A distributionally robust optimization approach for coordinating clinical and surgical appointments
针对专科手术中临床与手术预约的协调问题，提出分布鲁棒优化方法，应对患者是否需手术及手术时长的不确定性，案例表明可降低医生空闲和加班时间。
In this article, we address a two-stage scheduling problem that requires coordination between clinical and surgical appointments for specialized surgeries. First, patients have a clinical appointment with a surgeon to determine whether they are an appropriate candidate for the surgical procedure. Subsequently, if the decision to pursue the surgery is made the patient undergoes the procedure on a later date. However, the scheduling process aims to book both the clinical and surgical appointments for a patient at the time of the initial appointment request. Two sources of uncertainty make this scheduling process challenging: (i) the patient may or may not need surgery after the clinical appointment and (ii) the surgery duration for each patient and procedure is unknown. We present a Distributionally Robust Optimization (DRO) approach for coordinating clinical and surgical appointments under these uncertainties. A case study of the Transcatheter Aortic Valve Replacement procedure at Mayo Clinic, Rochester, MN is presented. Numerical results include comparisons with the current practice and four heuristic scheduling policies from the literature. Results show that the DRO-based scheduling policies lead to lower total surgeon idle-time and overtime per day. The proposed policies also restrict the under and over utilization of clinical capacity.