Evidence work for drug safety teams

A drug safety answer you can check.

Your data and every public source in one evidence base. Operators draft the work; your experts approve before anything leaves.

Talk to salesStart free

Medical information

Illustrative · real run on public data

A paediatrician asks

Could nightmares and agitation in a 14-year-old come from montelukast?

  1. Yes. Both are known, labelled effects, strongest in children.

    • 52.5%of nightmare reports are in children and adolescents
    • 46.3%of nightmare cases eased when the medicine was stopped
    • 34.9%of agitation cases eased when it was stopped

  2. “The types of events reported were highly variable, and included, but were not limited to, agitation, aggression, depression, sleep disturbances, suicidal thoughts and behavior (including suicide).”

    Found word for wordUS label · boxed warning, quoted word for word

  3. US label, boxed warningFDA reports, 2004–2026PubMed, 3 studies

  4. Data as of 2026 Q2, rerun as of any earlier quarter

  5. Waiting for the medical information reviewer

  6. Every run logged: input, output, tools, user

The reply sets out published evidence. It does not advise on treatment.

Enterprises running operators on Alomana

Since August 2025, every marketing-authorisation holder in the EU must monitor EudraVigilance for signals. EU Implementing Regulation 2025/1466

One evidence base under every answer

130M+

  • 40M+safety reports in WHO VigiBase, from more than 180 countries
  • 31Msafety reports in EudraVigilance
  • 20.6MFDA reports, deduplicated and counted as of any date
  • 40M+biomedical citations in PubMed
  • 607,160registered studies, every posted result read arm by arm
  • 262,883US label documents, read section by section
  • 163,791EU authorised product records, from 4,037 companies
  • 30EU and EEA national label registers, from AIFA to AEMPS

One platform, a suite for each team

  • Patient safety

    Case triage and line listings

  • Regulatory & labels

    EU decisions to label changes

  • Medical information

    Replies to doctors

  • Clinical evidence

    Trial results, arm by arm

A week of patient safety, worked

Patient safety · weekly review

Illustrative · real run on public FDA data, week 41

  1. 1Intake
  2. 2Follow-ups
  3. 3Coding
  4. 4Label check
  5. 5Priority
  6. 6Signals
  7. 7Evidence pack

This week’s cases

20

10 serious and not on the label, first in the queue

11 follow-ups joined to their first report

New important medical events

  • Pembrolizumab21
  • Apixaban9
  • Semaglutide4

34 events from the EMA important medical event list, across 3 medicines

For your safety physician

Select a card to see its evidence.

Weekly evidence pack for the safety physician

Dr. M. Ferri, safety physician

A flag is a reason to look, not a finding. Reports do not show cause.

Every EU decision, mapped to the labels it changes

Regulatory & labels

Illustrative · from public EU decisions

  1. 1Decision read
  2. 2Products found
  3. 3Labels compared
  4. 4Gap flagged
  5. 5Deadline tracked
  • 57EU safety decisions in ten months
  • 5,672product label changes they required
  • 850companies affected
  • 32countries, every one tracked

Public data · real run

Oxacillin and DRESS, a severe drug reaction

10 years earlier

  1. 2016signal criteria met, 3 reports
  2. 2026EMA recommends the label change

Replayed quarter by quarter on public FDA reports. A flag is a reason to look, not a finding.

Trials read arm by arm, beside the label

Serious adverse events, drug against placebo

Illustrative · real run on public trial results

51 posted placebo-controlled trials of semaglutide, each linked in the answer

  1. 1Trials found
  2. 2Results read
  3. 3Drug against placebo
  4. 4Checked against the label
  5. 5Set beside the reports
  • Cholelithiasis22SELECT0.50%0.35%Listed
  • Acute cholecystitis12SOUL0.64%0.39%Listed
  • Serious vomiting14FLOW0.62%0.28%Listed
  • Nephrolithiasis13SOUL0.41%0.27%Not listed

Checked against what regulators later did

1,892 of 2,420

side effects regulators later added to a label, found by the label reader, every quote verified

  • About nine in ten stated events read correctly, against public gold sets
  • Events coded at about 0.86 accuracy, with no invented terms

Your data, your standards

What the operators read
Safety databaseCore data sheets and labelsSOPs and templatesPast workWorks to your SOPWrites in your templateYour workspaceNever used to trainFull audit trail

Eight weeks to a measured result

  1. Weeks 1–2

    Connect

    Workspace set up for your teams

  2. Weeks 3–8

    Prove

    One team, its real weekly work

  3. Month 3 onward

    Scale

    Next suite on the same evidence base

  • Medical informationHours per reply
  • PharmacovigilanceHours per report
  • Regulatory affairsDays to impact note
  • Medical affairs and MSLsQuestions answered

A worked example, with its assumptions

  • A medical information team writes 614 standard responses a year and revises 1,352 (phactMI survey average).
  • At 8.7 hours to write one and 3.8 hours to revise one, that is about 10,500 hours a year.
  • At $60 an hour, drafting half of that work is worth about $315K a year.

Assumptions, not a result: your volumes, hours and rates replace these in week 1.

Nothing leaves without a yes.

  • Every reply, note and list waits for an expert's approval
  • A declined action is not retried
  • Every run logged: input, output, tools, user
Trusted for
ISO 27001GDPRNever trains on your dataSingle-tenant for enterprise

A custom operator typically goes live in up to 4 weeks.

Questions

What do the operators do?

They do the evidence work of patient safety, regulatory and medical information teams: they screen cases and signals against your labels, map each EU safety decision to the labels you must change, draft replies to doctors and read trial results arm by arm. Every number and quote links to its source, and your expert signs off.

Does it replace our safety database?

No. Your safety database stays the system of record. The operators read its case exports, such as E2B(R3) files from Argus, ArisGlobal or Veeva, and return worklists, evidence packs and drafts for your team.

Which data does it use?

Your own data, plus public sources: FDA adverse event reports, US labels, EU decisions and product registers, national labels, ClinicalTrials.gov results and PubMed. Sources with their own access terms, such as WHO VigiBase case data, are connected for your deployment under those terms.

How do we know a number or a quote is right?

Counts and measures are computed in code with standard published methods, and every answer states the measure and its limits. A label quote is shown only when it is found word for word. The label reader was checked against 2,420 side effects regulators later added to labels.

Does it decide whether something is a signal?

No. A flag is a reason to look, not a finding, and reports do not show cause. The operators rank what to look at first and draft the evidence; your safety physician decides.

Is our data private?

Alomana is ISO 27001 certified and GDPR-compliant, and your data is never used to train AI models. Enterprise deployments run single-tenant in an EU region, and every run is logged with input, output, tools and user.

Start with one team. Measure it in eight weeks.

Talk to salesStart free

Alomana is a secure enterprise AI platform where autonomous agents, called operators, run business processes end to end. Every run is fully auditable.