Improving the design and operation of WHO-EPI vaccine distribution networks

Vaccines have contributed significantly to the prevention of diseases. Yet millions of children, especially in low and middle income countries, remain unvaccinated and are exposed to preventable diseases such as typhoid, measles and tuberculosis. There are many reasons for this including personal belief systems, vaccine safety concerns, problems with vaccine availability, failures in the healthcare system, social barriers and economic constraints. International organizations are making continual efforts to increase vaccine coverage in these countries using various strategies. In this research we focus on the problems associated with poor design and operation of vaccine delivery systems and address these issues via four broad contributions. First, we present four quantitative models that can be used to optimize the selection of locations for vaccine outreach (where teams from clinics go to relatively remote places to administer vaccines), in order to maximize the number of residents that can be reached; each model addresses a different type of coverage possibility. The models are analyzed and contrasted using an example and adapted to address the situation when the coverage assumptions and demands are uncertain. Second, we propose modular vaccine packaging as an alternative to current packaging, which is not standardized and leads to inefficiencies when packing vaccines into a storage device; this in turn can result in vaccine shortages. We illustrate the benefits of modular packaging over current packaging schemes and storage devices that are commonly used in the field. Third, we suggest alternative ordering policies at the clinic level that are based on secondary vaccine packaging. The policies draw upon lean concepts that have been used in the manufacturing sector to simplify and improve inventory management. Since the ordering units are larger, storage space issues may occur at clinics or during vaccine transportation and the new ordering polices are analyzed in terms of their effect on storage. Lastly, we propose a mathematical model to redesign the vaccine distribution network from a central warehouse to individual health clinics and study algorithms to solve this difficult problem. We propose a hybrid algorithm based on mixed integer programming and an evolutionary strategy. We also describe how to improve the performance of the evolutionary strategy and how to use the results of the evolutionary strategy to reduce the calculation time of the integer programming model