Abstract
There are many different ways of calculating the impact of treatment on drug use; percentage of positive drug tests, probability of drug use, percentage of patients abstaining from any use, total number of days of use, daily probability of use and average days till next use, are some examples reported in the literature. We prefer average days till next use because (1) it allows intermittent drug use and relapse; (2) it fits the client's count of drug-free days, and (3) it simultaneously accounts for both tests results and time between tests. We show by way of an example, how conclusions arrived at using average days till next use are likely to be different from other measures in analysis of recent data from impact of online treatment on drug use.