What happens if you add milk to cola? We are conducting an unusual experiment. What happens if you mix cola with milk

Highly frequently asked question both to my channel (Youtube) and to the blog, this is the compatibility of coolants. Namely - what will happen if you mix antifreeze? And then there were variations - one manufacturer, but different colors. Same colors but different manufacturers. Different standards, such as G11, G12, G13, etc. In general, I constantly answer these questions, and they constantly ask me them. Therefore, today I want to write this article, in which I will answer everything at once, as there will usually be a video at the end. The information is helpful, so read on…

I would like to note right away that we will also talk about mixing, because despite the different names, these two liquids are also very similar.

What is the same in the compositions?

Well, what about me at the very beginning again I want to repeat - guys, all antifreezes, especially in the G11 and G12 standards, are very similar in their base. What I mean is, 80% of the composition of both red and green and blue antifreeze is the SAME. Usually it is ethylene glycol +. The remaining 20% ​​(and possibly less), these are already additives that characterize this or that behavior, I also have an article about this,.

That is, if you mix different antifreezes, different colors and categories, then they will be 80% similar.

What's different?

The difference, as I said above, is the additives. That is, various additives are added to the same liquid and it turns out desired composition. It is worth noting that additives in the compositions are needed precisely in order to remove the negative destructive effect of ethylene glycol and water, because this combination is extremely active and can destroy any metal surface, and even more so a tube or radiator wall. And additives hold back this ardor, removing the negative impact.

Roughly characterized, now there are only two types of additives:

• It's protective . They protect tubes and pipes inside, forming a film on their surface that prevents metal parts from collapsing. Mainly used in G11 and our TOSOLs.
• It is anti-corrosion . There is no film formation here, but it turns out that all the work happens when rust begins to appear. These additives block the hearth by simply sealing it. Applicable in G12 and G12+.

In fairness, now there is also a third type - hybrid additives (G13 antifreezes), this is when two protective and anti-corrosion effects are combined at once, that is, they are simply mixed in the right proportion.

About color

The color of antifreeze is more of a distinguishing component. As a rule, now, it does not carry any semantic load. Although many manufacturers, including Volkswagen, tried to introduce color differences in antifreezes, they even had their own recommended colors.

So G11 - almost always was green.

G12 - red (well, or bright orange)

G13 - purple

Although many manufacturers now do not follow the path of Volkswagen and paint antifreeze in the colors they please, this is because there is no strict standardization. So G11 can be either blue or reddish. G12 - green. .

Colors were introduced so that the driver could easily and easily distinguish between antifreezes, but the confusion between manufacturers is now becoming more and more confusing for the average layman.

What happens if you mix green, red, yellow (purple) antifreeze from different manufacturers?

YES actually nothing will happen, you can pour and not even be afraid, if they maintain their standard, then ABSOLUTELY NOTHING WRONG WILL HAPPEN. Such questions arise when, say, your branded antifreeze is over, there is nowhere to buy it (for example, you are on a trip), but red ones from another manufacturer are sold.

So G11 green (of one manufacturer) can be mixed with G11 green of another manufacturer. The main thing is that the standards are similar.

As you may have guessed, G12 can be mixed with G12 from another manufacturer.

You can do the same with G13, that is, pour yellow into yellow or purple into purple.

What happens if you mix different colors

We recall what I said above - the characteristic may be one, but the colors may differ. For example, I have personally seen the G11 in both blue and green. THEY CAN BE MIXED, NOTHING WRONG WILL HAPPEN.

Here I would also like to note that our Russian TOSOL is nothing more than G11 antifreeze and can be mixed with green or blue from another manufacturer.

Everything is the same, you can do it with G12. If they have different colors, then this does not mean anything! The characteristics are the same, which means mixing is possible.

G13 is now very much fooling the brain of my readers. The thing is that there are only two primary colors - yellow and purple. And many are simply afraid to mix them. Guys, if there is an inscription of the standard, then there is simply nothing to be afraid of, do not be afraid to mix. Color is just a dye.

Is it possible to mix different characteristics, for exampleG11 andG12

Here you already need to think, although again, nothing terrible, most likely will not happen. BUT you need to understand there is a subgroup of G11 and G12, but there is a completely different option - this is G13.

If we take the first subgroup, then mixing will lead to the fact that in the final liquid there will be additives, both protective and anticorrosive. Although properly you will not be able to control the mixing. Precipitation most likely will not fall out, BUT you need to understand that adding other additives not designed for your car, in particular radiators, can worsen cooling. WHY? YES, because green antifreezes envelop the tubes with a film inside, which prevents the engine and other units from cooling. That is, if you suddenly added to red antifreeze, say green or blue, then be prepared for the fact that temperature regime may fall. All this will also happen and vice versa, if you add blue or green (G11) to red (G12), then the characteristics of the liquid also fall.

Is it possible to mix yellow and green (red) antifreeze

There is already a slightly different situation, namely the yellow and purple compounds of G13, these are completely different substances. What I mean.

In red and blue (green) versions - the main mass fraction It's distilled water + ethylene glycol.

And in the yellow and purple version - the main mass fraction is propylene glycol + distilled water.

THAT is here even the base is different! Ethylene glycol (toxic) was replaced with propylene glycol (safe), these are two monohydric alcohols, they were changed only due to the removal of the toxic component.

Also, G13 contains two types of additives - immediately anti-corrosion + protective.

What happens when you add red or green to yellow antifreeze?

Nothing really good:

• WE just do not know how ethylene glycol and propylene glycol will react with each other, even if they are similar, BUT IT IS NOT THE SAME.
• Additives that are in the yellow (purple) version are designed for propylene glycol, how they behave with ethylene glycol is not known!
• Also, we do not know if additives are compatible in general!

Hence the moral to mixG11 andG12 (G12+) withG13, I wouldn't! There is Great chance that a precipitate may fall out, because nevertheless this is not the same thing.

But G13 purple and yellow, YOU CAN mix WITHOUT FEARING!

Fake antifreeze

Actually, it seems to me, the topic is fully disclosed! BUT what I want to say in the end - have you ever wondered why some formulations, say branded or serious companies, are expensive, but there are formulations that are very cheap?

The same G13 antifreeze, made under the patronage of Volkswagen (usually purple), can cost 300 rubles per liter

And the yellow one, let's say the same G13, produced in some unknown place and not clear by whom, can cost the same money only for a 5 liter canister.

The whole point here is a “fake”, because a high-quality liquid will not be cheap, it will also meet all the characteristics and safety requirements, such as:

• Will boil over 100 degrees (typically 105 to 110 degrees)
• When boiling, will not ignite
• Will not lose performance

When buying cheap antifreeze, you simply cannot check it, and it is not uncommon when it boils much below 100 degrees! Because of what, your engine can simply overheat and fail.

And when mixing cheap and expensive options, always think, HAVE YOU CHECKED THE QUALITY OF AN INEXPENSIVE PRODUCT? Does it comply with the regulations? Otherwise, you will simply spoil the company's expensive and high-quality liquid!

Therefore, most antifreezes can be mixed, but everything must be done wisely!

This concludes, read our AUTOBLOG.

The unique taste of Coca-Cola is known, perhaps, by every inhabitant of our country. After all, this drink simply flooded domestic supermarkets in the early 90s. However, few people know how and from what it is prepared. delicious drink. In this article, you will learn what happens if you add milk to cola. After all, this experiment will help open your eyes to the composition of the famous soda.

Coca-Cola - a taste familiar from childhood

Non-alcoholic unusual black color was born back in 1886. At the same time, the well-known logo was invented, made in calligraphic font.

As early as 1902, Coca-Cola became the most popular drink in America. But his fame came to our country much later. In 1988, the first production of Coca-Cola was opened at the Moskvoretsky brewery. After that, a whole era of Western products began in Russia and the CIS countries.

But have you ever wondered what this carbonated drink is made from? What happens if you add milk to cola? After all, this simple experiment is able to show what happens to the drink after drinking and what its real color is.

What happens if you add milk to cola?

To conduct the experiment, you will need a bottle of carbonated drink and some milk. It is enough to mix two products, and after 40 minutes you will see the result. If cola is added to milk, then chemical reaction. The caffeine and the coloring in the soda will coagulate, leaving a cloudy yellowish liquid in the bottle.

This is due to the content in Coca-Cola, when interacting with which milk curdles. That is why during the experiment, you can also notice the formation of foam at the neck of the bottle. Thus, if you add milk to Coca-Cola, you get a liquid of a very unpleasant consistency.

The manufacturer indicates that the drink consists of sugar, caffeine, acid, and flavorings. That is, there is no question of natural ingredients, and we use an ordinary sweet chemical mixture. And if, after going to McDonald's, you order milk shake? That's right, the same unpleasant picture is formed in the stomach as after. After all, any alkaline environment when mixed with an acid will give a predictable reaction.

Will you be able to drink Coca-Cola after such an experiment?

Anyone who has ever carried out such a thing is firmly convinced that a carbonated drink is very unhealthy. After all, what happens if you add milk to cola? Get the real color of everyone's favorite drink. That is, in fact, Coca-Cola is a cloudy yellowish liquid that is simply colored black.

When ingested, the drink gives the same reaction as when mixed with milk. In addition, we cannot know for sure what happens when it interacts with other products. But even the fact that one glass of Coca-Cola contains ten teaspoons of sugar makes you wonder. 'Cause it's more daily allowance for an adult.

Other experiments with Coca-Cola

Now you know what will happen if you add milk to cola. A photo of this unusual experiment is presented in our article. But what other experiments can prove that this drink is harmful to our body?

For example, the experiment with meat and Coca-Cola is very popular. It has been proven that if you put a piece of meat in a glass of carbonated drink, it will dissolve after 24 hours. This once again proves that the effect of cola on is not the most favorable.

Not so long ago, we observed a whole sensation produced by the experiment with Mentos dragees. Indeed, when mixed with Coca-Cola, a very violent reaction occurs with the release of a large amount carbon dioxide. Some claim it can be fatal.

It is quite difficult for people to give up the usual carbonated drinks, because without them life seems not so sweet. But simple experiments prove how detrimental Coca-Cola affects our internal organs. Therefore, before heading to the supermarket for another portion of colored soda, remember the experience with milk. Perhaps this will save you from a pointless and very harmful purchase.

Most people are big fans of Coca-Cola products. And surely most of them know that these products are far from harmless. There are an incredible number of legends about how Coca-Cola corrodes meat products, washes away scale and rust, removes blood stains, cleans battery contacts, removes grease stains from clothes, or even kills sperm. Most of these facts have already been verified by legend busters. What is Coca-Cola really? Red poison or safe food?

More recently, I happened to be faced with one fact that milk can discolor cola, leaving only natural ingredients. Then I wanted to check it myself:

I diluted the most ordinary bottle of Coca-Cola with milk (3.6% fat, but this is not so important).

While waiting for the result, I happened to find several entertaining facts about Coca-Cola products.

What will happen to your body after drinking Coca-Cola?

After 10 minutes:

10 teaspoons of sugar will hit your system (this is the daily recommended amount).

In 40 minutes:

Caffeine absorption is complete. Your pupils will dilate. Blood pressure will increase because the liver is throwing out more sugar into the blood. Receptors are blocked, thereby preventing drowsiness.

After an hour:

Phosphoric acid binds calcium, magnesium, and zinc in your intestines, boosting your metabolism. Increased excretion of calcium through the urine.

More than an hour and a half later:

You become irritable or lethargic. All water contained in Coca-Cola is excreted through urine.

An hour later, I decided to look into my bottle...

I would like to note that the components of the drink are:

Dye: Sugar color (E150)

orthophosphoric acid(E338)

—Caffeine

natural flavors

Carbon dioxide (E290)

Caffeine and acid in the composition suggest that cola - acid environment. Milk is an alkaline environment. Molecules of phosphoric acid enter into an oxidative reaction with milk and precipitate, as a result of which cola becomes clear liquid. And we can observe floating flakes (sediment).

Result after 3 hours of inactivity...

By the way, moonshine is also cleaned from "fusel" (fusel oils) with potassium permanganate and gauze.

The cooling system is an integral part of any engine internal combustion. Without it, the motor will simply boil. It is the cooling system that takes all the heat from the engine and removes it to the outside. Thus, a stable temperature of the internal combustion engine is maintained. An important component of this system is the coolant. It can be antifreeze or antifreeze. These fluids serve for years, but in the event of a repair of the SOD (for example, replacing a thermostat), their level often drops. To resume it, you need to add new fluid. But not always coolant with such characteristics can be found on the store shelf after 3-5 years. Then the question arises for car owners: what will happen if you mix antifreeze with antifreeze? Let's try to deal with this in our today's article.

Characteristics of antifreeze

So, first consider what each of the liquids is. Antifreeze is a coolant that was developed in the 70s of the last century in the Soviet Union. Moreover, it was developed specifically for the Soviet Zhiguli. VAZ "Kopeyka" - the first car in the USSR, in the cooling system of which not water was used, but antifreeze brand A-40.

Compound

This liquid consists of ethylene glycol and an additive package. also in in large numbers(about 50 percent) it contains distilled water. Such a liquid has the ability not to freeze at a temperature of -40 degrees. But today, most manufacturers are guided by an indicator of -35 degrees Celsius.

About antifreeze

Antifreeze is a purely foreign development. The properties of these liquids are almost the same. The composition contains distilled water, a package of additives and ethylene glycol.

There are some differences in antifreezes of the 13th group. It uses organic, not chemical additives. But anyway the main task liquids - to ensure high-quality heat dissipation and not freeze when low temperatures. Usually antifreeze does not freeze at -40 ° C. And the boiling point is about 120 C o. For comparison, antifreeze already boils at 110 C o.

About additives

It is worth noting that each coolant manufacturer uses a different package of additives in different volumes. Therefore, it is not always possible to mix antifreeze with antifreeze. In total, three additive packages are used in the coolant:

• Anti-corrosion. Usually these are acidic compounds designed to block the centers of corrosion and prevent them. further development.
• Protective. They form a thin film inside radiators, cooling jackets and pipes.
• Antifoam. To prevent the composition from foaming during operation, these additives are used. Why should these additives be used in every coolant? The presence of foam can lead to rapid engine overheating and cylinder head failure.

hybrid additives

We also note that in modern types antifreeze (does not apply to antifreeze), hybrid additives are used. They contain both protective and anti-corrosion components.

About types of antifreeze

It is important to know that this liquid, unlike antifreeze, is divided into several groups:

• G11. This is the earliest group. Now such antifreezes are rarely found on sale. What color are they painted? The liquid from this group is dyed in green color or blue (the same as that of antifreeze). The composition contains both organic and chemical additives, including carboxylic acid. The latter effectively fights corrosion foci. However, the liquid forms a very thick protective layer, due to which the thermal conductivity deteriorates. Also, the compositions of this group have the shortest service life among all types of antifreeze. The resource of green and blue liquid is no more than 60 thousand kilometers. The same period for domestic antifreeze. If the replacement is not made in time, the additives will crumble and clog all the small channels.
• G12. This is one of the most popular bands. Antifreezes of this brand are used on most modern cars. They are usually colored red. They are a more advanced liquid, the service life of which is about 150 thousand kilometers. Antifreeze of this group has good heat-conducting properties. However, it does not boil high temperatures and also does not form flakes at the end of its service life. This coolant consists of ethylene glycol and a package of organic additives. Among the inorganic, it is worth noting only carboxylic acid (and even then in a small amount). This group is great for copper and brass radiators. However, it is not suitable for aluminum, since rust is more likely to form there due to a lack of carboxylic acid.

Please note: Japanese G12 antifreeze has slightly different characteristics than the European one. Such a coolant is not designed for operation in harsh conditions, so its crystallization temperature is -25 degrees. This must be taken into account when buying a car from Japan, which will be operated in the northern regions of Russia.

• G13. This group appeared relatively recently, in 2012. Its main difference from other antifreezes lies in the composition. This coolant does not contain the usual ethylene glycol. Here it contains propylene glycol, which is less aggressive with respect to environment. But the additives for G13 antifreeze remained the same. It is an organic and part of a carboxylic acid. Usually antifreezes of this group are painted purple.

As you can see, each group has its own nuances in composition. Somewhere there are fewer additives, somewhere there are more of them. But what happens if you mix antifreeze with antifreeze? The effect will be different. The result depends on which antifreeze to mix with antifreeze. We will consider the consequences below.

Mix antifreeze G11 with antifreeze

Hypothetically, coolant data have the same "base". But you need to remember that the manufacturers here are different and each produces liquid according to its own technology. What happens if you mix antifreeze and green antifreeze from this group? As tests have shown, these liquids do not foam and do not change their characteristics. Since the composition is the same, this coolant can be used later on the engine. For example, what happens if you mix blue antifreeze and blue antifreeze. Such a liquid is absolutely suitable for use, does not foam and does not boil prematurely.

However, you need to remember about the service life. Do not think that when adding new antifreeze, the liquid will last longer. Part of the additive package will remain the same. Only their percentage of content in general will decrease. Therefore, the replacement period for the updated fluid is the same as for the previously filled one. For example, if a new antifreeze was mixed with a year-old antifreeze, such a coolant should be changed after 12 months. Otherwise, the additives will not work well and will precipitate. Flakes are formed that will clog the cavities of the cooling jacket.

We dilute antifreeze with the G12 group

Now let's talk about red antifreeze. As we said earlier, it has a slightly different additive package. If you want to mix red antifreeze with antifreeze, you should not do this. Especially if most the liquid in the system is precisely the domestic coolant (A-40). What happens if you mix antifreeze with antifreeze? When diluted, the liquid will acquire dark color. There will also be small oil marks on the surface. This indicates that a reaction of different additive packages has occurred. The fact is that the liquid from the twelfth group envelops the walls with a protective film in a slightly different way. Therefore, it is impossible to predict the effect of different additives when they are mixed. Those who have already tried mixing these liquids say that there is no detrimental effect. But still, the composition of such coolants is different, and you should not risk adding antifreeze to such a composition.

Is it possible to mix antifreeze and antifreeze of the 13th group?

Often, such coolant is not bought for topping up, since it is several times more expensive than you. It's easier to buy two new A-40 canisters and pour them into the SOD. But still, what will happen if you mix antifreeze with antifreeze from the 13th group? Since the composition of these products is significantly different, when mixed, both liquids will react. Some will say that propylene and ethylene glycol have the same base (monohydric alcohols), and this will not cause any harm.

But it's not. No one can guarantee the performance of additives. And it is visually noticeable how these liquids change color and concentration. What happens if you mix antifreeze with antifreeze from the 13th group? The hybrid additive packages of the latter are simply destroyed, and such a liquid is not able to properly remove heat. In addition, the radiator will corrode, and the channels of the cooling jacket will become clogged with sediment and flakes. What happens if you mix antifreeze with G13 antifreeze? The boiling point of such coolant will drop to +100 degrees. For some modern cars, this is still the operating temperature. Therefore, when using a liquid with such an unusual composition (propylene and ethylene glycol mixed together), the motor may simply boil. This is fraught with deformation of the cylinder head. On 90% of the car, it is aluminum and difficult to repair. The cost of restoring such an internal combustion engine is measured in tens of thousands of rubles.

Actions in case of incorrect mixing

If I mixed antifreeze with antifreeze, what should I do? If the domestic liquid was added to the coolant from the 12th or 13th group, it is no longer subject to further use. This concentrate must be drained from the system. How to do it? To do this, you need an empty container of 8-10 liters. We substitute it under the bottom of the radiator and unscrew the tap on the last one (on some cars it may be a bolt). So that a vacuum does not form in the system and the mixture comes out faster, unscrew the cap of the expansion tank.

What's next?

Then it would be possible to start filling in a new coolant, but since we had an unknown mixture, it is recommended to rinse the container from the inside. How to do it? We twist the faucet or bolt on the radiator and fill it with distilled water to an average level.

Then we start the engine and let it run for 5-10 minutes. In order for the liquid to circulate better through the system (and, accordingly, to bring out more of the old “slurry”), you can increase the speed of the pump by pressing the accelerator pedal or turning on the stove. Ideally, flushing should be extended over a day. That is, with such a liquid it is necessary to travel a couple of times around the city.

At the next stage, you need to drain this water in the same way, substituting an empty container under the bottom, and unscrew the tap. And after that, you can fill in a new coolant. It's antifreeze or antifreeze, it doesn't matter. The main thing is that the product does not contradict the manufacturer's recommendations and is not mixed with another liquid. The exact amount of coolant used should be found in the operating instructions. Having filled the liquid to the maximum level, we warm up the engine again. Some of the antifreeze will leave the tank, and this is normal. We turn off the engine and restore the liquid to normal level(this is the mark between "MIN" and "MAX").

Summing up

So, we found out what happens if you mix antifreeze with antifreeze different brands. As you can see, these fluids react differently to mixing. Therefore, you need to dilute very carefully.

If a we are talking about adding antifreeze to antifreeze, only a green or blue product from the 11th group should be used. At the same time, topping up does not lead to an increase in the service life of the coolant. Conversely, if there is a domestic liquid in the system, only G11 antifreeze needs to be added. Products from the 12th and 13th groups should not be mixed with antifreeze. This may lead to sad consequences, which will lead to best case to the need to flush the SOD channels. At worst, the motor itself will overheat, and then repair of the block head will be inevitable. In general, it is safest to mix the coolant with distilled water. It is in every group of antifreeze, and even more so in antifreeze. The difference will be only in the freezing temperature. Therefore, at high concentrations, such a liquid should be replaced on the eve of winter.

This pill for pressure, this - oral contraceptives and this lowers cholesterol. Are you sure that all these medicines can be taken together?

0:11
You go to the doctor and get tested. The doctor tells you that you high cholesterol and that you need to take certain medications. You receive a vial of pills. You have no doubt: the doctor is sure that the medicine will work. The company that produces it has conducted research, has been certified, the drug has been carefully studied and approved. They have general idea about its action, a rough idea of ​​the side effects. Should be fine. You talked to your doctor a little more, and he became worried, because you are depressed, you do not feel very good, you have ceased to enjoy the usual things. He says: “You know, I think you have. I'll prescribe more pills for you."

0:59
So now we are talking about two drugs. And the second is the same - millions of people take it, the company has done research, they have passed all the tests - everything is in order. There shouldn't seem to be any problems. There shouldn't seem to be any problems. But wait. How long have we been studying their interaction?

1:19
It is very difficult. In fact, no one usually does this. We are completely dependent on the so-called "post-marketing surveillance" after the drug has entered the market. But how can we find out that these two drugs are interacting somehow wrong? These three? Five? Seven? Ask loved ones with multiple diagnoses how many drugs they take.

1:43
Why am I concerned about this issue? She worries me a lot. I am a computer scientist and data analyst, and in my opinion, the only hope to understand these interactions is to raise great amount sources of information to find out which drugs are safe to use together and which are not.

2:03
I want to tell a story about data analysis. It all started with my student Nick. Let's call him "Nick" because that's his name. (Laugh)

2:11
He was a young student. I said, "You know, Nick, we need to understand how drugs work, how they work together, how they work separately, because we understand this very poorly." And now the FDA has posted a great database. This is a database of cases related to side effects. They just put it on the Internet - it's available, you can download it right now: hundreds of thousands of reports and complaints from patients, doctors, companies, pharmacists. And they are all quite simple: they say what diseases the patient has, what drugs he took, and that’s it. Negative consequences that they have experienced. This is not all that is happening in America today, but it is hundreds and hundreds of thousands of drugs.

2:53
So I said to Nick, “Take glucose, for example. Glucose is very important, we know it is linked to diabetes. Let's see if we understand the backlash to it." I sent Nick to work. Nick came back.

3:07
"Russ," he tells me, "I created a classifier to look at the side effects of a drug, according to this database, and it shows whether the drug will change glucose levels or not."

3:20
He did it. It was pretty easy. He took all the drugs that change glucose levels and a group of drugs that don't, and he thought, "What's the difference between their side effects? Changes in fatigue? Appetite? Urination?" All this together allowed him to create a very good determinant. He said, "Russ, I can tell with 93% accuracy whether a drug will change glucose."

3:42
I say, "Nick, that's great." He is a young student, you need to give him self-confidence. “Nick, but there is a problem. All the doctors in the world are aware of drugs that change glucose levels, because this is the basis of their practice. So it's great, great work, but not so interesting, you definitely won't publish it. (Laugh)

4:01
He said, "I know, Russ, I knew you'd say that." Nick is a smart guy. “I knew you would say that, but I did another experiment. I looked in the database for people taking two drugs, and began to look for signs of a change in glucose in such patients. At the same time, drugs alone did not change glucose, but together - I saw that they do it.

4:25
I said, "Well done great idea. Show me the list." And there are a lot of drugs, not very interesting. But what caught my attention was that there were two drugs on the list: paroxetine, or Paxil, an antidepressant, and pravastatin, or Pravacol, a cholesterol drug.

4:42
And I said, “Hmm. Millions of Americans take them at the same time." We later found that 15 million Americans take paroxetine, 15 million take pravastatin, and we estimate that a million take both at the same time. And that's a million people who could have a glucose problem if this shamanism that he did with the FDA database really works. But I said, “It’s still not published, I really like what you did with machine learning there, but this is not proof that we are right. So we need something else. Let's look at Stanford's electronic medical records, we have a copy for research - we've removed all identification information. And I said, "Let's see if there are problems with glucose levels in people taking both drugs."

5:30
So thousands and thousands of people are taking paroxetine and pravastatin, according to the Stanford records. But we needed special patients: those who took one of the drugs and measured the glucose level, and then started taking the second and measured the glucose level again, and in a certain period of time - say, two months. And when we did, we found 10 patients. However, eight of those 10 had a spike in glucose when they got the second "P" - we call them "P" - when they got the second "P". Any of them can be the first, they start taking the second - glucose rises by 20 mg per deciliter. I would like to remind you that in ordinary life if you are not diabetic, your glucose level is around 90. And if it rises to 120-125, your doctor may start to think that you have diabetes. So a jump of 20 points is quite significant.

6:21
I said, “Nick, this is really cool. But, unfortunately, it is still not worth publishing, because 10 patients, whatever one may say, this is not enough.”

6:30
And we thought - what can we do? Let's call our friends from Harvard and Vanderbilt who have - Harvard in Boston, Vanderbilt in Nashville - who have the same medical history as we do. Let's see if they can find similar patients with one "P", with another "P", glucose measurements in the range we need.

6:47
Thank God, 40 such patients were found at Vanderbilt University in a week, the dynamics are the same. 100 patients were found at Harvard, the dynamics are the same. As a result, we had 150 patients from three different medical centers, and everything indicated that those taking these two drugs had significant spikes in glucose levels.

7:09
Interestingly, we didn't include diabetics because diabetics don't have good glucose levels anyway. When we looked at those rates in diabetics, they went up by as much as 60 mg per deciliter, not 20. It was serious, and we thought we should publish it. We've submitted a job. Everything was based only on data - data from the FDA, data from Stanford, Vanderbilt, Harvard. We haven't done any experiment.

7:35
And we were nervous. And Nick, when our work was being considered, went to the laboratory. We found someone who does this. I don't know anything about it. I deal with patients, not pipettes. They taught us how to give drugs to mice. We took mice and gave them one "P" - paroxetine. And others were given pravastatin. And the third group was given both. And who would have thought, glucose rose from 20 to 60 mg per deciliter in these mice.

8:06
The data-only work was accepted, but at the end we added, “By the way, if you give drugs to mice, glucose rises.”

8:14
It was great, and everything could end there. But I still have six and a half minutes. (Laugh)

8:21
And we sat and thought about all this, and I don't remember who came up with it, but someone said, “I wonder if any of these patients who were taking both drugs noticed the side effect of hyperglycemia? They could and they should have. How could we determine this?

9:21
Well, yes. Super. I felt like... (Laughter)

9:26
It was like I was talking to Nick again. He works for one of the biggest companies in the world, and I'm already trying to cheer him up. But then he says, “No, Russ, you probably don't get it. We not only have Bing queries, but Google, Yahoo, Bing queries - from everywhere - made from Internet Explorer. We then store this information for 18 months for research purposes only.” And I say: “This is another matter!” It was Eric Horwitz, my friend at Microsoft.

9:51
And we did a study where we identified 50 words that a common person may gain if he has hyperglycemia, such as "fatigue", "loss of appetite", "going to the toilet a lot", "pissing a lot" - I'm sorry, this is one of the things they can look for. And we had 50 phrases, we called them "diabetic words." And first we looked at the source data. It turned out that from 0.5 to 1% of all searches on the Internet include one of these words. This is our starting point. If people search for "paroxetine" or "Paxil" - these are synonyms - and one of those words, the frequency rises to about 2% of "diabetic words" if you already know that the word "paroxetine" is there. If it's pravastatin, the frequency rises to about 3% of baseline. If both "paroxetine" and "pravastatin" are in a query, it rises to 10% - a huge three to four times increase in queries with drugs of interest and "diabetic" or "hyperglycemic" words.

10:55
We published it and it got attention. The reason this deserves attention is because patients don't tell us about their side effects directly, but through search. We sent it to the FDA. They got interested. They set up surveillance social networks to cooperate with Microsoft, with their technical capabilities for this task, as well as for the feed on Twitter, on Facebook, for statistics search queries to reveal early signs that drugs, alone or together, are causing problems.

11:27
And what did I understand? Why am I telling this? First of all, data can now help provide a better understanding of the interaction. medicines and, in fact, their action. How they work This will create, and has already created, a new ecosystem for understanding how drugs work and how to optimize their use. Nick did not stop - he is now a professor at Columbia University. He did the same in his doctoral thesis with hundreds of pairs of drugs. He found some very important interactions, so we replicated the experiment and showed that it actually works for finding drug interactions.

12:05
However, there are a couple more points. We don't just use drugs in pairs. As I said, there are patients taking three, five, seven, nine. Has anyone studied their nine-way interaction? Yes, we can consider them in pairs - A and B, A and C, A and D, but what about A, B, C, D, E, F, G - all together, taken by one patient? Perhaps they interact, decreasing or increasing each other's effectiveness, or causing unexpected side effects? We really don't know. This is an open field where we can use our data to try to understand these interactions.

12:45
Two more lessons. I want you to think about the opportunities we have, thanks to those people who have told their complaints to pharmacists, to themselves, to their doctors, thanks to those who allow the use of databases at Stanford, Harvard, Vanderbilt for research. People are worried about their data. They're worried about privacy, security - that's the way it should be. We need secure systems. But we don't need systems that restrict access to data, which is a rich source of inspiration, innovation, and new discoveries to create new things in medicine.

13:23
And in conclusion, I want to say that in our case we found a couple of drugs, and the results were sad. They really caused problems. They raised the glucose. They could give someone diabetes that wouldn't have happened without them, so you have to use the two drugs very carefully, maybe not together, maybe prescribe something else to the patient. But there was another possibility. We could find a couple or three drugs that would provide a beneficial interaction. We could discover new effects of drugs that they would not have on their own, but together, instead of side effect they could offer perfectly new therapy diseases that were considered incurable or difficult to treat. In modern drug therapy all significant discoveries - for HIV, tuberculosis, depression, diabetes - are always a cocktail of different drugs.

14:15
The good news is, and this is a topic for another TED talk, that we can use the same data sources to find positive effects combinations of drugs that will provide us with new treatment regimens, A New Look on how the drugs work, and the ability to care for patients even better.

14:34
Thanks a lot.

Read on Zozhnik: