A STATISTICAL APPROACH TO THE PREDICTION OF VERIFIABLE HEROIN USE FROM TOTAL CODEINE AND TOTAL MORPHINE CONCENTRATIONS IN URINE

There has been much debate in urine drug testing over what criteria sh ould be applied to total codeine and total morphine concentration data to determine the likelihood that a urine donor has used heroin and wh ether such use can be demonstrated by the presence of 6-acetylmorphine , After determining that the stability of 6-acetylmorphine in frozen u rine is adequate for a period of at least two years, a database of ove r 100 codeine and/or morphine positive urine specimens was subjected t o relative operating characteristic analysis to identify a criterion t hat would indicate a high probability of detecting 6-acetylmorphine in a specimen and thus confirming heroin use. A two-fold criterion was i dentified. Ey using a criterion that requires the total morphine conce ntration to be greater than 5.000 mg/L and the total codeine to total morphine ratio to be less than 0.125, one can predict the presence of 6-acetylmorphine with a sensitivity of 92%, a specificity of 79%, and an overall accuracy of 73%. Although this criterion is statistically t he most accurate in terms of both sensitivity and specificity for the data analyzed by the author, the results of ether criteria are present ed to aid toxicologists and medical review officers in determining if analysis for 6-acetylmorphine is likely to produce useful results.