What is Rate Comparison?
Risk indicator threshold type: Rate Comparison — performs a specialized statistical test to see if the current metric is significantly different from the overall study population.
- Calculate post-rate — an estimate of the rate for the whole study:
(1+sum(events_k))/(1/
pre-rate+sum(times_k)), where events_k and times_k are numbers of events (for example, AEs) and time-units (for example, visits) for k-th study subdivision (for example, a site), and pre-rate is an initial estimate of the rate for the whole study. - Calculate a confidence interval for the quantile of Poisson distribution of the number of events. The distribution parameters are the mean equal to post-rate * times (here "times" is the number of time-units) and the user-defined significance level "1-alpha". The meaning of the significance level is that the actual rate doesn't exceed the distribution's mean with the probability alpha if the comparison below is true.
- The number of events is compared with the lower boundary of the confidence interval if the direction of comparison is "lower" and with the higher boundary otherwise.