Vaccine Supply Chains
In addition to computational models of disease transmission dynamics, the VMI is developing models that represent vaccine supply chains in countries including cold chains. These models help public health authorities in decision making on how to optimally use existing resources for the delivery of effective vaccine to the clinics where vaccination occurs. In addition, these models can provide estimates on resources that would be required to achieve certain end goals, such as a certain vaccination coverage rate. The VMI is currently working on models for Niger and Thailand that include all vaccines in the Expanded Program of Immunization (EPI) with particular interest in the consequences for supply chain management of potential new vaccine introductions, such as rotavirus.
In collaboration with the World Health Organization and the Niger Ministry of Health, we have developed a representation of the Niger supply chain. VMI investigators have visited Niger twice in the last year and work on incorporating model results in policy making is ongoing. In Thailand, the VMI is working with the Southern Vaccine Research Team at Prince of Songkla University (PSU) to model the vaccine supply chain for one province (Trang) in Thailand which supplies ten district and eighty-seven sub-district stores.
Three generic types of extensible supply chain models have been developed: (1) an Equation-Based Model (EBM), (2) a discrete event simulation model based on the commercially available ARENA software, and (3) our own custom-designed discrete event simulator software package we call HERMES. These three simulators are being tested using supply chain data from Niger and Thailand. Each model can rapidly create a model of any vaccine supply chain containing any number of different vaccines, answer a different set of overlapping questions, and cross-validate the other models.
The Equation-Based Model (EBM) is a deterministic, mathematical EBM that represents movement of vaccines throughout the supply chain with a set of mathematical equations. The EBM can help to answer broad policy questions but does not capture many stochastic elements and deviations from operational policies. Therefore, in tandem, the VMI developed two discrete event simulation models that can incorporate greater detail and random events. These simulation models can serve as virtual laboratories to test various changes in vaccine and supply chain characteristics and circumstances (e.g., environmental conditions, personnel behavior).
One of the largest challenges for vaccine supply chains is vaccine wastage that occurs during shipments due to heat or cold exposure or that occurs at the health clinic if not all doses in a vial are used in one day and vaccine exceeds the expiration date. We have explored the effects of changing the number of doses per vial for all heat-sensitive EPI vaccines on vaccine wastage at the periphery of the vaccine supply chain. Runs of experiments have been completed for introduction of rotavirus vaccine and various thermostable vaccines to Niger.
Eventually, vaccine supply chain models will be linked to disease dynamic models that will allow experiments simulating the effects of candidate vaccine technologies on epidemic control.