Vancouver vending machine markets in everything

Health advocates say a safe supply of opioids is critical to help prevent people from overdosing on tainted street drugs.

Now, a pilot project in Vancouver’s Downtown Eastside provides some high-risk users with access to an automated machine that dispenses opioids prescribed by a doctor…

The machine, called MySafe, is stocked with hydromorphone tablets that are released on a pre-determined schedule to high-risk opioid users. A user must scan their palm on the machine to identify themselves. The machine recognizes each individual by verifying the vein pattern in their hand and then dispenses their prescription.

Made of steel and bolted to the floor, MySafe resembles an ATM or vending machine. It logs every package that is released and sends that information to a web feed that only program administrators can access.

Here is the full article, please solve for the equilibrium.  Via Michelle Dawson.

Comments

"Health advocates" is quite a euphemism.

From the linked article:

Participants can use the machine up to four times a day, but they must wait a few hours between each use.

The difference between this machine and a pharmacist is that it only dispenses single doses. This solves three drug problems: 1. unknown ingredients, 2. unknown dose, 3. inability of patient/consumer to follow indications.

It takes a high degree of cynicism for a true health advocate not to pay attention.

My surprise is that the focus is oral opiates. I thought the inner-city crisis was mostly injected opiates (i.e. heroin and its fentanyl alternatives).

If you are an ambulance chasing trial lawyer the bad opioids are prescription drugs. Because that's where the money is.

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

I think Bernheim and Rangel, AER 2004, have solved for the equilibrium in these kinds of situations (though their application are methadone dispensaries).

Respond

Add Comment

How much different is this than getting a prescription filled over the counter by a pharmacist? I've always thought that pharmacists are generally overpaid for just filling prescriptions. In what percentage of cases do they actually add value other than making a retail sale? Perhaps vending machines should be expanded to dispense other types of prescription drugs. Quite possibly, this would reduce the overall cost of health care.

Depends on what you mean by adding value. Pharmacists routinely inform doctors about prescriptions from other providers that are important to know about in the future even if the contraindication is not serious this time.

In any event, the better question is how much do they save when they catch something. A single case can easily cost the healthcare system over a million dollars (e.g. knowing something that prevents iatrogenic liver failure leading to a saved transplant).

This is most of medicine. 90% of the time the patient gets better regardless of what happens. The value added is in irregular chunks of exceedingly high value. Pharmacy is not my gig so I cannot speak intelligently to overall rates, but I can say that I still see enough catastrophic trouble from prescription drugs that they could be a net positive on the bottom line.

“Pharmacists routinely inform...”. But, if you read the article, the machine not only routinely informs but it always informs. And , the check on identity seems more robust.

Nothing you wrote, including “catching something” convinced me that pharmacists are always a necessary cost in the supply chain.

The big thing a pharmacist does is check over the patient file and ensure that nothing complicated is happening.

Most polypharmacy patients are taking medications from multiple prescribers. Most docs (myself included, though less so than average) know only a small fraction of the drugs and are utterly ignorant of interactions among less common classes. Pharmacists either know them all or (should be) checking them with every fill. It is idiotically common for an infectious disease specialist to prescribe something that plays poorly with what a gastroenterologist might already have the patient taking. The system is built to have the pharmacist catch this. I have watched them save lives this way in the hospital very often.

In the community role you might be able to replace a lot of this safety double check with an automated system, but in my experience those tend to start breaking down when patients have their fifth or so concurrent medication.

E.g. you get diagnosed with TB (rare in the US) so the doc prescribes rifabutin (preferred, but less familiar) and does not notice that this will lead to CYP 450 inhibition which might muck around with patient's Plavix. It is no knock on the doc, they might not even know that the patient is taking an anti-platelet. But that does not stop the patient from ending up dead when they develop a thrombotic embolus.

How common are such things? Well taking your later estimate of $160,000 per year in salaries and bumping it up to $200,000 per year for out the door costs, then we need a pharmacist to catch about 1 Plavix induced embolus per year. I know pharmacists who hit 500 scripts a shift, so annually we are talking about something about preventing one major screw up in every hundred thousand prescriptions.

Certainly looking at the harm pharmacists can do in errors suggests that a crappy pharmacist can kill a lot of people. I will put a lot more money down on the pharmacist correcting physician mistakes than I would on pharmacists introducing errors to correct physician scripts.

Prescription writing and dispensing is complex process. Pretty much all the easy cases are already automated with mail order pharmacy. Even with that we really only need pharmacists to bring our catastrophic error rate down by 1 part in hundred thousand to justify their cost.

Do they actually do that? I don't know. I mostly prescribe out of a hospital pharmacy and my patients are vastly sicker on average … yet for what comparison value there is, I can think of dozen of patients who would not have survived without a pharmacist double check.

I trust a drug interaction database over human memory. This should be automated, and would be, were it not for rent seeking guilds.

The one advantage they had over standard retail staff is they have more to lose and thus are less likely to put their career at risk by doing anything dodgy. This vein reading machine seems even less likely.

Databases are fine for once the information is in the system. The real trick is finding out if the information is valid and which interactions are just the cost of fixing something serious.

For instance, basically everything modulates autonomic responses and the automated systems will throw an interaction warning at me that at least 9 times out of 10 should be completely ignored. The patient needs both drugs for something like not being psychotic and being able to breathe; they may just have to live with wonky blood pressure and the restrictions that places on life. Likewise, I care an awful lot less about nephrotoxic interactions when my patient is already on dialysis.

In theory, sure, you can build out the rule set on the computer to catch even this sort of thing. But the reality is that you deal with a lot of incomplete information and huge amounts of GIGO.

The healthcare has spent literally decades and trillions trying to get those records to be accurate and machine readable. Last I checked, they still were too incomplete and poor to get by with out a human common sense test.

In any event the real answer is that we already use the automated system, we use the pharmacist and automated combination because it provides superior results to either alone.

Again, maybe the numbers have finally flipped, but this is not some nefarious plot, we have tried to eliminate pharmacist hours (and have gotten pretty close for some mail order outfits), but the real world implementation has not lived up to the hype yet.

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Also, while pharmacy is also not my gig, as an infrequent customer, I’ve yet to have a pharmacist do anything but look at the prescription and hand over the goods (frequently done by an assistant). The average annual wage of a pharmacist in the US is about $125k and assistants about $40k. They add a lot of cost to distribution.

Although I appreciate your emphasis on productivity gains through automation, I think you are missing the point of the article. This is a machine that safely dispenses single dose oral medication in a public place. It is closer in purpose to patient-controlled analgesia techniques but repurposed for high-risk addicts.

In terms of automating pharmacies, you are making the assumption that the only barrier to adoption is human shortsightedness. I'd argue it is purely a technical issue. Once the technical systems are available, perhaps your focus on human agency will prove to be the remaining barrier.

I'm not sure the article had much of a "point" in that it was a news report, not an editorial, of a recent development in Vancouver---to wit, the delivery of certain prescription drugs via vending machines to drug addicts. *My point*; if you care to address it, was why not extend this to other types of pharmaceuticals? If you care to address that, please do so with a little less sarcasm.

There was zero sarcasm intended in my response. I was trying to politely point out that you are putting zero effort into thinking through the problem like an engineer would. Think through the problem of automating prescriptions. Think through the problem of expanding the number of products dispensed. The answer to "why not extend?" is "because it is a non-trivial technical challenge".

Respond

Add Comment

Unless I miss my guess, you are most likely healthy, young (i.e. <60), and not a polypharmacy patient. Even if I screw up, royally, you are highly unlikely to die or even be all that seriously inconvenienced. You are also, maybe burning through 3 scripts a month. Getting rid of the safety checks a pharmacist provides for someone like you might be cost effective, but it might not. The NNT either way would be huge.

For the people who consume the vast bulk of pharmaceuticals, things are far more dicey. Having a safety check when you a kidney drug, a liver drug, a brain drug, and a heart drug on board (not at all uncommon) is going to save more people than a safety check when you have a bog standard, exceedingly well studied blood pressure pill.

As far as dispensing to addicts, well the dangers of an improper prescription are a lot less dangerous to them because they are much more likely to die from heroin than any medical errors. Absent treatment, life expectancy for heroin use disorder patients is something like 10 years. A regular supply of methadone will likely mean fewer exposures to street heroin with God knows what adulterants in it. You will have far fewer adverse drug interactions with even a high error rate prescription system than with any sort of exposure to rat poison or whatever fun they cut drugs with in your neck of the woods.

In any event, where exactly did we conclude that these "vending machines" actually cut out the pharmacist? I would assume that they operate the same as a mail order pharmacy - doc writes script, pharmacist double checks it, and the computer counts it out.

You totally screwed up (>65) but I hope you are right (highly unlikely to die). I also burn on average less than one script per month, but then again I’m not in the film industry.

“We” did not conclude and neither did I that we should get rid of pharmacists completely. However, I do believe that in most cases they are an unnecessary middle person between the prescribing doc and the patient. And, let me guess—you make money writing scripts in the hope they will not be burned😁

Nah, like I said, most of my scripts are filled by a hospital based pharmacy. I actively try to avoid ones for community fills as those can be a royal PITA with opioid restrictions, reporting requirements, and the need to do a thorough med rec before handing out something (hence I why I prefer to fill stuff in our pharmacy or send them back to primary care).

As far as making money, writing scripts is very inefficient for RVUs and I, thankfully, am not paid based on Press Ganey Scoring. Scripts do not actually touch my bottom line.

For various reasons, I am currently writing fewer scripts than I once did. This makes me value having a second, more knowledgeable, set of eyes go over them more when I do. My error rate is unlikely below 1/100,000 on my own, but the conjoint error rate for two of us should be in that range or lower.

A third set of eyes would likely reduce your error rate even further! Does that mean the government should require it?

"Scripts do not actually touch my bottom line."

That's a very narrow way to look at things. The ability to write scripts is basically what differentiates an MD from other health professionals. It is not necessarily the actual writing of scripts that enhances your bottom line, but the ability to write them that contributes to the wage premium.

Like all error correcting processes, there is a point where diminishing returns make them not worthwhile. Typically my scripts go through 4 sets of eyes, though only I and the pharmacist have the top level training to catch the rare stuff.

And you are behind the times on prescriptions; prescribing power has been extended to: nurse practitioners, advanced practice nurses, physician assistants, podiatrists, dentists, clinical pharmacists, optometrists, medical psychologists, and a few others (all of this varies by state and can have limitations on class and refill counts).

In my department we make heavy, heavy use of physician extenders so the vast majority of our scripts are not even written by MDs. The wage premium is that I trained longer and have a wider/deeper scope of understanding when it comes to certain areas.

The premium for the pharmacists come in that their area of expertise are the drugs themselves. I can manage your pulmonary embolism and keep you alive, even things start getting wonky. They can catch drug interactions that I might miss.

If you are otherwise healthy, either of us could manage the PE at least 95% of the time. Likewise if you are otherwise healthy either of us could manage the med rec. But 5% of the time (and it is hard to predict who will suddenly fall into that 5%) you want the guy who can quickly place chest tubes, traches, and central lines and manage the resulting afib and hypotension because your PE is difficult. Similarly 5% of the time you really want to have the guy who knows the actual differences between drugs within classes or the exact amount of CYP 450 inhibition.

I legit do not know at what point community pharmacists become redundant; but I would be uncomfortable writing scripts without a safety check at the pharmacy. What I can assure you about, having seen the times things went wrong, is that the vast, vast bulk of the cost savings are going to come from a small minority of patients.

Thanks for the interesting discussion. This is my last comment on this thread.

Regarding non-MD's writing prescriptions, first, I considered dentists and the like equivalent to be MD's. While I may not be completely up to date on the ability of NP's and PA's to write prescriptions, my understanding is that they don't have the same authority as MD's. That is, most often they need to have a supervisory relationship with an MD (that means six set of eyes and not four).

As regards dimishing returns, that is indeed the issue. It is my belief, that the "reforms" you have cited have come after hard battles with established interests, such as pharmacists, the AMA, etc. I'm not convinced that every drug that needs a prescription needs to pass through the careful eyes of a pharmacist even after considering your comments. In fact, it is my suspicion that they don't really review everything as it stands even though in theory they are supposed to.

The issue of diminishing returns crops up in a number of analogous places. For example, how much security is enough at the airport? Is all that cost worth the benefit?

Bottom line: I'm not for eliminating pharmacists, but I think the distribution channel of pharmaceuticals could be streamlined much, much more to everyone's overall benefit. The dollars saved if applied intelligently elsewhere would likely provide greater health improvement and cost savings. Money is not infinite so it needs to be applied where it serves the best purpose. Since you seem to know where the risks are, start with eliminating redundant reviews with drugs that require a prescription but don't present enough danger to warrant all the middle men and women.

Prescribing authority for the lowest tiers varies heavily by state. Some allow for independent prescriptions others don't. I suspect they will all be independent in another 10 years or so.

It is not really the drugs that determine the risk, but the patient. Adding a "harmless" drug to a risky one gets into basically the same state as adding a risky one to a harmless one. The places where automation is the safest are for maintenance prescriptions with patients who are not having complications. We have highly automated those via the mail. The biggest threat there is actually the government fighting the pharmacists, with the former repeatedly passing laws limiting the number of scripts that can be overseen by one pharmacist and the latter fighting to abolish or increase any cap. Typically this results when there is a one in a million fatality and people freak out hearing how little time a pharmacist spends on checking a script thanks to the huge amount of automation already built into the system.

Bring the mail supplier's efficiency to CVS is going to be much harder. CVS takes all comers and retail outfits are notorious for having incomplete med recs (PSA: use exactly one retail pharmacy for everything whenever possible).

I can agree with the sentiment, but an automated pill dispenser is not really going to change a lot of the economics underlying pharmacists role in medication delivery.

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

People often debate the different tropes in Orwell's "1984" vs. Huxley's "Brave New World", arguing which one is a better allegory for modern life. Well, they're both right.

One is the how-to manual, and the other is the reason why.

Respond

Add Comment

Crude attempt at solving for the equilibrium: this would push the supply curve to the right while demand stays about the same. The supply curve is pushed to the right because the government is artificially providing opioids to the market for free. The demand curve stays about the same because people generally are not drawn to the idea of becoming addicted to opioids.
As a result of this policy, we expect price of opioids to decrease, which would further disincentivize opioid production, both illicit and legal. From there, we could see additional benefits in terms of better public health, fewer overdoses and so on.
Also considering how it could go wrong: if it doesn't budge the supply curve to the right much, then price would stay about the same. As a result, drug dealers can milk their customers for profit for much longer (assuming fewer overdoses), and malevolent doctors pushing opioids purely for profit can sleep easier at night knowing patients will be able to get opioids with much less risk. Just a thought experiment.
It'll be interesting to see how this plays out.

In Vancouver, I'd be thinking: Oh, all I have to do is convince some motivated doctor that I have a heroin problem and forever afterwards, a machine will start dispensing for me free opiates with considerable street value? Gee, thanks Vancouver!

from the linked article:

Participants are opioid users who have a history of overdosing. They must undergo a medical evaluation before being prescribed hydromorphone and agree to regular followups with a health professional. During the pre-screening, a prospective participant must have fentanyl detected in their urine to be eligible.

>agree to regular followups with a health professional

Not a doctor, mind you. A "health professional."

Which includes acupuncturists, massage therapists, and people who tell you that their essential oils will cure your cancer.

This is all on you, Canada. No sympathy here.

Respond

Add Comment

"history off overdosing" is interesting.

What are the economics here? How much does an "overdose" cost the medical system? How many times does an addict overdose before they manage to kill themselves? Whats the end-point distribution for this population (death directly or indirectly due to drugs, or clean and productive life off drugs)?

Respond

Add Comment

Wait, so to get on this gravy train I have to overdose on heroin and get some fentanyl into my urine? Well, that's harder than just asking, but if Vancouver insists, I'm in. At worst, the doctor will tell me "You're not enough of a junkie yet - try harder, eh?!" and I'll try harder.

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

"The demand curve stays about the same because people generally are not drawn to the idea of becoming addicted to opioids."

Dealers aren't selling addiction, they're selling the experience of getting high. There are lots of people drawn to that idea, who will consume more if the price drops.

Respond

Add Comment

Respond

Add Comment

"Now, a pilot project in Vancouver’s Downtown Eastside provides some high-risk users with access to an automated machine that dispenses opioids prescribed by a doctor."

How could it go wrong?

So valuable medicine will be let unguarded so that criminals can loot?

One pill at a time just annoys some people, nice pic:

https://www.ozarksfirst.com/local-news/local-news-local-news/truck-crashed-into-marionville-pharmacy-suspect-steals-narcotics/

That, and dealers who are quite comfortable with removing competitors.

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

I don't think it's particularly useful to think about this policy initiative in market terms. Vancouver's downtown eastside is home to some of the worst human misery found in Canada. (I won't say the very worst only because some communities in the Far North may deserve that sorry title.) The downtown eastside is a congested nightmare of life-long drug addiction and infectious disease. This effort to provide free, easy access to opioids is to reduce the spread of disease, not to disrupt the drug market. In fact, I'm quite certain that serving some members in this community with free drugs has almost no impact on the overall drug trade in Vancouver.

Respond

Add Comment

... innovative Canadian & US government drug regulation solved the heroin, cocaine , marijuana, etc drug/health problems decades ago --- so this little Opioid problem will easily be eliminated by the same type of highly successful government policies.

"Health Advocates" within the government always know what is best.

These kind of policies actually did work exceptionally well in Switzerland in the 80ies and 90ies. Much of the success is certainly due to the fact that these policies were extremely pragmatic, and largely free of ideology.
See https://www.thenation.com/article/archive/switzerland-addiction-prescribed-heroin/

Indeed. In Australia the conservative approach would be a return to heroin being available by prescription, as it was in Grandad's day.

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

This is how our new robotic overlords will put us at their heels, not Schwarzenegger.

Respond

Add Comment

As other commenters have pointed out, calling this a "vending machine" is inaccurate: it has more in common with a time-release safe. Trolling for outrage again.

Even more in common with those timed cat feeders, that cats appear to worship.

Respond

Add Comment

Respond

Add Comment

If the private sector did this they’d go to jail.

Worse than that, they'd go broke.

Respond

Add Comment

Respond

Add Comment

In Australia heroin was available by prescription until 1953. Oddly enough, its use continued after it was effectively banned in that year and people have actually gone to jail for using or selling it. This is unfortunate because, when you weight things up, going to a doctor isn't as bad as going to jail. Less costly too.

Respond

Add Comment

Then they will steal the floor.

Respond

Add Comment

Lower the cost of being a junkie, and you'll get more junkies. QED.

(Saying this as someone who has observed every harm-reduction measure leading to more bedlam and misery on the DTES' streets)

Respond

Add Comment

This is anecdotal, but Vancouver was the most despondent North American or European city I've ever been in. There was one neighborhoood where people were just laying in the middle of the street, no traffic to speak of, I'm assuming people with cars avoid the area. I saw one person I thought was dead, but it was hard to tell. Just lots of open drug use, lots of fighting, it was shocking to see. Whatever they are doing seems to be failing, I am skeptical this would help.

Portland is similar.

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment