Stochastic integer programming based algorithms for adaptable open block surgery scheduling

We develop algorithms for adaptable schedule problems with patient waiting time, surgeon waiting time, OR idle time and overtime costs. Open block surgery scheduling of multiple surgeons operating in multiple operating rooms (ORs) motivates the work. We investigate creating an “adaptable” schedule of surgeries under knowledge that this schedule will change (be rescheduled) during execution due to uncertain surgery times and other disruptions. The problem we address consists of finding an initial “day before” OR schedule (surgery starting times, surgery sequence in each OR and surgery sequence for each surgeon), and is formulated as a multi-stage stochastic integer program (with the initial schedule determined in phase 1) using sample average approximation. A small (due to complexity) computational study is conducted. We find optimal initial schedules under simple right shift rescheduling policies, and further demonstrate that more complex rescheduling policies (i.e. recourse) can be effective in improving performance given a variety of initial schedules.

Article

Download

View Stochastic integer programming based algorithms for adaptable open block surgery scheduling