Clinical Trial Site Opportunity Monitor

• buyer: clinical research site business-development and feasibility teams
• result: one source-linked recent recruiting trial fit or material correction per paid event
• official source: ClinicalTrials.gov API v2 (no key required)
• first useful run: the bounded melanoma/FL preview below with baselineMode=emit_matches
• maximum charge: $0.18 with maxEvents=1
• unchanged replay: $0.00; an identical rerun emits no dataset row

bounded preview (buyer-visible)

Copy this input exactly. It is the storefront's bounded preview: one melanoma watchlist, Florida sites, a 30-day first-post window, at most one paid event, and a hard $0.18 ceiling. qualityTestMode=false means this is a real buyer run — it queries ClinicalTrials.gov, advances state, and charges only for emitted events. (The Apify Store smoke test uses the same input with qualityTestMode=true, which emits one unbilled summary and touches no source or state.)

{
  "watchlistId": "melanoma-fl-site-preview",
  "therapeuticAreas": ["melanoma"],
  "states": ["FL"],
  "countries": ["United States"],
  "lookbackDays": 30,
  "baselineMode": "emit_matches",
  "maxEvents": 1,
  "maxTotalChargeUsd": 0.18,
  "maxSourceRecords": 100,
  "stateStoreName": "melanoma-fl-site-preview",
  "resetState": false,
  "qualityTestMode": false
}

source-backed example

Fetched from the live ClinicalTrials.gov API v2 on 2026-11-03: NCT07621614, "A Multi-Institutional Central Pathology and Molecular Risk-Based Stratification Study of Surgical Management for Melanoma, Atypical Spitz/Spitzoid Tumors, and Other Atypical Melanocytic Neoplasms in Pediatric Patients." Recruiting, with Florida sites at Nemours Children's Clinic-Jacksonville (Jacksonville) and Nemours Children's Hospital (Orlando). Against the preview input it matches on therapeutic_area (melanoma), state (FL), and country (United States). Its first post date is 2026-06-02, outside a 30-day lookback from today, so a fresh preview run records it as baseline state at $0.00 instead of charging — the example proves matching against a real record; whether a given record charges depends on its first-post recency.

workflow

1. Preview: run the bounded input above once. You pay at most $0.18 and see exactly what a paid row looks like, including matched evidence and the official study link.
2. Baseline: rerun with baselineMode=silent (same watchlistId, settings, and stateStoreName) so historical recruiting trials establish state without paid events.
3. Schedule: keep the identical input on a daily or weekly schedule. Only new recent matches (NEW_MATCH, $0.18) and material changes to already-seen trials (SOURCE_CORRECTION, $0.05) are charged; unchanged replays are free.
4. Route: point dataset-item webhooks at the site's feasibility or outreach queue.

matching

At least one therapeutic area is required to bound the source query:

• therapeuticAreas: literal phrase match against trial conditions and keywords
• states: two-letter US abbreviations, matched after token normalization (lowercase, non-alphanumerics stripped) against trial location states — the API emits full names ("Florida") or abbreviations ("FL"), and both sides are normalized identically before exact comparison, so valid sites are not discarded by formatting differences
• countries: country names, matched with the same normalization against trial location countries

When therapeutic areas and geography are both supplied, a trial must satisfy both. lookbackDays gates only recent matches: a recruiting trial first posted outside the window is not paid as a recent match, but a material change to an already-seen trial still emits SOURCE_CORRECTION. Reuse the same watchlistId, settings, and stateStoreName for scheduled monitoring.

output

Paid dataset rows are either:

• NEW_MATCH — $0.18 for a recruiting trial first posted within the configured lookback
• SOURCE_CORRECTION — $0.05

Unmatched records, unchanged replays, silent baselines, malformed input, and failed source fetches are not paid events. Every row carries the official study page link (https://clinicaltrials.gov/study/<NCT>), matched evidence, source fingerprint, dates, and disclaimer. A deterministic run report is stored as REPORT with source queries, counts, receipts, state size, and claim scope.

limitations

• This is an internal site-development screening aid: it does not guarantee enrollment, issue recruitment advice, or imply sponsor endorsement. Verify against ClinicalTrials.gov before outreach.
• It does not detect a status transition or the opening of a new physical site; SOURCE_CORRECTION means the source record materially changed since you last saw it.
• Geography matching covers US states and country names only; territories and non-US administrative regions are not matched.
• The source query is bounded to RECRUITING trials per therapeutic area, sorted by first-post date descending; a trial that later changes status is only re-evaluated while it remains recruiting in the fetch window.

60-second quick start

Start with the bounded preview above to see one paid row, then switch to a silent baseline so historical recruiting trials do not create paid events, and schedule the identical input daily or weekly.

{
  "watchlistId": "melanoma-fl-site-preview",
  "therapeuticAreas": ["melanoma"],
  "states": ["FL"],
  "countries": ["United States"],
  "lookbackDays": 30,
  "baselineMode": "silent"
}

Illustrative output shape

Illustrative — not a run receipt. A delivered row contains eventType, NCT identifier, official study link, sponsor, matched therapeutic/geography evidence, sourceRecordUrl, fingerprint, and disclaimer.

Pricing

The bounded preview permits at most one $0.18 recent-fit event. Corrections cost $0.05; unmatched, silent-baseline, and unchanged records create no paid event.

Schedule, webhook, and API recipe

After previewing, keep the watchlist and state store stable, select silent for a clean recurring baseline, schedule daily or weekly, and route dataset-item webhooks to the site's feasibility or outreach queue.

Empty-run troubleshooting

Confirm that the therapeutic phrase exists in current recruiting records, remove geography temporarily to diagnose over-filtering, inspect REPORT, and widen lookbackDays within the schema limit. Do not reset state just to force a paid row.

Buyer bundle

Pair with ClinicalTrials.gov Material Change Monitor: this Actor finds recent recruiting fit; the change monitor follows enrolled records for subsequent material updates.