International Clinical Trials Registry Platform search guide

The International Clinical Trials Registry Platform (ICTRP) is run by the World Health Organization. Its aim is that every clinical trial is registered and that the main facts about it are publicly accessible, wherever in the world it runs.

WHO does not register trials itself. Trials are registered in national or regional registries, and the ICTRP collects their records and makes them searchable in one place through its search portal.

This matters because many trials never appear on ClinicalTrials.gov. Studies run only in China, India, Iran, Japan, Germany, Australia, Brazil or many other countries are often registered locally, and a search limited to one US registry misses them.

• One search portal across many registries
• A common minimum dataset for each trial
• Links to the full record in the source registry

The ICTRP works with a network of registries. Primary registries meet WHO criteria for content, quality, accessibility, unique identification, technical capacity and administration. Data providers send their records to the search portal on a regular schedule.

Examples of registries that feed the portal include ClinicalTrials.gov, the EU Clinical Trials Register, ISRCTN, the Australian New Zealand Clinical Trials Registry, the Chinese Clinical Trial Registry, the Clinical Trials Registry of India, the German Clinical Trials Register, the Japanese registries, the Iranian Registry of Clinical Trials, the Pan African Clinical Trials Registry and the Brazilian Clinical Trials Registry.

Each registry has its own identifier format, its own fields and its own update rhythm. The portal maps their records to a shared set of fields, often called the WHO Trial Registration Data Set, which covers items such as the main ID, secondary IDs, sponsor, condition, intervention, key outcomes, recruitment status and countries.

The source registry remains the authority. When the portal and the source disagree, the record in the source registry is the one to trust and to cite.

European trials have their own history, and it shows up in ICTRP results.

• The EU Clinical Trials Register (EU CTR) publishes trials authorised under the former Clinical Trials Directive, based on the EudraCT database. EudraCT numbers follow the pattern of a year, six digits and a check number.
• The Clinical Trials Information System (CTIS) is the system used under the EU Clinical Trials Regulation. Trials authorised through CTIS get an EU trial number, and CTIS has its own public search.
• Many trials moved from the old system to CTIS during the transition, so one study can carry both a EudraCT number and an EU trial number.
• In the EU CTR, a multinational trial can appear as several country specific protocol records, which is why the same EudraCT number may show up more than once.

Our page on the EU clinical trial register explains how ClinicalEvidence covers European records alongside other registries.

Large trials are often registered in more than one registry: on ClinicalTrials.gov for US sites, in the EU for European sites, and in a national registry for a third country. Each registration becomes its own record in the portal.

The ICTRP tries to link records that describe the same trial. It uses the secondary IDs that sponsors enter, and the WHO Universal Trial Number exists precisely to give one trial one identifier across registries. When sponsors fill these fields in, the portal can group the records. When they do not, the same trial appears as separate, unlinked results.

Practical consequences for anyone counting trials:

• A raw count of ICTRP results overstates the number of distinct trials.
• Country specific EU records multiply one European trial into several rows.
• Statuses can disagree between registries, because each record is updated on its own schedule.
• Enrolment figures may describe the whole trial in one record and only national sites in another.

Before you report a number, deduplicate. Match on secondary IDs first, then compare titles, sponsor, intervention, phase and dates for records that remain, and decide which record you treat as primary.

The portal offers a basic search box and an advanced search. These habits make both more reliable.

• Start broad. Search the condition alone, read what comes back, then add the intervention. Narrow queries in a multi registry portal miss records that use different wording.
• Collect synonyms. Registries in different countries describe the same disease and drug in different ways, so include generic names, code names, abbreviations and spelling variants.
• Use the advanced fields. Filter by condition, intervention, recruitment status, phase, country and date of registration where the fields are available, and remember that not every registry fills every field.
• Check the source registry. Open the original record for every trial you plan to cite, because it is more complete and more current than the portal summary.
• Note the update rhythm. The portal imports records periodically, so a very recent change in a source registry may not appear yet.
• Export and deduplicate. Download your results and remove duplicates before any count or chart.
• Record your search. Keep the date, the terms and the filters, so a colleague can repeat the search.

Writing your question first with the PICOS framework helps you decide which terms belong in the search and which filters to apply.

The ICTRP is a registry of registries, not a results database. Most records describe what a trial planned to do, and only some registries carry posted results.

Publications are not part of the portal. A finished trial may have a peer reviewed paper in PubMed that the portal does not link to, so a complete evidence review checks the literature as well.

Data quality depends on the source. Free text fields, missing secondary IDs and stale statuses are common, and the portal can only show what registries send it.

Language can be a barrier too. Some registries hold records with titles or details in a national language, and English fields may be shorter than the original. For trials you plan to rely on, read the full source record and, where needed, ask a colleague who reads the language.

Finally, the portal is a search tool rather than an analysis tool. For counts, charts or a systematic review, plan to export results and clean them in your own workflow.

ClinicalEvidence searches ClinicalTrials.gov, PubMed, PubMed Central, the EU Clinical Trials Register and the WHO ICTRP where licences allow, in one question. The answer is written only from retrieved passages, and every sentence carries the NCT number or PMID of its source.

• Semantic search finds trials described with different words in different registries.
• The clinical trial table shows phase, status, sponsor, enrolment, countries and dates for each matching study, ready to filter.
• The Sources searched panel lists only the sources actually searched for your question.
• Each citation is verified against the index, and an answer without verified sources is not shown.

If you work mainly with the US registry, read our guide on how to search ClinicalTrials.gov, and for a full overview of what we index see the clinical trials database page.