As a medical trainee starting your first research project, you’ve probably heard of the PICO format for writing a research question. It’s a staple in evidence-based medicine and the go-to structure for clinical trials and systematic reviews.

But if your project is based on retrospective data—like a national survey or claims database—PICO might not be the right fit.

Instead, you should be thinking in terms of PEO:
Population, Exposure, Outcome.

This framework is specifically designed for observational research, especially when you’re studying naturally occurring variables rather than introducing a new treatment or policy. In this article, we’ll walk you through how PEO differs from PICO, when to use each, and how to apply them in real-world research questions.

First, What Is PICO?

Let’s start with what you might already be familiar with.

PICO stands for:

• Population – Who are you studying?
  
• Intervention – What intervention or treatment are you introducing?
  
• Comparison – What are you comparing the intervention against?
  
• Outcome – What are you measuring?
  

This format is most appropriate for prospective studies—when researchers are actively changing something in the environment and measuring its effects.

Example using PICO:

In US adults with poor vision, does receiving free prescription eye-wear (intervention) compared to no intervention (comparison) increase the likelihood of telehealth use (outcome)?

This question assumes a controlled intervention—in this case, a policy that distributes free glasses. You're testing a causal impact that arises from something new being introduced by the researcher or policy.

So, What Is PEO?

PEO stands for:

• Population – Who are you studying?
  
• Exposure – What exposure or environmental factor naturally exists?
  
• Outcome – What are you measuring?
  

This format is designed for retrospective studies, where you’re analyzing data that already exists and where the variables you’re studying occur naturally, without any researcher interference.

Example using PEO:

In US adults, is poor vision (exposure) associated with a lower likelihood of telehealth use (outcome)?

Here, no intervention is introduced. You're simply looking at a naturally occurring condition (poor vision) and its association with an outcome (telehealth use), using already collected data.

Why the Distinction Matters

Understanding whether your research question involves an exposure or an intervention is critical to selecting the right framework.

Scenario: You're analyzing retrospective data like surveys or EHRs
Framework: PEO
‍
Scenario: You're running or evaluating a clinical trial
Framework: PICO
‍
Scenario: The factor you're studying occurs naturally (e.g., smoking, low income, poor vision)
Framework: PEO
‍
Scenario: The factor you're introducing is a new drug, device, or policy
Framework: PICO
‍
Using the wrong framework won’t just confuse your question—it can mislead your analysis and your reviewers.

Does PEO Include a Comparison Group?

Good question.

Unlike PICO, PEO doesn't explicitly include a "Comparison" element. But that doesn’t mean there isn’t one.

When you test associations statistically, you’re inherently comparing groups—those with and without the exposure.

In our earlier example:

In US adults, is poor vision associated with lower telehealth use?

You’re comparing those with poor vision to those with good vision. Even if “Comparison” isn’t listed in the acronym, it’s baked into the analysis. Without a comparison group, you can’t determine whether the exposure had any measurable effect.

Which Framework Should You Use?

Here’s a simple rule of thumb:

If the factor you’re studying already exists in your dataset, use PEO. If the factor is something being introduced or tested by a researcher, use PICO.

Example Research Question: Is poor vision associated with lower telehealth use?
Retrospective or prospective?: Restrospective
Framework: PEO

Example Research Question: Does providing free glasses increase telehealth use?
Retrospective or prospective?: Prospective
Framework: PICO
‍
Example Research Question: Does income level affect statin adherence?
Retrospective or prospective?: Retrospective
Framework: PEO

Example Research Question: Does a new app improve statin adherence?
Retrospective or prospective?: Prospective
Framework: PICO

Final Thoughts

Choosing the right research framework is one of the most overlooked steps in designing a study. But getting this right early can save you a lot of time—and dramatically increase your chances of writing a strong abstract, getting published, or presenting at a national conference.

If you’re working with retrospective datasets (like NHANES, EHRs, or insurance claims), PEO is almost always the better fit. Focus on clearly defining your population, your exposure, and your outcome. Then make sure your dataset allows for a meaningful comparison between groups.

At Lumono, we help medical trainees like you turn retrospective data into publishable insights. Whether you’re just forming your research question or ready to analyze data, our tools and guidance are built to support the entire process.