Shortages of generic drugs now and during the COVID-19 pandemic illustrate the vulnerabilities of the U.S. pharmaceutical supply chain. The risks of overconcentrating upstream pharmaceutical manufacturing in China and downstream formulation in India has highlighted the need for better understanding of and transparency in the global pharmaceutical supply chain.
The United States, European Union, and other countries are establishing measures to improve supply chain transparency that will help people understand where raw materials are being sourced, where upstream active pharmaceutical ingredient producers (API) are located, and where the finish product is manufactured.
Despite ongoing attempts to add transparency to the U.S. supply chain, such information is not systematically available. Reasons are plentiful, but fragmented and incomplete datasets along with concerns about data being privately owned by companies prevent the use of such information to improve understanding of upstream supply chain risks that can inform policy decisions.
Global health agencies, on the other hand, have achieved considerable success over the last decade with mapping each step of the supply chain and understanding the risks at each step—for HIV/AIDS, malaria, and tuberculosis (TB) medicines.
Any policy instrument, regulatory action, or financing support requires understanding where a supply chain's potential weak points are. This process starts with knowing who supplies the finished product for each therapeutic category, where those suppliers are located, from whom they purchase their active pharmaceutical ingredients (APIs), and from whom the API manufacturers buy their key starting materials (KSM).
Despite ongoing attempts to add transparency to the U.S. supply chain, such information is not systematically available
The supply chains for HIV/AIDS, malaria, and TB medicines face various shortcomings at different stages in the supply chain. Those challenges include limited government funding; patients' inability to pay higher prices for finished products; small and unpredictable market sizes, as seen with second-line TB drugs; and small or shrinking demand, as seen with pediatric HIV treatments where prevention of mother-to-child transmission program success reduced the number of children needing antiretrovirals (ARVs).
These deficiencies often manifest as supply shortages and higher prices because fewer manufacturers invest in adequate production capacity. As new products are launched, whether supply capacity can meet market demand is uncertain, raising questions about whether rollout should be slowed to align with supply constraints. Budget constraints have forced ministries of health and global health agencies to find opportunities to improve efficiency using data.
Together, small market sizes and limited budgets made global health agencies more thoughtful and serious about understanding their supply side markets, including upstream supply markets.
Between 2013 and 2015, with funding from Unitaid, a team of researchers and I began mapping the supply chain for HIV/AIDS, malaria, and TB medicines from the finished product to the APIs, and in some cases even the KSMs.
You Can't Fix What You Don't See
At the start, we faced the same challenges—lack of data—but by combining trade flow data with procurement data, regulatory data, and information gathered at site visits and from interviews with manufacturers, the team created detailed supply maps such as the following.
To accomplish this goal, our analysis followed the supply chain as if it were a river, starting from its mouth (the finished product) to the source of its tributaries (the KSM producers). We started by creating a simple map based on what could be imputed from existing trade flow data. Then we went to each of the junctions and sources of the tributaries (API manufacturers) to verify flow (production) in that tributary.
We depict that with different dotted lines in the map to be explicit about our degree of confidence in the links between finished product, API, and KSM producers.
Upstream Supply Chains
This map shows which active ingredient manufacturers supply the finished product manufacturers of HIV/AIDS, malaria, and tuberculosis treatments, based on trade flow data.

WHO-PQ or SRA approval
Sourcing linkage credence
High
Medium
Low
Approved
Under review
Not approved
Finished Product
Manufacturer
Finished Product
Manufacturer
Finished Product
Manufacturer
Finished Product
Manufacturer
Finished Product
Manufacturer
Active Ingredient
Manufacturer
Active Ingredient
Manufacturer
Active Ingredient
Manufacturer
The chart is a simplified example of a supply chain mapping exercise.
WHO-PQ: World Health Organization prequalification. SRA: stringent regulatory authority. Approved denotes a manufacturer with
WHO-PQ or SRA approval. Under review denotes WHO-PQ or SRA applications are under review. Not approved includes manufacturers
with products under development.
High denotes more than two sources of verification or high confidence in the linkage. Medium denotes two sources of verification or
medium confidence. Low denotes a single source of verification or low confidence.
Chart: CFR/Allison Krugman • Source: API MDIS Project 2014, WDI at the University of Michigan

WHO-PQ or SRA approval
Approved
Under review
Not approved
Sourcing linkage credence
High
Medium
Low
Finished Product
Manufacturer
Finished Product
Manufacturer
Finished Product
Manufacturer
Active Ingredient
Manufacturer
Active Ingredient
Manufacturer
The chart is a simplified example of a supply chain
mapping exercise.
WHO-PQ: World Health Organization prequalification.
SRA: stringent regulatory authority. Approved denotes a
manufacturer with WHO-PQ or SRA approval. Under
review denotes WHO-PQ or SRA applications are under
review. Not approved includes manufacturers
with products under development.
High denotes more than two sources of verification or
high confidence in the linkage. Medium denotes two
sources of verification or medium confidence. Low
denotes a single source of verification or low confidence.
Chart: CFR/Allison Krugman
Source: API MDIS Project 2014,
WDI at the University of Michigan
Trade flow data and interviews were also instrumental in estimating the production capacity of each plant. Although this task was particularly challenging for multiproduct plants, basic information such as plant size, reactor capacity, and tableting line capacity provided a solid foundation for establishing supply estimates. The regulatory status of each facility, including approvals from the U.S. Food and Drug Administration, World Health Organization prequalification, and local country regulators were also documented.
The volume flow and supplier mapping allowed us to estimate the risks of market concentration in suppliers by geography. The Global Fund [PDF], U.S. Agency for International Development (USAID), and the Global Drug Facility for TB have carried out similar initiatives to analyze the upstream market. The extent to which these efforts incorporated systematic mapping and capacity assessment, however, varied across the projects.