PubMed KOL Profile Builder — Medical Affairs Edition

Given an indication ("non-small cell lung cancer", "GLP-1 obesity", "CAR-T lymphoma"), return a ranked list of Key Opinion Leaders with affiliation, recent publication volume, h-index estimate, co-author network breadth, and geography — built from PubMed, no commercial KOL database licence required.

For Pharma Medical Affairs, MSL planning, and KOL identification workflows.

---

What this Actor does

Data source: NCBI E-utilities (esearch + efetch on PubMed) — the official NLM API. No key required.

On each run:

1. Searches PubMed for your indication over the last N years (default 5).
2. Fetches up to maxPubmedRecords (default 500) — author lists with affiliations.
3. Scores every author with the composite KOL signal (recency-weighted publication volume + log-scaled citation count + co-author network breadth + journal diversity).
4. Returns the top-N authors as ranked dataset items, each with the azureblue/medical-core envelope so you can join with our PubMed-Abstract, ClinicalTrials, Cochrane, or Conference actors on sourceUrl / dataHash.

---

Use Cases

1. Pre-launch KOL identification for a Medical Affairs team

A pharma Medical Affairs lead preparing a Phase-3 launch in NSCLC needs a defensible shortlist of 50 high-impact KOLs in Europe to seed advisory boards and MSL outreach. One run with indication: "non-small cell lung cancer", country: "DE" returns the top 50 German oncologists ranked by 5-year publication impact in the indication. Replaces ~$25k of bought KOL-list reports + 2 weeks of MSL desk research.

2. MSL territory mapping

A pharma MSL manager wants to assign 6 MSLs to optimal territories for a rare-disease portfolio. Running this Actor per country with topN: 30 produces objective KOL distributions per country in under 5 minutes. Removes the bias of legacy MSL relationships from territory planning — the data tells you where the science actually is.

3. Investigator-led-study (IIS) site sourcing

A clinical-development lead scouts academic sites for an IIS in CAR-T. Filter on indication: "CAR-T lymphoma", look for hIndexInQuery > 10 AND recentPubCount >= 5 — that's the publication signal you want before sending a Letter of Interest. Compresses 4-6 weeks of site-feasibility prep into 1 hour of dataset review.

---

Input

{
  "indication": "non-small cell lung cancer",
  "country": "DE",
  "yearsBack": 5,
  "topN": 50
}

mode: "trial" returns 5 profiles free — sanity-check the ranking before subscribing.

---

Output (sample item)

{
  "indication": "non-small cell lung cancer",
  "rank": 1,
  "name": "Reck Martin",
  "affiliation": "LungenClinic Grosshansdorf, Member of the German Center for Lung Research (DZL), Grosshansdorf, Germany.",
  "country": "DE",
  "kolScore": 4.2731,
  "pubCount": 38,
  "recentPubCount": 24,
  "hIndexInQuery": 14,
  "totalCitations": 0,
  "coAuthorCount": 217,
  "firstPubDate": "2020-02-14",
  "lastPubDate": "2025-09-08",
  "journals": ["Journal of Clinical Oncology", "Lancet Oncology", "New England Journal of Medicine", "Lung Cancer", "Annals of Oncology"],
  "scrapedAt": "2026-05-12T08:30:00.000Z",
  "sourceUrl": "https://pubmed.ncbi.nlm.nih.gov/?term=Reck%20Martin%20AND%20non-small%20cell%20lung%20cancer",
  "sourceDomain": "pubmed.ncbi.nlm.nih.gov",
  "actorVersion": "1.0.0",
  "dataHash": "8a2f...d0"
}

totalCitations: 0 in this sample because openFDA/Europe-PMC citation enrichment is queued for v1.1. Ranking still works — the score weights recency-volume + co-author count + journal breadth.

---

Pricing

Event	Price	When it fires
Monthly subscription	$149	Once per calendar month per user, on first delta-mode run — covers unlimited indication queries within the month
Per KOL profile	$0.05	Per ranked profile pushed to the dataset (snapshot or delta mode beyond the monthly cap)

Trial mode (mode: "trial"): 5 profiles free, no charges, no state writes. Use to evaluate ranking quality on YOUR indication before subscribing.

For Medical Affairs teams: 1 subscribed user × 8 indications × 50 profiles/indication = ~400 profiles/month at flat $149. Compare to commercial KOL lists ($15k–$50k/year for static data).

---

Compliance

• Public-data only — sourced from NCBI E-utilities, the NLM's official public API. No login walls, no PHI, no paywalled content.
• Author affiliations are public — the data ships exactly as PubMed displays it; no email scraping, no contact-detail enrichment.
• GDPR Art. 6(1)(f) / equivalent — buyer is responsible for downstream use of KOL data (legitimate interest as Pharma Medical Affairs research basis). Outreach to identified KOLs must follow the buyer's standard professional-engagement procedures.
• No anti-bot scraping — NCBI E-utilities is the intended programmatic interface.

---

Sister Actors — complementary coverage from azureblue

• pubmed-abstract-scraper — the underlying paper-level abstracts behind every KOL ranking
• clinical-trials-scraper — cross-reference identified KOLs against their PI activity on ClinicalTrials.gov
• medical-conference-scraper — match KOLs to their conference late-breaker activity for current scientific signal
• medical-university-rankings — score the KOL's institution alongside the KOL personally
• ema-drug-approval-watch — when a new EMA approval lands in your indication, this builder gives you the KOLs to brief

---

Roadmap (v1.x)

• v1.1: citation enrichment via Europe PMC citationsBySource — populates totalCitations and improves the kolScore accuracy.
• v1.2: integration with pubmed-author-network-mapper — feed a KOL into the network mapper for co-author graph + collaboration intensity.
• v1.3: conference-activity enrichment via medical-conference-scraper — distinguish "high publication volume" KOLs from "high stage presence" KOLs.

---

Changelog

See CHANGELOG.md in this Actor.