Yes, part of the problem we had in the early days of analyzing the positive test results was from individuals who were on prescribed medications; those would be showing up.
The random testing approach we use is such that every month the names of 5% of the offender population are randomly generated. Those individuals are then tested. A positive test might come back showing that they are positive for something. But when we did further analysis, we found that a portion of those positive tests were linked to individuals who were on prescribed medication.