Online Program Home
  My Program

Abstract Details

Activity Number: 506 - Statisticians' Approaches to the Realities of Clinical Trials
Type: Contributed
Date/Time: Wednesday, August 2, 2017 : 10:30 AM to 12:20 PM
Sponsor: Biopharmaceutical Section
Abstract #323578 View Presentation
Title: Approaches to Analysis of Treatment Interruptions Among Cancer Patients on Oral Oncolytic Agents
Author(s): Asish Banik* and Charles W. Given and Barbara Given and Atreyee Majumder and Alla Sikorskii and Eric Vachon
Companies: Michigan State University and Michigan State University and Michigan State University and Michigan State University and University of Arizona and Michigan State University
Keywords: Clinical trial ; missing data ; repeated measure ; generalized estimating equations ; generalized linear mixed effects models ; Binomial errors
Abstract:

This report describes the approaches to analysis and findings from a multisite clinical trial where cancer patients starting an initial course of oral oncolytic agents were randomized to either 8 weeks of automated telephone symptom assessment (control, N=135) or 8 weeks of automated symptom management and daily reminders to take oral agents (experimental, N=137). To address missing data, temporary and permanent stoppages in oral agents were determined using electronic health records for three 4-week periods. Repeated measures of summed severity of 18 symptoms at weeks 4, 8, and 12 were obtained from telephone interviews and had missing data due to attrition. For the outcomes of permanent or temporary stoppages in oral agents, generalized estimating equations approach was used to estimate the marginal (population-averaged effects) of trial arm, age, drug category, and time-varying symptom severity, while subject-specific effects were estimated using generalized linear mixed effects models with Binomial errors. No differences by trial arm or age were found; drug category and symptom severity were significant predictors of permanent and temporary stoppages over time.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2017 program

 
 
Copyright © American Statistical Association