Urine Drug Monitoring in Pain Therapy

During the past decade more attention rightly has been placed on effective pain management due to its role in contributing to the healing process and improving the quality of patients’ lives. At the same time, however, prescription pain medication overdoses have skyrocketed and are at epidemic levels, according to the U.S. Centers for Disease Control and Prevention. Given the potential for both benefit from and abuse of these drugs, professional associations and various regulatory bodies recommend urine drug monitoring for chronic pain patients, but none specify particulars about whom to test, when to test, how and how often to test, and the crucial issue of how to interpret test results. Seeking to fill this knowledge gap and start a broader discussion that one day might lead to formal guidelines, a group of 11 pain management and addiction specialists recently proposed a series of recommendations around urine drug monitoring for patients on long-term opioid therapy.

“We thought about all these issues and tried to come up with thoughtful incisive recommendations for physicians, recognizing that a formal guideline requires a much more robust evidence base than currently exists,” said co-author Perry Fine, MD. “Our hope is that people won’t be lost over this but will be able to read through and think, for every individual patient, is this a patient for whom I can optimize care and improve safety by using urine drug testing? Then, depending on what I find with that urine drug testing, how do I interpret the result and move forward in order to strengthen the therapeutic outcomes rather than weaken them.” Fine is a professor of anesthesiology at the University of Utah School of Medicine in Salt Lake City.

Filling the Knowledge Gap

The panelists and other experts agreed that the need for such guidance is high. Studies have shown that physicians generally have difficulty navigating the complicated terrain of urine drug testing. This is an even more daunting challenge for internists and family practitioners, who are more likely than ever before to be overseeing care of patients with chronic pain who often have other comorbidities and may be taking medications that could affect urine drug test results. Indeed, when the authors drafted the recommendations they particularly had primary care practitioners in mind, according to panelist Joseph Couto, PharmD, MBA, assistant professor of health economics and outcomes research at the Thomas Jefferson University School of Population Health in Philadelphia. “Ideally, it’s meant to be a resource for both specialists and primary care physicians. While we don’t have great data on how many pain physicians are using drug monitoring, just anecdotally, we think it’s quite a few. We don’t think they’re as naïve to this as primary care physicians who may have used it once or twice but aren’t as conversant with it. So it’s maybe more a resource to the latter.”

The knowledge gap about how to use and interpret urine drug monitoring results is so substantial that it affords laboratorians an excellent opportunity to shine as informed consultants, not only to pain management specialists but also internists and family practitioners, experts said. Pain management physicians often use toxicology reference labs, but that may not be the case with primary care providers. “So many hospitals have an opportunity to reach out to them and provide guidance in the selection of tests and interpretation of results. That’s a huge opportunity for AACC members,” said Gwendolyn McMillin, PhD, chair of AACC’s therapeutic drug monitoring and toxicology division. “The opportunity is there for our hospital labs to support their clinics if they want to, by consulting and/or providing point-of-care testing—perhaps by providing the personnel to operate a small analyzer in the office—or by doing the testing in the hospital lab.” McMillin is medical director of clinical toxicology and trace elements at ARUP Laboratories and associate professor of pathology at the University of Utah in Salt Lake City.

Guidance on Key Questions

The authors used a modified delphi consensus-building process to address five key issues related to urine drug monitoring in chronic patients, including whom to test, how to conduct testing, when testing should be started and repeated, how to interpret results, and how to deal with discrepant results. The panel recommended that all patients on opioids for more than 3 months should be tested. The authors also suggested that physicians might wish to adopt written treatment agreements and that the practice’s drug testing policy should be specified during the patient’s first office visit.

The panel called for comprehensive urine drug testing that covers not only commonly prescribed opioids and other prescription medications with potential for abuse, but also illicit drugs. Ideally, test results should be available on the day of the office visit; CLIA-compliant point-of-care testing (POCT) is an acceptable method of accomplishing this. However, if POCT results are inconsistent with prescribed therapy, the patient’s urine sample should be sent to a lab for further analysis, preferably with gas chromatography mass spectrometry (GC/MS) or liquid chromatography tandem MS (LC/MS/MS).

The panel recommended adjusting testing frequency based on each patient’s risk level. Patients at low-risk of misuse should be tested at least once every 6 months, while those at moderate-to-high risk should be tested at least quarterly. Offices that use POCT also should assess low-risk patients at least annually with comprehensive GC/MS or LC/MS/MS testing, and high-risk patients every 6 months.

Earlier this year, the authors presented their recommendations at the annual meeting of the American Academy of Pain Medicine and have submitted their findings for publication. They emphasized that due to a paucity of evidence, their recommendations amount to expert opinion only. “This is all based on very weak evidence,” said co-author John Peppin, DO, director of the clinical research division of The Pain Treatment Center of the Bluegrass in Lexington, Ky. “We hope the pain community and other interested individuals will begin to do the research to give us a much clearer understanding of when, where, how often to test, and whether that testing actually makes a difference in terms of abuse and diversion of prescription pain medications. Certainly we look forward to research that would either verify or refute our recommendations. We’d be supportive of either direction if we could get a clear understanding of how these should be used.”

Patterns of Drug Use

What is clear is that patients don’t strictly adhere to their medication regimens. In one study, Couto found that results from nearly 1 million patient test samples indicated that three-quarters of patients probably were not taking their medications in keeping with their prescription (Popul Health Manag 2009;12:185-190). Overall, 38% did not have detectable levels of their prescribed medication, while 27% had higher drug levels than would be expected, and 15% had lower-than-expected levels.

“This was a skewed sample representing only people who were tested, and each clinic could have had a different testing policy,” Couto explained. “But it does go to show that if you’re testing people, chances are you’ll find something that’s worthy of a conversation with patients.”

Couto stressed that in their recommendations the authors strove to balance fiscal realities with the legitimate interest in improving patient outcomes through urine drug testing. “Practices have to try and balance what they’d like to do clinically with what the healthcare system realistically can afford. This is a public health problem, but it’s not a problem that comes with cheap solutions.”

Beyond the Gottcha Moment

The panelists and other experts cautioned that urine drug monitoring in chronic pain populations should not be about gottcha moments with patients. “I think urine drug testing is best used as a tool to support behaviors that appear to be healthy and to challenge patients who don’t appear to be on track. It should give clinicians insight into what might be going on in their lives,” said Douglas Gourlay, MD, MSc, educational consultant for the Wasser Pain Management Center at Mt. Sinai Hospital in Toronto. He has written extensively about urine drug testing, and he and Howard Heit, MD, in 2005 proposed the concept of universal precautions in urine drug monitoring, similar to universal precautions used for infection control purposes (See Box, below) (Pain Med 2005;6: 107–112). “The theme of universal precautions is you can’t judge a book by its cover. The idea behind it was, let’s apply a reasonably thought-through strategy of risk containment until we can get a handle on the patient’s actual risk on an individual level. We also recognized that risk is dynamic and may need to be periodically reassessed in the context of the information the individual is providing to the clinician. It’s all about balance.”

Gourlay’s campaign to educate colleagues about not only the technical ins-and-outs of urine drug testing but also the philosophy behind it began after he learned about an experienced pain management specialist misinterpreting a test result and discharging the patient in question, who subsequently committed suicide. “The stakes here are very high. Urine drug testing is not a benign practice in so far as it can adversely impact on the doctor-patient relationship with respect to trust. When it’s overly used it can literally cripple the subject of the testing by over-medicalizing their already complicated life,” he contended.

The Complexity of Interpreting Results

Opportunities abound for misinterpreting test results, from analytical issues to a host of factors that influence the absorption, distribution, metabolism, and elimination of drugs. Examples of the latter include drug-drug and drug-food interactions, and the patient’s body mass index, age, genetic polymorphisms, and renal and liver function. “There are so many variables up in the air that to use quantitative testing as a way of sorting out who is or isn’t complying with treatment, is something I couldn’t support scientifically or clinically,” said Gourlay.

Complicating result interpretation further, pain prescriptions often give patients leeway in adjusting the dosage or dosage frequency depending on their level of pain. Recently, algorithms have been proposed to account for some of these factors and extrapolate dose adherence, but experts suggested such approaches are not ready for primetime just yet.

The POCT Challenge

From the analytical perspective, single-use POC devices with built-in immunoassays pose particular opportunities for misinterpretation. These devices, commonly used in pain management offices, are low-cost and have rapid turnaround times, which enable clinicians to discuss results with patients before they leave the office. However, they also have some distinct disadvantages, such as targeting only certain drugs in a class and cross-reacting to a variety of other prescription and over-the-counter medications. How well these subtleties are understood is unclear. For instance, “a common misconception is that an opiate screen will include all opiates and opioids. However, in general, opiate immunoassay screens will not reliably detect oxycodone, oxymorphone, meperidine, and fentanyl,” explained Amadeo Pesce, PhD, DABCC, lab director of Millennium Laboratories and principal investigator for the Millennium Research Institute in San Diego.

Peppin pointed out that despite its popularity, little research has been done on POCT urine drug monitoring. “When it comes to point-of-care testing, the literature is almost non-existent when we talk about monitoring patients on opiates. My opinion is that it should be used as a screen, if you will, with the results verified by GCMS,” he said. “The whole point of point-of-care testing is that you expect to have a lot of false-positives and few false-negatives because that’s how it’s set up. I’d be concerned about a practice that was only using point-of-care testing.”

One research team recently compared POCT against LC/MS/MS in detecting benzodiazepines and opioids and illicit drugs, respectively. In the case of benzodiazepines, they found that POCT had an overall efficiency—a measure of how often the test was right—of 78.4% in comparison to LC/MS/MS. For opioids and illicit drugs, test efficiency in comparison to LC/MS/MS ranged from 90% for oxycodone to 99.4% for cocaine (Pain Physician 2011;14:175–1187 and 259–270).

Forensic Versus Therapeutic Cutoffs

Experts also emphasized the importance of determining appropriate cutoffs for the population in question. Hospital labs tend to use higher cutoffs based on a forensic model that assumes most specimens will test negative for the drugs of interest. In contrast, urine drug testing for therapeutic monitoring assumes most samples will be positive and uses much lower cutoffs.

As an example, the standard cutoff for morphine, the target analyte for codeine/morphine testing, is 2,000 ng/mL. However, Pesce’s lab, which recently evaluated optimal cutoffs to identify 97.5% of the pain patient population, recommends a 59 ng/mL cutoff for morphine in a therapeutic monitoring paradigm (See Table, below) (Pain Medicine 2012; DOI: 10.1111/j.1526–4637.2012.01350.x).

Urine Drug Testing Cutoffs Standard forensic models of urine drug testing assume most specimens will test negative for the drugs of interest and therefore use higher cutoffs. Therapeutic monitoring in chronic pain patients assumes most samples will be positive and uses much lower cutoffs. Researchers recently proposed cutoffs designed to identify 97.5% of a chronic pain population for therapeutic monitoring purposes.
Analyte	Cutoff
	Standard	Proposed
	(values in ng/mL)
Amphetamine	1,000	76
Benzodiazepines	300
7-Amino-Clonazepam		19
Alpha-hydroxyalprazolam		15
Lorazepam		30
Opiate metabolites	2,000
Codeine		29
Morphine		59
Oxycodone	100	25
Methadone	300	89
Source: Pain Medicine 2012; DOI: 10.1111/j.1526-4637.2012.01350.x.

“A 2,000 ng/mL cutoff for opiates is not good for a pain doctor. You want to find out, since the patients often take as little medicine as possible or they’re at end of their dose, what cutoff really captures as much of the population as possible,” he explained. “We published our way of looking to determine what the appropriate cutoff should be in this population and to try to capture 97.5 percent on any particular medication. You can’t let the manufacturer set the cutoffs. The lab has to set the cutoffs.”

What Labs Can Do

Within the next couple of years, the panelists hope to review the evidence base and revisit their recommendations. In the meantime, they and other experts stressed that labs would do well to reach out to both pain specialists and generalists to be a trusted resource in this arena. “I’ve been doing this for years and I try to keep up with the literature, but I don’t claim at all to be an expert in interpreting urine drug test results,” said Peppin. “They’re very complicated and I frequently call the lab and say, what does this mean? So encouraging doctors to call whatever company or lab they’re using and try to get an understanding as much as possible of the drugs and metabolites and what the cutoffs mean—all those types of things would be good. This is something doctors would be very interested in.”

McMillin suggested that in addition to educating providers, labs might help practices find the best POCT solution for their needs. “We’ve evaluated several different point-of-care options for clients. We base selection of a device on how well its performance characteristics match the needs of the clinic in question. A high level of expertise from the lab and cooperation of the clinic is required to evaluate the options and select the best approach,” she explained.

Gourlay stressed that strong physician-lab working relationships are essential for physicians to understand the nuances of interpreting test results and improving patient care. “The time has come for value-added lab medicine, where the clinician is able to ask challenging questions of the lab director and the lab director is able to offer information about how the clinician is attempting to use the information. Together, by asking more challenging questions, we gain a deeper understanding of how best to order tests and interpret and act on the results,” he said. “I hope the panel’s recommendations will create that kind of dialogue. It’s time to dismiss the notion that either physician or lab can do their job in isolation.”