The New “Opioid Crisis”: Scientific Bias,Media Attention,and Potential Harms for Patients with Refractory Dyspnea

Opioids have the potential to cause harm when misused. In 2015 alone, 33,091 Americans died from an opioid overdose, ¹ and the opioid crisis shows no signs of abating. But as we struggle to prevent these tragic deaths and reduce the misuse of opioids, we need to be careful not to stigmatize opioids themselves. People have long associated opioids with death and unpleasant side effects, and opioid phobia remains a substantial barrier to symptom control among patients with end-stage illness. The scientific community has a responsibility to correct myths and misperceptions surrounding opioids, and to apply rigorous scrutiny to any study that claims to show prescribing of opioids is harmful.

Two recent studies of opioids in patients with chronic obstructive pulmonary disease (COPD) are worthy of such scrutiny. In the first, Vozoris et al. reported the results of a large population database-derived cohort study showing an association between new opioid prescriptions and mortality in patients with COPD. ² This finding is not surprising; opioid use is associated with mortality in many populations, ³ and the incidence of opioid and sedative use increases as patients approach death. ⁴ But the obvious explanation for this association is not causation but rather confounding by indication. COPD patients who have more severe illness are likely to receive more intensive treatments and, when comparing the interventions, the more intensive intervention will be associated with poorer outcomes. Dyspnea is a strong predictor of mortality ⁵ and proximity to death as confirmed by Campbell et al. in this edition of the Journal of Palliative Medicine, and because dyspnea is not captured in any health administrative database, it cannot be adjusted for in the analysis. Incompletely controlled “residual” confounding may persist in clinical investigations despite study design and statistical procedures aimed at eliminating this form of bias. The initiation of opioids is an appropriate response to refractory dyspnea in COPD patients with advanced disease^6,7 and is, therefore, a marker but not a cause of mortality.

Confounding by indication is well known to occur in observational pharmacoepidemiologic studies, ⁸ and there are published examples of beneficial therapies that lower mortality in a randomized-controlled trial while being associated with higher mortality in observational studies. ⁸ Nevertheless, Dr. Vozoris' study received national media attention in an article entitled “Opioids prescribed for COPD may harm patients: study”. ⁹ In the article, Dr. Vozoris warned that “Physicians and patients need to be a lot more careful and judicious about [the use of opioids].…” The Canadian national newspaper article mentioned the possibility of confounding only at the very end, and made no mention of the literature showing the safety and efficacy of opioids for dyspnea in patients with COPD,^7,10,11 or the fact that the Canadian Thoracic Society recommends the use of opioids for symptomatic treatment of end-stage COPD. ⁶ The authors of the related article ² also failed to acknowledge the substantial and increasing burden of dyspnea in patients with COPD as their disease becomes more severe or they approach end of life, or to mention the tendency for patients to under-report symptoms and for clinicians to undertreat this problem as Campbell et al. confirm in an article in this edition of the Journal of Palliative Medicine. We should acknowledge that a subsequent publication from Doctor Vozoris was more balanced ¹² but the damage had been done.

We might be tempted to lay blame with the media, always keen to find a good story to attract readers. But in this case we cannot, because the interpretation appeared to come from the authors rather than the journalist. More recently, Vozoris et al. published another observational study of opioid use in COPD, entitled “Adverse cardiac events associated with incident opioid drug use among older adults with COPD.” ¹³ Based on the title, one might expect this study to have shown a higher rate of cardiac events among patients who received opioids compared with those who did not. In fact, the data showed precisely the opposite—among community-dwelling COPD patients (90% of the cohort), new opioid users had a significantly lower rate of hospital admissions for congestive heart failure, with no increase in mortality from any cardiac cause. The higher rates of cardiac events were seen only among the subgroup of long-term care residents (a mere 10% of the cohort) and among the subgroup of patients who were prescribed opioids that were not combined with a nonopioid agent (again, 10% of the cohort). In other words, the title reflected the findings from small subgroups that would certainly have disappeared in the overall cohort, and contradicted the most notable finding from the entire cohort studied. We can only speculate whether this was a reflection of previously held biases, but the authors clearly failed to place epidemiological findings responsibly within an appropriate clinical context. To compound the problem, the results were interpreted with reference (among others) to two publications about the mechanisms of harm in opium addiction,^14,15 without any balanced review of how opioids safely relieve dyspnea in patients with COPD.^16–18

It is good medical practice to provide effective palliation for dyspnea in advanced COPD, and effective palliation for refractory dyspnea includes careful initiation and titration of opioids. Palliative measures can improve symptoms and health-related quality of life, ⁷ avoid unnecessary hospital admissions, and result in cost savings.^19,20 Dr. Currow, a member of the editorial board of this journal, has pitched the palliation of dyspnea as a human right. ²¹ Clinicians who treat patients with advanced COPD should not be put off by the findings and interpretations of the data published by Dr. Vozoris et al. and can take some solace from the findings from a large Swedish epidemiological study. This study attested to the lack of a mortality association with opioids in oxygen-dependent patients taking the morphine equivalent of <30 mg per day. ²² For those looking for insights beyond who takes an opioid to how opioids relieve dyspnea, a recent report of effects of morphine on exercise within a small RCT ²³ provides an updated and comprehensive extension of our understanding of mechanisms beyond the small studies that have informed the most recent systematic review on the topic. ²⁴ As stated in an accompanying editorial, ²⁵ we need to balance all the evidence and “cannot afford to ignore the patient and caregivers' experiences, and that requires further rethinking of how best to conduct research, especially in the field of palliation for multidimensional symptoms such as dyspnea”. ²⁵ Administrative databases cannot provide those experiential nuances and we should be wary of their influence on decisions regarding palliation for an individual patient in the appropriate clinical context.

There are overlapping approaches to the complexities of management of refractory dyspnea and we must remain cognizant of proven nonpharmacological interventions such as the appropriate use of oxygen, handheld fans, and supportive interdisciplinary programs.^26,27 Nevertheless, there is a time and a place for opioid therapy when other measures fail to bring sufficient relief.

Confirmation bias is a well-recognized phenomenon in both research and clinical practice. Certainly, we have all been guilty from time to time of looking diligently for data that validate our “hunches,” while overlooking disconfirming data. In this case, the harms of failing to correct these shortcomings ²⁸ are exacerbated by a historical predisposition toward fear of opioids, and a society currently primed by the epidemic of recreational opioid overdoses fueled by illicitly manufactured and distributed opioids. ¹ Already, there are reports that increased medicolegal concerns over opioid prescription have made physicians more hesitant to use them for any indication, ²⁹ prompting Canada's Palliative Care community to issue a warning about such unintended consequences. ³⁰ It has taken years for our professional societies to embrace the concepts of effective palliation for advanced COPD^31,32 and we now need to focus efforts and education to correct misperceptions and biases created by authors and media alike. We are reminded of a quote attributed to Mark Twain: “A lie can travel half way around the world while the truth is putting on its shoes.”