Sequential drug decision problems in long-term medical conditions: a case Study of Primary Hypertension Eunju Kim ba, ma, msc

3.6.2Enumeration

The algorithm was repeated to examine the six policies included in the search space SS (see Figure ‎3.). The total net benefit TotalReward was calculated through the successive decision-tree model function_SemiMarkov. The variations of each sequential treatment policy πx were applied depending on the subsequent health states. Taking an example of π₁=(drug₁,drug₂,drug₃), the variations are:

πx={ (drug₁,drug₁,drug₁),(drug₁,drug₁,drug₂),(drug₁,drug₂,drug₂),(drug₁,drug₂,drug₃) }

The optimal solution OptSol was determined by directly comparing the total net benefits of six sequential treatment policies.

T = 3; % The number of time periods. A = 6; % The number of possible treatment sequences. S = 27; % The number of possible disease pathways. % Possible treatment pathways where 1=drug1, 2=drug2 and 3=drug3. SS = [1,2,3;1,3,2;2,1,3;2,3,1;3,1,2;3,2,1]; % Estimate the total net benefit of each sequential treatment policy using ‘function_SemiMarkov’.