Skip Navigation Bar

ClinicalTrials.Gov Background

Session 1: Background and Legislative Requirements Background

Dr. Deborah A. Zarin Team

Jerry Sheehan

Betsy Humphreys


  • Deborah Zarin

Project Leads

  • Annice Bergeris
  • Nicholas Ide
  • Alison Robbins
  • Tony Tse
  • Rebecca Williams

Quality Assurance

  • John Frye
  • Cherryl Macalintal
  • Alex Valentine

Systems Development

  • Jane Fun
  • John Gillen
  • Alex Kostyukovsky
  • Russell Loane
  • Allison Yu

Domain Expert

  • William Harlan


  • Tamia Whitfield

Ethical and Scientific Rationale for Increased Clinical Trial Transparency

  • Subjects put themselves at risk
  • Subject’s right to be informed
    • All available options, including ongoing trials
    • Previous research, including completed trials
  • Avoid redundant trials
  • Support Evidenced-Based Medicine (EBM)
  • Detect reporting problems
    • Lack of publication
    • Unexplained changes to protocol

Recent Events: Lack of Transparency in Clinical Research

Image of title page of Special Article from the New England Journal of Medicine: “Selective Publication of Antidepressant Trials and Its Influence on Apparent Efficacy” Erick H. Turner, et al, 2008
Figure from Turner et al, 2008 showing the relationship between FDA decisions on antidepressant drugs and results reported in the published literature: FDA decision positive – 97% of studies published and agree with FDA decision; 3% not published. FDA decision questionable – 50% published and conflict with FDA decision; 50% not published. FDA decision negative – 12% published and agree with FDA decision; 21% published and conflict with FDA decision, 67% not published.

Source: Figure 1A. Turner et al. (NEJM, 2008)

Image of News Release from House Committee on Energy and Commerce, January 14, 2008 “Dingell, Stupak to Continue ENHANCE Trial Investigation” Second Image of News Release from House Committee on Energy and Commerce, January 14, 2008 “Dingell, Stupak to Continue ENHANCE Trial Investigation”

The investigation was launched following concerns...

  • …although the ENHANCE trial ended in April 2006, the data had not yet been released.
  • …[the sponsors] did not register the clinical trial in a timely manner.
  • …[the sponsors] attempted to change the study endpoints, and thus the study results, prior to the public release of the results.
Image of article from JAMA-Express, Original Contribution “Gastrointenstinal Toxicity with Celecoxib vs nonsteroidal Anti-inflammatory Drugs for Osteoarthritis and Rheumatoid Arthritis – The CLASS Study: A Randomized Controlled Trial” FE Silverstein, et al. JAMA. 2000 Sep 13;284(10):1247-55. Indicates that the main outcome measure was related to 6-month treatment period.

Source: Silverstein FE et al. JAMA. 2000 Sep 13;284(10):1247-55.

Figure from an article in the British Medical Journal showing that difference in performance between Celecoxic, Diclofenac, and Ibuprofen was much larger at 6-months than at 12-months.

Kaplan-Meier estimates for ulcer complications according to traditional definition. Results are truncated after 12 months, no ulcer complications occurred after this period. Adapted from Lu 2001.

Source: Jüni P, Rutjes AW, Dieppe PA. BMJ. 2002 Jun 1;324(7349):1287-8.

  • Prospective registry
  • Interventional and observational “trials”
    • Approved by IRB (or equivalent)
    • Conform to regulations of national health authorities
  • All sponsor types
  • All intervention types
  • Summary of key protocol items
    • Consistent with ICMJE and WHO
  • Links to publications, other info
  • Does not include unpublished results

Policies and Users

Many policies motivate registration in Public Law 110-85, FDAMA, Section 113, Best Pharmaceuticals for Children Act, Maine law, Policies of sponsoring/funding agencies (e.g., NIH, VA), International Committee of Medical Journal Editors (ICMJE), Ottawa Statement, WHO. Many users of Recruitment (e.g., patients, physicians), Journal editors, Research funders, Systematic reviewers, Health Policy Makers

International Registries and Number of Trials (as of 2/4/08)

RegistryNo. of Trials 50,564
ISRCTN (UK) 6,514
Australian New Zealand Clinical Trial Registry (ANZCTR) 1,973
Netherlands Trial Registry 1,111
UMIN Clinical Trials Registry (Japan) 943
Chinese Clinical Trial Register (ChiCTR) 34
Clinical Trials Registry – India (CTRI) 16

Characteristics of

  • Web-based registration system (PRS)
  • Register through organizational accounts
    • Key part of validation
  • XML data upload (<20% records)
  • Data are dynamic—”data provider” can modify data at any time
    • Changes are tracked

Homepage -

Screenshot of the homepage <>
Screenshot of the Advanced Search page on <>
Screenshot showing results of a search of for search with the following criteria: Recruitment status: Open studies, Study type:  Interventional studies, Conditions:  Heart disease, Interventions: Stent, Sponsors: NHLBI, Locations: United States, Massachusetts. The search returns two studies

Screenshot of the Archive Site <> showing the history of changes to trial NCT00430612.
Screenshot showing how changes to a registration record are displayed in the archive. General Statistics

Graph showing the number of records registered in and the number of organizational accounts between January 1, 2004 and December 1, 2007. As of December 1 there were more than 49,000 registered trials from more than 4,700 organizations. The number of registration records and the number of organizational accounts increased steadily after September 2005. The number of registration records and the number of organizational accounts increased dramatically between May 2005 and September 2005 due to the implementation of the ICMJE registration policy. Statistics (as of February 4, 2008)

Total Registered Trials 50,564 100%
Type of Trial*
Observational 7,701 15%
Interventional 42,836 85%
Drug & Biologic 32,835  
Surgical Procedure 10,856  
Behavioral, Gene Transfer, Other 4,809  
Device 2,628  
International Sites
US Only 25,485 50%
Non-US only 15,982 32%
US & Non-US mixed 3,934 8%
Missing 5,163 10%

* 27 records missing Study Type information Statistics Continued (as of February 4, 2008)

Trials by Data Provider
US Federal (including NIH) 16,430 33%
University, other 19,803 39%
Industry 14,331 28%
User Statistics
Page Views per month 20 Million  
Unique visitors per month 500,000  

Organizational Accounts

Fed accounts 71
NIH accounts 26
Industry 1,783
Other 2,912

Validation of Data for Registry

  • Issues
    • Quality Assurance: Are the entries as accurate and informative as possible?
    • Validation (w/o access to protocols): Is the fundamental information true?
  • QA Process at
    • System of automated and manual checks
    • Staff works with data provider to correct/improve records
    • Links inserted
Table 4. Attributes of Entries in "Primary Outcome Measure" Field.
AttributeFrequency (N=657)* %Examples from
Vague 17 Clinical response
Domain without specific measure 19 Glucose regulation
Severity of symptoms of schizophrenia
Specific measure without time frame 23 Intravenous glucose-tolerance test
Structured clinical interview -- positive and negative syndrome scale
No. of hospitalizations
Time frame without specific measure 10 Tumor response at 3 mo
Freedom from progression at 2 yr
Improvement in glucose-control over 16-wk period
Specific measure and time frame 31 Change in glycosylated hemoglobin from baseline to 6 mo
Mortality from any cause at 30 days

*Frequencies are based on a review of 657 records from the top 10 drug companies, ranked according to data from IMS Health on the volume of U.S. sales.23 Phase 2, 3, and 4 Trials were included.


Image showing three different locations on the carotid artery that were initially identified in the ENHANCE study as sites for the primary outcome measurement. Source:  Kastelein JJ, Sager PT, de Groot E, Veltri E. Am Heart J. 2005 Feb;149(2):234-9.

Source: Kastelein JJ, Sager PT, de Groot E, Veltri E. Am Heart J. 2005 Feb;149(2):234-9.

Screenshot of the registration record for the ENHANCE trial, showing that the stated primary outcome measure -- "change in mean carotid artery intima media thickness" -- is not specific enough to identify the three initially chosen locations.