Problematic naloxone guidelines

This is going to be my first attempt at writing an essay concerning a largely pharmacy-specific issue. I’m trying hard to write it generally enough to allow more widespread readers to understand what I’m criticizing, but it’s a delicate balance to walk in this regard because I don’t want to make my argument irrelevant or unconvincing to healthcare professionals in the process. I’ll do my best to relegate most of the heavy analysis of evidence to the notes, and hopefully I won’t scare everyone away by the end.

The first Canadian national consensus guidelines for the dispensing of take home naloxone (THN) by pharmacists were published by Canadian Pharmacists Journal at the end of last year.¹ I was informed by a representative of Emergent BioSolutions, the makers of Narcan (naloxone) nasal spray around the time of publication, but I didn’t find the time to read them in detail until more recently. When I first heard of them, I was surprised by the recommendation that “All patients receiving an opioid should be dispensed take home naloxone and counselled by a pharmacist,” going against typical patient selection guidelines targeting patients at high risk for opioid-induced respiratory depression (OIRD), as was the focus of the accompanying Practice Brief, in the same issue, written by almost all of the same authors.^2,i Upon finishing multiple readings of the guidelines and referenced materials, some of my suspicions were confirmed: Justification is provided for this recommendation, but it isn’t backed up by good evidence,ⁱⁱ and I wanted to shine a bit of light on this fact here.

The main point of justification that THN be provided to all patients receiving opioid scripts is that “risk factor information is not easy to acquire,”¹ and I worry that this was based on assumptions made regarding reports on the causes of opioid-related deaths. The Practice Brief identifies four important factors putting patients at high risk of OIRD: patients with a history of OIRD, opioid-related substance use disorder, high-dosage opioid prescription, and taking a benzodiazepine (such as Xanax or Valium) concomitantly with an opioid.² The authors of the guidelines suggest that “the stigma related to substance use disorders may prevent people from disclosing this information accurately,”¹ but in the absence of evidence in support of this claim this becomes open to a number of points of criticism. For one, developing a healthy pharmacist-patient relationship should allow for more honest and open dialogue. In this regard, I’ve had many patients disclose things to me such as illicit drug use and impulsively taking their prescribed medications at doses higher than prescribed after they got to know me and felt safe doing so, though I understand that there’s potential that a large number of patients withhold this information from me. However, I don’t know the proportion of patients who disclose this information to me compared to those who don’t because this evidence doesn’t exist, as far as I know. And, along the lines of most of what I will be arguing in this essay, the guideline authors imply conclusions from such inexistent evidence.

As a point that’s potentially more important than the relationship with patients, I’d suggest that many of the risk factors are readily identifiable. The doses of opioids can be calculated and concomitant CNS depressants should be flagged by dispensing software. The authors identify patients using multiple pharmacies as a barrier, but many jurisdictions have systems that monitor controlled medications––including the Narcotic Monitoring System in Ontario and PharmaNet in British Columbia––which should aid practitioners in gathering this kind of information. And anyway I question what percentage of patients with opioid prescriptions engage in this kind of behaviour; some, certainly, but is it a high enough proportion to justify providing THN to everyone just in case? I don’t exactly know the answer to that, just as the authors don’t provide an answer, though they likely assume it’s at a higher proportion than I do.

The risk factor of opioid-use disorder (OUD) is another one that I’d suggest is more readily identifiable than the authors imply. Part of the problem I see is the potentially misleading nature of the statistics given in the point: “About one-quarter to one-third of opioid-related deaths in Ontario involve prescription opioids. In addition, approximately 80% of people assessed at an emergency department for opioid toxicity had received an opioid prescription in the previous 3 years.”¹ To me, this comes across implying that a large percentage of general opioid prescriptions for conditions separate from OUD, like pain, are associated with OIRD and opioid-related deaths, though the references used don’t demonstrate this.^iii,iv,v And studies in other jurisdictions showed a much lower percentage of patients experiencing OIRD linked only to prescribed medications.^vi

But all this likely matters little if we accept the authors’ assertion that “The benefit of having naloxone kits available in cases of emergency outweighs the drawbacks of dispensing those that end up not being used.”¹ The biggest point of criticism I have for this has to do with the appropriate allocation of healthcare dollars. Can the recommendation that THN be provided to all patients with opioid prescriptions truly be justified if the expected health benefit is unknown and may not be more than marginally better than appropriately screening and actively providing THN to those identified as being at high risk for OIRD, especially if the only marginal improvement is likely to be associated with a substantially higher cost? I suspect there are those who will take the view that a marginal benefit still means lives are saved, so it’s worth the cost, but it’s not that simple. In healthcare, allocation of money to ineffective services also costs lives, as funds need to be taken away from other necessary, life-saving, evidence-based services as a necessity. And pushing for a potentially wasteful practice environment could harm a THN program more generally. If decision makers see that the program is a huge drain on healthcare dollars without delivering substantial benefit, cuts can be justified. This can ultimately take funds away from even an evidence-based program targeting high risk individuals, which we can with reasonable confidence say will cost lives.

And keep in mind that Emergent BioSolutions provided funding through an unrestricted grant for the making of these guidelines, they stand to profit enormously from the recommendations, and this cost is effectively hidden from a wide section of the public because patients don’t have pay out of pocket for THN in a number of Canadian jurisdictions, including in the most populous province. While I fully agree with the need for pharmacist participation and intervention in addressing the opioid epidemic and helping patients thoughtfully and without worsening the stigma commonly attached to OUD and OIRD, accepting guidelines that have a reasonable potential for bias sets a dangerous precedent to the practice environment. I worry that the recommendations promote thoughtlessness in our practice, at the support of business-centric practice that hides behind a patient-centred ideology, keeping in mind that pharmacies also stand to profit from dispensing THN.

For pharmacists who stuck with me this far, the big takeaway I want to impart on you is that it’s important to question if it is in fact appropriate to actively provide THN to all patients with opioid scripts––especially for those with short-term scripts for acute conditions and lacking risk factors for OIRD––and that we owe it to our patients to do so if we truly want to provide ethical care. As for other readers, I don’t want you to come away from this feeling that you shouldn’t trust your healthcare providers, but I do want you to understand that providers with the reputation of upholding evidence-based care will promote things that are nothing of the sort, and I think it’s reasonable for you to expect good justification for their therapeutic decisions.

Notes

i. The most recent Canadian Pharmacists Journal also contained both an editorial and practice tool—both written in part by authors of the guidelines—that reinforce the main recommendation of the guidelines. In each, the focus was more heavily directed to reducing stigma associated with naloxone, which was only briefly discussed in the original guidelines.^3,4

ii. Studies cited in the guidelines in favour of the effectiveness of THN all involve programs targeting individuals at high risk of OIRD or those close to high risk individuals, such as friends or family.^5-7 The only cost-effectiveness study cited involved high risk individuals.⁸

iii. The one-quarter proportion apparently comes from a report put out by the Ontario Drug Policy Research Network.⁹ (I say “apparently” here because A PLOS ONE article was actually cited in the guidelines,¹⁰ but the only thing I could find in the article that was relevant to the stats cited in the guidelines was the single statement that referenced the Ontario Drug Policy Research Network report.) In it, it was found that 25.9% of accidental opioid-related deaths were from pharmaceutical (prescribed) origin. Included in this definition are both opioid agonist therapy (including methadone and buprenorphine) for OUD and prescribed opioids that were diverted and ingested by someone other than their intended recipient, though the study doesn’t break this figure down further to show the proportions of each. This means that some unknown percentage of the 25.9% would be readily identifiable as high-risk by nature of an obvious indication of therapy, and it also means that we would not be able to prevent opioid-related death in some other unknown percentage because we would not be providing THN to the eventual recipient––unless we are to assume those diverting their doses would be kind enough to also divert the naloxone.

iv. Both the one-third proportion and the 80% statistic come from a British Medical Journal study,¹¹ which found that a range of between 32.5 to 38.2% of people who died from an opioid-related cause had an active opioid script at the time of death, varying by year analyzed. The type of prescribed opioids were partially reported when only the prescribed opioid was present in the toxicology report at the time of death. Looking at the statistics reported, this means that there’s a largely unknown range of patients who were prescribed opioids for OUD: somewhere between 7 and 64% of those who had active opioid scripts at the time of death. (Methadone was reported in between 7 and 17% of cases where only the prescribed opioid was present in the toxicology report. Buprenorphine was included in the “other” group, at ≤5% across all years. The prescribed opioid was not reported in those in which non-prescribed opioids were found, with the range of 36 to 42%, with an unknown amount of methadone and buprenorphine prescribed for these individuals. OUD was reported as the indication for methadone therapy in >97% of cases where it was prescribed.) In addition, of those who had an active prescription for an opioid at the time of death, between 27.6 and 33.3% also had an active prescription for a benzodiazepine. This means that a hugely unknown number of these patients could be readily identified as being high risk for OIRD, and it really doesn’t tell us the true potential need we have for intervening for every single patient rather than only intervening with those at high risk for OIRD.

v. As for the fact that 80% of those assessed in emergency departments received an opioid prescription within the previous 3 years,¹¹ I think it’s important to clearly state a couple of things. This doesn’t mean that a large percentage of people who receive opioid prescriptions experience OIRD. However, it does mean that a large proportion of those who eventually experience OIRD come into contact with the healthcare system, and it represents an important opportunity for assessment or intervention.

vi. As an example, in a B.C. study referenced in the guidelines in support of the effectiveness of THN programs,⁷ deaths caused only by opioids prescribed to the deceased individuals, as identified in the toxicology reports, accounted for less than 5% of cases, and were excluded from the analysis. Keep in mind also that the indications, doses, and concomitant non-opioids were not reported for this group, and so an unknown number of individuals within that 5% potentially had readily identifiable risk factors for OIRD.

References

1. Tsuyuki R, Arora V, Barnes M, et al. Canadian national consensus guidelines for naloxone prescribing by pharmacists. Can Pharm J (Ott) 2020;153(6)347-51.
2. So R, Al Hamarneh Y, Barnes M, et al. The status of naloxone in community pharmacies across Canada. Can Pharm J (Ott) 2020;153(6)352-56.
3. Beazely MA and Tsuyuki R. The opioid crisis: Naloxone and pharmacists to the rescue (editorial). Can Pharm J (Ott) 2021;154(5)289-90.
4. Cid A, Patten A, Grindrod K, et al. Frequently asked questions about naloxone: Part 1. Can Pharm J (Ott) 2021;154(5)301-4.
5. Walley AY, Xuan Z, Hackman HH, et al. Opioid overdose rates and implementation of overdose education and nasal naloxone distribution in Massachusetts: interrputed time series analysis. BMJ. 2013;346:f174.
6. McDonald R, Strang J. Are take-home naloxone programmes effective? Systematic review utilizing application of the Bradford Hill criteria. Addiction. 2016;111(7):1177-87.
7. Young S, Williams S, Otterstatter M, et al. Lessons learned from ramping up a Canadian take home naloxone programme during a public health emergency: a mixed-methods study. BMJ Open. 2019;9(10):e030046.
8. Langham S, Wright A, Kenworthy J, et al. Cost-effectiveness of take-home naloxone for the prevention of overdose fatalities among heroin users in the United Kingdom. Value Health. 2018;21(4):407-54.
9. Ontario Agency for Health Protection and Promotion (Public Health Ontario); Office of the Chief Coroner; Ontario Forensic Pathology Service; Ontario Drug Policy Research Network. Opioid mortality surveillance report: analysis of opioid-related deaths in Ontario July 2017-June 2018. Toronto, ON: Queen’s Printer for Ontario; 2019.
10. Choremis B, Campbell T, Tadrous M, et al. The uptake of the pharmacy-dispensed naloxone kit program in Ontario: a population-based study. PloS One. 2019;14(10):e0223589-e.
11. Gomes T, Khuu W, Martins D, et al. Contributions of prescribed and non-prescribed opioids to opioid related deaths: population based cohort study in Ontario, Canada. BMJ. 2018;362:k3207.