Methods to minimize bias in meta-analyses of randomized controlled trials with treatment switching

Our project aims to better understand how cancer research studies, specifically those that test the effectiveness of treatments, can be improved when patients switch treatments during the study. This switch, which happens often in cancer treatments, can make it difficult to tell how effective the original treatment is. Currently, the way researchers combine and analyze data from multiple studies doesn’t take this switching into account very well. We plan to use computer simulations to test two main approaches to handling this issue. One approach pretends the switches never happened, while the other stops counting the data at the time of the switch. Our goal is to see how these approaches might change the results and reliability of combined study findings, focusing on measures like how long patients live without the cancer getting worse and overall survival. By doing this, we hope to provide a clearer picture of what these treatment studies really show, which will help in making better decisions about cancer treatment. This will ultimately benefit the organization by making the research they rely on more accurate and useful.

Faculty Supervisor:

Thomas Loughin;Haolun Shi;Rachel Altman

Student:

Partner:

Core Clinical Sciences

Discipline:

Mathematics

Sector:

Professional, scientific and technical services

University:

Simon Fraser University

Program: