How do scientists estimate drug dosage based on studies?

[physio]

Scientists usually conduct a variety of in vitro and in vivo studies on rodent and their cell cultures and they show the concentration of a particular drug is effective against the disorder for both in vitro and in vivo for the rodents. My question is how do they translate that to a dose for the same disorder in humans. For instance I was reading an article that said X uM drug concentration was effective in bringing the desired effect in vitro and Y uM was effective in bringing about the same effect that is ameliorating the said condition. What will be the dosage that will be effective to bring about the same change based on these studies? In short, how do scientists proceed from these studies after the optimal dose has been established in rodents? Thank you in advance for your replies.

 

[jim mcnamara]

You missed a step. Drug dosages are tested on humans, too. Any drug has to have a validated LD50 value - the dosage (usually in mg/Kg of body weight) that kills 50% of the test subjects. NEVER on people. Some drugs behave very differently from species to species. There are established protocols for developing dosage values before human testing is even considered.

Here is an overview of the FDA's drug approval process - dosage and evaluation criteria are part of it.
http://www.fda.gov/ForPatients/Approvals/default.htm

Before a new drug is offered on the market, information of drug overdosage, patient care, etc, is disseminated to Poison Control Centers, for example. Further clinical experience with patients who abuse the drug, accidentally or otherwise, expand physicians knowledge base on the drug. You cannot ethically or reasonably give a deliberate large overdose of a drug to humans. You have to wait for patients and their families to do that for you. This sounds cynical but is pretty much how it happens.

Here is epdemiological data on prescription drug overdose deaths/mortality in New Mexico (USA):
http://nmhealth.org/about/erd/ibeb/sap/

 

[Ygggdrasil]

To add to jim's comment, Phase I clinical trials are done to determine the safe dosages of new medicine. Typically, researchers will give healthy volunteers very small dosages (a fraction of the corresponding dose found to be safe in animal studies) then gradually increase the dosages while monitoring for side effects.

In other words, safe dosages are determined empirically. The procedure for determining the safe dosage is guided by animal studies, but ultimately, there is no general formula for converting the safe or effective dose in animals to humans.

 

[Pythagorean]

To add more, drugs even affect different people differently, particularly people with atypical brains, who often need those drugs.

Even drugs that pass trials still need to be prescribed empirically to psychiatric patients. The psychiatrist will sometimes try several different drug cocktails on a single patient until one works to reduce bad symptoms while minimizing side effects.

The problem here though, is that we don't really understand the mechanisms of mental disorders. Their nosology is all symptom based. This means that it's possible to have two or three different mechanisms underlying a symptom in three different patients, yet we group them all together as a single mental disease and throw the same drugs at them until one of the drugs works.