Rumke offered a simple rule of thumb for identifying the upper confidence limit for a treatment causing adverse events if none are seen in a clinical trial.


n = size of a study group for which no adverse events were noted

Upper Confidence Limit for the Occurrence of an Adverse Event

Expression in Terms of the Study Group Size


3 / n


4.61 / n


5.30 / n


To read more or access our algorithms and calculators, please log in or register.