You are having 100 guests for an outdoor picnic and need to make the decision: do you buy big bottles of soda (potentially messier but less expensive), smaller bottles (potentially more convenient but also more expensive), individual cans, or a compromise between the two? Your decision will depend on several factors including available refrigerator space, price, convenience, expected number of guests, and probability of leftover soda. You may also consider the servers (adult or child), the risks (e.g. mess, amount consumed) availability (will there be enough, is there a need to ration?), and waste produced (e.g., aluminum cans versus plastic bottles).
Sound familiar? It’s a very similar decision process for vaccines – which must consider the price to the country, if there is enough storage space, how difficult is it to transport all of the required product to where it needs to be, how much wastage will there be, if is there enough for everyone, and what is the risk that some containers will remain unopened?*
Looking at shelves in a grocery store, it is clear marketers understand the dynamics of these everyday decisions. Companies like Coca-Cola fully research how consumers make decisions and weigh trade-offs of price, size, and convenience. They vary offerings by region and by store, limiting choice in some places and providing every option under the sun in others.
Interestingly enough, where the stakes are very high – in vaccines – there doesn’t seem to be the same level of consideration. No one gets sick if a bottle of Coke is not available. We wish the same were true with a vaccine. Yet countries, donors, and manufacturers often do not fully weigh the consequences of container choice on cost, availability, likelihood of usage, and safety.
IVAC’s new report, Coverage, Cost and Safety Impacts of Primary Container Choice, details these tradeoffs and calls on the public health community to take a more data-oriented approach to supporting decisions for each individual vaccine. The report uses various scenarios run on the Highly Extensible Resource for Modeling Supply Chains (HERMES) model of the Benin vaccine supply chain built by the HERMES Logistics Modeling Team, working with colleagues from the Logivac Project and Benin. The HERMES Benin model allowed us to change the size of a vaccine container and simulate the resulting effects (e.g., cost per dose delivered and vaccine availability). The effects depend on a number of factors, including other vaccines in the supply chain, current constraints, and wastage policies (e.g., is a vaccine vial opened or not depending on number of children in a session).
We also considered what could happen at particular locations and clinics...not just at the country or global level. This is important since country-wide measures of routine vaccine coverage may be misleading. Often, country-wide routine vaccine coverage is high (>90%) even when coverage in remote areas (where vaccines are needed most urgently) remains low. Remote areas also may be very difficult to reach, have limited refrigerator space to store vaccines, and offer poor access to health care, all resulting in limited vaccine availability and high disease risk. At the same time, fewer children may show up to be vaccinated on a particular day, making healthcare workers loath to open a vial if they want to avoid wasting unused doses and country policies may compel them to do so. There is also an additional scenario to be considered for countries or even communities that are risk averse. Some product presentations inherently carry with them an increased risk for contamination (particularly multi-dose vials if not handled appropriately) or error (e.g, when a lyophilized vaccine is reconstituted with product other than diluent), or even contain a preservative that may be perceived to have safety issues (as is the case with thiomersal in some high- and middle-income countries). In this scenario, containers not requiring preservative or single-use presentations may be preferred, even when space is limited and/or cost may be higher.
At a global level we must start looking at the unique characteristics of each new antigen, providing guidance well in advance of product introductions, and looking across a range of scenarios. Although it is tempting for the manufacturer to limit the number of product presentations, building a case based on simulation experiments from a variety of countries, considering implications on coverage, cost, safety, and feasibility could help manufacturers better understand the investments needed and the “cost” of getting it wrong. Furthermore, the global community can better appreciate why certain costs may be higher for some populations and evaluate the need for more specific guidance on individual vaccines across a variety of scenarios. Container choice is not as straight forward as looking at price, wastage, and space. When humans are involved and there are imperfect choices, we should be approaching the decision with a framework that considers global, local, and manufacturing implications to make better-informed choices for every product introduced.
A number of efforts are underway that have improved forecasting and enabled a more systematic view of cold chain space. Now is the time to invest in ensuring we understand the full picture of cost per dose delivered, availability, and safety for each individual product and scenario to make better informed decisions.
*When there is a mismatch between number of doses in a vial or container and expected session size (e.g. the health worker sees two children per day and has a 10-dose vial of vaccine) and there is concern about wastage or availability of vaccine for future sessions, some health workers may not open the vial and vaccinate those children that particular day. The multi-dose vial policy, which enables the vaccine to remain opened for a designated period of time, can help mitigate wastage concerns, but it isn’t always followed, and not all vaccines are eligible.
Lois Privor-Dumm, MIBS, is Director of Policy, Advocacy & Communications and Bruce Lee, MD, MBA, is Director of Operations Research at IVAC.