In 1996, former Federal Reserve Board chair Alan Greenspan warned that “irrational exuberance” may have inflated stock asset values to unsustainable heights. Greenspan was concerned with economics, not health care, but he would surely see the parallels with some recent medical enthusiasms. The rise in prescription opioid use is one example. According to the Centers for Disease Control (CDC), on an average day in 2014, more than 650,000 opioid prescriptions were dispensed, 3900 people initiated nonmedical use of prescription opioids, and 78 people died from an opioid-related overdose (https://www.hhs.gov/sites/default/files/Factsheet-opioids-061516.pdf). Though 2015 data show a decline in total opioid prescriptions for the first time in years, the problem continues to command widespread attention.

Primary care physicians are admittedly part of the problem. We failed to question for too long the discredited notion that there is no upper limit to the safe prescribing of opioids for chronic non-cancer pain (CNCP); we casually accepted the mantra of pain as the fifth vital sign; we continued to refill opioid prescriptions for patients with CNCP without carefully re-evaluating indications for treatment. But in our defense, the initial opioid prescriber is often a dentist, an orthopedist, or an emergency room doctor.1 A disproportionate number of risky prescriptions are issued by a small number of practitioners.2 Pharmacologic alternatives to opioids like acetaminophen and NSAIDs are themselves not particularly effective.3 And many of the most promising modalities for addressing chronic pain (cognitive behavioral therapy, physical therapy, complementary modalities, mindfulness, and more) are either out of reach or unacceptable to many of our patients. Perhaps most importantly, dependence on prescription opioids is often a symptom of deep despair arising from economic insecurity; lack of jobs for non-college graduates that do not involve some form of physical labor; social isolation; community fragmentation; and lack of purpose. Weaning patients off opioids will reduce overdoses. But unless we find ways to strengthen families and communities, despair will just seek other outlets.

In this issue of JGIM, Nugent et al.4 report on yet another way we fail our opioid-taking patients. From a sample of 600 VA patients, they identified 169 whose opioid prescriptions were discontinued due to a urine drug test that was positive for alcohol, cannabis, or other illicit or non-prescribed controlled substance. Of these, 73 patients (43%) were referred for substance use disorder treatment, and 34 patients (20%) actually made it into treatment. These sobering results were obtained within a system where substance use disorder treatment is widely available; outside of a well-integrated system of care, the results would assuredly be worse. In addition to highlighting the problem of inadequate support for patients unable to safely take long-term opioids, this study is a cautionary tale for guidelines developers. Physicians are good at ordering tests, and if the federal government says to obtain urine from prescription opioid recipients, most doctors will comply. However, unless systems are in place to help physicians and patients deal with the results, the potential for benefit is vastly diminished.

The problem of testing with inadequate follow-up is closely paralleled by the problem of excessive follow-up testing, a problem highlighted in the article by White VanGompel et al.5 In this study, the authors examined records of almost 6000 women who received initial osteoporosis screening with dual-energy x-ray absorptiometry (DXA). Federal guidelines recommend that women whose initial screen is normal or low-risk (no more than mild osteopenia) should not be screened again for at least 2 years (and possibly more, based on data suggesting a risk of progression to osteoporosis among these women of <10% in 15 years). Of the 3564 women in this category, 8% were screened within 2 years and 43% within 5 years. Smokers and African-Americans were less likely than the average patient to be re-screened, whereas women making frequent primary care or specialty visits were more likely to be re-screened. These findings raise several questions. Who is driving this arguably unnecessary utilization, the physician or the patient? What is the role of continuity or discontinuity of care? Can electronic health records be enlisted to caution against overuse, just as they currently issue reminders to order needed screenings?

General internists are not solely responsible for the opioid epidemic, nor for the broader problem of health care services overuse. But we have played our part. The good news is that we can also be part of the solution. We can advocate for more broadly available addiction services. We can push back against guidelines unsupported by high-quality evidence (like mandatory urine drug screens), while fighting for better systems of addiction care. We can master techniques for gently demurring on patient’s medically unindicated requests. And we can be resolute in our demand that science, evidence, and reason are enlisted to keep irrational exuberance at bay. Even when the exuberance is our own.