© The Author(s) 2022
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/17151635221139829
Pharmacists seeking to improve their provision of care are likely to be faced with a myriad of recommendations on further education, new services to provide, optimizations of workflow, and so on. However, given limited time and resources, it can be overwhelming to decide which efforts are the most worthwhile. It is within this context that this commentary seeks to propose a simple change to communication with patients that is justified by both empirical research and ethical considerations, namely quantification-based counselling.
Before diving into the practicalities, it is first necessary to provide a brief background on medical ethics. Medical ethics has 4 foundational principles: respect for autonomy, beneficence, nonmaleficence and justice.1 Among these, respect for autonomy is the primary principle guiding medical ethics.2 Respect for autonomy implies that ethical treatment decisions align with patient values and preferences. As such, informed and voluntary consent is a key piece of the provision of ethical care. This not only has a firm ethical justification but also helps to establish trusting relationships with patients.3
Unfortunately, there are gaps between what patients expect in terms of respect for their autonomy and what is provided.4,5 While most of this research takes place in other countries, there is evidence that Canadian patients receive unsatisfactory care in this regard as well.6 Given the importance of respect for autonomy within our ethical framework, it is essential for health care workers to strive to improve practice in this area.
There are many ways that pharmacists can contribute to this objective depending on their setting and scope of practice (e.g., pharmacists in primary care clinics implementing the use of decision aids). However, our focus here is on how pharmacists across direct patient care roles can contribute to this goal. One way to achieve this is by ensuring informed consent during medication counselling through better providing patients with information regarding the benefits and harms of their medications in a way that they broadly understand.
When counselling on a medication, the norm of practice is typically to describe the medication’s benefits and harms in a qualitative manner. For example, a patient prescribed a new beta-blocker may be told that it is common for it to cause fatigue, or it may be said that an allergic reaction to a vaccine is very rare. These qualitative descriptions (“common†and “very rareâ€) are often assumed to adequately inform a patient.
Unfortunately, qualitative descriptions are prone to misunderstanding. While the word “common†has no set meaning in usual conversations, it does have a European Union–defined frequency of 1% to 10%.7 Assuming that the pharmacist had roughly this value in mind when counselling, this could lead to a major misunderstanding. This is because, per the results of a study on patient interpretation of qualitative frequencies, on average patients thought that “common†meant a 45% chance of occurring.7 This is problematic, since informed consent requires someone to receive enough information to make a choice as to whether to proceed with treatment. In this case, assuming the provider had in mind the 1% to 10% adverse effect range, then the patient would be drastically overestimating the chance of experiencing fatigue with a beta-blocker (by at least 35%, assuming they had the average interpretation). While there is some leeway (patients don’t need to know if the risk of a side effect is exactly 8.4%), this degree of misunderstanding is enough to hamper their ability to make an informed choice about whether they want the medication. Similar overestimates were also seen with other qualitative descriptions; for example, “very rare†is classified by the European Union as <0.01% but was interpreted to mean 4% (what we would classify as “commonâ€), an at least 400 times overestimation.7 Unfortunately, the solution is not as simple as changing which qualifier we use (e.g., very very rare), as there was also great variability between patients in their interpretations.7
Fortunately, pharmacists have excellent potential to provide value in this area by making quantitative language a norm of practice when counselling on common medications. Instead of using problematic qualitative descriptions, there is compelling evidence that the use of quantitative descriptors provides a better understanding of risk.8 Returning to our example, this could mean informing the patient there was an approximately 3% chance of fatigue with a beta-blocker.9 Not only is this more likely to accurately convey the information (facilitating informed consent), it has the added benefit of increasing patient satisfaction with the information received and the likelihood of medication use.8 The increased rate of medication use with quantitative descriptions is unsurprising; since patients typically overestimate harms with qualitative descriptions, providing this information quantitatively reassures them that adverse effects are less frequent than they may think. While our discussion has focused on harms, there is likewise a responsibility to accurately convey information regarding benefits as well (e.g., the probability that a statin prevents a heart attack), which are also commonly misunderstood by patients.10
By engaging in these practices, pharmacists can use their medication expertise to help address health care system weaknesses in respecting patient autonomy. There are, however, practical limitations to this: A pharmacist, outside of a specialized setting, may counsel on hundreds of medications and it is not reasonable to know the numbers associated with each one. Fortunately, there exists a set of medications that are very common in generalist settings (e.g., cardiovascular or diabetes medications). For these medications, it is possible to learn the numbers and to include them in counselling. Commonalities between medications in the same class (e.g., common blood pressure medications have similar chances of preventing cardiovascular events or causing dizziness) may help simplify this learning process as well.11 Many resources are available to find this information, whether it be directly from primary resources (e.g., Cochrane Reviews)12 or from tertiary resources (e.g., Compendium of Pharmaceuticals and Specialties).13 Importantly, although this does require some time for background research, it should not add any additional time to counselling sessions, making it practical even for settings where time is precious.
Although I (Blair MacDonald) had previously learned about this approach in an academic setting, seeing the benefits first-hand during a routine counselling session was what truly convinced me to use this style of communication: A 62-yearold male with “blood pressure slightly above target†was receiving a new prescription of amlodipine 5 mg po daily, to be taken in addition to his ramipril 5 mg po twice daily. I asked if he was aware of the increased risk of heart attack or stroke from the celecoxib that he had been filling regularly for several years. He said his doctor was aware he was on the celecoxib for ongoing pain (it was the only analgesic he found effective and tolerable) and had mentioned it could raise his blood pressure, but the patient was visibly concerned to learn about these cardiovascular risks. However, when I further informed him that the ballpark estimate was a 3% risk over the next 5 years, he immediately looked relieved (having clearly feared the risk was higher). He said he was happy to continue the medication because that risk was well worth the pain relief.
It goes to show that just a few numbers here and there can go a long way towards improving patient care.
From the Faculty of Pharmaceutical Sciences (MacDonald), the University of British Columbia; BC Emergency Health Services (Galozo), Vancouver, British Columbia. Contact blair.macdonald@ubc.ca.
Author Contributions: Blair MacDonald initiated the project and was the primary author for each iteration of the manuscript. Bryson Galozo reviewed and provided feedback on each iteration of the manuscript.
Financial Acknowledgments: No funding was received in relation to this manuscript.
Industry Sponsorship: There was no industry involvement in this manuscript.
ORCID iDs: Blair MacDonald https://orcid.org/0000-0002-4666-1790
Bryson Galozo https://orcid.org/0000-0001-6390-4873
