UDT Study Misinterprets Opioid Analgesic Misuse

Urine Drug Testing (UDT) for monitoring opioid therapy compliance can be a useful clinical tool when judiciously applied in select patients and properly interpreted. However, a large study of UDT results claiming that three-fourths of patients prescribed opioids for chronic pain are noncompliant with the therapy appears to raise more questions than it answers. Inadvertently, the study also reveals significant limitations of relying on UDT as a guide for patient care.

The study, reported in a front-page feature article in the August 2009 edition of Pain Medicine News [Pizzi 2009], was originally presented in poster format at a recent meeting of the International Society of Pharmacoeconomics and Outcomes Research in Orlando, Florida. Researchers retrospectively examined UDT results from 700,000 patients (about 940,000 test samples) with chronic pain who were prescribed opioid analgesics during more than a 33-month period and they found alarming rates of purported medication noncompliance, abuse, or diversion. The outcomes appear to support advantages of universal UDT monitoring in all patients prescribed opioids; however, a closer look suggests potential misinterpretations of the data.

To begin, a few caveats should be considered. First, funding for the study was afforded by Ameritox, which is a supplier of commercial prescription monitoring services and they also provided the UDT assay data. Second, one of the lead investigators in the study is a key executive at Ameritox. Third, there was a probable selection bias in the data, according to the Pain Medicine News article, since the UDT assays were more likely ordered by practitioners for patients who were high risk or suspected of opioid medication misuse, substance abuse, or other irregularities (specific reasons for requesting testing were not noted). This last concern is very important — the data should not be interpreted as representing a random sampling of patients with chronic pain prescribed opioid analgesics and this raises questions about the validity of the study’s observations for everyday practice. Many additional questions are raised by the data, but unanswered, and these are instructive for pointing out the limitations of what can and cannot be learned via UDT monitoring.

• Prescribed pain medications were absent in 39% of all UDT samples. The study authors believe this may indicate substantial diversion of opioids; however, there are other possibilities. Are patients running out of medication early or unable to afford taking the medication as frequently as prescribed? Also, a proportion of patients may be ultrarapid metabolizers of certain opioids, which is often not appreciated (or not believed) by healthcare providers.
  
  
• Non-prescribed opioids and/or other controlled substances, such as benzodiazepines, were found in 29% of the samples. Such results do not indicate how many of these medications, unknown to the practitioner ordering the UDT, were legitimately prescribed by other healthcare providers. Or, were patients supplementing their prescribed opioid regimens with other drugs to self-medicate unrelieved pain? Furthermore, the study did not assess alcohol use as a pain-coping mechanism, which is prevalent in some patient populations and can be an important factor influencing therapeutic noncompliance and hazardous drug interactions.
  
  
• In 11% of urine samples illicit drugs were present. This seems rather ominous; however, it must be remembered that the data probably reflect a large percentage of UDT assays ordered specifically for patients suspected of substance abuse. Also, what were the specific drugs found and, especially, the proportion of samples positive for illicit marijuana? High rates of marijuana smoking have been found among some patients with pain, used for analgesic rather than recreational purposes. So, is this broadly-stated result more indicative of the self-medication of unresolved pain in the sample population or flagrant substance abuse or addiction?

Overall, the above data suggest that therapeutic noncompliance was evident in 78% of the samples tested. Quantitatively, prescribed-opioid levels were reported as higher than expected in 27% of “noncompliant” patients and lower than expected in 15%. However, are these data more indicative of metabolic variability among patients, failures to understand and follow the prescribed dosing regimens, or intentional noncompliance for surreptitious reasons? And, while Ameritox claims to have proprietary methods for estimating what the opioid levels “should be” based on typically prescribed dosing regimens of specific opioids, the validity and reliability of such quantitative assays need further, independent investigation.

We believe that data from this study should not be interpreted as denoting the prevalence of opioid misuse, abuse, diversion, or other irregularities in typical clinical populations of patients with pain. Furthermore, while UDT findings that are unexpected or abnormal could indicate problematic medication-use behaviors by a patient, they are unhelpful for understanding the motivations behind those behaviors, or their extent and legitimacy. The study authors believe their findings demonstrate significant clinical concerns and “confirm the importance of periodic urine drug screening for the population prescribed long-term opioid therapy,” but they may have commercial interests in this.

Whether the time, inconvenience, and cost of routine UDT in ALL patients prescribed long-term opioids is the very best approach seems worthy of debate. Even in the best of circumstances, urine drug testing is a limited tool that still requires the clinical skills of healthcare providers for interpreting results in ways that are tailored to each patient’s medical condition, socioeconomic situation, and treatment needs. Perhaps, there are other clinical approaches for gathering patient-centered information for medication monitoring that are of equal or greater clinical utility and less costly? What do you think? Leave a comment below.

Reference: Pizzi DM. Study: substantial number of opioid patients not taking drugs as prescribed. Pain Medicine News. 2009;7(08):1. [Available here – free registration required.]

ADDENDUM 8/28/09: This study also was just published in journal format: Couto JE, Romney MC, Leider HL, Sharma S, Goldfarb NI. High rates of inappropriate drug use in the chronic pain population. Population Health Management. 2009(Aug);12(4):185-190 [see abstract]. NOTE: There are some discrepancies between data presented in various news reports and this journal article; however, the essential findings remain the same.