Non-interventional studies are different from randomized controlled trials (RCTs). Instead of actively intervening in a patient’s healthcare, researchers just observe and record what happens naturally during the patients normal (routine) healthcare journey. The data from these studies is useful because it helps generate ideas for further research, flag potential safety concerns, or even provide additional evidence for regulatory decisions, like updating product labels or expanding the use of approved medical products.

Now, if you’re thinking of submitting this data to the FDA, there are certain regulations (rules) you must follow. In the U.S., there are two main pathways:

[1] For studies that will be submitted to the FDA: There’s a ‘checklist’ to follow, such as complying with certain sections of the Code of Federal Regulations (21 CFR 11, 21 CFR 50, 21 CFR 56, etc.) and listing your study on public databases like clinicaltrials.gov.

[2] For studies that won’t be submitted to the FDA: The requirements are a bit different, focusing more on the Common Rule (45 CFR 46), but still following similar safety rules, such as 21 CFR 314 (drugs) and 21 CFR 600 (biologics).

Despite the differences, there are common themes between the two. Both need ethical green lights in the form of Institutional Review Board (IRB) approvals, and participant informed consent is a must. With health data involved, ensuring privacy is non-negotiable, meaning regulations like HIPAA come into play. And regardless of the regulatory pathway, the study should be well-designed and use rigorous methodology to ensure data quality, integrity, and reliability. This includes clear definitions of exposures, outcomes, confounders, and other variables.

Because it’s hard to predict if non-interventional study’s results (RWE) might be submitted to the FDA, it’s a good idea to play it safe and align with FDA regulatory requirements and guidelines from the start.

In both cases, it’s essential to follow good epidemiological and statistical practices to ensure that the findings are robust, valid, and interpretable. Non-interventional studies, by their nature, are more susceptible to confounding and bias than RCTs, so it’s crucial to apply methods that address or mitigate these potential pitfalls.