A continuation protocol is the paperwork that keeps human research alive after its first approval or first study runs out. The phrase has two closely related uses in research oversight. In one, it names the continuation submission a principal investigator files with an institutional review board (IRB) so that an approved study can keep running for another period. In the other, it names a separate study plan, more often called a rollover protocol, that lets participants keep receiving an investigational drug after the trial they joined has ended. Both exist for the same reason: research outlives the approvals and the studies that start it, and someone has to decide, on the record, whether it may go on.

The phrase is not a standard term in advertising, ad tech or artificial intelligence. No specification body in those fields defines it, searches of PPC Land found no article using it, and the closest technical usages, covered further down, describe continuation mechanisms rather than a protocol of that name. The research-oversight sense is the only one in which "continuation protocol" labels a formal, regulated object, so this explainer leads with it.

How a continuation works in research oversight

In the United States, two sets of rules govern most human research. The Common Rule, codified for the Department of Health and Human Services at 45 CFR 46, covers federally funded studies. The Food and Drug Administration (FDA) applies its own IRB regulation, 21 CFR 56, to clinical investigations of drugs, devices and biologics. Both require an IRB to re-examine approved research "at intervals appropriate to the degree of risk, but not less than once per year". That re-examination is continuing review, and the investigator's filing that triggers it is the continuation.

University research-administration systems treat it as a distinct record type. The Kuali Coeus glossary used at Indiana University defines a "Continuation (Protocol)" as a report the investigator submits to the IRB, or to an animal-care committee, before the protocol expires, which once approved extends the research for up to 364 days. A full fresh review is called a renewal instead.

A typical continuation reports how many participants have enrolled and withdrawn, any adverse events and unanticipated problems, complaints, amendments since the last approval, interim findings and the current consent form and recruitment materials. The IRB then decides whether the balance of risk and benefit still holds. If approval lapses before the board acts, the HHS Office for Human Research Protections (OHRP) says in its 2010 continuing review guidance that research activity must stop, unless the IRB finds that continuing treatment is in the best interests of individual participants already enrolled.

Rollover and continued-access protocols

The second sense operates at the level of the study itself. A parent trial ends on a fixed date, but some participants are still benefiting from the drug under test, which may be years from approval. A rollover protocol enrols those people into a new, usually open-label study whose main purpose is continued supply plus safety monitoring.

Sponsors often write a single master rollover protocol that accepts participants from several parent studies of one compound. Public registry entries illustrate the pattern: a rollover protocol for tivozanib (NCT01369433), a Novartis roll-over protocol for everolimus patients who completed earlier Novartis studies (NCT01789281), and a master rollover protocol "for continued safety assessment of study drug" (NCT02632994). When no rollover exists, US patients can seek the drug through FDA expanded access under 21 CFR 312 subpart I, or through the 2018 federal Right to Try Act.

Origin and evolution

Continuing oversight grew out of the post-war ethics codes. The World Medical Association adopted the Declaration of Helsinki in June 1964. In the US, the National Research Act of 1974 created the commission whose Belmont Report, published in 1979, set out the principles IRBs still apply. The FDA issued its IRB regulation on January 27, 1981, and 15 federal departments adopted the Common Rule on June 18, 1991, both carrying the annual review floor.

Continued access developed in parallel. The FDA's treatment investigational new drug (IND) rule of 1987 let patients outside trials receive promising drugs, and an expanded access final rule, published on August 13, 2009, organised that pathway into its current form. The Declaration of Helsinki added post-trial access language in its 2000 revision and strengthened it in 2013, when it asked sponsors, researchers and host governments to plan for it before a trial begins.

The most significant change to continuation came in the revised Common Rule, published on January 19, 2017, with a general compliance date of January 21, 2019 after two delays. Under the new 45 CFR 46.109(f), continuing review is no longer required for research eligible for expedited review, for research approved through limited IRB review, or for studies that have moved on to data analysis or to collecting follow-up data from standard clinical care, unless the IRB documents a reason to keep it. The FDA did not adopt the change. It proposed rules on September 28, 2022 to harmonise parts of 21 CFR 56 with the revised Common Rule, but drug and device trials remain subject to annual review.

Why the term matters for marketing

For most marketers, the term arrives through health and pharmaceutical work. IRBs review recruitment advertising as part of the consent process, and an FDA information sheet on recruiting study subjects asks that ads avoid implying a certain or favourable outcome, avoid calling an investigational product a "new treatment" without saying it is experimental, and avoid making payment the dominant message. Changes to that material go back to the board, often inside the next continuation. A clinical-trial recruitment campaign can therefore be bound by the same approval cycle as the trial behind it.

Federal rules also bar sponsors from promoting an investigational drug as safe or effective under 21 CFR 312.7, and a rollover study does not change that status. Platform rules add a separate layer. Google has revised its prescription drug advertising policy for three markets and given programmatic buyers new freedom on prescription drug ads, while LegitScript, a third-party certifier, has passed 10,000 certified websites as a health ad gatekeeper.

The concept also matters as a template. Experiments on platform users usually sit outside IRB oversight because no federal funding or FDA-regulated product is involved. The 2014 Facebook study that altered the news feeds of 689,003 users for a week in January 2012, published in the Proceedings of the National Academy of Sciences, drew criticism partly because no such review applied. Regulators building AI sandboxes now use similar gated continuation. Brazil's data protection authority, the ANPD, let Synapse continue in its AI sandbox with conditions despite a score of 45 out of 100, below the 50-point floor its framework sets for continuing.

Limitations and disputes

Annual continuing review has long been criticised as paperwork with little protective value. The 2017 rule's preamble accepted much of that argument for low-risk work, which is why the requirement was dropped there. The FDA's divergence means many institutions now run two regimes, one with continuation and one without, depending on whether a study involves a regulated product.

Lapses remain a practical problem: a late filing pauses enrolment and can make data collected during the gap unusable.

Rollover studies raise their own disputes. Post-trial access is costly, and the Declaration of Helsinki states an ethical expectation rather than a legal duty in most countries. Sponsors can close a rollover once a drug is approved or a programme is dropped, leaving participants to insurers or expanded access. Critics also point out that open-label extensions generate safety data with no control group, which limits what they can show.

Not the same as

Continuation application. In US patent law, a continuation is a later application that claims the filing date of an earlier one under 35 U.S.C. 120. It concerns intellectual property, not research ethics.

HTTP/2 CONTINUATION frame. The HTTP/2 standard, now RFC 9113 from June 2022, defines a CONTINUATION frame that carries header data too large for one frame. In April 2024, researcher Bartek Nowotarski disclosed a group of "CONTINUATION flood" denial-of-service flaws in several HTTP/2 implementations. That is a protocol-level continuation mechanism, not a continuation protocol.

AI session continuation and handoff. Assistants increasingly let a conversation or task carry on elsewhere. Duck.ai share links let a recipient continue a shared chat on their own branch, and Google introduced handoffs from AI Overviews to AI Mode in January 2026. Agent standards such as Google's Agent2Agent protocol, and the advertising-specific Ad Context Protocol, handle long-running tasks that pause and resume. None of these names its mechanism a "continuation protocol".

Business continuation in bankruptcy. Sale processes speak of a buyer providing for the continuation of a business. Character.ai's filings say the platform will be wound down if the sale process does not produce a purchase that continues it.

Recent developments

The World Medical Association adopted a revised Declaration of Helsinki in October 2024 at its General Assembly in Helsinki, keeping post-trial provisions in place. The International Council for Harmonisation adopted the third revision of its good clinical practice guideline, ICH E6(R3), on January 6, 2025. In November 2025, the Multi-Regional Clinical Trials Center at Brigham and Women's Hospital and Harvard published a resource for ethics committees on post-trial continued access.

Outside medicine, gated continuation is spreading to AI regulation. The ANPD's second partial report on its sandbox, covered by PPC Land on October 10, 2026, set Cycle 3 conditions for continuing participants, including formal go or no-go criteria for each cycle. The structure mirrors an IRB continuation: a periodic report, a scored review and a decision.

Timeline

  • June 1964 - World Medical Association adopts the Declaration of Helsinki.
  • 1974 - US National Research Act creates the national commission on research ethics.
  • 1979 - Belmont Report published.
  • January 27, 1981 - FDA issues its IRB regulation, 21 CFR 56, with annual continuing review.
  • 1987 - FDA treatment IND rule opens access to investigational drugs outside trials.
  • June 18, 1991 - Common Rule adopted by 15 federal departments.
  • 2000 - Declaration of Helsinki adds post-trial access language.
  • August 13, 2009 - FDA publishes expanded access final rule.
  • November 10, 2010 - OHRP issues continuing review guidance.
  • 2013 - Declaration of Helsinki strengthens post-trial provisions.
  • January 19, 2017 - Revised Common Rule published.
  • May 30, 2018 - US Right to Try Act signed.
  • January 21, 2019 - Revised Common Rule compliance date; continuing review dropped for many minimal-risk studies.
  • September 28, 2022 - FDA proposes harmonising its IRB rules with the revised Common Rule.
  • October 2024 - Revised Declaration of Helsinki adopted.
  • January 6, 2025 - ICH E6(R3) adopted.
  • November 2025 - MRCT Center publishes post-trial continued access resource for ethics committees.

Summary

Who: Principal investigators and institutional review boards on the research side, and pharmaceutical sponsors running rollover studies for trial participants. In the US, OHRP and the FDA set the rules, and the World Medical Association sets the international ethical baseline.

What: A continuation protocol is either the continuation submission that extends an approved study for up to a year, or a rollover study plan that gives participants continued access to an investigational drug after their trial ends. It is not an established term in advertising, ad tech or AI.

When: Annual review dates to the FDA's 1981 IRB rule and the 1991 Common Rule. The revised Common Rule dropped it for many low-risk studies from January 21, 2019, while FDA-regulated trials still require it.

Where: Mainly in US human-research regulation, with post-trial access shaped globally by the Declaration of Helsinki, most recently revised in October 2024.

Why: Research routinely outlasts its first approval and its first study. A continuation forces a fresh, documented decision on risk, and for marketers it governs how clinical-trial recruitment advertising is approved and revised.