You are on the right track I think, the tolerability of the drug is much more about what side-effects will stop the patient taking it and much less about toxicity addressed by the LD50. So classifying into symptomatic groups is a good start and comparing frequency of each class across drugs will give you comparison you need.

Dear Mostafa. the classification in A, B, C, D, E and F wont give you the safety of the drug because this depends on the incidence of each of this adverse drug reactions. example, a type A can be hypoglicemia induced by insulin. but the important is if the type of insulin has an incidence of 1%, 10% or 50% of hypoglicemia and the serverity of it. this classification is about the phisiopathologic mechanism of the adverse effect (not exactly at all). in the example that you mention the hemolytic anemia, the relevant is how often ibuprofen causes this adverse effect. this is really a very infrequent adverse event. in the other hand methyl dopa induce in near 10% this adverse effect.

Isn't the type indicative of the frequency? For example, if constipation is seen in 15% of the population taking drug X, wouldn't we classify this as type A adverse reactions because drug X is thought to have another action that causes constipation. But, on the other hand, if skin rash is seen just in 0.1% of the population, we would classify this as type B adverse reaction because we will think it is patient dependent maybe this group of patients lack a specific enzyme, etc.. So just the frequency of the reaction is indicative of which type this reaction would fall under?

You already have classification of adverse events by their severity regardless of their nature. You can look at any oncology clinical trial and you will find it.

Yes, adverse events (AEs) were classified to primarily help us understand and segregate the profile of AEs. As we all know, the first classification of adverse events was proposed by Rawlins and Thompson in 1977: Type A related to pharmacological mechanism and Type B, specific characteristic reaction. Subsequently, Uppsala Monitoring Centre made a new classification which retained Type A as proposed by Rawlins and Thompson, but included the following additional types: B (Bug reaction), C (Chemical reaction), D (Delivery reaction), E (Exit reaction), F (Familial reaction), G (Genotoxicity reaction), H (Hypersensitivity reaction), and U (Unclassified reaction).

In general, type A reactions are related to drug action and the others related to patient reactions due to immunogenicity or allergy, idiosyncrasy etc.

If the objective of the exercise is to undertand the nature/mechanism of AEs, classification may help. However, as discussed in the earlier comments, it will not give any information regarding incidence of AEs.

Very nice topic to discuss and participate

1. Classification using A, B, C, D for other issue is not suit able as these are already adopted by FDA safety categories to use drugs in pregnancy.

2. Even if one uses such classification that will be much more complex because a single drug has multi system side effects .

3. If you consider nausea as class A, Diarrhoea B, Constipation C, Vomiting D, Gastric Ulcer E, Stomatitis F, Metalic test G, Anorexia H, Insomnia I, Dizziness J, Fits K, Vertigo L, Headache M, Treamors N,

3.I mean to say long list of side effects they not be classified as you suggest even it will further complicate pharmacology of drugs to learn and remember

So, How can I view the side effects of any drug and make sense of it, and gather some insights? Will I have to know the adverse drug reaction type? Or the frequency of the reaction?