Mitigating Bias in Colorectal Cancer Research: Assessing the Limitations of the Cox Proportional Hazards Model and Considering Alternative Methods
We read with great interest the recent article by Wang et al1 titled “An Empirical Dietary Pattern Associated with the Gut Microbial Features in Relation to Colorectal Cancer Risk” in Gastroenterology. The study finds that the colorectal cancer (CRC) microbial dietary score was associated with increased CRC risk. However, we note several biases in the use of the Cox proportional hazards (CoxPH) model that the authors did not address.
