Written and maintained by CASRAI Editorial Board
Last updated
If your trial collects patient-reported outcome (PRO) data, the base CONSORT and SPIRIT checklists are not sufficient on their own — two dedicated extensions exist because PRO data has failure modes the general-purpose checklists don’t ask about, and the biggest of these is missing data. A patient who feels worse is more likely to skip a quality-of-life questionnaire than one who feels fine, so an incomplete PRO dataset is rarely missing at random the way a lost lab sample might be. Both extensions were built, the same way every EQUATOR-listed guideline is built, by adding PRO-specific items to an existing checklist rather than starting over: SPIRIT gets extended for protocols, CONSORT 2010 gets extended for results.
Which extension applies to your document?
| Your document is… | Use | Stage | Items | Published |
|---|---|---|---|---|
| A trial protocol that specifies a PRO as an endpoint, written before enrollment starts | SPIRIT-PRO | Protocol | 16-item extension | JAMA, 2018 (Calvert et al.) |
| The results paper of a completed trial that measured a PRO | CONSORT-PRO | Results reporting | 5-item extension | JAMA, 2013 (Calvert et al.) |
Both are led by the same core group (Melanie Calvert and colleagues at the University of Birmingham’s Centre for Patient Reported Outcomes Research), which is why the two checklists share a consistent view of what a PRO-inclusive trial needs to document at each stage rather than reflecting two independently-invented sets of requirements.
SPIRIT-PRO: what the protocol has to specify
SPIRIT-PRO’s 16 items sit on top of the base SPIRIT 2013 checklist and focus on decisions that have to be locked in before a single participant enrolls, because they can’t be reconstructed afterward. In practice, reviewers and PRO methodologists look for:
- Rationale and hypothesis — why the PRO matters for this population and condition, and what direction/magnitude of change the trial is powered to detect, not just “quality of life will be measured.”
- Instrument selection with justification — which validated PRO measure was chosen, evidence of its validity and reliability in a comparable population, and an explicit justification if no fully validated instrument existed for the study population.
- Domain specification — which specific domain(s) of a multi-domain instrument (e.g., one subscale of a quality-of-life questionnaire, not the composite score) are the actual endpoint.
- Assessment schedule — PRO timepoints stated relative to clinical assessments and to key trial events (e.g., before a clinic visit that might itself bias how a patient answers).
- Who completes it — self-report by default, with explicit rules for when and how proxy-report is permitted (e.g., a caregiver completing it for a cognitively impaired participant) and how proxy responses are distinguished from self-report in the analysis.
- Managing missing data prospectively — the methods the trial will use to promote completion (reminders, electronic capture, staff training) and the statistical approach the analysis plan will use if data are missing anyway.
CONSORT-PRO: what the results paper has to report
CONSORT-PRO’s 5 items extend the base CONSORT 2010 checklist and apply once the trial has read out. They ask a completed manuscript to:
- Identify a PRO as a primary or secondary outcome directly in the abstract, not only in the methods section, if it was one.
- Describe the PRO hypothesis and the relevant domain(s) being tested, matching what the protocol pre-specified.
- Provide or cite evidence of the PRO instrument’s validity and reliability — the same instrument-quality bar SPIRIT-PRO asks for at the protocol stage, now expected to appear in the published report too.
- State the statistical approach used for missing PRO data explicitly — not folded into a general “intention-to-treat” statement about the trial’s primary clinical endpoint, because PRO missingness patterns and clinical-endpoint missingness patterns are frequently different within the same trial.
- Discuss PRO-specific limitations and generalizability, separate from the trial’s overall limitations discussion.
Why missing PRO data gets a dedicated checklist item in both
This is the piece that makes both extensions worth reading together rather than in isolation, and it’s the item most often under-specified in practice. PRO missingness is frequently informative: participants who are hospitalized, whose disease has progressed, or who are close to death are systematically less likely to complete a questionnaire than participants who are doing well. A naive complete-case analysis of PRO scores — simply averaging whatever responses came in — can therefore make a treatment look better than it is, because the patients who would have reported the worst outcomes are disproportionately the ones missing from the dataset. This is well-established in the PRO/quality-of-life methodology literature as a distinct problem from generic trial attrition, which is why SPIRIT-PRO asks for a prospective missing-data plan and CONSORT-PRO asks for the statistical method actually used to be stated on its own, separately from whatever missing-data approach the paper used for its primary clinical endpoint.
How this fits with the rest of the PRO/COA vocabulary
PRO is one of four FDA-defined clinical outcome assessment types, alongside clinician-reported, observer-reported, and performance outcomes — a PRO specifically means the report comes directly from the patient, without clinician or observer amendment. Before a PRO instrument reaches the protocol-writing stage these two extensions govern, it typically has to clear a separate selection-and-validation process; see CASRAI’s guides on choosing and validating a PROM and clinical outcome assessment validation for that earlier step. Electronic PRO (ePRO) capture, covered in CASRAI’s ePRO dictionary entry, is one of the practical tools trial teams use to hit the completion rates SPIRIT-PRO’s missing-data item is asking them to plan for.
Frequently asked questions
Do I need to follow both SPIRIT-PRO and CONSORT-PRO, or just one?
Both, at different points in the same trial’s lifecycle: SPIRIT-PRO applies while you’re writing the protocol, before enrollment; CONSORT-PRO applies when you write up the completed trial’s results. A trial that used a PRO instrument should be checked against SPIRIT-PRO at the protocol stage and against CONSORT-PRO when the results manuscript is drafted — they are a matched pair, not alternatives.
What if PRO is only a secondary or exploratory outcome, not the primary one?
Both extensions still apply, though the level of detail journals expect typically scales with how central the PRO endpoint is to the trial’s conclusions. CONSORT-PRO’s item on identifying PRO status in the abstract explicitly applies whether it’s a primary or secondary outcome, and instrument-validity evidence and missing-data handling are expected regardless of endpoint hierarchy.
Does SPIRIT-PRO or CONSORT-PRO replace the base SPIRIT or CONSORT checklist?
No. Both are extensions layered on top of the base checklist, the same pattern used across the EQUATOR Network’s extension family (the AI-specific extensions, the non-inferiority extension, and others). A trial with a PRO endpoint still has to satisfy the full base SPIRIT 2013 or CONSORT 2010 checklist in addition to the PRO-specific items.
Where can I find the actual checklists to use?
Both are catalogued in the EQUATOR Network’s Reporting Guidelines Library, and the original publications (Calvert et al., JAMA 2013 for CONSORT-PRO; Calvert et al., JAMA 2018 for SPIRIT-PRO) include the full item-by-item checklist and explanation.
Is there more detailed guidance on the missing-data item specifically?
Both original papers and their companion explanation-and-elaboration articles discuss it, but neither hands you a single mandated statistical method — the requirement is that your protocol pre-specifies a plan and your results paper states what was actually done, not that you use one particular technique. Multiple imputation and mixed models for repeated measures are commonly used approaches in the PRO literature precisely because they can account for informative missingness better than a simple complete-case analysis, but the checklist item is about transparency of method, not endorsement of a specific one.
Last verified: 31 August 2026, against the CONSORT/SPIRIT extensions registry (consort-spirit.org) and the EQUATOR Network’s reporting guidelines library.
