Quality Improvement Project vs Human Subject Research: What’s the Difference?

It’s not always easy for research teams to determine whether a project should be defined as quality improvement (QI) or if it truly is research involving human subjects. I recently discussed this confusion with some colleagues, and I’ve shared highlights from that conversation in this blog post. I hope this helps shed some light on this common challenge.

Let’s start with the basic definitions:

• Research defined by the federal regulations is a systematic investigation designed to develop or contribute to generalizable knowledge.
• Quality improvement is generally deemed to be a systematic process that involves data activities designed to bring about immediate improvements to healthcare delivery in a particular setting.

In research, the question we’re trying to answer is usually some version of “does this work?” or “is this better than other approaches?” The question is relatively narrow and straightforward. In quality improvement, we have a variety of activities that look quite different but all fall under the umbrella of quality improvement. QI can be practical problem solving, addressing an issue within a single hospital or unit like whether nightly bed checks at regular intervals reduce patients falls. Or it can be broader in scope, such as implementing an established intervention and then collecting data to evaluate whether it improves care. In general, quality improvement does not seek to create generalizable knowledge; rather, it evaluates programs specific to a particular organizational setting.

When determining whether a project is research or quality improvement, consider whether you’re seeking an answer that will contribute to generalizable knowledge—or whether you’re trying to improve practice in your local setting by implementing a practical solution to a problem or evaluating something that is already considered an established practice.

Risk and Research

If the project involves randomization, it’s usually research, as randomization implies that an answer is not already known. Randomization is also a potential risk to study participants, and generally, there is no risk involved in quality improvement activities.

When you start looking at identifiable participant data, you cross the line into research. Even if you’re only recording de-identifiable data, you initially have to go into a database to look at identifiers, and this can increase risk to human participants.

Exempt from IRB Review

The regulations outline specific requirements for research to be exempt from IRB review. This is different from quality improvement not requiring IRB review, because QI is by definition not research. If it is quality improvement, it does not require IRB review.

This can get tricky, however, if you want to publish the findings of your quality improvement project. Many publications ask about IRB review prior to accepting an article. In this situation, I suggest you be clear with the publisher that this is not research and avoid using the word “research” in describing the QI project. You might say something like, “This is not research but rather a quality improvement program designed to assess a standing program in our clinic.”

Note that the IRB has no input on whether a project has to be published. Additionally, intent to publish does not in and of itself turn a QI project into research.

Still not sure if it’s quality improvement or research? Contact the Advarra team for assistance.

Back to Resources