Whose problem is the diversion of opioids?

Okay. In the last post I mentioned that I get annoyed when concerns about opioids being diverted pop up in discussions about when opioids are indicated treatments. It's not that I don't think diversion is an important concern in drug policy. It's just that I feel like it shouldn't be part of discussions of when opioids are good treatments.

Anyway, I was annoyed by not knowing I feel this way.* So, the following is some off-the-cuff noodling about when concerns are relevant in decisions about the use of opioid drugs. I'm not at all sure about how much of it I want to stand behind. But hopefully it might be useful for sparking some discussion.

*I remember someone once telling me that philosophy starts with a sense of wonder. I have since found that, for me, it usually starts with annoyance; it ends in wonder.

-------------

In general, I tend to think that the dangers of opioid diversion --opioids ending up outside of the patient's hands-- get too much weight in discussions of drug policy (although some recent statistics on overdose and death rates involving opioids are giving me some pause in my beliefs about the severity of diversion's harms).

But in addition to questions about how severe the consequences of diversion are, we also need to know whose problem it is. A comprehensive drug policy spans many different areas including, inter alia, the law in its criminal, civil, and regulatory forms; professional determinations of best clinical practices; and individual clinicians' decisions about how to treat individual patients. Thus we need to know whether preventing diversion should have the same importance for everyone involved in the prescription drug arena.

I'm going to suggest that preventing diversion can be a legitimate concern at the more general levels. And they may inform doctors practices in a general way. But, I suspect, the potential harms resulting from diversion should not factor into a doctor's decisions about what medications to prescribe a patient.

My claims here will rest on the supposition that a clinician's ethical responsibilities arise from her patient's individual welfare. Her professional obligation is not the promotion of the general welfare via her interactions with a certain individual. The clinician's responsibility is to alleviate her patient's suffering in the safest and most effective way available.

A rough analogy may help bring out this distinction between duties based in the promotion of the general welfare and duties based in the promotion of an individual's welfare. In an adversarial system of criminal justice like we have in the United States, the role of a defense attorney is to advocate for her clients interests as best she can. Even if she recognized that her client's conviction may benefit the public at large, she is obligated to ignore that fact in doing her job. This doesn't mean that the job of defense attorney is entirely removed from the enterprise of promoting the public good. It's that a system in which a party is assigned to look out only for the interests of the defendant is more likely to be better overall. (One major disanalogy here is that my supposition about the source of the doctor's duties need not appeal to claims about what would best promote the general welfare.)

If we a clinician's duties as tied her patient's welfare in this way, concerns about the welfare of others are thus (nearly always) irrelevant to decisions about what substances to prescribe her patient. This suggests that even though the clinician may foresee that others may be harmed through diversion if she prescribes an opioid to a patient, this possibility should have no weight in her decision about what to prescribe. Her duty arises from and is directed at the health of her patients, not the health of people in general.

Obviously, this has its limits. Massive harms to others may trump this obligation. And it may be that if two treatments were exactly equal in their efficacy and safety, then considerations of the general good or other effects on others may break the tie.

Nor does this mean that the doctor must completely ignore the possibility that the drug will be diverted. Other public entities' interests in preventing diversion are based in their obligation to protect public health overall. But given the source of her professional obligations, the clinician's concerns about diversion should be limited to its effects on her patient's health.

Clearly, a responsible clinician must be attuned to the possibility that the patient herself will divert the drug. But her vigilance is not demanded by the need to prevent harms to the recipients of the diversion. It comes from her responsibilities to the patient. The clinician's treatment decisions must be based on the supposition that the patient will comply with the prescribed regime. She cannot aim to promote an individual's welfare by prescribing her a substance that she believes that the patient will not take. Therefore, the belief that the patient will take the drug as prescribed is a necessary condition of justifiably prescribing an opioid.

Suppose that a patient is accompanied by a stoned adolescent whose T-shirt reads "I love drugs!" Does this necessary condition imply that she ought to take into consideration the likelihood that the son will divert the drugs?

The answer seems to be yes. She cannot prescribe a medication to benefit her patient if she believes that the patient won't take the drug because someone else will steal it. Of course, it's unlikely that the suspicion in this case would justify her refusing to prescribe an otherwise indicated opioid Much will hang on the strength of her conviction that the drug will be diverted. In the drug diverting adolescent case, the clinician may be required to put special emphasis on the need to keep control of the medication in counseling the patient. But as long as she can be satisfied that the patient will be reasonably vigilant, she will be justified in writing the prescription. Her uncertainty about the likelihood of diversion combined with the need to respect the patient's autonomy will set the bar for reasonable vigilance pretty low.

Cases in which she should altogether refuse to prescribe on these grounds will likely be rare. But they are easy to imagine. Suppose that a disabled patient is completely dependent on her caretaker for all of her medications. If the clinician was convinced that the caretaker would divert a significant portion of the prescribed opioid, then she should not write the prescription. Indeed, doing so would be tantamount to writing the prescription for the caretaker. Though, she may have some obligation to seek other ways of getting the indicated treatment to the patient (e.g., recommending at home nursing visits, and patient treatment).

What's important is the way concerns about diversion are figuring in here. A clinician should be cautious of diversion insofar as it would interfere with her patient's treatment. Her responsibilities do not depend on how the recipients of the diverted substance may be affected. Those dangers of diversion give her reason to, for example, keep her cabinets locked. But they should be irrelevant to her decisions about patients' treatments.

This is not to say that a comprehensive drug policy should not be concerned about the harms to non-patients who gain access to opioids through diversion. It is a fact that the availability of opioids in legitimate channels will involve some diversion and some non- patients will be harmed. While the clinician's responsibility is based in her individual patient's welfare, government policies are properly attuned to protecting welfare across the board. Thus entities (in the US) like the FDA, the Department of Justice and the DEA are justified in creating policies and enforcement practices which will minimize the amount of diversion.

But this picture of the clinician's obligations does create tension between the government's proper aims creating drug policy and the duties of clinicians. We should thus want a principled way of resolving these kinds of inevitable conflict. One possibility is that one set of considerations will always trump the other (that is, the first set is lexically prior to the other).

To see the implications of a lexical ordering of these considerations suppose that the paramount consideration in shaping drug policy was ensuring clinicians' abilities to carry out their duties to their patients. This would have implications for how we decide conflicts. Such a partial lexical ordering would entail that the protection of access to safe and effective drugs cannot be trumped by considerations about diversion. More generally, this might mean that any proposed policy that would promote the general good could be vetoed if it unreasonably affected the ability of clinicians to treat their patients.

This ordering of concerns would be unlikely to undermine reasonable restrictions on the use and prescription of opioids. For example, this is compatible with a well regulated and organized system for inventory control in the manufacturing, shipping, and distribution of opioids. The same is true for methods of verifying the legitimacy of prescriptions and the identity of patients. But some apparently relatively mild restrictions on prescribing ability may not be compatible with this set of drug policy priorities.

For example, the FDA is presently considering requiring all clinicians who prescribe powerful long-acting opioids to have a special certification. Many general practice clinicians who currently prescribe such medications may be unwilling to go through the hassle of obtaining and maintaining the certification. If the certification process was unduly difficult, many clinicians would be unable to prescribe the medications that they thought were best indicated for their patients conditions. Such a regulation would likely decrease the number of deaths from diversion. But no matter how many diversion related deaths would be prevented, it should be rejected if we believe that the clinician's abilities to treat their patients should always trump any other consideration.

So, in sum, here's what I've suggested: If we think about the source and nature of clinicians' professional obligations in a particular way, then concerns about diversion should not play a role in determining whether to prescribe an opioid (outside of diversion undermining the treatment regime). Direct focus on preventing diversion is instead the job of regulatory agencies whose mission is the common public good.

I haven't given any argument in favor of the further idea that concerns about diversion should always be subordinated to clinicians ability to prescribe opioids as they see fit. Though I am definitely attracted to this view. We can leave that a subject for another post.