Cancer remains one of the leading causes of human death, and early detection is the key to reducing mortality. To detect cancer in the early stages, two-stage screening programs are widely adopted in practice. Individuals receiving positive outcomes in the first-stage (initial) test are recommended to undergo a second-stage test for further diagnosis. The initial test design—i.e., selecting cutoffs to report test outcomes—is crucial for screening effectiveness (i.e., cancer detection) and efficiency (i.e., second-stage capacity costs). However, not all individuals who receive positive outcomes follow up with the second-stage test; evidence shows that adherence behavior is closely associated with the cutoff used in the initial test. This paper studies the initial test design in the context of colorectal cancer (CRC) screening to balance the trade-off between screening effectiveness and efficiency and takes into account individuals’ guideline adherence behavior.

We adopt a Bayesian persuasion framework with information avoidance to model the initial test design and individuals’ response to screening guidelines. We analytically prove that under certain conditions, an initial test using a single cutoff (i.e., a dichotomous test) is optimal for screening follow-up maximization, and a continuous test (i.e., showing exact readings of the biomarker) is optimal for screening effectiveness maximization. We apply the framework to Singapore’s CRC screening guideline design and calibrate the model using various sources of data, including a nationwide survey in Singapore. Our results suggest that compared with the current practice, increasing the cutoff to the level that maximizes expected follow-ups by cancer patients can detect 969 more CRC incidences and prevent 37,820 colonoscopies, which are the second-stage test for CRC screening. Aiming only for high-sensitivity initial tests using lower cutoffs (as in the current practice) can backfire and lead to large numbers of unnecessary colonoscopies and low follow-up rates from cancer patients. We further explore the benefits of using different cutoffs for different subpopulations and use an interpretable clustering technique to construct implementable rules for partitioning the population. We demonstrate that using a lower cutoff for males older than 60 and females older than 70 (high-risk and high-adherence groups) and a higher cutoff for the rest of the screening population (low-risk and low-adherence groups) can further improve screening effectiveness and efficiency.

In many countries, the COVID-19 vaccination program faces great challenges, from the management of limited and irregular supply of vaccines, uncertain vaccine take-up rate from the population, and appropriate appointment booking management to reduce congestion, etc. In addition, the feature of a two-dose regimen of most COVID-19 vaccines poses a unique operational challenge, the “blocking phenomenon,” where the need to reserve vaccines for the second-dose appointment may “block” the take-up rate for the first-dose appointment. Determining the appropriate volume of vaccines to be kept in reserve in case of disruption to the supply schedule is an important operational problem for vaccine rollout.

In this paper, we use the concept of “booking curve” (from the revenue management literature) to develop a practical tool that jointly determines the vaccine appointment booking limits that control the administration of the first and second doses, followed by an invitation schedule that decision-makers can use to regulate the appointment bookings (demand). The optimization framework aims to design the vaccine rollout policy to maximize the vaccination rate, while ensuring that the aggregate appointment waiting time in the system remains minimal, with sufficient cushion to account for supply disruption and uncertain take-up of appointments. We show that the optimization problems can be efficiently solved by linear and conic programs. A vaccine rollout numerical study based on the Singapore vaccination program is presented to demonstrate the novelty and advantage of the optimization framework. The optimization framework has been developed into an open-source tool to assist the policymakers in designing an effective and adaptive COVID-19 vaccine rollout policy facing the evolving challenges in fighting against the pandemic.

Optimization problems posed in the context of queueing networks are notoriously difficult to model and solve. However, they crop up frequently in the service management context, such as healthcare operations management, platform operations, distributed systems, and also in areas such as manufacturing and inventory management. Existing techniques in queueing either do not cater to the finite time transient context that such optimization problems arise in, or rapidly run into the curse of dimensionality, that may require significant approximations or simplifying assumptions to surmount. In this work, we develop new primitives to modelling queues altogether in the discrete finite-time transient setting. Our framework can handle general service and arrival distributions, wide variety of network structures (including loops), and in particular, can be utilized to optimize for capacity, routing within the network, and scheduling, and is able to accept targets and objectives such as queue metrics of queue length and waiting times. Moreover, we apply robust optimization techniques to recover a formulation that is quasi-convex, and in many examples, reduces to solving a sequence of linear programs. In numerical tests, we have found them to be superior to fluid approximations and approximate dynamic programming, two of the most common and promising approaches seen in the literature.

Gar Goei Loke is currently an Assistant Professor with the Department of Technology and Operations Management at Rotterdam School of Management, Erasmus University. Before, he is a Visiting Assistant Professor with the Department of Analytics & Operations at the National University of Singapore (NUS) Business School. His research interest is in the application of robustness optimization, or satisficing, to the solution of flow control problems in network-based settings. His research falls under the application domains of healthcare, manpower planning, public policy and resource allocation. Recently, he has started on a new stream of research that looks at cross-pollinating ideas between Machine Learning and Operations Research, such as contextual stochastic optimization. Gar Goei had spent more than 5 years in the Singapore Government, leading teams that work on both analytics and operation research projects.