Robertsville is a small town. Like many small towns, it is famous for one thing only. In Robertsville’s case, it is for having recently granted citizens rights to self-medicate. Although Robertsville doesn’t have many amenities, it does have a doctors surgery and a pharmacist.
In Robertsville, people are now able to purchase recreational drugs, untested pharmaceuticals, and there are no longer prescription requirements. Although pharmacists are free to dispense them without a prescription, this doesn’t mean everyone can get hold of them. Pharmacists must assess people’s competence before dispensing drugs and engage in mandatory disclosures about the risks and benefits of consuming them, asking people to answer questions about the reasons for their decisions and whether they understand the information disclosed to them. People who cannot demonstrate they are competent at the time of purchase will not be dispensed medications.
The aim of all of these policies in Robertsville is to ensure that people don’t harm themselves unknowingly or as a result of their incompetence. To illustrate how they work in conjunction, let us consider an example:
Bruxism: For the last couple of weeks, Frank has been having terrible headaches when he wakes up in the morning. When he woke up today, he found a piece of his tooth-filling in his mouth and his face felt tense. Worried, Franks goes online and starts searching the NHS website for advice on whether to see a doctor. Online Frank discovers a condition called bruxism he had never heard of. The main symptoms are teeth grinding, which in turn leads to broken fillings and headaches. Frank also discovers bruxism is related to stress and can be a coping mechanism for frustration. Frank often feels both of these things, as he has a terrible boss. Convinced he has bruxism, Frank looks at how to treat it. One way to stop his headaches is to try and reduce his stress. Although Frank would like to do this, he has failed in the past because his fluctuating shift patterns at work make it difficult to engage with cognitive behavioural therapy. Another way to solve it is to adopt better sleep hygiene, which he can’t do because of his shifts. The third option is to give up on alcohol and recreational drugs such as MDMA or cocaine. Although Frank has cut down on these since his raving days, the occasional night out in a club is the only thing left in his life that still makes him feel young. Frank concludes that, given his inflexibility, there isn’t much use visiting the GP, who he presumes will encourage him to treat his underlying stress and lead a generally healthier life-style. Convinced there must be a way to reduce the symptoms he is feeling without dealing with the underlying issue, Frank embarks on more research. On the NHS website, Frank discovers that GPs occasionally prescribe muscle relaxants to treat bruxism. Having used Valium recreationally as a teenager, Frank feels it would help relieve his tension and decides to visit his pharmacists to get some Valium.
In Bruxism, Frank runs a high risk of misdiagnosing himself and, consequently, of choosing ineffective means to his ends; in this case, the absence of pain. Headaches are very common complaints and, in the majority of cases, aren’t caused by anxious teeth grinding. Some of the other causes of Frank’s headaches might be dehydration, eyesight problems, or not eating regular meals. If these are the true underlying causes, and they are left untreated by using Valium as a muscle relaxant, Franks headaches will likely continue. More worryingly Frank could find himself ignoring more serious problems which cause headaches (such as inflamed arteries in the head). Moreover, as Valium has psychoactive effects, using it to mask headaches could put Frank in a position in which he can’t identify further symptoms (such as drowsiness) that would be indicative of another explanation for his headaches.
In short, due to the potential for misdiagnosis and choice of inadequate treatment, Frank could be unknowingly harming himself by deciding to consume Valium to treat his headache. The question is: could the safe-guards set out in section "Some Unanswered Questions" help Frank avoid harming himself through ignorance or incompetence. To illustrate, it will be useful to expand on the example considered above:
Misdiagnosis: Frank has just made it to the pharmacists on Robertsville High Street. Frank goes inside and is greeted by the pharmacist, Mr. Nice. “What can I do for you?” asks Mr. Nice. Frank explains he would like to purchase some Valium. As these medicines are behind the counter, Mr. Nice tells Frank they will have to conduct a brief competence assessment. To start off, Mr. Nice asks Frank whether he is suffering from anxiety or depression, or whether this is for recreational use. “Neither,” Frank says, “I’ve had a terrible headache every morning for weeks and I think I’ve been grinding my teeth”. Frank tells Mr. Nice he thinks he has bruxism. Mr. Nice asks Frank whether a physician diagnosed him or whether he’d come to this conclusion himself, to which Franks responds he read it on the NHS website. “You should probably go to the doctor as your headaches seem persistent” says Mr. Nice, “but it is unlikely that this is bruxism so the Valium probably won’t help. Moreover, given that it makes you sleepy and makes your breathing shallow the best thing to do for your headache is to just go home, drink water and take some paracetamol.” Frank explains he hasn’t got the time to go home and convalesce, as he needs his headaches gone so he can go back to work and carry on with his life. Mr. Nice tells Frank he might find that harder on Valium than he currently does as it will make him sleepy and lethargic. Frank is worried about this possibility, but over the years of working shifts he has become quite adept at working even though he is tired. If it gets too bad, he will stop taking them.
Frank explains he doesn’t like going to the doctor and will look into potential other causes of his headache if it doesn’t go away after 2 or 3 weeks on Valium. He understands that there is a risk of misdiagnosis, but there isn’t really any way of completely excluding that risk either. When medical problems are complex, it can take more than one visit to a physician to get an accurate diagnosis. Getting an appointment at the GP is difficult, so he will try fixing it himself first and go to them if it doesn’t improve.
Mr. Nice tells Frank he shouldn’t drink alcohol when taking Valium, as they are both depressants. If taken together, they can lead to shortness of breath and, in worst case scenarios, death. Frank has always drunk every day, not excessively, but he does drink double or sometimes triple the UK national guidelines. Mr. Nice tells Frank it is unlikely he is going to stop drinking but that he is worried about the combined effect of the alcohol and Valium. Although he won’t go tee total, he will try and moderate his drinking.
Following this informational disclosure, Mr. Nice needs to assess the extent to which Frank is competent. The first thing to do is establish the competence threshold Frank must meet. During their conversation, Frank expressed reservations when confronted with the possibility of misdiagnosis, was clearly worried about the risk of combining Valium and alcohol, and was concerned about Valium interfering with his ability to work by making him sleepy and lethargic. Given that Frank is unlikely to stop drinking, change his mind about visiting the doctor, or trying less risky treatments (like resting and rehydrating, treating his underlying stress or simply wearing a mouthguard to prevent teeth-grinding), Mr. Nice concludes it is risky for Frank to consume Valium and sets a moderately high threshold of competence.Footnote 4
Having agreed on the threshold, Mr. Nice needs to ascertain whether Frank can meet it. To do this, Mr. Nice needs to test the extent to which Frank possesses the Core Capacities (i.e., knowledge, rationality, and a life plan). To check whether Frank can acquire knowledge, Mr. Nice asks him to explain what bruxism is and asks him questions about the effects he thinks Valium will have, the outcome he intends to achieve by using Valium, and the risks he is taking by doing so. Frank responds that bruxism is teeth grinding, that this can lead to headaches, and that muscle relaxants are sometimes used to treat it. For Frank, the goal is to relax his jaw, thus lessening the headaches. Although he understands that there may be other causes for his headache, Frank insists that he wants to try Valium first. If that doesn’t work he will try something else or consult a doctor to see what they recommend. Frank is aware of the risks of misdiagnosis and has set a 2-3 week deadline on his attempts to self-medicate with Valium as a way of reducing the risk of him harming himself.
To test whether Frank possesses the capacity for instrumental rationality and the capacity to revise a life plan, Mr. Nice asks Frank to explain why he thinks it is so important to get back to work and why he can’t simply take some time off to recover. Mr. Nice tells Frank that his goal of getting on with his life and getting back to work may actually be hindered by using Valium to self-medicate, as it will make him sleepier. Getting some rest and trying to lead a generally healthier life, on the other hand, might actually help him further his goals more reliably for longer. Frank acknowledges that, in the long run, leading a healthier life would help him pursue his goals more effectively and that getting hooked on Valium would make it harder to get on with life. The problem is that leading a healthier life would require a lot of will power. Frank is moderately happy with how he lives his life. At the moment, it is just particularly stressful and Frank needs a short term solution which is easy to fit in to his routine. Valium seems, to Frank, to be the obvious answer.
Following their conversation, Mr. Nice determines that Frank is competent and dispenses Frank’s medications, informs him about recommended dosages and points out the number for a self-medication help-line printed on the packaging. “The helpline is staffed 24/7 by qualified professionals who can offer advice on how to use pharmaceuticals safely.” says Mr. Nice. “Call them if you have any questions or, alternatively, pop back in to the pharmacy.” With his medications in hand, Frank leaves the pharmacist.
The question we need to answer in Misdiagnosis is: if Frank came to harm from taking these medications, could this be attributed to his ignorance or incompetence? It seems to me that, if Frank does eventually come to harm from his Valium use, this cannot be said to have done so unknowingly or due to a lack of competence.
Prior to dispensing the medication, Mr. Nice informed Frank of the risks of consuming drugs and that of the existence of less harmful solutions to his headaches (e.g., rest, relaxation and hydration). Importantly, this information was disclosed to Frank at the time he was making the decision in a format he could understand. In other words, the information was made accessible to him (as opposed to simply being made available). Moreover, the fact that competence assessments require Frank’s active participation serves to guard against the possibility of Frank simply sitting through the disclosure, letting it go in one ear and out the other. Given that Frank must demonstrate that he has the capacity to acquire knowledge before Mr. Nice can dispense the medication, Frank has to engage with the informational disclosure and take it into account in his deliberations.
If, for example, Frank didn’t understand that he could be mistaken about the cause of his headaches and that, as a consequence, the Valium might not help, he would have been found to be incompetent by Mr. Nice and Mr. Nice would not have dispensed the medications Frank had asked for. This would also be the case if Frank’s lack of competence were down to the fact that Frank couldn’t demonstrate he possessed the capacity for instrumental rationality when choosing treatment options, or if Franks behaviour was completely purposeless in that he didn’t have a goal he was seeking to achieve by taking Valium.
In short, informational disclosure and competence assessments could help protect people from unknowingly or incompetently harming themselves without abridging people’s self-medication rights. If extensive safe-guards are in place to ensure that people purchasing drugs are both competent and have the information about risks when they need it, the idea that people should be entitled to access dangerous pharmaceuticals without a prescription is less counter-intuitive. Under a regime of liberalised access, such as the hypothetical one I propose here, individuals who wish to avoid harming themselves have ample opportunities to engage with experts who can disclose information to them and answer any questions they may have to help them understand the risks they are taking. Once information has been made accessible to people and their decision-making competence has been rigorously assessed, any harm that may result from their actions cannot be attributed to ignorance or incompetence.
It could be objected at this point, that the hypothetical scenario I have outlined above is overly optimistic in a couple of ways. Firstly, it might objected that it is unlikely, in practice, that pharmacists will take the time to have conversations about the drugs and ascertain competence. Isn’t it more likely they will simply dispense the medication, no questions asked? Secondly, it could be objected that I haven’t provided any data to justify the optimistic view of how things would progress in Robertsville and, as a consequence, it is still unclear whether the safeguards outlined above would help prevent people from harming themselves with drugs. In what remains of this section, I take up these objections.
Let us take the first objection first. Assessing competence and disclosing information in an accessible format is time-consuming. Implementing the extensions to Flanigan’s proposal I am arguing for will, therefore, mean that dispensing medications in Robertsville is likely to be a more protracted process than simply having a prescription filled. Now, the problem is that time is in short supply in many healthcare systems,Footnote 5 which could make compliance difficult to achieve (Hibbert et al. 2002, p. 55, Berger 2009). Community pharmacists in particular, are under commercial pressure to fill prescriptions quickly to increase the amount of patients they can see (Berger 2009; Latif 2000; Wingfield et al. 2004; Hibbert et al. 2002; Prayle and Brazier 1998; Resnik et al. 2000). The question is then: how likely would it be in practice that pharmacists would engage in thorough competence assessments and information disclosure before dispensing drugs?
The answer to this question will depend on a number of factors including (among others): whether dispensers have the resources to meet these requirements, whether having these conversations is a legal requirement (Resnik et al. 2000), what the sanctions for non-compliance are, how effectively these are enforced, how well the skills are taught (Coulter and Ellins 2006, p. 68), how much emphasis is placed on students learning them during training (Roche and Kelliher 2009; Kettle 2003; Wingfield et al. 2004), the institutional culture of the workplace (Latif 2000; Resnik et al. 2000), and whether there is a widespread expectation amongst the public that pharmacists routinely disclose information about drugs and assess competence. In short, how likely compliance will be will depend on the specifics of how the proposal outlined above is implemented on the ground.
The question of how to implement policies to achieve optimal compliance is beyond the scope of this paper. The reader is therefore invited to introduce their preferred view of how to enforce the proposal outlined above. As a consequence, I’m forced to leave the question of how likely compliance with my proposal will be in practice unanswered in this paper. That said, what this short discussion of enforceability does reveal, however, is that how likely compliance will be with any given proposal is not a fixed, immutable, characteristic. Instead policymakers have a number of tools at their disposal which they can use to increase the likelihood of compliance.
The requirements that pharmacists assess competence and disclose information before dispensing drugs mirrors informed consent requirements imposed on other front-line healthcare professionals. Seen in this light, achieving compliance with the policies outlined in this paper doesn’t seem especially problematic. If backed by appropriate enforcement mechanisms (whatever those turn out to be), lack of compliance with the policy can be made as likely as compliance with informed consent requirements in other contexts. Although there are still hurdles to be overcome when implementing informed consent requirements consistently in practice (Coulter and Ellins 2006; Evans et al 2007; Jackson and Warner 2002; Kim 2010, p. 59), these are not generally considered insurmountable or a reason to cease trying to implement the requirements of informed consent. Instead, they are seen as problems to be overcome by, for example, redirecting resources or developing new and betters ways of ensuring compliance. This is also the approach we should take to the problem of ensuring pharmacists comply with their obligations to assess competence and inform people before dispensing drugs.
Having responded to the first objection, it is now time to consider the second objection: the fact that I have not provided data to support my account of how things would progress in Robertsville. The crux of this objection is that, without data on how effective the policies outlined above are at protecting people from harm, we can’t be sure that the deontological and the epistemilogical-consequentialist arguments will pull in broadly the same direction. This is a problem for defenders of drug liberalisation, because unless we can allay concerns that increased access to drugs will lead to catastrophic consequences for significant numbers of people, it will be hard to generate the levels of public support for drug liberalisation necessary to implement the policy in a democratic society.
The short response to why I haven’t provided any data on how effective the proposals outlined above would be at ensuring that people don’t harm themselves as a result of ignorance of their own incompetence is that we simply don’t have it. The reason we don’t have any directly relevant data is that, to my knowledge, the proposals I am outlining haven’t been implemented anywhere in the world. Given the lack of direct data, if we are to try and allay concerns about the liberalisation of drugs, we need to search for analogous evidence that might give us an indication of whether the proposals outlined are workable in practice. However, once we choose to go down this route we are immediately confronted with the problem of determining what counts as "analogous evidence", which turns out to be far from easy.
To illustrate, one option would be to try and compare the drug poisoning rates in countries with stringently enforced prescription requirements with the rates in countries with more lax approaches to prescriptions (Peltzman 1987). The idea behind making these comparisons is that people in countries with lax enforcement of prescription requirements are living in a de facto (but not de jure) liberalised market, which is the closest we can get to a certification scheme. If there were significantly higher drug poisoning rates in countries with de facto liberalised markets, we could take this as an indication that liberalising access to pharmaceuticals could have severe negative consequences.
The problem with this approach is that, although data is available for some countries accurate data on drug poisonings rates in countries which do not rigorously enforce prescription requirements is hard to come by, making it impossible to establish fair comparisons between the two groups. Furthermore, even if the data concerning mortality and morbidity rates did consistently exist and showed that both were higher in countries which don’t enforce prescription requirements, it is not clear that alone would show the harm reduction measures I am arguing for are ineffective. In order for the data to establish that, we would need good evidence that pharmacists dispensing medications actually abide by the harm reductions outlined above. In the absence of said evidence, higher mortality and morbidity rates don’t show us that disclosure and competence assessments are insufficient to protect people from harm.
A second option is to look for data on how the decriminalisation of drugs affects consumption, overdose, and death rates. The idea is that the decriminalisation of recreational drugs is partly analogous to a certification system because both proposals make it easier to access drugs. If decriminalisation has disastrous consequences, we might argue, this is a reason to be cautious of all proposals aimed at liberalising access to drugs.
According to Flanigan, “When Portugal de-criminalised all recreational drugs in 2000, rates of abuse, overdose, and HIV infection fell, and greater numbers of users sought rehabilitation and treatment” (Flanigan 2017, p. 80). Data from Czechia also points to decreased use after the decriminalisation of recreational drugs in 2009 and drug death rates which are below the EU average (EMCDDA 2019). We might think, therefore, that the experience of decriminalisation provides evidence to allay the concerns opponents of drug liberalisation might have to implementing a certification scheme.
The problem with these comparisons stems from the fact that that decriminalisation is only partly analogous to the regulatory regime I am advocating. Firstly, decriminalisation only applies to recreational drugs. As a consequence, the data doesn’t give us a clear indication of what to expect from abolishing prescription requirements across the board. Secondly, with decriminalisation, trade in recreational substances still takes place in an unregulated marketplace where, for all intents and purposes, it is impossible to enforce the requirement that dealers establish competence and disclose information about the drugs they distribute. As a consequence, it is unclear what data concerning prevalence of use and overdose rates under a decriminalised system means for the viability of the proposals outlined above.
A third option would be to look at the effects disclosure and competence assessments have in other healthcare settings, such as clinical consultations between physicians and patients. The data here is mixed. There is some evidence to suggest that better communication practices during the process of disclosure can lead to better medication adherence and improved patient knowledge, but have mixed results on health outcomes (Coulter and Ellins 2006, p. 71; Adams 2010, p. 66) and the prevention of adverse events (Coulter and Ellins 2006, p. 143). With regard to the effectiveness of competence assessments, there is some evidence to suggest that higher levels of decision-making competence correlate with the avoidance of negative outcomes (Parker et al. 2015; Bruine de Bruin et al. 2007), which could be seen to lend support to the idea that properly conducted competence assessments could reduce adverse drug events, such as accidental poisonings.
Given the mixed results, it is difficult to assess how effective mandatory disclosure and competence assessments would be at reducing harms resulting from either ignorance or incompetence. Here, as before, we also face the problem of applying the insights from data gathered during clinical interactions in doctors offices and hospitals to interactions in the pharmacy, which are not necessarily comparable.
To summarise, due to the fact that the regulatory regime outlined above hasn’t been implemented in practice, we can’t be sure how effective the proposals outlined above will be at ensuring that people don’t harm themselves as a result of their own ignorance or incompetence. In other words, there is an inherent uncertainty as to what the effects of liberalising access to drugs would be in practice. As a consequence of this lack of empirical evidence, it will prove difficult to allay the concerns of the staunchest opponents of liberalising access to drugs. The problem is that, unless the proposal is actually implemented, it will be impossible to ever generate the kind of robust evidence that would convince those most opposed to drug liberalisation that the deontological and consequentialist arguments can be made to pull in broadly the same direction.
Flanigan’s solution to this Catch-22 situation is to argue that self-medication rights ought to be insulated from democratic politics. Rights to access drugs, like other important rights, should be enforced by the judiciary (Flanigan 2017, p. 163). My reservation with this solution is that, although using the judiciary to enforce self-medication rights is a means of side-stepping the (potentially biased) opposition to liberalisation, it does nothing to resolve the genuine uncertainty surrounding the effects of liberalising access to drugs or actually allay concerns surrounding liberalisation.
If we genuinely want to allay the concerns of opponents of drug liberalisation, we have to actually resolve the uncertainty surrounding the effects of moving to a certification scheme. Judicial decisions won’t help here. What we need is high quality data on the effects of removing prescription requirements. Until we have this data, we cannot be sure that protecting people’s right to self-medicate won’t have negative consequences for society (Dunn 1997). Under these conditions of uncertainty, opponents of drug liberalisation have genuine reasons to be concerned which must be taken seriously.
As I see it, there are two ways of getting the data we need to ascertain whether or not the deontological and consequentialist arguments can be made to pull in broadly the same direction. The first is to wait for a natural experiment to occur and use the results to extrapolate to other situations. The second is to conduct a policy trial consisting in implementing a certification scheme in a limited geographical area for a limited period of time for the purposes of tracking key measures such as overdose deaths, prevalence of use of drugs, health outcomes, what healthcare services are being used and healthcare cost. This data could also be supplemented with qualitative data gathered through interviews of surveys of a subset of the population.
My preference would be for the latter. There are two reasons for this. Firstly, the results of natural experiments can be difficult to interpret and, as a consequence, it might be harder to establish precisely what the experiment has shown. Conducting a policy experiment allows us more control over the regulatory regime being tested and, therefore, helps ensure that the results are applicable to the problem at hand (McDermott 2002, p. 39). Secondly, there is no guarantee that a natural experiment will ever occur and, if it does, we might find ourselves waiting for a long time. Given how high the stakes are when it comes to drug regulation, we need data sooner rather than later.