Number of clinical trial registrations by location, disease, phase of development, age and sex of trial participants (1999-2021)

Published: February 2022

The number of trials listed in the WHO International Clinical Trials Registry Platform (ICTRP) is reported by year, location (WHO region and country), disease (or condition), phase of development, age, sex and number of participants, for the period 1999–2021. Note that the ICTRP comprises both interventional and observational trials. See more on the scope of the ICTRP below.

See also:

Number of clinical trials by year, country, WHO region and income group
Percent of clinical trials with enrollment year preceding registration year

What you see | Scope and limitations | Data sources |  Current version

What you see

The data visualization above shows the number of trials listed on the ICTRP from 1999 to 2021 by:

  • year, worldwide (chart A)
  • phase of development (chart B)
  • year and WHO region (chart C)
  • health category (chart D)
  • country or area (chart E; colour coded by WHO region)
  • sex of trial participants (chart F)
  • age and life stage grouping of trial participants (chart G1 and G2)
  • number of trial participants (chart H)

Chart A shows the total number of trials, regardless of the number of trial sites involved. A multicountry trial is counted once for each participating region in chart C but once for each participating country in chart E. The default for other charts is one trial counted once regardless of the number of sites. When a selection is made, either by filtering a region or country, the count will reflect the same principle described above. The year corresponds to the date of enrollment of the first trial participant; in a small number of cases when this is not available the registration year is used. With respect to age of trial participants, charts G1 and G2 count trials once for each age group in their inclusion criteria.

The data can be visualized separately for interventional or observational trials by selecting the relevant tick box at the top left of the visualization. With respect to disease or conditions focus, trials on ' neglected tropical diseases' or ' R&D Blueprint pathogens', or 'trials on pregnant women or pregnancy-related trials' can also be viewed separately (select desired options using the circular buttons at the top left or top right of the data visualization).

Points to note:

  • The United States of America had the highest total number of trials registered during 1999-2021 (157,618), followed by China (80,333) and Japan (57,754) (chart E).
  • Of trials with an identified phase of development, the largest number of trials is in phase II (88,530) (chart B).
  • Data mining techniques were used to assign a primary disease to each trial (see approach under analysis below). It was not possible to match 107,915 trials (16%) to any disease. For some trials, it was only possible to classify them into broader categories.
  • Of the trials that were categorized into a disease or condition, 81% were for noncommunicable diseases, 15% for communicable, maternal, perinatal and nutritional conditions, and 4% for injuries (chart D).
  • Among the trials on the R&D Blueprint pathogens (tick the corresponding box at the top right to filter), the top three were on COVID-19 (10,142), severe acute respiratory syndrome (137) and Ebola virus disease (131) (chart D.2).
  • Of all trials, most (78%) included both males and females (chart F).
  • Of all trials, most (73%) included adults in their inclusion criteria while 13% included children; this information was unknown in 18% of trials (chart G1). Of the trials that specified children in their inclusion criteria, 7% were open to neonates in the early or late neonatal periods (hover the cursor on the corresponding bar in chart G2 and see information in the popup window).

To explore the data further

  • Tick the box (top left) to filter the results for only interventional or observational trials.
  • To filter the results for one or a combination of the following: year, WHO region, country or area, phase of development, disease or condition, or participant number, click on the relevant data element (for example on a region from the key in chart C; a point on the trend line for the year; or a bar beside the country, phase or disease of interest) in the relevant charts.
    -- For any of the above selections, information in the other charts will update accordingly, as relevant.
  • Hover the cursor on a data element of interest (for example a bar or a point on a trend line) to see more information in a popup window.
  • Hold the Ctrl key to select more than one option, for example two regions.
  • Undo a selection by clicking ‘undo’ or ‘reset’ near the bottom of the page or by clicking the same element again.

Analysis and limitations of the data analysis

  • There are several gaps in the ICTRP data source which required data cleaning to uniformly classify data elements (when possible). In some cases, data was missing/unavailable, e.g. on the country in which the trial is conducted or the phase of development of a trial.
  • Automated data mining was used to generate information on the primary disease investigated in each trial using text-based data fields.
    -- A list of disease synonyms was compiled using as a base the Unified Medical Language System (UMLS) and the International Classification of Diseases. This was complemented by synonyms drawn from the data, mostly to account for errors in data entry such as spelling errors or the use of abbreviations.
    -- An automated algorithm was applied to two data fields using the list of disease synonyms to generate the uniform disease classification field used in this analysis. The first field is based on free-text keywords provided by the registrant. The second is the scientific title of the trial. If the first field provided a match the second was not used.
    -- The first match that was closest to the beginning of the text field was selected. This was considered the primary disease investigated by the trial.
    -- It is possible that a trial has more than one disease focus; this is not captured in this analysis.
    -- The algorithm was refined through various iterations but as with any automated algorithm, it is likely that some trials were not correctly matched. A full description of the methods and approach is available in this paper:  Resource allocation for biomedical research: analysis of investments by major funders.
    -- This automated data mining method resulted in matching around 84% of the trials to a disease or condition.
  • The data presented in this visualization utilizes classifications that are not mutually exclusive. For example, a registered trial can recruit participants from multiple countries and regions. In this case, the trial will be counted once per region in chart C but once per country in chart E. The total number of trials across the two charts is therefore not comparable.
  • Classification of trials by age group and life stage follows the standardized age-disaggregation classification. Charts G1 and G2 count trials once for each age group and life stage in their inclusion criteria.
  • This analysis will be updated at regular time points but time lags with the scheduled updates by the data sources are inevitable. Accuracy and completeness of the information is the responsibility of the data source, see terms and conditions of use.