In a randomized controlled trial, missing data led to biased results regarding anxiety

Area of Science:

• Clinical Trials
• Psychosocial Intervention
• Health Outcomes Research

Background:

• Randomization in clinical trials does not prevent bias from missing observations.
• Missing data due to death or nonresponse can lead to invalid study results.
• Understanding reasons for missing data is critical for accurate interpretation.

Purpose of the Study:

• To illustrate how bias due to missing observations can threaten randomized intervention studies.
• To examine the impact of death and nonresponse on study outcomes.
• To highlight the importance of addressing missing data in well-being research.

Main Methods:

• A randomized clinical trial investigating psychosocial intervention effects on well-being post-colorectal cancer surgery.
• Data collection through patient interviews at 3, 6, 12, and 24 months post-discharge.
• Analysis of nonresponse and mortality probabilities in relation to patient characteristics and intervention status.

Main Results:

• Nonresponse probability varied by anxiety scores differently in intervention and control groups, indicating potential bias.
• Lower physical functioning and global health status were associated with increased mortality risk.
• Observed associations between anxiety/depression and mortality could be explained by factors related to dying.

Conclusions:

• The study underscores the necessity of specific missing data analyses in well-being research.
• Addressing missing data is vital for the validity of findings in randomized intervention studies.
• Careful consideration of missing data mechanisms is essential for reliable health outcomes research.