Sample dataset: 20 complete trials, one Parquet file per table

README.md

@@ -0,0 +1,456 @@
+---
+license: cc-by-nc-4.0
+task_categories:
+  - tabular-classification
+  - tabular-regression
+language:
+  - en
+tags:
+  - clinical-trials
+  - clinicaltrials.gov
+  - biotech
+  - pharmaceutical
+  - drug-development
+  - longitudinal
+  - protocol-amendments
+  - sample
+pretty_name: ClinicalTrials.gov Version History (Sample)
+size_categories:
+  - 100K<n<1M
+configs:
+  - config_name: core
+    data_files: "core/*.parquet"
+  - config_name: interventions
+    data_files: "interventions/*.parquet"
+  - config_name: annotation_unposted_events
+    data_files: "annotation_unposted_events/*.parquet"
+  - config_name: annotation_violation_events
+    data_files: "annotation_violation_events/*.parquet"
+  - config_name: arm_groups
+    data_files: "arm_groups/*.parquet"
+  - config_name: avail_ipds
+    data_files: "avail_ipds/*.parquet"
+  - config_name: central_contacts
+    data_files: "central_contacts/*.parquet"
+  - config_name: collaborators
+    data_files: "collaborators/*.parquet"
+  - config_name: conditions
+    data_files: "conditions/*.parquet"
+  - config_name: documents
+    data_files: "documents/*.parquet"
+  - config_name: history_change_unposted_events
+    data_files: "history_change_unposted_events/*.parquet"
+  - config_name: history_metadata
+    data_files: "history_metadata/*.parquet"
+  - config_name: history_original_other_outcomes
+    data_files: "history_original_other_outcomes/*.parquet"
+  - config_name: history_original_primary_outcomes
+    data_files: "history_original_primary_outcomes/*.parquet"
+  - config_name: history_original_secondary_outcomes
+    data_files: "history_original_secondary_outcomes/*.parquet"
+  - config_name: id_aliases
+    data_files: "id_aliases/*.parquet"
+  - config_name: intervention_arm_links
+    data_files: "intervention_arm_links/*.parquet"
+  - config_name: intervention_other_names
+    data_files: "intervention_other_names/*.parquet"
+  - config_name: ipd_info_types
+    data_files: "ipd_info_types/*.parquet"
+  - config_name: keywords
+    data_files: "keywords/*.parquet"
+  - config_name: location_contacts
+    data_files: "location_contacts/*.parquet"
+  - config_name: locations
+    data_files: "locations/*.parquet"
+  - config_name: mesh_condition_ancestors
+    data_files: "mesh_condition_ancestors/*.parquet"
+  - config_name: mesh_conditions
+    data_files: "mesh_conditions/*.parquet"
+  - config_name: mesh_intervention_ancestors
+    data_files: "mesh_intervention_ancestors/*.parquet"
+  - config_name: mesh_interventions
+    data_files: "mesh_interventions/*.parquet"
+  - config_name: misc_removed_countries
+    data_files: "misc_removed_countries/*.parquet"
+  - config_name: other_outcomes
+    data_files: "other_outcomes/*.parquet"
+  - config_name: overall_officials
+    data_files: "overall_officials/*.parquet"
+  - config_name: phases
+    data_files: "phases/*.parquet"
+  - config_name: primary_outcomes
+    data_files: "primary_outcomes/*.parquet"
+  - config_name: reference_retractions
+    data_files: "reference_retractions/*.parquet"
+  - config_name: references
+    data_files: "references/*.parquet"
+  - config_name: results_adverse_event_groups
+    data_files: "results_adverse_event_groups/*.parquet"
+  - config_name: results_adverse_events
+    data_files: "results_adverse_events/*.parquet"
+  - config_name: results_adverse_other_event_stats
+    data_files: "results_adverse_other_event_stats/*.parquet"
+  - config_name: results_adverse_other_events
+    data_files: "results_adverse_other_events/*.parquet"
+  - config_name: results_adverse_serious_event_stats
+    data_files: "results_adverse_serious_event_stats/*.parquet"
+  - config_name: results_adverse_serious_events
+    data_files: "results_adverse_serious_events/*.parquet"
+  - config_name: results_baseline
+    data_files: "results_baseline/*.parquet"
+  - config_name: results_baseline_categories
+    data_files: "results_baseline_categories/*.parquet"
+  - config_name: results_baseline_class_denom_counts
+    data_files: "results_baseline_class_denom_counts/*.parquet"
+  - config_name: results_baseline_class_denoms
+    data_files: "results_baseline_class_denoms/*.parquet"
+  - config_name: results_baseline_denom_counts
+    data_files: "results_baseline_denom_counts/*.parquet"
+  - config_name: results_baseline_denoms
+    data_files: "results_baseline_denoms/*.parquet"
+  - config_name: results_baseline_groups
+    data_files: "results_baseline_groups/*.parquet"
+  - config_name: results_baseline_measure_classes
+    data_files: "results_baseline_measure_classes/*.parquet"
+  - config_name: results_baseline_measure_denom_counts
+    data_files: "results_baseline_measure_denom_counts/*.parquet"
+  - config_name: results_baseline_measure_denoms
+    data_files: "results_baseline_measure_denoms/*.parquet"
+  - config_name: results_baseline_measurements
+    data_files: "results_baseline_measurements/*.parquet"
+  - config_name: results_baseline_measures
+    data_files: "results_baseline_measures/*.parquet"
+  - config_name: results_more_info
+    data_files: "results_more_info/*.parquet"
+  - config_name: results_outcome_analyses
+    data_files: "results_outcome_analyses/*.parquet"
+  - config_name: results_outcome_analysis_groups
+    data_files: "results_outcome_analysis_groups/*.parquet"
+  - config_name: results_outcome_categories
+    data_files: "results_outcome_categories/*.parquet"
+  - config_name: results_outcome_class_denom_counts
+    data_files: "results_outcome_class_denom_counts/*.parquet"
+  - config_name: results_outcome_class_denoms
+    data_files: "results_outcome_class_denoms/*.parquet"
+  - config_name: results_outcome_classes
+    data_files: "results_outcome_classes/*.parquet"
+  - config_name: results_outcome_denom_counts
+    data_files: "results_outcome_denom_counts/*.parquet"
+  - config_name: results_outcome_denoms
+    data_files: "results_outcome_denoms/*.parquet"
+  - config_name: results_outcome_groups
+    data_files: "results_outcome_groups/*.parquet"
+  - config_name: results_outcome_measurements
+    data_files: "results_outcome_measurements/*.parquet"
+  - config_name: results_outcome_measures
+    data_files: "results_outcome_measures/*.parquet"
+  - config_name: results_pflow
+    data_files: "results_pflow/*.parquet"
+  - config_name: results_pflow_drop_withdraw_reasons
+    data_files: "results_pflow_drop_withdraw_reasons/*.parquet"
+  - config_name: results_pflow_drop_withdraws
+    data_files: "results_pflow_drop_withdraws/*.parquet"
+  - config_name: results_pflow_groups
+    data_files: "results_pflow_groups/*.parquet"
+  - config_name: results_pflow_milestone_achievements
+    data_files: "results_pflow_milestone_achievements/*.parquet"
+  - config_name: results_pflow_milestones
+    data_files: "results_pflow_milestones/*.parquet"
+  - config_name: results_pflow_periods
+    data_files: "results_pflow_periods/*.parquet"
+  - config_name: review_reset_reasons
+    data_files: "review_reset_reasons/*.parquet"
+  - config_name: secondary_ids
+    data_files: "secondary_ids/*.parquet"
+  - config_name: secondary_outcomes
+    data_files: "secondary_outcomes/*.parquet"
+  - config_name: see_also_links
+    data_files: "see_also_links/*.parquet"
+  - config_name: std_ages
+    data_files: "std_ages/*.parquet"
+  - config_name: submission_infos
+    data_files: "submission_infos/*.parquet"
+  - config_name: topics
+    data_files: "topics/*.parquet"
+  - config_name: version_history
+    data_files: "version_history/*.parquet"
+  - config_name: who_masked
+    data_files: "who_masked/*.parquet"
+---
+
+# ClinicalTrials.gov Version History: Sample
+
+A free sample of the **ClinicalTrials.gov Version History** dataset by
+[Anamnesis Data](https://anamnesisdata.com): **20 complete trials, every version of each, in
+79 tables**. It shows the shape, grain and joins of the full dataset, which covers
+~605K trials across their full version histories from 1999 to today.
+
+ClinicalTrials.gov shows only the latest state of a trial. This dataset keeps every version
+the registry has published, so you can see what a trial said on any past date, when a sponsor
+moved a primary completion date, or when the arms and outcomes were changed.
+
+- **20 trials**, 741 versions in total (from 8 to 162 per trial).
+- **79 tables**, 221,561 rows. Each table is its own config, so it loads on its own.
+- **Complete:** every row of every table for these trials is included, so joins across tables work as
+  they do in the full dataset.
+- **Chosen for coverage:** the trials were picked so that every table has rows, including the
+  rarer ones (results, adverse events, documents, retractions, unposted annotations).
+- The newest change in this sample was posted by ClinicalTrials.gov on **2026-08-28**.
+
+For the full dataset, see [anamnesisdata.com](https://anamnesisdata.com) or contact
+[licensing@anamnesisdata.com](mailto:licensing@anamnesisdata.com).
+
+## Quick start
+
+Download the repository (or clone it); every table is a folder of Parquet files. With
+[DuckDB](https://duckdb.org), from the repository folder:
+
+```sql
+.read setup.sql          -- one view per table, named after its folder
+.read sample_queries.sql -- 12 example queries, each with a comment explaining it
+
+SELECT nct_id, nct_version, overall_status, lead_sponsor_name
+FROM core
+QUALIFY row_number() OVER (PARTITION BY nct_id ORDER BY nct_version DESC) = 1;
+```
+
+Or load one table with the `datasets` library:
+
+```python
+from datasets import load_dataset
+
+core = load_dataset("anamnesis-data/clinical_trials_history_sample", "core", split="train")
+```
+
+DuckDB can also read the files in place, without downloading them:
+
+```sql
+SELECT overall_status, count(*) AS versions
+FROM 'hf://datasets/anamnesis-data/clinical_trials_history_sample/core/*.parquet'
+GROUP BY overall_status;
+```
+
+## Example queries
+
+`sample_queries.sql` holds 12 queries, from basic to multi-table.
+
+**Basic**
+
+| Query | What it shows |
+|---|---|
+| `current_record` | The current record of every trial (one row per trial) |
+| `versions_per_trial` | Versions per trial and the span of posting dates |
+| `status_changes_for_one_trial` | Every version where one trial's status changed |
+| `current_interventions` | Interventions of each trial's current version (`core` + `interventions`) |
+| `current_trials_by_phase_and_status` | Phase and status counts (`core` + `phases`) |
+
+**Multi-table**
+
+| Query | Tables joined |
+|---|---|
+| `trial_profile` | `core`, `phases`, `conditions`, `interventions`, `locations`, `collaborators` |
+| `registry_as_of_date` | What the registry said on a past date, next to today's record (`core`, `conditions`) |
+| `primary_completion_date_slips_by_phase` | How often sponsors push out completion dates, by phase (`core`, `version_history`, `phases`) |
+| `primary_outcome_results` | Results for primary outcomes (`results_outcome_measures`, `_classes`, `_categories`, `_measurements`, `_groups`) |
+| `serious_adverse_event_rates_by_group` | Serious adverse events by arm (`results_adverse_serious_events`, `_stats`, `results_adverse_event_groups`) |
+| `arms_and_their_interventions` | Which intervention each arm receives (`arm_groups`, `intervention_arm_links`, `interventions`) |
+| `sponsor_site_footprint` | Where each sponsor's trials run (`core`, `locations`) |
+
+## Reading the data
+
+- **Versions.** `nct_version` is ClinicalTrials.gov's own 0-based version number for a trial. Almost
+  every table has one set of rows per `(nct_id, nct_version)`; join tables to `core` on both columns.
+  Tables without `nct_version` are per trial (join on `nct_id`); `version_history` uses
+  `ct_gov_version`, which equals `nct_version`.
+- **There is no "current" flag.** The current record is the highest `nct_version` held for a
+  trial, as in the `current_record` query. That is the latest version in this dataset, which can be
+  older than the live registry record if the trial has changed since it was last collected.
+- **Obsolete IDs.** Every table ends with `is_obsolete` and `obsolete_by`. `is_obsolete` is true
+  when ClinicalTrials.gov now resolves the `nct_id` to a different registration (for example a merged
+  duplicate); `obsolete_by` is the surviving ID when known. None of the trials in this sample is obsolete.
+- **Raw text.** Text fields are shown as collected from ClinicalTrials.gov and are not cleaned. Many
+  contain HTML-escaped characters, for example `&amp;`, `&#x27;` and `&#x2F;` in sponsor names, group
+  titles and descriptions, and some descriptions contain HTML tags. Decode them before display or matching.
+- **Key columns.** The tables below give each table's key.
+
+## Tables
+
+| Config | Rows | Key |
+|---|---|---|
+| `core` | 741 | `nct_id`, `nct_version` |
+| `interventions` | 2,593 | `nct_id`, `nct_version`, `intervention_type`, `intervention_name` |
+| `annotation_unposted_events` | 63 | `nct_id`, `nct_version`, `ord` |
+| `annotation_violation_events` | 30 | `nct_id`, `nct_version`, `ord` |
+| `arm_groups` | 2,348 | `nct_id`, `nct_version`, `arm_label` |
+| `avail_ipds` | 108 | `nct_id`, `nct_version`, `ord` |
+| `central_contacts` | 544 | `nct_id`, `nct_version`, `ord` |
+| `collaborators` | 790 | `nct_id`, `nct_version`, `collaborator_name` |
+| `conditions` | 1,839 | `nct_id`, `nct_version`, `condition` |
+| `documents` | 152 | `nct_id`, `nct_version`, `ord` |
+| `history_change_unposted_events` | 63 | `nct_id`, `ct_gov_version`, `ord` |
+| `history_metadata` | 20 | `nct_id` |
+| `history_original_other_outcomes` | 58 | `nct_id`, `ord` |
+| `history_original_primary_outcomes` | 49 | `nct_id`, `ord` |
+| `history_original_secondary_outcomes` | 159 | `nct_id`, `ord` |
+| `id_aliases` | 75 | `nct_id`, `nct_version`, `alias_nct_id` |
+| `intervention_arm_links` | 4,159 | `nct_id`, `nct_version`, `intervention_name`, `intervention_type`, `arm_label` |
+| `intervention_other_names` | 626 | `nct_id`, `nct_version`, `intervention_name`, `intervention_type`, `other_name` |
+| `ipd_info_types` | 284 | `nct_id`, `nct_version`, `info_type` |
+| `keywords` | 2,280 | `nct_id`, `nct_version`, `keyword` |
+| `location_contacts` | 2,741 | `nct_id`, `nct_version`, `location_ord`, `contact_ord` |
+| `locations` | 65,638 | `nct_id`, `nct_version`, `ord` |
+| `mesh_condition_ancestors` | 169 | `nct_id`, `mesh_id` |
+| `mesh_conditions` | 46 | `nct_id`, `mesh_id` |
+| `mesh_intervention_ancestors` | 184 | `nct_id`, `mesh_id` |
+| `mesh_interventions` | 52 | `nct_id`, `mesh_id` |
+| `misc_removed_countries` | 44 | `nct_id`, `country` |
+| `other_outcomes` | 863 | `nct_id`, `nct_version`, `ord` |
+| `overall_officials` | 801 | `nct_id`, `nct_version`, `name`, `role` |
+| `phases` | 753 | `nct_id`, `nct_version`, `phase` |
+| `primary_outcomes` | 2,212 | `nct_id`, `nct_version`, `ord` |
+| `reference_retractions` | 162 | `nct_id`, `nct_version`, `reference_ord`, `retraction_ord` |
+| `references` | 2,387 | `nct_id`, `nct_version`, `ord` |
+| `results_adverse_event_groups` | 299 | `nct_id`, `nct_version`, `group_id` |
+| `results_adverse_events` | 91 | `nct_id`, `nct_version` |
+| `results_adverse_other_event_stats` | 23,011 | `nct_id`, `nct_version`, `event_ord`, `stat_ord` |
+| `results_adverse_other_events` | 5,380 | `nct_id`, `nct_version`, `event_ord` |
+| `results_adverse_serious_event_stats` | 17,325 | `nct_id`, `nct_version`, `event_ord`, `stat_ord` |
+| `results_adverse_serious_events` | 5,120 | `nct_id`, `nct_version`, `event_ord` |
+| `results_baseline` | 91 | `nct_id`, `nct_version` |
+| `results_baseline_categories` | 1,719 | `nct_id`, `nct_version`, `measure_ord`, `class_ord`, `category_ord` |
+| `results_baseline_class_denom_counts` | 2,605 | `nct_id`, `nct_version`, `measure_ord`, `class_ord`, `denom_ord`, `count_ord` |
+| `results_baseline_class_denoms` | 665 | `nct_id`, `nct_version`, `measure_ord`, `class_ord`, `denom_ord` |
+| `results_baseline_denom_counts` | 360 | `nct_id`, `nct_version`, `denom_ord`, `count_ord` |
+| `results_baseline_denoms` | 94 | `nct_id`, `nct_version`, `denom_ord` |
+| `results_baseline_groups` | 351 | `nct_id`, `nct_version`, `group_id` |
+| `results_baseline_measure_classes` | 931 | `nct_id`, `nct_version`, `measure_ord`, `class_ord` |
+| `results_baseline_measure_denom_counts` | 270 | `nct_id`, `nct_version`, `measure_ord`, `denom_ord`, `count_ord` |
+| `results_baseline_measure_denoms` | 90 | `nct_id`, `nct_version`, `measure_ord`, `denom_ord` |
+| `results_baseline_measurements` | 6,639 | `nct_id`, `nct_version`, `measure_ord`, `class_ord`, `category_ord`, `measurement_ord` |
+| `results_baseline_measures` | 597 | `nct_id`, `nct_version`, `measure_ord` |
+| `results_more_info` | 91 | `nct_id`, `nct_version` |
+| `results_outcome_analyses` | 1,275 | `nct_id`, `nct_version`, `outcome_ord`, `analysis_ord` |
+| `results_outcome_analysis_groups` | 2,509 | `nct_id`, `nct_version`, `outcome_ord`, `analysis_ord`, `group_id` |
+| `results_outcome_categories` | 5,254 | `nct_id`, `nct_version`, `outcome_ord`, `class_ord`, `category_ord` |
+| `results_outcome_class_denom_counts` | 4,569 | `nct_id`, `nct_version`, `outcome_ord`, `class_ord`, `denom_ord`, `count_ord` |
+| `results_outcome_class_denoms` | 1,974 | `nct_id`, `nct_version`, `outcome_ord`, `class_ord`, `denom_ord` |
+| `results_outcome_classes` | 4,466 | `nct_id`, `nct_version`, `outcome_ord`, `class_ord` |
+| `results_outcome_denom_counts` | 3,641 | `nct_id`, `nct_version`, `outcome_ord`, `denom_ord`, `count_ord` |
+| `results_outcome_denoms` | 1,614 | `nct_id`, `nct_version`, `outcome_ord`, `denom_ord` |
+| `results_outcome_groups` | 3,641 | `nct_id`, `nct_version`, `outcome_ord`, `group_id` |
+| `results_outcome_measurements` | 12,764 | `nct_id`, `nct_version`, `outcome_ord`, `class_ord`, `category_ord`, `measurement_ord` |
+| `results_outcome_measures` | 1,660 | `nct_id`, `nct_version`, `outcome_ord` |
+| `results_pflow` | 91 | `nct_id`, `nct_version` |
+| `results_pflow_drop_withdraw_reasons` | 2,282 | `nct_id`, `nct_version`, `period_ord`, `dw_ord`, `reason_ord` |
+| `results_pflow_drop_withdraws` | 576 | `nct_id`, `nct_version`, `period_ord`, `dw_ord` |
+| `results_pflow_groups` | 265 | `nct_id`, `nct_version`, `group_id` |
+| `results_pflow_milestone_achievements` | 1,974 | `nct_id`, `nct_version`, `period_ord`, `milestone_ord`, `achievement_ord` |
+| `results_pflow_milestones` | 501 | `nct_id`, `nct_version`, `period_ord`, `milestone_ord` |
+| `results_pflow_periods` | 145 | `nct_id`, `nct_version`, `period_ord` |
+| `review_reset_reasons` | 267 | `nct_id`, `nct_version`, `scope`, `scope_path`, `ord` |
+| `secondary_ids` | 494 | `nct_id`, `nct_version`, `secondary_id` |
+| `secondary_outcomes` | 8,930 | `nct_id`, `nct_version`, `ord` |
+| `see_also_links` | 207 | `nct_id`, `nct_version`, `ord` |
+| `std_ages` | 1,487 | `nct_id`, `nct_version`, `std_age` |
+| `submission_infos` | 23 | `nct_id`, `ord` |
+| `topics` | 91 | `nct_id`, `topic_source`, `url` |
+| `version_history` | 741 | `nct_id`, `ct_gov_version` |
+| `who_masked` | 1,358 | `nct_id`, `nct_version`, `role` |
+
+## The trials
+
+| NCT ID | Lead sponsor | Versions | Current status |
+|---|---|---|---|
+| [NCT00356369](https://clinicaltrials.gov/study/NCT00356369) | GlaxoSmithKline | 15 | COMPLETED |
+| [NCT01224106](https://clinicaltrials.gov/study/NCT01224106) | Hoffmann-La Roche | 162 | COMPLETED |
+| [NCT01727336](https://clinicaltrials.gov/study/NCT01727336) | Acceleron Pharma, Inc., a wholly-owned subsidiary of Merck &amp; Co., Inc., Rahway, NJ USA | 26 | TERMINATED |
+| [NCT01844986](https://clinicaltrials.gov/study/NCT01844986) | AstraZeneca | 79 | ACTIVE_NOT_RECRUITING |
+| [NCT02148952](https://clinicaltrials.gov/study/NCT02148952) | Harvard School of Public Health (HSPH) | 11 | COMPLETED |
+| [NCT02403674](https://clinicaltrials.gov/study/NCT02403674) | Merck Sharp &amp; Dohme LLC | 52 | COMPLETED |
+| [NCT03013751](https://clinicaltrials.gov/study/NCT03013751) | Mezzion Pharma Co. Ltd | 11 | COMPLETED |
+| [NCT03045406](https://clinicaltrials.gov/study/NCT03045406) | Fadoi Foundation, Italy | 24 | COMPLETED |
+| [NCT03052816](https://clinicaltrials.gov/study/NCT03052816) | University Hospitals Cleveland Medical Center | 8 | COMPLETED |
+| [NCT03061734](https://clinicaltrials.gov/study/NCT03061734) | Allodynic Therapeutics, Inc | 26 | COMPLETED |
+| [NCT03094715](https://clinicaltrials.gov/study/NCT03094715) | University Hospital Heidelberg | 12 | TERMINATED |
+| [NCT03204318](https://clinicaltrials.gov/study/NCT03204318) | Myovant Sciences GmbH | 36 | COMPLETED |
+| [NCT03204331](https://clinicaltrials.gov/study/NCT03204331) | Myovant Sciences GmbH | 34 | COMPLETED |
+| [NCT03709680](https://clinicaltrials.gov/study/NCT03709680) | Pfizer | 78 | COMPLETED |
+| [NCT03785340](https://clinicaltrials.gov/study/NCT03785340) | Ocugen | 10 | COMPLETED |
+| [NCT03800134](https://clinicaltrials.gov/study/NCT03800134) | AstraZeneca | 55 | ACTIVE_NOT_RECRUITING |
+| [NCT04363801](https://clinicaltrials.gov/study/NCT04363801) | Leap Therapeutics, Inc. | 31 | TERMINATED |
+| [NCT04518228](https://clinicaltrials.gov/study/NCT04518228) | National Institute of Allergy and Infectious Diseases (NIAID) | 42 | TERMINATED |
+| [NCT04605978](https://clinicaltrials.gov/study/NCT04605978) | Institut de Recherches Internationales Servier | 14 | COMPLETED |
+| [NCT05080764](https://clinicaltrials.gov/study/NCT05080764) | Biofrontera Inc. | 15 | COMPLETED |
+
+## Source
+
+The trial records in this dataset come from [ClinicalTrials.gov](https://clinicaltrials.gov),
+a database operated by the U.S. National Library of Medicine (NLM) at the
+National Institutes of Health.
+
+**Data currency:** the newest change in this sample was posted by ClinicalTrials.gov on
+**2026-08-28**. ClinicalTrials.gov is updated daily; check
+[ClinicalTrials.gov](https://clinicaltrials.gov) for current information.
+
+### Modifications made to the ClinicalTrials.gov data
+
+- **Restructured:** records were reshaped into tables, one per config.
+- **Subset of fields:** each config publishes a subset of the ClinicalTrials.gov
+  fields; the Parquet schema of each config lists exactly which.
+- **Version index:** `nct_version` equals ClinicalTrials.gov's own 0-based
+  version number.
+- **Partial dates:** dates reported only as `YYYY-MM` or `YYYY` are stored as the
+  first day of that period (`YYYY-MM-01`, `YYYY-01-01`); where a `*_precision`
+  column exists, it records the original precision. Date values in any other
+  format are stored as null.
+- **Interventions:** entries missing a name or a type are omitted, and duplicate
+  `(intervention_name, intervention_type)` entries within one version are
+  collapsed to the first occurrence (the later entries' descriptions are not
+  kept).
+- **Added columns:** `is_obsolete` and `obsolete_by` on every table.
+- **Always-null columns:** in `core`, `expanded_access_status_for_nct_id`,
+  `version_holder`, `first_mcp_post_date`, `first_mcp_post_date_type` and
+  `estimated_results_first_submit_date` are null on every row.
+- **Sample selection:** 20 trials out of the full dataset, with all of their rows.
+- **Text encoding:** not modified (see "Raw text" above).
+
+Apart from the changes listed above, field values are carried over as returned
+by ClinicalTrials.gov.
+
+## License
+
+### The underlying ClinicalTrials.gov records
+
+The trial records themselves (registration and results information submitted
+by study sponsors) are from ClinicalTrials.gov, a United States Government
+database. **No proprietary rights are claimed in them.** They are available to
+anyone at no charge directly from [ClinicalTrials.gov](https://clinicaltrials.gov),
+subject to the [ClinicalTrials.gov Terms and Conditions](https://clinicaltrials.gov/about-site/terms-conditions),
+which note that some ClinicalTrials.gov data may be subject to the copyright of
+third parties. Nothing on this page restricts your use of data you obtain from
+ClinicalTrials.gov.
+
+### This dataset
+
+The compilation published here (its relational structure, schema,
+version history, derived fields, and documentation) is licensed
+under **[CC BY-NC 4.0](https://creativecommons.org/licenses/by-nc/4.0/)** to the
+extent the author holds rights in it. This sample is public and is not gated. Please cite
+it (see [Citation](#citation)) and attribute the underlying data to ClinicalTrials.gov.
+
+For commercial licensing, contact
+[licensing@anamnesisdata.com](mailto:licensing@anamnesisdata.com) or visit
+[anamnesisdata.com](https://anamnesisdata.com).
+
+## Citation
+
+If you use this dataset in research, please cite:
+
+```bibtex
+@misc{clinicaltrials_history_sample_2026,
+  title  = {ClinicalTrials.gov Version History: Sample},
+  author = {Bret Gaulin},
+  year   = {2026},
+  url    = {https://huggingface.co/datasets/anamnesis-data/clinical_trials_history_sample}
+}
+```

annotation_unposted_events/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8cca708200f8e7b475e706459f268ba82bc1c673696ed1532541d1bdc84ebc16
+size 2846

annotation_violation_events/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:471319b61a9acc1cdf2f9bd8b6748d9a0446f7f50944dcb7c4cef9b078f05716
+size 4454

arm_groups/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b76a59dab27dba93c8b73cb6cbf146224a9acc8d6ae668e5c8fddf89acbf42ee
+size 11776

avail_ipds/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9173945452ebb040366c09aa066fea677ff4560eb7fc95ae25b9e488729f7b73
+size 3644

central_contacts/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:445855842574bc4eb228e1b4d1d03cacf845c8a54549a2c05c600fb39f1e874b
+size 5576

collaborators/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dcefb9331acfe3d17e050a04f604027be07f77f83370d3ca0d5e23e0c244ac3f
+size 3779

conditions/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a4f5c2064a1e9673aa826aaf7aa0ea4f720b1e06cbdbec130fd39c7707048d3e
+size 3470

core/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b0c36963bafa4af8c53a4952d24f5a2fde609f123f9e7c0f3a723b779b688bbb
+size 126125

documents/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:741ef51b4370a44d0888f7f867eab3ffc4a7ee47c2b58128a1f1c8f65a00657a
+size 5737

history_change_unposted_events/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3903d51294111c70b2eb9f9668c33e3925cbb452fc822fab3e163c4dee83d4de
+size 2864

history_metadata/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fb3cdaf70f6d9a221cf12cc878966ea8517ca747d20361430747fee1c1d67a28
+size 12528

history_original_other_outcomes/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b0c86c5779b2c4247130573c8774c3619a0e8b4a75f509c9937b088838b6860
+size 5755

history_original_primary_outcomes/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f91668cb82b2f27d5ce4345f1a258f5f771385786d9d57137e228435a3c63b11
+size 6652

history_original_secondary_outcomes/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:754fafcf0fee5bbf368e0eca02f5f2a05729efa0fda421b2072221a090c5e579
+size 13686

id_aliases/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f85a9617a892649327f144f041fd520d08f66dab25d005172d5620eaeba05786
+size 1705

intervention_arm_links/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:62bb25689dbb942989e7d308e71bec647e0f24c085774a42ff7014b5407405c4
+size 8095

intervention_other_names/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1d5d4f56d021e847bab6b09f2061ae1d655db51a2a97726d1601a2ff0c8fa9cf
+size 4061

interventions/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:64ba34312e15eb6fe261a6fb359a60f723e088edc28e128558f9da5b7fb36e4b
+size 9490

ipd_info_types/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dff68f1246dc050195a7fe2615ae2b9ca2031a5b6e3ee4a13e7724d1a3ee279f
+size 2013

keywords/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:35f5f72d6d515442e551ca028f02e3e0ea0a0d8a0b435ea5b6f2c14b71a42c6b
+size 4224

location_contacts/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7e278c7589b25f2962b769b758f40889fa7acf88ebb9bc4fe9f2fa7cfa867336
+size 17444

locations/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:679144df21f34dabc7fefbddc5daaf93c047c7c2ba7d58d11ee1c4a76bbef26e
+size 170343

mesh_condition_ancestors/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd3e24ed8b5c5bccca13b19bf5b0bcb57782f8db994f01a8ddf3aac27aff318d
+size 3524

mesh_conditions/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d3fc635509bd3363f4be266291845f6e993f8ce410c211c734f29602ca2f5331
+size 2357

mesh_intervention_ancestors/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bb342095e8a0aeb361fc3ec99cd5a8a948d3aac646a56376eba5c6b3848082ba
+size 3392

mesh_interventions/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:52f1af5cf6e07f0ab7693d9219b348f5de847aa0f3daafa05b7d1e2c4b0998ec
+size 2500

misc_removed_countries/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b7c3ef4edd02e999c50622185b8a085971d4b2a4c498310bcb196f0106837935
+size 1588

other_outcomes/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a8d4f7b65aca904cc808c5e6f9f3f2466fef75a369bf526b5fe2606874dfad8
+size 10205

overall_officials/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:deb64e4a466cf7f1f53294ba26c4f7be3582ae123856d86816a02fd0578a4f99
+size 4203

phases/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:db19e54b41ab4b16e665b340523fbb8fdd32ee9b6eb1190ac89fecaf20c6bf24
+size 2320

primary_outcomes/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c64d4a36dd6d9efacf585673d4870badd34679a8eaec6030096ec14693d75a0f
+size 20325

reference_retractions/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:447490507abd1b7560509c9ec4b44d9db73c4a407b2b0b66c296959c48d34aab
+size 3053

references/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fcf1813a88d19c2c7461304dbab8cdf96557ca276b6e51f66bfdcaa481626b4d
+size 16855

results_adverse_event_groups/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:77c3ebf72d9dbea677a76a23ad32606b480fe26a8b9ce4f888f95e2fd35b186a
+size 14928

results_adverse_events/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6036f91489bdb9dec2cb7c5a8828c241160a78a7aa06886e4576f5875caa885b
+size 7975

results_adverse_other_event_stats/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eb9fd7452bd7c76ba2e4fafa354f97774a73cd82ef0c5c2929c2127a99eee611
+size 22440

results_adverse_other_events/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a61d1bfdabea0e65c978b614315fb40414347d96cff23eb8e268770951b7c0f
+size 23322

results_adverse_serious_event_stats/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a7b58f5de370651612859d3964e8899c34d0d93fbd282a31a55775d13b10821d
+size 13658

results_adverse_serious_events/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:945ada300974be89eb87b6b3091a6ac1b925de4e520b2ae44a575184b0271441
+size 18948

results_baseline/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c7a466aa2f4cee0256f11d9b9fd7a021644f4096e6e524019c01c4c375415f4
+size 4547

results_baseline_categories/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d3c63b53e7dc4b533c412ae15c008d671fad5f845b8226962817c65f1a89ba94
+size 4658

results_baseline_class_denom_counts/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2e26947b3b864f913b5788771eac9174aa8680dda813f5ca1e0582d14eb77caa
+size 4805

results_baseline_class_denoms/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2a88393a09103eb1b532ac82539ffd8a675684d81cee7a8b26928f69a87715b8
+size 3132

results_baseline_denom_counts/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2236b4ed1b68072b117bef9ad4ae67ecac6953f1228085a522beb0ae6ccf234d
+size 3633

results_baseline_denoms/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cbdb4dcb924f7e120f0b0b6d6f0de566c5e02ffa502c6a7341daa12afeb01541
+size 2249

results_baseline_groups/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d808dbee90fa7daea30d020d390be75c15443c5764c36c1780ede10e59c61217
+size 10698

results_baseline_measure_classes/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:96e482c03c34284510424cd0eba5a33e061ff5936a4ea118981b979de3990d21
+size 4037

results_baseline_measure_denom_counts/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:80126743a49598b7bdd38ff8202d237b19332200945b132b44aab0f260b4af4d
+size 3011

results_baseline_measure_denoms/part-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:df016cc7d4241243e8834425be371eb3820cf59e226dbd606eac6c94135c9327
+size 2315