Skip to contents

Applies the Zhang & Yu (1998) formula to convert odds ratios from logistic regression to approximate relative risks: $$RR = \frac{OR}{(1 - p_0) + p_0 \times OR}$$ Applied to the point estimate and both CI bounds. Works directly on the output of fit_mortality_rr_logistic().

Usage

convert_or_to_rr(
  or_data,
  p0,
  or_col = "OR_death",
  ci_lower_col = "CI_lower",
  ci_upper_col = "CI_upper"
)

Arguments

or_data

Data frame with OR columns.

p0

Numeric scalar. From compute_p0()$p0.

or_col

Character. Default "OR_death".

ci_lower_col

Character. Default "CI_lower".

ci_upper_col

Character. Default "CI_upper".

Value

or_data with columns RR_death, RR_lower, RR_upper appended.

References

Zhang J, Yu KF. What's the relative risk? JAMA. 1998;280(19):1690-1.