BRENNEN

Industries / Health / Pharmaceuticals

Clinical research

Clinical research is the work of testing drugs, devices and other treatments in people, in planned studies called trials, to learn whether they are safe and whether they work.

1747

On 20 May 1747, aboard the British warship Salisbury, the naval surgeon James Lind picked twelve sailors with scurvy and compared six treatments on them, two men to each; by the end of May the two given oranges and lemons were almost recovered.

jameslindlibrary.org

42

Milestones achieved across 6 domains.

15

To watch for: announced, not yet confirmed outside.

17

Open deliverables. Nobody has shipped them.

In one minute

Where it started

In 1747 James Lind gave twelve sailors with scurvy six different treatments, side by side, and the two given citrus fruit recovered. Trials grew into a trade of their own once regulators demanded them: US rules reissued in 1970 set out what an "adequate and well-controlled" study must contain, and in 1982 the statistician Dennis Gillings founded Quintiles to run trials for drug makers for a fee. By 2010, Forbes reported, such contract research organizations had some role in almost all large trials.

Where it stands

ClinicalTrials.gov, the US public register, which also lists trials run elsewhere, held 612,549 study records when I queried it on 18 November 2026. Of these, 467,469 were trials of a treatment, 47,603 of those were recruiting, and 302,515 listed no site in the United States. Most drugs tested in trials are never approved: an MIT team counted 13.8% of drug development programmes from 2000 to 2015 that ended in approval. Rules now let sponsors aim their checks at the data that matter most instead of checking every entry at every site, and a few trials take their data straight from hospital records.

What is open

Many of the checks that keep trials honest are still done by hand, one study at a time: comparing what a trial promised with what it published, counting harms, reading inspectors' letters. ClinicalTrials.gov held posted results for 81,088 studies on 18 November 2026, and current AI models can read long protocols and papers. The jobs below run such checks across whole registries and score them against answer keys made by people, without contact with any participant.

Domains

AchievedWatch forOpen

Open in clinical research

3 of 17 open deliverables

Open · Trial design

Audit estimands at scale

A public table, with its code, covering every phase 3 trial on ClinicalTrials.gov that started in 2020 or later and has its statistical analysis plan attached: 2,130 trials on 18 November 2026. For each trial's main outcome, record the five parts of the estimand (population, treatment, outcome, how events such as stopping the drug are handled, and the summary measure), and flag the parts that are missing or unclear.

Built with what exists

the ClinicalTrials.gov API, filtered with AREA[LargeDocHasSAP]true AND AREA[Phase]PHASE3 AND AREA[StartDate]RANGE[2020-01-01,MAX]; pdftotext or PyMuPDF to pull text from the attached PDFs; Claude or Gemini, which can read a protocol and analysis plan of 200 pages in one pass; Kahan and colleagues' coding scheme.

Credit · unclaimed

Open · Finding and enrolling participants

Eligibility rules as data

For every recruiting trial for one common condition, such as type 2 diabetes, turn the free-text eligibility criteria into structured rules (age, laboratory cut-offs, required and excluded diagnoses and drugs) with standard medical codes. Publish the rules, the code and a hand-made answer key for a random sample, and rerun monthly.

Built with what exists

the ClinicalTrials.gov API; Chia's annotation scheme as a template; Claude or Gemini for a first pass; two people to build the answer key; Python.

Credit · unclaimed

Open · Sites and monitoring

Inspection findings linked to papers

A public table of every FDA warning letter to a clinical investigator from 2014 to 2026, labelled by kind of finding (false data, unreported harms, protocol not followed, poor records, consent and patient protection), linked to the trial's registry number and the papers from that trial, with a column saying whether each paper mentions the problem. Publish at trial level.

Built with what exists

the FDA's Clinical Investigator Inspection List on data.gov; the FDA warning letters search; the ClinicalTrials.gov API; the Europe PMC API; Claude or Gemini to read each letter and find trial numbers in it.

Credit · unclaimed

Where it sits