Examples of Good PICO Questions in Medicine

Good PICO questions are specific about who is studied, what is done, what it is compared with and what is measured. The examples below are invented for teaching and show the same topic written weakly and then strongly.

Therapy

• Weak: "Is metformin good for prediabetes?"
• Better: "In adults with prediabetes (P), does metformin (I) compared with lifestyle advice alone (C) reduce progression to type 2 diabetes (O) over 3 years?"

Diagnosis

• Weak: "Is ultrasound good for appendicitis?"
• Better: "In children with right lower quadrant pain (P), what are the sensitivity and specificity of point-of-care ultrasound (I) compared with surgical or follow-up diagnosis (reference standard) for appendicitis (O)?"

Prognosis

• Weak: "Does smoking affect surgery?"
• Better: "In adults undergoing elective hip replacement (P), is current smoking (I) compared with non-smoking (C) associated with surgical-site infection within 90 days (O)?"

Harm

• Weak: "Are statins dangerous?"
• Better: "In adults over 70 (P), is statin use (I) compared with non-use (C) associated with new-onset muscle symptoms recorded in primary care (O) over 2 years?"

Illustrative example: improving a question step by step

Invented for teaching. Start: "Does telemedicine work?" Add population: "in adults with heart failure". Add intervention and comparator: "weekly nurse telemonitoring versus usual follow-up". Add outcome and time: "30-day readmission". Result: "In adults discharged after heart failure admission, does weekly nurse telemonitoring versus usual follow-up reduce 30-day readmission?"

A short checklist for your own question

Before you finalise a question, ask: Can I name the population in terms a colleague could use to recruit patients? Is the comparator realistic in my setting? Is the outcome defined by a specific measurement or code? Is the time frame stated? Would the answer change what a clinician does? A "no" to any of these means the question needs another revision.

Common pitfalls

• Using "effective" or "good" as the outcome without defining it.
• Ignoring the time frame.

Key takeaways

• Fill every element with a measurable detail.
• Match the question type (therapy, diagnosis, prognosis, harm) to a suitable design.

This page is general educational information about research methods. It is not medical, legal or regulatory advice; follow your institution's ethics board and your target journal's instructions.