health
a dollar a dose, three suppliers
Three Chinese factories make the active ingredient for the only drug that prevents congenital Syphilis, it sells for under a dollar a Dose, and no one can see global demand — so supply breaks every few years
Problem statement
Benzathine benzylpenicillin (BPG) is a seventy-year-old long-acting injectable penicillin that is "the sole treatment recommended by the World Health Organization" to prevent mother-to-child transmission of syphilis and the standard secondary prophylaxis for rheumatic heart disease. Authors from WHO, the U.S. CDC, the Clinton Health Access Initiative and the World Heart Federation report that "by 2016, only three Chinese API manufacturers remained, continuing to supply to the global market today," while CHAI estimated at least 30 companies were producing the finished product in 2016, most for low- and middle-income markets where "prices for BPG in these markets are low, typically less than US$1 per dose." Because "commercially producing BPG is not considered profitable, leading to prolonged order times and large minimum orders," and because "the global community ... rarely had a comprehensive view" of demand, the market swings between glut and shortage: between 2014 and 2016 at least 39 countries reported BPG shortages, and in 2023 the United States — where "one supplier controls the entire market" — was again short during a syphilis surge. The unsolved problem is a structurally thin, low-margin, three-supplier API market for a medicine that more than 40 million people a year depend on, with no mechanism that lets buyers see demand or lets suppliers trust it.
Why this matters
Mother-to-child syphilis transmission "causes >350 000 adverse birth outcomes" a year, two-thirds of them stillbirths or neonatal deaths, and more than 39 million people live with rheumatic heart disease with about 320,000 deaths annually — both concentrated in low-resource settings. When BPG runs out, "patients receiv[e] non-recommended therapies," and the authors link the 2014–2016 shortages to increases in congenital syphilis. The market is also far smaller than the need: CHAI's estimate of global demand, 74–100 million 1.2-million-unit doses in 2016, is "less than half of what is required," with an estimated 200 million doses a year needed for rheumatic heart disease alone — so any effort to close the treatment gap would immediately collide with a supply base that cannot flex. Concentration compounds fragility: "API manufacturing is concentrated in two regions of China where environmental regulations have disrupted production in the past," and "limited compliance with good manufacturing practices further increases the risk of production suspension by international regulatory authorities."
What’s been tried and why it hasn’t worked
The 2014–2016 shortages "were triggered by concurrent and unrelated events that disrupted the API supply from two of the three active manufacturers," and the responses since have been visibility and quality measures rather than market structure: WHO surveys of national availability (a 2019 survey found six countries still short; WHO was again soliciting country data in 2023), WHO prequalification technical assistance (yet "as of July 2023, only one BPG product has received WHO prequalification"), and donor procurement channels such as the Global Fund and PEPFAR guidance requiring syphilis treatment availability. The authors are explicit that "while proposed mitigation actions contribute to the response, they fail to offer a structural solution to prevent shortages." The reasons are economic and informational: price ceilings, competition among many small buyers of non-quality-assured product, and low volumes "undermine economic incentives to supply quality-assured product that is costly to produce"; producers show "inflexibility in filling orders"; and even where a prequalified product exists, "availability of a quality-assured product may not stimulate new demand or buyers that opt for a switch from non-PQ products unless accompanied by wider market intervention." National-level obstacles add friction — South Africa has procured BPG since 2016 on temporary Section 21 authorisations because new suppliers struggle to register.
What would unlock progress
The authors' own prescription is market shaping: country-level forecasting built on disease surveillance ("robust disease surveillance data to quantify national demand, particularly for RHD and syphilis"), monitoring of stock levels and the number of authorised suppliers ("single-supplier nations are more susceptible to shortages"), global market transparency with direct manufacturer engagement, faster shortage notification, a maintained response plan with access to a backstop supply, and — structurally — a large payer or coordinated group of payers that aggregates demand for quality-assured product, as was done in HIV and vaccine markets. The transferable pattern is vaccine-style demand pooling and long-term supply agreements applied to an off-patent, sub-dollar antibiotic; the missing piece is the demand-visibility layer that would let a fourth or fifth API producer, or a stockpile, be financed on evidence rather than hope.
Entry points for student teams
A data team could prototype the demand-visibility tool the paper says does not exist: an open model that converts published syphilis (including antenatal) and rheumatic heart disease burden estimates plus public procurement records into country-level BPG dose demand and supplier-count risk scores, and shows where a single-supplier country meets a rising caseload. A market-design team could model the economics of BPG API production — minimum order sizes, batch cadence, price per dose — to size the offtake commitment or stockpile that would make a fourth quality-assured API line viable, drawing on published vaccine advance-market-commitment designs. A regulatory team could map registration pathways for BPG in a set of high-burden countries and design a collaborative-registration route for prequalified product. Relevant skills: epidemiology and health-data analysis, health economics/market shaping, pharmaceutical regulation, supply-chain modelling.
Genome — every gene is a door
Structural cousins — same reason stuck, other fields
Sources
Seghers F., Taylor M.M., Storey A., Dong J., Wi T.C., Wyber R., Ralston K., Dongmo Nguimfack B., "Securing the supply of benzathine penicillin: a global perspective on risks and mitigation strategies to prevent future shortages," International Health 16(3):279–282 (May 2024), doi:10.1093/inthealth/ihad087, author manuscript at accessed 2026-08-18, re-verified 2026-08-21 go to source ↗
verification notes (working record)
The collection team’s own sourcing notes for this brief, kept verbatim:
All quotations and figures (three Chinese API manufacturers by 2016 and today; ≥30 finished-product companies; <US$1 per dose; 39 countries short in 2014–2016; 74–100 million doses demand vs. ~200 million needed for RHD and >6 million 2.4-MU doses for adult syphilis; one WHO-prequalified product as of July 2023; >350,000 adverse birth outcomes; >39 million with RHD and ~320,000 deaths; U.S. single-supplier shortage from May 2023; South Africa Section 21) are from the author manuscript of Seghers et al. 2024 (International Health) read in full on 2026-08-18; the paper is a commentary by WHO/CDC/CHAI/WHF authors and thus a tier-1 agency-authored gap statement funded by the Gates Foundation. A search-result summary of another source claimed a fourth API manufacturer in Austria producing under certified GMP; that claim was not verified against a primary text and is not used in the brief — flag for verification. The U.S. Bicillin L-A shortage continuing into 2026 was reported in a search summary (CDC/urgent-care sources) and is likewise not asserted here. `constraint:supply-chain` is applied literally (API concentrated in three producers in two Chinese regions). `constraint:coordination` was considered and rejected on filter (2): removing coordination failure would not fix a sub-dollar price and unviable API economics — the binding constraints are economic and informational. `failure:unviable-economics` (humanitarian/commodity variant) is the primary failure mode; `failure:not-attempted` was rejected because WHO prequalification, Global Fund/PEPFAR channels and WHO surveys are serious attempts. `stakeholders:multi-institution` passes: WHO (prequalification, surveillance), donors (demand aggregation), national ministries (registration and procurement), and API/FPP manufacturers each hold a non-substitutable piece. Related collection briefs: `health-sterile-injectable-manufacturing-fragility` (U.S. sterile-injectable market fragility in general) and `health-cryptococcal-meningitis-flucytosine-access` (an essential drug unregistered where needed); this brief is the global API-concentration and demand-invisibility problem for one essential antibiotic.
Source type: Self-articulated (global-health agencies describing a market they steward)
Verified at intake 2026-08-18: gate (net) + adversarial source check + contested-tag second coding.
Related briefs (distinct sub-problems, cross-referenced 2026-08-18): `health-sterile-injectable-manufacturing-fragility`.
Reconciliation 2026-08-21: Flagged score-3 in the citation-drift triage sweep; full re-verification against the PMC author manuscript (https://pmc.ncbi.nlm.nih.gov/articles/PMC10987389/, fetched 2026-08-21) found the brief substantially clean — the citation metadata is exact (Seghers F, Taylor MM, Storey A, Dong J, Wi TC, Wyber R, Ralston K, Dongmo Nguimfack B; Int Health 16(3):279–282, May 2024; doi:10.1093/inthealth/ihad087), and every checked figure verified (three Chinese API manufacturers by 2016; ≥30 FPP companies in 2016; <US$1/dose; 39 countries 2014–2016; 74–100M doses vs. ~200M for RHD; one WHO-PQ product as of July 2023; >350,000 adverse birth outcomes, two-thirds stillbirth or neonatal death; >39M with RHD, ~320,000 deaths; US single-supplier shortage from May 2023; 2019 survey six countries; South Africa Section 21; ">40 million people globally depend on BPG every year"). One error found and fixed: the three-manufacturers quotation was paraphrase presented as quotation — the brief had "remained to supply the global market, a status they maintain today," where the paper reads "By 2016, only three Chinese API manufacturers remained, continuing to supply to the global market today."; corrected to the verbatim string, and the adjacent 30-companies clause re-anchored to the paper's actual claim (CHAI estimate, 2016, past tense). All 18 other quoted strings in the brief matched the paper character-for-character (allowing bracketed "receiv[e]" and marked ellipses). No claims removed.