01 · The Question
If Nothing Can Be Done Medically, Should the Participant Still Be Told?
Suppose researchers discover a reliable finding indicating that a participant has a substantially increased risk of a serious disease. If effective surveillance, prevention, or treatment exists, the argument for returning that information seems relatively straightforward.
Now remove the intervention. The finding is still valid. The condition may still be serious. The participant may strongly want to know. But medicine cannot currently prevent the disease, alter its course, or offer useful surveillance.
Does the absence of clinical actionability mean the finding should not be returned? This is where an apparently simple criterion becomes an important debate about what returning research findings is actually for.
03 · What You Need to Know
Actionability Answers an Important Question, but Not the Only Question
What does clinically actionable actually mean?
A finding is generally described as clinically actionable when knowing it creates an opportunity to take an established medical action with meaningful potential to improve health outcomes. Depending on the condition, that action might involve prevention, increased surveillance, earlier diagnosis, treatment, altered treatment, or another intervention capable of changing the condition's onset, course, or consequences.
For example, a genetic finding that identifies a substantially increased risk of a serious disease may be actionable if established surveillance can detect that disease earlier or preventive intervention can materially reduce the risk.
Actionability therefore means more than “a doctor can discuss this with you.” There should be a meaningful connection between knowing the information and an available response that can potentially improve clinical management.
Why does actionability carry so much ethical weight?
Actionability creates a particularly strong benefit-based argument for return. If researchers possess reliable information that allows a participant to reduce a substantial risk of serious harm, withholding the information becomes harder to justify.
This reasoning appears prominently in genomic research guidance. NHGRI notes that many genomic researchers consider medically actionable incidental findings a minimum category for return following appropriate informed consent, while also acknowledging continuing debate about what counts as actionable and which additional findings should be returned.
Consensus recommendations for biobank research have similarly proposed that researchers generally offer findings that are analytically valid, reveal an established and substantial risk of a serious health condition, are clinically actionable, comply with applicable requirements, and are wanted by the participant.
Actionability is not the same as treatability
Researchers sometimes interpret “actionable” too narrowly. A disease does not necessarily need to have a cure for information about it to be clinically actionable.
Action might include heightened surveillance, preventive measures, earlier diagnostic evaluation, modification of treatment, or another intervention that can alter the disease course. Population-research guidance has similarly defined actionability in terms of recognized therapeutic or preventive interventions or other available actions capable of changing the clinical course.
Treatable
An intervention can treat the disease or its manifestations.
Clinically actionable
Knowing the finding enables a meaningful clinical response, which may include treatment but can also include prevention, surveillance, early diagnosis, or altered management.
A condition can therefore be actionable even when it cannot be cured.
Why isn't actionability always an absolute requirement?
The disagreement begins with a deeper question: whose perspective determines whether information is useful?
A clinically oriented standard asks whether a healthcare professional can use the finding to improve the participant's medical course. A broader participant-oriented standard asks whether reliable information can provide meaningful benefit to the person even when medicine cannot change the disease itself.
Wolf and colleagues describe this as an unresolved debate between tethering return to what clinicians can use and considering what participants themselves may find valuable. Their work recognizes that health, reproductive, and personal utility can sometimes support return beyond the strictest clinical-actionability threshold.
Reproductive utility can exist without treatment
A finding may have implications for reproductive decisions even when it provides no treatment for the participant. Carrier status is an obvious example. The participant may be healthy and need no medical intervention, yet the information could affect reproductive planning.
Some return frameworks have therefore treated reproductive importance as relevant to return decisions, although there has been disagreement over whether such findings belong in the category that researchers should return or the broader category they may return.
This disagreement matters. It shows that clinical actionability is not the only defensible conception of participant benefit.
Personal utility can also matter
Information may influence important life decisions even when it cannot change medical management. A participant might use reliable information about a serious future condition to make financial arrangements, reconsider employment or retirement plans, discuss the issue with family, prepare caregiving arrangements, or make other personal decisions.
This is often described as personal utility. It is ethically more controversial than straightforward clinical actionability because personal value varies among individuals and can be difficult for researchers to predict.
Still, influential scholarship on return of results recognizes personal utility as one possible basis for return. One proposed framework permits researchers to offer analytically valid findings that reveal an established and substantial likelihood of health or reproductive importance or personal utility when return is likely to provide net benefit from the participant's perspective.
Non-actionable does not mean automatically beneficial to disclose
The argument should not be reversed too far. Once clinical actionability is removed, the benefits of disclosure may become less certain while its burdens remain real.
A participant who learns about a serious untreatable condition may experience anxiety or altered self-perception. Information may have implications for relatives. Follow-up can generate costs. Some participants would value the opportunity to prepare, while others may strongly prefer not to know.
For that reason, saying that non-actionable findings may sometimes be returned is quite different from saying that all valid non-actionable findings should be returned.
| Finding |
Potential utility |
General ethical implication |
| Serious, valid, clinically actionable finding |
Prevention, surveillance, treatment, or altered clinical management |
Usually presents the strongest case for offering return |
| Valid finding with reproductive utility |
May inform reproductive decisions |
May justify return under frameworks recognizing reproductive utility |
| Valid finding with substantial personal utility |
May inform important life decisions even without medical intervention |
Can support return when likely net benefit and participant preference justify it |
| Valid but minimally useful finding |
Little plausible benefit from knowing |
Existence of the information alone provides a weak reason for return |
| Highly uncertain finding |
Utility itself may be unclear because the meaning is uncertain |
Requires separate assessment before actionability becomes the central question |
Actionability exists on a spectrum
Even apparently actionable findings are not equally actionable. An intervention might almost completely prevent a condition, modestly reduce its probability, permit earlier detection without guaranteeing improved outcomes, or merely influence how symptoms would later be investigated.
Likewise, the burden and risks of the intervention matter. A simple surveillance test is not ethically equivalent to a major preventive operation carrying substantial complications.
ClinGen's work on clinical actionability reflects this multidimensional character by assessing evidence related to disease severity, likelihood, effectiveness of intervention, and the nature and burden of the intervention rather than treating actionability as a simple yes-or-no property.
Actionability cannot rescue an unreliable finding
A potentially lifesaving intervention does not make weak evidence acceptable. Before asking what can be done about a finding, researchers need confidence that the finding is analytically valid and that its relationship to the relevant condition is sufficiently established.
There has been broad agreement in the return-of-results literature that acceptable analytic validity is a fundamental condition for responsible return, even while debate continues over other thresholds.
This means the broader criteria for deciding whether an incidental finding is important enough to return cannot be replaced with an actionability test alone.
Participant preference becomes especially important for non-actionable findings
When a finding offers clear opportunities to prevent serious harm, the expected benefit of knowing may be substantial. When no clinical action is available, the value of knowing can depend much more heavily on the participant's own preferences and circumstances.
A person who wants information for planning purposes may view disclosure as beneficial. Another person may reasonably prefer not to live with knowledge of an untreatable future risk.
Researchers should therefore address participant choices about receiving incidental findings prospectively whenever such findings are reasonably foreseeable.
Watch Out
Do not describe a finding as “actionable” merely because some response is technically possible. The relevant question is whether an established action offers meaningful potential benefit, considering its effectiveness, risks, burdens, and the evidence supporting the finding.
04 · A Practical Example
Two Reliable Findings, but Only One Has a Medical Intervention
Hypothetical Example
A genomic study identifies two unrelated findings
During a study, researchers encounter two independently confirmed genomic findings outside the research question. Both have sufficiently established interpretations, and participants previously indicated that they wished to receive qualifying findings.
Finding A
The first finding indicates a substantial risk of a serious condition for which established surveillance and preventive management can materially reduce harm.
Return assessment
The finding is analytically valid, serious, and clinically actionable. It falls within the strongest conventional justification for offering return.
Finding B
The second finding indicates a serious future condition for which no established treatment, prevention, or surveillance can alter the clinical course.
Return assessment
The absence of clinical actionability weakens the argument that the finding must be returned, but it does not automatically prohibit return. The team considers its approved return policy, the participant's preferences, certainty of the result, potential personal utility, and likely benefits and harms of knowing.
The important difference is not that Finding B has suddenly become meaningless. Rather, the ethical justification for returning it must come from somewhere other than the possibility of improving clinical management.
07 · A Quick Checklist
Before Calling an Incidental Finding Actionable
Assess the finding by checking:
Is the finding analytically valid and sufficiently well established to support the proposed interpretation?
What specific action becomes available because the participant knows the finding?
Could that action meaningfully prevent disease, reduce risk, improve surveillance, enable earlier diagnosis, or alter treatment?
How strong is the evidence that the intervention actually provides benefit?
What risks, burdens, and costs accompany the proposed intervention?
If clinical actionability is absent, does the finding have established reproductive or substantial personal utility under the approved return policy?
Has the participant chosen to receive this category of information?
Is return likely to provide net benefit rather than simply more information?
