A non-interventional study (NIS) is a study of patients who receive a medicine or device as part of their normal care, where the researchers observe and record what happens but do not decide the treatment, assign patients to it, or add tests beyond routine practice. The European Union defines it in law as a clinical study that is not a clinical trial. In the US the same design is more often called an observational study. Pharmaceutical, biotech, and medical device companies use non-interventional studies to learn how a product performs in everyday practice after approval.
Quick answer: Under EU Regulation 536/2014, a study is a clinical trial if the treatment is assigned in advance outside normal practice, if the decision to prescribe is made together with the decision to enroll, or if extra diagnostic or monitoring procedures are added. A study that does none of those is non-interventional. Common designs include registries, cohort studies, case-control studies, chart reviews, and surveys. They generate real-world evidence and are often run by medical affairs after launch.
The clearest definition is in the European Union's Clinical Trials Regulation, (EU) No 536/2014. Article 2 defines a clinical trial as a clinical study that meets any one of three conditions:
A non-interventional study is then simply a clinical study that is not a clinical trial. The test is about what the study changes in the patient's care. If the physician would have prescribed the same product and ordered the same tests anyway, and the study only records the results, it is non-interventional.
FDA approaches the same idea from the evidence side. Its March 2024 draft guidance, Real-World Evidence: Considerations Regarding Non-Interventional Studies for Drug and Biological Products, describes a non-interventional, or observational, study as one in which patients receive the marketed drug during routine medical practice and are not assigned to an intervention by a protocol. It gives cohort, case-control, and self-controlled studies as examples.
In a clinical trial, the protocol decides who gets which treatment, often by randomization, and the study adds visits and tests to measure the result. That control is what lets a trial show cause and effect, and it is why trials are the main evidence for approval.
A non-interventional study gives up that control in exchange for realism. It includes the patients a trial might exclude, such as older patients or those with other conditions, and it follows them in ordinary settings for longer. The trade-off is bias. Patients who receive a treatment in practice may differ from those who do not, so the analysis has to account for confounding. FDA's 2024 draft calls identifying and addressing confounding and other bias critical when planning these studies.
Regulation follows the design. In the EU, clinical trials need authorization under the Clinical Trials Regulation. Non-interventional studies fall outside it, although national rules, ethics review, data protection law, and safety reporting duties still apply.
Non-interventional studies are a main source of real-world evidence. Companies use them to:
Medical affairs usually owns company-sponsored non-interventional studies after launch, planned against the evidence gaps in the integrated evidence plan. An external investigator can also run a non-interventional study as an investigator-initiated study, with company support.
A study can drift into being a trial without anyone deciding it should. A protocol that requires a scan the patient would not normally get, or that enrolls patients only if they start the product, can meet the trial definition. Compliance teams also check that a non-interventional study has a genuine scientific purpose, because paying physicians to record data on patients they put on a product can look like a reward for prescribing if the study has no real research question. Fees to investigators should reflect the work involved.
A medical affairs team may run registries, chart reviews, and investigator studies in parallel across countries. TikaStudies governs company-sponsored non-interventional and real-world evidence programs in the same portfolio as investigator-initiated studies, with milestones, documents, and spend on each study. It does not recruit patients or capture study data.
A non-interventional study observes patients who receive a medicine or device as part of their normal care. The study does not assign the treatment or add procedures beyond routine practice. EU Regulation 536/2014 defines it as a clinical study that is not a clinical trial.
In a clinical trial, the protocol assigns the treatment and may add tests or visits. In a non-interventional study, the treatment decision is made in normal practice, independent of the study, and the study only records what happens.
Largely, yes. Non-interventional study is the term used in EU law. FDA's 2024 draft guidance uses non-interventional and observational interchangeably for studies in which patients receive a marketed drug in routine practice and are not assigned to an intervention by a protocol.
After approval, medical affairs often owns company-sponsored non-interventional studies, with input from HEOR, safety, and market access. External investigators can also run them as investigator-initiated studies with company support.
If your team runs registries and real-world studies across several countries, we can walk through how the portfolio looks with your own study list.
Book a 20-minute walkthrough →