Clinical · Live
Clinical Study Reports (ICH E3)
Every section drafted and numbered to the guideline from the study's source data, with each statement cited to the passage it was read from.
The platform
Decision work, with the evidence attached.
AI for life sciences.
Altio turns the regulated documents and data you rely on, internal and external, into a longitudinal record that becomes one institutional memory. Built once at the core, that record carries forward as decisions move from safety to clinical, regulatory, quality, and beyond. Every answer traces back to the exact passage that required it, so review is checking the work rather than redoing it.
Not sure which use case to start with? Our advisory services help you choose it and build it.
Deliverables
Ask in plain language. Read the receipts.
Four workflows, one way of working. Pick one to see the question, the answer, and the passage behind every statement.
Does this repeat something we already assessed?
Altio opens the assessment with what the record already holds: the prior signal, the position taken, and the reasoning behind it. The question becomes whether anything has changed, not what happened last time.
- Prior assessments surface without anyone retrieving them
- Case narratives read as evidence, not just coded fields
- The earlier rationale carried forward with its source
Is this signal new, or the same one we assessed eighteen months ago under a different name?
Internal
It is the same signal. The pattern flagged this week matches the assessment closed in February 2025, recorded then as “cytokine storm” in the case narratives, and the committee's rationale is carried forward on the product record.
Read from
-
1
Signal Assessment SA-2025-004Internal Narrative review confirmed that events reported as “cytokine storm” met the case definition applied to cytokine release syndrome in this assessment.
-
2
Safety Committee Minutes, Feb 2025Internal The observed pattern is consistent with the established CRS profile; signal closed with routine monitoring.
Select a citation to open the passage it was read from.
One section per agency question.
Each question in the request gets its own section, answered from the sources in scope and dual-signed before it leaves. Nothing is asserted that a reviewer cannot open and check for themselves.
- One section per question, numbered to the request
- Dual sign-off enforced before release
- Every statement traceable to a source passage
What do the labels say about cytokine release syndrome?
External
The US Prescribing Information states that EPKINLY can cause cytokine release syndrome, including serious or life-threatening reactions. The EU Summary of Product Characteristics for the same molecule lists the most common adverse reactions, neutropenia among them.
Read from
-
1
EPKINLY US Prescribing InformationPublic EPKINLY can cause cytokine release syndrome (CRS), including serious or life-threatening reactions.
-
2
TEPKINLY EU Summary of Product CharacteristicsPublic the most common (≥ 20%) adverse reactions were neutropenia, rash, upper respiratory…
Select a citation to open the passage it was read from.
Assembled from the record, reproducible after the meeting.
The pack is built from the same sources the assessments used, so a figure on a slide can be traced back months later without reconstructing anyone's spreadsheet.
- Several lookups answered in a single pass
- Each figure anchored to the row it came from
- Reproducible long after the meeting closes
Do five separate lookups, one search step for each of these terms, then report what each found.
External
Neutropenia: decreased neutrophils occurred in 62% (Grade 4, 13%) of patients in the monotherapy arm, and febrile neutropenia in 6% (Grade 4, 2.1%). Pneumonia: 24% all grades and 16% Grade 3 or 4 in the combination arm.
Read from
-
1
EPKINLY US Prescribing InformationPublic 62% (Grade 4, 13%), decreased hemoglobin in 7%, and decreased platelets in 10% (Grade 4, 4.1%) of patients.
-
2
EPKINLY US Prescribing InformationPublic Febrile neutropenia occurred in 6% (Grade 4, 2.1%).
-
3
EPKINLY US Prescribing InformationPublic Pneumonia 24 · 16 · 8 · 4.6
Select a citation to open the passage it was read from.
Numbered to the guideline, section by section.
Sections draft against ICH E3 numbering, with tables and figures built from the data rather than described in prose, so the reviewer's job is judgment rather than formatting.
- All sections drafted to ICH E3 numbering
- Tables and figures built from the data, not from prose
- Statistics run as approved, versioned code
For the adverse events section, what does the FL monotherapy cohort report for pyrexia and diarrhea?
External
In the FL monotherapy table (N=127), pyrexia occurred in 26% all grades and 2% Grade 3 or 4. In the monotherapy arm (N=157), diarrhea occurred in 20% all grades.
Read from
-
1
EPKINLY US Prescribing InformationPublic Pyrexia 26 · 2
-
2
EPKINLY US Prescribing InformationPublic Diarrhea 20 · 0
Select a citation to open the passage it was read from.
Where it fits
An intelligence layer above your validated systems.
Altio sits above the systems you already run. It ingests documents and data across sources, builds the longitudinal record, and produces audit-ready outputs, replacing uncontrolled tooling with traceable, reproducible work.
It reads
- ICSRs and case narratives
- PSURs, DSURs, and literature
- Labels and health-authority queries
- Your own assessments and minutes
It produces
- Signal assessments
- Health-authority responses
- Safety committee packs
- Clinical study report sections
Each one drawn from the same institutional memory, extending to CTD and regulatory work, then medical affairs and evidence, as modules come online, with PBRER as a natural progression built with a design partner.
Connects to the stores you already run. Intake draws on public data such as FAERS alongside the documents you bring from your safety database, literature, and document management system, and signed deliverables flow downstream to Veeva Vault by API.
1
IngestICSRs, PSURs, literature, and HA queries, from the sources you bring and public data such as FAERS
2
AnalyzeAssess, draft, and cite every statement to its passage
3
ReviewNamed reviewer, human-in-the-loop sign-off
4
OutputAudit-ready deliverables, exported back to your document store on an immutable trail
How the work is controlled
Nothing leaves without a named reviewer.
The same discipline applies whichever deliverable you are working on.
- DraftedIn your own template, every statement cited to its source.
- ReviewedPassed to a named reviewer, with comments recorded and resolved.
- SignedPart 11 e-signature, each version kept on an immutable trail.
- It proves its quotesEvery citation is checked back against the source document before the answer is delivered.
- It refuses rather than guessNo safety figure from model memory, and no statistic from data it does not hold.
- It says what is missingIf a source or data point the answer needs is not in scope, the gap is stated in the deliverable so the reviewer knows what to obtain.
- Numbers come from codeStatistics run as versioned, deterministic code once a person approves the plan.
Beyond safety
Patterns across the R&D lifecycle.
Safety & regulatory · With a design partner
PBRER and CTD Authoring
Periodic benefit-risk reports (PBRER/PSUR) are not built today; they are the natural progression of the signal record, built with a design partner. CTD modules follow the same pattern, drafted from the same record.
Regulatory affairs · Suggested use case
Regulatory Intelligence & Label-Change Monitoring
Tracking of HA guidance, label changes, and competitive intelligence across global markets, built with a design partner.