Identification of cancer-related symptom clusters: an empirical comparison of exploratory factor analysis methods.

CONTEXT: Symptom clusters, important for symptom management strategies, have been determined empirically by various analytical methods. Guidance to select methods from the options available in standard statistical packages is limited. OBJECTIVES: To compare alternative common factor analysis (FA) extraction methods appropriate to the data, to assess whether or not they determine similar symptom clusters, and to propose analytical approaches that are useful in this clinical context. METHODS: Within one month of commencing chemotherapy, outpatients from oncology and hematology clinics (n = 202) reported their symptom experience on a modified Rotterdam Symptom Checklist. Symptom distress levels in the past week were rated on a scale of one (not at all) to four (very much). In a secondary data analysis of 42 symptoms, the associations between symptoms and factors were determined using alternative common FA methods: principal axis factoring, unweighted least squares, image factor analysis, and alpha factor analysis (AFA). Symptom inclusion in a cluster was based on the interpretation of pattern and structure coefficients, and importantly, clinical relevance of the grouping. RESULTS: Five symptom clusters were commonly identified across methods: musculoskeletal discomforts/lethargy, oral discomforts, upper gastrointestinal discomforts, vasomotor symptoms, and gastrointestinal toxicities. In AFA, three additional clusters were lethargy, somatic symptoms, and treatment-related symptom clusters. CONCLUSION: The most parsimonious solution resulted from principal axis factoring, but for large numbers of symptoms, AFA may be superior by identifying symptom clusters more useful for symptom management. Interpreting complex symptom relationships may lead to the investigation of pathophysiological mechanisms and intervention opportunities. Future studies should include psychological and cognitive symptoms.

Cancer-related symptom clusters for symptom management in outpatients after commencing adjuvant chemotherapy, at 6 months, and 12 months.

GOALS OF WORK: The aim of this secondary data analysis was to investigate symptom clusters over time for symptom management of a patient group after commencing adjuvant chemotherapy. MATERIALS AND METHODS: A prospective longitudinal study of 219 cancer outpatients conducted within 1 month of commencing chemotherapy (T1), 6 months (T2), and 12 months (T3) later. Patients' distress levels were assessed for 42 physical symptoms on a clinician-modified Rotterdam Symptom Checklist. Symptom clusters were identified in exploratory factor analyses at each time. Symptom inclusion in clusters was determined from structure coefficients. Symptoms could be associated with multiple clusters. Stability over time was determined from symptom cluster composition and the proportion of symptoms in the initial symptom clusters replicated at later times. MAIN RESULTS: Fatigue and daytime sleepiness were the most prevalent distressing symptoms over time. The median number of concurrent distressing symptoms approximated 7, over time. Five consistent clusters were identified at T1, T2, and T3. An additional two clusters were identified at 12 months, possibly due to less variation in distress levels. Weakness and fatigue were each associated with two, four, and five symptom clusters at T1, T2, and T3, respectively, potentially suggesting different causal mechanisms. CONCLUSION: Stability is a necessary attribute of symptom clusters, but definitional clarification is required. We propose that a core set of concurrent symptoms identifies each symptom cluster, signifying a common cause. Additional related symptoms may be included over time. Further longitudinal investigation is required to identify symptom clusters and the underlying causes.

Multivariate methods to identify cancer-related symptom clusters.

Multivariate methods are required to assess the interrelationships among multiple, concurrent symptoms. We examined the conceptual and contextual appropriateness of commonly used multivariate methods for cancer symptom cluster identification. From 178 publications identified in an online database search of Medline, CINAHL, and PsycINFO, limited to articles published in English, 10 years prior to March 2007, 13 cross-sectional studies met the inclusion criteria. Conceptually, common factor analysis (FA) and hierarchical cluster analysis (HCA) are appropriate for symptom cluster identification, not principal component analysis. As a basis for new directions in symptom management, FA methods are more appropriate than HCA. Principal axis factoring or maximum likelihood factoring, the scree plot, oblique rotation, and clinical interpretation are recommended approaches to symptom cluster identification.

Barriers to effective cancer pain management: a survey of hospitalized cancer patients in Australia.

The purpose of this study was to examine attitudinal barriers to effective pain management in a consecutively recruited cohort of 114 cancer patients from four Australian hospitals. When surveyed, 48% of this sample reported experiencing pain within the previous 24 hours. Of these, 56% reported this pain to be "distressing, horrible or excruciating," with large proportions indicating that this pain had affected their movement, sleep and emotional well-being. Three factors were identified as potentially impacting on patients' responses to pain-poor levels of patient knowledge about pain, low perceived control over pain, and a deficit in communication about pain. A trend for older patients to experience more severe pain was also identified. These older patients reported being more willing to tolerate pain and perceive less control over their pain. Suggestions are made for developing patient education programs and further research using concepts drawn from broader social and behavioral models.