Charted: Top-3 Countries Hold 34% of Late-Stage Trial Country-Uses — 52% of Phase III Oncology Facilities

Pivotal trials are where evidence becomes a label claim. The geography of those sites is therefore not a logistics footnote — it is a claim about whose patients, whose clinics, and whose regulators shape the evidence base the rest of the world inherits. This desk restates industry late-stage site geography from IQVIA country-prioritization shares, ClinicalTrials.gov / WHO ICTRP site listings, and a 2026 Phase III oncology facility census, then places those shares against WHO-style disease-burden (DALY) anchors.

The interactive dashboard above is built as a mismatch lens, not a simple league table. Toggle country-use vs oncology facility metrics, filter by region, switch therapy geography, and flip Phase III initiation year. The punchline is deliberately double-booked. On pipeline country-uses (2021–23), the United States holds about 16%, China about 13%, and Spain about 5% — a Top-3 near 34%, with the Top-10 near 58%. On unique Phase III oncology research facilities, the United States alone is about 40%, and US + China + Germany clear roughly 52%. Same question, two meters: initiation geography is spreading toward China; facility stock still looks American.

Top-3 country site share is the wrong comfort number

Desks like Top-3 because it fits a slide. The useful version of the number is which Top-3 — and which perimeter.

On IQVIA's country-use framing, utilization consolidated toward the United States and China even as Western Europe remained the largest regional bloc. Western Europe's share of global country-uses fell from about 32% in 2019 to 25% in 2023 (a 21% relative decline). Eastern Europe fell harder — about 17% to 11%. North America rose from roughly 19% to 23%. China rose from about 10% to 15% (+57% relative). That is not "Europe left the building." It is Europe still first as a bloc, while the country leaderboard and the facility leaderboard disagree about who sits second and third.

Spain's outsized role in European Phase III initiations (often third globally on Citeline/ABPI ranks) is a reminder that country-use, initiation counts, and unique facilities are not interchangeable. Germany can lead Europe on facility stock while Spain leads on trial starts. Treating "Europe" as one row hides that internal rotation.

The table's lesson is procedural: do not splice initiation ranks into facility shares. China overtaking the United States on Phase III starts in 2024 is a real signal about where protocols are opening. It does not erase a facility census in which US institutions still dominate unique site identity — especially in oncology.

US, China, and Europe versus the burden they claim to treat

The core question is not "who has the most sites?" It is whether site geography tracks the disease burden those trials claim to address.

Approximate DALY shares (WHO GBD / Global Health Estimates style anchors, carried for gap math) put the United States near 7% of global burden and China near 18%. Sub-Saharan Africa is on the order of one-quarter of DALYs. South Asia is not far behind. Against those anchors, US country-use share (~16%) implies a gap ratio around 2.2× (sites over burden). Western Europe's ~25% country-use share against roughly 8–9% of DALYs is closer to 3×. China's ~13% country-use against ~18% DALYs is a mild under-index (~0.7×) — surprising to readers who only watch initiation headlines. Sub-Saharan Africa's ~1.8% site share against ~25% DALYs is a gap ratio near 0.07×: the largest mismatch in the set.

That mismatch is not proof of bad faith. Industry site selection optimizes for enrollment speed, investigator density, data quality, IP enforcement, and regulatory predictability. Those criteria correlate with high-income clinic infrastructure. The analytical claim is narrower: if pivotal evidence is generated where patients are easiest to instrument, external validity for high-burden regions is a residual, not a design target.

Cardiometabolic and infectious-disease programs show slightly more Rest-of-World share than rare-disease and CNS programs in the therapy lens. Rare/orphan and many oncology regimens still lean US–Europe. The dashboard's therapy control is a stylized roll-up of disclosed patterns — use it as a directional scanner, not a protocol census.

China's rise is real — and still not a burden match

Two narratives collide. Narrative A: China is taking over late-stage research. Narrative B: trials still ignore Asia's disease burden. Both can be locally true.

On starts, China more than doubled Phase III initiations into 2024 and overtook the United States on ABPI/Citeline ranks. On country-uses, China's share rose from ~10% to ~15% in five years. On oncology facilities, China is about 7% of unique research facilities — meaningful, not hegemonic. Domestic China-only programs (especially from China-headquartered sponsors, which IQVIA puts near 30% of global trial starts in 2024, many single-country) inflate start counts without exporting site diversity.

For global multi-regional trials aimed at FDA/EMA labels, China is increasingly a must-have region for enrollment math and for China NMPA strategy — not a substitute for US and EU evidence packages. The burden gap remains: China is under-weighted on country-uses relative to DALYs even while it leads some initiation tables. That is what a fast-growing base from a low share looks like.

Europe's share compression is a concentration story, not an exit

Western Europe's relative decline gets read as "Europe is dead for trials." The levels still say otherwise. At 25% of country-uses, Western Europe remains the largest regional bloc — larger than North America or China alone. What changed is the dispersion inside Europe and the opportunity cost against US/China growth.

Central and Eastern Europe's drop from ~17% to ~11% is the sharper operational story for CROs: geopolitics, slower start-up in some markets, and competition from Spain, Germany, UK, and France for the protocols that still want European diversity. Spain's Phase III initiation strength is the clearest European counter-example to decline fatalism — a country that punches far above its DALY weight (gap ratio >7× on our carried anchors).

For evidence policy, Europe's compression matters because EMA-relevant populations still need representation, while global protocols reallocate slots to US mega-sites and China. The risk is not zero European sites; it is thinner, more repetitive European site lists that look multinational on paper and monocultural in practice.

Facility stock versus initiation flow

Facility censuses and initiation ranks answer different desk questions.

• Initiations ask: where are new Phase III protocols opening this year?
• Country-uses ask: how often does a country appear across the active/starting pipeline?
• Unique facilities ask: how concentrated is the physical institutional base that can run those protocols?

The 2026 oncology facility mapping (roughly 6,600 unique facilities across 84 countries) finds the United States at ~40% of facilities and a strong correlation between facility counts and trial participation. Outside high-income settings and China, capacity remains thin and skewed toward industry-sponsored, multiregional, systemic-therapy work. Latin America, South Asia, and Sub-Saharan Africa show high shares of facilities that only appear on multiregional industry protocols — a dependency signature, not an indigenous Phase III engine.

That is why Top-3 facility share (~52%) is a harsher concentration meter than Top-3 country-uses (~34%). Starts can rotate; buildings and trained networks move slowly.

What the burden gap implies for regulators and sponsors

If US and Western European patients are systematically over-represented in pivotal packages, two failures follow.

First, effect sizes and adverse-event profiles may not travel. Pharmacogenomics, comorbidity mixes, concomitant medicines, and care pathways differ. Second, access narratives diverge from evidence narratives: high-burden countries are asked to trust — and eventually pay for — medicines validated elsewhere.

None of this requires a conspiracy. It requires noticing that "global trial" often means "global enough for the agencies that matter commercially." WHO ICTRP and ClinicalTrials.gov make the geography auditable. The audit is rarely done against DALYs.

Practical responses that show up in protocol design already: stratified regional enrollment floors, Africa and South Asia hub investments, decentralized elements that reduce site dependence, and indication-specific burden matching (infectious disease and some cardiometabolic programs lead; rare disease lags). The dashboard's Africa and South Asia gap ratios are the accountability metric those efforts have to move.

Caveats and how to read the meters

This print is a desk synthesis, not a primary registry scrape shipped as a paper. Country-use percentages follow IQVIA Institute regional and top-country restatements; facility percentages follow a published Phase III oncology facility census; initiation ranks follow ABPI/Citeline reporting; DALY shares are WHO-style regional anchors carried for gap ratios. Gap ratio = site share ÷ DALY share. Values near 1.0 are matched; values >>1 are over-trialed relative to burden; values <<1 are under-trialed.

Do not treat therapy-mix bars as audit-grade. They are directional composites. Do not compare 2024 initiation ranks to 2021–23 country-use shares as if they were the same vintage. Do not read China-only domestic starts as multi-regional diversity. Oncology facility geography is not identical to cardiometabolic or vaccine geography.

HHI figures in the dashboard are analytical on stated country buckets (0–10,000 scale). They are concentration meters for the disclosed table, not filings.

Reading the dashboard

Start with the country ladder on country-use, then flip to oncology facilities — watch the US bar jump and the Top-3 pie thicken. Use the regional path to see Europe's compression and China's rise on the same axis. The scatter puts site share against DALY share: points far above a 1:1 mental diagonal (high burden, low sites) are the equity problem; points far below (high sites, low burden) are the validity problem for exporting evidence. The bloc mismatch bars make the same point without requiring scatter literacy. Therapy and initiation-year controls are for stress-testing whether the story is oncology-only or initiation-year-only. It is not.

Bottom line: Top-3 country site share is real — about one-third of country-uses and half-plus of oncology facilities — but the ethically and scientifically sharper headline is the gap ratio. The US and Western Europe remain heavily over-indexed to burden; Sub-Saharan Africa and South Asia remain nearly absent from the pivotal map that decides what counts as evidence.