The hazard ratio would be 2, indicating higher hazard of death from the treatment.

Regression models are used to obtain hazard ratios and their confidence intervals.

If the hazard ratio between groups remain constant, this is not a problem for interpretation.

However, interpretation of hazard ratios become impossible when selection bias exists between group.

The researchers' decision about when to follow-up is arbitrary and may lead to very different reported hazard ratios.

It should be clear that the hazard ratio is a relative measure of effect and tells us nothing about absolute risk.

The hazard ratio does not convey information about how soon the death will occur.

Treatment effect depends on the underlying disease related survival function, not just the hazard ratio.

The relationship between treatment effect and the hazard ratio is given as .

Usage of hazard ratio can be used somewhat similarly to relative risk.