Ask Mallard

Pragmatic trials

A pragmatic trial is designed to find out whether an intervention works under ordinary conditions of care, rather than whether it can work under ideal ones.
NHMRC EVIDENCE LEVEL · 2009Level II · InterventionProvisional design classification. Study quality and certainty of findings require separate appraisal.
Why this level?Participants or clusters are randomly allocated to concurrent comparison groups. NHMRC 2009 source

Pragmatic trials randomise. What makes them pragmatic is the setting, the eligibility and the outcome, none of which is what the hierarchy rates.

What question does this help answer?

It answers the question a health service actually faces: what happens if we adopt this, with the patients we have, delivered by the staff we have, with the adherence we will get.

Pragmatic is a spectrum, not a label

A trial is more or less pragmatic along several separate dimensions: eligibility, recruitment setting, how the intervention is delivered and by whom, adherence measures, follow-up intensity, outcome relevance, and how the analysis handles deviation. A trial can be pragmatic on some and explanatory on others, and describing where it sits on each is more informative than the word itself.

The PRECIS-2 domains

The nine PRECIS-2 domains, each placed on its own spectrum from very explanatory to very pragmatic, unscored. A trial can sit at different points on different domains.Pragmatic is a spectrum per domain, not a label for the whole trialEach domain is judged on its own spectrum; nothing is scored hereVeryexplanatoryVery pragmaticEligibilityRecruitmentSettingOrganisationFlexibility of deliveryFlexibility of adherenceFollow-upPrimary outcomePrimary analysisThe domains are those of the PRECIS-2 tool cited on this page. A trial can be pragmatic in its eligibilityand explanatory in how it measures the outcome.
Conceptual diagram. It shows the structure of the idea, not data from any study.

When to use it

Use it when efficacy is already reasonably established and the question is adoption, when the intervention is complex and delivered by ordinary staff, or when the decision the trial informs is a service decision rather than a licensing one.

When not to use it

Avoid it when the biological question is unsettled: a pragmatic trial that finds no effect cannot distinguish an ineffective treatment from one that was not delivered well.

Why Mallard may recommend it

Mallard may recommend a pragmatic design when the investigator describes evaluating something already in use, in the setting where it is used, with routinely collected outcomes.

Key assumptions and design requirements

  • Where the trial sits on each pragmatic dimension is stated deliberately rather than implied.
  • Broad eligibility is matched by an analysis plan that can cope with heterogeneity.
  • The primary analysis follows the randomized assignment, since deviation is part of what is being measured.
  • Outcomes are ones the decision maker actually cares about, not surrogates.
  • Routine data sources are checked for completeness and differential ascertainment between arms.

What to watch out for

Pragmatic is a spectrum, not a label: a trial can be pragmatic in its eligibility and explanatory in its outcome measurement, and calling a trial pragmatic does not license loose conduct. Broad eligibility also buys heterogeneity, so an average effect across a wide population may not describe anyone in it.

Clinical example

Question: Should the health service adopt a pharmacist-led discharge review? Design: Randomize wards or patients in ordinary hospitals with broad eligibility, deliver the review through existing staff, and measure readmission from routine records rather than a study-specific assessment.

How to report it

CONSORT extension for pragmatic trials is the guideline for this design. It asks where each design choice sits on the explanatory to pragmatic spectrum: how participants and sites were selected, how far the intervention was standardised, what usual care actually consisted of, and how much follow-up departed from routine practice. A reporting guideline is not a design manual: use it to make the study transparent, not as a substitute for careful design. Check the target journal's own requirements as well, which this page cannot know.

Recent studies that used this
Searching PubMed…
Related Learn pages
Sources & further reading
  1. The PRECIS-2 tool: designing trials that are fit for purpose (opens in a new tab) · BMJ · doi 10.1136/bmj.h2147
  2. Overview of clinical study designs (opens in a new tab) · Clin Exp Emerg Med · doi 10.15441/ceem.23.036
  3. Improving the reporting of pragmatic trials: an extension of the CONSORT statement (opens in a new tab) · BMJ · doi 10.1136/bmj.a2390
PRECIS-2 is the tool that makes the pragmatic-explanatory spectrum concrete, scoring a trial on nine domains. It is a design aid rather than an analysis method.
Last source review: 2026-09-06
AI-generated educational content
Mallard uses reputable methodological and reporting sources, but this material may contain errors or omit important context. Verify important methodological decisions with the cited sources and, when appropriate, a qualified research methodologist or statistician.