When real life intrudes on the protocol
The RCT gave you a beautiful result under ideal conditions. Now you are in your ICU — where nurses rotate, patients arrive at night, labs come back late, and the pharmacy runs out of the study drug. Does the trial result still apply?
This is the question pragmatic trials ask — and the framework you must use when translating evidence to bedside practice.
What makes a trial pragmatic?
Feature of explanatory trial | Feature of pragmatic trial |
Highly selected population (strict inclusion/exclusion) | Broad, inclusive eligibility (mirror real practice) |
Intensive monitoring, frequent follow‑up | Routine care; minimal extra visits |
Strict protocol adherence enforced by study staff | Treatment delivered by usual clinical staff |
Primary outcome is a surrogate or biomarker | Primary outcome is a patient‑centered, clinically meaningful endpoint |
Protocol deviations are excluded or penalised | Protocol deviations are part of the analysis (intention‑to‑treat) |
A pragmatic trial asks: “Does this work when we implement it the way we actually implement things?” It is designed for external validity from the start.
CONSORT‑Pragmatic Extension
The CONSORT‑Pragmatic Extension (2018) adds items to the standard CONSORT checklist specifically for pragmatic trials:
- Eligibility criteria must be broad and match the target population.
- Setting must reflect where the intervention will be used (community hospitals, academic ICUs, mixed settings).
- Intervention must be deliverable by usual staff without special resources.
- Comparison should be usual care, not placebo, unless placebo is standard.
- Outcome should be important to patients (mortality, quality of life, functional status) — not just biomarker change.
- Analysis should include all randomised patients (ITT), with subgroup analyses by site, severity, or comorbidity.
Appraisal checklist for pragmatic trials
- Population: Are inclusion/exclusion criteria broad enough to include the patients you see?
- Setting: Is the trial site similar to yours (academic vs community, high vs low volume)?
- Intervention: Is the intervention delivered by usual staff with usual resources?
- Outcome: Is the primary outcome patient‑centered and clinically meaningful?
- Adherence: Was adherence reported (not just in the ideal subgroup)?
- Analysis: Was ITT the primary analysis? Were subgroup analyses pre‑specified?
The pragmatic–explanatory continuum
No trial is purely one or the other; they sit on a continuum. Most ICU trials blend both. Your job is to identify where on the continuum the trial sits — and whether its position matches the clinical question you are trying to answer.
- If you want to know whether an intervention works in principle (efficacy), prefer explanatory trials with strict controls.
- If you want to know whether an intervention works in your ICU (effectiveness), prefer pragmatic trials with broad eligibility and routine delivery.
Go deeper
- EQUATOR Network – CONSORT‑Pragmatic Extension (free PDF): checklist and explanation of pragmatic trial reporting guidelines. https://www.equator-network.org/reporting-guidelines/consort-pragmatic-extension/
- StatPearls – “Pragmatic Clinical Trials” (NBK539695): free, covers design, advantages, limitations, and the pragmatic continuum. https://www.ncbi.nlm.nih.gov/books/NBK539695/
- PMC9874983 – “Designing pragmatic trials in critical care” (free full text): ICU‑focused guide to designing and appraising pragmatic ICU trials, with CONSORT‑Pragmatic examples. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9874983/
- BMJ – “Pragmatic trials and real‑world evidence” (free): discussion of how pragmatic trials bridge the gap between RCT and routine practice. https://bmj.com/content/368/bmj.m460 (open access)
Next: Chapter 16 — Putting it together: a step‑by‑step walkthrough of an ICU paper