Colorful nonpareils can uniquely identify drug capsules and counterfeit fashions:
While most of us were baking sourdough bread and watching "Tiger King" to stay sane during the pandemic shutdown, UC Riverside bioengineering professor William Grover kept busy counting the colorful candy sprinkles perched on top of chocolate drops. In the process, he discovered a simple way to prevent pharmaceutical fraud.
The technique, which he calls CandyCode and uses tiny multicolored candy nonpareils or "hundreds and thousands" as a uniquely identifiable coating for pharmaceutical capsules and pills, is published in Scientific Reports.
[...] "The inspiration for this came from the little colorful chocolate candies. Each candy has an average of 92 nonpareils attached randomly, and the nonpareils have eight different colors. I started wondering how many different patterns of colored nonpareils were possible on these candies," said Grover. "It turns out that the odds of a randomly generated candy pattern ever repeating itself are basically zero, so each of these candies is unique and will never be duplicated by chance."
[...] To test this idea, Grover used edible cake decorating glue to coat Tylenol capsules with nonpareils and developed an algorithm that converts a photo of a CandyCoded pill into a set of text strings suitable for storing in a computer database and querying by consumers. He used this algorithm to analyze a set of CandyCode photos and found they function as universally unique identifiers, even after subjecting the CandyCoded pills to physical abuse that simulates the wear-and-tear of shipping.
"Using a computer simulation of even larger CandyCode libraries, I found that a company could produce 10^17 CandyCoded pills—enough for 41 million pills for each person on earth—and still be able to uniquely identify each CandyCoded pill," Grover said.
[...] "Anecdotally, I found that CandyCoded caplets were more pleasant to swallow than plain caplets, confirming Mary Poppins' classic observation about the relationship between sugar and medicine," said Grover.
Journal Reference:
Grover, W.H. CandyCodes: simple universally unique edible identifiers for confirming the authenticity of pharmaceuticals [open] Sci Rep 12, 7452 (2022).
DOI: 10.1038/s41598-022-11234-4
(Score: 5, Insightful) by Opportunist on Tuesday May 10 2022, @07:34AM (5 children)
Until the first 4 year old comes across a medication box. If pills don't look enough like candy already, this should seal the deal.
(Score: 3, Interesting) by pkrasimirov on Tuesday May 10 2022, @09:31AM (4 children)
All pills here have something bitter added so the 4 year old will spit it at the instant. And the coating does not need to make it eye-candy. Personally I think it is fantastic idea. My concerns are mostly on the deliberately abused side, as in insurance companies gaining somehow more leverage from this new feature.
(Score: 3, Interesting) by DeathMonkey on Tuesday May 10 2022, @03:24PM
Not sure how this is supposed to prevent fraud considering it's just a QR code printed on a pill but I think it could still have some beneficial and interesting uses!
(Score: 0) by Anonymous Coward on Tuesday May 10 2022, @04:38PM (2 children)
Not sure where "here" is. But, in the US, I had to specifically request a child proof cap once. My regular pharmacy always used them, but I was in Los Angeles for medical treatment, and the pharmacy I went to there said, they only used them if requested. It was a bottle of opioid pain killers for after a surgery.
(Score: 0) by Anonymous Coward on Tuesday May 10 2022, @05:33PM (1 child)
The problem with childproof caps is that they are also resistant to opening by old people with hand problems: arthritis, weakness, Parkinson's, etc. Most medications are taken by older people and many have an issue like I mentioned. The newer (not that new, let's call them "current") child resistant caps only require one hand to open, unlike the early ones.
(Score: 0) by Anonymous Coward on Tuesday May 10 2022, @07:05PM
I have a plan that always solves this.
Part A: a large stack of empty altoids tins.
Part B: a mitre saw.
Cap comes off, pills go into tin.
Add some painter's tape and a marker pen, and you have a label.
Then tins go into steel locking cabinet with nice, easy-to-grasp handles.
(Score: 2, Funny) by stretch611 on Tuesday May 10 2022, @07:57AM (1 child)
Let's put all meds in a candy coated layer...
Just to spice things up, lets put diabetic medications in two layers!!!
Now with 5 covid vaccine shots/boosters altering my DNA :P
(Score: 3, Informative) by quietus on Tuesday May 10 2022, @10:29AM
(Score: 2) by FatPhil on Tuesday May 10 2022, @08:55AM (12 children)
But even then, how does any of this prevent fraud? What is the precise problem to which this is the solution?
Great minds discuss ideas; average minds discuss events; small minds discuss people; the smallest discuss themselves
(Score: 2) by pkrasimirov on Tuesday May 10 2022, @09:32AM (1 child)
The problem is you find a pill and you want to know what is it. Not an extremely big problem during everyday life but might make a world difference in some cases (crimes? I don't know).
(Score: 1, Informative) by Anonymous Coward on Tuesday May 10 2022, @04:44PM
That is already trivially easy. The color/shape/imprint is unique, you can google a given pill and find out what it is. I generally do when i get a different pill from the pharmacy. I had a girlfriend who got Diazepam/Diltiazem mixed up from the pharmacy and have been suspicious since.
(Score: 2) by Fluffeh on Tuesday May 10 2022, @09:37AM (1 child)
Hard to put a QR code onto a pill-shaped product is my guess for that.
It's a glorified serial number that a company could track and say "Yes, we made this...". Not really amazing.
A summary of this article could be along the lines of:
Non-mathematician discovers topology and multi-coloured spots, amazed to find it can make loads of combinations, thinks it could make some shitty serial numbers on pills.
I'd call out, for a customer, it's not showing in any meaningful way whether the pill they are taking is a knock-off or legitimately made by who they expect. I ain't scanning pills any more than QR codes - which my phone already does. I'm clearly not as excited as the duder in the story on this one.
(Score: 0) by Anonymous Coward on Tuesday May 10 2022, @02:13PM
The customers here, are government officials. They appreciate any excuse to create a little private business on the side, which could then be made mandatory through "think of the children" or any other such bullshit. Russia is the model testing ground of this bureaucratic monetization of office.
https://www.mondaq.com/russianfederation/food-and-drugs-law/675050/faking-it-can-qr-codes-combat-counterfeit-medicines [mondaq.com]
https://www.ccis-expertise.com/en/news/272-russia-data-matrix-code-of-the-national-marking-system-chestny-znak-becomes-mandatory-for-some-products [ccis-expertise.com]
https://rfxcel.com/dairy-product-labeling-russia/ [rfxcel.com]
(Score: 0) by Anonymous Coward on Tuesday May 10 2022, @09:51AM (1 child)
In the past they cleaned money. These days we clean pills.
(Score: 0) by Anonymous Coward on Tuesday May 10 2022, @09:53AM
giving new meaning to "knock off pills"
(Score: 2) by driverless on Tuesday May 10 2022, @12:22PM (5 children)
Because it's not a PUF (physically unclonable function, which is what this is) and therefore not cool. PUFs are defiitely cool, they've been rediscovered every few years since the late 1980s so they must be good.
(Score: 2) by FatPhil on Tuesday May 10 2022, @01:11PM (4 children)
So the problem's that of repudiation - they want to be able to say that a fake pill's not an official one? Well, they can *say* that, but it's not proof, because if they don't publish the non-repudiatable patterns in some form we can't know they're not repudiating a real one. (And if they are publishing the patterns in advance, there might be a way to clone them, with only a small work-factor disadvantage - heck, you could probably pay an Indonesian 7-year-old 2 cents to clone a pill in watercolours once the pattern's known.)
The cryptographic solution to the problem they have is watermarking. PUFs, and technicolour titty sprinkles, aren't watermarking. So I'm still a bit confused. Maybe they want people confused, as confused people buy things.
Great minds discuss ideas; average minds discuss events; small minds discuss people; the smallest discuss themselves
(Score: 2) by HiThere on Tuesday May 10 2022, @01:20PM (3 children)
My guess is that the target is druggists, not end users. This sure wouldn't help end users identify pills, but it might help pharmacists.
My problem is that already too many of the pills I take look too similar, so when I set up my weeks pills and drop one, I can't tell which one I dropped. (Yes, I'm clumsy, and take a lot of meds.)
Javascript is what you use to allow unknown third parties to run software you have no idea about on your computer.
(Score: 2) by FatPhil on Tuesday May 10 2022, @01:43PM (1 child)
Great minds discuss ideas; average minds discuss events; small minds discuss people; the smallest discuss themselves
(Score: 0) by Anonymous Coward on Tuesday May 10 2022, @02:32PM
Timmy, did you eat all of Mommy's Percocets in the Mickey Mouse dispenser? I do NOT want to take you to the ER *again*!
(Score: 0) by Anonymous Coward on Tuesday May 10 2022, @04:10PM
That and law enforcement
(Score: 1, Insightful) by Anonymous Coward on Tuesday May 10 2022, @11:30AM (11 children)
How does this in any way reduce pharmaceutical fraud? Are they suggesting that someone making "fake" pills can't duplicate the "candy coating"? I guess maybe they're going to rely on the DMCA and send take-down notices?
(Score: 1, Informative) by Anonymous Coward on Tuesday May 10 2022, @12:06PM (3 children)
Each individual pill has a unique randomly generated pattern that serves as a serial number. If you aren't certain of the provenance of a batch of pills you can contact the manufacturer to verify the 'serial numbers'. A counterfeiter can copy the technique but the cost of copying a specific pattern should be significantly greater than the price of the pill, and if the pattern isn't in their database then the legit manufacturer will say 'nope, not ours'.
Capsules might still be vulnerable if the counterfeiter can profitably extract and fence the original contents, but that is a pretty niche market and I'm not convinced that it would be profitable.
(Score: 2) by FatPhil on Tuesday May 10 2022, @01:13PM (2 children)
You can't know they're not lying.
Top Tip: The planet where drugs manufacturers don't routinely lie, in particular when $$$$ are at stake, does not exist.
Great minds discuss ideas; average minds discuss events; small minds discuss people; the smallest discuss themselves
(Score: 3, Insightful) by Immerman on Tuesday May 10 2022, @09:05PM (1 child)
One option to reduce that would be a independent third party (FDA?) that held the "valid" pill IDs "in escrow". You want to know the source of a pill, you submit a photo to the FDA, they look it up and tell you exactly what it is, who made it, on what day, which batch, the expiration date, etc.
(Score: 2) by FatPhil on Wednesday May 11 2022, @04:04PM
Great minds discuss ideas; average minds discuss events; small minds discuss people; the smallest discuss themselves
(Score: 3, Informative) by mr_mischief on Tuesday May 10 2022, @02:29PM (6 children)
It's not the candy coating that stops the fraud. There's no standard pattern per type of pill. It's the photographed random pattern on each pill. They take a picture, turn it into a string of codes, and put it in a database. Then a smart phone app can look up the code for each individual pill in a bottle.
It's an ingenious solution, but a data handling and web app nightmare.
(Score: 2) by DeathMonkey on Tuesday May 10 2022, @05:36PM (3 children)
Would that be considered security through obscurity though?
The pattern and/or the random number it decodes to are only secure so long as they remain secret.
(Score: 2) by Immerman on Tuesday May 10 2022, @09:35PM (2 children)
Not really - after all there's very little obscurity involved. A fraudster could just buy a bottle of legitimate pills and have a bunch of valid codes to work with.
I would say it's more security through disproportionate labor requirements, because *using* those codes is another matter altogether. Think of the labor involved in putting a pattern of 92 colored marks in *exactly* the right position on every single pill you manufacture. That's likely to be far more expensive than making the pill itself, severely cutting in to the profit margins that make selling counterfeit pills attractive in the first place.
Plus, if you assume the average person doing a check just dumps out a bunch of pills and snaps a picture, then the counterfeiter needs to use enough different valid codes to make it very unlikely there will be any duplicates in that photo since even a single duplicate would be a dead giveaway of counterfeiting. And an honest pharmacist doing spot checks on their inventory might have thousands of pills in a single photo - that'd take a LOT of copied codes to avoid detection.
Furthermore, if you assume the lookup database ties each code to a specific batch number (which would be useful for tracing any contamination to specific bad batches), then you've got an expiration date as well, which means the counterfeiter needs to continuously acquire new codes for non-expired pills.
(Score: 0) by Anonymous Coward on Wednesday May 11 2022, @08:47PM (1 child)
I read the article as meaning that the pills take on the patterns they get from dipping them in the colored dots, and then you image and make a code for them while they are coming off of the production line, so every single pill is unique and has an entry in a database.
(Score: 2) by Immerman on Wednesday May 11 2022, @10:20PM
That's how I read it too - but that only works for the original manufacturer.
A counterfeiter can't put codes in the database. So they'd have to clone existing codes, a much more expensive proposition.
(Score: 2) by Immerman on Tuesday May 10 2022, @09:53PM (1 child)
>It's an ingenious solution, but a data handling and web app nightmare.
Data handling for sure - a decade ago opioids alone accounted for ~12 billion pills sold in the US per year, while a 10^17 possibility space means at least 57 bits of data per pill, or ~86GB of pill code data per year. Add in everything else and you're looking at a pretty large database to dig through, especially if you want to allow partial-code matches for user convenience - e.g. pills photographed from a single side, or perhaps while sitting atop a thick mirror that will let you get maybe 80+% of the surface in a single photo, or where a couple dots may have come off during shipping.
Of course, maybe you only use this technique on relatively rare and expensive medicines rather than mass-market addict-candy. That would reduce the data size considerably.
The web app though? What's hard about that? At worst you need to embed a variant of the same image-to-code conversion that you use to record pills at the factory, and then send a simple database query for that code. The simpler method would be to simply submit an image and generate the code(s) on the server side, where there's far less chance the "secret sauce" will be reverse engineered. (but comes at increased computational cost to the company)
(Score: 2) by mr_mischief on Friday May 13 2022, @02:55PM
Web apps from actual web tech companies have outages or security issues often enough. This is asking a drug company to have something of huge scale and potentially life-critical import be reliable. It's not impossible, but it's not going to be easy.
(Score: 2, Insightful) by Anonymous Coward on Tuesday May 10 2022, @12:45PM (8 children)
Article said:
"While most of us were baking sourdough bread and watching "Tiger King" to stay sane during the pandemic shutdown..."
Author thinks he/she is like most people, but has no idea of people outside his/her bubble. This is a huge problem with "journalists."
(Score: 2) by cmdrklarg on Tuesday May 10 2022, @01:56PM
Hear, hear. I really could not care less about 'Tiger King'.
Answer now is don't give in; aim for a new tomorrow.
(Score: 2) by DeathMonkey on Tuesday May 10 2022, @03:28PM (2 children)
Except you know exactly what Tiger King is and how it relates to the pandemic.....
Even if you didn't watch it you still understand the reference which is all that actually matters.
(Score: 0) by Anonymous Coward on Tuesday May 10 2022, @04:54PM (1 child)
No, i have no idea WTF Tiger king is.
(Score: 0) by Anonymous Coward on Tuesday May 10 2022, @07:08PM
I think it's a Disney direct-to-video followup to Lion King, but I could be wrong.
I don't watch Disney stuff much.
(Score: 0) by Anonymous Coward on Tuesday May 10 2022, @04:12PM
I was too busy working my ass off keeping people fed to do that stuff. But, people have already forgotten. It's a bit of a kiss that the folks in the supply chain don't get any recognition, but nurses, and for some reason teachers, do.
(Score: 0) by Anonymous Coward on Tuesday May 10 2022, @05:13PM (2 children)
Or you are taking a small quip way too seriously. Search/replace sourdough and tiger king with your new food/show, suddenly relevant!
Why did that reference upset you so?
(Score: 2) by DeathMonkey on Tuesday May 10 2022, @05:33PM
There is literally nothing too ridiculous for them to get triggered over!
I remember when "victimhood mentality" was a bad thing. These guys have managed to turn a Tiger King reference into some kind kind of oppression!
(Score: 0) by Anonymous Coward on Tuesday May 10 2022, @05:38PM
No, it is incorrect and sounds stupid. All it would take is a single word change to make it correct. Replace the word "most" with "some" or even "many." It's just bad writing that reflects an insular viewpoint.
(Score: 3, Interesting) by Immerman on Tuesday May 10 2022, @10:02PM (2 children)
I see one potential problem with this strategy:
Why are so many pills blue? It's not just for marketing reasons, it's because even when tested under laboratory conditions, blue pills reliably work better than the exact same pill in a different color. They're harnessing the placebo effect to increase the effectiveness of medication beyond what the medicine itself can accomplish.
And it's not like the placebo effect is small - it can actually exceed the effect of the medication itself, which is why drug studies have to carefully control for it.
I suspect most people would find it hard to take a "funfetti colored" pill seriously, and thus the placebo effect would instead partially (or even totally!) neutralize the legitimate effectiveness of the medication. Which means higher dosages required, and more severe side effects.
I suppose they could use colors outside the visual spectrum instead, but that would mean you'd need a special camera to photograph them. Perhaps they could use various shades of blue?
(Score: 0) by Anonymous Coward on Tuesday May 10 2022, @10:58PM (1 child)
Or a test subject in a study fires up the app and knows they got a placebo.
(Score: 2) by Immerman on Wednesday May 11 2022, @10:25PM
That's a tiny use case, easily solved by:
- removing the dots
- putting an obscuring shell over the dots
- getting test pills off the production line before they're coded
- removing test pills from the database
The last two are likely the easiest considering that most studies into new applications of a drug that's already on the market are done on the behalf of the company producing them.