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

4.4.5Reward

The reward in the hypertension SDDP model is the total net benefit, which is the health benefit expressed in the monetary units, minus the total costs associated with sequential treatment. Each drug selected for s at time t leads to an expected cost or treatment benefit, which is estimated in terms of QALYs. The costs and QALYs assigned to each transition, r(s_t+1|s_t,a_t,) can be interpreted as the one-step (or immediate) reward. The global reward for the given policy, π=(d₁,d₂,d₃,d₄), is the weighted sum of the one-step rewards with the probabilistic structure of the underlying evaluation model; this includes the costs and QALYs, calculated both in the short-term drug switching model and the long-term CVD model.