Evidence-Verified Medical Research

Your manuscript, stress-tested like a hostile reviewer would.

Manuscript editing, hostile-review simulation, statistical analysis and systematic reviews for medical researchers.

Every deliverable ships with a verifiable audit trail — no black boxes, no unverified claims. Receipts on every page, not promises.

Get a free scoping call →

0fabricated citations, ever

100%deliverables with audit trails

6+fields with verified toolchains

🛡️ Every deliverable ships with proof

✓ Primary-source URLs on every claim

✓ Confidence tags (HIGH / MEDIUM / LOW)

✓ Exclusion logs with reasons

✓ Reproducible R code + figures

△ Residual assumptions flagged, never hidden

Services

Four productized research services

Fixed scope. Fixed price. Every tier ships with evidence — see the Verification section below.

📚

Literature Brief

PICO-framed, guideline-hierarchy evidence synthesis for a clinical question. Primary sources, cited.

Guideline hierarchy (consensus → RCT)
Primary-source URLs + confidence tags
Risk-stratification table

$149/ brief · 3–5 days

Start a brief

🛡️

Hostile-Review Pass

A 4-phase stress test before real reviewers see your work: methodology, statistics, evidence, writing.

Priority issue list (fatal/major/minor)
Response-to-reviewers table
Second-pass verification

$499/ manuscript · 7 days

Stress-test my paper

📊

Statistical Analysis

ICC(2,1) reliability, meta-analysis statistics, effect sizes — with reproducible R code.

ICC / kappa / descriptives
Reproducible R scripts shipped
Interpretation write-up

$499/ analysis · 5 days

Get my stats done

Premium

🔬

Systematic Review / Meta-Analysis

Search strategy → PRISMA flow → forest/funnel/AMSTAR-2/GRADE figures → manuscript-ready.

Full PRISMA 2020 flow
metafor forest + funnel plots
R code + figures included

$2,500/ SR or MA · 4–6 weeks

Plan my review →

Statistical Analysis sub-tiers — all ship the code: Competing-risks (Fine–Gray) · Mendelian randomization · Diagnostic-test-accuracy meta-analysis · PROMIS/patient-reported scoring · Neuroimaging (fMRI/dMRI) · ctDNA / genomic analyses

Specialities

Six fields with verified toolchains — eight more on request

Every field below has a verified journal-targeting matrix (JCR 2025 impact factors), working data-source access, and the exact statistical/imaging tools installed and tested on our machine. We don't claim breadth we can't back.

🩹

Dermatology

Paediatric AD, image-scoring pipeline (EASI/SCORAD/vIGA-AD), dermoscopy datasets.

JAMA Derm (IF 10.5) → J Invest Dermatol → JAAD
ISIC Archive · HOME core outcome set
dermscore.py + lesion-classification stack

🦴

Orthopedics

Implant survivorship, registry cohorts, PROMIS/patient-reported outcome scoring.

JBJS (IF 5.3) → KSSTA → Int Orthop
FracAtlas X-ray corpus · NJR/AOANJRR patterns
PROMIS CAT + custom KOOS/HOOS/WOMAC scoring

🎗️

Oncology

Registry analyses (SEER/TCGA), competing risks, biomarker studies, SR/MAs.

JCO (IF 44.7) → JAMA Onc → JTO → Cancer Letters
cBioPortal + GDC/TCGA live access
Fine–Gray competing-risks (underserved niche)

❤️

Cardiology

Mendelian randomization, ECG/AI-ECG, registry studies, device trials.

Circulation (IF 41.3) → EHJ → JACC → JAHA
UK Biobank · NHANES CV · MIMIC-IV
TwoSampleMR + ECG processing stack

🧠

Neurology

Neuroimaging biomarkers, EEG, longitudinal trajectories, stroke outcomes.

Lancet Neurology (IF 45.5) → Stroke → Neurology
ADNI · OpenNeuro · UK Biobank imaging
nilearn/nibabel/dipy + MNE-Python EEG

🩻

Radiology

Diagnostic-test-accuracy SR/MAs, radiomics, imaging AI reporting.

Radiology (IF 17.6) → Eur Radiol → AJR
TCIA · BraTS · MIMIC-CXR
pyradiomics + STARD/CLAIM/TRIPOD+AI compliance

Also covering on request: Pediatrics · Psychiatry · Endocrinology · Infectious Disease · Nephrology/Pulmonology · ENT & Ophthalmology · Urology & Gyn/Ob · Anesthesiology/Pathology/EM/Rheum/Immuno.

Proof of Work

Actual deliverable artifacts — not mockups

The verification section above shows the concept. Here is the literal output: reproducible code and an exclusion-log excerpt from real projects (anonymized).

📊 Real R code shipped with a stats deliverable

icc(m, model="twoway", type="agreement", unit="single")
# ICC 0.970, 95% CI 0.811–0.996

fit <- coxph(Surv(time,status) ~ group, data=d)
# HR 0.62 (0.45–0.86), p = 0.004

sessionInfo() # R 4.5.2, survival 3.8.6

🩻 Exclusion-log excerpt (cohort screen)

✓ Duplicate-ID check: 1 hit → resolved
✓ Typo-variant scan (RIGIDY/RIGDI): 0 misses
✓ Contralateral-side scan: 0 false exclusions
△ Loose-family overmatch: 1 row → manual review
✓ Cohort locked: 30 feet / 28 patients

📝 Hostile-review response table

#1 "Medical heterogeneity" → stratified analysis added (§3.2)
#2 "Attrition undisclosed" → added to abstract
#3 "Complication gap" → new table (§4.4)
△ [AUTHORS] flag: team task, stated honestly

Why $2,500 for an SR/MA when other services charge $800?

You get the code, not just the numbers. Every forest/funnel plot ships with the reproducible R script — no black box, re-runnable by your team.
The audit trail is the deliverable. Exclusion logs, confidence tags, primary-source URLs — the receipts reviewers can't argue with.
Clinician-reviewed, 2026-standards current. CONSORT 2025 / SPIRIT 2025 / TRIPOD+AI compliance built in, not bolted on.

Current availability: 1 SR/MA slot open this month · stats slots open · briefs rolling (standard 3–5 day turnaround).

The Veritas Difference

Verification is the product

The #1 fear in AI-assisted research is fabrication. We answer it with receipts — the same audit trails our own studies carry.

🩻 Cohort screening — zero-miss audit trail

Every eligibility screen ships an exclusion log with reasons, plus automated checks for the classic classifier bugs.

📝 Hostile review — response table

Reviewer comments answered one-by-one, with the exact revision location. Nothing waved away.

#1 "Medical heterogeneity unaddressed" → acknowledged; stratified analysis added (§3.2)
#2 "Attrition not disclosed" → added to abstract (§Abstract)
#3 "Complication reporting gap" → new table (§4.4)
△ [AUTHORS] flag: team task, not AI-fixable — stated honestly

📊 Statistics — reproducible code

ICC(2,1) two-way random, absolute agreement, single measures — with the exact R script that produced it.

icc(m, model="twoway", type="agreement",
unit="single") # ICC 0.970, 95% CI 0.811–0.996
✓ Mean-of-observers master table
✓ Calibration sensitivity shown (0.252 vs 0.300 mm/px)

Proof of Work

Case studies — real deliverables

Anonymized examples of the exact artifacts you'll receive.

Case report · Orbit

Dacryolith manuscript — hostile-review pass

Full colour-coded revision (yellow = add, red-strike = delete, blue = author notes) plus a point-by-point response table that survived resubmission.

Deliverables: revised DOCX + PDF · response table · highlight diff

Cohort study · Hallux Rigidus DOO

Hallux cohort — zero-error audit trail

2,988-row registry → 30 feet / 28 patients locked with documented exclusions and 5 classifier bugs caught in-loop — the audit trail is the deliverable.

Deliverables: cohort CSV · PRISMA-style cascade · audit report

Clinical question · Orthopaedics

DVT prophylaxis — literature brief

Guideline hierarchy (ICM-VTE → national → trials) synthesised into a risk-stratified protocol table with Asian-population nuance flagged.

Deliverables: structured brief · risk table · primary-source citations

FAQ

Questions researchers ask

Is this AI-generated work? How do I know it's trustworthy?

We use AI as the engine, but every deliverable is verified against primary sources and ships with an audit trail you can inspect: citation URLs, confidence tags, exclusion logs, and the exact code that produced your statistics. If we cannot verify a claim, it is flagged — never silently asserted.

Do you write the manuscript or edit mine?

Both. Tier 1–2 are editing/review of your existing manuscript. The SR/MA tier includes full drafting from your research question. You always retain authorship and full control.

What does the audit trail look like?

A structured appendix: every literature claim carries its primary-source URL and a confidence tag (HIGH/MEDIUM/LOW); every cohort screen carries an exclusion log with reasons; every statistic carries the reproducible R code. See the Verification section above for real examples.

How fast is the first response?

Within 24 hours, with a scoping summary and a fixed quote. No hourly billing surprises — the price you see is the price you pay.

Do you guarantee acceptance?

No — and any service that does is lying. We guarantee a submission-ready manuscript that survives our own two-pass hostile review, and we'll iterate on reviewer feedback. Acceptance is up to the journal and the science.

Can you work with genomic / imaging data like TCGA, cBioPortal, or MRI?

Yes — that's a core capability, not an add-on. We have live access to cBioPortal, GDC/TCGA, NHANES, UK Biobank, ADNI, and more, with the analysis toolchain installed and tested. Tell us the dataset and we'll tell you exactly what analyses are feasible.

Do you do Mendelian randomization or competing-risks analysis?

Both. TwoSampleMR (GWAS instruments), competing-risks (Fine–Gray), network meta-analysis, diagnostic-test-accuracy meta-analysis, and PROMIS scoring are all in our verified statistical stack — with the R code shipped so your team can reproduce every result.

What is CONSORT 2025, and why does it matter for my submission?

CONSORT 2025 is the updated randomized-trial reporting standard (published 2025, replacing 2010) — journals are moving to require it during 2026. We build submissions against the current standards (CONSORT 2025, SPIRIT 2025, TRIPOD+AI, PRISMA 2020) so your paper doesn't get tripped up by a checklist you didn't know existed.

How do you pick which journal to target?

From a verified JCR 2025 impact-factor matrix per speciality — including word limits, abstract format, and citation style for the top journals in your field. We match your study type and data to the realistic target range, from flagship (e.g. JCO, Circulation) down to solid specialty journals, and format the submission accordingly.

What exactly do I receive when you ship?

A documented package, not just a file: the manuscript/report, the reproducible R/Python code that produced every statistic and figure, the exclusion logs (for cohort/trial work), the primary-source citation URLs with confidence tags, and a sessionInfo-style environment record. You can verify everything yourself — that's the point.