Skip to content
Trial Agents

European site-network evidence

European clinical trial sites: where activity concentrates and what site networks look like

European site activity is concentrated enough that the same institutions repeatedly appear across different trial types, but not concentrated enough for a short list of famous centres to explain the market. The strongest networks combine repeat high-performing institutions with a much broader long tail of sites.

AP-HP, Vall d’Hebron and other major academic systems recur across disease and modality cohorts, yet the top ten institutions represented only 16.1% of identifiable glioma site-trial participations and 11.1% in Phase III infectious disease. Country leadership also changes by therapeutic area, and the country with the most sites is not always the country with the most planned patients.
What the evidence shows

Recurring findings across European trials

1

A small group of institutions recur across very different trial portfolios

Assistance Publique–Hôpitaux de Paris led the RNA Phase I–II cohort with participation in 9 of 31 trials and the Phase III infectious-disease cohort with 43 trial participations. Vall d’Hebron appeared in 8 of 31 RNA trials and tied for the lead in glioblastoma/glioma with 13 trial participations. These repeated appearances make large academic systems useful anchor candidates, but not a complete site strategy.

2

Even the leading institutions explain only a minority of the usable network

The top ten listed sites accounted for 16.1% of identifiable site-trial participations in glioma and 11.1% in Phase III infectious disease. In RNA Phase I–II, the leading site appeared in 29.0% of trials, but the cohort still involved 281 distinct European sites. Site selection therefore needs both repeat performers and deliberate long-tail coverage.

3

The leading country changes materially by therapeutic area

France was the largest platform in the infectious-disease Phase III cohort with 950 site entries and in glioma with 190 site placements. Poland led IBD with 611 site slots, while Spain had the largest RNA Phase I–II site footprint with 100 of 442 site slots. A generic ‘best countries for trials’ ranking therefore hides the indication-specific network.

4

The country with the most sites is not always the country with the most planned patients

In RNA Phase I–II, Spain had the largest site footprint with 100 of 442 site slots, while France had the largest planned participant allocation with 427 of 2,601 participants. Site density and expected recruitment contribution should therefore be evaluated as separate variables rather than using site count as a recruitment proxy.

5

Disease and modality reshape the network rather than simply changing its size

IBD showed a broad multi-country network led by Poland and repeat commercial/academic centres, while glioma concentrated more visibly around specialist academic hospitals such as Heidelberg and Vall d’Hebron. RNA Phase I–II still reached 281 distinct sites despite being an advanced-modality cohort. The practical choice is not ‘centralised versus broad’ in the abstract; it depends on the patient pathway, specialist capability and recruitment burden of the programme.

6

Institution-level activity is a stronger public signal than a simple investigator leaderboard

The most consistent CTIS evidence across these cohorts is the recurring site or institution, country footprint and planned participant allocation. Investigator-level records can be useful, but they need person-level normalization and role consistency before they should be compared as a public ranking. For operational planning, institution recurrence is the more robust first screen.

What this changes in trial planning

  • Separate site recurrence, country footprint and planned patient allocation when ranking countries and institutions.
  • Use major recurring institutions as anchors, then build long-tail coverage around indication-specific recruitment needs.
  • Do not infer recruitment capacity from site count alone; planned participant allocation can point somewhere else.
  • Prefer institution-level evidence until investigator identities and roles are normalized consistently enough for fair comparison.

Scope note. This page synthesizes findings from separate Trial Agents analyses of European CTIS cohorts. Each statistic remains tied to the phase, disease, modality and time window of its underlying analysis. Cross-cohort patterns are used as planning signals, not pooled estimates or causal claims.

Underlying evidence

Read the analyses behind these findings

More evidence

Related European trial intelligence

Never miss a new clinical trial insight

We continuously analyze new and updated clinical trial data and publish practical, evidence-led reports for clinical leaders.