Practical approaches to minimize problems with missing quality of life data

Area of Science:

• Oncology
• Clinical Trials Methodology
• Health Outcomes Research

Background:

• Missing quality of life (QOL) data is a significant challenge in cancer clinical trials, particularly for patients with advanced disease.
• Patient clinical deterioration can lead to non-response in QOL assessments, potentially biasing results.
• Non-respondents in QOL assessments often exhibit poorer health status compared to respondents.

Purpose of the Study:

• To address the issue of missing QOL data in cancer trials.
• To explore the utility of auxiliary outcome variables as proxies for missing QOL data.
• To illustrate methods for assessing and imputing missing QOL data in advanced cancer trials.

Main Methods:

• Analysis of data from four clinical trials, comparing respondents and non-respondents to QOL assessments.
• Utilizing auxiliary outcome variables (e.g., health status) as proxies for QOL.
• Illustrating a method for imputing missing QOL data using auxiliary variables in palliative and advanced breast cancer trials.

Main Results:

• Non-respondents to QOL assessments in cancer trials generally have poorer health status than respondents.
• Auxiliary outcome variables can serve as useful proxies to evaluate the impact of missing QOL data on treatment effect estimations.
• Methods for imputation of missing QOL data using auxiliary variables can be applied.

Conclusions:

• Missing QOL data in cancer trials, especially in advanced disease, requires careful management.
• Auxiliary outcome variables offer a viable strategy to mitigate bias from missing QOL data.
• Proactive trial design incorporating preventative strategies and auxiliary QOL data collection is crucial for minimizing missing data issues.