STROBE Checklist and Guidelines: All 22 Items Explained
The 22 STROBE items for cohort, case-control and cross-sectional studies, what reviewers object to in each block, and a checker for your manuscript.
STROBE is the guideline for reporting an observational study (cohort, case-control or cross-sectional) with what is needed to judge it. It has 22 items: 18 common to the three designs and 4 with their own wording for each one (participants, statistical methods, descriptive data and outcome data). The statement dates from 2007.
By Víctor Ciudad-Fernández, PhD · Updated: October 6, 2026 · Source: von Elm et al. (2007)
What STROBE is and when to use it
STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) brings together what the article of an observational study has to contain so that the risk of bias and the validity of the conclusions can be judged. Von Elm and colleagues published it in 2007, in several journals at once, with an explanation and elaboration paper that gives the reasoning for each item.
The official website offers one combined checklist for the three designs and one per design (cohort, case-control, cross-sectional), plus another for conference abstracts. Journals usually ask for it completed at submission, with the page where each item is met.
A warning the authors themselves repeat: STROBE is a guideline for reporting the study, not a scale for scoring its quality. Using it as a quality tool in a systematic review is a mistake reviewers point out.
Cohort studies, prospective or retrospective.
Case-control studies.
Cross-sectional studies, surveys included.
If the data come from routinely collected records (health records, registries), STROBE plus its RECORD extension; for a Mendelian randomization study, STROBE-MR.
A randomized trial does not go with STROBE (it is CONSORT), nor does a diagnostic accuracy study (it is STARD).
The 22 items, section by section
Items 6, 12, 14 and 15 change with the design; here you have a summary of what they ask in each case. What follows explains each item in plain words; for submission, use the official checklist for your design. Under each section you have what reviewers usually object to in it.
Title and abstract
1a. Name the design with a commonly used term in the title or the abstract.
1b. An informative and balanced summary of what was done and what was found.
What reviewers usually object to:
The title does not name the design, or names it wrongly: a cross-sectional analysis presented as a "cohort study" because the data come from a cohort.
The abstract talks about "effect" or "impact" in a design that only allows you to talk about association.
Introduction
2. The scientific background and the rationale for the research.
3. The specific objectives, including any prespecified hypotheses.
What reviewers usually object to:
The hypotheses were not set before the analysis and it shows: the objectives look tailored to whatever came out significant.
Methods
4. The key elements of the design, early in the article.
5. The setting, locations and relevant dates: recruitment, exposure, follow-up and data collection. (checked by the checker)
6a. The eligibility criteria and how participants were selected; for cohorts, also follow-up, and for case-control studies, how cases and controls were defined and chosen. (checked by the checker)
6b. If there was matching: the criteria and, depending on the design, the number of exposed and unexposed or of controls per case.
7. The definition of every outcome, exposure, predictor, potential confounder and effect modifier, with diagnostic criteria if applicable.
8. For each variable, where the data come from and how it was measured, and whether measurement is comparable across groups.
9. What was done to address potential sources of bias. (checked by the checker)
10. How the study size was arrived at. (checked by the checker)
11. How quantitative variables were handled in the analysis and, if they were grouped, into which categories and why.
12a. All statistical methods, including those used to control for confounding.
12b. The methods used to examine subgroups and interactions.
12c. How missing data were addressed. (checked by the checker)
12d. Depending on the design: how loss to follow-up (cohort), matching (case-control) or the sampling strategy (cross-sectional) was addressed.
12e. Any sensitivity analyses.
What reviewers usually object to:
Exposure, outcome and potential confounders are not defined, and the adjusted model looks chosen after seeing the data.
The model adjusts for variables on the causal pathway (mediators) and loses the effect it meant to estimate.
Study size is justified as a "convenience sample" and non-significant results are then read as absence of effect.
A continuous variable (age, body mass index) is categorized without justifying the cut-points.
Missing data are handled by deleting cases and the number lost is not reported.
Results
13a. The number of people at each stage: potentially eligible, examined, confirmed eligible, included, with complete follow-up and analyzed.
13b. The reasons for non-participation at each stage.
13c. Consider a flow diagram.
14a. The characteristics of participants (demographic, clinical, social) and information on exposures and potential confounders.
14b. How many participants have missing data for each variable of interest.
14c. For cohorts, a summary of follow-up time.
15. Outcome data: number of events or summary measures (cohort and cross-sectional) or numbers in each exposure category (case-control).
16a. Unadjusted and, if applicable, confounder-adjusted estimates with their precision (confidence interval), and which factors were adjusted for and why.
16b. The category boundaries when continuous variables are categorized.
16c. If relevant, translating relative risk into absolute risk over a meaningful period.
17. Other analyses: subgroups, interactions and sensitivity.
What reviewers usually object to:
It is impossible to reconstruct how many people were lost between the eligible and the analyzed.
Only adjusted estimates are given, without the crude ones, or the adjustment variables are not stated.
Odds ratios are computed for a common outcome and read as if they were relative risks.
Discussion
18. The key results with reference to the objectives.
19. The limitations, with the sources of bias or imprecision and the direction and size of any potential bias.
20. A cautious interpretation that considers the objectives, the limitations, multiplicity of analyses, results from other studies and other evidence.
21. The generalizability (external validity) of the results.
What reviewers usually object to:
Causation is concluded ("X reduces Y") from a cross-sectional design.
The limitations do not say in which direction selection bias or residual confounding would move the result.
Other information
22. The sources of funding and the role of the funders in this study and, if applicable, in the original study it builds on.
What reviewers usually object to:
The funder’s role is missing, especially when the database belongs to a company or an insurer.
Check your manuscript against STROBE
Paste the article (Methods and Results at a minimum). The checker searches the text for 5 of the 22 STROBE items, the ones a text search can decide without reading (inclusion and exclusion criteria, sample size justification, sources of bias or confounding, missing data or losses to follow-up, recruitment or data-collection period), and shows you the sentence that supports each one. It runs in your browser: nothing is uploaded or stored.
Checklist checker: Detects the design and picks the guideline for you.
Previum · AI peer review. A checklist tells you whether something appears in the text; it does not tell you whether it was done right. Once the manuscript is complete, see what Reviewer 2 would object to before you submit: findings anchored to your text, in the margin. Review my manuscript
Variants and extensions
STROBE has not been replaced by a new version: the 2007 checklist is the one on the official website, in a combined version and one per design. What has grown is the set of extensions for specific cases.
The most requested: RECORD for studies using routinely collected health data (health records, registries, administrative databases) and STROBE-MR for Mendelian randomization studies. They are filled in on top of STROBE, not instead of it.
How to cite the guideline
If your Methods section says you followed STROBE, cite the statement (and the explanation and elaboration if you rely on it), not this page.
von Elm, E., Altman, D. G., Egger, M., Pocock, S. J., Gøtzsche, P. C., & Vandenbroucke, J. P. (2007). The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: Guidelines for reporting observational studies. PLoS Medicine, 4(10), e296. https://doi.org/10.1371/journal.pmed.0040296
Vandenbroucke, J. P., von Elm, E., Altman, D. G., Gøtzsche, P. C., Mulrow, C. D., Pocock, S. J., Poole, C., Schlesselman, J. J., & Egger, M. (2007). Strengthening the Reporting of Observational Studies in Epidemiology (STROBE): Explanation and elaboration. PLoS Medicine, 4(10), e297. https://doi.org/10.1371/journal.pmed.0040297
Twenty-two. Eighteen are common to cohort, case-control and cross-sectional studies; items 6, 12, 14 and 15 are worded differently depending on the design.
Can STROBE be used to rate the quality of a study?
No. It is a guideline for reporting the study. To assess the risk of bias of observational studies in a review, use tools built for that, such as ROBINS-E or the Newcastle-Ottawa scale.
Which checklist do I use for a cross-sectional study?
The STROBE version for cross-sectional studies, which the official website offers separately. If your journal links the combined one, that works too: fill in, for each item, the part that matches your design.
What if my data come from health records or a registry?
Then, on top of STROBE, you need RECORD, the extension for routinely collected health data. Among other things, it asks how the population was identified and how the codes used were validated.