Editorial: The Opioid Epidemic and Orthopaedic Surgery—No Pain, Who Gains?

The available information in the lay press and medical journals about the opioid epidemic is both contradictory and confusing. One media outlet suggests the problem is both increasing [12] and decreasing [9], seemingly at the same time. A large federal healthcare agency claims pain should be considered the “5th Vital Sign” [6], even as The Joint Commission—which earlier had pushed for more-robust pain management—says it shouldn't be a vital sign after all [3]. And the latest data from the Centers for Disease Control and Prevention note that while opioid prescribing decreased in 2015, overall it remains higher than it was in 1999 [10]. But some statistics can help to clarify the real scope of the problem. For example, the United States represents only 4.6% of the world's population, but Americans consume 80% of the global opioid supply and 99% of the hydrocodone [18]. In 2015, more than 52,000 people in the United States died from drug overdoses, and some 15,000 of those overdoses involved a prescribed opioid [10]. The CDC reports that prescription-opioid abuse, dependence, and overdose costs the US economy an estimated USD 78.5 billion each year [10]. What seems clear enough is that the opioid epidemic is large, and that orthopaedic surgeons have been and are still a part of the problem. On the bright side, orthopaedic surgeons also can be a big part of the solution. The causes of opioid abuse in the US are many and they are complicated. Some trace the genesis of this problem to a decades-old letter to the editor in The New England Journal of Medicine (NEJM) [23] that offered overly reassuring and insufficiently researched claims about opioid safety. The letter, which now includes a warning label on the NEJM website [22], has been uncritically cited innumerable times—often to advantage by drug companies [31]. Others attribute the problem to advocacy groups that recommended “getting pain to zero,” and to aggressive marketing by pharmaceutical companies that unrealistically minimized the risks associated with these drugs [20]. And as much as we might like to attribute the problem to family doctors who prescribe narcotics for chronic conditions like arthritis, or to for-pay pain clinics that spring up like mushrooms after the rain, in fact our specialty is deeply involved [14, 29], and orthopaedic patients are at risk. Orthopaedic surgery is painful, and orthopaedic surgeons prescribe more narcotics than physicians in almost any other specialty [29]. Of course, no one wants to see his or her patients in pain, but is it reasonable to expect to eliminate postoperative pain entirely? And even if it is, does that goal justify exposing opioid-naïve patients to long-acting narcotics? Certainly numerous small studies have recommended use of preemptive long-acting narcotics in just this way [5, 8, 15]. But a careful read of these papers finds that they do so on the basis of barely perceptible differences in joint ROM, and improvements in pain that are so small that they typical patient is unlikely to notice the difference (defined as an effect size below the minimum clinically important difference [13]). For decisions like this, statistical significance is not enough; we need to focus on treatment-effect sizes [17]. Rather astoundingly, one study recommended the use of long-acting narcotics in these patients even in the absence of discernible analgesic differences [11], and another trial endorsed fentanyl patches for outpatient arthroscopic shoulder surgery [19], a long step onto a very-slippery slope [28]. Our task as clinicians is to assess the risks and benefits of opioid utilization as thoughtfully as we can, and our duty as clinician-scientists is to study the problem so that robust, evidence-based guidelines can be developed. Authors often make claims that multimodal pain management approaches are “safe and effective,” when in fact it appears they may be neither. Studies of a few dozen to a few hundred patients are far too small to prove safety [16], especially when the serious complications we worry most about (here, drug overdose or addiction following surgery) are uncommon or rare. Having said that, it appears that exposure to narcotics after a surgical procedure will cause a surprisingly high proportion of previously narcotic-naïve patients to continue to use these drugs well after recovery is complete [1], and long-acting narcotics may be especially dangerous [24]. Efficacy of long-acting narcotics as part of multimodal analgesic approaches likewise is questionable; as alluded to earlier, patients do not perceive improvements of <2 cm on a 10-cm VAS pain scale to be clinically important [13]; yet this is the effect size typically associated with the use of these drugs for this indication [5, 8, 15]. And on the subject of indications: The large majority of long-acting narcotics are not FDA approved for acute pain [2, 21]. The FDA specifically recommends against using OxyContin or MS Contin on an as-needed basis or in the immediate post-operative period, instead providing a black-box warning for both drugs, restricting them to situations “when a continuous, around-the-clock opioid analgesic is needed for an extended period of time” [26, 27]. Although surgeons are free to use drugs for off-label indications, they should do so only after careful consideration, both for safety and medicolegal reasons [4]. Some important endpoints—such as the fate of excess bottles of pills sitting on orthopaedic patients’ shelves—simply have not been sufficiently studied. We do know that orthopaedic surgeons prescribe far more narcotics than patients use [14]; the opioids for sale on the street come from somewhere. Too often, we just don't see the harms we cause. While a single 10- or 20-mg dose of OxyContin before a shoulder replacement is unlikely to lead to addiction, we believe the more-permissive off-label use of preemptive long-acting narcotics in populations of narcotic-naïve patients has lowered the bar for surgeons to continue to prescribe these drugs after surgery. Some of these patients will continue to use these drugs over the longer term [1], while others will have leftover pills that can fall into the wrong hands [14]. Neither of those outcomes is desirable. We need to remember that the patients who receive our prescriptions are not the ones who engage in the majority of nonmedical narcotic use [7] and that giving patients more opioids than they need means more will be available for experimentation by family members, and for sale on the street. The illicit market for OxyContin tablets varies geographically across the United States, but lately runs about a dollar a milligram [25]. The reasons that surgeons add long-acting narcotics to preemptive regimens in opioid-naïve patient populations or send patients home with more narcotics than they need probably vary. As noted, some may erroneously conflate statistical significance with clinically important effects of treatment. Others may be more focused on increasing the proportion of patients who report being satisfied with their care [30]. Most probably just don't want patients to feel pain unnecessarily. Regardless, we need to remember that these practice patterns cause some patients to become narcotic-dependent [1], and probably facilitate the illegal diversion of narcotics to others who profit from their resale [14]. The American Academy of Orthopaedic Surgeons provides a Pain Relief Toolkit (https://www.aaos.org/Quality/PainReliefToolkit/) that can help surgeons discuss these complex topics with patients. It also has links to instruments that assess patients’ coping strategies, screen for depression, and guide pain-management planning for patients in chronic pain. This page is worth a visit. Beyond that, we need high-quality clinical research, and from that, we need to develop guidelines that identify the analgesic approaches that best keep our patients comfortable while minimizing the risks of addiction, overdose, and narcotic diversion. Clinical Orthopaedics and Related Research® is interested in original studies on this pressing topic. In the meantime, we should write smaller prescriptions for shorter periods of time, think twice before using long-acting opioid medications in narcotic-naïve patients, and help our patients to set realistic expectations about pain after surgery.