Skip to content

Registry Policies

How the IDAG Registry handles registrations, from submission to permanent record.

1. Scope and Purpose

The IDAG Registry is a public registry of systematic review protocols. It records what a review team plans to do before the review is carried out, so that plans can be compared with the eventual review and unnecessary duplication can be avoided.

More background is on the About IDAG page.

2. Eligibility for Registration

IDAG is intended for registering protocols of structured evidence-synthesis projects within the review categories the registry supports. It currently supports reviews of human-health evidence and systematic reviews of preclinical animal studies; see Review Types.

Basic requirements

A submission should generally:

  • have a clearly defined review objective or question;
  • use a structured, reproducible evidence-synthesis method appropriate to its review type;
  • state its eligibility criteria;
  • describe the planned information sources and search methods;
  • describe the planned approach to study selection;
  • describe the planned approach to evidence synthesis;
  • give enough information for the registry record;
  • identify the review team;
  • disclose relevant funding and conflicts of interest as the registry requires.

Not normally eligible

The following are not normally eligible for IDAG registration:

  • narrative reviews without a systematic method for identifying evidence;
  • opinion pieces, editorials and commentaries;
  • individual primary research studies, including individual clinical trials;
  • case reports or case series submitted as primary studies;
  • protocols unrelated to evidence synthesis;
  • clearly duplicate submissions of the same review by the same review team without justification.

Completed reviews

IDAG is primarily intended for registration before a systematic review is completed. Starting work before registration does not by itself make a submission ineligible. However, a review that was already fully completed before submission should not normally be eligible for prospective protocol registration.

Registrations elsewhere

An existing registration in another registry does not automatically prevent IDAG registration. Authors should disclose the registry name, the registration number and its relationship to the IDAG submission. The registration form already asks for this under previous registration.

Multiple reports and updated reviews

Several manuscripts or publications arising from the same systematic review should normally stay linked to the same IDAG registration rather than being registered as separate reviews. Genuinely distinct review questions may need separate registrations.

An update of an existing review should not automatically be treated as an unrelated new review. If the review is already registered in IDAG, authors should use the existing living-review update or amendment process where applicable.

Scope decisions and borderline cases

IDAG may return or decline submissions that clearly fall outside the registry's scope. Such decisions concern suitability for the registry and completeness against its requirements. They are not scientific endorsement or rejection of the work.

For unusual designs that do not fit an existing category neatly, authors should choose the closest supported review type and give enough methodological information.

3. Registration Timing

Prospective registration is strongly encouraged. Authors should register as early as reasonably possible after the review protocol and methods have been defined, and preferably before formal study screening or data extraction has substantially progressed.

Starting a review before registration does not automatically make a submission unacceptable. Instead, the stage reached should be declared transparently and may be considered during administrative screening. This approach avoids rewarding inaccurate reporting.

Review stage at submission

Rather than a single "prospective" or "retrospective" label, IDAG asks authors to report the review stage at the time of initial submission. The stages are:

  1. Protocol development
  2. Search strategy under development
  3. Preliminary or pilot searches only
  4. Formal literature searching started
  5. Formal literature searching completed
  6. Title/abstract screening started
  7. Title/abstract screening completed
  8. Full-text screening started
  9. Full-text screening completed
  10. Data extraction started
  11. Data extraction completed
  12. Risk-of-bias / critical appraisal started
  13. Evidence synthesis / meta-analysis started
  14. Evidence synthesis substantially completed
  15. Review completed

Future timing classification

IDAG may in future classify registration timing from the declared stage, for example as early prospective registration, registration after searching began, registration after screening began, late registration, or review already completed. These categories are not in use yet; their exact thresholds need institutional approval.

4. Submission and Administrative Screening

Authors prepare and save a draft, then submit it once all required questions are complete. The submitted version is locked and cannot be edited.

A Registry Administrator then screens the submission. They may send it on for protocol review, return it to the authors for corrections, or decline it. How to prepare a submission is explained in Registration Guidance.

A Registry Administrator who is the account owner or a listed author of a protocol cannot screen it, invite its Protocol Reviewers, approve or publish it, or decide its amendments or lifecycle requests. Another Registry Administrator must act.

5. Protocol Review

Following administrative screening, a submission may undergo protocol review where applicable. This is protocol review for registry purposes, not journal, manuscript or publication peer review.

Purpose

Protocol Reviewers assess the scientific and methodological quality, completeness and clarity of submitted systematic review and meta-analysis protocols before registration and publication. Protocol review helps the Registry Administrator identify information needing clarification or revision.

It does not certify that the systematic review is scientifically correct, guarantee methodological quality or publication, or constitute endorsement by IDAG, ICMR, DHR or AIIMS Guwahati.

Selection and assignment

Protocol Reviewer registration requires Administrator or Super Administrator approval. Each submitted protocol should normally be assessed by two approved Protocol Reviewers.

Protocol Reviewers should be selected for relevant expertise where reasonably possible, such as systematic-review methodology, evidence synthesis, statistics or meta-analysis, subject expertise, or expertise in the specific review type.

A Registry Administrator invites Protocol Reviewers, assigns the protocol version, controls the review round and receives the assessments. Authors do not select Protocol Reviewers for their own submission. Each assessment applies to the specific protocol version assigned for review, not another draft or version.

Recommendations and decisions

A Protocol Reviewer recommends either Approve or Revise with Comments. Recommendations are advisory to the Registry Administrator. A Protocol Reviewer cannot register or publish a protocol, and a recommendation does not by itself register, publish, reject or withdraw one. The Registry Administrator remains responsible for registry workflow decisions.

Where Protocol Reviewers disagree, there is no voting or automatic majority rule. The Registry Administrator considers the assessments and decides the next step, for example requesting clarification or revision, seeking an additional assessment, or proceeding on the information available where appropriate.

Feedback to authors

Assessments go to the Registry Administrator, not directly to authors. The Administrator decides what feedback to send. Feedback may draw on Protocol Reviewer assessments but is an administrative communication from the registry; authors are not entitled to the full internal assessment.

Authors respond to requested changes through the revision process, where they can answer each revision comment and explain any disagreement to the Registry Administrator. There is no direct contact between authors and Protocol Reviewers.

Protocol Reviewer conduct

Protocol Reviewers should:

  • assess the assigned protocol objectively;
  • focus comments on the protocol and registry requirements;
  • give clear, constructive and relevant comments;
  • avoid inappropriate or discriminatory language;
  • declare conflicts and maintain confidentiality;
  • complete the assessment within the agreed timeframe where possible.

IDAG may take responsiveness, timeliness, the relevance of assessments, and adherence to confidentiality and conflict requirements into account when making future assignments. A Registry Administrator can remove the Protocol Reviewer role from an account. Protocol Reviewer performance is not made public.

6. Registration and Publication

An IDAG registration number is assigned only after the protocol has completed the required registry workflow, received final approval and been published by a Registry Administrator. Numbers use the format IDAG-YYYY-NNNNN and are issued in sequence for each year.

Publication creates Version 1 of the public record (the Original Registration) and makes it searchable. Registration concerns the protocol only. IDAG does not peer-review or publish the eventual systematic review.

7. Amendments and Version History

Substantive changes to a registered protocol are made through an amendment request. The amendment is checked by a Registry Administrator.

An approved amendment creates a new public version. Earlier public versions are preserved and stay visible in the record's version history.

8. Living Reviews

Reviews registered as living reviews can report updates through the living-review update process.

  • Evidence-only updates, such as a new search or newly included studies with unchanged methods, can be approved as a living-review update and create a new labelled version.
  • Methodological or protocol changes are not handled as living-review updates. They must go through the amendment process.

9. Review Completion

Authors can report the progress of a registered review, including when it has been completed. Progress changes are shown in the record's public status history. Reporting progress does not create a new protocol version.

10. Resulting Publications

Authors can submit links to publications resulting from the review. A publication link becomes public only after it has been verified by a Registry Administrator.

IDAG does not assign a DOI to the resulting journal article. Any DOI belongs to the article and is issued by its publisher.

11. Discontinuation

If a registered review is stopped before completion, the authors can request discontinuation, which a Registry Administrator reviews.

A discontinued registration remains publicly accessible and keeps its IDAG number, registration date and version history. Its public status clearly shows DISCONTINUED.

12. Withdrawal

Authors can request withdrawal of a registration, which a Registry Administrator reviews. A withdrawn registration is not deleted. It remains publicly accessible, searchable, citable and historically preserved, and its public status clearly shows WITHDRAWN.

Only the approved public withdrawal statement is shown. Private reasons stay internal.

13. Duplicate Registrations

IDAG may point out registrations that look similar to a new submission. Similarity is only a warning and does not by itself mean a submission is a duplicate. Authors and Registry Administrators must judge the relationship. Where appropriate, related registrations may be linked.

You can check for existing registrations using Search. How suspected duplicates are assessed is described under Complaints / Queries.

14. Administrative Corrections

Clerical errors in the public record, including typing errors in an existing identifier (such as an ORCID, DOI, PMID, external registration ID or web address), may be corrected administratively when the correction refers to the same person, protocol, publication or registration and is not a substantive protocol change. Substantive changes are managed through the amendment and version process.

Evidence Domain: changing a registration between Human and Preclinical Animal after registration is not permitted; a new IDAG registration is required.

Discontinued or withdrawn registrations: no Protocol Amendment may be published while a registration is Discontinued or Withdrawn.

Reports of inaccurate public information are covered under Complaints / Queries.

15. Protocol Reviewer Confidentiality

Protocol materials made available to a Protocol Reviewer for registry review are confidential unless already publicly available. Protocol Reviewers should not:

  • publicly share unpublished protocol content;
  • pass the protocol to unauthorised persons;
  • disclose confidential review discussions;
  • use confidential information for inappropriate personal or professional advantage.

As an interim confidentiality safeguard, Protocol Reviewers should not upload confidential or unpublished protocol information to external AI systems or other third-party services unless IDAG policy explicitly permits it.

Protocol Reviewer identities are not publicly displayed, and authors do not automatically receive them. Protocol Reviewer identities, assignments, conflict declarations and assessments are internal registry information, except where disclosure is required by law or by formally approved institutional policy.

16. Conflicts of Interest

Review authors: every submission must state whether any review author has financial or non-financial interests relevant to the review. If so, each declared interest is recorded with the author concerned, the nature of the interest, any related organisation and a description. Authors must also write an overall conflict-of-interest statement in their own words. Declarations intended for public display appear on the public record.

Protocol Reviewers

Protocol Reviewers should disclose any interest or relationship that could reasonably affect, or appear to affect, their impartiality in reviewing the assigned protocol. Examples, where relevant, include:

  • being an author or collaborator on the submitted review;
  • recent or ongoing close collaboration with members of the review team;
  • direct involvement in the underlying project;
  • significant academic competition or dispute;
  • relevant financial interests or commercial relationships;
  • supervisory or dependent professional relationships;
  • personal relationships that could reasonably affect impartiality;
  • other circumstances the Protocol Reviewer believes the Registry Administrator should know.

These are examples, not automatic disqualifications. Protocol Reviewers are asked about potential conflicts when invited and can report a conflict discovered later during an assignment. The Registry Administrator evaluates the nature and relevance of each disclosure: some conflicts may require replacement, others may be manageable.

Recusal: a Protocol Reviewer who believes they cannot give an impartial assessment should not continue and should inform the Registry Administrator. The Administrator may cancel the invitation or assignment, appoint another Protocol Reviewer, or take other appropriate registry action.

Conflicts found after an assessment: the Registry Administrator should consider the nature of the conflict, its possible effect on the assessment, whether another assessment is needed, and whether the existing assessment should be set aside for decision-making. Historical assessments are not deleted, and the record history is preserved.

Individual conflict disclosures are internal and are not public.

17. Public and Private Information

Public records may include appropriate protocol and authorship information. The following are never shown publicly:

  • login or account email addresses
  • internal user identifiers
  • Protocol Reviewer identities
  • confidential Protocol Reviewer assessments
  • administrator notes
  • private withdrawal or discontinuation reasons
  • audit logs

See the Privacy page for details.

18. Record Permanence

Published IDAG registrations form part of the permanent registry record. Withdrawal or discontinuation does not delete a registration, and earlier public versions remain part of its history. Complaints and concerns do not cause a published registration to be silently removed; see Complaints / Queries.

19. Registry Disclaimer

Registration does not mean:

  • journal publication
  • endorsement by IDAG, ICMR, DHR or AIIMS Guwahati
  • a guarantee of methodological quality
  • a guarantee of eventual publication

Authors remain responsible for the information they submit. See also the Terms of Use.

20. Complaints / Queries

IDAG registration is an administrative registry process. It is not journal peer review, institutional ethics review, research-misconduct adjudication, authorship arbitration or legal adjudication. IDAG may examine matters needed to protect the accuracy, integrity and operation of the registry.

Questions and complaints

Routine questions, for example about the registration process, accounts, choosing a review type, submission status, amendments, resulting publications or public records, should be sent through Contact IDAG. Common questions are answered in Help & Guidance. The official registry contact email is still to be confirmed.

Anyone, including authors, Protocol Reviewers, registered users, readers of the public registry, institutions and journals, may raise a concern about the registry process, registered information, duplicate registration, authorship information, confidentiality or privacy, inappropriate public content, possible misuse of the registry, or other matters directly relevant to IDAG. An IDAG account is not needed to raise a legitimate concern about a public record. No fixed response time has been set.

Request for Reconsideration

Authors may make a Request for Reconsideration of an administrative registry decision if they believe relevant information was overlooked, the submission was misunderstood, a factual error affected the decision, or they can provide material clarification.

A request does not automatically reverse, suspend or invalidate the original decision. The Registry Administrator may uphold the decision, seek clarification, seek additional review where appropriate, or revise the decision where justified. Where practicable, a reconsideration should be reviewed by an Administrator who was not solely responsible for the original decision.

Disagreement with Protocol Reviewer feedback should normally be addressed first through the revision process, where authors can explain why a requested change was not made. Disagreeing with Protocol Reviewer comments alone is not a complaint, and there is no direct contact between authors and Protocol Reviewers.

Duplicate or overlapping registrations

Suspected duplicates are assessed on the substance of the reviews, not on title similarity alone. Relevant points may include the review question, population, the intervention, exposure, index test or construct as applicable, outcomes, eligibility criteria, methods, review team, any external registration, and whether one record is an update or expansion of another.

A published IDAG registration is not silently deleted if it is found to duplicate another. Depending on the circumstances, records may be linked, their relationship identified publicly, or one withdrawn where justified. Registering the same or a related review in another registry is not in itself wrongdoing; authors should disclose the relationship, and IDAG may link or describe related registrations.

Inaccurate public information

If public registry information may be factually inaccurate, the matter may be reviewed. Scientific or methodological changes go through the amendment and version process. Clerical or factual corrections, such as spelling, author-name or affiliation errors, broken contact details, incorrect links or obvious transcription errors, are handled through Administrative Correction.

Authorship changes after registration

After a registration is published, adding, removing or reordering authors, or changing the corresponding author, is a substantive change. It can only be made through a Protocol Amendment, which the Registry Administrator must approve. Administrative Correction may correct an existing author's name spelling, affiliation or ORCID, but cannot add, remove, reorder or substitute authors or change who the corresponding author is.

When an amendment changes authorship, IDAG identifies the change automatically and the submitting author must complete an Authorship Change Declaration: the reason for the change, the contribution or circumstances of any added or removed author, and whether agreement was obtained from all affected authors, was not obtained from all of them, or cannot reasonably be obtained (with an explanation). This declaration is the submitting author's own statement; IDAG does not independently verify author consent. The declaration is private and is not published.

Before approving, the Registry Administrator reviews the current and proposed author lists, the justification and the agreement declaration, and records an acknowledgment of that review. Where full agreement was not declared, the Administrator must also record a decision note. The acknowledgment is an administrative review, not proof of consent. If the declaration no longer matches the proposed author list, or the registration has changed since the amendment was prepared, approval is refused until the amendment is reconciled.

An approved change is published as a new protocol version. Earlier versions keep their original author lists and author order, and the citation of each version uses that version's authors. Changing the corresponding author does not transfer management of the registration, which stays with the account that registered it.

Authorship concerns

IDAG is not the primary arbiter of complex authorship disputes. Where authorship information in a record is challenged, or agreement of affected authors is missing, IDAG may ask for clarification or supporting confirmation, return the amendment, or decline it. Substantive disputes may need to be resolved through the parties' institution, journal or another appropriate authority.

Research-integrity concerns

Concerns may include fabricated information in a submission, falsified declarations, deliberate misrepresentation of review progress, impersonation or misuse of another person's identity, a serious undisclosed conflict affecting registry integrity, or other deliberate misuse of the registry. IDAG may review information needed to protect the integrity of the registry, but it does not conduct formal research-misconduct investigations.

Where a concern falls outside IDAG's authority or needs formal investigation, IDAG may refer it, or advise referral, to an appropriate body such as the relevant institution, an ethics or research-integrity office, a journal or publisher, or another competent authority.

While a concern is assessed

A complaint does not automatically freeze a registration. IDAG may take proportionate administrative action where necessary to protect registry integrity while a serious concern is assessed. A concern never causes a published registration to disappear silently: the IDAG number, version history, and any withdrawal or discontinuation remain visible. Where a concern leads to a change to a published registration, the public record should show the resulting action where applicable.

Confidentiality and records

IDAG will handle complaint and reconsideration information with appropriate confidentiality, but information may need to be shared with relevant persons or authorities where necessary to assess or address the matter. Complaint details are not published. Significant registry decisions arising from complaints or reconsideration should be appropriately documented.

IDAG may decline to act on communications that are clearly abusive, spam, unrelated to registry functions, or repeated without substantive new information. Complainants do not need to prove misconduct before contacting IDAG.