We Should Track Global Vaccine Manufacturing Better - Here Is Why, and How

At the start of the COVID-19 pandemic in early 2020, the world didn’t know the global vaccine and biologics manufacturing capacity with much precision. Eighteen months later, there are still no accurate estimates of the currently installed capacity, nor do we know much about the future capacity to produce COVID-19 vaccines.

Multiple groups (Castillo et al, Wouters et al, AirFinity, Reed and Agarwal) have provided and utilized different capacity estimates—sometimes suggesting there is sufficient manufacturing capacity, sometimes alerting there is a wide shortfall. 

Estimates of vaccine production are not just inadequate on a global level, but nationally too, where governments didn’t have a good understanding of the capacity available within their jurisdiction. An example of this is India, where the government overestimated the number of vaccines that the country could produce (and production capacity estimates still remain unclear).

Manufacturing capacity increased greatly since the first COVID-19 vaccines were approved in December 2020, but our estimates of how much it increased by, or how much the capacity currently is, are based largely on collating public media announcements. With COVAX recently cutting its supply projections, it’s clear that there are very real problems with vaccine supply chains. To what extent are these caused by production problems, procurement issues, or export bans is not easy to discern.

Why is tracking global vaccine manufacturing capacity so crucial?

Estimating global manufacturing capacity provides several important benefits:

1. It supports adequate planning for procurement and vaccine delivery. International procurement organizations, such as COVAX, and low- and middle-income country governments are still not clear on when vaccines will be delivered to them. Planning for vaccine procurement and deployment would be easier with more reliable estimate of manufacturing capacity. Estimates of manufacturer’s order books combined with precise estimates of manufacturing capacity can help create more robust scenarios of delivery schedules.

2. It identifies problem input shortages ahead of time. Production of COVID-19 vaccines moved from one bottleneck to the next—from fears of glass vial shortages to shortages of single use bio-reactor bags. Measuring the global capacity to produce key materials for vaccine production would help industry and policy makers plan production and encourage scale up of future bottlenecks ahead of time. Imprecision in estimates has also led to an inefficient allocation of inputs. Some countries (and companies) have probably held onto more key ingredients than they need for fear of losing out, whilst in other parts of the world shortages of the same component causes binding contrictions. A better understanding of the market for components and production capacity for vaccines would allow them to be distributed more efficiently, which would build on the work already done by the COVAX manufacturing task force and others.

3. It identifies the right incentives and investments. Uncertainty causes great problems for producers and policy makers interested in increasing production. Without knowing current production capacity, nor the ability to adapt current manufacturing facilities to make vaccines, it is hard to know whether production should be expanded, and if so, where. The private sector is particularly weary of footing the bill for infrastructures that will not be used. Donors are also sceptical about repeating the “build and decay” stories seen in the past. Better estimates of finished product manufacturing capacity will improve certainty and will direct investments to the areas and geographies where there is greatest value.

4. It makes the polarizing debates on IP and IP waivers more objective and evidence based. In addition to current capacity, not knowing what potential capacity exists—which can be reprogrammed and repurposed to make COVID-19 vaccines in the short to medium term—further exacerbates and polarizes the debate on intellectual property, tech transfer and other scaleup approaches. Knowing whether there were factories able to manufacture these products sitting idle, and which bottlenecks were binding production, would make it easier to remedy the problem.

5. It supports better pandemic preparedness and response.  It is not possible to predict what kind of capacity or what platform will be needed until it is known what the pandemic organism is and, for novel organisms like COVI-19D was, which platforms work well against it. Identifying capacity in advance of a pandemic—particularly surge capacity—will be better for planning and quicker manufacturing in a future crisis.

What are the current efforts to track manufacturing capacity?

Since the start of last year, important work has been done to track COVID-19 vaccine manufacturing capacity. These efforts used one of two approaches.

The first approach is tracking announced capacity for different vaccines. This technique has been used by Airfinity, who have led much of the global conversation around vaccine supply, and by UNICEF, who have launched a dashboard that allows easy tracking of publicly announced capacity. Researchers at Duke University have also used a similar approach. Here, researchers track the announcements for vaccine production by different companies—as well as data for historical production—and estimate how quickly production plans can be scaled up. This data set is kept up to date and it capitalizes on easy-to-source information because it’s tracking public announcements, which companies are incentivized to make. This approach has led to useful estimates for the number of COVID-19 vaccines that will be produced in the next year. However, like governments, companies have greatly underestimated the time it would take to scale up manufacturing, and there may be incentives to overestimate production for good headlines or to win contracts. Estimates also only work for companies close to manufacturing a vaccine, but not for tracking all manufacturing capacity. This is because companies that do not yet have a vaccine or are further away from production will not be making announcements about vaccine capacity for said vaccine. This approach is likely to be more reliable in the shorter term, as longer-term estimates will miss new investments in facilities and new vaccines coming on stream. It also only gives an insight into COVID-19 vaccine capacity, does not track specific inputs for COVID-19 vaccines and granular, full datasets are only available for those who pay for it.

A second approach is surveys of existing and potential vaccine manufacturers. CEPI has led the way by conducting two global surveys of vaccine manufacturers’ bioreactor capacity and the type of vaccines they can produce. These have given invaluable insight into manufacturing capacity. UNICEF is also investigating undertaking their own survey. However, low survey response rates underestimate the available capacity, and the status of those not replying might differ substantially from those who do. Surveys also only provide a snapshot of production. For example, CEPI’s 2020 survey found there was almost no mRNA vaccine manufacturing capacity in the world, which was true at the time, but mRNA manufacturing was scaled up quickly and they eventually produced a plurality of the world’s COVID-19 vaccines. Capacity can change quite rapidly where existing facilities are either committed to a product (another vaccine or non-vaccine) or become available. For this reason, capacity could drop as the initial crisis wanes and plants are switched to other, possibly more profitable, products. Manufacturers of newer vaccine technologies that use biotechnology facilities can be more vulnerable to rapid changes in product. The process for drug product manufacture (fill-finish) is also more universal and therefore prone to use by other products. A survey taken once a year will not capture this.

Three actions to improve tracking of vaccine manufacturing capacity

There is now much better information about global vaccine capacity. Both production and policy are undoubtedly improved because of the work of CEPI, Airfinity, UNICEF, Duke University and others. However, there should be a broader approach than just looking at finished product manufacturing capacity for COVID-19 vaccines or capacity in LMICs only. Three actions would improve the status quo, by building on the work conducted so far:

1. Create robust estimates for the world’s vaccine manufacturing capacity—stratified by antigen, type of capacity, and region—by leveraging the best of statistical sampling, expert interviews, surveys, and public announcements. Estimates should include information on how quickly production can be scaled up during a crisis and capture important vaccine inputs (such as glass vials, single use bioreactor bags, filters and filtration equipment, primary seeds, culture media, syringes, lipids). This is crucial for COVID-19, but even more importantly for future pathogens: it will help plan additional capacity where needed, identify bottlenecks, and scale up production more quickly.

   This research should be undertaken with sustainability in mind, and collected regularly, so that this information is up to date during any future crisis.

2. Rely on rich ecosystems of researchers to get more reliable results. When different approaches generate similar results, there can be greater confidence in them. When they differ, it will force all groups to re-examine their approach, help identify methodological flaws and lead to better answers.

3. Cast a wider net, instead of assessing capacity for a single product. There is a need to understand capacity for the whole platform (the underlying technology for different types of vaccines). Fill and finish capacity can be obtained by repurposing sterile injectable manufacturing lines; drug substance capacity can also be repurposed. Essentially, there are three capacities that should be understood: capacity available now, capacity which can be reconfigured with some modest changes, and capacity that can be reconfigured with more sizable investments. This will help in maximizing the efficiency of the current system and understanding where best to strengthen it.

Conclusion

Poor understanding of vaccine capacity was one of the reasons why the world did not have the infrastructure needed to manufacture sufficient doses. This slowed down the global response to COVID-19, with tremendous human and economic costs. Not knowing what the current capacity is, also makes it harder to plan future capacity. Tracking manufacturing capacity is a starting point to create robust estimates for the world’s vaccine manufacturing capacity. This information is critical to inform current conversations on the infrastructure needed both to make COVID-19 vaccine manufacturing more robust and the system better able to deal with future pandemics, as well as the policy solutions that should be put in place to get us there.

Compared to the costs of a pandemic or of building new infrastructure, research into how to optimize manufacturing is exceptionally cheap. The world should make this small but crucial investment.