Population adjustment reshuffled state clinical-trial listing counts

California had the largest raw count in MedDataUSA's October 2026 ClinicalTrials.gov snapshot: 5,310 unique studies with at least one in-state location carrying an included recruitment status. After dividing the same study–state counts by Census Vintage 2025 population, California had 13.5 listings per 100,000 residents and did not rank among the ten highest rates.

The District of Columbia had the highest population-adjusted result, at 108.3 study–state presences per 100,000 residents, followed by Massachusetts at 41.5 and Rhode Island at 40.8. The shift shows how strongly a state population denominator changes the descriptive comparison. It does not rank patient access, physical research sites, open seats, or research quality.

Quick read

• California led the raw count with 5,310 study–state presences; Texas had 4,654 and New York 4,295.
• D.C. had 751 presences and 693,645 residents, producing the highest population-adjusted rate at 108.3 per 100,000.
• The source was already restricted to studies whose overall study status was Recruiting, Not yet recruiting, or Enrolling by invitation. The analysis then separately required an individual U.S. location status to be one of those same three values.
• Population adjustment does not measure participant access, travel burden, eligibility, site capacity, physical sites, or available seats.

The denominator changed the picture

Large states dominated the unadjusted counts. California, Texas, New York, Florida, and Ohio formed the top five. Those figures count each NCT identifier once within a state when at least one listed location passes the location-status filter; a multistate study can contribute to several states.

Dividing by resident population brought smaller jurisdictions with substantial research activity to the top. After D.C., the highest rates were Massachusetts, Rhode Island, Maryland, and Minnesota. California's 13.5 per 100,000 and Texas's 14.7 were far below their raw-count positions.

This is the article's distinct comparison: it shows what changes when the same governed numerator is normalized by population. The earlier MedDataUSA article explained why registry location counts do not measure open seats. This analysis retains those limits and adds a Census denominator rather than repeating the raw-count story.

Two inherited status filters

The preserved snapshot did not begin with every ClinicalTrials.gov record. Its API query filtered overall study status to Recruiting, Not yet recruiting, or Enrolling by invitation and matched records to the United States.

Within that already-filtered study population, MedDataUSA applied a second rule: a state counted only when an individual U.S. location had one of the same three statuses. “Not yet recruiting” means recruitment has not started, while “Enrolling by invitation” is not general public recruitment. Even “Recruiting” does not establish a remaining seat for a particular person.

Keeping both filters explicit matters. The result describes study–state presences within the inherited overall-status population after an additional location-status test. It is not a measure of all registered research activity.

Why per-capita is not access

Population is a transparent general denominator, but it is not an eligible-patient denominator. D.C. draws participants across jurisdictional borders and contains major research institutions relative to its resident population. Compact northeastern states have similar cross-border effects.

The calculation does not adjust for disease prevalence, age, sponsor strategy, trial phase, travel time, public transportation, contact responsiveness, eligibility, or the number of locations within a state. Multiple qualifying locations for the same study still collapse to one study–state presence.

For those reasons, the rate can describe listing intensity relative to population but cannot say where it is easiest for a patient to enroll.

Methodology and limitations

MedDataUSA reused the preserved ClinicalTrials.gov API v2 snapshot retrieved October 7, 2026. The acquisition query required an overall study status of Recruiting, Not yet recruiting, or Enrolling by invitation and a United States location match. From that result, we retained U.S. locations whose individual location status was Recruiting, Not yet recruiting, or Enrolling by invitation, then counted each NCT ID once per state.

We joined state names to Census Vintage 2025 resident population estimates for July 1, 2025 and calculated study–state presences per 100,000 residents. All 50 states and D.C. matched a population row. The 69,914 national denominator is the sum of state-specific unique-study counts; because multistate studies contribute to multiple states, it is not a count of unique studies nationally.

Registry records are sponsor-submitted and dynamic. The population rates are descriptive normalizations, not estimates of physical sites, seats, enrollment, eligibility, capacity, geographic accessibility, or research quality.

Sources

• National Library of Medicine, ClinicalTrials.gov API v2.
• ClinicalTrials.gov, study data structure.
• U.S. Census Bureau, Vintage 2025 state population totals.
• Reproducible preserved extracts and analysis: the preserved internal analysis package.