Researchers have identified symptom clusters as provocative because of the implications for symptom management. However, the state of the science in symptom clusters seems to be at the descriptive stage. The current literature has focused on how many symptoms a patient must have in order to "qualify" as having a symptom cluster. This approach typically classifies patients according to cutoff scores (e.g. the patient is either sleep disturbed or not). Although this approach is informative in a descriptive sense, the research has yet to focus on the reasons why symptoms cluster. An alternative approach to the study of symptom clusters is to think of the symptom cluster as system of symptoms (measured on a continuum) that covary. In this approach, the focus changes from looking at the number or the combinations of symptoms to the study of the sources of the covariation among symptoms. From this perspective a number of analytical options become viable. One may explore the structure of symptom clusters by using hierarchical cluster analysis to derive groups of patients that are relatively homogenous in their symptom experience (and ask if these groups differ on other dimensions). Or one may employ path analysis to test hypotheses about how the relationship between two variables might be mediated by other variables. This presentation will provide examples of how these types of statistical analyses can be used to explore the structure of symptom clusters and the underlying associations among processes that lead to the phenomena of symptom clusters.
Back to Symptoms as Clusters
Back to The Advancing Nursing Practice Excellence: State of the Science