13 Research methods, statistics, critical review and evidence-based practice13.3 Experimental studies
13.3.4 Pragmatic trials and validity
2 min read
Written by Resident Doctor Committee · Published
New in the 2027 syllabus
The College added this topic for exams from January 2027.
What the trial is for
Pragmatic trials estimate how well a treatment works in conditions like everyday care. They help decide whether to use it in practice.
Explanatory trials estimate effects under more controlled conditions, often with selected participants and intensive support.
Pragmatism is a continuum across design choices, not all or nothing. An open-label study is not automatically pragmatic, and a pragmatic trial can still use randomisation.
How the designs differ
Dimension | More explanatory | More pragmatic |
|---|---|---|
Participants | Narrow eligibility and selected volunteers | Eligibility closer to the intended clinical population |
Delivery | Specialist staff and a tightly specified treatment | Routine staff with ordinary flexibility |
Comparator | May be placebo or a tightly controlled treatment | A relevant alternative from usual care |
Follow-up | Extra visits and intensive measurement | Follow-up closer to routine practice |
Outcome | May focus on a biological measure or a scale score | Focuses on outcomes that matter to patients and decisions |
PRECIS-2 helps place a trial on this continuum. It looks at eligibility, recruitment, setting, organisation, flexibility of delivery, flexibility of adherence, follow-up, primary outcome and primary analysis.
Watch for resources the trial supplies. They can make a treatment that looks routine hard to reproduce outside research.
Internal and external validity
Concept | Question it asks | What threatens it |
|---|---|---|
Internal validity | Is the comparison believable for the people studied? | Bias, confounding and the wrong analysis |
External validity | Does it apply to the intended population and service? | Different patients, treatment delivery or baseline risk |
A pragmatic trial still needs good internal validity to be useful. Broad recruitment does not fix biased measurement or data that go missing selectively.
Two safeguards still work when full blinding is not possible:
Allocation concealment (keeping the next allocation hidden until the patient is enrolled) can be kept even when patients and clinicians cannot be blinded after allocation.
Blinded outcome assessment may reduce bias for subjective outcomes when the intervention itself is visible.
Reading the result
Intention-to-treat analysis estimates the effect of being assigned to treatment, under the trial's own treatment and adherence conditions. This holds only if missing outcomes are handled appropriately.
Usual care should be described. Services differ, and adding an intervention may change the other care people receive.
Non-adherence can shrink the difference between the allocated groups. It does not justify reading an uncontrolled as-treated comparison as a randomised effect.
Applicability means checking recruitment, exclusions, adherence, harms, follow-up and whether local services can deliver the intervention.