Table 2.

Steps involved in methodological approach

Steps	Approaches	Methods
Step 1	Use available knowledge	Bibliography
		(i) Pathophysiology, diagnosis, therapeutics, pharmacology
		(ii)Discursive and mathematical models for the disease
		(iii)Discursive and mathematical models for the drug effect(s)
		Individual epidemiological and RCT databases
		(i) Statistical approaches for analysis
		(ii) Identify prognostic biomarkers
		(iii) Identify N potential drugs (or therapeutic strategies) for evaluation in phase III RCTs
		(iv) Identify predictive biomarkers for these N drugs (e.g. interactions between patient characteristics and drug efficacy)
		(v) Validate drug-disease models
Step 2	Drug-disease modeling for the N treatments identified above	Treatment 1: (Disease model + Drug effect model 1)
		Treatment i: (Disease model + Drug effect model i)
		Treatment N: (Disease model + Drug effect model N)
Step 3	Drug-disease modeling for the N treatments above in patients whose characteristics may interact with drug efficacy	Treatment 1: (Disease model + Drug effect model 1) in patients whose specific characteristics interact with treatment effect 1
		Treatment i: (Disease model + Drug effect model i) in patients whose specific characteristics interact with treatment effect i
		Treatment N: (Disease model + Drug effect model N) in patients whose specific characteristics interact with treatment effect N
Step 4	Experimental RCT design modeling (including orthogonal approaches) for N conditions above	P experimental designs *N situations
Step 5	Simulate these N*P options	Results ordered in terms of potential efficacy, adverse events, number of needed patients, cost (including trial duration)
Step 6	Identify the most relevant drugs to be evaluated in phase III RCTs and the RCT design to be used for each of them	Multiple-criteria decision analysis approaches