Friday, September 16, 2011

Myth-Representations of Opioids & Their Risks

Editor's Notebook After more than a decade of increasing opioid prescribing for chronic pain, the pendulum may be swinging the other way. A recent editorial in a major journal appears to reflect common misunderstandings and negative perspectives on the effectiveness and potential harms of opioids. So, it seems essential at this time to dispel the myths and shift the dialog back toward patient needs and benefits of opioids.

Writing in the September 17, 2011 edition of the Archives of Internal Medicine, Deborah Grady, MD, MPH, and colleagues from the University of California at San Francisco and Los Angeles, discuss “Opioids for Chronic Pain” [Grady et al. 2011]. However, a more apt title might have been, “Why Not to Use Opioids for Chronic Pain,” as they present negatively biased views about risks and harms of these medications. While we do not question that this editorial was well-intended, in the interest of fair balance there are numerous misrepresentations — expressing a false mythology surrounding opioid analgesics — that must be challenged.

The “myth-representations” expressed by these authors in their editorial are of some concern and importance for two reasons:

  1. It is a message going to the broader medical community in a major journal, with a potential for inspiring “opiophobia” that could negatively affect beliefs and opioid-prescribing practices, and

  2. the expressed biases and misinterpretations of research are not new and are typical of similar perspectives on the subject appearing in journals and being presented by speakers at medical conferences.

To begin their editorial, Grady and colleagues acknowledge that chronic pain affects 20% to 40% of adults [and the latest estimate is that 116 million adult Americans suffer chronic pain; see UPDATE here]. Furthermore, they note, 15% to 20% of office visits in the United States include an opioid prescription and 4 million of those are for a long-acting opioid. [In other data, the U.S. CDC (Centers for Disease Control and Prevention) estimates there are 10-12 million Americans on long-term opioid therapy.] The editorialists concede that this preponderance of opioid prescribing naturally evolved from an increasing recognition of pain and the need for its adequate treatment. However, they further state:

“This situation would be acceptable if the benefits of opioid treatment clearly outweighed the risks. Astonishingly, little is known regarding the long-term efficacy of opioid therapy for chronic pain. Of the 25 recommendations included in the Opioid Treatment Guidelines of the American Pain Society [APS] and the American Academy of Pain Medicine [AAPM], none is supported by high-quality evidence.”

We have previously criticized the APS/AAPM “Guideline for the Use of Chronic Opioid Therapy in Chronic Noncancer Pain” [access document here] and its failings in proffering strong recommendations based on weak evidence [see e-Briefing PDF here]. However, the implication in the editorial is that benefits definitely do not outweigh risks, which is a remarkably absolutist assertion in the face of insufficient and/or deficient evidence.

It reflects what is called an argumentum ad ignoratum, or appeal to ignorance, by connoting that opioids incur inordinate risks and are ineffective long-term because they have not been proven otherwise, when, in fact, the high-quality research has not been done to establish the facts one way or the other. Another way of stating this is the old maxim, “the absence of evidence is not evidence of absence” [see Altman and Bland, 1995].

The authors also protest that most efficacy trials of opioids for chronic pain have been of small size, of short duration, have included select groups of patients, and have been placebo-controlled rather than comparisons with other active therapies. They further observe, In general, these trials report that opioid therapy results in approximately a 30% reduction in pain scores. Effects on functional outcomes are more mixed and the benefits more modest.”

Again, the lack of high-quality evidence in this case does not confirm with certainty that opioids have no place in the long-term treatment of chronic pain. And, the assertion that opioids produce only a 30% reduction in pain is an overreaching generalization, as many trials of individual agents have demonstrated far better outcomes for select conditions in terms of both pain relief and functionality. Complaining that comparisons have only been made with placebo is again over-generalizing but, even so, randomized placebo-controlled trials are a high standard of evidence and there are no direct, head-to-head comparisons of a great many drugs used in everyday clinical practice. Why should opioids be held to a different standard?

As for the demonstrated long-term effectiveness and relative safety of opioids, there actually is evidence. As we discussed in an UPDATE [here], Forest Tennant, MD, has assembled an extensive array of documented cases in patients with chronic pain, ranging in age from 30 to 83 years, who have responded well to and thrived on opioid analgesic therapy for from 10 to 35 years. He observed relatively few complications of the therapy, and those were easily managed. The fact that more comprehensive, higher-quality, clinical studies of this nature have not been done — perhaps using extensive data-mining approaches that are often otherwise used to detect the more negative aspects of opioid therapy — may reflect an inherent bias against opioids among researchers and funding agencies.

The editorialists continue…

“…the harms associated with opioid therapy, especially high-dose therapy, are clear. Many patients require ever-increasing doses of opioids to attain the same pain relief. It is unclear whether this need for increased dosage is due to disease progression, development of opioid tolerance, or addiction. Nevertheless, many patients eventually take opioid dosages equivalent to more than 100 mg of morphine per day for many years.”

This argument ignores the basic pharmacology of opioids while proposing an arbitrary 100 mg/day of morphine or equivalent dose ceiling. Needing increasing amounts of analgesia is associated with harm by the authors; whereas, the development of opioid tolerance is a natural biological phenomenon, which may be further influenced by interacting comedications, individual metabolism, disease progression, and other factors. The authors further raise the specter of addiction, which rarely occurs de novo in patients taking prescribed opioids, is a diagnosable condition, and should be clearly detectable by appropriately trained practitioners [see discussion in e-Briefing, p 6, here].

Physiologically, there is no ceiling for opioid analgesic effects (with the exception of opioid agonist/antagonist agents, such as buprenorphine), and some patients with intractable severe pain may require multiple grams per day of morphine or equivalent dose (MED). The editorialists’ suggestion of a maximum safe threshold amount, 100 mg/day MED in their case, has become a widespread and pernicious myth (explained in the text further below).

Grady et al. further state…

“Because federal and state regulations require that these prescriptions be filled on a monthly basis, health care professionals and pharmacists spend an inordinate amount of time dealing with narcotic prescriptions. Also, many patients express chronic anxiety regarding their next refill and expend enormous effort to obtain their medications.”

The implication here is that “narcotic” [an unfortunately biased word choice, see UPDATE here] prescriptions are simply too much bother for busy healthcare providers, which is a rather disappointing perspective. And, an important reason that patients have “chronic anxiety” and obsess about a continued supply of their vital opioid medication might be due to those cavalier attitudes of some prescribers and pharmacies.

The editorialists go on to observe that opioids are associated with “significant risks of addiction, tolerance, opioid-induced abnormal pain sensitivity, constipation, nausea, somnolence, and immune suppression.” Of these, constipation is the only adverse effect that has been consistently demonstrated to occur with opioids, and it can be medically managed relatively easily. As noted above, addiction is a rare occurrence and tolerance is a natural, often beneficial phenomenon. For example, nausea and somnolence are usually transient and resolve as tolerance to properly dosed opioid medication develops. Immune suppressive effects of certain opioids have been demonstrated in vitro, at ultra-high doses, but clinical manifestations of this in humans needs further research.

“Opioid-induced abnormal pain sensitivity,” more commonly described in the pain literature as “opioid-induced hyperalgesia,” or OIH, has been demonstrated in animal models, but as a common and significant adverse effect of opioids in humans it remains a controversial and unresolved topic of ongoing investigation. Unfortunately, presumed OIH is sometimes used as an excuse to discontinue opioid analgesia when a patient requests more of the medication.

Grady and colleagues continue their case against opioids by citing typical facts relating to opioid involvement in substance misuse (“abuse”), overdose, and deaths. They note that prescription “painkillers” [another biased word choice] are the most commonly misused drugs, second only to marijuana, and have accounted for increasing admissions to substance abuse programs in the United States. This, indeed, is of great concern; however, the absolute rates of opioid misuse in the entire population are relatively miniscule compared to the large numbers of patients who compliantly take opioids daily as prescribed and do not misuse or become addicted to these medications [eg, see UPDATE here]. The authors additionally argue…

“Of most concern, opioids are associated with a substantial risk for overdose and death. In 2007, prescription opioid overdose was related to 11,499 deaths, a number greater than that of the combination of deaths from heroin and cocaine. Deaths from opioid overdose have risen steadily since 1990 in parallel with increasing prescription of these drugs.”

It is important to consider that, not only have there been increases in opioid prescribing during the timeframes noted above, but also concurrently increased prescribing of many psychotropic drugs (eg, for anxiety or depression, or as adjunctive pain therapy) along with soaring rates of alcohol and other substance misuse. Data regarding opioid-related overdoses and fatalities have been burdened by unclear forensic definitions of such occurrences and biased studies that have distorted incidence rates.

For example, in many reported cases, prescribed opioids have been present in a lethal cocktail containing other prescribed medications, alcohol, and illicit substances; yet, the incidents are recorded as opioid-related overdose or death, which skews the data toward opioids. In one report from the U.S. CDC, they noted that more than one drug was listed in 72.3% of death certificates of alleged opioid-related fatalities, including benzodiazepines (in 21%), antidepressants (32%), and/or illicit drugs (22%) [MMWR 2009].

The editorialists, themselves, concede that persons with depression [possibly receiving medications for this] and concurrent users of benzodiazepines experience increased risks of opioid overdose. Furthermore, they claim…

“Risk of overdose and death increases with higher doses of opioids. In 1 study, patients taking more than 100 mg per day of morphine-equivalent opioids displayed a 9-fold increased risk of overdose compared with that among patients taking less than 20 mg; the risk was approximately 2% per year among patients using high-dose opioids.”

Their referenced source for this statement is a study by Dunn and colleagues, published in 2010, and this misleading report [discussed in a Pain-Topics UPDATE here] has served as the basis for the 100 mg MED safety threshold myth adopted without critical assessment by many other authors and conference speakers. In this retrospective data-mining research, spanning 9 years and including 9,940 patients receiving long-term opioid analgesics for chronic noncancer pain, there were only 51 opioid-related overdoses (which included 6 deaths), or a 0.057% yearly incidence rate. Those receiving ≥100 mg/d MED appeared to have a greater overdose risk but they comprised only 22% of all overdose cases; so, the greatest frequency of events in this population involved doses below this level.

Further complicating the Dunn et al. data, roughly three-quarters of patients (74.4%) had been prescribed sedative hypnotics, including benzodiazepines. Eight cases of overdose were due to accidental excess ingestion of opioids, 6 involved suicide attempts, and 11 involved drug misuse or illicit opioids. So attributions of cause, and particularly implications of higher-dose Rx-opioids as most culpable, are highly questionable — yet, this myth has endured.

Some additional evidence, overlooked by Grady et al., should be mentioned…

  • A recently-reported study — published in JAMA, April 6, 2011, by Bohnert et al. [discussed in an UPDATE here] — compared retrospectively culled records of 154,684 patients (32% of whom were prescribed long-term opioids for chronic noncancer pain during a 5-year period) compared with 750 cases of unintentional Rx-opioid overdose deaths in the population (a total fatality rate calculated by the researchers as 0.04%). The overdose rate was greater for opioid doses ≥100 mg/d MED as compared with lower doses (1 to <20 mg/d); however, most deaths (43.5%) were inexplicably in persons listed as receiving 0 (zero) mg/d and/or patients with substance-use disorders. As with the Dunn et al. study, there were many potentially confounding factors unaccounted for, including the presence of co-medications or illicit drugs, polysubstance misuse, and co-occurring medical conditions.

  • Another recently-published report, by Gomes et al. [2011] in Canada, retrospectively examined more than 600,000 persons prescribed opioid analgesics for chronic noncancer pain during a 10 year period and discovered 593 associated fatalities, or an annualized death rate of only 0.01%. They found a greater risk with >200 mg/d MED than in persons taking <20 mg/d MED; however, among the many confounding factors in this research that biased outcomes, benzodiazepines also were present in 85% of all decedents taking opioids. [See UPDATE here.] 

  • It is interesting to note that the APS/AAPM Opioid Guidelines (noted above) state: “By panel consensus, a reasonable definition for high dose opioid therapy is >200 mg daily of oral morphine (or equivalent), based on maximum opioid doses studied in randomized trials and average opioid doses observed in observational studies. Some studies suggest that hyperalgesia, neuroendocrinologic dysfunction, and possibly immunosuppression may be more likely at higher opioid doses, though more evidence is needed to define these risks, their relationship to dose, and their relationship to clinical outcomes [emphasis added].”

    It is important to consider that the >200 mg/d MED was a “best guess” of the panel based on very limited evidence and that the purported adverse effects at this “high dose,” as well as their clinical significance, are admittedly still undetermined.

  • Of further interest, new opioid prescribing rules soon going into effect in Washington state adopted a 120 mg/d MED benchmark ceiling of opioid prescribing for chronic noncancer pain, and higher doses would require special considerations or actions by the healthcare providers prior to continuing therapy [see UPDATE here]. Other than a reference [in PDF document here] to the poor-quality study by Dunn et al. (noted above and specifying 100 mg/d MED), the evidence base in support of this 120 mg/d MED number is unclear and the specific number seems to have been set rather arbitrarily. Curiously, state authorities also concede, “There is no clearly defined ‘safe’ opioid dose. The guideline does not say that a dose above 120 mg/day of morphine equivalents is necessarily unsafe, nor that doses below this are guaranteed to be safe” [emphasis added, also see Washington state guideline document PDF here].

It seems intuitively pragmatic that as doses of drugs like opioids are increased there may be correspondingly greater risks of overdose and drug-related death. However, the research evidence to date suggests that the picture is much more complex and doses ≥100 mg/d MED, as suggested in the Grady et al. editorial, are probably only a secondary factor at most. And, as Bohnert et al. conclude in their study, the estimated overall risk of opioid overdose among individuals properly treated with opioids at any dose is quite small and opioid-related overdose deaths are an important but rare outcome.

In their editorial, Grady and colleagues assert, “Compared with shorter-acting preparations, long-acting opioids such as sustained-release morphine, sustained-release oxycodone, and methadone also increase risk of overdose.” This, too, seems commonsensical; however, in the study by Dunn et al. that the authors reference, fewer than 1 in 10 (9.6%) of all subjects had been prescribed long-acting opioids and there was no indication that these formulations played a greater role in overdose incidents than shorter-acting agents.

The authors comment on attractive features of methadone for chronic pain, but that it is “more dangerous than other drugs because it has complicated and unpredictable pharmacokinetics and pharmacodynamics and is associated with corrected QT-interval prolongation and cardiac arrhythmias.” Exaggerations of methadone’s alleged cardiotoxicity have gained much traction in the pain management and addiction treatment communities, but this myth is largely based on very limited data, along with biased and questionable interpretations of poor-quality research evidence [discussed in prior UPDATES here and here].

What actions do Grady and colleagues recommend in their editorial?

First, they suggest that the trend to prescribe opioids for patients with chronic pain should be reversed. Healthcare providers should first try to treat the underlying condition using nondrug modalities such as “physical therapy, cognitive behavioral therapy, pain management techniques, and appropriate assistive devices.” While this seems reasonable, many cases of chronic noncancer pain do not respond to such therapies and many patients exhibit no readily diagnosed underlying etiology or pathology. Although nondrug therapies can be important, research evidence for their possible effectiveness in chronic pain is still emerging.

The editorialists further recommend, First-line drug therapy for chronic pain should include aspirin, acetaminophen, nonsteroidal anti-inflammatory drugs, and appropriate corticosteroid injections.” However, this completely overlooks the fact that these agents have toxicities and risks that in many patients are greater than those of even the strongest opioid analgesics [discussed in an UPDATEs series here].

Grady and colleagues go on to state…

“Health care professionals should evaluate pain and functional status at each visit. Patients who had no improvement in pain or function should not receive higher doses of opioids, and those taking dosages as high as 100 mg of morphine or an equivalent drug per day should not have their dose increased further.”

The authors again invoke the mythical 100 mg/d MED ceiling, without considering that a reason some patients may not realize improvements in pain or function is due to insufficient analgesia in the first place. Instead, they recommend that healthcare professionals should consider tapering or stopping opioids. Would practitioners similarly limit or discontinue insulin, rather than increase the dose, in patients who fail to achieve an arbitrary level of glucose control?

Another option, rather than increasing the opioid dose, the editorialists say, is referral to a pain specialist with a focus on treatments such as nerve blocks, corticosteroid injections, and behavioral therapy. However, there is probably even less good research evidence to support the long-term benefits of these interventional or cognitive therapies than there is for ongoing opioid therapy, at any dose. Thus, it seems clear by this point that Grady and colleagues prefer any option, no matter how questionable or risky, rather than opioids.

Finally, to synopsize their concluding statements…

“Many physicians will tell you they would be happy to stop prescribing opioids if a better option existed for treating chronic pain.  ...[but] the lack of alternatives does not justify treatment with agents for which the risks outweigh the benefits. However painful it is to explain to patients who are experiencing chronic pain that we do not have effective and safe treatments for their pain, we should be willing to do so when appropriate.  . . . .  Health care professionals need to feel empowered to say no to the unnecessary, often risky, ever-increasing use of opioids without worrying that their judgment could be challenged.  . . . .  …we should not continue to prescribe high-dose opioids with little evidence of long-term benefit and clear evidence of substantial harm.”

These assertions from the authors, encapsulating the overall bias of this editorial, leave both practitioners and their patients with pain in a predicament. In our opinion, evidence of “substantial harm” from opioids, at any dose when properly prescribed for chronic noncancer pain, is far from clear. What does seem clear is that, just as their physicians might feel about prescribing opioids, almost all patients would happily forego opioids and accept alternative therapies for their pain — IF those alternatives were as effective as opioids.

Many other discussions of opioid harms and recommended solutions, coming from many directions, have echoed the perspectives in this editorial by Grady et al. Often, the stated objectives are something like, “Reduce abuse and overdose of prescription opioids while ensuring patients with pain are safely and effectively treated and have access to the drugs that they need.” While this seems noble and appropriate, the statement also might reflect a myth of beneficence: that is, it appears to have the best interests of patients in mind while, at the same time, encouraging actions or endorsing rules/regulations that are questionable in terms of their actual benefit to patients.

Meanwhile, healthcare providers have legitimate concerns about new rules and regulatory-agency scrutiny of their opioid prescribing practices, and of being duped by persons seeking opioids for dishonest and/or unlawful purposes. A certain amount of “opiophobia” on the parts of practitioners is understandable; especially, considering the lack of a strong, extensive base of high-quality research evidence and validated opioid-prescribing guidelines. At the same time, it is important to understand that decisions regarding opioid analgesics are being increasingly driven by demagoguery, and fear, and the myth-representations all too often found in communications like the editorial critiqued above.

Our own bias is that it is time to shift the dialog regarding opioid analgesics. While these medications may not be curative of a chronic pain condition, they can be a means to an end in providing essential relief from pain so that a patient might pursue additional therapeutic modalities to help promote enduring functionality and facilitate a better quality of life — albeit, a life that may not always be completely pain free. For many patients, opioids are a temporary measure — at whatever dose and period of time that are necessary — but, for some, these drugs may be a vital and relatively safe lifelong therapy for maintaining a more normal, comfortable existence; much like insulin, antihypertensives, antidepressants, or other medications are necessary in some patients. Rarely, if ever, would patients choose opioids (or any medication) over nonpharmacologic approaches, if the latter were readily accessible, affordable, and could provide comparable, lasting relief from pain and suffering.

We agree with others* that the greatest problem facing healthcare providers and society today is not too much pain medication or its misuse by a small minority of individuals; rather, the problem is too little pain relief for the millions of patients in need. Readers are invited to add their own opinions in comments below.

*Thank you to Lynn Webster, MD, for inspiring this thought.

DISCLOSURE: Pain Treatment Topics is currently supported by medical education grants from Purdue Pharma L.P. and Endo Pharmaceuticals, manufacturers of opioid analgesics. However, these organizations had absolutely no role in the inception, development, review, or approval of this UPDATE. The opinions expressed are those of the author.

REFERENCES (not otherwise linked within the text):
> Altman DG, Bland JM. Absence of evidence is not evidence of absence. BMJ. 1995;311(7003):485 [
PDF here].
> Gomes T, Mamdani MM, Dhalla IA, et al. Opioid Dose and Drug-Related Mortality in Patients With Nonmalignant Pain. Arch Intern Med. 2011;171(7):686-691 [
abstract here].
> Grady D, Berkowitz S, Katz MH. Opioids for Chronic Pain. Arch Intern Med. 2011;171:1426-1427 [
access here].
> MMWR [2009;58:1171-1175]. Reported in: Overdose death involving prescription opioids among Medicaid enrollees — Washington, 2004-2007. JAMA. 2010;303(1);21-22.