Risk ratio can avoid selection bias. Assuming that the death occurred in concentration in the above example, if the time node for calculating the relative risk is selected before the occurrence of concentrated death, then the relative risk cannot objectively reflect the role of the drug, but the risk ratio can reflect the risk of the drug at each time point.
Extended data:
Matters needing attention
The risk ratio reflects the difference between the two risk rates. This difference is caused by various exogenous variables, such as different types of intervention, the influence of gender (male or female) and so on. Firstly, the risk rate of a benchmark is determined, and then the influence of various exogenous variables on the risk rate is measured by regression equation.
By fixing all other variables, such as gender, age, environment, location, etc. We can pay attention to the impact of intervention types on risk, such as comparing the experimental group using specific drugs with the control group using placebo. If some confounding factors can't be fixed, then use randomized controlled trials to offset the influence of these confounding factors.
Baidu encyclopedia-risk rate