Amadeo Pesce (Presenter)
UCSD
Bio: Volunteer faculty UCSD School of Medicine Emeritus Professor University of Cincinnati
Authorship: Pesce AJ1, Metushi I1, Gonzales E2, and Fitzgerald R1;
UCSD School of Medicine Dept of Pathology and Laboratory Medicine1, and USC School of Pharmacy2.
| Short Abstract The advent of quantitative methods of determining medications and their metabolites in urine using LS-MS/MS analysis presents a quandary of establishing an appropriate cutoff to assess medication compliance. We suggest a simple frequency distribution plot after log transformation will make the data approximately Gaussian. Extrapolation of the lower values allows estimate of a 2.5% cutoff. Quantitative urinary excretion of hydrocodone (734 patients) and hydromorphone (732) were evaluated. Values less than 20ng/mL were excluded. The frequency distribution plot of hydrocodone roughly followed a Gaussian distribution with an estimated 2.5% cutoff of 20ng/mL. However, the hydromorphone was not Gaussian. |
Long Abstract
Background
The advent of quantitative methods of determining medications and their metabolites in urine using sensitive LC-MS/MS analysis has presented a quandary of how to establish an appropriate cutoff to assess medication compliance. We have suggested a simple frequency distribution plot after log transformation of the quantitative data will make the data approximately Gaussian. Extrapolation of the lower values allows estimate of a 2.5% cutoff. (References 1 and 2).
Methods
This study was approved by the UCSD IRB committee. De-identified patient data and their quantitative urinary excretion of hydrocodone and hydromorphone determined by UPLC-Xevo-TQ-S were obtained on 734 patients for hydrocodone and 732 for hydromorphone. The majority of the patient values were less than 20ng/mL. The remaining positive data (N=191 for hydrocodone and N=252 for hydromorphone) were transformed to their base 10 logarithm and plotted as a frequency distribution.
Results
Two graphs were made. The one labeled hydrocodone (see below) roughly follows a Gaussian distribution (predicted by the algorithm) and 2.5% cutoff would be about 20ng/mL. However, the one labeled hydromorphone does not have a Gaussian distribution. A similar estimate of the 2.5% cutoff is not possible.
Discussion
The excretion of hydrocodone is primarily as the unchanged drug while the hydromorphone data reflect the metabolism of hydrocodone to hydromorphone as well as ingestion of hydromorphone. The lower concentrations of hydromorphone shown likely represent metabolism of hydrocodone. We interpret these graphs to state that the >20ng/mL cutoff for hydrocodone is appropriate because one can extrapolate these values to estimate a 2.5% cutoff, while one cannot do that for the hydromorphone data. Lower cutoffs will not change the shape of the curve which appears bimodal. A limitation of the study is that prescription data for hydromorphone was not obtained. Some of the higher values hydromorphone values may reflect prescription rather than metabolism excretion.
References & Acknowledgements:
References
1. Pesce, A., West,C., West,R; Crews,B, Mikel C., Almazan,P., Latyshev, S; Rosenthal, M.,Horn, P Reference intervals: A novel approach to detect drug abuse in a pain patient population J. Opioid Management 2010; 6: 341-350
2. Pesce A, West C, West R, Crews B, Mikel C, Rosenthal M, Almazan P, and Latyshev S. Determination of medication cutoff values in a pain patient population. J. Opioid Manag. 2011;7(2):117-122.
| Description | Y/N | Source |
| Grants | no | |
| Salary | no | |
| Board Member | no | |
| Stock | no | |
| Expenses | no |
IP Royalty: no
| Planning to mention or discuss specific products or technology of the company(ies) listed above: | no |