Paving the way for more accessible cancer care in low-income countries with optimization.

Cancers are a growing cause of morbidity and mortality in low-income countries. Geographic access plays a key role in both timely diagnosis and successful treatment. In areas lacking well-developed road networks, seasonal weather events can lengthen already long travel times to access care. Expanding facilities to offer cancer care is expensive and requires staffing by skilled medical professionals, which are often in short supply. In this article, we propose a mathematical model to improve geographic access to cancer care by jointly considering expansions to care facilities and improvements to the road network. We model this as a multi-period stochastic facility location network design problem. For each period, a decision maker must simultaneously choose a set of facilities at which to add tertiary cancer services and a set of roads to improve while facing demand and travel time uncertainty. Once demand for cancer care and weather events are realized, patients observe road conditions and use the transportation network to travel towards the closest facility with available cancer services. We create a new path-based formulation of this problem and develop a new branch-price-and-cut algorithm with acceleration techniques that take advantage of this formulation's structure. We demonstrate our approach using Rwanda as a case study and show that the reductions in travel time to cancer care directly attributable to the road network improvements can be as high as one hour.