How Do They Get Small Amounts Into Meds?

[Uncayimmy]

I've often wondered how they get such small amounts into medicines. For example, hydrocodone (Vicodin) comes in 5mg, 7.5mg, and 10mg pills. As far as I can tell, a milligram is pretty small. It would seem pretty tricky to control the amounts.

Anybody have any idea of how it's done during manufacture?

[Graham Jackman]

Originally Posted by UncaYimmy

I've often wondered how they get such small amounts into medicines. For example, hydrocodone (Vicodin) comes in 5mg, 7.5mg, and 10mg pills. As far as I can tell, a milligram is pretty small. It would seem pretty tricky to control the amounts.

Anybody have any idea of how it's done during manufacture?

The drug substance is mixed with a combination of inert ingredients that allow the formation of a stable tablet, while also ensuring that the drug is released reasonably quickly once swallowed. So a 5 mg pill may well contain 100mg or more of inert ingredients. In manufacture, care is taken to ensure that the drug is uniformly mixed with the other ingredients before the pills are pressed.

[Uncayimmy]

Originally Posted by Graham Jackman

The drug substance is mixed with a combination of inert ingredients that allow the formation of a stable tablet, while also ensuring that the drug is released reasonably quickly once swallowed. So a 5 mg pill may well contain 100mg or more of inert ingredients. In manufacture, care is taken to ensure that the drug is uniformly mixed with the other ingredients before the pills are pressed.

I understand that, but how do they do it? In my example we're talking about (roughly) a few grains of sand per pill. The pill also contains about 100X as much acetaminophen as well as whatever inert ingredients go along with it.

[NewtonTrino]

1. Get a big mixer
2. Dump 5 kilos of the active ingredient and 100kilos of the inert stuff.
3. Mix it up really good.
4. Press it into pill shapes.
5. Profit!!

[Dymanic]

Botox is the one that gets me. Most toxic protein known to man.

[Uncayimmy]

Originally Posted by NewtonTrino

1. Get a big mixer
2. Dump 5 kilos of the active ingredient and 100kilos of the inert stuff.
3. Mix it up really good.
4. Press it into pill shapes.
5. Profit!!

I may be thinking about this too hard, but in my mind I am considering randomness. A random distribution in the long run is not the same as an even distribution in the short run. A prescription of 30 pills may be close to the 20:1 ratio of inert to medicine, but I doubt every pill will have that ratio. So when they say it contains 5mg of whatever, is that on average or really and truly per pill?

[Miss_Kitt]

UncaYimmy == I suspect that the FDA has information on this. I'm sure they require that manufactured drugs fall within a certain range of accuracy.

One possible means would be to use a solution with the correct amount of drug(s) that is then put into each pill mold/material before or during the pressing process.

Anybody here actually in the industry??

--MK

PS I do think it's possible you sometimes get a varied dose. At least, the first Vicodin I ever took alleviated my pain, but it also got me totally wasted. They have not had that effect on me since.

[NewtonTrino]

If you mix it well enough the distribution will be fine (as long as the particles are reasonably small). This really isn't a problem.

And Miss Kitt, it's because you built up a tolerance for the vicodin.

[JJM]

They can, and do, analyze individual tablets. Of course the active ingredient can vary slightly from one to another. The requirement is that the tablets must be +- 5%. So, a 100 mg tablet will have between 95 to 105 mg of drug. That is the batch-to-batch variation, within one batch, the tablets vary much less; but I don't recall the typical figures.

[Merko]

Realize that 5 mg still contains bazillions of molecules. Think about it like pouring a spoon of red colour in a bucket of blue colour and mixing. Random chance just won't cause some corner in your bucket to go all red.

[Miss_Kitt]

Originally Posted by NewtonTrino

...
And Miss Kitt, it's because you built up a tolerance for the vicodin.

You know, NewtonTrino, I'd buy that if I had taken them on an ongoing basis. But I can't believe that a tolerance builds up from ONE use, and the second one had a markedly more minor effect than the first. After that acute injury healed, two years ago, I've only taken one perhaps every 6 or 8 weeks if I have a bad migraine. I don't think tolerance works for something you haven't had in months.

It might have just been that for whatever reason I was already not in the pink on that first occasion--for one thing, I had been in serious pain for hours before we got things diagnosed. But I don't think tolerance per se can be an issue.

Just my thoughts, Miss Kitt

[MRC_Hans]

Ahh, tiny dilutions? Serial dilution, just like the homeopaths do. So you want a 1/10,000 dilution? Then you first make a 1/100 dilution, then make a 1/100 dilution of that. Presto, a nice 1/10,000 dilution.

Hans

[ArmillarySphere]

Originally Posted by MRC_Hans

Ahh, tiny dilutions? Serial dilution, just like the homeopaths do. So you want a 1/10,000 dilution? Then you first make a 1/100 dilution, then make a 1/100 dilution of that. Presto, a nice 1/10,000 dilution.

Just remember to shake vigorously for 100 times in each direction...

For some reason, I have a mental image of a lab-ful of chemist boogieing wildly with test tubes in their hands.

[MRC_Hans]

Originally Posted by ArmillarySphere

Just remember to shake vigorously for 100 times in each direction...

For some reason, I have a mental image of a lab-ful of chemist boogieing wildly with test tubes in their hands.

Well, yes. Effective mixing is also necessary. You know all this is very old hat, so not even Hahnemann invented it, he just dubbed it 'potentization' and assigned magic powers to it.

Hans

[ponderingturtle]

Originally Posted by UncaYimmy

I may be thinking about this too hard, but in my mind I am considering randomness. A random distribution in the long run is not the same as an even distribution in the short run. A prescription of 30 pills may be close to the 20:1 ratio of inert to medicine, but I doubt every pill will have that ratio. So when they say it contains 5mg of whatever, is that on average or really and truly per pill?

It is an average but is very tightly controled, by measureing samples.

Why do you question that ratio being pretty uniform?

[paximperium]

Originally Posted by UncaYimmy

I've often wondered how they get such small amounts into medicines. For example, hydrocodone (Vicodin) comes in 5mg, 7.5mg, and 10mg pills. As far as I can tell, a milligram is pretty small. It would seem pretty tricky to control the amounts.

Anybody have any idea of how it's done during manufacture?

We use very small people to manufacture these pills. Pixies are the cheapest but liliputians are decent although the Republicans have issues with their visas and illegal immigrant status.

[ElMondoHummus]

Originally Posted by Miss_Kitt

You know, NewtonTrino, I'd buy that if I had taken them on an ongoing basis. But I can't believe that a tolerance builds up from ONE use, and the second one had a markedly more minor effect than the first. After that acute injury healed, two years ago, I've only taken one perhaps every 6 or 8 weeks if I have a bad migraine. I don't think tolerance works for something you haven't had in months.

It might have just been that for whatever reason I was already not in the pink on that first occasion--for one thing, I had been in serious pain for hours before we got things diagnosed. But I don't think tolerance per se can be an issue.

Just my thoughts, Miss Kitt

Maybe it was that whiskey shooter you chased it with.

Whoa! I was kidding! I was kidding!!

[ElMondoHummus]

On a more serious note: From what little I can find with a quick whack at Google, it seems as though the accuracy of individual doses is something that's accomplished in two ways:

1. Really, really accurate measurement instrumentation; I found an ad for some gas flow meter used in pharmaceutical manufacturing that boasted:
   Quote:

   +/-0.05% mass and volume flow accuracy, and up to =?- 0.0002 g/cc density accuracy...
2. Lots of in-process measurement checks. This I conclude from the plethora of
   1. Job postings for such positions
   2. Sources like this that explicitly state:
      Quote:

      Each critical step in the manufacturing process should be done by a responsible individual and checked by a second responsible individual. if such steps int he processing are controlled by automatic mechanical or electronic equipment, its performance should be verified.

Now, granted, your question was on the specifics of how they do it. I confess, I found nothing that got down into the actual processes. All I found was that there's a high degree of automation - which makes sense - and a high degree of human accuracy checking at each "critical step" of manufacture. I'm guessing at this, but I presume many of the processes that guarantee that 7, not 5 or 9 milligrams of active ingredient end up in each pill are either so highly specialized that information on it is only available through insider sources of information (i.e. industry publications, which I haven't perused), or are manufacturers "secrets". Either way, all I could tell is that the procedures appear to this outsider to be rigorous and highly automated. How they guarantee dose accuracy is something that has to be deduced; I can't find anything directly addressing it.

[EHocking]

Originally Posted by UncaYimmy

I've often wondered how they get such small amounts into medicines. For example, hydrocodone (Vicodin) comes in 5mg, 7.5mg, and 10mg pills. As far as I can tell, a milligram is pretty small. It would seem pretty tricky to control the amounts.

Anybody have any idea of how it's done during manufacture?

You could try this site.

www.manufacturingchemist.com

There are a number of technical articles that would probably come close to answering your question.

[Capsid]

5mg is still something you can see and weigh out on a fine balance for most chemicals.

[Uncayimmy]

Originally Posted by Miss_Kitt

One possible means would be to use a solution with the correct amount of drug(s) that is then put into each pill mold/material before or during the pressing process.

I thought about that, too. Seems like a good way to be absolutely sure about concentration.

[joobz]

There are multiple ways of distributing the API (active pharmaceutical ingredient) in a dosage form. You have granulation processes, wet granulation, extrusion, gel formulations...

Each have their own set of pluses and minuses.


To Pondering turtle, the idea of uniform distribution isn't a given for a power mixture. The surface forces at play in a powder can result in a high degree of cohesive (like particles sticking together), which hinder an even mix.

If interested, I can go into more details of the formulation process and how industry is changing to meet with modern needs.

I'm not in the industry, but I do know quite a bit about it.

[Madalch]

Originally Posted by Capsid

5mg is still something you can see and weigh out on a fine balance for most chemicals.

True, but you wouldn't want to have to get exactly 5 mg.

I used to regularly weigh out similar quantities of compounds, and I would simply add an appropriately small spatulaful of compound to the reaction vessel (an NMR tube), write down the mass I had actually put in, and adjust the other components to match. If it was below 5 mg, I'd add more, but I wasn't about to take any out if I'd put in 9.7 mg. This worked fine for my project, but that wouldn't work for pills.

Liquid compounds are easier to measure small quantities of, since microlitre syringes are common.

For real precision work, you need to make up a standard solution. To make the pills, they'd make a homogeneous mixture of active ingredients and inert material. And they'd put a lot of work into making it homogeneous, so that there would be very little variation from pill to pill.

[the_smasher]

It can be tricky to control, but it really comes down to ensuring a good formulation before starting. For example, magnesium stearate is an excipient that‘s used a lot in tablet manufacture because it lubricates the blend and helps prevent clumping and sticking. There are so many different techniques and tricks it would be tough to list them all. Once a formulation is set it can be as simple as dumping (in certain order and amount based on your formulation) the excipient and API into a big “v” shaped blender and mixing it up. Analytical tests can be performed on the blend prior to compression or encapsulation to make sure it’s uniform. If blend uniformity fails the batch gets destroyed.

We don’t make a lot of commercial product where I work but we manufacture a lot of material for clinical trials. Each lot gets a specification and is tested prior to release. We calculate a % of label claim and content uniformity result using HPLC. Spec for tablets and capsules is usually 90-110% of label claim so your 5mg tablet could have between 4.5 and 5.5mg and still make it out the door.

Our microbalance is readable to 0.001mg so 5mg is generally not a problem for us, but a manufacturing run might be hundreds of thousands of tablets….they don’t need to worry about small scale weights.

[Zeuzzz]

Originally Posted by UncaYimmy

I've often wondered how they get such small amounts into medicines. For example, hydrocodone (Vicodin) comes in 5mg, 7.5mg, and 10mg pills. As far as I can tell, a milligram is pretty small. It would seem pretty tricky to control the amounts.

For amounts this size high precision scales can be used. I use them all the time, up to 1mg accuracy is pretty common. You can get ones that measure 100 ug, and even 10ug (really expensive), but I haven't seen any go down to one ug. I think that down at those weights things like vibrations from varying heat, wind, atmospheric pressure and other things effect the readings too much. You would need to use a vacuum.

People who need 1-100ug amount of chemicals usually use a material that they know how much of a substance it will absorb per given surface area and just use this without weighing anything. Thus why people use blotters for LSD, etc.

[shadron]

Two things I can think of:

First, have you ever watched a small cement mixer work? That is roughly the scale of operations for this, and such a mixer, when mixing dry substances, can achieve really thorough mixing after running for ten minutes. I would imagine that there are studies done by the equipment manufacturer that specifies the length of run needed to assure any given standard deviation from the mean proportionality, giben the grain size and dryness, etc.

Second: If needbe, they can start by mixing the active ingredient with it's own mass of inert ingredient, and after that is mixed, then add in that much more of the inert, mixing at every step in the process.

[Miss_Kitt]

I love learning stuff here!!

And El Mondo -- If you're coming to TAM, you owe me a cuppa!

[PbFoot]

In high school I did some experiments measuring the amount of ASA in various brands of asprin. In every case, it was exactly as specified on the label to 3 decimal places, which was the limit of the equipment we had.

- PbFoot

[wollery]

In regards to the thread title, I suggest they use very small measuring spoons.

[ElMondoHummus]

Originally Posted by Miss_Kitt

I love learning stuff here!!

And El Mondo -- If you're coming to TAM, you owe me a cuppa!

... cuppa whiskey??

[Uncayimmy]

Thanks for all the responses. I managed to find a book called, Pharmaceutical Dosage Forms: Tablets, Third Edition Volume 1: Unit Operations and Mechanical Properties By Larry L. Augsburger, Stephen W. Hoag Edition: 3, illustrated

If you think rocket surgery is complex, take a look at this stuff! There's all sorts of statistical math involved along with analyzing particle sizes and shapes (among other factors) to figure out how things need to be blended and made into tablets. There are lots of different sampling techniques.

It's not quite like baking a cake.

[KingMerv00]

Originally Posted by JJM

They can, and do, analyze individual tablets. Of course the active ingredient can vary slightly from one to another. The requirement is that the tablets must be +- 5%. So, a 100 mg tablet will have between 95 to 105 mg of drug. That is the batch-to-batch variation, within one batch, the tablets vary much less; but I don't recall the typical figures.

Measuring pills and solutions was my job for a few years.

The percent range varies a lot depending on the drug in question. I've seen limitations as high as 10% and as low as 2%. That being said, I've never seen an out-of-spec result in a non-experimental batch.

[Miss_Kitt]

This thread is a wonderful example of what the JREF is about!!

[joobz]

The big push in Pharma now is "process analytical technology", PAT for short.

The standard of analysis was offline. You'd take a statistical sample (pull randomly a couple pills*) and evalaute thier dissolution profile, their friability, and thier mass. The dissolution profile amounts to taking the pill, dropping it into a vessel of specific size, geometry and containing a buffer solution, and then measuring the amount of API coming out over time. You have some standard curve that you'd compare to to elvauate if it is within spec. Also, you'd have some maximum dose concentration that would let you know your loading efficiency.

The obvious problem to this is that 1.) it's only effective for batch processes and 2.) it's slow.

That's were PAT comes in. PAT, you do some online measure of the dosage form which allows you to measure EVERY pill rather than a statistical sample. You also gain the advantage that if you have PAT on a mixing step, you can switch to continous processing, rather than batch processing.

PAT relies heavily on multivariant statistical analysis and novel spectroscopic methods. The cool one that I saw was a near-IR detector placed on the lid of a V-blender. Then the blender was switched on. Each time the vessel was inverted, the powder mixture would slam on the surface of the lid and you can measure the near-IR signal. This way, you simply keep rotating the mixer until the signal no longer varied within some tollerance range.



*I'm using pill, but in truth it's tablet or gel cap or some other dosage form. A "pill" is a through-back to old-school pharmacy when the pharmacist would mix up the drugs. He'd make a thick dough (think playdough) roll it out to a certain thickness and then cut it to the right length. This would be your dosage.

[skeptsci]

Originally Posted by UncaYimmy

I've often wondered how they get such small amounts into medicines. For example, hydrocodone (Vicodin) comes in 5mg, 7.5mg, and 10mg pills. As far as I can tell, a milligram is pretty small. It would seem pretty tricky to control the amounts.

Anybody have any idea of how it's done during manufacture?

Mixing solids to receive a fairly homogeneous mix is not obvious especially if done on a hundred kilos scale. However, this is not "rocket science" and there are special pharmaceutical mixers with performance that comply with Quality Assurance qualifications. As to the randomness of the process, the amount of the active ingredient amids the excipients (the non-active ingredients) is an average with tight enough variance (say +- 5%). So, voila!

[joobz]

Originally Posted by skeptsci

Mixing solids to receive a fairly homogeneous mix is not obvious especially if done on a hundred kilos scale. However, this is not "rocket science" and there are special pharmaceutical mixers with performance that comply with Quality Assurance qualifications.

Assuming you worked out the formulation standards for your API.

Originally Posted by skeptsci

As to the randomness of the process, the amount of the active ingredient amids the excipients (the non-active ingredients) is an average with tight enough variance (say +- 5%). So, voila!

5% is tight variance???? Haven't you heard of 6sigma?

[skeptsci]

Originally Posted by joobz

5% is tight variance???? Haven't you heard of 6sigma?

I am talking real life pharma and not sci-Fi. 30mg or 31mg, same thing. There is something called therapeutic window (this is an index for estimation of drug dosage which can treat disease effectively while staying within safety limits). In most cases we're talking about a window of ten meaning 10 orders of magnitude! You wouldn't like it to be otherwise.

[luchog]

Originally Posted by Miss_Kitt

You know, NewtonTrino, I'd buy that if I had taken them on an ongoing basis. But I can't believe that a tolerance builds up from ONE use, and the second one had a markedly more minor effect than the first. After that acute injury healed, two years ago, I've only taken one perhaps every 6 or 8 weeks if I have a bad migraine. I don't think tolerance works for something you haven't had in months.

There are any number of individual factors that can vary not only between people, but in the same person over time, even a very short period of time.

And yes, one single dose can precipitate a tolerance or sensitization effect in some individuals. Not necessarily a significant one, but combined with other factors, the end result can be significant. Medicine is not a precise science.

[Uncayimmy]

Originally Posted by skeptsci

I am talking real life pharma and not sci-Fi. 30mg or 31mg, same thing. There is something called therapeutic window (this is an index for estimation of drug dosage which can treat disease effectively while staying within safety limits). In most cases we're talking about a window of ten meaning 10 orders of magnitude! You wouldn't like it to be otherwise.

10 orders of magnitude? I don't think so. A factor of 10? Sure, but that's one order of magnitude. According to Wikie diazepam is fairly forgiving with a therapeutic index of 100 (two orders of magnitude).

According to Wiki the therapeutic index is TD50/ED50 where TD is the Toxic Dose in 50% of the people and ED is the Effective Does in 50% of the people.

BTW, the dosages of Vicodin are 5mg, 7.5mg, and 10mg. I can attest to there being a big difference between 5mg and 10mg, which is what got me thinking about it.

As for your comment about just buying a machine and dumping in some chemicals, it's way more than that. Check it out in Google Books.

[skeptsci]

Originally Posted by UncaYimmy

10 orders of magnitude? I don't think so. A factor of 10? Sure, but that's one order of magnitude. According to Wikie diazepam is fairly forgiving with a therapeutic index of 100 (two orders of magnitude).

According to Wiki the therapeutic index is TD50/ED50 where TD is the Toxic Dose in 50% of the people and ED is the Effective Does in 50% of the people.

BTW, the dosages of Vicodin are 5mg, 7.5mg, and 10mg. I can attest to there being a big difference between 5mg and 10mg, which is what got me thinking about it.

As for your comment about just buying a machine and dumping in some chemicals, it's way more than that. Check it out in Google Books.

Sure sure I stand corrected we're talking about a factor of ten - sorry slip of the pen. 5mg and 7.5mg is a 50% difference but 5 and and 5.25 - same thing or they cannot make a product out of it. And as to the mixing I never put it the way you did "just buying a machine and dumping in some chemicals," although it's not much more than a smart powder mixer (can get pretty smart but that's not the point) and a very controlled production schedule. I come from the production floor (in my past). Nothing there that you cannot read about and you did.
What's more interesting is the equipment used in biology labs to handle liquids. They have pipettors that can (routinely) accurately and precisely enough handle volumes of 1 micro-liter (10^-3 ml) and this can be either manual or robotic.