Calibrate Nominal Study Estimates Using Empirical Null Distribution

Implements OHDSI HADES empirical calibration, estimating null distribution mean and variance from negative control estimates and calibrating p-values and confidence intervals.

Usage

calibrate_estimates(null_estimates, observed_effect, observed_se = 0.1)

Arguments

null_estimates

Data frame with columns log_rr (or effect) and se_log_rr (or se), or a numeric vector / matrix of negative control estimates.

observed_effect

Numeric point estimate on the log scale (e.g. log RR or log HR).

observed_se

Numeric standard error of the observed point estimate (default 0.10).

Value

A list of class c("omop_empirical_calibration", "list") with null parameters and calibrated statistics.