The people building onconephrology

Eight renal-scoped PubMed queries define the corpus: onco-nephrology, anticancer drug nephrotoxicity, checkpoint-inhibitor kidney injury, MGRS & paraprotein disease, myeloma cast nephropathy, anti-VEGF/TKI renal effects, tumor-lysis & electrolytes, and cellular-therapy renal effects. Every matching article is retrieved in full (no sampling) and its author list parsed. All counts are tallied in code, never estimated, from real records via the PubMed E-utilities API.

Clinical papers only. Because this is a clinical atlas, recognition counts only human/clinical articles — those carrying the PubMed Humans MeSH heading or a clinical publication type (trial, case report, cohort, guideline, review). Pure preclinical (animal-model or cell-culture) studies are excluded — as far as PubMed’s indexing allows; an animal study indexed under Humans, or a review of animal models, can still pass — so a prolific bench scientist doesn’t outrank the clinicians and pathologists doing patient-facing onconephrology. This run kept 15,541 clinical articles and set aside 5,591 preclinical ones.

Ranking blends volume and impact. Contributors are ordered by a 50/50 blend of publication volume (papers) and citation impact, each normalized to the pool. Impact is a position-weighted sum of the Relative Citation Ratio (RCR, from NIH iCite) — a field- and time-normalized citation metric where 1.0 is the median NIH-funded paper. First (lead) and last (senior) authors get full credit; middle authors a fraction, so the score rewards driving high-impact work, not merely appearing on it. Each paper’s RCR is capped (winsorized at 8× the field median) so a single blockbuster — say a phase-3 oncology trial that merely reports renal side effects — can’t crown its authors; impact measures sustained above-median contribution. We use RCR rather than journal impact factor (proprietary, and a cruder proxy), and we do not compute an h-index — a real one needs each author’s full global bibliography and reliable disambiguation, which these records can’t guarantee.

Known limits. Authors are grouped by surname plus given name; PubMed carries no universal author id, so this deterministic key can occasionally split one person across name variants or merge two who share a name. Corresponding authorship is not reliably machine-extractable, so it is omitted rather than invented. Senior-author credit is withheld on 50+ author consortium papers, and RCR accrues over years, so very recent work contributes mainly through the volume half of the blend. This is recognition of contribution within these queries — not a verdict on scientific merit or influence.

Regenerated from PubMed + NIH iCite on 2026-10-02. Educational content — not medical advice.