Vector control, particularly distribution of insecticide-treated bed nets (ITNs), constitutes one of the major pathways to prevent and reduce malaria transmission. ITN distribution campaigns face several challenges, such as inadequate funding, budgetary constraints, hard-to-reach areas, limited transportation, and market and price volatility. While long-term agreements and proper planning can effectively overcome such challenges, those options may not be available for all humanitarian organizations and governments. To gain a better understanding of such tradeoffs we develop a robust optimization model that minimize ITN distribution costs while taking into consideration protection against financial, market and logistical uncertainties. The simultaneous account of such uncertainties is rarely seen on the humanitarian supply chain design literature. The proposed robust model explores data-driven adaptive uncertainty sets that capture the dependence structure among procurement and distribution costs, leading to plausible uncertain scenarios. In addition, we develop a hierarchical optimization approach to ease the burden of setting a specific robustness level for each constraint, when uncertainties are related to the independent terms. We study a United Nations Children's Fund (UNICEF) ITN distribution campaign in Ivory Coast, observing that (1) total costs increase with campaign robustness, as expected, and (2) campaign robustness comprises of improved supply chain flexibility, which might minimize efforts if it becomes necessary to adjust procurement and transportation plans when uncertainty arises. In addition, assessing robust solutions through Monte Carlo simulations against several realizations of uncertain parameter values indicates that, as desired, robust plan feasibility increases with the specified level of conservatism.

No abstract