The authors’ reply

We thank DePriest et al. (1) for their comments, with specific mention of the reported noroxycodone prevalence of 49% by Elder et al. (2). As we re-evaluated the original data set, an error was determined at the start of the analysis. The data set included de-identified urine specimens prior to 2010. Testing for noroxycodone did not start until September 2010. A significant number of specimens were recorded as 'negative' for noroxyco-done and should not have been included in the analysis because testing was not performed. We have repeated the analysis and updated our results. There were 73,835 specimens that met inclusion and exclusion criteria. Further, 92% (n ¼ 68,141), 94% (n ¼ 69,603) and 97% (n ¼ 71,822) of specimens had detectable oxycodone, oxy-morphone and noroxycodone urine concentrations, respectively. Eighty-nine percent (n ¼ 65,353) of specimens were positive for all three analytes. Descriptive statistics for oxycodone, oxy-morphone and noroxycodone mole fractions are summarized in Table I.