Optimisation of medical equipment replacement using stochastic dynamic programming

In this paper, the medical equipment replacement strategy is optimised using a multistage stochastic dynamic programming (SDP) approach. The outcome is an optimal path which shows whether to keep an existing piece of medical equipment (defender) or replace it with a more economical alternative (challenger). We assume that each decision can result in a number of different possible outcomes, each with a known probability. Contrary to deterministic dynamic programming, the state at the next stage is not completely determined by the state and policy decision at the current stage. Instead, the next stage depends on the operation and maintenance cost which is modelled as a stochastic variable. A Keep-Replace sequence of the highest returns (lowest costs) is the result of solving the problem using forward decision making. The benefit of the SDP solution versus that of keeping medical equipment until the end of its expected life is investigated for three scenarios: (1) no revenue for the defender and the challenger, (2) equal revenues for both, and (3) higher revenue for the challenger. The percentage of benefits relative to the current acquisition cost for the three scenarios are 616.9%, 728.2%, and 789.29%, respectively. Each percentage represents the relative difference between the equipment life cycle cost of the optimal sequence and that of the conventional sequence.