Opium derivatives—and later, synthetic opioids—have probably been used for millennia for the relief of pain. Given human biology, they’ve probably been abused for just as long. Opiate use disorders are a daily fact for primary care physicians; the use of these drugs has become more and more common for chronic non-cancer pain. These medications are very effective in the treatment of pain, but come with a lot of undesired effects, not least among them the potential of developing a substance use disorder. They also have considerable street value, with Vicodin selling for $5-$10 per tablet on the illicit market.

But our options for the treatment of pain are not unlimited. Non-steroidal anti-inflammatory medications such as ibuprofen are not safe in all patients, and are not always effective. A multi-modal approach to the treatment of chronic pain can be very helpful, but many patients do not have access to this expensive treatment, and many more simply want instant relief, something which opiates can provide, but with a price.

The abuse of prescription opiates is on the rise. Continuing with Vicodin as an example, 9.3% of American 12th graders reported using Vicodin illicitly in a recent survey. From 1994 to 2002, the mention of hydrocodone—the narcotic in vicodin–in emergency center charts increased 170%. This is a big problem.

So we have two big problems: chronic pain, and narcotic abuse. How can we treat chronic pain and avoid contributing to substance use disorders and drug diversion? One strategy has been the use of so-called narcotic contracts, which we’ve discussed at length. But absent from that discussion was the evidence.

Before we look at this evidence, we must re-examine our reasons for using these contracts. In my own practice, we generally use them to protect ourselves from becoming involuntary drug dealers, and to prevent patients from abusing the narcotics we prescribe. So how are we doing with that?
I can’t answer the first question, but the second was subjected to a systematic review published in the current issue of Annals of Internal Medicine. One of the primary findings of this review was that this question has not been well-studied. The few studies that are out there do not measure some of the most important end-points, such as abuse, dependence, overdose, and death. They also don’t focus on primary care offices, the setting in which these drugs are often prescribed. The limited data available point toward a reduction in narcotic misuse with the use of treatment contracts. They conclude:

Our systematic review reveals that weak evidence supports the use of opioid treatment agreements and urine drug testing to reduce opioid misuse, despite the theoretical benefits of these strategies. This lack of evidence may explain in part why they have not been widely adopted in primary care.

I’m not as optimistic as the authors that it is the lack of evidence driving practice here. Leaving that aside, they make some interesting points regarding plausibility, attitudes, and the use of evidence. With regard to narcotic treatment contracts and urine drug testing they write:

Even in the absence of strong evidence, several compelling reasons for physicians to consider implementing these strategies exist. First, primary care providers who use opioid treatment agreements report improved satisfaction, comfort, and sense of mastery in managing chronic pain. Second, management strategies that include treatment agreements have been associated with reductions in emergency department visits in observational studies. Third, cross-sectional studies and a case series have demonstrated that urine drug testing is a valuable tool to detect use of nonprescribed drugs and confirm adherence to prescribed medications beyond that identified by patient self-report or impression of the treating physician. Finally, implementing routine urine drug testing may improve the provider-patient relationship and clinic morale, as suggested in a letter to the editor.

This is a clearly written and subtle approach to the use of a plausible but not-yet-proved modality, and is a nice example of one way to approach the dark zones of data in science-based medicine. They give a rationale for pursuing further research (the importance and scale of the problem of narcotic misuse, and the dearth of good evidence for current practices). And they give some plausible reasons why we might continue to use this as-yet unproved modality. But they do not overplay the current state of research, or make hyperbolic conclusions.

Science-based medicine does not always give us clear guidelines to care, but often leaves us with more questions to answer. This is one way to approach a difficult problem with incomplete data.


Starrels JL, Becker WC, Alford DP, Kapoor A, Williams AR, & Turner BJ (2010). Systematic review: treatment agreements and urine drug testing to reduce opioid misuse in patients with chronic pain. Annals of internal medicine, 152 (11), 712-20 PMID: 20513829

Posted by Peter Lipson

Peter A. Lipson, MD is a practicing internist and teaching physician in Southeast Michigan.  After graduating from Rush Medical College in Chicago, he completed his Internal Medicine residency at Northwestern Memorial Hospital. He currently maintains a private practice, and serves as a teaching physician at a large community hospital He also maintains appointments as a Clinical Assistant Professor of Medicine at Wayne State University School of Medicine and at Oakland University William Beaumont School of Medicine, the first being a large, established medical school, the latter being a newly-formed medical school which will soon be accepting its first class of students.  He blogs at White Coat Underground at the Scientopia blog network. A primary goal of his writing is to illuminate the differences between science-based medicine and everything else.  His perspective as a primary care physician and his daily interaction with real patients gives him what he hopes is special insight into the current "De-lightenment" in medicine.  As new media evolve, pseudo-scientific, deceptive, and immoral health practices become more and more available to patients, making his job all that much more difficult---and all that much more interesting. Disclaimer: The views in all of of Dr. Lipson's writing are his alone.  They do not represent in any way his practice, hospital, employers, or anyone else. Any medical information is general and should not be applied to specific personal medical decisions.  Any medical questions should be directed to your personal physician.  Dr. Lipson will not answer any specific medical questions, and any emails and comments should be assumed public. Dr. Lipson receives no compensation for his writing. Dr. Lipson's posts for Science-Based Medicine are archived here.