What are competing risks and how do they affect cohort analyses?

• A. They are rare events that can be ignored in most analyses
• B. They increase the observed incidence of the primary outcome
• C. They preclude the outcome and may bias standard analyses
• D. They only affect exposure assessment

Answer: (C)

Competing risks are events that prevent the occurrence of the outcome you’re studying. In a cohort, if someone experiences a competing event—like death from another cause, or another irreversible endpoint—they can no longer develop the primary outcome of interest. This matters because standard survival analyses treat competing events as if individuals could still experience the primary outcome later, effectively censoring them in a nonrandom way. That leads to biased estimates: the risk of the primary outcome can be misrepresented, often overestimated when using methods like Kaplan-Meier, and hazard ratios from traditional Cox models can mislead if the competing risk isn’t properly accounted for.

To handle this, analysts use methods designed for competing risks, such as the cumulative incidence function that directly incorporates competing events, or subdistribution hazard models (Fine-Gray) that provide interpretation of risk in the presence of competing risks. The key point is that competing risks preclude the outcome and can bias standard analyses if not properly addressed.