Comparison groups built from data: a plain guide to synthetic control arms

Many clinical trials still need a comparison group, but a growing number of them build that comparison group from health information that already exists rather than from newly enrolled volunteers. That approach is called a synthetic control arm. For a person weighing whether to join a study, the idea matters because it can change how many volunteers are needed for a comparison group and, in some designs, can raise the chance that an enrolled volunteer receives the product being studied. This article explains what a synthetic control arm is, why researchers use it, how the science is kept fair, and what safeguards exist.

What a control arm does in a clinical trial

A control arm is the comparison group in a study. To judge whether a new approach works, researchers need something to measure it against. In a randomized controlled trial, which is a study that assigns volunteers to groups by chance so the groups are balanced, one group receives the study intervention and another group serves as the control. The control group may receive a placebo, which is an inactive substance with no active ingredient, or it may receive the standard of care, meaning the usual care that clinicians already provide for the condition.

Comparing the two groups helps researchers separate the real effect of the investigational product, which is the product being studied and not yet approved for general use, from changes that would have happened anyway. Randomization and comparison remain the reason results can be trusted. A closer look at how clinical trials are structured helps explain why comparison groups are central to modern research.

What a synthetic control arm is

A synthetic control arm is a comparison group assembled from patient information that already exists, rather than from people newly enrolled and assigned to a control group. It is one form of what researchers call an external control arm, meaning a comparison group drawn from outside the current study.

The information comes from sources such as real-world data, which is health information collected during routine care; electronic health records, which are the digital medical charts kept by clinics and hospitals; patient registries, which are organized collections of health data about people who share a particular condition; and historical trial data, which is information gathered from earlier completed studies. Researchers select records from these sources to represent what typically happens to comparable patients receiving prior or standard care, then compare those outcomes against the group receiving the investigational product. Advances in AI and virtual tools shaping clinical trial enrollment have made it easier to organize and analyze this kind of existing health information at scale.

How a synthetic control arm differs from a traditional control group

The core difference is timing and source. A traditional control group is made up of people enrolled in the same study at the same time and followed forward. A synthetic control arm is built from data on other people, often treated at an earlier time or in another setting. The U.S. Food and Drug Administration describes the concept plainly in its proposed guidance on the subject: in an externally controlled trial, outcomes in participants receiving the study intervention according to a protocol are compared to outcomes in a group of people outside the trial who had not received the same intervention.

A traditional randomized design still offers the strongest protection against bias, because assigning people to groups by chance balances both known and unknown differences. A synthetic control arm gives up some of that protection in exchange for other advantages, which is why it is used selectively rather than universally. Study designs vary in other ways too, such as in open-label versus blinded clinical trials, where the difference lies in who knows which group a participant is in.

Why researchers turn to synthetic control arms

Several situations make a traditional control group hard to justify or hard to fill. In rare diseases, very few patients exist, so enrolling enough people to fill both a study-intervention group and a control group can delay a study for years. In serious or life-threatening conditions, assigning volunteers to a placebo can raise ethical concerns, especially when no adequate standard of care exists. In some cancer research and in pediatric research, families and clinicians are reluctant to accept a design in which a participant might receive only a placebo.

A synthetic control arm can reduce or remove the need to enroll a separate control group, which can speed a study, lower the total number of volunteers required, and allow more of the enrolled volunteers to receive the investigational product. This is particularly relevant for the rare disease research community, where small patient populations make traditional trial designs especially challenging.

How researchers keep the comparison fair

A comparison is only meaningful if the two groups resemble each other. To achieve that, researchers use statistical methods to match the outside records to the enrolled volunteers on characteristics that affect outcomes, such as age, disease severity, and prior care. The most common method is propensity score matching, a technique that pairs each enrolled volunteer with one or more outside individuals who had a similar likelihood of receiving the same care, so that the groups look alike at the starting point.

Researchers plan these methods before the study begins, document which data sources were used and why others were excluded, and test how sensitive the results are to hidden differences. Ongoing oversight during the study, provided by groups such as a data safety monitoring board, adds another layer of independent review. Newer approaches, including designs that draw partly on a concurrent control group and partly on outside data, aim to combine the strengths of both.

What the U.S. government and the FDA say

Interest from the U.S. government in using existing health data grew after a 2016 law directed the FDA to create a framework for evaluating real-world evidence in regulatory decisions. The FDA has since issued a series of documents on the topic, including proposed guidance on externally controlled trials. That guidance remains in draft form and has not been finalized, so it represents the agency's current thinking rather than a fixed rule.

The FDA stresses that suitability must be judged individually, describing the design as one that warrants a case-by-case assessment. The agency is candid about the risks, noting that unmeasured differences between groups, lack of blinding, and other sources of bias cannot be eliminated in externally controlled trials, and that careful analysis is required to reduce their impact. Regulators have accepted this kind of evidence for a range of products, particularly for rare diseases and cancers, while continuing to state that a randomized study is preferred whenever it is feasible. A parallel example of tailored regulation is the 1983 law that reshaped rare disease research, which shows how policy adapts when standard approaches fall short.

What this means for someone considering a trial

For a volunteer, the most direct effect is on the odds of receiving the investigational product. When a study uses a synthetic control arm, fewer enrolled volunteers, or in some designs none, are assigned to a placebo or to standard care alone, which can raise the chance that a participant receives the product under study. The scientific oversight does not disappear. Studies that use this approach are still reviewed, still governed by informed consent, and still required to show their evidence holds up.

A person considering a study can ask the research site team how the comparison group is formed, whether a placebo is involved, and what the chances are of receiving the investigational product. Those questions are a normal part of the conversation before enrollment, and knowing how to read the informed consent form before joining a clinical trial is one of the best ways to prepare for that conversation.

Limits and safeguards: when this approach fits and when it does not

A synthetic control arm is not a universal replacement for randomization. Its accuracy depends on the quality and completeness of the outside data, and gaps or inconsistent records can weaken it. A central concern is hidden difference, meaning a factor that separates the two groups that no one measured or accounted for, which can make a product look better or worse than it truly is. The comparison can also drift when the standard of care changes over time, so that older records no longer reflect current outcomes.

For these reasons, the approach fits best when outcomes are objective and clearly measured, when the disease course is well understood, and when a traditional study is genuinely impractical. Regulators have at times reversed course on individual programs and asked for a randomized study instead, which shows that acceptance is decided case by case. The safeguard for volunteers is that these judgments are made by independent scientific and regulatory review, not by any single party with an interest in the outcome. This complements the broader eligibility framework that determines who can join a trial, which is itself designed to protect both participants and the integrity of the science.

How DecenTrialz supports people exploring clinical trials

DecenTrialz helps people take the first steps toward finding a study that may fit their situation. The process is straightforward: a person shares some information, the platform uses AI-assisted matching to identify studies that may be relevant, and a registered nurse completes an initial pre-screening review before any referral is made.

A registered nurse conducting pre-screening does not determine final eligibility, provide medical advice, or make enrollment decisions.

When a potential match is identified, DecenTrialz refers the person to the research site team, and that team handles the study walk-through, final eligibility determination, informed consent, and the enrollment decision. This structure lets a person learn about opportunities, including studies that use different comparison-group designs, without pressure and with a clear path to the professionals who run the study. Explore current options at decentrialz.com.

Frequently asked questions

Does a synthetic control arm mean no one receives a placebo?

Not always. Some studies remove the placebo group entirely, while others keep a smaller comparison group and add outside data. The informed consent process explains the specific design of any given study, including whether a placebo is involved.

Is a study that uses existing data less rigorous?

It is different, not automatically weaker. A randomized study offers the strongest protection against bias, but a well-designed study using outside data applies careful statistical matching and independent review. Rigor depends on the quality of the data and the methods, which regulators evaluate individually.

Can a volunteer ask how the comparison group is built?

Yes. A person can ask the research site team how the control group is formed, whether a placebo is used, and what the chance is of receiving the investigational product. These are standard questions during the consent conversation.

Learn more about clinical trial options

Understanding how comparison groups work can make the decision to explore a clinical trial feel clearer and less intimidating. Synthetic control arms are one of several ways researchers are trying to reduce the burden on volunteers while keeping the science sound. For people who want to see what studies may be available and speak with a registered nurse during an initial pre-screening review, DecenTrialz offers a simple starting point at decentrialz.com.