The problem of scheduling pre-operative assessment clinic (PAC) consists of assigning patients to a day for the exams needed before a surgical procedure, taking into account patients with different priority levels, due dates and operators availability. Realizing a satisfying schedule is of upmost importance for a hospital, since delay in PAC can cause delay in the subsequent phases, thus lowering patients’ satisfaction. In this paper, we propose a two-phase solution to the PAC problem: in the first phase, patients are assigned to a day taking into account a default list of exams; then, in the second phase, having the actual list of exams needed by each patient, we use the results of the first phase to assign a starting time to each exam. We first present a mathematical formulation for both problems. Further, we present a solution where modeling and solving are done via answer set programming. We then introduce a rescheduling solution that may come into play when the scheduling solution cannot be applied fully. Experiments employing synthetic benchmarks on both scheduling and rescheduling show that both solutions provide satisfying results in short time. We finally show the implementation and usage of a web application that allows to run our scheduling solution and analyze the results graphically in a transparent way.

The problem of scheduling Pre-Operative Assessment Clinic (PAC) consists of assigning patients to a day for the exams needed before a surgical procedure, taking into account patients with different priority levels, due dates, and operators availability. Realizing a satisfying schedule is of upmost importance for a clinic, since delay in PAC can cause delay in the subsequent phases, causing a decrease in patients' satisfaction. In this paper, we divide the problem in two sub-problems: In the first sub-problem patients are assigned to a day taking into account a default list of exams; then, in the second sub-problem, having the actual list of exams needed by each patient, we use the results of the first sub-problem to assign a starting time to each exam. We first present a mathematical formulation for both problems. Then, we present solutions based on Answer Set Programming (ASP): The first solution is a genuine ASP encoding of the sub-problems, while the second introduces domain-specific optimizations. Experiments show that both solutions provide satisfying results in short time, while the second is able to prove optimality faster.