|Grant Number:||5R01CA114263-02 Interpret this number|
|Primary Investigator:||Dudley, William|
|Organization:||University Of Utah|
|Project Title:||Analysis of Individual Trajectories of Oral Mucositis|
DESCRIPTION (provided by applicant): Oral mucositis (OM) is a common side effect of high-dose chemotherapy, affecting about 400,000 patients annually. OM is typically accompanied by moderate to severe acute oral pain; often leads to other morbidities, such as infection, bleeding, nutritional compromise, and psychosocial sequelae and has a significant impact on quality of life, utilization of resources, and cost of care. If severe and unremitting, OM may prevent delivery of optimal antineoplastic therapy. The importance of OM is evidenced by a plethora of translational and clinical research seeking to understand its underlying pathobiologic mechanisms and develop more effective clinical assessment and management. The dominant biological model of OM proposes that individual trajectories of clinical manifestations (erythema, ulceration, and pain) are due to complex dynamic processes and that variability in these individual trajectories are due, in part, to individual demographic, biological, and clinical factors. Unfortunately, traditional statistical analyses, such as repeated measures Analysis of Variance (ANOVA), cannot capture individual variables in trajectories and traditional clinical assessment may not facilitate clinical management of OM when multiple variables are involved. The long-term goal is to study individual trajectories of OM with multivariate statistical and visual representation techniques that capture individual patterns of change across multiple measures. This study will use Sonis' model as a basis for innovative statistical analyses of change and multivariate visual representations of change in OM. The primary aim is to apply alternative statistical approaches such as Individual Growth Curve Analysis (IGCA) and hierarchical cluster analysis, to the study of individual trajectories of OM and associate patient-related, therapy-related and clinical variables. The secondary aim is to develop a clinically useful OM visualization tool using the Information Visualization (InfoViz) technique which visually synthesizes change across many variables. This study is a secondary analysis of a longitudinal (N=153), multivariate data set collected in an intervention study of cancer patients treated with high-dose chemotherapy. The IGCA will be used to estimate parameters of change in OM which will be examined for associations with individual factors predicted by Sonis. InfoViz methods will be used to graphically represent information across OM biological and clinical domains. Representations will be submitted to a panel of experts and two groups of clinicians for evaluation and refinement. This study will add to an understanding of OM through application of statistical models and visualization techniques that capture the multivariate nature of individual symptom trajectories.