Operating a biomedical samples laboratories network under stochastic demand

Abstract

Management of biomedical samples plays a central role in an efficient healthcare system and requires important resources. This paper addresses an assignment problem where several types of samples are collected at clinics and other collection centers, and then transported to medical laboratories to be analysed. For practical reasons, samples of the same type collected at the same collection center are consolidated and sent to the same laboratory. However, a collection center may send different types of samples to different laboratories. Moreover, demand for each type of sample at each collection center is uncertain but modelled by a known probability distribution. In this context, collection centers need to be allocated to laboratories in order to balance the workload between them while minimizing the total collecting distance. To tackle this problem, we first formulate it as linear integer model assuming deterministic demand. Then, a stochastic counterpart is presented, along with a solving method based on the sample average approximation technique (SAA). Numerical experiments inspired by a real-life case are conducted to illustrate how the proposed approach may be contribute to help decision makers to better manage the laboratories' capacity, increasing the system efficiency.