Avenzoar Mizan

Mizan, an Avenzoar application

What the evidence says. And what it doesn't say yet.

Mizan checks whether a consensus answers your clinical question, summarizes recent literature using GRADE, and brings your peers together when the answer does not exist yet.

Join the pilot program See how it works

Mizan informs decisions with the current state of knowledge. It does not diagnose, does not prescribe, and does not replace clinical judgment.

Three possible answers, always dated

Starting an SGLT2 inhibitor in patients aged 80 and over with type 2 diabetes and an eGFR around 35 mL/min

P T2D, 80 and over, eGFR 30 to 44I SGLT2 inhibitorC Usual careO Renal, CV, tolerability
No up-to-date consensus for this population
Certainty
Very low
Graded review
Available, recent studies selected
Alert
Closest guideline: more than 5 years old

Question sent to the nephrology panel. You will be notified when the recommendation is published.

Three blind spots prescribers face

Every day, treatment decisions are made without knowing whether the reference used is still valid.

5.8 yrs

Median time a clinical practice guideline stays valid before it needs revision.

Shekelle et al., JAMA, 2001
11%

Median share of ACC/AHA cardiology recommendations based on the highest level of evidence.

Tricoci et al., JAMA, 2009

Grey zones with no answer suited to local context

Drug availability, reimbursement and epidemiology vary from one country to the next. International guidelines do not always account for them.

A closed loop, from question to recommendation

The physician asks a question in plain language. Mizan does the rest, and says at each step what it knows and what it does not.

  1. 1

    Question

    Asked freely, then structured as PICO: population, intervention, comparator, outcomes.

  2. 2

    Check

    Search across versioned guidelines, with the date of their last update.

  3. 3

    Graded review

    Summary of recent literature using GRADE, built on verified sources.

  4. 4

    Peer consensus

    Targeted statements submitted to a panel of prescribers using an anonymous Delphi method.

  5. 5

    Publication

    The validated recommendation is published, versioned, and sent to those who asked the question.

Steps 3 to 5 start only when no up-to-date consensus answers the question.

Three answers, always transparent

Established consensus

The recommendation, its level of evidence, its source and the date of its last update. An alert appears if it is old.

Diverging positions

The positions of professional societies, side by side, with their levels of evidence. Mizan shows, it does not decide.

No consensus

A graded review of recent literature, then a consensus process launched with your peers.

An AI that knows how to say “I don't know”

Most assistants answer even when the evidence is weak. Mizan holds back, says so clearly, and turns that gap into a recommendation built by physicians.

GRADE

Every answer shows the certainty of the evidence, from very low to high.

PRISMA

Reviews are reported to recognized standards: sources, selection criteria, excluded studies.

ACCORD

Consensus follows a transparent method: thresholds set in advance, documented rounds.

Verified references

No invented references. Every DOI and PMID is checked against bibliographic databases.

Human review

Reviews and recommendations are checked by Avenzoar's methodologists before publication.

Independence

Conflicts of interest declared by every expert, academic governance.

What Mizan does

  • Tells you whether a consensus exists, and since when
  • Shows the level of evidence for every answer
  • Presents disagreements instead of hiding them
  • Brings your peers together when the answer is missing

What it does not do

  • Diagnose or write a prescription
  • Make a decision for a specific patient
  • Ask for data that identifies a patient
  • Replace clinical judgment

Value for everyone involved in care

Prescribers

A reliable answer within minutes, and recognized participation in consensus work.

Professional societies

Faster consensus, driven by the real questions asked in practice.

Healthcare institutions

Shared references across teams and a history of the questions asked.

Agencies, universities and research

A map of evidence gaps to guide priorities and research topics.

Practitioner

  • Unlimited questions
  • Access to graded reviews
  • Panel participation
Recommended

Institution

  • Team licenses
  • Question dashboard
  • Consensus shared internally

Professional society

  • Consensus management
  • Delphi panel management
  • Publication under your name

Pricing available on request during the pilot phase.

Why Mizan

Mizan, ميزان, means “the scale” in Arabic. That is what the application does: it weighs the evidence and sets diverging positions side by side.

Mizan builds on the methods Avenzoar applies in its systematic reviews and expert consensus work.

Join the pilot program

Six months, one specialty, your real questions. We are looking for professional societies, hospital departments and prescribers ready to evaluate Mizan with us.

RelevanceAgreement between Mizan and experts
QualityReviews compared with manual reviews
SpeedTime from question to recommendation