challenge. As every clinician knows, patients rarely present with only one symptom. Patients with cancer often experience multiple symptoms simultaneously. Several groups of symptom management researchers (4,5,14), including our own group at the University of California, San Francisco, are beginning to explore this area of symptom management research; namely, an evaluation of symptom clusters. In the remainder of this article, I predict some of the key areas that need to be considered in determining future research trajectories on symptom clusters. I feel that I can predict with some degree of certainty that each of the presenters at the recent State of the Science Conference on Symptom Management in Cancer who were charged with addressing some aspect of symptom cluster research will present three resounding and recurring themes: This area of research is extremely important, there are very few data in this area, and more research is warranted to answer critical questions that will improve the care of patients with cancer as well as the QOL of their family caregivers. Terms One of the initial challenges in research regarding multiple symptoms is the terminology that should be used to describe this concept. Terms that could or have been used include symptom clusters,’ ‘symptom constellations,’ and the ‘co-occurrence of symptoms.’ According to the dictionary (15), a “cluster” is defined as “a number of things grouped together.” One of the definitions of a “constellation,” specifically with reference to the discipline of psychology, is given as “a group of related thoughts regarded as a cluster about one central idea” (15). Neither of these dictionary definitions appears to fit the concept of a symptom cluster in symptom management research. Dodd, Miaskowski, and Paul (5) define a “symptom cluster” as three or more concurrent symptoms that are related to each other. This definition specifies two important characteristics of a symptom cluster: it must be three or more symptoms, and the symptoms must be related to each other. However, it should be noted that the meaning of “related to each other” was not specified in the Dodd et al. paper. Several possibilities for what the phrase means can be supposed, including that the symptoms in the cluster share a common mechanism or a common etiology, that measurements of the intensity/severity of the symptoms are correlated with each other, or that the occurrence of the symptom cluster produces different outcomes compared with each of the individual symptoms. Therefore, one of the critical steps in the development of this conceptual approach to symptom management research will be to clarify this concept and determine the critical elements that need to be met to establish a symptom cluster. Nature and the Prevalence of Symptom Clusters Research on symptom clusters is still in its infancy. One of the critically important areas for this aspect of symptom management research is determining the nature of clinically significant symptom clusters and their associated prevalence rates. Undoubtedly, this task will not be an easy one. For example, in a recent study of the symptoms of advanced cancer (16), patients (n = 1000) reported a median of 11 symptoms (range = 1-27) on admission to a palliative care unit. The prevalence rates for the various presenting symptoms that affected over 40% of the patients in this study are listed in Table 1. One can see a large number of potential combinations for a variety of symptom clusters, and it will be imperative that criteria be developed to determine what constitutes a symptom cluster and which symptom clusters are clinically significant. Additional questions that can be asked and answered regarding the prevalence of clinically significant symptom clusters and how these clusters change over time, depending on a wide variety of factors, are as follows: Does the prevalence of clinically significant symptom clusters vary depending on age, gender, ethnicity, socioeconomic status, stage of disease, type of cancer, or type of cancer treatment? Does the prevalence of clinically significant symptom clusters vary over time (e.g., during a course of chemotherapy or radiation therapy, following the completion of cancer treatment)? Preliminary work from our own group has noted that the symptom cluster of pain, fatigue, and sleep disturbance has deleterious effects on patient outcomes (specifically, depressive symptoms, functional status, and QOL). Perhaps one criterion that can be used to determine whether a symptom cluster is clinically significant is whether the symptom cluster has a negative effect on one or more important patient outcomes. Assessment of Symptom Clusters One of the major challenges in the assessment of symptom clusters is to determine what types of tools or instruments will provide the most valid and reliable data. Using a more focused approach to the assessment of a single symptom, it is feasible to use a multidimensional tool, or more than one tool, to evaluate a particular symptom. For example, in the area of pain research, symptom assessment tools often evaluated multiple dimensions of the pain experience, including description, location, severity/intensity, aggravating and relieving factors, interference with various functional activities, and distress (17). The assessment of fatigue and depression has followed a similar approach, and valid and reliable instruments to evaluate multiple dimensions of the symptom experience have been developed (18–21). Researchers will need to determine which dimensions of a symptom are critical for the assessment of a symptom within a symptom cluster. Respondent burden will be a major consideration as we begin to conduct this type of research, particularly when we study the prevalence, severity, and effect of a symptom cluster in a longitudinal fashion. Cutoff Points for Symptoms within Symptom Clusters Another consideration that will need to be grappled with is whether a symptom is a symptom is a symptom. In other words, if a patient reports the presence of a symptom, does that qualify the symptom for entry into the cluster, or does the symptom have to achieve some level of intensity or distress for it to be considered a part of a symptom cluster? Said another way, do cutoff point scores for symptom severity or distress need to be established to qualify a symptom for inclusion in a symptom cluster? Research will need to be done to determine whether different levels of severity or distress of a symptom (e.g., mild, moderate, or severe) change its relationship to other symptoms within a cluster and, more important, whether the level of severity or distress of the symptom within a cluster has a differential effect on patient outcomes. Focus of the Assessment Related to the assessment of a symptom is the need at this time to focus on both an assessment of the prevalence of various symptom clusters as well as on the severity of the symptoms within a specific cluster. The focus of the assessment (i.e., prevalence or severity) will influence the choice of instrument that is used and the sample of patients that will need to be evaluated. Outcomes Measures for Symptoms Clusters Another area that requires exploration is the enumeration of the appropriate outcome measures that need to be used to evaluate the effect of symptom clusters. The two most common outcome measures that are used in clinical trials research of cancer treatments are mortality and disease progression. To date, no studies have been done on the effect of symptom clusters on mortality. However, it remains very reasonable to ask what this effect is. Additional outcome measures could include QOL, functional status, and alterations in mood states. However, one could argue that an alteration in mood (e.g., depression, anxiety) is a symptom, not an outcome. Indeed, future research will need to determine the most appropriate outcome measures for various symptom clusters both in relationship to their effect on the patient as well as in relation to judging the effectiveness of pharmacologic and nonpharmacologic interventions for symptom clusters. Intervention Studies for Symptom Clusters Undoubtedly, the absence of large-scale, randomized clinical trials that have evaluated the effectiveness of various pharmacologic and nonpharmacologic strategies for the management of pain, fatigue, and depression associated with cancer and cancer treatment will be a recurrent theme in this issue. Having worked on the development of the Clinical Practice Guideline on Cancer Pain Management, published by the Agency for Health Care Policy and Research in 1994 (22), and as the current co-chair for the revision of that guideline for the American Pain Society, I can testify to the paucity of clinical research in cancer pain management. In fact, the evidence report prepared by the New England Medical Center's Evidence-Based Practice Center for the Agency for Healthcare Research and Quality (23) stated that “the lack of high-quality evidence required that nearly every key question be answered on the basis of suboptimal or incomplete data” (p. 145). An equally distressing picture is found when one examines the paucity of intervention studies for cancer-related fatigue and depression. Again, testing the effectiveness of interventions for symptom clusters will be challenging. One approach to this challenge would be to design an intervention for one symptom within the cluster and to evaluate whether this intervention has an effect on the other symptoms within the cluster. This approach would be the easiest and perhaps the simplest. However, the dose of the intervention, particularly for cognitive-behavioral interventions, will need to be carefully defined to evaluate the effect of the intervention on the targeted symptom, as well as on the associated symptoms within the cluster. In addition, longitudinal studies will need to be done to evaluate the effect of the intervention over time, as well as the sustainability of the intervention. Another approach to intervention studies for symptom clusters would be to develop a multimodal intervention for the symptom cluster. In the area of pharmacologic interventions, concerns about polypharmacy and drug interactions will be of the utmost importance. In the realm of cognitive-behavioral interventions, the challenge will be to determine the active ingredient or ingredients present within the intervention. Equally important will be the need to develop clinically feasible and clinically useful interventions that will not overburden oncology patients and their family caregivers. Given the paucity of research on effective interventions for pain, fatigue, or depression as single symptoms, the outlook for intervention studies for symptom clusters appears grim indeed. One would think that the easiest way to approach the design of intervention studies for symptom clusters would be to combine effective interventions for single symptoms. However, these clinical trials remain to be done. Additional Considerations Related to Research on Symptom Clusters Although the list of research questions/issues in the area of symptom clusters is already quite voluminous, several additional areas are worth noting (even though this list is not exhaustive, it should stimulate discussion throughout the meeting). Determination of the Mechanisms Underlying the Development of Symptom Clusters Chronic pain is now viewed as a medical condition. Although more research is needed, unrelieved chronic pain associated with cancer and cancer treatment can be classified, based on its mechanisms, into two main types; namely, nociceptive and neuropathic pain. Knowledge of the underlying mechanisms for a specific type of pain is extremely important because it has implications for treatment. In addition, an understanding of the molecular and biochemical mechanisms for cancer pain, for example, has implications for the development of new treatments for a pain problem that is not responsive to current therapies. A similar model needs to be developed and applied to the determination of the underlying mechanisms for symptom clusters. One can hypothesize that a symptom cluster that includes pain, depression, and fatigue may share some common molecular and neurochemical pathways. Undoubtedly, research on the mechanisms underlying symptom clusters will require the development and refinement of animal models to help unravel the mysteries surrounding single or multiple mechanisms for symptom clusters. Timing of Measurements for Symptom Clusters Most of the research studies on the prevalence of pain, fatigue, and depression are descriptive and cross-sectional in nature. Initial studies were done with hospitalized patients, and then as oncology care moved to the outpatient setting, cross-sectional studies were done in that setting. Assessments of the prevalence of symptom clusters should begin with cross-sectional evaluations. However, as research in this area progresses, the timing of the assessments will be a critical variable in any study that evaluates for changes in the types and severity of symptom clusters as patients undergo cancer treatment and recover from therapy, as well as for changes in the effect of a symptom cluster on patient outcomes. Statistical Challenges in the Evaluation of Symptom Clusters Our colleagues in biostatistics will need to develop new models and refine existing models for the evaluation and analysis of symptom clusters. The need to evaluate for changes in symptom clusters over time, changes in the severity of the symptoms within the cluster and their relationship to each other over time, and the effect of a symptom cluster on multiple patient outcomes will present new statistical challenges. Approaches To Advance the Science of Symptom Cluster Research Research on symptom clusters in patients with cancer is cutting-edge science and a new frontier in symptom management research, and it needs to be done in tandem with research on single symptoms. Data from a recent Institute of Medicine report titled Improving Palliative Care for Cancer (24) explains why the science of symptom management for patients with cancer has not progressed at a more rapid rate. Foley and colleagues noted in the Institute of Medicine report (24) that in 1999 the National Cancer Institute (NCI) spent less than 1% of its 2.9 billion dollar budget on research and training in symptom management and palliative care. If the science of symptom management research is to advance faster than a snail's pace, then the NCI needs to invest more dollars to support symptom management research. Unrelieved pain, fatigue, and depression affect a huge number of cancer patients and cancer survivors. These symptoms also have significant negative effects on mood states, functional status, and QOL. The time has come for the NCI to make a substantial financial investment in symptom management and palliative care research and training. Existing funding mechanisms within the NCI need to be used to support research and training in this area, and research dollars need to be set aside from the individual-investigator grant pool to support descriptive and intervention studies in symptom management. In addition, the program project grant mechanism needs to be used to support symptom management research. This level of grant support to established research teams would move the science of symptom management forward exponentially. Particularly in the area of symptom clusters, program project grants integrated around the theme of symptom management could focus a set of integration aims on the evaluation of symptom clusters in heterogeneous patient populations. This approach would allow for comparisons across patient groups in terms of the prevalence of symptom clusters, changes in the prevalence and severity of the clusters over time, and the effect of the clusters on patient outcomes. Given the state of the science in symptom clusters research, the approach of using heterogeneous patient populations is similar to the approach used in a phase I study of a new chemotherapeutic drug. In a phase I study, the safety and maximum tolerated dose of the chemotherapy drug is determined in patients with a variety of cancer diagnoses. Using heterogeneous populations of patients in the initial investigations of symptom clusters will allow investigators to determine whether the same symptom clusters occur across different patient samples. If the same clusters do occur across different patient populations, these data would provide important clues regarding common mechanisms that may underlie the development of a specific symptom cluster. Equally important, if the same symptom clusters do not occur across different patient populations (e.g., patients with different cancer diagnoses, patients with other chronic illnesses), then these data may suggest different underlying mechanisms. It should be noted that to achieve the goal of increased funding of symptom management research within the RO1 and PO1 pools, a critical mass of experts in symptom management research is needed on study sections that review these types of grant applications. Another approach that the NCI could use to advance the science of symptom management research is the cooperative groups. Symptom management studies should be incorporated into cooperative group trials. These studies should go beyond the usual evaluations of the NCI common toxicity criteria or QOL. Although the inclusion of a QOL measure in cooperative group trials is an important initial step, these data are rarely published. However, this situation could be corrected easily with the inclusion of symptom management researchers on the committees that design cooperative group trials. Another avenue that the NCI could consider is the establishment of a cooperative group to do symptom management research. This level of commitment and financial investment by the NCI would substantiate the importance of this work within the “war on cancer.” Another mechanism that could be used to advance the science of symptom management is the network of NCI-designated Comprehensive Cancer Centers. The University of California, San Francisco, Comprehensive Cancer Center placed within its core grant renewal application a developing program in Symptom Management and Palliative Care. I believe that if the university is successful in establishing this program within the University of California, San Francisco, Comprehensive Cancer Center, it will be the first center in the country to achieve this goal. The NCI should encourage and support this approach in all NCI-designated Comprehensive Cancer Centers. One final approach that can be considered is the development of public and private partnerships in symptom management research. The NCI should take the lead in facilitating the development of partnerships among government (particularly the Food and Drug Administration), industry, and symptom management researchers to improve the science of symptom management and, ultimately, the quality of care that cancer patients and their family caregivers receive. Summary The study of symptom clusters in patients with cancer represents a new frontier in symptom management research. Although numerous challenges exist as we approach this area of scientific investigation, the rewards for patients and family caregivers will be substantial. A firm commitment is needed to provide the financial, as well as the infrastructure, support to conduct research on the occurrence of, assessment of, and interventions for pain, fatigue, and depression either as single symptoms or as symptom clusters. Table 1. Prevalence of symptoms in a sample of patients (n = 1000) on admission to a palliative care unit Symptom  Percentage  Pain  84  Easy fatigue  69  Weakness  66  Anorexia  66  Lack of energy  61  Dry mouth  57  Constipation  52  Early satiety  51  Dyspnea  50  >10% weight loss  50  Sleep problems  49  Depression  41  Symptom  Percentage  Pain  84  Easy fatigue  69  Weakness  66  Anorexia  66  Lack of energy  61  Dry mouth  57  Constipation  52  Early satiety  51  Dyspnea  50  >10% weight loss  50  Sleep problems  49  Depression  41  Adapted from reference (16). View Large References 1 Dodd MJ, Miaskowski C. The PRO-SELF© Program: a self-care intervention program for patients receiving cancer treatment. Semin Oncol Nurs  2000; 16: 300–8. Google Scholar 2 Glover J, Miaskowski C, Dibble S, Dodd MJ. Mood states of oncology outpatients: does pain make a difference? J Pain Symptom Manage  1995; 10: 120–8. Google Scholar 3 Burrows M, Dibble SL, Miaskowski C. Type of cancer pain influences patient outcomes. Oncol Nurs Forum  1998; 25: 735–41. Google Scholar 4 Miaskowski C, Lee KA. Pain, fatigue, and sleep disturbances in oncology outpatients receiving radiation therapy for bone metastasis: a pilot study. J Pain Symptom Manage  1999; 17: 320–2. Google Scholar 5 Dodd MJ, Miaskowski C, Paul SM. Symptom clusters and their effect on the functional status of patients with cancer. Oncol Nurs Forum  2001; 28: 465–70. Google Scholar 6 Von Roenn JH, Cleeland CS, Gonin R, Hatfield AK, Panda KJ. Physician attitudes and practice in cancer pain management: a survey from the Eastern Cooperative Oncology Group. Ann Intern Med  1993; 119: 121–6. Google Scholar 7 Ward SE, Goldberg N, Miller-McCauley V, Mueller C, Nolan A, Pawlik-Plank D, et al. Patient-related barriers to management of cancer pain. Pain  1993; 52: 319–24. Google Scholar 8 Cleeland CS, Gonin R, Hatfield AK, Edmonson JH, Blum RH, Stewart JA, et al. Pain and its treatment in outpatients with metastatic cancer. N Engl J Med  1994; 330: 592–6. Google Scholar 9 Cleeland CS. Undertreatment of cancer pain in elderly patients. JAMA  1998; 279: 1877–82. Google Scholar 10 Rieger PT. Assessment and epidemiologic issues related to fatigue. Cancer  2001; 92(6 Suppl): 1732– 6. Google Scholar 11 Stone P, Richardson A, Ream E, Smith AG, Kerr DJ, Kearney N. Cancer-related fatigue: inevitable, unimportant, untreatable? Results of a multi-centre patient survey. Cancer Fatigue Forum. Ann Oncol  2000; 11: 971–5. Google Scholar 12 Breitbart, W. Psycho-oncology: depression, anxiety, delirium. Semin Oncol  1994; 21: 754–69. Google Scholar 13 Breitbart, W. Identifying patients at risk for, and treatment of major psychiatric complications of cancer. Support Care Cancer  1995; 3: 45–60. Google Scholar 14 Kurtz ME, Given B, Kurtz JC, Given CW. The interaction of age, symptoms, and survival status on physical and mental health of patients with cancer and their families. Cancer  1994; 74: 2071–8. Google Scholar 15 Webster's New Twentieth Century Dictionary of the English Language. 2nd ed. New York (NY): Simon and Schuster; 1979 . Google Scholar 16 Walsh D, Donnelly S, Rybicki L. The symptoms of advanced cancer: relationship to age, gender, and performance status in 1000 patients. Support Care Cancer  2000; 8: 175–9. Google Scholar 17 Cleeland CS, Ryan KM. Pain assessment: global use of the Brief Pain Inventory. Ann Acad Med Singapore  1994; 23: 129–38. Google Scholar 18 Piper, BF, Dibble SL, Dodd MJ, Weiss MC, Slaughter RE, Paul SM. The revised Piper Fatigue Scale: psychometric evaluation in women with breast cancer. Oncol Nurs Forum  1998; 25: 677–84. Google Scholar 19 Meek Pm, Nail LM, Barsevik A, Schwartz AL, Stephen S, Whitmer K, et al., Psychometric testing of fatigue instruments for use with cancer patients. Nurs Res  2000; 49: 181–90. Google Scholar 20 Mendoza TR, Wang XS, Cleeland CS, Morrissey M, Johnson BA, Wendt JK, et al. The rapid assessment of fatigue severity in cancer patients: use of the Brief Fatigue Inventory. Cancer  1999; 85: 1186–89. Google Scholar 21 Yellen SB, Cella DF, Webster K, Blendowski C, Kaplan E. Measuring fatigue and other anemia-related symptoms with the Functional Assessment of Cancer Therapy (FACT) measurement system. J Pain Symptom Manage  1997; 13: 63–74. Google Scholar 22 Jacox A, Carr, DB, Payne R, Berde CB, Breitbart B, Cain JM, et al. Management of Cancer Pain.  Clinical Practice Guideline 9. AHCPR Publication 94-0592. Rockville, MD: Agency for Health Care Policy and Research, U.S. Department of Health and Human Services, Public Health Service; March 1994 . Google Scholar 23 Goudas L, Carr DB, Bloch R, Balk E, Ioannidis J, Terrin N, et al. Management of cancer pain.  Evidence Report/Technology Assessment 35 (Prepared by the New England Medical Center Evidenced-based Practice Center under Contract 290-97-0019). AHRQ Publication 02-E002. Rockville, MD: Agency for Healthcare Research and Quality; October 2001 . Google Scholar 24 Foley KH, Gelband H, editors. Improving palliative care for cancer—Summary and recommendations.  Washington (DC): National Academy Press; 2001 . Google Scholar © Oxford University Press

Journal

JNCI Monographs – Oxford University Press

Published: Jul 1, 2004

There are no references for this article.