Estimating Adverse Outcome Incidence Rate Imputing Partial Status with Application to a Phase IV Cancer Trial

• Activity Number: 248
• Type: Contributed
• Date/Time: Monday, August 4, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistical Computing
• Author(s): Jianmin Pan*+ and Shesh N. Rai and Xiaoyong Wu and Pradeep Singh and Melissa M. Hudson and Deo K. Srivastava
• Companies: University of Louisville JG Brown Cancer Center and University of Louisville and University of Louisville and Southeast Missouri State University and St. Jude Children's Research Hospital and St. Jude Children's Research Hospital
• Keywords: Phase IV clinical trial ; Imputation ; Cross-section survey data ; Interval censored data ; K-M Method ; Missing Value

Abstract

Phase IV clinical trials are designed to monitor the long-term toxic effects of drugs in cancer survivors. Evaluations to study the long-term effects of the cancer treatment are often made in cross-sectional surveys. In addition to finding prognostic factors for log-term survival outcome, estimating the cumulative incidence rates of adverse outcomes is also desired. Such data pose many issues: incomplete data, competing risks and selection bias. For example, one such study was designed to study the effect of anthracyclines exposure, received as part of treatment for childhood cancer, to cardiotoxicity.1 In this paper, we resort to imputing the missing current status using regression method, under some parametric assumptions, and then combining with methods previously described in Rai et al.2 to estimate the cumulative incidence rates in an illness-death/failure model. A comprehensive simulation study suggests that the results obtained using the imputation approach is significantly more efficient than those obtained without imputation. We further apply the proposed approach to the data reported in Rai et al.2 and compare the results reported there to our approach that utilizes imputation.